Oceanography

An Illustrated Guide

C.P. Summerhayes and S.A. Thorpe

Southampton Oceanography Centre,
Southampton, UK

Foreword by Robert D. Ballard

Center for Marine Exploration
Woods Hole Oceanographic Institution,
Woods Hole, MA, USA

MANSON
PUBLISHING
 The Contributors
All contributors are based at the Southampton
Oceanography Centre, apart from Dr A. Gebruk (P.P.
Shirshov Institute of Oceanography, Moscow, Russia),
and Dr C.M. Young (Harbor Branch Oceanographic
Institution, Ft Pierce, Florida, USA).

Copyright © 1996 Manson Publishing Ltd

ISBN 1–874545–38–3 (hard cover edition)
ISBN 1–874545–37–5 (soft cover edition)

All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system or transmitted in any form or by any means without the written
permission of the copyright holder or in accordance with the provisions of the Copyright
Act 1956 (as amended), or under the terms of any licence permitting
limited copying issued by the Copyright Licensing Agency, 33–34 Alfred Place,
London WC1E 7DP.

Any person who does any unauthorised act in relation to this publication may be liable to
criminal prosecution and civil claims for damages.

A CIP catalogue record for this book is available from the British Library.

For full details of all Manson Publishing Ltd titles please write to Manson
Publishing Ltd, 73 Corringham Road, London NW11 7DL, UK.

Text organisation and supervision: John Ormiston

Design and layout: Patrick Daly
Line artwork: Kate Davis
Proofreading and index: Michael Forder
Colour separations by: Tenon & Polert, Hong Kong
Printed by: Grafos SA, Barcelona, Spain
Contents
Foreword 4 11 The Ocean: A Global Chemical
Robert D. Ballard System 165
Prologue 4 J.D. Burton
John G. Shepherd 12 The Marine Carbonate System 182
Preface 5 M. Varney
The Global Oceans 6 13 A Walk on the Deep Side: Animals
in the Deep Sea 195
The Geological Time Scale 7 P.A. Tyler, A.L. Rice, C.M. Young,
Standard International (SI) Units 8 and A. Gebruk
Commonly Used SI Units 8 14 Light, Colour, and Vision in the
Ocean 212
1 How the Science of Oceanography P.J. Herring
Developed 9
M.B. Deacon 15 Ocean Diversity 228
M.V. Angel
2 The Atmosphere and the Ocean 27
H. Charnock 16 Life in Estuaries, Salt Marshes,
Lagoons, and Coastal Waters 244
3 The Role of Ocean Circulation A.P.M. Lockwood, M. Sheader,
in the Changing Climate 41 and J.A. Williams
N.C. Wells, W.J. Gould, and
A.E.S. Kemp 17 Artificial Reefs 259
K.J. Collins and A.C. Jensen
4 Ocean Weather – Eddies in the Sea 59
K.J. Richards and W.J. Gould 18 Scientific Diving 273
J.J. Mallinson, A.C. Jensen,
5 Observing Oceans from Space 69 N.C. Flemming, and K.J. Collins
I.S. Robinson and T. Guymer
19 Marine Instrumentation 280
6 Marine Phytoplankton Blooms 89 G. Griffiths and S.A. Thorpe
D.A. Purdie
20 The Sea Floor – Exploring
7 Snow Falls in the Open Ocean 96 a Hidden World 300
R.S. Lampitt P. Riddy and D.G. Masson
8 The Evolution and Structure 21 Ocean Resources 314
of Ocean Basins 113 C.P. Summerhayes
R.B. Whitmarsh, J.M. Bull,
R.G. Rothwell, and J. Thomson 22 Waste Disposal in the Deep Ocean 338
M.V. Angel
9 Slides, Debris Flows, and
Turbidity Currents 136
D.G. Masson, N.H. Kenyon,
and P.P.E. Weaver Some Commonly Used Words and Terms 346
10 Mid-Ocean Ridges and Acronyms 347
Hydrothermal Activity 152 Index 348
C.R. German, L.M. Parson,
and R.A. Mills
Foreword
The latter half of the twentieth century has seen the passes inside the Ridge and out every six to eight
science and technology of Oceanography progress million years.
from a primitive state of understanding to a highly During this same period, we have seen the primi-
sophisticated science, although many believe tive sampling techniques of surface-towed dredges
oceanography is still in its infancy. and plankton nets give way to manned sub-
It was not until 1960, for example, that mersibles and, more recently, to advanced remotely
oceanographers first recognized the size and signifi- operated vehicle (ROV) systems. These ROV sys-
cance of the Mid-Ocean Ridge, a great mountain tems not only provide scientists with around-the-
range that stretches through the ocean basins of the clock access to the deepest reaches of the ocean
world for a distance of 70,000 km and covers close floor, but can also be operated from shore via satel-
to 23% of the Earth’s total surface area. Even more lites. Clearly, the information highway of the twen-
ironic is the fact that astronauts walked on the sur- ty-first century will enable scientists world-wide to
face of the Moon before Earth scientists explored have easy access to the oceans of our planet.
the Ridge’s rift valley for the first time, in 1973, Despite this better understanding, few outside
using manned submersibles. the field of Oceanography are aware of these new
Following this lowly start, oceanographers have findings. It is with such thoughts in mind that the
not only learned the significance of this great staff of the Southampton Oceanography Centre in
undersea mountain range to the genesis of oceanic the UK have joined forces to create this introducto-
crust, but have also discovered the existence of ry book on Oceanography.
hydrothermal events along the axis of the Ridge, We believe that this book will help everyone bet-
surrounded by important mineral deposits and ter appreciate this exciting new field of science,
exotic life forms that live independently of the especially the next generation who will most likely
Sun’s life-supporting energy. explore more of the ocean floor and learn more
The discovery of volcanism and hydrothermal about its hidden secrets than all the previous gener-
circulation within the Mid-Ocean Ridge not only ations combined.
had an impact on the biological and geological sci- Dr Robert D. Ballard
ences, it also helped us to better understand the Director, Center for Marine Exploration
chemistry of the oceans. We now know, for exam- Woods Hole Oceanographic Institution,
ple, that the entire volume of the world’s oceans Woods Hole, MA, USA

Prologue
This book is published to mark the occasion of a science of the oceans he loved so much to a wider
major event in Oceanography in Britain, the forma- audience. The book thus marks the passing of an
tion of the Southampton Oceanography Centre era dominated by great oceanographers, including
(SOC). This Centre brings together the Natural John’s American friend and colleague Henry
Environment Research Council’s Institute of Stommel, who also died quite recently. We hope
Oceanographic Sciences and Research Vessel that its publication marks the commencement of a
Services, with its three ships, Discovery, Charles new era, in which we can continue their work with
Darwin, and Challenger. In a fine new building equal vigour and honesty, and maintain the spirit
alongside the Empress Dock in Southampton, the of friendship and co-operation in which they
SOC is one of the largest institutions devoted to the worked.
study of the Earth and its oceans in Europe and in It is therefore also a great pleasure and entirely
the world. appropriate to include a Foreword by Bob Ballard
The book is a collection of contributions largely from the Woods Hole Oceanographic Institution,
by the staff of the SOC and their colleagues. It where Henry Stommel worked for so many years,
reflects the range of their interests, and I hope that and which John Swallow regarded as his second
it conveys something of their excitement and enthu- home.
siasm for their subject. There is, after all, only one ocean, and we must
All of us were saddened by the death of our emi- work together to understand its secrets.
nent colleague John Swallow during the prepara- Professor John G. Shepherd
tion of the book. We are sure that he would have Director, Southampton Oceanography Centre,
approved of this venture, which aims to bring the Southampton, UK

4
Preface
The original motivation for producing this book extraction from the sea bed, to coastal protection,
was a comment and question from a young student to tidal or wave power, and to waste disposal. It is
visiting the Department of Oceanography at the of great relevance to defence and to the proper
University of Southampton; ‘I couldn’t find any- management and regulation of the use of the seas,
thing in the bookshops to tell me what oceanogra- especially when this may involve pollution or
phy is, so what is it?’ Our purpose here is to pro- degradation of the marine environment. The global
vide some answers, to explain the science, and to ocean, because it responds slowly to changes in
express some of the delight and the privilege we temperature or atmospheric composition, acts as a
feel in being involved in it. Chapters are written in flywheel to climate change, and the prediction of
a variety of styles, but all, we hope, are at a level at future climate is presently a major driving force
which a science undergraduate should have no dif- affecting the direction of a significant part of the
ficulty in understanding; the book contains a science.
wealth of information and guides to further read- In this book we describe, as individuals or with
ing, which all should find of interest and informa- colleagues, those facets of the science in which we
tive. This book is not directed at any particular have found particular delight and satisfaction and
teaching course, but provides much which will sup- for which we have special expertise and enthusi-
plement courses in environmental science. asm. It touches on many of the present areas of
So, what is Oceanography? Oceanography greatest activity in the science today and, novel in a
begins at the shoreline. It is the science of the book of this nature, it describes the equipment used
oceans, their interaction with the atmosphere above in research, some of the methodology, and some of
and with the underlying sea-floor sediments and the applications of the science – the discovery and
oceanic crust, their chemical and biological compo- assessment of resources and ocean wealth, and the
nents, their physical properties and motion, their use of the oceans for the disposal of waste.
geology, their creation, past history, and develop- There is excitement, stimulus, and satisfaction in
ment, their present state, and their future. finding, often quite unexpectedly, that discoveries
Oceanography is founded on the basic scientific in one area of the science have important conse-
disciplines of biology, chemistry, geology, physics, quences for another. Oceanography is not the geog-
and mathematics. Many of the problems addressed raphy of the oceans, it does not have all the glam-
by oceanographers are interdisciplinary, so their our of Jacques Cousteau’s films or of diving in the
solution demands a breadth of knowledge that warm waters of a coral reef – it is good, often
crosses conventional scientific boundaries and applicable and useful, science, hard and mind-chal-
requires multidisciplinary team collaboration. The lenging work which provides great personal satis-
science uses the range of facilities of all these basic faction – and a lot of fun.
sciences, for example advanced computers and the The reader should keep in mind the toil, fatigue,
analytical laboratory instruments of biologists and and effort needed to gather information about the
chemists. Oceanography also relies heavily upon a ocean – the heaving and pitching of a vessel in
range of technologies, for example computing, elec- severe weather – and recognise that part of the
tronics, optics, and acoustics which, together with oceanographer’s reward and satisfaction derives
the engineering involved in the design and con- from the occasionally successful (would that it were
struction of winches and wires, enable the scientist always so!) combat with a harsh, uncompromising
to make observations and sample the remote ocean and hostile environment. This book may be seen as
depths – it is this remoteness which makes a tribute to the dedication of those who have wrest-
Oceanography akin to Space Science. It requires ed that information from the sea.
elaborate and robust instruments to survive the We are most grateful to Dr Bob Ballard of the
hostile environment, one which by its very nature Woods Hole Oceanographic Institute, US, for the
spans the Globe. Because the collection of material Foreword. Dr Ballard is well-known for his investi-
and measurements from the sea is essential, it is an gations of the deep ocean, for discovering the
expensive and high-risk science, using ships, air- wreck of Titanic, and for his initiative in bringing
craft, satellites, and submersibles. International col- the science of oceanography live by satellite links to
laboration is often necessary to provide the school children.
resources and ships for major experiments and pro- Dr C.P. Summerhayes,
grammes of research. Professor S.A. Thorpe, FRS,
Oceanography is a science with application to Southampton Oceanography Centre,
areas of considerable economic importance; for Southampton, UK
example, to fisheries, to hydrocarbon or mineral

5
6

The Global Oceans

1. Bahamas 12. Gulf of Mexico 23. Madeira Abyssal Plain 35. Porcupine Seabight 47. Storegga slide
2. Baltic Sea 13. Gulf Stream 24. Mediterranean Sea 36. Porcupine Abyssal Plain 48. Weddell Sea
3. Bay of Biscay 14. Hatteras Abyssal Plain 25. Mid-Atlantic Ridge 37. Red Sea
4. Bermuda 15. Hawaiian Islands 26. Monterey, California 38. Rio Grande Rift Based on the General Bathymetric
5. Black Sea 16. Hawaiian–Emperor 27. Newfoundland 39. River Zaire Chart of the Oceans (GEBCO),
6. Bosporus Seamounts 28. Ninetyeast Ridge 40. River Seine published by the Canadian
7. Canary Islands 17. Iberian Peninsula 29. North Sea 41. Rockall Bank Hydrographic Service, Ottawa,
8. East Pacific Rise 18. Indian Ocean 30. North Atlantic Ocean 42. Santa Barbara Basin Canada, 1984; reproduced with
permission of the International
9. East African Rift 19. Irish Sea 31. Norwegian Sea 43. Sardinia
Hydrographic Organization and
Valley 20. Isthmus of Panama 32. Pacific Ocean 44. Somali Current the Intergovernmental
10. Florida 21. Kuroshio 33. Philippines 45. Southampton Water Oceanographic Commission (of
11. Grand Banks 22. Labrador Sea 34. Poole Bay, UK 46. Southern Ocean UNESCO).
 a,b
The Geological Time Scale
Era Sub-era, Period, Sub-period Epoch Age (Myr)
Quaternary Sub-era Holocene
0.01
Pleistocene
Cenozoic

1.64
Tertiary Sub-era Pliocene
5.20
Neogene Period Miocene
23.3
Oligocene
35.4
Palaeogene Eocene
56.5
Palaeocene
65.0
Senonian
88.5
Cretaceous
Gallic
131.8
Mesozoic

Neocomian
145.6
Malm
157.1
Jurassic Period Dogger
178.0
Lias
208.0
Triassic 3
235.0
Triassic Period Triassic 2
241.1
Scythian
245.0
Permian Period Zechstein
256.1
Rotliegendes
290.0
Gzelian
295.1
Pennsylvanian Kasimovian
303.0
Carboniferous Sub-period Moscovian
311.3
Period Bashkirian
322.8
Serpukhovian
332.9
Mississippian Visean
349.5
Sub-period Tournaisian
362.5
Devonian 3
Palaeozoic

377.4
Devonian Period Devonian 2
386.0
Devonian 1
408.5
Pridoli
410.7
Silurian Period Ludlow
424.0
Wenlock
430.4
Llandovery
439.0
Ashgill
443.1
Caradoc
463.9
Ordovician Period Llandeilo
468.6
Llanvirn
476.1
Arenig
493.0
Tremadoc
510.0
Merioneth
517.2
Cambrian Period St. David’s
536.0
Caerfai
570.0

a
After Harland, W.B., Armstrong, R.L., Cox, A.V., Craig, L.E., Smith, A.G., and Smith, D.G. (1990), A Geologic
Time Scale. Cambridge University Press, Cambridge.

b
These Eras comprise the Phanaerozoic Eon. Preceding it is the PreCambrian, dating back to the origin of the Earth,
at around 4600.0 M.

7
Standard International (SI) Units
Wherever possible the units used are those of the International System of Units known as SI. Oceanographers have tradi-
tionally used other units, such as the litre, which often cannot be avoided because of their common usage. Despite the
recommendations periodically published by international committees as to what constitutes a standardised scientific ter-
minology, agreement is still rather poor. Conversion between units often requires great care.

SI Unit Prefixes
Name Symbol Multiplying factor 10N, N is given below
peta P 15
tera T 12
giga G 9
mega M 6
kilo k 3
hecto h 2
deca da 1
deci d –1
centi c –2
milli m –3
micro μ –6
nano n –9
pico p –12
femto ƒ –15
atto a –18

Commonly Used SI Units

SI units
Name Symbol Name Equivalent cgs
Force N Newton kg m/s2
Pressure Pa Pascal kg/m s2
Energy/work J Joule kg m2/s2
Power/energy flux W Watt kg m2/s3
Irradiance E/m2s Einstein/m2s mol photons/m2s

The expression of gas concentrations is a particularly problematic area. The SI unit for pressure is the Pascal (1 Pa =
1 N/m2). Although the bar (1 bar = 105 Pa) is also retained for the time being, it does not belong to the SI system.
Various texts and scientific papers still refer to gas pressure in units of the torr (symbol: Torr), the bar, the conven-
tional millimetre of mercury (symbol: mmHg), atmospheres (symbol: atm), and pounds per square inch (symbol: psi) –
although these units will gradually disappear. Irradiance is also measured in W/m2. Note; 1 mol photon = 6.02 × 1023
photons.
The SI unit used for the amount of substance is the mole (symbol: mol), and for volume the SI unit is the cubic
metre (symbol: m3). It is technically correct, therefore, to refer to concentration in units of mol/m3. However, because
of the volumetric change that sea water experiences with depth, marine chemists prefer to express sea water concen-
trations in molal units, mol/kg.

8
CHAPTER 1:

How the Science of

Oceanography Developed
M.B. Deacon

Early Ideas about the Sea suggestions were plausible given the state of knowl-
Oceanography is a young science with a long his- edge at the time. Up to the end of the seventeenth
tory. Scientists in the seventeenth and eighteenth century it was widely held that ground water was
centuries tried to study the sea, but were often absorbed by the land directly from the sea, and that
frustrated by its sheer size and complexity and the this formed the source of wells and springs. Then
practical difficulties involved. During the nine- measurements made by the French scientist Edmé
teenth century technological advances made Mariotte, showing that the rainfall in the Paris area
systematic exploration of the deep sea possible for was, in fact, sufficient to account for the flow of
the first time, and oceanography became an inde- water in the Seine, removed the objection that rain-
pendent scientific discipline. However, the greatest fall was insufficient to account for large rivers.
strides toward understanding the sea and its
importance, both as a feature which governs the The Age of the Discoveries
Earth as we know it and one that influences To account for the movement of the heavens,
human activities in many ways, have been made Aristotle suggested that the Sun, Moon, and stars
during the twentieth century. Oceanography today revolved around the Earth attached to concentric,
is very different from what it was 50 years ago, let crystalline (and therefore invisible) spheres. These,
alone 200 years or more, so is there any good rea- he said, derived their movement from an outermost
son why anyone, apart from historians, should be sphere, which became known to medieval science as
interested in its past? Science is a continuum; as an the primum mobile, or prime mover. When
activity it grows out of its past, even if that some- Columbus made his voyages of discovery across the
times means rejecting outmoded ideas or Atlantic in the 1490s, he experienced a westward-
unreliable data. A look at how oceanography has flowing current in the tropics, and similar currents
developed can be a valuable way of helping to were also identified in the Pacific and Indian
understand the modern science, both in terms of a Oceans. Only in the Indian Ocean, north of the
set of ideas and as an institution. equator, where the proximity of the Asian land
Primitive societies developed complex mytholo- mass creates strong seasonal variations in winds and
gies to explain the workings of the universe, and weather (the monsoons), is this flow subject to peri-
invoked deities to account for natural phenomena. odic reversals in direction, a fact already known to
By Greek and Roman times, however, philosophers Arab geographers of the ninth century AD. The
were beginning to look for natural causes for things westward flow near the equator (there is actually an
about them, from the movement of the heavens8 to eastward-flowing equatorial counter-current divid-
the waves of the sea5. Of these, Aristotle (in the ing it into two streams, but this was not identified
fourth century BC) most influenced later European until the early nineteenth century) was thought by
science. He wrote widely on natural science, as well many sixteenth century writers to be due to motion
as politics and philosophy, and his works contain transmitted to the Earth’s fluid envelope by the pri-
much to interest oceanographers, including the first mum mobile3. However, after Copernicus suggested
known observations on marine biology. We find that the Sun was at the centre of the Universe and
him considering such diverse topics as how winds that the Earth rotated around it, this explanation
cause waves, water movements in straits, and the had to be adapted. The new version supposed that
water balance of the ocean. Aristotle believed that the westward movement, thought to exist through-
the presence of water vapour in the atmosphere, the out the oceans, although most marked near the
source of rain, was due to evaporation, principally equator, was due to inertia – as the Earth rotated
from the sea. Rainfall supplied rivers and these daily on its axis, the sea lagged behind.
flowed into the sea, so the level of the ocean was During the seventeenth century we find alterna-
maintained. This seems obvious today, but other tive explanations beginning to appear4. These were

9
 M.B. Deacon

1.1 Figure 1.1 Chart showing ocean currents from Mundus

Subterraneus by Athanasius Kircher14, a Jesuit mathemati-
cian who taught at Rome. This chart was probably the
first to attempt to show ocean currents in the major
oceans, whose geographical limits were by then quite
well-understood, except in the polar regions (where far
more land is shown than actually exists). Kircher specu-
lated that the ocean was connected with water masses in
the interior of the Earth (his subterranean world) through
openings in the sea floor. The siting of the abysses he
shows on his map was not entirely fanciful, some being
located on sites of reputed whirlpools. For example,
Kircher supposed that water was sucked in at the
Maelstrom – in the Atlantic off Norway – and flowed via
a subterranean tunnel into the Baltic. This had no actual
basis in fact, but modern oceanographers are finding that
mainly the work of Roman Catholic philosophers sea water under the high pressures that exist at the sea
who, after Galileo’s condemnation, were forbidden bed penetrates the rocks and sediments of the sea floor
to express Copernican ideas. One of these, and migrates through them. At mid-ocean rift systems,
Athanasius Kircher14 (Figure 1.1) suggested that as this sea water, charged with chemicals leached from the
the Sun travels over the sea its heat evaporates the rocks through which it has passed at high temperatures,
water below. This creates a depression in the sea is forced back into the ocean in a scenario that would, if
surface so that currents flow in from either side to suggested, have seemed as exotic as Kircher’s abysses
until relatively recently (see Chapter 10). (Courtesy of The
restore its level. The evaporated water falls as rain Royal Society, London, England.)
in higher latitudes, so that the circulation is main-
tained. Isaac Vos, a Dutch scholar who later settled
in England, objected that this would actually lead The Scientific Revolution of the
to a current flowing eastward, from the part of the Seventeenth Century
ocean which the Sun had not yet reached 34. He The Treasure for Travellers by William Bourne2,
agreed that it was the Sun’s heat which was respon- which contains an interesting account of the geogra-
sible for currents, but said it operated by expanding phy of the sea as seen by a representative of the
the water, so that the level of the sea rose slightly newly educated professional classes, is a good
as the Sun moved across it. This was sufficient to example of a change in attitude. Bourne takes his
cause a flow toward a lower level. Vos’ theory was own observations as a starting point to suggest how
one of the most original and well-supported to common coastal features, such as cliffs, beaches,
appear up to that time (1663). He was less convinc- and stacks, might have come into existence instead
ing when he tried to introduce effects from the of, as conventional scholars would have done, con-
Earth’s annual rotation in its orbit around the Sun centrating on discussion of previous ideas, which
in order to explain tides as a by-product. The idea could often be traced back to Aristotle. The achieve-
that surface currents are actually caused by winds ment of the scientific revolution of the seventeenth
occurs in seventeenth-century literature, but few century was to highlight the need to advance science
scientists took it seriously at that time. by experiment and observation, in conjunction with
The cause of tides had been keenly discussed theory.
since Greek philosophers first learned of their exis- This was the philosophy which lay behind the
tence (because the Mediterranean is so enclosed, its foundation of the Royal Society in 1660, but the
tides are generally small and unnoticed.) The much Fellows’ interest in the science of the sea5 was also
larger rise and fall of the tides on ocean coasts influenced by the growing importance of maritime
(twice daily in most places) and the monthly springs affairs in English national life during the sixteenth
and neaps (high and low tidal ranges) seemed to be and seventeenth centuries. The projects they under-
linked to the Moon and its phases, but how did this took, therefore, had a dual purpose. The wish to
come about? Throughout the Middle Ages and the further knowledge of the natural world, in collabo-
Renaissance numerous explanations were made, ration with scientists in other countries, was com-
usually linked to contemporary thinking on cosmol- bined with the hope of information which would
ogy8 but without, apparently, any attempt to obtain be of practical benefit to seafarers, an intention
more accurate knowledge of tides themselves. that was partly humanitarian, but which also had
However, these were clearly well-known to seafar- its roots in the desire for national economic and
ers and during the sixteenth century we find details, strategic advantage.
usually of the ‘establishment’ of ports (i.e., time of Much of the work done by individual Fellows of
high water relative to the Moon’s passage over- the Society was directed toward devising apparatus
head), appearing in printed works on navigation. that sailors could use on voyages. They particularly

10
 1: How the Science of Oceanography Developed

hoped to obtain information about the depth of the reading by anyone interested in the science of the
sea, but were worried that soundings made in the sea1. However, Newton percipiently remarked in
ordinary way, with lead and line, might be inaccu- one of his letters that, rather than sailors sending
rate because of subsurface water movements bend- information to mathematicians at home, it would
ing the line out of the vertical. To overcome this a be more fruitful to send the mathematicians to sea.
device suggested in earlier literature was adopted, Early scientific interest in the sea was not, of
consisting of a weight with a float attached to it in course, confined to Britain. Perhaps the most inter-
such a way that when the weight hit the sea bed the esting contribution at this time was made by L.F.
float disengaged and rose to the surface through its Marsigli33. A native of Bologna, his university stud-
own buoyancy. The depth of water was calculated ies seem to have given direction and method to a
from the time this took. Robert Hooke improved boundless curiosity about the world in general. As a
the basic design of the apparatus (Figure 1.2); dur-
ing the next 150 years much energy was devoted to
1.2
improving this method. Although Hooke’s device
performed well when tested in shallow water,
unfortunately both it and its successors suffered
from an unsuspected design fault that made it use-
less in deep water. Pressure of water increases in
proportion to its depth, so that in the open sea the
float, which was made of wood, became water-
logged and could not rise to the surface. Hooke
also designed a sampler to collect sea water from
different depths. This was intended to find out if
the sea was only salty at the surface, as Aristotle
was supposed to have said. If this idea was correct
(it was not) then fresh water would be brought up
from the depths.
By the provision of instructions and encourage-
ment, the Royal Society31 had some success in per-
suading its followers to collect information about
the sea, in spite of difficulties experienced with the
apparatus. Notably, some of the tidal observations
thus obtained were used by Sir Isaac Newton to
illustrate his theory of the Universe. He was able to
show that tides are due to the gravitational attrac-
tion of the Moon and, to a lesser extent, the Sun.
This claim proved highly controversial, since the
idea of gravity operating through empty space had
been rejected by thinkers earlier in the seventeenth
century. They were trying to show that nature is
governed by physical laws and were therefore
Figure 1.2 Sounding machine (Fig 2), water sampler (Fig.
reluctant to employ a concept that appeared no 8), and hydrometer (Fig. 7); this plate was published in
more soundly based than discredited ideas about 1667 by The Royal Society31, founded in 1660, which
astrological ‘influence’. The fact that Newton was wanted to collect information on scientific topics as
able to express the effect of gravity in mathemati- widely as possible, so these instructions were drawn up
cally demonstrable laws led to the gradual accep- to show overseas travellers the kind of observations that
tance of his views, but his tidal theory was also cru- were wanted, and how to make them. Among other
cial in this process. His friend Edmond Halley11 felt things, they hoped to find how the depth of the sea
so strongly about it that he wrote an article sum- altered from place to place, and whether the sea was salt
marising Newton’s arguments in language and throughout, or only at the surface. The Society experi-
terms that could be understood by the layman (the mented with apparatus for measuring depth and bringing
up water from the lower layers of the sea, but had vari-
Principia had been published in Latin). Information able success. Robert Hooke, the society’s curator, pro-
on tides and currents (see Figure 1.3) also came duced designs which were an improvement on earlier
from British travellers abroad. Another famous sci- models. Anita McConnell16 has pointed out, however,
entist, Robert Boyle, relied on such observations that these woodcuts, which appeared in the journal, were
when he wrote three essays on the salinity, depth, not faithful copies of his drawings and would not actually
and temperature of the sea in the 1670s, still worth work! (Courtesy of The Royal Society, London, England.)

11
M.B. Deacon

1.3

Figure 1.3 Richard Bolland’s sounding lead for tides and currents, and sounding boat for cur-
rents. These are reproduced from Bolland’s manuscript Mediterranean Journall of 1675 (in the
Pepys Library at Magdalene College, Cambridge). Bolland was stationed at Tangier during its
brief period as a British possession (it was part of the dowry of Charles II’s queen, Catherine of
Braganza). During the British occupation, extensive works were carried out to provide a safe
anchorage for shipping. Bolland took part in this work and made a chart of the Strait of
Gibraltar, showing its tides and surface-water movements5. The existence of a current flowing
into the Mediterranean through the Strait had long been known. Sailors believed that there was a
compensatory outflow into the Atlantic below. Bolland wanted to obtain proof of the existence
of the undercurrent and devised a method of doing this. The sounding lead had a small float (D)
attached and a mechanism to release it on striking the bottom. He hoped that by seeing where
the float came to the surface, it would be possible to work out the speed of the undercurrent.
However, he realised that it would be necessary to allow for the effect of the surface current
throughout its depth (which he supposed might be as much as 100 fathoms – about 180 m). To
establish this he intended using the drogue attached to the boat, which could be lowered to the
desired depth, and the speed of the boat measured relative to that of the surface water. These are
quite sophisticated ideas and nothing comparable was achieved until the nineteenth century.
However, Bolland, and other supporters of the undercurrent, could offer no explanation of how
it was generated. Other people thought it was only a seaman’s yarn. A colleague of Bolland’s at
Tangier, Sir Henry Sheeres, wrote an essay to prove that the inflow from the Atlantic into the
Mediterranean was maintained by the climate of the area. Low rainfall and hot sun meant that
evaporation exceeded the input of water from rain and rivers (see Chapter 2), and the sea’s level
would otherwise have fallen below that of the ocean outside. Edmond Halley used this explana-
tion, widely accepted during the next two centuries. However, it did not account for what hap-
pened to the salt that is left behind when sea or other salt water evaporates. A German scientist,
J.S. von Waitz, who was connected with the salt industry (which relies on this principle), pointed
out in the mid-eighteenth century that, as water became more salty, and therefore heavier, it
would sink and that, as the depths of the Mediterranean filled with this more saline water, it
would flow out into the Atlantic (see text). When more detailed physical surveys of the
Mediterranean and the Atlantic came to be made in the late nineteenth and twentieth centuries,
these showed that saline water does, indeed, spill out over the lip of the Strait of Gibraltar and
spread out into the Atlantic, where it has a significant effect on the wider oceanic circulation
(see Chapter 11). Currents can now be measured by moored instruments or acoustic remote
sensing (see Chapter 19). (Reproduced by kind permission of the Master and Fellows, Magdalene
College, Cambridge, England.)

12
 1: How the Science of Oceanography Developed

young man, he accompanied a diplomatic mission ple, and apparatus, but state funding of science is a
to Constantinople and, while there, investigated comparatively recent innovation. There were also
reports of a counter-current in the Bosphorus, technical obstacles, especially to the exploration of
beneath the surface current flowing out of the Black the deep ocean. Above all, in spite of the work that
Sea. He showed that the depths of the strait were had already been done, there was at this time no
occupied by more saline, and therefore heavier, recognised ‘science of the sea’. Nevertheless, impor-
water of Mediterranean origin, and that this water tant advances in understanding were made during
must reach the Black Sea, which would otherwise be the eighteenth century. At the same time, a number
entirely fresh because of the rivers flowing into it of related developments contributed to laying the
(Figure 1.3). He demonstrated the way this could foundation of oceanography as we know it today.
happen with an experiment showing how, when These included improvements in navigation and
two liquids of differing specific gravity were intro- marine surveying, in particular the discovery of
duced into a container, they would form layers with methods to measure longitude at sea, which made
the heavier liquid below and the lighter one above. it possible for the first time to fix a ship’s position
Marsigli spent many years on military service in accurately when out of sight of land – an essential
eastern Europe before returning to the science of the prerequisite for studying the ocean. These improve-
sea in later life. He wrote about his researches off ments were exploited first on official voyages of
the southern coast of France in his book, Histoire exploration, expeditions despatched by govern-
Physique de la Mer15. Earlier works on geography ments with political and economic objectives in
and navigation had contained sections on tides and mind, but from the time of Captain James Cook
currents, and occasionally other aspects of the sea, onward (the circumnavigations he commanded
but this was the first book on a truly oceanographic spanned the years 1768–1780) they became
theme. However, its title is misleading as it is largely increasingly scientific in nature. A considerable
about marine invertebrates. These interested amount of oceanographic work was done, especial-
Marsigli because they had so far attracted little ly by French and Russian expeditions, in the early
attention – there were earlier works on fish. He nineteenth century. Their observations of surface
himself was particularly interested in coral, valued temperature and salinity were used by geographers
for use in jewellery and decorative objects, but like Humboldt in studies of the world climate.
wrongly classed it as plant rather than an animal, However, as deep-water observations were more
because of what he concluded were its ‘flowers’17. difficult, few were made. Much depended on indi-
vidual scientists being fortunate enough to have the
The Eighteenth and Early opportunity of observing for themselves, or
Nineteenth Centuries through interested laymen, especially naval officers.
In spite of this promising beginning, marine science Even so, sufficient new information was
did not develop as rapidly as one might have obtained to encourage the development of ideas
expected during the eighteenth century. An indica- about the interior of the ocean. Toward the middle
tion of why this was so is found in Marsigli’s of the eighteenth century Stephen Hales10, who had
book15. where he points out that science at sea is already tried to improve Hooke’s sounding
beyond the resources available to individuals and machine (Figure 1.4), also produced an apparatus
that further progress would only be made with gov- designed to measure temperature by raising water
ernment aid. This is because oceanographic from ocean depths. Such attempts were not entirely
research demands expensive items, like ships, peo- new. Hooke had proposed a design for a deep-sea

Figure 1.4 Deep-sea sounding machine devised by Stephen Hales10. A clergyman and sci- 1.4
entist, Stephen Hales (1677–1761) befriended the naturalist Gilbert White. In the early years
of the eighteenth century, he and J.T. Desaguliers, the Royal Society’s curator, tried to devel-
op new ways of measuring the depth of the sea which would be more reliable than those
suggested by Hooke half-a-century earlier. They relied on the compression of air inside the
apparatus, but most were made of glass and would not have been sufficiently robust for use
in the marine environment. An example of this type is held in the George III collection at
the Science Museum in London. The more durable apparatus illustrated here employed a
rifle barrel (Fig. II, K–Z), with a removable rod (Fig. I, A–B) inside. Hales intended using
coloured oil to mark the height reached by the water inside, which would enable the pres-
sure, and therefore the depth of water, to be calculated. He did not appreciate the weakness
of this design in that it still relied on a float (Fig. II, i) made of wood to bring it to the surface.
In the sea’s depths, pressure would force water into the pores of the wood so that it would
lose its buoyancy and not bob up again. When this machine was tried in mid-ocean it never
returned to the surface.

13
M.B. Deacon

thermometer and Marsigli15 had measured sea tem- evidence for this view, in particular those made by
peratures – until his only thermometer was broken the French scientist François Péron. He used de
in a raid by pirates. Hales’ apparatus10 consisted of Saussure’s method to reveal the existence of low
an enclosed bucket with hinged flaps, opening temperatures in the depths of seas in warm lati-
upward only, in both top and bottom. This tudes. Since it could not have formed there, it was
allowed water to flow freely through the device on argued that this colder water must have originated
the way down, but trapped a sample inside as soon in polar regions.
as the observer began to haul it up. Its temperature This idea was widely accepted on the continent,
was measured at the surface. However, the objec- but, in Britain in particular, the supposition that sea
tion could be made that the water’s temperature water, like fresh water, expands before freezing led
might have altered on the way up. To overcome to the widely held belief that water in the depths of
this problem, the Swiss scientist H.B. de Saussure the sea could not fall below 4˚C, the temperature of
insulated the thermometer itself and left it down maximum density of fresh water. In 1819 Alexander
long enough to take on the temperature of the Marcet, a Swiss physician living in London, pub-
water at that depth. This took such a long time that lished an important paper on the salinity of sea
the method was not much used, though it proved water in different parts of the world5. In this he
the most reliable at the time. Self-registering ther- showed that sea water of average salinity behaves
mometers, pioneered by James Six in the 1780s, differently from fresh water, and that its density
were more popular, but had disadvantages that did increases with cold until it freezes. This meant that
not become apparent until later. in a theoretical ocean, where salinity was uniform
The reason for this growing interest in tempera- and density was a function of temperature only, the
ture measurement was initially connected with geo- coldest water would always sink to the bottom. The
logical debate – was the interior of the Earth hot or delay in accepting Marcet’s findings, which were
cold? It was some time before people began to spec- later confirmed by other scientists, was due princi-
ulate what this work was telling them about the sea. pally to poor communication.
In the mid-eighteenth century a German scientist, There was no recognised science of oceanography
J.S. von Waitz6 pointed out that Marsigli’s argu- during the first half of the nineteenth century and
ments15 about the Bosphorus could equally well individuals interested in marine science at that time
apply to the Strait of Gibraltar. Beneath the surface came from a variety of backgrounds. Another reason
current from the Atlantic there must be an outflow for the continuing confusion over ocean circulation
of more saline water (see Figure 1.3, caption), oth- was that the Six self-registering thermometers (see
erwise the Mediterranean’s salinity would be far earlier), then widely used to measure deep-sea tem-
higher. Waitz then suggested that similar imbal- peratures, were not sufficiently protected and gave
ances gave rise to currents in the ocean. These were readings distorted by the effect of pressure. On his
due to differences in density between equatorial voyage of discovery in the Southern Ocean between
regions, where the Sun’s heat caused evaporation 1839 and 1843, Sir James Clark Ross30 measured
and increased salinity, and high latitudes where deep-sea temperatures assiduously, but was not sur-
rainfall would lessen it. Saline water in the tropics prised that they never apparently fell below 4˚C. As
sank and spread toward the poles in the ocean a naval officer, although one who was active in sci-
depths, while lighter, fresher water at the surface entific research, Ross could not be expected to be
flowed toward the equator to replace it. The situa- fully up to date in all the branches of science repre-
tion is more complex than Waitz supposed. The sented on the expedition (its primary task was obser-
density of sea water depends on temperature as well vations on terrestrial magnetism in the southern
as salinity and one can counteract the other. Heavy hemisphere). He should have been better advised by
rain falls at the equator, and though melting ice scientific colleagues at home, but the information he
reduces salinity in high latitudes in summer, in win- needed in this instance was possessed by chemists
ter brine is released as the sea freezes. However, he and physicists with whom he had no direct contact.
was the first, as far as we know, to suggest the exis- It was not until the events leading up to the
tence of an internal circulation in the ocean. Challenger expedition (see later) that the misunder-
Waitz’s suggestion had little immediate impact, standing was exposed, but such widely held misap-
but toward the close of the eighteenth century we prehensions are hard to eradicate and the 4˚C error
find similar ideas appearing, but with an alteration can be found in some twentieth century publications,
in emphasis that had temperature rather than salin- including Hansard12 (1961), in a reply to a question
ity differences being responsible for maintaining about the operation of Royal Navy submarines in
circulation. More ambitious deep-sea temperature the Arctic!
measurements made on voyages of exploration in The situation just described had partly arisen
the early nineteenth century provided supporting because naval surveyors, geographers, and some

14
 1: How the Science of Oceanography Developed

Figure 1.5 Chart of Atlantic currents by James

Rennell, from Rennell’s book, An Investigation
of the Currents of the Atlantic Ocean, 183223. 1.5
Rennell was the first to produce charts of
winds and currents, based on observation, for
an entire ocean and to show how the course
of ocean currents was largely shaped by pre-
vailing winds. A former naval officer and sur-
veyor for the East India Company, he was
already interested in currents before his return
to England in 1778. The ship in which he was
travelling narrowly escaped loss on the Scilly
Islands, notorious for shipwrecks. He suggest-
ed that part of the problem might be an
unidentified current, since known as Rennell’s
Current, flowing out of the Bay of Biscay. Later
literature has generally discounted this, but
some recent models of North Atlantic circula-
tion provide a possible explanation for such a
feature. Rennell devoted the last 50 years of
his life to geographical research, including
work on ocean currents throughout the world.
Only part of his work, a volume of charts and
accompanying memoir on the Atlantic, was
published in 1832, two years after his death.
These charts were based either on observa-
tions made by seafaring friends or derived
from ships’ log-books. Chapter 4 describes
recent discoveries about ocean currents.
(Courtesy of the SOC, Southampton, England.)
scientists were working in a different tradition, improved if ships from all nations systematically
linked to hydrography and the interests and needs collected and recorded details of wind and weather.
of seafarers, rather than to the physical sciences. As a result of his efforts, an international confer-
Though charts of the Gulf Stream had been pub- ence was held in Brussels in 1853 at which govern-
lished somewhat earlier by Benjamin Franklin27 and ments agreed to adopt a standardised scheme of
his less well-known predecessor, W.G. de Brahm7, observations. Scientists had always recognised the
it was the introduction of chronometers toward the importance of co-operation and exchange of infor-
end of the eighteenth century which made it possi- mation and ideas with colleagues in other coun-
ble to collect information on ocean currents on a tries. The Brussels meeting was a milestone in the
wider scale. This was because, once ships could fix development of maritime meteorology, which is
their position when out of sight of land, the infor- closely linked to several branches of modern
mation contained in their log-books enabled the oceanography (in particular the study of air–sea
effect of currents upon them to be calculated. The interaction, see Chapter 2). It also introduced the
first to take advantage of this was James Rennell23 idea of scientific co-operation between govern-
(Figure 1.5). Other hydrographers followed his ments, which has been of particular significance in
example, so mid-nineteenth century sailing direc- the development of modern oceanography.
tions contained good accounts of the surface cur- Maury’s initiative came at a time of rapid
rents of the major oceans. Rennell had shown how change – sail was already giving way to steam for
closely such water movements were allied to the naval and commercial purposes. The technological
direction of winds blowing over the sea surface, developments of the nineteenth century revolu-
something that had probably always been self-evi- tionised the opportunities for scientific research,
dent to seafarers, but which was slow to take root and the study of the oceans in particular. It was as
in the scientific literature. During the 1840s, a result of this that the establishment of oceanogra-
Matthew Fontaine Maury, Superintendent of the phy as a separate discipline took place, and much
United States Navy’s Depot of Charts and of the impetus came from increasing maritime
Instruments, produced seasonal wind and current activity. The construction of more and larger ships,
charts20, based on averaging data from log-books of ports and harbours to receive them, and of light-
and designed to speed the passages of sailing ships. houses to guide their passage made it necessary, for
He believed that these charts could be further example, to obtain better knowledge of tides

15
 M.B. Deacon

1.6 (Figure 1.6) and waves. But perhaps the most

important development for marine science came
through the technology developed in response to
the challenge of laying deep-sea telegraph cables, as
this made scientific investigation of the ocean
depths possible for the first time.

The Origins of Deep-Sea Exploration

The first functioning submarine telegraph cable
was laid across the Straits of Dover in November
1851. From that time the prospect of extending this
new means of rapid communication between conti-
nents was a powerful incentive to governments and
industry alike. New technology had to be devel-
oped, not only to protect and lower cables to the
sea bed and to raise and repair them if the need
arose, but also to find out about the deep-sea envi-
ronment – knowledge essential for routing and
operating the cables16. The nature and contours of
the sea bed had to be established, and also the tem-
perature of the water. Up to this time, deep sound-
ings had rarely been attempted because of the great
effort involved, particularly if line and instruments
were to be retrieved (Figure 1.7). The introduction
of steam power made such operations possible on a
more routine basis for the first time, although they
were still laborious and time-consuming. By the
mid-nineteenth century, hydrographic surveying
was already a specialised activity in most navies, so
new techniques and apparatus were developed
rapidly for use in the deep sea. It was the combina-
tion of this new technology, and the accompanying
professional expertise, with scientific thought that
made possible further advance. Yet there was some
delay before this happened – it was marine biolo-
gists rather than physical scientists who were the
first to make use of the new opportunities that had
been created.
Familiarity with the marine life of coastal and
surface waters had been greatly extended during
Figure 1.6 A tide gauge (from Nautical Magazine, 1832, the latter part of the eighteenth and early nine-
1, 401–404), an apparatus for measuring the rise and fall
teenth centuries, as biologists sought to expand
of the tide devised by an engineer named Mitchell at
Sheerness Dockyard. Similar gauges were being installed their knowledge of living creatures and establish
elsewhere at the time, but Mitchell’s gauge incorporated their affinities through schemes of classification38.
an important innovation – it was self-registering, so did By the middle of the nineteenth century many
not require the presence of an observer. Tides are caused European and American zoologists specialised in
by the gravitational pull of the Moon (and to a lesser marine work and their interest was further stimu-
extent of the Sun), which varies with distance. Monthly lated by the publication of Darwin’s theory of evo-
and annual cycles can be detected, but tidal heights are lution in 1859. Thus, the sea-shore collecting
also affected by weather. By the 1830s the Industrial which became a popular craze among Victorian
Revolution was in full swing in Britain, and engineering holiday-makers (Figure 1.8) served a more serious
projects needed precise information of such phenomena.
purpose among the scientific fraternity. They were
The information was also welcomed by scientists who
were trying to work out how the gravitational forces, well- not only interested in discovering new species, but
known from the work of Newton and his successors in the also in learning more about the physiology and life
eighteenth century, were translated into actual movement history of individual organisms. This required
in the ocean. Tides can now be measured by instruments working space and equipment, so seasonal labora-
set on the sea bed in mid-ocean (Chapter 19). tories were set up by the sea-shore, from which

16
 1: How the Science of Oceanography Developed

Figure 1.7 Deep Sound- 1.7

ings; or, no Bottom with
4600 Fathoms, a woodcut
from Sir James Clark Ross29.
This shows how laborious a
task making deep-sea
soundings was in sailing
ships. Ross was an experi-
enced polar explorer –
before going to the Ant-
arctic he had accompanied
his uncle Sir John Ross on
naval and private expedi-
tions in search of the North
West Passage and the North
Magnetic Pole. The primary
aim in this expedition, in
H.M.S. Erebus and H.M.S.
Terror, was to measure
magnetic variation in the
southern hemisphere and
locate the South Magnetic
Pole, in conjunction with a
survey of terrestrial magnetism being made by scientists from many nations. He was also interested in ocean science
and made several deep soundings and many sea-temperature measurements. In fact, his soundings greatly exaggerated
the depth of water in the areas he covered. During the next few decades surveys of routes for submarine cables using
steam vessels resulted in the development of new techniques and greater precision in such observations, and made the
scientific study of the deep sea practical, 200 years after it had been proposed by the Royal Society.

more permanent institutions, the marine biological tists and academics, but in the mid-nineteenth cen-
laboratories, emerged. One of the most famous tury amateur collectors were still numerous and
and influential of these, though not the first to be made an important contribution. Some were keen
established, was the Stazione Zoologica at Naples, yachtsmen and began to extend their interests to
founded by a German zoologist and follower of the waters of the continental shelf. Dredging was
Darwin, Anton Dorhn, in the early 1870s. Others also supported by the newly formed British
followed – on both sides of the Atlantic. Both the Association for the Advancement of Science 21.
Woods Hole Biological Laboratory on Cape Cod Some of the impetus for this work came from geo-
and the Scripps Oceanographic Institution, part of logical discoveries. When modern relatives of fossil
the University of California, started life in remains were discovered in seas further to the
this way. north or south, this suggested that climate might
When these developments took place the lead- have undergone considerable shifts (see Chapter
ing figures were, increasingly, professional scien- 3). This finding was of much interest in the light of

1.8

Figure 1.8 Pegwell Bay, Kent – a

Recollection of October 5th 1858, a paint-
ing by William Dyce, shows a scene that
would have been familiar at the time. The
growing interest of nineteenth-century biol-
ogists in marine life-forms was for a short
while reflected in a more widespread
enthusiasm for sea-shore collecting of nat-
ural history specimens. This became a fea-
ture of seaside holidays, taken increasingly
by the expanding middle class as rail trans-
port made travel cheaper. From the 1850s
onward many semi-scientific books and
guides were written to cater for this market.
(Courtesy of the Tate Gallery, London, UK.)

17
 M.B. Deacon

evidence being put forward, particularly by Swiss ters gave a more accurate picture of the distribution
geologists, for periods of extensive glaciation in of temperature with depth, and more was known
the past. about the behaviour of sea water at low tempera-
Almost without exception, those engaged in this tures. Carpenter adopted the idea that density dif-
work failed at first to appreciate the new opportu- ferences between equatorial and polar regions cause
nities which deep-sea surveying work was creating. internal circulation in the ocean, with warm, light
This was because it was widely believed that life water moving poleward at the surface to compen-
could not exist in the conditions of darkness, cold, sate for colder, denser water spreading toward the
and immense pressure that exist in the depths of equator in the depths5. Such an idea was largely
the sea. There was then no conception of the range unfamiliar to a British audience brought up on
of adaptations (which modern biologists are still Rennell and his successors, so it had a mixed recep-
discovering) that enable creatures to live in such tion. Carpenter’s fiercest critic was James Croll,
environments (see Chapters 13 and 15). The find- who had recently put forward a theory to account
ings of Edward Forbes, who had worked in the for the ice ages which held that shifts in the pattern
Mediterranean in the 1840s, were often quoted to of trade winds, and the ocean currents which they
support the idea of an azoic (without life) zone generate, were responsible for climate change.
below 400 fathoms (780 m). There was, it was Carpenter believed that if he could obtain informa-
true, some evidence from elsewhere that seemed tion from the other oceans he would have
contrary to this view, but for some years most peo- irrefutable proof of his theory. A respected elder sci-
ple regarded it as unconvincing25. However, in the entist, he used his contacts with other scientists and
late 1860s accumulating observations from a num- politicians to win support for a large-scale expedi-
ber of sources suggested to workers in several tion – the first to have marine science as its primary
countries that the supposed limit was erroneous. In objective. This resulted in the round-the-world voy-
1868, two British biologists, W.B. Carpenter and age of H.M.S. Challenger between 1872 and 1876
C. Wyville Thomson, backed by the Royal Society, (Figure 1.9), with a naval crew and a team of civil-
persuaded the Admiralty to allow them the use of ian scientists led by Wyville Thomson.
one of its survey ships so that they could dredge in The voyage of the Challenger (Figure 1.10) was
deep water. In three voyages, first in the Lightning a major landmark in the development of oceanog-
and then in the Porcupine24, they obtained incon- raphy, both in essence and in its findings. The
trovertible evidence that the deep sea was populat- observations of temperature and salinity (Figure
ed by a thriving community of creatures previously 1.11) showed hitherto unsuspected features, like
unknown to science. the spread of saline water from the Mediterranean
into the North Atlantic. Neither instrumentation16
The Voyage of H.M.S. Challenger nor theory were yet good enough to enable a
As the work went on, Carpenter’s attention increas- detailed picture of ocean circulation to be made,
ingly turned to the physical observations being but its existence in some form could no longer be
made by the naval personnel. Improved thermome- doubted (Figure 1.12).

1.9

Figure 1.9 The route followed by

H.M.S. Challenger during her oceano-
graphic voyage round the world of
1872–1876, taken from Wild36. Wild
was the expedition’s artist and secre-
tary to the scientific leader, C.
Wyville Thomson. Other circumnavi-
gations, including the voyages of
Captain Cook in the eighteenth
century, French expeditions, and
Russian voyages in the early nine-
teenth century, had added much to
geographical and scientific knowl-
edge, but the Challenger expedition
was the first large-scale expedition
devoted primarily to the science of
oceanography. For a non-scientist’s
view of the voyage, see the recently
published letters of Joseph Matkin22.

18
 1: How the Science of Oceanography Developed

1.10 1.11

Figure 1.10 Challenger in the ice (reproduced from

Wild37). (Courtesy of the Southampton Oceanography
Centre, Southampton, England.)

Figure 1.11 This drawing, by Elizabeth Gulland, was one

of a series commissioned as illustrations for the narrative
volumes of the Challenger Report. It shows a scientist
and members of the crew taking readings from deep-sea
thermometers. [Courtesy of Edinburgh University Library
(Special Collections), Edinburgh, Scotland.]

Figure 1.12 This diagram, showing the deep basins of the 1.12
Atlantic Ocean, is taken from H.M.S. Challenger, No. 7.
Report on Ocean Soundings and Temperatures, Atlantic
Ocean, 1876 (plate VI). This was the last of a series of pre-
liminary reports issued by the Admiralty on the expedi-
tion’s hydrographic work. It shows how the presence of
submarine ridges in the Atlantic influences the distribution
of bottom temperatures. The lowest temperature is found
in the southwest Atlantic, where cold Antarctic bottom
water flows northward. It cannot penetrate the eastern
basin of the Atlantic, or the North Atlantic, because its way
is blocked by submarine ridges, so the bottom temperature
in these basins is slightly warmer. The pattern of topogra-
phy shown here was partly known from deep-sea sound-
ings, but partly inferred from temperature measurements.
They had only a few observations to work from, so the fea-
tures shown only bear a generalised resemblance to what
would be seen on a modern chart – the techniques avail-
able to oceanographers today enable the sea bed to be
mapped in fine detail. The interesting point to nineteenth-
century scientists was that the information contained in
this chart could be held to support the idea of internal
ocean circulation due to density differences (both the tem-
perature and salinity of sea water affect its specific gravity),
as opposed to the pattern of largely wind-driven currents at
the surface. This idea was being hotly contested at the time
and was one of the reasons for the Challenger expedition.
This principle has now long been accepted, but the chart
also shows a feature which has aroused much excitement
within the working lifetime of present-day oceanogra-
phers. This is what is now known as the Mid-Atlantic Ridge. Nineteenth-century cable surveyors were surprised to find
that the greatest depths in the North Atlantic Ocean were not in mid-ocean, but to either side. A combination of sound-
ings and deep-sea temperature observations suggested to the Challenger staff that the ridge might continue into the South
Atlantic, and on the voyage home in 1876 they carried out soundings that showed this was so. Geological theory at the
time could not easily explain such a feature and it was not until the 1960s that it became widely accepted that it is, in
fact, a spreading centre at which new ocean floor is being created, and part of a world-wide system whose dynamics are
explained by the theory of plate tectonics (see Chapter 8). Connecting ridges, like the Walvis Ridge, whose approximate
course is shown here linking to southern Africa, are now thought to be due to hot-spot activity.
19
 M.B. Deacon

1.13a 1.13b

Figure 1.13 (a) Globigerina ooze and (b) Pteropod ooze, from Murray and Renard19, plate XI, Figures 5 and 6, respec-
tively. One of the principal scientific achievements of the Challenger expedition was to produce the first global map of
what the sea bed is made of. John Murray, who edited the Challenger Report after the death of Wyville Thomson in
1882, had made observations during the voyage which enabled him to show that, in most parts of the ocean, sedi-
ments reflect the composition of marine life in the surface layers (plankton). Globigerina ooze was named after the
remains of microscopic calcareous species of Foraminifera which form it. In the Southern Ocean, siliceous remains of
single-celled phytoplankton (plants) predominate in diatom ooze. In deep water, far from land, where calcium carbon-
ate dissolves and there are no terrigenous sediments (derived from land), with the exception of volcanic pumice which
can float long distances, the Challenger found slowly accumulating ‘red clay’ and manganese nodules (see Chapters 6
and 7). (Courtesy of the Southampton Oceanography Centre, Southampton, England.)

Sediment samples (Figure 1.13) collected during types of oceanographic equipment, such as current
the expedition formed the basis for the first world meters, reversing thermometers, and self-closing
chart of sea-floor deposits. However, the emphasis nets, became standard13,16. There was considerable
both during the voyage and afterward was on variation in the provision for oceanographic work
marine biology. Of the 50 volumes of the from one country to another, depending on local
Challenger Report, edited by John Murray after customs and arrangements for supporting science.
Thomson’s early death, 33 were reports on the zoo- In some countries, as, for example, Germany,
logical collections, contributed by specialists from which had only recently been unified, state funding
both Europe and America. Most of the species was relatively generous. The German government
described were new to science and the reports are supported the Stazione Zoologica at Naples (whose
still a basic work of reference for oceanographers. founder, Anton Dohrn, was German), as well as
The expedition also aided the growth of marine sci- oceanographic expeditions. A number of govern-
ence in a more general sense. Scientists in other ment bodies were involved in various aspects of
countries used the example of the Challenger to marine science and a research institute (the Institut
obtain government support for their work. As this für Meereskunde), attached to Berlin University,
progressed, the common ground between their was set up in 1900. In America, Maury’s methods
researches gave them a sense of identity strong and ideas had been criticised by more orthodox sci-
enough to override links with their various parent entists, but when he supported the South in the
sciences, such as physics, chemistry, or biology. American Civil War, marine science suffered
They called the new science ‘Oceanography’. through the loss of what Schlee32 has described as
‘his stubborn and passionate interest in the sea and
Oceanography from the 1880s his ability to channel funds toward its exploration’.
to the 1930s Schlee32 shows how oceanographic work carried
While single-ship expeditions continued to make out by US government agencies actually declined in
important contributions to oceanography40, unlike the latter part of the nineteenth century, with the
the Challenger, which had been exploratory in exception of Pillsbury’s survey of the Gulf
nature, they tended to concentrate more on a par- Stream (in the Coast Guard steamer Blake) in the
ticular area or problem. The Challenger expedition early 1880s.
had been organised without much time to prepare. Prince Albert I of Monaco, a wealthy patron of
On the whole, it made use of well-tried techniques science and himself an active oceanographer13,24
which were already a little old-fashioned. It was (Figures 1.14–1.17), established an oceanographic
during the post-Challenger period that many basic institute in Paris and the Musée Océanographique

20
 1: How the Science of Oceanography Developed

1.14 1.15

Figures 1.14–1.17 Prince Albert I of 1.16

Monaco – his statue, by Françoise Cogné,
stands in the gardens adjacent to the Musée
Océanographique, Monaco (1.14) – used
his wealth to further a number of sciences,
but oceanography was his principal
interest13. He was an enthusiastic yachts-
man and made almost annual voyages in
his own research ships 24 from the 1880s
until the outbreak of war in 1914 (he died
in 1922). 1.15 shows one of his vessels, the
Princesse Alice I (courtesy of Musée Océan-
ographique, Monaco). With the help of spe-
cialist assistants, and visiting colleagues
from France and other countries, he investi-
gated a wide range of physical and biologi-
cal problems during these cruises. Much
attention was paid to the improvement of
existing apparatus and to the development
of new methods 16 . One of the Monaco 1.17
inventions was the ‘nasse triédrique’ 26 ,
(1.16) a baited trap which could be lowered
to predetermined depths to catch creatures
that evaded traditional nets and trawls.
Prince Albert benefited the French oceano-
graphic community by founding an institute
for research and teaching in Paris. His
superbly situated Musée Océanographique
at Monaco (1.17), inaugurated in 1910,
continues to provide a valuable resource for
visitors and students interested in oceanog-
raphy, and its history (courtesy of Musée
Océanographique, Monaco).

21
 M.B. Deacon

1.18 Figure 1.18 Alexander Agassiz on board the US Fisheries

vessel Albatross. Agassiz was a mining engineer, and one
of the leading marine zoologists of his day (the Swiss
zoologist Jean Louis Agassiz was his father13). He used his
wealth to build up his father’s Museum of Comparative
Zoology at Harvard and undertake research expeditions,
many of them to study coral reefs, in the hope of throw-
ing new light on the origin of atolls. In the late 1870s he
made three dredging cruises to the Caribbean in the US
Coast Survey Ship Blake. The US Fish Commission was
founded in the early 1870s, one of a number of such
national organisations to come into being at that time
(see text). Its ocean-going research vessel, the Albatross,
built in 1882, was used by Agassiz on several expedi-
tions, for which he paid part of the expenses. The first
cruise, in 1891, was undertaken to compare deep-sea
fauna on the Pacific side of the Isthmus of Panama with
Caribbean forms and perhaps arrive at an approximate
date for the closure of the sea-way between North and
South America, which his earlier researches on the
Atlantic side had suggested must have persisted until
comparatively recent geological times. (Courtesy of the
Southampton Oceanography Centre, Southampton,
England.)

at Monaco in the first decade of the twentieth cen- Murray’s Scottish Marine Station for Scientific
tury. In the US the situation outlined above had led Research in the 1880s had been a short-lived
to a greater reliance on private funding. One of the attempt to create a more diversified institution.
outstanding marine scientists of the late nineteenth In the early twentieth century, with more atten-
century was Alexander Agassiz (Figure 1.18), who tion being paid to the importance of physical
financed expeditions in the Pacific13,32. The Woods oceanography than to marine biology, laboratories
Hole Oceanographic Institution was set up in 1930 began to widen their interests. The Scripps
through the agency of the National Academy of Institution of Oceanography in the US took this
Sciences, with an endowment from the Rockefeller step to its logical conclusion when it transformed
Foundation32. Only a few such specialised research itself from a marine station in 192532.
institutes for oceanography were founded before During this period, whatever the prevailing atti-
World War II, though some others, such as the tude to support for science, most technologically
Geophysics Institute at Bergen in Norway, estab- advanced nations began paying more attention to
lished in 1917, also did important marine work. marine research because of a variety of economic
Existing educational and administrative struc- and political needs. They also discovered the bene-
tures did not generally lend themselves to such fits of international co-operation. By 1900 much of
developments, and private wealth was not usually the marine survey work connected with submarine
available on the scale required for such enterprises. cables was being done by the cable companies
In spite of this and the somewhat differing attitudes themselves, but naval hydrographers also continued
to science in different countries, marine science their interest in deep-sea work 28. At the Inter-
developed through a variety of local, national, and national Geographical Congress held in Berlin in
international agencies during the period from the 1899, it was decided that details of all these sound-
1880s to the 1930s. This happened partly through ings should be collected onto a continually updated
the expansion of higher education. University General Bathymetric Chart of the Ocean. This
departments of oceanography were created – in the work was undertaken by Prince Albert of Monaco,
UK there was one at Liverpool, established in 1919, and after his death, by the International
and one at Hull, established in 1928. Other scientif- Hydrographic Bureau which had been established
ic departments also did marine work. A consider- in Monaco in 1921.
able number of marine biological laboratories were Owing to the growing economic importance of
also attached to universities, although some, like the fisheries, scientific research in this area developed
Marine Biological Association’s laboratory at rapidly during the late nineteenth century. National
Plymouth, were maintained by private bodies. John research organisations were set up in many coun-

22
 1: How the Science of Oceanography Developed

tries13 to study the life histories of food fish, in the oceanography. In the 1870s the American meteo-
hope of reviving dwindling fisheries and creating rologist, William Ferrel, drew attention to the
new ones. From the start it became apparent that it deflecting effect of the Earth’s rotation (the Coriolis
was not sufficient just to study individual species; force) on ocean currents32, but it was the circula-
more needed to be known about biological diversi- tion theorem of Vilhelm Bjerknes and work by
ty in selected regions, and about the physical envi- Bjorn Helland-Hansen and Fridtjof Nansen on cur-
ronment and its influence on populations. This was rents in the Norwegian sea, carried out in the fish-
the starting point for much local activity, such as ery steamer Michael Sars in the early 1900s, that
H.B. Bigelow’s biological survey of the Gulf of formed the basis for modern dynamical oceanogra-
Maine in the early 1900s32. However, the wider phy. The behaviour of oceanic winds and currents
questions posed had a considerable impact on the had been of crucial importance for Nansen’s
development of oceanography in the broad sense. famous attempt to drift to the North Pole in the
In the 1890s, Scandinavian oceanographers pro- Fram (1893–1896). Observations he made then
posed that there should be a joint programme of also formed the basis of two important discoveries
observations and this resulted in the setting up of by V.W. Ekman35. One was the existence of inter-
the first intergovernmental body for marine science, nal waves, now known to occur naturally through-
the International Council for the Exploration of the out the ocean, at the interface between layers of
Sea (ICES), in 1902. For several years the Council water of differing density. In northern seas such
maintained a Central Laboratory in Norway, waves are generated when a ship moves through a
where important work was done toward improving shallow layer of fresh water, originating from rivers
oceanographic apparatus. Surveys of important or melting ice, that overlays normal sea water and
fishing grounds were carried out in North America, hampers the ship’s progress, a phenomenon known
Australia, and by the European nations, both at to sailors as ‘dead water’. The ‘Ekman spiral’ is the
home and in their colonies around the world. name now given to the discovery that the direction
British scientists working for the Discovery of near-surface currents is increasingly deflected
Committee carried out research in the Southern with depth. This results from the Earth’s rotation
Ocean (Figure 1.19) during the 1920s and 1930s, and from frictional forces, and causes a mean cur-
with the aim of putting the whaling industry on a rent drift to the right of the wind in the northern
sustainable basis. hemisphere (and to the left in the southern
Work undertaken by these various organisa- hemisphere).
tions, as well as by individual institutions and expe- Important contributions were also made at this
ditions, contributed to the growing knowledge of time to our knowledge of general oceanic circula-
many aspects of the oceans during the late nine- tion. In his report on the work of the German
teenth and early twentieth centuries. The two most research ship Meteor in the 1920s, Georg Wüst39
important areas of research developed during these incorporated data from earlier expeditions to show
years were ocean circulation studies and biological the origin and distribution of the main Atlantic

Figure 1.19 R.R.S. Discovery II, at Port 1.19

Lockroy, Wiencke Island, Palmer Archipelago,
off the Antarctic Peninsula, in January 1931.
The Discovery II was built in 1929 for the
Discovery Committee, to replace the sailing
vessel Discovery (originally built for Captain
Scott in 1901) which had been used for initial
work in the seas around South Georgia from
1925–1927. Discovery II worked throughout
the Southern Ocean during the 1930s, con-
tributing to an understanding of ocean circula-
tion and the marine environment. A central
theme was the study of the distribution and life
history of Antarctic krill, the principal food of
southern hemisphere baleen whales. After
World War II, Discovery II became the
research vessel of the newly formed National
Institute of Oceanography (later Institute of
Oceanographic Sciences), until replaced by
the modern R.R.S. Discovery (built 1962).
(Courtesy of the Southampton Oceanography
Centre, Southampton, England.)

23
 M.B. Deacon

1.20

Figure 1.20 Diagram (Plate 33) by Georg Wüst39 showing the longitudinal distribution of salinity in the Atlantic Ocean.
Wüst was one of the leading physical oceanographers of the mid-twentieth century. As a young man in the 1920s he
participated in the cruise of the German research vessel Meteor and used data collected during the voyage, together
with observations made by other expeditions, from the Challenger onward, to show how density differences, due to
salinity and temperature variations, are responsible for the movement of water masses within the body of the ocean.

Modern Oceanography
water masses (Figure 1.20). His picture was extend- When Thomas Wayland Vaughan, the Director of
ed by Discovery Committee scientists working in the Scripps Institution of Oceanography, surveyed
the Southern Ocean in the 1920s and 1930s24. world oceanographic activity on behalf of the US
Fisheries science was also responsible for the National Academy of Sciences in the 1930s, he
growth of interest in marine productivity. This field found nearly 250 institutions devoted to marine
was developed by German scientists of the ‘Kiel research throughout the world, from Russia to
School’ 18 in the late nineteenth century. Their Japan and from Australia to Argentina. He had
researches showed that life in the sea depends on pri- been asked to undertake this survey because of con-
mary production – the phytoplankton, single-celled cern that not enough was being done to encourage
plants grazed by the zooplankton on which all other the study of the sea, which, as we have seen,
creatures in the sea depend, either directly or indi- requires special conditions to make progress. The
rectly, for food (modern research also highlights the trend toward increasing state support for science
importance of bacteria in primary production – see was soon to be greatly accelerated by the demands
Chapter 6). Victor Hensen18 developed vertical nets of war, and of the Cold War which followed, and
for sampling plankton in order to obtain quantita- by the role of science in post-war economic growth.
tive data on the productivity of the ocean. In the Oceanography developed rapidly during and
Plankton Expedition of 1889 he extended his work after World War II. It shared in the general expan-
to the North Atlantic in the ship National. Karl sion of science, as the impact of science and tech-
Brandt18 showed that the growth of phytoplankton nology on almost every aspect of modern life has
is controlled by the supply of available nutrients. led to increased funding for research. National

24
 1: How the Science of Oceanography Developed

Figure 1.21 Launching the clover-leaf buoy from R.R.S. 1.21

Discovery in 1967. The 1950s and 1960s saw a dramatic
expansion in oceanography, encompassing people, ideas,
and methods. This apparatus was one of a number of dif-
ferent types designed and made at the National Institute of
Oceanography for studying properties of waves. (Courtesy
of Laurence and Pamela Draper, Rossshire, Scotland.)

defence needs and economic concerns, particularly

the search for new energy sources (offshore oil and
gas), as well as renewed concern about fisheries
stocks, are among the factors that have led to a
wider interest in marine research. Such needs have
increased rather than diminished with time. While
the political international situation has eased over-
all, environmental worries have come to the fore, in During the past 50 years such developments,
particular the problem of climate change, in which and others described in the following chapters,
the oceans must play an important part. have profoundly transformed our knowledge of the
However, it is interesting to see how many of the oceans. The main subject areas – marine physics,
ideas and techniques that are important in present- including ocean-circulation studies, knowledge of
day oceanography have their roots in the first half of the sea floor (which has played a major role in the
the twentieth century. For example, work on under- revolution of the earth sciences leading to modern
water sound was begun before World War I9, but theories of plate tectonics), marine chemistry, and
then developed rapidly in the search to perfect a biological oceanography – have all made important
means of submarine detection by echo location advances. The development and use of more
(sonar). The invention of the hydrophone had many sophisticated apparatus has been assisted by the
other scientific and peace-time applications. Echo- introduction of computers and satellites, which per-
sounding by ships provided more detailed informa- mit the gathering, transmission, and analysis of
tion about the topography of the sea floor. Seismic data in quantities that would have been inconceiv-
surveying, to investigate the internal structure of the able a generation ago, let alone to the first scientific
sea bed, was first employed by Maurice Ewing in the observers of the sea in the 1660s. Oceanography is
1930s. After 1945, this work played an important still an expensive science, with the ship remaining a
part in obtaining the information on which modern fundamental tool, although this too may change in
ideas about sea-floor spreading and plate tectonics the future. This expense has provided a strong
are based32. Magnetic surveys also contributed to incentive for co-operation and sharing on a more
these developments, as did gravity measurements, formal basis, so joint expeditions have become an
first obtained at sea in submarines in the 1930s by important aspect of modern oceanography, from
the Dutch scientist, F.A. Vening-Meinesz. Modern the International Indian Ocean Expedition of the
oceanographers rely heavily on continuous-recording early 1960s to the World Ocean Circulation
instruments, which in many areas have replaced the Experiment (WOCE), designed to throw new light
older single-observation measuring devices (see on the relation between the oceans and climate, in
Chapter 19). Such devices were foreshadowed by the the 1990s.
bathythermograph, invented by US scientists in the
1930s to measure the temperature of the upper lay-
ers of the ocean. This and similar ideas were taken
up by scientists during World War II. General References
Other fields of study originated at that time, Deacon, M.B. (1971), Scientists and the Sea, 1650–1900: A
including wave research. Until then no way had Study of Marine Science, Academic Press, London
and New York, 445 pp.
been found to study sea waves that was not purely Herdman, W.A. (1923), Founders of Oceanography and their
descriptive32. A major breakthrough occurred when Work: An Introduction to the Science of the Sea,
a war-time research team based at the Admiralty Edward Arnold, London, 340 pp.
Research Laboratory, in Teddington, England, McConnell, A. (1982), No Sea Too Deep: The History of
Oceanographic Instruments, Adam Hilger, Bristol,
developed a method of analysing wave spectra that 162 pp.
enabled their components to be identified. This Rice, A.L. (1986), British Oceanographic Vessels, 1800–1950,
work continued after the war at the newly estab- The Ray Society and Natural History Museum,
lished National Institute of Oceanography, where London, 193 pp.
new kinds of wave recorders were developed to Schlee, S. (1973), The Edge of an Unfamiliar World: A History
of Oceanography, E.P. Dutton, New York, 398 pp.
measure waves at sea (Figure 1.21).

25
M.B. Deacon

References
1. Birch, T. (ed.) (1744), The Works of the Honourable Robert 22. Rehbock, P.F. (ed.) (1992), At Sea with the Scientifics: The
Boyle, A. Millar, London, Vol. 3, pp 105–113 and Challenger Letters of Joseph Matkin, University of
378–388. Hawaii Press, Honolulu, 415 pp.
2. Bourne, W. (1578), A Booke Called the Treasure for 23. Rennell, J. (1832), An Investigation of the Currents of the
Traveilers, Thomas Woodcocke, London, 269 pp. Atlantic Ocean, J.G. and F. Rivington, 359 pp.
3. Burstyn, H.L. (1966), Early explanations of the role of the 24. Rice, A.L. (1986), British Oceanographic Vessels,
Earth’s rotation in the circulation of the atmosphere 1800–1950, The Ray Society and Natural History
and the ocean, Isis, 57(2), 167–187. Museum, London, 193 pp.
4. Burstyn, H.L. (1971), Theories of winds and ocean currents 25. Rice, A.L., Burstyn, H.L., and Jones, A.G.E. (1976), G.C.
from the discoveries to the end of the seventeenth cen- Wallich, M.D. – megalomaniac or mis-used oceano-
tury, Terrae Incognitae, 3, 7–31. graphic genius?’ J. Soc. Bibliogr. Natur. Hist., 7,
5. Deacon, M.B. (1971) Scientists and the Sea, 1650–1900: a 423–450.
Study of Marine Science, Academic Press, London 26. Richard, J. (1910), Les Campagnes Scientifiques de S.A.S. le
and New York, 445 pp. Prince Albert Ier de Monaco, Imprimerie de Monaco,
6. Deacon, M.B. (1985), An early theory of ocean circulation: p. 33.
J.S. von Waitz and his explanation of the currents in 27. Richardson, P.F. (1980), The Benjamin Franklin and
the Strait of Gibraltar, Progr. Oceanogr., 14, 89–101. Timothy Folger charts of the Gulf Stream, in
7. De Vorsey, Jr, L. (1976), Pioneer charting of the Gulf Oceanography: The Past, Sears, M. and Merriman,
Stream: the contributions of Benjamin Franklin and D. (eds), Springer, New York, pp 703–717.
William Gerard de Brahm, Imago Mundi, 28, 28. Ritchie, G.S. (1967), The Admiralty Chart. British Naval
105–120. Hydrography in the Nineteenth Century, Hollis and
8. Duhem, P. (1913–1959), Le Système du Monde: Histoire des Carter, London, 388 pp. (Reprinted 1995 by Pentland
Doctrines Cosmologiques de Platon à Copernic, 9 Press, Edinburgh.)
vols, Hermann, Paris. 29. Ross, Sir J.C. (1847), A Voyage of Discovery and Research
9. Hackmann, W. (1984), Seek and Strike: Sonar, Anti- in the Southern and Antarctic Regions, during the
Submarine Warfare and the Royal Navy, 1914–54, years 1839–43, Vol. 2, John Murray, London, facing
HMSO, London, 487 pp. p. 354.
10. Hales, S. (1754), A descripton of a sea gage, to measure 30. Ross, M.J. (1982), Ross in the Antarctic: The Voyages of
unfathomable depths, Gentleman’s Magazine, 24, James Clark Ross in H.M. Ships Erebus and Terror,
215–219. 1839–43, Caedmon of Whitby, Whitby, Yorkshire,
11. Halley, E. (1697), The true theory of the tides, extracted 276 pp.
from that admired treatise of Mr Isaac Newton, enti- 31. Royal Society (1667), Directions for observations and exper-
tled Philosophiae Naturalis Principia Mathematica; iments to be made by masters of ships, pilots, and
being a discourse presented with that book to the late other fit persons in their sea-voyages, Phil. Trans.
King James, Phil. Trans. Roy. Soc. Lond., 19, Roy. Soc., Lond., 2, 433–448.
445–457. 32. Schlee, S. (1973), The Edge of an Unfamiliar World: A
12. Hansard (1961), Parliamentary Debates, Fifth Series, 638, History of Oceanography, E.P. Dutton, New York,
235. 398 pp.
13. Herdman, W.A. (1923), Founders of Oceanography and 33. Stoye, J. (1994), Marsigli’s Europe, 1680–1730: The Life
their Work: An Introduction to the Science of the Sea, and Times of Luigi Ferdinando Marsigli, Soldier and
Edward Arnold, London, 340 pp. Virtuoso, Yale University Press, New Haven and
14. Kircher, A. (1678), Mundus Subterraneus, Vol 1, 3rd edn, London, 356 pp.
Apud Joannem Janssonium à Waesberge & Filios, 34. Vossius, I. (1993), A Treatise Concerning the Motions of the
Amstelodami, pp 134–135. Seas and the Winds, together with De Motu Marium
15. Marsigli, L.F. (1725), Histoire Physique de la Mer, Aux et Ventorum Liber, Deacon, M.B. (ed.), Scholars’
dépens de la Compagnie, Amsterdam, 195 pp. Facsimiles and Reprints, Delmar, New York, for the
16. McConnell, A. (1982), No Sea Too Deep: The History of John Carter Brown Library, 376 pp.
Oceanographic Instruments, Adam Hilger, Bristol, 35. Walker, J.M. (1991), Farthest North: Dead water and the
162 pp. Ekman spiral, Weather, 46(4), 103–107; 46(6),
17. McConnell, A. (1990), The flowers of coral – some unpub- 158–164.
lished conflicts from Montepellier and Paris during 36. Wild, J.J. (1877), Thalassa: An Essay on the Depth,
the early 18th century, Hist. Phil. Life Sci., 12, Temperature, and Currents of the Sea, Marcus Ward,
51–66. London, facing p. 14.
18. Mills, E.L. (1989), Biological Oceanography: An Early 37. Wild, J.J. (1878), At Anchor. A Narrative of Experiences
History, 1870–1900, Cornell University Press, Ithaca Afloat and Ashore during the Voyage of H.M.S.
and London, 378 pp. Challenger from 1872 to 1876, Marcus Ward,
19. Murray, J. and Renard, A.F. (1891), Deep-sea deposits, London, 198 pp.
Report on the Scientific Results of the Voyage of 38. Winsor, M.P. (1976), Starfish, Jellyfish, and the Order of
H.M.S. Challenger during the Years 1872–76, Life: Issues in Nineteenth Century Science, Yale
HMSO, London, Plate XI. University Press, New Haven and London, 288 pp.
20. Pinsel, M.I. (1981), The wind and current chart series pro- 39. Wüst, G. (1928), Der Ursprung der Atlantischen
duced by Matthew Fontaine Maury, Navigation, Tiefenwässer, Zeitschrift der Gesellschaft für
28(2), 123–137. Erdkunde zu Berlin Sonderband zur Hundertjahrfeier
21. Rehbock, P.F. (1979), The early dredgers: ‘naturalizing’ in der Gesellschaft, pp 506–534.
British seas, 1830–1850, J. Hist. Biol., 12(2), 40. Wüst, G. (1964), The major deep-sea expeditions and research
293–368. vessels, 1873–1960, Progr. Oceanogr., 2, 1–52.

26
CHAPTER 2:

The Atmosphere
and the Ocean
H. Charnock

Introduction and of the underlying ocean surface, together with

The atmosphere and the ocean are held on the the limited but increasing observations of the deep
Earth by gravity and irradiated by the Sun. Both ocean, to study the global transfers of heat, of fresh
are shallow relative to the radius of the Earth and water, and of momentum in the coupled atmos-
the motions within them are slow relative to that phere–ocean system.
due to the Earth’s rotation: they have similar
dynamics. As they share a common boundary it is Energy and Water Exchanges in the
attractive to treat them as a single, coupled, system, Atmosphere–Ocean System
but the physical and chemical properties of air and The general circulation of the atmosphere and
water are so very different that meteorology and ocean, regarded as a single system, is determined by
oceanography have developed separately and at dif- the distribution of its sources and sinks of energy.
ferent rates. Electromagnetic radiation (light, Much the dominant external source is the absorp-
microwaves, radar, radio …) travels easily through tion of energy from the Sun (solar radiation of
the atmosphere and this, together with the com- wavelength between 0.2–4 μm). The near constan-
mercial and economic benefit of weather-forecast- cy of the temperature of the system requires an
ing, has led to the existence of a global network of equal outgoing flux of energy in the form of terres-
meteorological observing stations, making and trial radiation (long-wave radiation of wavelength
transmitting regular routine surface and upper-air 4–100 μm).
observations, as well as increasing information The distribution and transformation, within the
from sensors on satellites. system, of the incoming solar radiation is compli-
Most meteorological stations are on land, but cated. Typical global mean values of the major
winds, waves, currents, and weather affect ships, so components are shown in Figure 2.1, which indi-
have been observed by mariners from time cates that nearly half the absorbed solar radiation
immemorial: some selected merchant ships now reaches the Earth’s surface, most of which is
report their observations as part of the global mete- ocean. This sunlight does not penetrate far into
orological system. the ocean (even in the clearest water 99% is
Observation of the ocean away from the surface absorbed in the upper 150 m): heat gains and
has developed more slowly (Chapter 4): water is losses take place at and close to the surface, so the
almost opaque to electromagnetic radiation so ocean is relatively inefficient thermodynamically.
oceanographers are essentially restricted to Directly driven motions, due to the cooling of sur-
indirect observation of a fluid through which they face water at high latitudes (the thermohaline cir-
cannot see; the period from the Challenger culation) are slow. Indirectly driven motions (the
Expedition of 1872 has been described as a ‘centu- wind-driven circulation) arise from the transfer of
ry of undersampling’. heat (and water vapour) from the ocean to the
More recent technological development has clar- atmosphere, where it is converted by complicated
ified some processes in restricted areas: it is now processes into depressions and anticyclones, the
accepted that, like the atmosphere, the ocean is a winds of which provide energy to generate ocean
three-dimensional turbulent fluid, with interacting waves and drive ocean currents. Their energy in
motions and processes on all time- and space- turn is dissipated into heat by small-scale (vis-
scales. Understanding it requires an observing sys- cous) processes and re-radiated. It is a complex,
tem which does not yet exist, one that may evolve inefficient system with many interlocking compo-
if national governments perceive a need to forecast nents of different scale.
conditions in the ocean to the same extent as those It can be seen from Figure 2.1 that of the
in the atmosphere. In the meantime, we can 153 W/m2 received at the Earth’s surface as solar
attempt to use our knowledge of the atmosphere radiation, 54 W/m2 is lost as long-wave radiation

27
 H. Charnock

2.1 Reflected from

Transmitted through
Space
clouds and
Earth's surface 340 Solar radiation atmosphere
187 51
17 Outer limits of
10 75
atmosphere

Scattered Atmosphere
by atmosphere Absorbed
85 Infra-red by atmosphere
Absorbed radiation 326
by atmosphere (510) 'Sensible'
emitted by Latent heat heat
Diffuse 62 37
Diffuse Direct radiation radiation atmosphere
(from clouds) to surface (blue sky)
44 82 27

153 323 –377 –62 –37

Solar radiation Infra-red radiation
absorbed at ground absorbed at ground
Sea and
Short-wave radiation Long-wave radiation Heat transferred
by evaporation
land
and conduction

+476 –476
Energy received Energy lost
by land from land
and sea surface and sea surface

and 37 W/m2 is transferred by the conduction of Figure 2.1 The average radiation balance (in W/m2) for
heat (sensible heat) from a warmer sea to cooler air the Earth as a whole (based on Nieburger et al.11).
above. Energy is also used in evaporation of water
vapour from the sea surface to the drier air above,
heat (latent heat) being required to convert a liquid are ultimately powered by solar heating, the impor-
into a gas. Water has a high latent heat, 62 W/m2 tant working substance of the global heat-engine is
being used in evaporation. The latent heat is water; as vapour, as liquid, and as solid ice. The
released to the atmosphere only when the water infra-red characteristics of water vapour make it a
vapour condenses back into clouds, of liquid water major agent of long-wave radiative heat transfers
or ice, often far from where it evaporated from the (see Figure 2.1) The solar radiation absorbed at the
ocean into the atmosphere. Much of the cloud for- surface is used to evaporate water, the large latent
mation happens at relatively low altitudes, justify- heat of which is released to the atmosphere when
ing the statement that the atmosphere can be and where condensation occurs; the distribution of
regarded as heated from below, making it an active evaporation, precipitation, continental run-off, and
thermodynamic system with important vertical as ice are crucial to the determination of the salinity
well as horizontal motion. and so to the watermass structure and to the ther-
Although both atmospheric and oceanic motions mohaline circulation of the ocean. For these and

2.2 Figure 2.2 Estimates of the amount

of water (in 103 km3) in the atmos-
Atmosphere 16 phere, the land, and the ocean, and
of the fluxes between them (1 Sv =
;

106 m3/s) (from Schmitt15; the figures

;

Evapo- Precipitation Evaporation Precipitation

;

transpiration 3.5 Sv 13.5 Sv 12.2 Sv are derived from Baumgartner and

;

2.2 Sv Reichel2).

Land Rivers
Oceans
59,000 1.3 Sv 1,400,000

28
 2: The Atmosphere and the Ocean

for many other reasons the basic question posed by Net incoming radiation 2.3
weather and climate – ‘what happens to the sun- 300 300
shine?’ – must be supplemented by asking – ‘what Outgoing radiation
happens to the water?’
The distribution of evaporation and precipita-
200 200
tion over the ocean is clearly vital to understanding
climate: unfortunately, this part of the hydrological

Radiation (W/m2)
cycle is not well known. Most treatments of the
water cycle concentrate on exchanges over land 100 100
Net radiation
(they are concerned with man’s use of water, for
agriculture and industry as well as human con-
sumption), but it is estimated that 78% of the glob-
al precipitation goes into and 86% of the global 0 0
90˚ 60˚ 30˚ 0˚ 30˚ 60˚ 90˚
evaporation comes from the ocean (see Figure 2.2). N Equator S
The corresponding transfer of (latent) heat repre-
sents a major component of the heat balance of the –100 –100
atmosphere and of the ocean.
Meridional Fluxes
The near-constancy of the global mean temperature Figure 2.3 Zonal averages of radiation at the top of the
implies a balance between the absorbed solar radia- atmosphere. Net incoming radiation peaks in the tropical
tion and the outgoing long-wave radiation, as indi- regions; outgoing long-wave radiation varies less with lat-
cated in Figure 2.1, but their variation with latitude itude. To maintain a constant temperature, the excess of
is significantly different. The solar radiation is radiative heating within 35° of the equator is transferred
poleward by atmospheric and oceanic motions to com-
absorbed mainly in the tropics: the long-wave radi-
pensate for the deficit of radiative heating nearer the
ation, determined mainly by the radiative proper- poles. A uniform bias of 9 W/m2 has been subtracted
ties of the atmosphere and the underlying surface, from the net incoming radiation to ensure a balance
is observed to be much less dependent on latitude between the total incoming and the total outgoing radia-
(Figure 2.3). It follows that there is a flux of heat tion (from Bryden4; the figures are derived from satellite
from the tropics to the poles. Measuring this flux is observations reported by Stephens et al.18).
important: it is fundamental to the maintenance of

120˚E 180˚ 120˚W 60˚W 0˚ 60˚E 120˚E 2.4

70˚N 70˚N
60˚ 0 60˚
50˚ 0
50˚
-80
40˚ -40 0 40˚
0 40
30˚ 40 30˚
120
20˚ 80 80 120 20˚
0
120

10˚ 0 10˚
0

-40

80
0˚ 40
0˚
0 -40
10˚ 0 10˚
120
40
20˚ 120 20˚
40 80
0
30˚ 120 30˚
80 80
40˚ 0 40 40˚
0

0
0
50˚ -40 50˚
-40 -40 -40
60˚ -40 60˚
70˚S 70˚S
120˚E 180˚ 120˚W 60˚W 0˚ 60˚E 120˚E

Figure 2.4 A chart of evaporation minus precipitation (E-P) over the ocean. Units are cm/yr; solid lines indicate E>P,
dashed lines E<P (from Schmitt and Wijffels17; the figures are derived from Schmitt et al.16).

29
 H. Charnock

2.5 80˚ 80˚ 80˚ 80˚ 80˚ 80˚

60˚ 60˚ 60˚ 60˚ 60˚ 34 35 60˚

32
32 35
33 35
40˚ 40˚ 40˚ 40˚ 40˚ 40˚
34
40 35
30 > 37
20˚ 41 20˚ 20˚ 20˚
40 33 35
38 32 36 34
34 28 34
36 34 33 35 35 30
0˚ 30 0˚
35 36
37
35 34 36 > 37
20˚ 20˚ 20 20˚
> 36
36 36
35 35 35
40˚ 40˚ 40˚ 40˚ 40˚
35 33 34
34 34
< 34 34 34 < 34
60˚ 60˚ 60˚ 60˚ 60˚ 60˚
< 34
80˚ 80˚ 80˚ 80˚ 80˚ 80˚

Salinity greater than 36 parts per thousand

Salinity 34–36 parts per thousand

Salinity less than 34 parts per thousand

climate and its magnitude provides a significant Figure 2.5 Salinity of surface waters during the northern
constraint on atmospheric, oceanic, and coupled summer (from Gross 6 ; the figures are derived from
general circulation models. Understanding how the Sverdrup et al.20 and from later sources).
meridional heat flux is maintained requires a
detailed knowledge of atmospheric and oceanic years been measured as part of a major programme
structures and processes. (the Earth Radiation Budget Experiment, ERBE)
Although the distribution of evaporation and using orbiting satellites fitted with radiometers sen-
(especially) of precipitation over the ocean is not sitive to solar and to long-wave radiation.
well known, major features of the net water flux Although the measurements are technically
between atmosphere and ocean (evaporation minus demanding and present difficult problems of data
precipitation, E–P) can be recognised (Figure 2.4). analysis, recent results imply a near-balance
The distribution is roughly zonal (except for the between the global annual incoming solar radiation
North Indian Ocean); precipitation dominates in and the corresponding outgoing long-wave radia-
subpolar regions and especially in the Intertropical tion. The imbalance was small (less than 10 W/m2),
Convergence Zone (the ITCZ) at the thermal equa- but its distribution produces some uncertainty in
tor and in the South Pacific Convergence Zone to the derived values of the meridional heat transfer
the northeast of Australia. Elsewhere, the subtrop- by the atmosphere and ocean combined. Figure 2.6
ics have an excess of evaporation. There is a clear shows that it peaks near latitudes 30°N and 40°S
association with the surface salinity of the ocean where the flux amounts to almost 6 PW
(Figure 2.5); the latitudes where (E–P) is high are (1 petawatt, PW = 1015 W).
associated with high salinity at the ocean surface Estimates of the meridional flux of water and
(and with deserts on land). The general pattern is heat (and momentum) by the atmosphere can be
of net annual precipitation at high and at low lati- made using the observations made daily from mete-
tudes, with net annual evaporation between. That orological upper-air stations, where balloon-borne
the mean structure is not changing implies merid- instruments measure the wind, the temperature,
ional transports of fresh water by the ocean. As and the humidity in relation to pressure (height).
river transports are negligible in comparison, equal The local meridional flux of, say, water vapour is
and opposite flows of water must occur in the carried by the northward component of the wind,
atmosphere. These, in turn, transfer significant V, and is measured by lVq where q is the specific
quantities of heat. humidity and l the air density. The total meridion-
To assign numerical values to these important al flux is given by [lVq] where the square brackets
meridional fluxes presents difficulties. The radia- represent a mean value over the depth of the
tion balance of the Earth as a whole has for many atmosphere and around a latitude circle, over a

30
 2: The Atmosphere and the Ocean

2.6 Jacobwitz et al. 8 (and therefore density), at all depths. The difficulty
ERBE – four validation
6.0 months1
in determining deep-ocean currents geostrophically
Vonder Haar and Ellis21 is that although the density field is known at all
Gruber7 depths the pressure field is not known at any one,
4.0
Ellis and Vonder Haar5
so the calculation of currents using the geostrophic
Energy transport (PW)

2.0 balance requires a knowledge of the total transport.

Given measurements or reliable estimates of the
0.0 transport of western boundary currents, and mak-
ing allowance for the near-surface currents due to
–2.0 the frictional drag of the wind, convincing esti-
mates of the meridional heat flux by the ocean can
– 4.0 be obtained. Estimates of the heat flux northward
across latitude 24°N in the Atlantic have been
–6.0
–90 –60 –30 0 30 60 90 made three times over the last 35 years and found
S
Latitude (˚)
N to be consistently close to 1.2 PW to the north. A
transoceanic hydrographic section made in 1985
Figure 2.6 The total meridional transport of energy esti-
has provided observations along 24°N in the
mated from ERBE data for April, July, and October 1975 Pacific from which an oceanic heat flux of 0.8 PW
and January 1986 (from Barkstrom et al.1; other satellite- was obtained. Since there is virtually no heat flux
based estimates are from Vonder Haar and Ellis21; Ellis across 24°N in the Indian Ocean the total ocean
and Vonder Haar5; Gruber7; Jacobwitz et al.8). heat-flux across 24°N amounts to some 2.0 PW to
the north.
If the atmospheric flux (Figure 2.7) across 24°N
time for which [V] vanishes. The calculations can is taken to be 2.3 PW, the total (atmosphere +
be made by using the upper-air observations direct- ocean) flux is 4.3 PW, significantly less than the
ly or by assimilating them together with other 5.7 PW required from the most recent analysis of
observations into a suitable atmospheric computer the ERBE results. Further transoceanic sections are
model and using interpolated values. being made as part of the World Ocean Circulation
The meridional flux of water vapour in the Experiment and should serve to clarify the climati-
atmosphere (which is discussed later) implies a cor- cally important meridional heat flux. They will be
responding flux of latent heat. To estimate the total particularly important in the southern hemisphere,
heat flux this must be supplemented by the sensible
heat flux given by [lCpVq], where Cp is the specific 8
heat of air at constant pressure and q the tempera- 2.7
ture. Some results are shown in Figure 2.7 for the 6
total heat flux (compare with Figure 2.2). The
atmospheric fluxes peak at about 40°N and 40°S 4
where the flux amounts to about 4 PW.
There is no oceanic equivalent of the meteoro- 2
(PW)

logical upper-air observing network, but it is possi- 0

ble to estimate the meridional flux of heat (and that
[vY]

of salt) by an analogous method: the heat transport –2

is estimated by calculating the covariance (Vq)
between the temperature and the inferred, as dis- –4
Michaud and Derome10
tinct from the directly measured, northward veloci- Oort and Peixoto13
ties. In the deep ocean away from the surface and –6
Masuda9
the sea floor, the currents can be treated as friction- –8
less and unaccelerated, so can be calculated as if 90S 60 30 0 30 60 90N
they are in geostrophic balance: that is, the force Latitude (˚)
due to the horizontal variation in pressure is bal-
anced by the force due to the Earth’s rotation,
Figure 2.7 Annual mean northward flux of energy by the
which is proportional to the speed of the current. atmosphere, December 1985 to November 1986 (from
In this case the currents flow along the isobars like Michaud and Derome10). The estimates of Michaud and
winds on a weather map. The observations used Derome10 are compared with those of Oort and Peixoto13
are from a high-quality east–west hydrographic sec- (observed values from upper-air observations) and of
tion between two continental land masses, giving Masuda 9 (ECMWF and GFD model assimilations for
accurate measurements of temperature and salinity 1978–1979).

31
H. Charnock

where the upper-air observing stations are even agreed with the ‘direct’ flux estimates but they are
more sparse than in the northern hemisphere and not consistent from ocean to ocean and do not
the estimates of atmospheric heat flux correspond- help to resolve the discrepancy between space,
ingly less certain. atmospheric, and oceanic estimates. Surprisingly,
Information on the meridional transport of heat perhaps, recent estimates from the water balance
and water can also be obtained by considering the show a gratifying anti-symmetry with the ‘direct’
heat and water balance of the ocean. A particular atmospheric values (Figure 2.8).
vertical column of the ocean gains heat from the The trans-Atlantic and trans-Pacific hydrograph-
absorption of solar radiation and loses it by the ic sections used to estimate meridional heat trans-
emission of long-wave radiation and as the latent port have also been used to estimate the corre-
heat used in evaporation: it gains or loses sensible sponding salt transport. Given the transport of
heat depending on whether the sea is colder or water from the North Pacific to the Arctic through
warmer than the air. Empirical formulae are used the Bering Strait these salt transports can be used to
to estimate the heat gains and losses using long calculate the corresponding freshwater fluxes
series of surface meteorological observations of across 24°N in the Atlantic and the Pacific. The
cloudiness, windspeed, air temperature, and sea results are 0.9 Sv southward in the Atlantic and
temperature made regularly from merchant ships. 0.6 Sv northward in the Pacific. The former is
In general there is a surplus or a deficit of heat; if about 0.1 Sv greater and the latter about 0.1 Sv
the imbalance is averaged over a time such that smaller than that calculated from the surface
heat associated with change of the heat content of (E–P–R) distribution, their sum being in reasonable
the ocean can be neglected, the remainder, errors agreement with it and with the global estimates
apart, is due to advection of heat by ocean cur- using upper-air observations: the global flux is
rents. From averages over a latitude circle between small, as would be expected for a latitude in the
continents the meridional oceanic heat flux is centre of the subtropical gyre.
obtained by integrating from the North or the Meridional flux calculations using climatological
South Pole, or from other latitudes where the heat data from merchant ships appear to have reached
flux is zero or known. their limits of accuracy. They have produced valu-
Estimating the meridional flux of fresh water by able patterns of heat and water exchange at the sea
ocean currents using surface meteorological obser- surface but the extent to which they can be used
vations is, in principle, simpler since water is effec- quantitatively is limited: ‘direct’ methods in both
tively confined to the ocean–ice–atmosphere sys- the atmosphere and the oceans seem inherently
tem. The meridional flux can be estimated by inte- more accurate. There is a need for more analysis of
grating zonal averages of Evaporation minus meteorological upper-air data and especially for
Precipitation minus Runoff (E–P–R) from some lat- more high-quality trans-oceanic hydrographic
itude where the flux is known. In practice estimates sections.
are uncertain since none of the terms is well
known, there being as yet no good way of measur- Transfer Mechanisms
ing rainfall at sea. Precipitation estimates are made The ‘direct’ methods estimate meridional fluxes by
using statistics of its frequency, type, and intensity calculating the covariance of temperature and of
from the ‘present weather’ reports from routine humidity with observed winds (in the atmosphere)
merchant ship weather reports. and with inferred currents (in the ocean). They
The accuracy of meridional flux estimates made therefore provide information both about the flux-
from heat and water exchanges at the sea surface es and also about the mechanisms by which they
appears to be inferior to those made by the ‘direct’ are brought about.
covariance technique: the surface ships’ observa- The structure of the much better observed
tions are not well distributed, they have random atmosphere is reasonably well known; it can be
and probably some systematic errors, and the thought of as a gigantic – and very inefficient –
empirical formulae used are necessarily crude. Also heat engine, absorbing heat in the hot equatorial
it is difficult, sometimes impossible, properly to belt and losing it nearer the poles. Since much of
allow for the transport brought about by flow the heat transfer is due to evaporation and conden-
from one ocean basin to another, notably the sation of water vapour, one can think of the mech-
Antarctic Circumpolar Current, and the anism as a steam engine and use such expressive
Indonesian Through Flow from the Pacific to the terms as ‘the equatorial firebox’ and regard the
Indian (see, e.g., Figure 3.3). The flow from the giant cumulo-nimbus cloud towers of the ITCZ as
North Pacific to the Arctic Basin is small but well the cylinders in which the steam is converted into
established. Some ocean heat budget analyses have liquid water drops and ice crystals. How does the

32
 2: The Atmosphere and the Ocean

1.0 2.8
0.8

0.6
Transport (x106 m3/s)

0.4
Atmosphere
0.2

0.0
Mississippi
–0.2
Ocean
–0.4

–0.6

–0.8

–1.0
90N 80 70 60 50 40 30 20 10 0 10 20 30 40 50 60 70 80 90S
Latitude (˚)
Figure 2.8 Northward transport of fresh water as a function of latitude in both the atmosphere and
the ocean (from Schmitt and Wijffels17; the ocean transports are derived from Wijffels et al.22 and
the atmospheric transports are derived from Peixoto and Oort 14). The transport of the River
Mississippi is shown for comparison.

atmosphere bring about this huge heat transfer which has risen from the equator tends to retain its
from the tropics to the polar regions? There are angular velocity as it moves poleward. When it
two main mechanisms, one which operates between reaches 30° or so it is moving much faster toward
the equator and about 30°N and S and another the east than the earth beneath it (Figures 2.9 and
which operates at higher latitudes. 2.10).
In the mechanism which operates at lower lati- The Hadley cells provide a good explanation for
tudes the essential feature is that the air rises at the many observed phenomena, and the mechanism of
thermal equator and then spreads northward and the general atmospheric circulation in low latitudes
southward toward the poles. Most of it sinks again is relatively well understood. In both hemispheres,
at about 30°N or S and returns toward the equa- poleward of 30°N and S, the heat transfer is not
tor, acquiring an easterly component due to the brought about by motions of the Hadley cell type
rotation of the Earth, and forming the Trade but by disturbances on a smaller scale. Here the
Winds. These carry less energy than the poleward prevailing winds are westerly at all levels, unlike
winds aloft so there is a net transport of energy those of the tropical regions where the Trade
from low latitudes to high. Winds at the surface have an easterly component.
The essential features of this circulation are the Embedded in the westerlies of middle latitudes are
two closed wind patterns north and south of the the travelling depressions which provide typically
equator; they were described by Hadley as long ago unsettled weather. They have a complicated wind
as 1735 and are known as Hadley cells. Their exis- and temperature structure, such that air which is
tence provides a rational explanation for the equa- warmer than average tends to be going poleward
torial rain belts (ascending air motion) and the sub- and air which is colder than average tends to be
tropical anticyclones around 30°N or 30°S. The going toward the equator. On average the heat is
subtropical highs are linked with descending air transferred in the required sense – from equator to
motions, little rainfall, and hence the desert areas pole. The disturbed westerly regime and the com-
on land and the deep blue and high salinity water plex travelling depressions are essential features of
at sea. The associated horizontal winds are the the heat transfer. The situation is obviously compli-
Trades, at the surface, and the subtropical jet cated but the overall features of the general circula-
stream aloft. The subtropical jet stream is a band of tion of the atmosphere are clear. Dynamical meteo-
strong westerly winds in latitudes 30° to 40° at a rologists are using computers to simulate the
height of about 12 km. It arises because the air behaviour of the atmosphere, with results

33
 H. Charnock

2.9
Earth's Rotation
CLIMATIC ZONES POLAR VERTICAL CIRCULATION
EASTERLIES
POLAR CELL
POLAR
k JET STREAM
k Low 60˚N Polar front
SUBPOLAR k k
k FERREL CELL
TEMPERATE
High 30˚N
SUBTROPICAL
'Horse
latitudes' NORTHEASTERLY TRADES HADLEY CELL
Clear skies k k
Low rainfall k
ITCZ Intertropical
convergence

k
k 0˚
EQUATORIAL
k k k zone (ITCZ)
Persistent
cloudiness SOUTHEASTERLY TRADES HADLEY CELL
High rainfall
SUBTROPICAL
'Horse latitudes' High 30˚S
k

TEMPERATE k
FERREL CELL
'Roaring Forties' k k
k
Frequent storms WESTERLIES JET STREAM
k 60˚S Antarctic front
k
k
POLAR CELL
Antarctica

Figure 2.9 Schematic representation of features of the general circulation of the atmosphere (from Gross6).

Pole
2.10
High
Mainly easterlies Cyclonic
}

60˚N
Low
Low Mainly westerlies
Low
Anticyclonic
30˚N
High High
High Trades

Equator
Low

Figure 2.10 Atmospheric pressure at the surface of an idealised Earth as it might be on a partic-
ular day in comparison with the long-period average of Figure 2.9 (from Sutcliffe19).

34
 2: The Atmosphere and the Ocean

80˚ 60˚ 40˚ 20˚ 0˚ 20˚ 40˚ 60˚ 80˚ 100˚ 120˚ 140˚ 160˚ 180˚ 160˚ 140˚ 120˚ 100˚ 80˚ 2.11

y
wa
or
N
60˚
A

io

las
tic

ash
tlan

k
th A

Oy

a
Nor Subarctic
40˚ North Pacific

Ca
shio
ry

lif
uro
na

or
Ca

ni
K
20˚

a
North Equatorial
North Equatorial
Sou Countercurrent
0˚ th Equato
ria North Equatorial
l Countercurrent South Equatorial
l
Brazi

20˚ South Equatorial

as
ulh

40˚ Ag
nd

Peru
kla

Wind Drift
Fal

West nd Drift
West Wi
60˚

Figure 2.11 Schematic chart of the climatological average of the surface currents of the ocean (based on NRDC12).

sufficiently life-like to make one feel that the essen- transmission of radiant energy, and it must be
tial physics of the situation is correct. recognised that their geometry also plays an impor-
The winds that bring about transfer of heat and tant role. There are no barriers in the atmosphere
water in the atmosphere have to be such as to con- which correspond to the continental barriers to the
serve angular momentum about the Earth’s axis, oceans.
and their frictional drag at the Earth’s surface such Momentum transfer in the ocean is less well
as to maintain the near-constancy of the rate of known but again it is found that large-scale eddies
rotation of the Earth. These considerations limit the in the ocean are important in transferring angular
latitudinal extent of the Hadley circulation, from its momentum from strong surface currents into ocean
ascent at the ITCZ to latitudes between 30°N and depths. The distribution of surface currents,
30°S (Figure 2.9). Poleward of this tradewind belt thought to be mainly wind-driven, has been com-
the transfer of momentum, as well as that of heat piled from ships’ reports of their drift from their
and water, is brought about by smaller scale fea- calculated course. The general features are shown
tures such as depressions and anticyclones. in Figure 2.11: although the North Pacific, and the
Calculations of momentum transfer using upper North Atlantic are quite different in shape they
air observations are analogous to those of heat and have a rather similar current pattern, or general cir-
water vapour transfer: they involve the covariance culation. There is an anticlockwise circulation (or
[lVq] of the northerly and easterly components of gyre) in their northern parts and a huge clockwise
the wind velocity. They demonstrate that the large- one in the south. This is conspicuously asymmetric,
scale eddies bring about most of the required trans- the currents being much stronger in a narrow
ports of momentum, except at low latitudes, where region near the western boundary of the North
meridional cells play an important role, especially Pacific and the North Atlantic (the situation in the
in the vertical exchange of momentum. The middle Indian Ocean is complicated by the seasonal varia-
latitude westerlies and the low latitude North-East tion due to the monsoon). These strong boundary
and South-East Trades are a consequence of the currents (the Atlantic Gulf Stream and the Pacific
conservation of angular momentum (Figure 2.10): Kuroshio) are the best known currents of the
the westerlies have to be the stronger to keep the ocean.
rotation of the Earth constant. Near the equator in all three oceans there are
Although the atmosphere and the ocean have two west-flowing Equatorial Currents. The South
certain basic similarities – both are vast bodies of Equatorial Current lies at or south of the equator
fluid on a rotating Earth – their differences must and the North Equatorial Current to the north of
be recognised. They have marked differences in it. In the Pacific and Indian Oceans, and in part of
physical properties, especially those controlling the the Atlantic, the two west-flowing Equatorial

35
 H. Charnock

2.12 80˚ 80˚ 80˚ 80˚ 80˚ 80˚

60˚ 60˚ 60˚ 60˚ 60˚ 60˚

#
40˚ 40˚ 40˚ 40˚ 40˚ 40˚

20˚ 20˚ 20˚ 20˚

0˚ 0˚

20˚ 20˚ 20 20˚

40˚ 40˚ 40˚ 40˚ 40˚

60˚ 60˚ 60˚ 60˚ 60˚ 60˚

80˚ 80˚ Less than 0˚C 80˚ 80˚ 80˚ 80˚

#
0–1˚C
1–2˚C
Greater than 2˚C
Inferred water movements

Currents are separated by an Equatorial Counter- Figure 2.12 An impression of the flow pattern at 4000 m
current flowing toward the east. in the ocean. The major inputs are the North Atlantic
In the Southern Ocean, surrounding the Deep Water (NADW), which enters at the northern end
Antarctic, there is no continental barrier (though of the Western Basin of the Atlantic, and the Weddell Sea
Bottom Water, which enters from the margin of the
the relatively narrow Drake Passage may have a
Antarctic continent adjacent to the South Atlantic (based
similar effect) and the main surface current flows on Gross6 and Broecker and Peng3).
round the Earth as an east-going flow referred to as
the Circumpolar Current or as the West Wind
Drift. ocean. A rough impression of the currents at
It must be emphasised that the charts of ocean 4000 m depth is given in Figure 2.12.
currents are climatological; that is, they are based The estimation of meridional fluxes from trans-
on averages of observations made over a long time. oceanic hydrographic sections has provided infor-
On any particular occasion a ship may find a cur- mation on major differences between the mecha-
rent very different from the average current por- nisms in the North Atlantic and the North Pacific.
trayed on the Pilot Chart. This is especially notice- Study of the hydrographic section at 24°N in the
able in the region of a fast western boundary cur- North Atlantic shows that the heat flux is mainly
rent, such as the Gulf Stream, which meanders and due to a deep vertical-meridional cell. Warm and
changes the position of its axis in an unpredictable relatively saline water flows north near the surface,
way (see Figure 4.8). In such a case the climatologi- ultimately losing enough heat in winter to sink to
cal chart can be misleading, for the observations at great depth and return southward. The warm Gulf
a particular place are averaged over a long period, Stream water flows north near the surface but does
irrespective of whether the current is present or not return south at similar depth; only after a high-
not. It can be seen that a strong narrow current latitude cooling process does it return to the south
which varies in position is represented on a clima- as deep water. The smaller northward heat transfer
tological chart as a broader but slower current. In at 24°N in the Pacific is due to a nearly horizontal
this way, what may be called the ‘climatological circulation: relatively warm water flows north in
Gulf Stream’ (as represented on a time-averaged cli- the Kuroshio on the western side and in the near
matological chart) is perhaps ten times wider, and surface layer, loses heat in the subtropical and sub-
considerably weaker, than the Gulf Stream on any polar North Pacific and returns southward in the
particular occasion. central and eastern Pacific, at colder temperatures
Currents at greater depth are much less well but still at depths less than 800 m. Unlike the
known; the mean currents are small and are North Atlantic, the North Pacific has no source of
obscured by the variability of the large-scale eddies deep water and its deep circulation is correspond-
that have been found to be ubiquitous in the deep ingly slower. These ocean circulation differences

36
 2: The Atmosphere and the Ocean

80˚ 80˚ 80˚ 80˚ 80˚ 80˚ 2.13

60˚ 60˚ 60˚ 60˚ 60˚ 5 60˚
0 10
0 5
5 10 15
40˚ 40˚ 40˚ 40˚ 40˚ 40˚
15
15
20 20
20 20 20
20˚ 25 20˚ 20˚ 20˚
25 25 25
25
25
0˚ 28 28 0˚
28
28 > 28 25
20˚ 20˚ 20 20˚
25 25
20 20 20
20
15 15
40˚ 10 40˚ 40˚ 40˚ 10 40˚
15
5 5
10
0 5 0
60˚ -1 60˚ 60˚ 0 60˚ 60˚ 60˚
-1 -1
80˚ 80˚ 80˚ 80˚ 80˚ 80˚

Temperature less than 10˚C

Temperature 10–25˚C
Temperature 25–28˚C
Temperature greater than 28˚C

Figure 2.13 Average temperature at the sea surface in lent boundary layers which exist near the Earth’s
February (from Gross 6; the values are derived from surface and at the sea floor. Most important is the
Sverdrup et al.20 and from later sources). coupled boundary layer of the atmosphere and
ocean which occupies a layer typically 1 km in
are consistent with the marked climatic differences height above and 100 m in depth below the sea sur-
in the climate of the North Atlantic and the North face. This is a region in which many energy
Pacific – the Atlantic is much warmer, especially at exchanges and transformations take place, processes
subpolar and polar latitudes (Figure 2.13). which determine the properties of the ocean and
The climate of the ocean–atmosphere system many of those in the atmosphere. Two processes of
depends on complicated interactions between the the air–sea boundary layer are of particular impor-
meridional fluxes, brought about by motions of rel- tance: the production of vertical velocities and the
atively large scale, and the near-surface vertical transfer of boundary layer air to the less turbulent
fluxes, brought about by motions on a much small- free atmosphere above, and of boundary water to
er scale in the complicated turbulent interacting the less turbulent deep ocean below.
air–sea boundary layer. Processes that determine The frictional stress of the wind on the sea sur-
the properties of the ocean take place at its surface: face, on a rotating Earth, drives a mass transport to
the resulting horizontal transfers within it are such the right (left) of stress direction in the northern
as to maintain the surface sources and sinks. How (southern) hemisphere. If the stress varies from
this is done is the central problem for observational place to place it produces convergences or diver-
and theoretical oceanographers alike. It is the tar- gences that lead to vertical motion (Figure 2.14); a
get for those making computer models of the ocean similar effect happens in the atmosphere. The
and especially for those attempting to model the resulting vertical motions are fundamental to the
coupled ocean and atmosphere: they have great dif- general circulation of the atmosphere and ocean
ficulty in matching the horizontal motions in the through their effect on the vorticity balance (see
two media to the flux of heat and water between Box 2.1). In the atmosphere they lead to the forma-
them. tion of cloud and rain, complications that are not
present in the ocean. In the ocean, however, verti-
The Coupled Atmosphere–Ocean cal motions (usually smaller than those due to wind
Boundary Layer stress) are also produced by the difference between
Evaporation and Precipitation (E–P). Evaporation
Most of the atmosphere and most of the ocean, and precipitation are also fundamental to the near-
most of the time, can be treated as frictionless and surface energy exchanges and so to establishing the
adiabatic. But in some places there are vital process- properties of the lower atmosphere and the upper
es that are more complicated: clouds, fronts in the ocean. These properties are communicated, by
atmosphere and the ocean, and especially the turbu- complicated and little understood processes, to the

37
 H. Charnock

Box 2.1. The Vorticity Balance of the Ocean

All the large-scale flow in the atmosphere and ocean is affected by the rotation of the Earth. The rela-
tively small-scale flows encountered in bathroom, kitchen, or laboratory are dominated by other
forces, so the effect of the rotation of the Earth is beyond our usual experience. The flow patterns it
produces in the atmosphere and ocean sometimes seem bizarre.
The basic notion is one of the vorticity, or spin, which anything on a rotating globe must have. If
one imagines a man standing astride the North Pole, for example, it is obvious that he will be rotating,
or spinning, about his own axis, at the same rate as the Earth – once per day. If the same man now
stands astride the equator, the Earth continues to rotate about its axis but he no longer rotates about
his. His local rate of rotation or spin is zero. So the spin which affects anything on the Earth is zero at
the equator and increases to one revolution per day at the poles.
In these examples, our hypothetical man – we could equally have considered a parcel of fluid – was
at rest relative to the Earth. He – or the fluid – could also have been rotating on his own axis relative to
the Earth. The total spin is obviously made up of two components – the spin relative to the Earth and
that due to the rotation of the Earth beneath it. It is a difficult but important concept – important
because, in relatively shallow fluids like the atmosphere and the ocean, a body of water moves in such
a way that its total spin, its vorticity, stays constant.
The vorticity relative to the Earth of a column of fluid increases if it is stretched: the stretching
decreases the diameter and the rotation increases to conserve the column’s angular momentum.
Conversely shrinking the column increases the diameter and the vorticity decreases. The spinning of
an ice skater provides a familiar example. To maintain its total vorticity constant, on the rotating Earth,
a shrinking column must move equatorward and a stretched column poleward.
In the upper ocean shrinking and stretching are brought about by a vertical gradient of vertical
velocity, produced near the surface by wind-stress convergence (Figure 2.14), and by E–P. Over much
of the North Atlantic the wind distribution is such as to produce shrinking, so motion toward the equa-
tor. The E–P distribution is such as to produce stretching, so poleward motion of smaller magnitude.
The resulting southward transport leads to the westward intensification of wind-driven currents, the
necessary poleward return flow being accomplished in narrow western boundary regions – like the
Gulf Stream – whose dynamics are more complicated.
In the deep ocean there is a very slow upward velocity to compensate for the sinking of deep water
at high latitudes. This stretches the water column, and would be expected to produce generally north-
ward transport with equatorial return flow being confined to a narrow region on the western bound-
ary. Such western boundary currents are observed but the northward transport is obscured by the
large-scale eddies and by the effects due to sea floor topography.

2.14 Wind air above and to the water below the coupled
stress
boundary layer: they determine the properties of
Near-surface the atmospheric airmasses and the oceanic water-
&

mass transport Balancing

Coriolis Force masses.
CONVERGENCE Unfortunately knowledge and understanding of
(downward motion) the complicated boundary layer processes is very
& poor. There is some empirical information about
Balancing Near-surface
Coriolis Force mass transport the marine atmospheric boundary layer, especially
DIVERGENCE its lowest 100 m or so (which allows the climato-
Wind (upward motion) Wind
stress stress logical estimates of heat transfer and evaporation)
but uncertainty remains about the interaction of
&

wind and waves. Processes near the top of the

Near-surface
mass transport
marine atmospheric boundary layer, where there is
frequently cloud, are much less understood. In the
upper layers of the ocean surface waves make
Balancing
Coriolis Force observations difficult: there is a need for more
information on the effect of spray and of bubbles,
Figure 2.14 The balance of forces in the upper 50 m of as well as on the living and non-living particulate
the ocean produces a depth mean transport at right angles matter that determines the transparency and the
to the wind stress (to the right in the northern hemi- absorption of solar radiation. The transfer proper-
sphere). Variations of wind stress from place to place pro- ties of the helical (Langmuir) circulation (Figure
duce convergence or divergence in the surface layers and 2.15) are still uncertain. In both atmosphere and
corresponding vertical velocities below the surface.

38
 2: The Atmosphere and the Ocean

Wind
2.15

Surface streak
k k
k k k Typically
k k k 5m
k k k
Typically
(20–30 m)
Figure 2.15 Helical (Langmuir)
circulations in the upper layers
of the ocean producing wind-
rows of floating material on the
surface.

ocean we need a reliable way of predicting the ver- provides an input (a haline buoyancy flux) to the
tical extent of the boundary layer and a better surface buoyancy flux which is sometimes compa-
understanding of how fluid is transferred from the rable to that of the thermal buoyancy flux. It has
boundary layer to the free atmosphere and the deep been suggested that climatic freshening of high-lati-
ocean. The problems are technically demanding, tude surface water could stop the formation of
and the theory underlying them notoriously diffi- deep water in the North Atlantic: the overturning
cult, but an improved knowledge of the coupled meridional cell is thought to be very sensitive to the
air–sea boundary layer is vital to our need to freshwater flux. Recent estimates show that the
understand and to model the atmosphere and the thermal buoyancy flux dominates the haline buoy-
ocean. ancy flux at high latitudes, suggesting that large
changes in (E–P) would be needed to bring about
Climate Studies what has been called the ‘haline catastrophe’.
Numerical modelling of the atmosphere is already However the effect of continental run-off, and
well advanced, both for weather forecasting and especially of the freezing and melting of ice (Figure
for simulating climate and climatic change. Ocean 2.16) could be significant, especially in the areas
modelling is advancing rapidly as increasing com-
puter power allows finer resolution to represent
ocean eddies, which are smaller than atmospheric 2.16
disturbances (see Chapter 4). The discrepancies
between meridional flux estimates using different
methods should soon be clarified: the ERBE project
continues, more trans-oceanic hydrographic sec-
tions are being made, and there is the prospect of
improved atmospheric flux estimates. The results
will provide valuable constraints and tests of cli-
Average Average extent
mate models simulating a coupled atmosphere and extent of
ocean. of permanent ice
Near-surface meteorological observations are winter ice
vital as input to weather forecast models and for the
verification of climate models. Their calcula-
tion relies on realistic simulation of the
atmosphere–ocean boundary layer, but rapid
progress in our knowledge of this complicated region is
not to be expected. There is also a great need for better
observation (preferably from space) and improved
simulation of precipitation.
The difference between evaporation and precipi-
tation (E–P) is important as providing a vertical Figure 2.16 Chart to indicate the extent of permanent
velocity and affecting the vorticity balance, and it and winter ice in the Arctic (from Oceanus, 37, 2, 1994).

39
H. Charnock

where deep water formation now occurs. There is 8. Jacobwitz, H., Smith, W.L., Howell, H.B., and Hagle, F.W.
increasing evidence that both (E–P) and thermal (1979), The first 18 months of planetary radiation
budget measurement from the Nimbus-6 ERB experi-
changes are closely connected to changes in ocean ment, J. Atmos. Sci., 36, 501–507.
circulation, and so to climate on decadal and 9. Masuda, K. (1988), Meridional heat transport by the atmos-
longer time-scales. The problem requires an phere and the ocean: analysis of FGGE data, Tellus,
improved understanding of how the motions of the 40A, 285–302.
10. Michaud, R. and Derome, J. (1991), On the mean meridion-
atmosphere and the ocean maintain the sources and al transport of energy in the atmosphere and oceans
sinks of heat and water, and of the coupling as derived from six years of ECMWF analyses, Tellus,
between the large scale horizontal winds and cur- 43A, 1–14.
rents and the small scale turbulent transfer process- 11. Nieburger, M., Edinger, J.G., and Bonner, W.D. (1982),
es of the coupled air–sea boundary layer. Understanding our Atmospheric Environment, W.H.
Freeman and Company, San Francisco.
12. NRDC (1946), Summary Technical Report, Division 6,
Office of Naval Research, Washington DC.
13. Oort, A.H. and Peixoto, J.P. (1983), Global angular
General References momentum and energy balance requirement from
observations, Advances in Geophysics, 25, 355–490.
Gill, A.E. (1982), Atmosphere–Ocean Dynamics, Academic
14. Peixoto, J.P. and Oort, A.H. (1983), The atmospheric
Press Inc., Orlando.
branch of the hydrological cycle and climate, in
Gross, M. Grant (1992), Oceanography, a View of Earth,
Variations in the Global Water Budget, Street Perrott,
Simon and Schuster, Englewood Cliffs, New Jersey.
A. (ed.), pp 5–65, Reidel, Dordrecht.
Peixoto, J.P. and Oort, A.H. (1992), Physics of Climate,
15. Schmitt, R.W. (1994), The Ocean Freshwater Cycle, JSC
American Institute of Physics (AID Press), Woodbury,
Ocean Observing System Development Panel, Texas
New York.
A&M University, College Station, Texas.
16. Schmitt, R.W., Bogden, P.S., and Dorman, C.E. (1989),
Evaporation minus precipitation and density fluxes
References for the North Atlantic, J. Phys. Oceanogr., 19,
1. Barkstrom, B.R., Harrison, E.F., and Lee, R.B. (1990), Earth 1208–1221.
Radiation Budget Experiment: preliminary seasonal 17. Schmitt, R.W. and Wijffels, S.E. (1993), The role of the
results, EOS, 71, 297–305. ocean in the global water cycle, in Interactions
2. Baumgartner, A. and Reichel, E. (1975), The World Water Between Global Climate Sub-systems: the Legacy of
Balance, Elsevier, New York. Hann, McBean, G.A. and Hantel, M. (eds),
3. Broecker, W.S. and Peng, T.-H. (1982), Tracers in the Sea, Geophysical Monographs, 75, 77–84.
Lamont–Doherty Geological Observatory, Columbia 18. Stephens, G.L., Campbell, G.C., and Vonder Haar, T.H.
University, New York. (1981), Earth radiation budgets, J. Geophys. Res., 86,
4. Bryden, H. (1993), Ocean heat transport across 24°N lati- 9739–9760.
tude, in Interactions Between Global Climate Sub- 19. Sutcliffe, R.C. (1966), Weather and Climate, Wiedenfield
systems: the Legacy of Hann, McBean, G.A. and and Nicolson, London.
Hantel, M. (eds), Geophysical Monographs, 75, 20. Sverdrup, H.U., Johnson, M.W., and Fleming, R.H. (1942),
65–75. The Oceans, their Physics, Chemistry and Biology,
5. Ellis, J. and Vonder Haar, T.H. (1976), Zonal Average Earth Prentice Hall, New York.
Radiation Budget Measurement from Satellites, 21. Vonder Haar, T.H. and Ellis, J. (1974), Atlas of Radiation
Atmos. Sci. Papers 240, Colorado State University, Budget Measurements from Satellites, Atmos. Sci.
Fort Collins, Colorado. Papers 231, Colorado State University, Fort Collins,
6. Gross, M. Grant (1992), Oceanography, a View of Earth, Colorado.
Simon and Schuster, Englewood Cliffs, New Jersey. 22. Wijffels, S.E., Schmitt, R.W., Bryden, H.L., and Stigebrandt,
7. Gruber, A. (1978), Determination of the Earth–Atmosphere A. (1992), Transport of fresh water by the oceans, J.
Radiation Budget from NOAA Satellite Data, NOAA Phys. Oceanogr., 22, 155–162.
Tech. Rep. NESS 76, Washington DC.

40
CHAPTER 3:

The Role of Ocean Circulation

in the Changing Climate
N.C. Wells, W.J. Gould, and A.E.S. Kemp

Overview of Climate System influences on climate, one of our great challenges in

The relatively warm and stable climate which we the closing decade of the twentieth century is to
have had for the past 10,000 years, since the end of measure and understand natural climate variability
the most recent glaciation, has been essential for across the globe, together with the role played by
the evolution of cultivated crops, which led to the the oceans.
development of human settlements rather than to a Given the limitations of the observing network,
nomadic existence, and hence to the development where does our scientific evidence for climate
of civilisation. change over longer periods come from?
Recent research has shown that changes in Indirect evidence in ice cores from Antarctica and
ocean circulation played a key role in controlling Greenland, and from sediment cores studied during
climate change and regulating the glacial–inter- the CLIMAP (Climate: Long Range Interpretation,
glacial cycles that have been a hallmark of the Mapping, and Prediction) initiative and, more
northern hemisphere climate for the past 2 million recently, from the Ocean Drilling Programme, have
years. Understanding the nature of this link provided reliable indicators of past climate varia-
between ocean circulation and climate change is tions. The measurement of the air trapped in the
now a key goal of research in this area. In particu- Vostok ice core, from Antarctica, has revealed the
lar, the identification of periods of very rapid cli- levels of carbon dioxide from the most recent inter-
mate change in the recent past has given a new glacial period (circa 125,000 years BP) to the pre-
urgency to these studies. sent day16, and thus has given a benchmark from
The measurement of climate change by scientific which the more recent anthropogenic contribution
instruments dates back to the invention of the ther- can be determined. Measurements of oxygen iso-
mometer by Galileo in the sixteenth century. It tope ratios, contained in bubbles trapped in the ice,
was, however, not until considerably later that sys- are now known to be a proxy for air temperatures
tematic methods were applied to the measurement (see Chapter 8, Box 8.3). Recent analyses of ice
of temperature and rainfall, which allowed the con- cores from Greenland have thus provided time
cept of climatology to develop. These measure- series of temperature from the most recent inter-
ments have been and are unevenly distributed over glacial to the present day. This has shown that the
the globe, with the majority covering the landmass- climate over the last 5,000 years has been remark-
es and, particularly, the well-populated areas of the able for its stability. The ice cores also provide evi-
northern hemisphere. The oceans, covering nearly dence for rapid changes in climate during the previ-
71% of the Earth’s surface, have not been neglect- ous interglacial, from temperatures similar to those
ed; indeed, one of the first responsibilities of the of today to glacial conditions on the remarkably
UK Meteorological Office, when it was formed in short time-scales of decades to centuries (see later).
1854, was to implement the systematic recording of Evidence of longer-term changes, those over mil-
surface observations from commercial ships plying lions of years, are found in marine sediments (see
the trade routes of the world. Despite such efforts, Chapter 8). The glacial and interglacial cycles that
measurements in the southern hemisphere, particu- occurred during the last 2 million years have been
larly over the Southern and Indian Oceans, remain linked to variations in solar radiation, associated
rather sparse. We shall see demonstrated that these with variations of the orbital parameters of the
poorly observed oceans are a key element in the Earth around the Sun, known as Milankovitch
Earth’s climate. cycles. Longer-term changes may be associated
Recent interpretation of ice cores from with the different configurations of the oceans and
Greenland15 has shown that the climate system is continents, due to continental drift, and with varia-
not as stable as was once thought and may undergo tions in continental uplift and mountain building.
extremely rapid changes (e.g., 5–7°C in a decade). In addition to these interpretations of past cli-
As we concern ourselves with possible man-made mate we now have new powerful methods which

41
 N.C. Wells, W.J. Gould, and A.E.S. Kemp

3.1 ATMOSPHERE
Ocean–atmosphere coupling
Terrestrial
radiation
Biomass
Atmosphere–ice coupling Evaporation
Solar Precipitation
radiation
Ice-sheets Snow Wind stress
Sea–ice
Heat
exchange

Heat storage

Deforestation Atmospheric Plankton

pollutants Ice–ocean OCEAN
coupling
Ocean circulation

EARTH

Figure 3.1 A schematic view of the Earth’s climate system showing the roles of land, atmosphere, oceans, and sea ice.

are being applied to understanding how the Earth’s abundant observations are available for the atmos-
climate works. These methods are based on an phere, comprehensive, synoptic ocean measure-
understanding of the different ‘components’ of the ments are few and far between. The World Ocean
climate system (air, water, and ice), and the interac- Circulation Experiment (WOCE; see later) will pro-
tions between them (Figure 3.1). vide critical ocean measurements, such as heat and
Each component responds to change over differ- fresh water transports, on a global scale.
ent periods of time. The atmosphere takes the Improvements in our ability to model the various
shortest time (of the order of a week to a month) to components of the climate system will allow the
communicate changes throughout its mass, because development of more comprehensive climate models
winds are very much faster than ocean currents and in which the components are coupled together.
mixing through turbulence is efficient. The domi- These models will aid our interpretation and under-
nant time-scale of the ocean is of the order of standing of the complex climate system, as well as
weeks to seasons in the surface layers and decades providing predictions of socioeconomic importance.
to centuries at abyssal depths. The ice caps of
Antarctica and Greenland are excellent measures of The Role of the Ocean in Climate
longer-term change as they grow slowly on time- The ocean and atmosphere together transfer heat
scales of centuries to millennia. from the tropics to the polar regions, at a rate of the
Mathematical computer models of each one of order of 5 PW at 30°N, (equivalent to the output of
the components, bounded, constrained, or driven by 5 million large power stations) in order to balance
interactions with others, allow a deeper understand- the deficit of incoming radiation in the mid-latitudes
ing of both the components themselves and their and polar regions with the excess of radiation in the
interactions with one another. For example, mathe- tropics. The partition of this transfer between
matical models of the ocean (see Boxes 3.1 and 3.2) atmosphere and ocean is not well-determined, but
are based on the dynamical laws which govern the recent ocean measurements at 24°N have shown
behaviour of the ocean. These ocean models can transports within the ocean of the order of 2 PW; a
reproduce the large-scale, wind-driven circulations value which is comparable with the atmospheric
of the ocean basin gyres, and the deep vertical trans- flux at this latitude (Figure 3.2)7; see Chapter 2 for
ports, or overturning, of the thermohaline circula- further discussion.
tion of the oceans driven by cooling at polar lati- The oceanic component of the heat flux is pro-
tudes. The advent of faster, more powerful comput- vided by ocean currents, which have much longer
ers has allowed models to reproduce some of the time-scales (of the order of 10 years for the wind-
smaller (ca 100 km), energetic ocean eddies (see driven subtropical gyres and of many decades to
Chapter 4), which are important for both the trans- centuries for the vertical overturning thermohaline
port of heat and fresh water in some regions of the circulation) than the atmosphere. When one con-
world’s oceans and for the maintenance of the large siders that a 2.5 m deep layer of sea water covering
ocean circulations. Observations are always needed the globe has the same thermal capacity as the
to initialise and test the models, but while relatively entire atmosphere, it can be appreciated that the

42
 3: The Role of Ocean Circulation in the Changing Climate

Figure 3.2 Poleward transfer of heat by: (a) ocean and (a) 6
atmosphere together (TA+TO), (b) atmosphere alone (TA), 3.2
+ +
and (c) ocean alone (TO). The total heat transfer (a) is 4 + +
derived from satellite measurements at the top of the +

TA + TO (PW), annual
2 +
atmosphere, that of the atmosphere alone (b) is obtained +
from measurements of the atmosphere, and (c) is calculat- 0
+
ed as the difference between (a) and (b) (1 PW = 1015 W). ✕ +
✕
(Based on Carrissimo et al.7; results from other investiga- -2 + Oort and Vonder Haar18
✕ ✕ Trenberth27
tions are added for comparison.) ✕
-4 Newell et al.17
✕ ✕
oceans make a very important contribution to the ✕ ✕
stabilisation of our climate system. -6
4
While, overall, the oceans transfer heat pole- (b)
+ + + +
ward, they also exchange heat between ocean basins

TA (PW), annual
2
+ + +
in a more complex fashion. For instance, the North + +
0
Atlantic Ocean loses more heat to the atmosphere ✕
✕
than it gains from incoming radiation, so there has ✕
-2 ✕ ✕
to be a net heat transfer from the Pacific and Indian ✕ ✕
✕
-4
Oceans into the South Atlantic Ocean and thence to 4
the North Atlantic to compensate for this deficit. (c)
+
The thermohaline circulation, driven by the produc-

TO (PW), annual
2 + +
+ +
tion of denser water at polar latitudes, is a mecha-
0 ✕ ✕ + + + +
nism by which heat is transferred within and ✕ ✕
between ocean basins. The oceans are interconnect- ✕
-2 ✕
✕
ed by what has come to be known as The Global ✕

Thermohaline Conveyor Belt (Figure 3.3)13. -4

80˚S 60˚ 40˚ 20˚ 0˚ 20˚ 40˚ 60˚ 80˚N

(a) 3.3

Figure 3.3 Schematic views

of two versions of the ‘con-
veyor’ circulation of the Warm
oceans. Warm water associ-
ated with the surface and
intermediate waters of the
oceans (upper 1000 m) fol-
lows a pathway toward the Cold
northern North Atlantic
Ocean, where it is subjected
to intense winter cooling.
This leads to the formation
of cold North Atlantic deep
water, which spreads south- (b)
ward into the Southern
Ocean and returns to the
Pacific Ocean. The convey-
or is responsible for a north-
ward transfer of heat
throughout the whole of the
Atlantic Ocean. (a) The Warm
upper water moving east-
ward from the Pacific into
the Atlantic. (b) The upper
warm water moving through
the Indonesian Archipelago Cold
into the Indian Ocean and
thence into the Atlantic
Ocean.

43
N.C. Wells, W.J. Gould, and A.E.S. Kemp

3.4

60˚N
0.95
0.77

0.43 0.88
30˚
Fp
Fp + 0.02 0.69
0˚
0.06
Fp – 0.51 Fp + 0.30 0.13
30˚

Fp + 0.24

60˚S FA+0.17 FA+Fp – 0.13 FA

0˚ 90˚E 180˚ 90˚W 0˚

Figure 3.4 An estimate of the transfer of fresh water (x 109 kg/s) in the world oceans. In general, in polar and
equatorial regions precipitation and river run-off exceed evaporation and hence there is an excess of fresh water, while
in the subtropical regions there is a water deficit. A transfer of fresh water is required from the regions of surplus to the
regions of deficit. For example, in the North Atlantic, there is southward flow of 950,000 t/s (t = metric tonne) of fresh
water at 60°N, while at the equator the southward flow is 690,000 t/s – hence 26,000 tonnes of fresh water per second
are evaporated in the North Atlantic. It can be seen that fresh water is exported from the North Pacific to the North
Atlantic, through the Arctic Ocean. FP and FA refer to the fresh water fluxes of the Pacific–Indian throughflow and of
the Antarctic Circumpolar Current in the Drake Passage, respectively.

The ocean also transports fresh water around the increase in salinity will increase the density. The
globe, another key element of the global conveyor low salinity layer at the surface of the North Pacific
(Figure 3.4). The total mass of salts in the ocean maintains a stable stratification (low density at the
remains unchanged on time-scales shorter than geo- surface), which cannot be destabilised (i.e., made to
logical time. In contrast, the fresh water content of have a higher density at the surface than deeper in
the oceans changes in response to precipitation, the water column) by surface cooling in the present
evaporation, freezing and melting of ice, and run-off climate state. Hence, there is no significant deep
from the land. All these factors influence the dilu- convection in the North Pacific and, consequently,
tion or concentration of ocean salt, and hence the little formation of cold deep water. By contrast, the
salinity. Differences in the input of fresh water into northern North Atlantic, with a higher surface
the ocean, from one region to another, have to be salinity, is destabilised by winter cooling, and pro-
balanced by a horizontal transport of fresh water by duces deep cold watermasses by vertical convec-
ocean currents and by sea ice in polar regions. tion. Polar watermasses, formed in the Green-
Generally, in the subtropics there is a deficit of fresh land/Norwegian sea and, to a lesser extent, in the
water which is reflected in higher surface salinity, Arctic Ocean, also enter the North Atlantic and
while in the higher latitudes there is an excess of provide additional cold deep water. The transport
fresh water and a lower surface salinity. The North of ice from the Arctic Ocean provides an additional
Pacific Ocean is less saline than the North Atlantic, source of fresh water to the North Atlantic, which
because of lower evaporation, and therefore is a through its variability can modulate the surface
source of fresh water. This water, in turn, is trans- salinity on decadal time-scales28.
ported into the North Atlantic by the conveyor cir-
culation, to make up for the deficit there. Evidence of Climate Change in the Ocean
The density of sea water depends on tempera- Despite its high thermal capacity, the ocean responds
ture and salinity. Heating or a decrease in salinity to exchanges of heat, fresh water, and momentum
will decrease the density, while cooling or an with the atmosphere on a range of time-scales from

44
 3: The Role of Ocean Circulation in the Changing Climate

3.5 Year Figure 3.5 Long-term changes in annual mean sea level
1820 1840 1860 1880 1900 1920 1940 1960 1980 at selected ports. The general upward trend is seen at all
but one station. The downward trend in sea level at Sitka,
Alaska, is due to the vertical movements of the land at
which sea level is measured (see text). The global mean
Sitka sea level rise is estimated to be about 1–3 mm/year.

Newlyn
Measurements of temperature and salinity with
Brest
useful accuracy have only been available for the
past 50 years or so, and there are few places where
Halifax NS
high-quality measurements have been made over
several decades. This severely limits our ability to
directly measure climate-scale change in the ocean.
50 mm
At the surface of the ocean there is a great deal of
San Francisco
temperature variability caused by daily and seasonal
Honolulu
heating and cooling, so here salinity is a better indi-
cator of decadal change. Salinity records made in the
area west of Scotland show a remarkable decrease
Bombay
that lasted from 1973–1979, with the lowest values
in 1975. The lowest value deviated from the average
Sidney by almost four times the typical variability.
Subsequent analysis of other salinity data from
1820 1840 1860 1880 1900 1920 1940 1960 1980
around the Atlantic showed that this salinity
Year
change, now referred to as the ‘Great Salinity
Anomaly’, was not confined to Scottish waters, but
a day and shorter right up to geological scales was a phenomenon that took over 10 years to
(glacial and interglacial periods of 10,000–100,000 propagate around the North Atlantic (Figure
years; see Chapter 8). The thermal inertia of the 3.6)10,12. There is still much conjecture about the
ocean, however, means that oceanic changes are of cause of the ‘Great Salinity Anomaly’. Recent evi-
much smaller amplitude than those seen in the dence has suggested that in the 1960s an unusually
atmosphere, but nevertheless they are easily measur- large export of ice into the North Atlantic from the
able on daily to seasonal scales and, with care, on Arctic Ocean may have caused the low salinity
decadal scales. The state of the ocean at any time is
predominantly a result of exchanges with the atmos-
phere over the previous 100 years or so. 70˚ 3.6
1979
By far the longest time series of ‘oceanographic’ 1978–1979
measurement is that for sea level (Figure 3.5).
Measurements made for the prediction of tides go 1981–1982
back over 100 years and, after the tidal signal is
removed, show long-term trends, rising in some 1969–1970
1968 1977–
1978
places, in others falling. It has to be remembered that 60˚
sea level is measured relative to a fixed point on land,
1976
so the changes seen are a summation of changes in 1976
1977–
1978
true sea level (due primarily to thermal expansion of
1979
the water, and melting of glaciers and ice caps) and 1975
the not insignificant vertical movement of the land 1972
50˚N
(isostasy). This movement can be caused by
1971–1972 1974?
‘rebound’ after release from the covering of ice dur- 1977

ing the most recent glacial period or by tectonic activ-

ity. Unravelling the changes due to each individual
factor is difficult21 and yet the prediction of sea level 40˚
change is of immediate and practical importance to
1977
low-lying areas and, in particular, to the inhabited
coral atolls of the Pacific and Indian Oceans.
30˚
Figure 3.6 Timing of the propagation of the ‘Great
Salinity Anomaly’ around the North Atlantic.
50˚ 40˚ 30˚W 20˚ 10˚

45
 N.C. Wells, W.J. Gould, and A.E.S. Kemp

3.7 (a) 0
-1.0 0.0 anomaly, which suppressed deep convection for

-1.0
0.5 -0.-1.0
more than a decade1.

-0.5
5
0.0

0.0.5
500 0.0

0
-1.
-0.
5
0
0.5
1.0 0.0
Whatever the cause, the significance of such
1,000
changes cannot be over-emphasised. The global ther-
Pressure (dbar)

0.0
1,500 mohaline conveyor belt is driven by deep convection
in high latitudes and salinity is, in many cases, the
0.0

2,000 Rockall
controlling factor that decides whether such deep con-

0.0
Plateau
0.0
2,500 vection is possible. In simplified terms, fresher water
Mid-Atlantic at the surface results in the formation of an insulating
3,000 Ridge
layer of ice in winter, rather than plumes of dense cold
30˚
saline water. So it is possible that extreme salinity
40˚ 35˚ 25˚ 20˚ 15˚ 10˚W
anomalies could be responsible for the transitions
between glacial and interglacial episodes. We give an
0
(b)
-0
2.0
1.0 example of this in the next section.
-2.0 -

.5
-0.

In the interior of the ocean, patterns of tempera-

1.0

500 0.5
5

1.0
0.0

1.0

0.0 ture and salinity change have been determined in

2.0

-2.0

1,000 0.5
three ways, from time series at fixed stations, by
Pressure (dbar)

-1 0.5
.0
-

1,500 the subtraction of averages of values accumulated,

0.
0

2,000 0.0 0.0

say, over one decade from those in another decade,
and from the comparison of repeated sections (lines
2,500
Rockall of measurements across an ocean). Each of these
Mid-
3,000 Atlantic
Plateau techniques has its drawbacks. There are very few
Ridge
time series stations (Bermuda, Hawaii, and off the
35˚ 30˚ 25˚ 20˚ 15˚W West Coast of Canada), the decadal averages are
subject to errors of non-uniform spatial sampling,
and the sections are instantaneous pictures that are
Figure 3.7 Cooling (blue areas, °C) along the section of contaminated by the presence of energetic transient
the subpolar North Atlantic between northwest Europe eddies (see Chapter 4).
and Greenland: (a) 1981–1991, (b) 1962–1991. Despite these limitations, results from each of
these lead to the conclusion that temperatures in
0 the ocean change by a few tenths of a degree over a
3.8
typical 10-year time-scale. The changes are largest
in the upper part of the water column and decrease
with depth (see Figures 3.7 and 3.8). The horizon-
1000
tal areas of such changes are large, comparable to
the width of an ocean basin19,23.
2000 Thermohaline Catastrophes
and the Younger Dryas
Pressure (dbar)

During the most recent ice age, the North Atlantic

3000 component of the Global Thermohaline Conveyer
was partially shut down and the ocean is thought
to have operated in a different mode to that of the
4000 present day. The northern North Atlantic was con-
siderably cooler and the transport of the North
Atlantic current (the northward extension of the
1992–1957
Gulf Stream) much reduced (Figure 3.9). The
5000
1992–1981 North Atlantic component of the conveyer was
1981–1957 reactivated at the end of the most recent glaciation,
at about 14,000 years BP.
6000 Recent research on deep-sea sediment cores has
–0.4 –0.3 –0.2 –0.1 0.0 0.1 0.2 0.3 0.4 shown that this reactivation of the conveyer was
Zonally averaged temperature difference (˚C) not without hiccups! Part of this conveyer stopped
abruptly at about 11,000 BP – a period known as
Figure 3.8 Temperature changes (°C) in the subtropical the Younger Dryas29. This led to a catastrophic
North Atlantic (24°N), 1957–1992. The measurements cooling of the North Atlantic region and caused the
have been averaged across 24°N between North Africa build up of small glaciers in the British mountains
and Florida. in what geographers call the Loch Lomond glacial

46
 3: The Role of Ocean Circulation in the Changing Climate

3.9 Increased heat Figure 3.9 (a) The ‘present-day’ North Atlantic with a
transport
(a) Polar High heat loss Heat gain Tropical vigorous conveyor (thermohaline) circulation, and a large
region region northward transfer of heat. (b) The ‘ice-age’ North

;
; Warm, saline Atlantic with a weak conveyor (thermohaline) circula-
tion, and a reduced northward heat transfer. The northern
North Atlantic was therefore cooler and fresher than the
Thermocline
‘present-day’ ocean. The extended sea-ice cover is asso-
ciated with a fresher surface layer, which stabilises the
Cold, less saline
water column and inhibits deep convection.

on the thermohaline system is an example of posi-

tive feedback. The fresh water input weakens the
Reduced heat
transport
thermohaline circulation, which makes the circula-
(b) Polar Heat gain Tropical
region Low heat loss region tion more susceptible to further weakening. This
;

process has been investigated in recent years using

;

Fresh, Ice
cold Warm, saline
mathematical models of ocean circulation. The
Thermocline
models show that there are a number of different
states of the thermohaline circulation, some of
which are stable (Figure 3.10). There are, however,
Cold
transitions between stable states, which occur over
periods as short as 40 years9. It has been speculated
that the present North Atlantic Ocean may be close
to one of these transitional states, of which the
Younger Dryas is an example. One study has sug-
re-advance. This cooling only appears to have last- gested that the transition between states is not nec-
ed a few centuries, but it developed very rapidly essarily symmetrical. The change from strong to
over decades. There are several theories about what weak thermohaline circulation may be more rapid
exactly led to the shut down of the conveyer. One (40 years) than the re-establishment of the strong
view is that a sudden influx of fresh melt-water circulation (500 years). Further study of the
from the Laurentide ice sheet into the North Younger Dryas and similar events in the palaeocli-
Atlantic could have stabilised the vertical stratifica- mate record may give us important clues to the
tion and reduced the rate of formation of North
Atlantic deep cold water. This, in turn, could have 20
shut down the North Atlantic conveyor circulation, (a) ON 3.10
resulting in a cooling of the surface waters of the
Conveyor belt

northern North Atlantic4.

x 106 m3/s

The effect of changes in fresh water fluctuations 1 2 3

Figure 3.10 A theoretical model for oscillations of the 0

OFF
ocean conveyor belt, continental ice volume, surface
water flux, and Atlantic Ocean salinity as a function of (b)
volume at last glacial maximum

time (1 unit = 1000 years). The oscillator model is based

Fraction of continental ice-

0.6
on bi-stable states for the conveyor circulation [(see 20
Figure 3.9(a)]. When the conveyor circulation is turned
off, northward heat transfer is reduced and cools (a), 0.4

which results in a growth in ice volume (b). The reduced 10

export of salt from the Atlantic, because of the ‘turning 0.2
off’ of the conveyor circulation, causes an increase of
salinity (c). The salinity increases to a threshold value, at 0.0 0
which point the deep overturning of the northern ocean 1 2 3

occurs, and the conveyor circulation is ‘turned on’ again. 35.4

The continental ice volume (b) decreases, because of the (c)
increase in melting associated with the increased north-
Atlantic salinity

ward heat flux into the North Atlantic. With more melting 35.2
(PSU)

and a greater export of salt (due to the ‘turning on’ of the

thermohaline circulation), the Atlantic salinity decreases. 35.0
1 2 3

These oscillations have a period of the order of 1000

years – the thermohaline circulation time-scale. (Based
on Birchfield and Broecker3.) 34.8

47
N.C. Wells, W.J. Gould, and A.E.S. Kemp

likely response of the present-day thermohaline cir- (a) 12.85 3.11

culation in the North Atlantic to global warming
(remaining stable or ‘flipping’ to another state).

WMDW potential temperature (˚C)

12.80
Monitoring Climate Change
So we know there are changes in the ocean that
occur over periods of tens of years. Similarly, we 12.75

suspect that effects of this type may be implicated

in longer-period and more extreme (glacial and
12.70
interglacial) changes. We have now to ask the ques-
tion whether anthropogenic activities can be detect-
ed in the ocean and whether new technologies can 12.65
help us to observe the ocean better.
There is at least one case in which repeated
observations in the ocean seem to point to a man- 12.60
made effect. The western Mediterranean Sea is an 1900 1950 2000

interesting ‘laboratory’ in which the process of

deep water convection can be observed in condi- (b) 38.50
tions that are less hostile than those found in the
high Arctic and Antarctic. The product of western
Mediterranean convection is a homogeneous,
WMDW salinity (PSU)
38.45
warm, saline watermass that ultimately leaves the
Mediterranean and enters the North Atlantic
through the Straits of Gibraltar. Measurements of
the properties of this water have been made since 38.40

early in the twentieth century, since when they

show that the temperature, salinity, and density of
the watermass have increased slowly up to the mid- 38.35
1950s and more quickly thereafter, Figure 3.11.
This change has been attributed to a reduction of
fresh water inflow into the Mediterranean caused
38.30
by the damming of the Nile and those rivers flow-
1900 1950 2000
ing to the Black Sea25.
Year
The detection of anthropogenic effects in the
open ocean is much more difficult. We know little Figure 3.11 Changes in (a) temperature and (b) salinity of
about the inherent ocean variability, so detection the western Mediterranean deep water (WMDW) during
requires a long-term commitment to systematic and the twentieth century.
careful monitoring of the state of the oceans. This
has been embarked upon already in the WOCE.
Over the period from 1990–1997, this experiment
will provide a ‘snapshot’ of the state of the oceans
(Figure 3.12), including the distribution of physical
and chemical properties and an assessment of the We already have some elements of a monitoring
role of ocean circulation in the transport of heat system in place. Sea level from coastal stations can be
and water. The WOCE measurements, Figure 3.13, monitored centrally in real time and is used to
are being used to test and improve ocean models analyse and predict the progress of El Niño events in
running on some of the largest computers now the Pacific (see later). Satellite-tracked drifters as indi-
available14. cators of surface currents are becoming more and
WOCE will produce a baseline picture of the more reliable and can measure not just water move-
oceans against which future measurements can be ment, but also temperature, salinity (being devel-
compared. Such future measurements will be made oped), and meteorological parameters (see Chapter
by the Global Ocean Observing System (GOOS), 19). Many merchant ships are equipped to measure
the ocean element of a Global Climate Observing temperature using expendable bathythermographs to
System (GCOS). WOCE observations are based on provide a set of observations of subsurface tempera-
the use of expensive and sophisticated research ture; but these cover only the major trade routes and
ships, and these cannot be expected to provide contribute almost nothing over the remainder of the
much longer-term routine monitoring; to do this globe.
we must look to other techniques. At present, satellites provide the only global rou-

48
 3: The Role of Ocean Circulation in the Changing Climate

80 80 3.12

60 60

40 40

20 20

0 0

20 20

40 40

60 100 140 60
20 60
80 40 0
180 140 100
80 80

Figure 3.12 The WOCE hydrographic survey grid. Physical and chemical properties will be measured from surface to
sea bed every 50 km along the red lines between 1990–1997.

3.13
Latitude

Longitude
Figure 3.13 The tracks of subsurface ALACE (Autonomous LAgrangian Circulation Explorer) floats
deployed in the Pacific Ocean during WOCE. The floats drift with the currents at a predetermined depth,
surface every 20 days to transmit their position and data to a satellite, and then return to their programmed
depth. This illustrates data points over 2.5 years.

49
 N.C. Wells, W.J. Gould, and A.E.S. Kemp

3.14 Figure 3.14 Sea surface topog-

raphy from the Topex–Poseidon
satellite altimeter. The range is
from –180 cm to +140 cm. Most
prominent is the large surface
height change across the
Antarctic Circumpolar Current
and the clear shape of the North
Atlantic and North Pacific ocean
gyres.

tine monitoring capability (e.g., Figure 3.14), but ocean temperatures routinely. ATOC (Acoustic
what they can observe about the ocean is limited to Thermometry of Ocean Climate; see Figure 19.30)
surface temperatures and sea surface elevations. relies on the fact that sound can be transmitted
However, they do this well (to an accuracy of 0.5°C over vast distances in the ocean, and that the speed
and less than 5 cm) and will continue to be part of of sound in sea water is, at any given pressure, pre-
any future ocean monitoring scheme. dominantly dependent on temperature. An experi-
For measurements of the interior of the ocean, ment in 1991 transmitted sound from Heard Island
new techniques will need to be developed (see in the Southern Ocean to receivers as far away as
Chapter 19). The UK is developing unmanned, the east and west coasts of North America
autonomous submersibles, capable of carrying out (16000 km). Based on these initial encouraging
many of the observations presently made from results, low frequency (70 Hz) sound sources off
research vessels. Such vehicles would be capable of Hawaii and California and receivers around the
traversing entire ocean basins and making measure- rim of the Pacific were deployed in early 1994. This
ments from surface to sea bed on a regular basis ATOC array, Figure 3.16, will show whether the
(Figure 3.15). kind of temperature anomalies that have been seen
Another novel technique is even now being from repeated hydrographic station measurements
deployed in the Pacific with a view to measuring can be reproduced and their evolution monitored.

3.15

Figure 3.15 An artist’s

impression of the AUTOSUB
vehicle carrying out a trans-
oceanic hydrographic survey
mission. The project involves
the development of a very
low drag body, an efficient
propulsion system, sensors
that can retain their accuracy
over long missions, and
deployment and recovery sys-
tems, as well as navigation
and data telemetry schemes.

50
 3: The Role of Ocean Circulation in the Changing Climate

Figure 3.16 A proposed array of 3.16

sources (north and south of Hawaii,
off the Californian coast, and off
Kamchatka, in the N.W. Pacific) and
receivers (small circles, triangles and
stars), to be deployed in the the ATOC
project. The sound paths, for each
source and receiver, are great circles. ✴

Model studies based on coupled ocean–atmosphere In the late 1960s K. Bryan5 published the first
models and a scenario of atmospheric increases in model of the general circulation of the ocean (see
greenhouse gases suggest that ATOC measurements Box 3.1). This was developed further by M. Cox8,
might even be able to detect signals due to global and has been used extensively by research groups
warming. around the world. The Fine Resolution Antarctic
Model (FRAM) project used this general circula-
Modelling of Climate Change tion model as its basis (see Box 3.2), and it is now
The realisation that many of the large-scale being used for other global ocean modelling pro-
processes involved in the ocean, atmosphere, and jects. The development of the AGCM went hand-
ice sheets may be described by a consistent set of in-hand with improvements in the power of the
dynamical equations has been with us since the computer – climate perturbation experiments for a
early twentieth century. L.F. Richardson 24 few decades can be now run in a matter of a few
described his experiment to predict the weather by weeks on a supercomputer.
numerical iteration of the equations of motion in Ocean circulation models, however, have been
1922. He also discussed how the method could be severely hampered by computer resources. First, the
extended to short-term climate prediction. Though response time-scales in the ocean vary from days
his experiment was not a success, the methods were and weeks in the surface layers to the order of cen-
later successfully developed for numerical weather turies in the deep ocean. This means that a global
prediction in the 1950s. These methods were also ocean model may have to be integrated for at least
used by N. Phillips20, who developed the first gen- a 100-year period to reach equilibrium after a cli-
eral circulation model of the global atmosphere. mate perturbation, compared with a period of a
The past 40 years have seen the further improve- few years for an AGCM. Second, in some cases
ment of atmospheric general circulation models there is a need to model ocean eddies (scales of
(AGCM) to include not only the dynamical 50–200 km; see Chapter 4), which in some parts of
processes, but also details of the radiations and the ocean are significant in the transport and mix-
their interaction with cloud, surface processes asso- ing of ocean properties. A finer grid resolution of
ciated with hydrology and vegetation, and the about 10 km is required to fully resolve these
ocean. Indeed, the atmospheric models have been oceanic eddies in ocean models, compared with the
at the forefront of estimating the response to the lower resolutions needed in AGCMs. It is clear that
predicted changes in greenhouse gases. computers of two orders of magnitude faster than

51
 N.C. Wells, W.J. Gould, and A.E.S. Kemp

Box 3.1 An Ocean General Circulation Model

An ocean general circulation model is composed of a set of mathematical equations which describe
the time-dependent dynamical flows in an ocean basin. The basin is discretised into a set of boxes of
uniform horizontal dimensions, but variable thickness in the vertical dimension. The horizontal flow
(northward and eastward components) is predicted by the momentum equation, Figure 3.17(a), at the
corners of each box (Figure 3.18).

3.17
(a) Momentum equation

Rate of change = Transport of momentum + Coriolis + Horizontal + Friction

of horizontal across faces of box acceleration pressure
momentum gradient
in box

(Rotation (Gradients in sea (Internal dissipation

of earth) level and in density) and at ocean boundaries)

(b) Temperature equation

Rate of change = Transport of temperature + Mixing of temperature between

of temperature across faces of box box and adjacent boxes
in box

(c) Salinity equation

Rate of change = Transport of salinity + Mixing of salinity between

of salinity across faces of box box and adjacent boxes
in box

Figure 3.17 The basic equations for an ocean general circulation model.

The forcing for the flow may come from the surface wind stress (the frictional term in the momen-
tum equation) or from surface buoyancy fluxes, arising from heat and fresh water (precipitation–evap-
oration) exchange with the atmosphere. These buoyancy fluxes change the temperature and salinity
in the surface layer of the ocean. However, the horizontal and vertical flow carry these properties far
into the interior of the ocean, where they tend to mix with other water masses.
This process of transport and mixing is described by the temperature and salinity equations, Figures
3.17(b) and 3.17(c), at the centre of each ocean
3.18 Heat loss
Earth's box (Figure 3.18). From these two equations the
rotation
sea water density and thence the pressure can be
obtained for each box. The horizontal pressure
gradient is then determined for the momentum
equation, while the vertical velocity is calculated
from the horizontal divergence of the flow. This
Wind set of time-dependent equations can then be used
to describe all the dynamical components of the
flow field, provided that suitable initial and
boundary conditions are specified.

u u
Figure 3.18 A schematic of the model boxes in
T
an ocean general circulation model. The equa-
u S u tions for momentum are solved at the corners of
the boxes (u), while the temperature (T), and
salinity (S) equations are solved at the centres of
the boxes. The model is forced by climatologi-
cal wind stress, surface heat, and fresh water
fluxes.

52
 3: The Role of Ocean Circulation in the Changing Climate

Box 3.2 The Fine Resolution Antarctic Model (FRAM)

FRAM was developed26 to investigate the role of eddy processes in the circulation of the Southern
Ocean, in particular the Antarctic Circumpolar Current. The Southern Ocean comprises 30% of the
global ocean and is an important region for the transfer of heat and fresh water between the Antarctic
Ice Sheet and the northern land masses. The dynamics of the Antarctic Circumpolar Current are not
well understood, although there is strong evidence to suggest that energetic ocean eddies, on hori-
zontal scales of the order of 100 km, play an important role in the dynamics of the current.
The FRAM model forms the basis for the development of a global ocean model, which will be used
for climate change experiments. This global model will resolve these energetic ocean eddies and the
major frontal zones of the ocean.
The FRAM model subdivides the ocean, south of 22°S to the Antarctic continent, into a regular set
of boxes. Each box has a horizontal length of 0.25° latitude by 0.5° longitude (approximately 27 km x
27 km at 60°S). Beneath each surface box a string of boxes reaches to the ocean floor. The thickness
of each box varies from 20 m in the surface layer to over 200 m in the deepest parts of the ocean.
Within each box, equations for the northward and eastward horizontal components of momentum,
temperature, and salinity are specified. There are 5 million boxes which represent the southern ocean,
and therefore 20 million prognostic variables to calculate. These variables are calculated by integra-
tion in time of the equations from an initial cold, saline motionless ocean. In the first 6 years of the
integration the model was forced by the annual mean wind stress and by the observed temperature
and salinity. After this period the model was free to run for a further 6 years, subject to seasonal wind
forcing, annual mean temperature and salinity at the ocean surface, and an open northern boundary.
An example of the model ‘output’ is shown in Figure 3.19 and a comparison with satellite data is
shown in Figure 3.20. Estimates of the meridional heat flux in the model are shown in Figure 3.21.

3.19 Figure 3.19 Contours of the instantaneous stream

function in FRAM. The stream function shows the
depth-averaged flow circulating clockwise around the
Antarctic continent. The flow is most intense in the
Antarctic Circumpolar Current (in the yellow and
neighbouring green regions). The flow is unsteady due
to the presence of ‘eddies’, mainly in the Antarctic
Current and Agulhas Current south of South Africa.

0.3 3.21

0.2

0.1 Latitude
30˚ 40˚ 50˚ 60˚ 70˚S
3.20
H (PW)

(a)
60˚W 40˚ 20˚ 0˚ 20˚E 0.0
S

30˚ –0.1

–0.2
40˚
Total
–0.3 Mean flow
Transient flow
50˚ –0.4
(b)
S

30˚
Figure 3.21 The latitudinal variation in meridional-
heat transfer (H) by the FRAM model. Northward
40˚
heat fluxes are positive. The total heat transport is
directed southward (negative) toward the pole,
50˚ although the time-mean circulation drives a heat
flow toward the north, between 37°S and 43°S, in
the region of the Antarctic Circumpolar Current.
Figure 3.20 (a) Tracks of eddies detected by the The ‘eddies’ in the flow, however, drive a stronger
Geosat satellite in the South Atlantic. (b) Tracks of heat flux toward the south, and thus result in a total
eddies in the FRAM. Note that the model produces a heat transport toward the pole. The eddies play an
more regular eddy track than the observations show. important climatological role in the model.

53
N.C. Wells, W.J. Gould, and A.E.S. Kemp

the present machines (which are capable of 109 (a) Upper Westerlies 3.22
floating point instructions per second) will be
required to manage this task. The recent develop-
Indonesia Pacific Ocean
ment of parallel computer systems is expected to
deliver this power by end of the twentieth century. Indian
W C
Equator
Ocean
Indeed, one of the important tasks of ocean mod- South East Trades South
America
ellers is to develop methods for the analysis and
interpretation of the huge quantity of data that
these models will produce. Because of their com-
(b)
plexity, there will be also be a need to develop sim-
pler models to investigate interactions more com-
pletely. To understand and predict the behaviour of
sea ice, for example, the sea-ice models will have to Indonesia

be coupled to models of the upper ocean and over- W W

Equator

lying atmosphere. Surface

Westerlies
Weak
South East Trades South
America
A third problem is the importance of ocean 120˚E 160˚W 80˚W
chemistry and biology to climate change. The
ocean is a depository for the greater part of the
Earth’s exchangeable fraction of carbon. It is not Figure 3.22 The tropical atmospheric circulation over
known how carbon is regulated by the ocean, the Pacific Ocean during (a) normal conditions and (b) El
though it is clear that phytoplankton blooms pro- Niño conditions. During normal conditions the surface
duce a lowering of the partial pressure of carbon pressure is low over Australia and Indonesia (high rain-
dioxide at the ocean surface and thereby have the fall) and high over the southeast Pacific, so the surface
trade wind circulation is strong and the Southern
ability to alter the flux between ocean and atmos- Oscillation index (PDARWIN – PTAHITI, where P is the seasonal
phere (see Chapters 6 and 12). Modelling of the surface pressure) is high. During El Niño conditions, the
ocean basin ecology and chemistry has commenced pressure is higher over Australia and Indonesia (low rain-
in recent years and it is expected that these process- fall) and lower in the southeast Pacific; consequently the
es, as they become understood, will be incorporat- trade wind circulation is weaker and the Southern
ed into the general climate models. Oscillation index is low (W, warm; C, cold).

The El Niño–Southern Oscillation

(ENSO) Phenomenon – An Example of
Ocean Prediction?
The ENSO phenomenon is now recognised as the of the tropical trade wind system, in particular the
largest contributor to the perturbation of the cli- southeast trades which occur on the eastern flank
mate on a global scale over a period of a few years, of the South Pacific anticyclone. The normal wind
and is known to be a natural oscillation of the circulation produces strong winds, which drive an
atmosphere–ocean system. It is a coupled interac- upwelling of cooler, nutrient-rich, and highly pro-
tion between the atmosphere and the upper layers ductive thermocline waters along the tropical coast
of the tropical Pacific Ocean, which can result in of South America and on the equatorial band of the
changes of global surface temperature of a few Eastern Pacific. When the trade winds weaken, the
tenths of a degree Celsius on a time-scale of one upwelling is reduced and the waters warm by as
year. This is a similar change in temperature to that much as 5°C, due to both the southward move-
attributed to the 25% increase in atmospheric car- ment of warmer equatorial waters and the high
bon dioxide in the past 100 years. solar radiation at the surface.
The El Niño is the ocean component of the At first sight this appears to be a one-way forc-
interaction; it is a general warming of the upper ing of the atmosphere by the surface wind on the
layer of the eastern and central Equatorial Pacific ocean, as is the case in most of world’s upwelling
Ocean. [The name El Niño comes from the fact regions. However, this area of the Pacific Ocean
that the impact is felt on the coast of South behaves rather differently because the rise in sea-
America around Christmas time and hence it is surface temperature over a vast area of ocean
referred to in Spanish as the (Christ) child.] A changes the distribution of atmospheric heat
major consequence of the oceanic warming is the sources and sinks, which in turn drives the trade
decline of biological productivity and hence of fish wind circulations. The trade winds carry water
stocks, which are a major source of livelihood for vapour, evaporated from the ocean, into areas of
the local population. tropical atmospheric convergence where high rain-
It is associated with a reduction in the strength fall occurs. These convergence zones tend to be

54
 3: The Role of Ocean Circulation in the Changing Climate

3.23

(a) (b)

Pacific Ocean Pacific Ocean

Equator L H H L Equator

Pacific Pacific
Ocean Ocean
Equator Equator Warm
Warm

Cooler Cooler

Warmest water H/L = Higher/Lower atmospheric pressure

Figure 3.23 The tropical Pacific Ocean during (a) normal and (b) El Niño years. During normal years the strong trade
winds drive the warm water westward and intensify upwelling of cooler subsurface waters in the east Pacific. During
El Niño years the weaker atmospheric circulation allows the warmer lighter water to flow eastward, replacing the cool-
er upwelling waters.

located in regions of maximum surface tempera- The monitoring of ENSO in recent decades, in
ture, that is to the north of the equator and in the particular during the exceptional episode of
tropical southwest Pacific. During El Niño events 1982–198322, has provided a stimulus to atmos-
these convergence zones tend to move southward pheric and ocean modellers. These efforts have pro-
across the equator and eastward (see Figures 3.22 vided good simulations of both the individual com-
and 3.23). The South Pacific anticyclone becomes ponents of the system (the ocean response to
weaker and the trade winds weaken. observed winds and the atmospheric response to
Simultaneously, the surface atmospheric pressure observed sea surface temperature).
over Indonesia and Australia tends to rise and rain- Coupled models of the tropical ocean and
fall decreases. atmosphere are showing promise for the simulation
This see-saw of surface pressure between the of the ENSO cycle and a number of Climate
southeast Pacific and Indonesia is known as the Centres in the world are now producing experi-
Southern Oscillation. Although its major influence mental forecasts for ENSO with some degree of
is in the tropical Pacific, its effect is felt throughout success.
the world. For example, in Zimbabwe both rainfall These models will be improved when the results
and maize yields are highest during El Niño years6. of the Tropical Ocean Global Atmosphere experi-
The ENSO phenomenon is a coupled interaction ment (an experiment to measure and understand
between the surface layers of the ocean and the some of the complex processes in the tropical
world wind systems, which occurs two or three atmosphere and ocean) in the western equatorial
times a decade. Pacific Ocean are analysed. The proposed mooring

55
 N.C. Wells, W.J. Gould, and A.E.S. Kemp

3.24

Figure 3.24 The tropical Pacific Ocean observing system. The orange diamonds are ATLAS surface buoys from which
continuous temperature measurements are obtained from the surface to a depth of 500 m. These observations together
with surface meteorological measurements (e.g., wind, temperature and humidity) are transmitted by satellite to a
receiving station. Currents (orange squares) are measured routinely along the equator. The Tide gauge network (yellow
circles) provides sea-level height observations, which can be used to calculate horizontal pressure gradients in the
oceans. These observing stations are supplemented by routine measurements from satellite-tracked drifting buoys
(arrows) and voluntary observing ships (light blue lines).

array (Figure 3.24) will provide routine measure- The Importance of Climate Change
ments of the upper 500 m of the tropical Pacific The most comprehensive study of the scientific evi-
Ocean and of surface winds to be used in the dence for climate change, its potential impact, and
ENSO forecast models. the strategies needed to ameliorate its impact has
There is also evidence that the intensity of been made by the Intergovernmental Panel on
ENSO varies on the longer decadal to centennial Climate Change (IPCC). While climate change is, in
time-scales. Longer-term variations in the intensity general, regarded as an atmospheric phenomenon,
of El Niño events affecting primary production off the IPCC reports make it clear that there are direct
California have been related to solar cycles (varia- impacts on and by the ocean and, furthermore, that
tions in the intensity of 11–22 year sunspot cycles knowledge of the behaviour of the oceans is central
that cause small changes in incident radiation). to any climate prediction capability. The potential
Within Californian continental margin sediments, socioeconomic costs of climate change far outweigh
decadal to millennial cycles of preservation of lami- the resources needed to make the measurements
nated sediments, driven by variation in the intensity and run the models required to improve our ability
of the oxygen minimum zone, have been ascribed to predict climate change. The key areas identified
to solar cycles affecting the longer-term alternation in the IPCC report, together with recommendations
of El Niño and anti-El Niño2. on improvements in observations and modelling,
Data on sea surface temperature variations from are given in Box 3.3.
oxygen isotope studies of Galapagos corals show- Clearly, in all these areas the oceans are impor-
ing 11 and 22 year periods11 lend support to sug- tant and understanding them represents one of the
gestions of solar cycle modulation of ENSO greatest challenges in the area of climate change
activity. prediction.

56
 3: The Role of Ocean Circulation in the Changing Climate

Box 3.3 The Scientific Uncertainties of Climate Change

IPCC identifies the key areas of scientific uncertainty as:
• Clouds: primarily cloud formulation, dissipation, and radiative properties, which influence the
response of the atmosphere to greenhouse forcing.
• Oceans: the exchange of energy between the oceans and the atmosphere, between the upper layers
of the ocean and the deep ocean, and transport within the ocean, all of which control the rate of global
climate change and the patterns of regional change.
• Greenhouse gases: quantification of the uptake and release of the greenhouse gases, their chemical
reactions in the atmosphere, and how these may be influenced by climate change.
• Polar ice sheets: affect predictions of sea level rise.
The main observational requirements are:
• The maintenance and improvement of observations (such as those from satellites) provided by the
World Weather Watch.
• The maintenance and enhancement of a programme of monitoring, both from satellite-based and
surface-based instruments, of key climate elements for which accurate measurements on a continu-
ous basis are required. These include the distribution of important atmospheric constituents, clouds,
the Earth’s radiation budget, precipitation, winds, sea surface temperatures, and the terrestrial ecosys-
tem extent, type, and productivity.
• The establishment of a Global Ocean Observing System to measure changes in such variables as
ocean surface topography, circulation, transport of heat and chemicals, and sea ice extent and thick-
ness.
• The development of new systems to obtain data on the oceans, atmosphere, and terrestrial ecosys-
tem using both satellite-based instruments and instruments based on the surface, on automated vehi-
cles in the ocean, on floating and deep sea buoys, and on aircraft and balloons.
• The use of palaeoclimatological and historical instrumental records to document natural variability
and changes in the climate system, and subsequent environmental response.
In the area of modelling the report concludes that any reduction in the uncertainties of climate pre-
diction will be dictated by progress in the areas of:
• Use of the fastest possible computers to take into account coupling of the atmosphere and the
oceans in models, and to provide sufficient resolution for regional predictions.
• Development of improved representation of small-scale processes within climate models, as a
result of the analysis of data from observational programmes to be conducted on a continuing basis
well into the twenty-first century.

General Reference 4. Broecker, W.S, Kennett, J.P., Flower, B.P., Teller, J.T.,
Wuethrich, B. (1995), El Niño goes critical, New Scientist, 145, Trunbone, S., Bonani, G., and Wolfi, W. (1989),
32–35. Routing of meltwater from the Laurentide Ice Sheet
during the Younger Dryas cold episode, Nature, 341,
318–321.
References 5. Bryan, K. (1969), A numerical model for the study of the
1. Aagaard, K. and Carmack, E.C. (1989), The role of sea ice world ocean, J. Computat. Phys., 4, 347–376.
and other fresh water in the Arctic Circulation, J. 6. Cane, M.A., Eshel, G., and Buckland, R.W. (1994)
Geophys. Res., 94(C5), 14485–14498. Forecasting Zimbabwean maize yield using east equa-
2. Anderson, R.Y., Linsley, B.K., and Gardner, J.V. (1990), torial Pacific sea surface temperatures, Nature, 370,
Expression of seasonal and ENSO forcing in climatic 204–205.
variability at lower than ENSO frequencies: evidence 7. Carrissimo, B.C., Oort, A.H., and Van de Harr, T.H.V.
from Pleistocene marine varves off California, (1985), Estimating the meridional energy transports
Palaeogeogr., Palaeoclimatol., Palaeoecol., 78, in the atmosphere and ocean, J. Phys. Oceanogr., 15,
287–300. 52–91.
3. Birchfield, G.E. and Broecker, W.S. (1990), A salt oscillator 8. Cox, M.D. (1984), A Primitive Equation: 3-Dimensional
in the glacial Atlantic? A ‘scale analysis’ model, Model of the Ocean, GFDL Ocean Group Technical
Paleoceanogr., 5, 835–843. Report No.1, GFDL/NOAA, Princeton University.

57
N. C. Wells, W. J. Gould, and A. E. S. Kemp

9. Delworth, T., Manabe, S., and Stouffer, R.J. (1993), 20. Phillips, N.A. (1956), The general circulation of the atmos-
Interdecadal variations of the thermohaline circula- phere: a numerical experiment, Quart. J. Roy.
tion, J. Climate, 6, 1993–2011. Meteor. Soc., 82, 124–164.
10. Dickson, R., Meinke, J., Malmberg, S., and Lee, A. (1988), 21. Pugh, D. (1987), Tides, Surges and Mean Sea Level: a hand-
The Great Salinity Anomaly in the northern North book for engineers and scientists, John Wiley and
Atlantic 1968–1982, Progr. Oceanogr., 20, 103–151. Sons, Chichester, 472 pp.
11. Dunbar, R.B., Wellington, G.M., Colgan, M.W., and Glynn, 22. Rasmusson, E.M. (1985), The 1982/83 El Niño Event,
P.W. (1994) Eastern Pacific sea surface temperature World Meteorological Organisation, Marine
since 1600 AD: the 618O record of climate variability Meteorology and Related Oceanographic Activities
in Galapagos corals, Paleoceanogr., 9, 291–315. Report, No. 14, 11–22.
12. Ellett, D.J. and Blindheim, J. (1992), Climate and hydro- 23. Read, J.F. and Gould, W.J. (1992), Cooling and freshening
graphic variability in the ICES area during the 1980s, of the subpolar North Atlantic Ocean since the
ICES Mar. Sci. Symp., 195, 11–31. 1960s, Nature, 360, 55–57.
13. Gordon, A.L. (1986), Interocean exchange of thermocline 24. Richardson, L.F. (1922), Weather Prediction by Numerical
water, J. Geophys. Res., 91, 5037–5046. Process, Cambridge University Press, Cambridge.
14. Gould, W.J. (1994), Update: World Ocean Circulation 25. Rohling, E.J. and Bryden, H. (1992), Man-induced salinity
Experiment, Sea Technology, 35(2), 25–32. and temperature increases in western Mediterranean
15. GRIP Project Members (1993), Climate instability during deep water, J. Geophys. Res., 97(C7), 11191–11198.
the last interglacial period recorded in the GRIP ice 26. The FRAM Group (1991), An eddy-resolving model of the
core, Nature, 364, 203–207. Southern Ocean, EOS Trans. AGU, 72(15), 169,
16. Lorius, C., Jouzel, J., and Reynaud, D. (1993), 174–175.
Glacials–interglacials in Vostok: climate and green- 27. Trenberth, K.E. (1979), Mean annual poleward energy
house gases, Global Planet. Change, 7, 131–143. transports by the oceans in the Southern Hemisphere,
17. Newell, R.E., Kidson, J.W., Vincent, D.G., and Boar, G.J. Dynam. Atmos. Ocean, 4, 57–64.
(1972), The General Circulation of the Tropical 28. Wijffels, S., Schmitt, R., Bryden, H., and Stigebrandt, A.
Atmosphere, Vol. 1, The MIT Press, 258 pp. (1992), Transport of fresh water by the oceans, J.
18. Oort, A.H. and Vonder Harr, T.H. (1976), On the observed Phys. Oceanogr., 22, 155–162.
annual cycle in the ocean–atmosphere heat balance 29. Zahn, R. (1992), Deep ocean circulation puzzle, Nature,
over the Northern Hemisphere, J. Physical. 356, 746.
Oceanogr., 6, 781–800.
19. Parrilla, G., Lavin, A., Bryden, H., Garcia, M., and Millard,
R. (1994), Rising temperatures in the Subtropical
North Atlantic Ocean, Nature, 369, 48–51.

58
CHAPTER 4:

Ocean Weather –
Eddies in the Sea
K.J. Richards and W.J. Gould
Introduction pheric counterparts and evolve on a much longer
Weather maps brought to us daily via television time-scale. It is only recently that we have been
and newspapers have made us all aware of the able to sample the richness in structure of the
changing state of the atmosphere (Figure 4.1). oceanic eddy field, and we have only just begun to
Cyclones and anticyclones (low and high pressure assess the importance of these oceanic eddies in
systems) evolve and interact, produce severe winds, shaping the large-scale ocean circulations and their
fronts with associated rain, and vary the weather impact on climate and the biology of the oceans.
we experience on time-scales from a few hours to
several days. Less well-known is that the ocean is Our View of the Ocean
populated by very similar systems. The oceanic Our understanding of the nature of the ocean circu-
equivalent of atmospheric highs and lows, the lation is determined by the tools at our disposal to
ocean weather is again an ever-changing pattern. observe it. Just as the development of astronomical
As in the atmosphere, intense storms can develop in telescopes (the most recent of which is the Hubble
the ocean to produce strong currents. Fronts sepa- Space Telescope) has given us successively deeper
rating warm and cold water masses are a common insights into the structure of the Universe, so has
occurrence. The oceanic systems, however, have a our view of the oceans altered as new measuring
much smaller horizontal scale than do their atmos- techniques have become available.

Figure 4.1 A familiar sight for those in the UK, a low- 4.1
pressure system and associated rain-bearing fronts
sweeping across the country. This illustrates the surface
pressure field for a day in January. The low-pressure sys-
tem, or cyclone, in the centre of the picture is approxi-
mately 2000 km across. The wind circles the low pres-
sure in an anticlockwise sense. Interaction between high
and low pressure systems dictates the weather we experi-
ence on a daily basis. The insert, expanded by a factor of
10, shows a detail from an infra-red image of the sea sur-
face taken from a satellite. Light areas correspond to cool
water and dark areas to warm water. Using the sea sur-
face temperature as a tracer, we can clearly see the
imprint of an ocean eddy, 100 km across, where the
ocean currents associated with the eddy have caused the
warm waters to the south and the cool waters to the
north to spiral around each other; an oceanic cyclone (in
the insert note also the sharp transition from warm to
cool water, an ocean front, and the smaller scale struc-
tures, 10 km, along the front). Oceanographers call the
cyclones and anticyclones of the ocean meso-scale
eddies. These ocean eddies are dynamically equivalent
to the weather systems in the atmosphere. However,
ocean eddies have very different space- and time-scales;
typically, horizontal scales of 10–200 km, current speeds
of a few tens of centimetres per second, and circulation
times of tens of days. In the vertical, some are restricted
to the upper levels of the ocean, while others extend to
the bottom.

59
 K.J. Richards and W.J. Gould

4.2 Figure 4.2 The earliest ideas of

ocean currents came from the
long-term drift of debris from
shipwrecks. Some hulks re-
mained water-logged and drifted
with the surface currents for
years on end. This compilation
of such trajectories from the
North Atlantic presents a picture
not greatly different from our
present best estimates of mean
ocean currents9.

Until well into the twentieth century, measure- Simultaneously, in the UK John Swallow 18
ments of ocean surface currents were made by com- (Figure 4.3) had the idea of using floats that would
piling observations of ship drift, made during the drift with the subsurface currents. The floats he
course of routine navigation, and of floating objects used were sealed aluminium tubes designed to be
(Figure 4.2). These served to delineate the major less compressible than sea water and contained
currents of the upper ocean with which we are what would now be regarded as a very primitive
familiar; the intense currents on the west side of acoustic beacon. By virtue of their compressibility
ocean basins (Gulf Stream, Kuroshio, and Somali being lower than that of sea water, they could be
Current), the Antarctic Circumpolar Current sys- ballasted to sink at the ocean surface and would
tem, and the equatorial currents. Maps drawn by
Rennell in the eighteenth century (see Figure 1.5)
and by Maury in the nineteenth century are not very 4.3
different from those that appear in today’s atlases.
Until 40 years ago the measurement of subsur-
face currents was well-nigh impossible. The only
direct method relied on the tracking of surface
buoys attached to subsurface drogues from atten-
dant ships, and was limited to short (no longer
than two-week) measurements at shallow depths,
which were also distorted by wind and wave forces
on the surface buoy. The development of the
dynamical method by Helland, Hansen, and
Nansen, published early in the twentieth century,
enabled the vertical structure of currents to be
inferred from the vertical profiles of density, but
the method was not capable of measuring the
changes in currents over long periods, their
absolute values, or detailed horizontal spatial
structures.
It was not until the 1950s and 1960s that Figure 4.3 The invention of the neutrally buoyant
attempts were made to make reliable, absolute Swallow float in the 1950s gave the first indication that
measurements of the deep subsurface currents of deep currents in the interior of the ocean were not the
then-predicted sluggish drift, but were dominated by
the ocean. Bill Richardson in the US had limited energetic (a few centimetres per second) currents associ-
success in developing and using recording current ated with meso-scale eddies. Here John Swallow is on
meters on moorings. The electronics were cumber- board the RRS Discovery II preparing an early float (made
some and unreliable and the deployment methods from scaffold tube) and watched intently by the ship’s cat
(on moorings) were equally vulnerable. and two of the crew.

60
 4: Ocean Weather – Eddies in the Sea

(b)
gain buoyancy relative to the water in which they (a)
A1
A2
4.4
A3
were located as they sank. Then, each individual B1 B2 B3 B4 B5 B6 B7
14

float would reach neutral buoyancy (its weight 29˚ C1 C2

C3
C4 C5 C6 C7 29˚ 7
15
N D1 D2 D3 D4 D5 N
equalling the weight of the displaced water; E1 E2
E3 E4 E5 E6 E7 E8 E9
E10 13
6 19
2
N4
Archimedes principle) at some depth determined by 28˚
F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11
28˚
3 25
1
8

G4 G5 G6 G7 G8 G9 100 km 200 km 5 18
its initial weight. The float would remain at this G1
G2 G3
H1 H2 H3 H4 H5
G10 G11
17 4
N1
N2

16
depth and drift with the currents (since if it moved 27˚
I1 I2 I3 I4 I5 I6 I7
27˚ 12
10
20
9 N3

upward from its balance depth the float would J1 J2 J3 J4

K3
J5 J6 J7

expand less than the water around it, become heav- K1 K2

24 11

71˚W 70˚ 69˚ 68˚

ier than the water, and sink back to the original 71˚W 70˚ 69˚ 68˚

depth, and vice versa for downward movements). (c)

The acoustic beacon enabled bearings to be taken
from the attendant ship and hence the position and 18˚
N
depth of the float could be estimated (see Chapter
19). These floats were designed to investigate the
predicted sluggish (of the order a few millimetres
per second) deep currents, but were also used to 16˚
confirm the existence of a more energetic counter-
current deep beneath the Gulf Stream, postulated
by Stommel and Arons17.
The measurements of the deep flow south of 35˚W 34˚ 33˚ 32˚

Bermuda made by Crease and Swallow19 in 1960

using Swallow’s float showed the surprising result
that the currents were far from sluggish. Indeed, Figure 4.4 In the 1970s arrays of moorings were set in
they were sufficiently energetic that floats rapidly the US MODE and the USSR Polygon experiments; they
escaped from the acoustic tracking range of the were used to map the current fields at a number of dis-
ketch Aries, from which the measurements had crete depths. These were the largest experiments ever
been made. The experiment was modified and undertaken and in the case of the Polygon array involved
recovering and replacing each mooring eight times5,15. (a)
revealed for the first time the totally unforeseen
MODE hydrographic grid; (b) MODE current meter array;
energetic deep currents. (c) Polygon-70 hydrographic grid (open circles) and cur-
These measurements made in the late 1950s and rent meter array (filled circles).
early 1960s gave tantalising glimpses of the large-
scale complexity of the deep flow, but the tech- scales, and became known to the Soviet scientists as
niques were not adequate to fully investigate the the synoptic scale.
nature the currents. This had to await the develop- By the early 1970s both the Soviet and western
ment, by the early 1970s, of reliable current- scientists were able to mount concerted investiga-
recording meters, the moorings to support them, tions of the spatial and temporal characteristics of
and the design and construction of neutrally buoy- what we call here meso-scale variability. In 1970
ant floats that could be tracked for hundreds rather Soviet scientists carried out the Polygon-70 experi-
than tens of days and over thousands rather than ment in the Atlantic North Equatorial Current and
tens of kilometres. in 1972 US and UK scientists conducted a 4-month
Gradually, the evidence for the nature of the experiment on a similar scale in the area south of
deep flow accumulated. Long-duration current Bermuda [the Mid-Ocean Dynamics Experiment,
measurements made with current meters at places MODE; Figures 4.4(a) – 4.4(c)].
like Woods Hole Oceanographic Institution Site D The MODE experiment was made possible by
on the continental slope north of the Gulf Stream the use of neutrally buoyant floats [some from the
showed a ‘spectral gap’ of low energy between the UK with ranges of up to 100 km and, more impor-
relatively energetic tidal–inertial periods (of the tantly, long-range SOFAR (SOund Fixing And
order of 1 day) and the much longer periods of tens Ranging) floats from the US that could be tracked
and hundreds of days2 that became known as the at distances of several hundred kilometres; see
meso-scale. In the Soviet Union during the 1960s a Chapter 19). The analyses of these experiments
number of long-term current measurements made using objective mapping techniques revealed the
in the Indian Ocean under the direction of existence of features in both the flow and density
Shtockman showed the same time- and space- fields with typically a 100 km horizontal scale

61
 K.J. Richards and W.J. Gould

4.5 1 –1 1 1 2 Figure 4.5 The measurements from

–2 1
–
0 0 –2

–3
–2
–1 L
0
–4
–6
3 MODE current meters and neutrally
29˚ 29˚ 29˚
L
0 L 0 buoyant SOFAR floats were mapped

–1
N 1 0 N –4 3 –2 N
– 1 4 –1
32
54 3
2 –6
–42
–
5
–2
–3
to produce a coherent picture of the
6 4
28˚ H 28˚ 5
H 28˚ 0 H –4 flow field, using a technique called

–5
H

6
L
8
7 6
5 4
32
H
8 7
6
5 4
3
2
2
4 objective analysis in which a flow
1 3

27˚
5 4
32
0
–1 27˚ 3
4
5
–
1
0 27˚
1
0
2
field was devised that had the same
1 –2 21 –4 – 1 –1
0
L
Day = 100 0 –
3 –2
–3
–2
Day = 130
statistical characteristics as the data
–1 L5 Day = 110
–3 L
–4 L themselves and fitted the observations
71˚W 70˚ 69˚ 68˚ 71˚W 70˚ 69˚ 68˚ 71˚W 70˚ 69˚ 68˚ with minimum error. There is little
confidence in the predicted flow field
4 H
at points outside the observational

–5
–4
3 0 H
–7 –1 –2 –3 4 –6 6
2 L 5
3
29˚ –6
L 1
0 29˚ L 2
1 29˚ –5
–4 0 12
3
4
array marked by the dashed lines14.
N –1 N 0 1 0
–1 N –3 1
–5 2 2 0 –1
–4 –2 –2 –2 3 –2
0 12345 3
–3 –3 4 –3 –1 4 –3
5 5
28˚ –2 –1 –4 6 28˚
8

7 6
H 28˚ H H
L
6
5 7
34 6
2 –4 5
0 1 4 3 2 1 0 –1 –2
–2

27˚ 27˚ L 27˚

L –3 –2
–1

–1 0
–2 –1 –3
–3 Day = 140 –1
–4 –4
Day = 155 L Day = 165
–5

71˚W 70˚ 69˚ 68˚ 71˚W 70˚ 69˚ 68˚ 71˚W 70˚ 69˚ 68˚

(Figure 4.5) and with a complex vertical structure. maintained by German scientists – the longest
Features were seen to propagate westward at direct current meter measurement series20).
speeds of about 5 km/day. The complexity of the The Polygon, MODE, and Polymode experi-
current field was amply illustrated by the compila- ments studied meso-scale features in detail, but
tion of the SOFAR float tracks into so-called from the early 1970s onward evidence accumulated
spaghetti plots (Figure 4.6), which show the how (from the newly developed observational tech-
the eddy-like meso-scale field acted to disperse the niques of SOFAR floats, reliable current meters,
floats10. satellite infra-red images of the ocean surface, and
Even though the areas of ocean covered by both satellite altimetry data; see Chapter 5, Box 5.4) for
of these experiments were 200 km across, this is the ubiquitous nature of the meso-scale eddy field
still a relatively small area compared to the size of (see, for example, Figure 4.7). All these observa-
ocean basins. The experiments, therefore, shed little tions confirmed the concentration of high kinetic
light on the geographical variability of the charac- energy of the time-varying currents (the eddy kinet-
teristics of the meso-scale currents. In order to ic energy, EKE) in regions near the major current
explore this a joint US–USSR experiment, carried systems.
out between August 1974 and April 1975 and So a view was formed, which still holds good
called Polymode (a coming together of names as
well as scientists), explored the energetics and
scales of meso-scale motions over a much wider 4.6
range of latitudes in the region of anticyclonic cir-
culating water southeast of the Gulf Stream, called
the western subtropical Atlantic gyre. In parallel, in
the east Atlantic a more limited exploration of eddy
variability using long-term moorings (the North
East Atlantic Dynamics Study, NEADS) was started
by European scientists (one of these current meter
sites between the Azores and Madeira is still being

Figure 4.6 A compilation of the tracks (9/1972–6/1976)

of all the SOFAR floats launched in the MODE array area
near 28°N 70°W (1° of latitude is approximately
111 km). The tracks last several years and show the dis-
persive nature of currents from a small area to eventually
fill much of the western subtropical gyre. In some areas,
for instance off the Bahamas and near the Gulf Stream,
eddy activity is high. The site chosen for the MODE
experiment turned out to one of very low eddy energy15.

62
 4: Ocean Weather – Eddies in the Sea

Figure 4.7 The statistics of 4.7

eddy energy derived from the
variability of sea surface slope
obtained from the Topex/
Poseidon satellite. High eddy
variability is seen in the regions
of strong meandering currents,
such as the Gulf Stream in the
Atlantic, the Kuroshio in the
North Pacific, the Agulhas
Retroflection off South Africa,
and the Antarctic Circumpolar
Current in the Southern Ocean.
(Courtesy of Prof C. Wunsch,
Massachusetts Institute of
Technology, Boston, USA.)

today, of the ocean populated with eddies, just as images (Figure 4.8). The spatial distribution of Gulf
the atmosphere is full of cyclones, anticyclones, and Stream rings has been delineated by several census-
frontal systems. es (see, for example, Figure 4.9), but none is truly
comprehensive.
Beasts in the Eddy Zoo By virtue of being water masses enclosed within
Meso-scale features in the ocean take on a number water with very different properties the rings can
of guises. Indeed, in the mid-1970s an article was be regarded as isolated ecosystems and the
published entitled New Animals for the Eddy Zoo, evolution of their physical, chemical, and biological
since at that time almost every eddy studied properties over lifetimes of several seasons can be
appeared to have different characteristics. For
some, there was clear evidence of their presence at 4.8
the sea surface, while others were confined within
the water column; their diameters ranged from over
200 km to about 10 km, and they appeared to have
different formation mechanisms. In a review such
as this we can only touch on the characteristics of
some of the more abundant animals in the eddy
zoo. (The reader is referred to Robinson13 for a col-
lection of papers giving an overview of our knowl-
edge at that time.)
Gulf Stream rings
Some of the best-documented eddies are those that
are formed by ‘pinching off’ meanders from ener-
getic current systems. The area that has been most
intensively studied is the Gulf Stream – its variabili-
ty downstream of Cape Hatteras is well-known.
Even as long ago as 1793 evidence for an isolated
body of warm water to the north of the stream was
noticed, and in the 1930s analysis of ships’ thermo-
graph records started to provide evidence of the
population of eddies associated with the Gulf
Figure 4.8 The meandering of a current and the location
Stream, known as Gulf Stream rings. of eddies detached from the current can be clearly seen
Undoubtedly, the detailed study of these features in satellite infra-red images of sea surface temperature.
and their formation was strongly influenced by the Warmer hues denote warmer temperatures. (Courtesy of
advent of infra-red sensors flown on satellites that O. Brown, R. Evans, and M. Carle, University of Miami
could easily identify both warm and cold core rings Rosenstiel School of Marine and Atmosphere Science,
north and south of the Gulf Stream on cloud-free Miami, USA.)

63
 K.J. Richards and W.J. Gould

Figure 4.9 (a) A visualistion of the process of ‘pinching

4.9 (a)
off’ a Gulf Stream ring. (b) The distribution of the centres
Slope water of Gulf Stream cold core rings8.
GULF

REAM

studied. For example, the thermal structure at the

centre of a Gulf Stream ring (Figure 4.9) is a result
ST

of the heating and cooling cycles to which it has

A
Sargasso Sea
B C been subjected8.
Gulf Sargasso Gulf Cold core
Slope water Stream Sea Stream ring

Warm core Mid-ocean eddies

Thermocline Away from the areas of high currents and strong
North–South sections
fronts, such as the Gulf Stream, things are no less
chaotic. In these quieter regions, meso-scale eddies
(b) dominate the flow. The ratio of the kinetic energy
of the eddies to the kinetic energy of the mean flow
m
Gu
lf S
trea
is typically of the order of 10, and often higher1. A
typical vertical section of density taken from the
northeast Atlantic using SeaSoar (Figure 19.6) is
shown in Figure 4.10(b). The tell-tale sign of the
Bermuda
presence of eddies is the undulating depth of sur-
0m
20

faces of constant density, implying variations in the

flow speed and direction on a horizontal scale of a
few tens of kilometres or less. Upwelling and
downwelling motions associated with the eddies,
Figure 4.10(a), can be as high as several tens of
metres a day, producing an enhanced exchange of
water and its properties between the surface and
80˚W 70˚ 60˚ 50˚
the deep. It is only recently that the technology was
developed to map out the density and velocity
structure on a fine-enough scale to resolve the
4.10 eddies at depth 7, and this is only to a depth of
around 500 m.
The surface signature of eddies can be viewed
from satellites using the sea surface height, sea sur-
face temperature, and ocean colour (Figure 4.11

Figure 4.10 A section along 20°W in the North Atlantic.

The plots show the vertical structure of both (a) water
density as a function of latitude and (b) chlorophyll con-
centration (a proxy for phytoplankton; see Chapter 6).
Large undulations are seen in the depth of surfaces of
constant density of several hundreds of metres. The
strong density gradient at 47.5°N marks the position of
the polar front, the boundary between the warmer sub-
tropical waters to the south and the colder polar waters
to the north. South of this are a number of eddy features
with a scale of tens of kilometres. The flow at these scales
is very much influenced by the earth’s rotation, so that in
the north hemisphere a density surface shallowing to the
north implies a flow to the east and likewise a deepening
implies a westward flow (relative to the flow deeper
down). Concurrent velocity measurements confirm this.
There are important large vertical movements of water
induced by the action of the eddies, as evidenced by the
filament of high chlorophyll reaching down to 400 m.
(This vertical movement must be have been relatively
swift as chlorophyll quickly degrades away from the sun-
lit surface waters.) (Courtesy of Mr G. Griffiths, SOC.)

64
 4: Ocean Weather – Eddies in the Sea

4.11a 4.11b

Figure 4.11 The use of biological tracers to examine eddying motions. (a) The false colour picture shows the level of
visible reflectance as viewed by a satellite using AVHRR (see Chapter 5). Orange and red show areas of high
reflectance, while blue and violet indicate low levels. The high reflectance levels are caused by the great abundance
of the phytoplankton species, Emiliania huxleyi, which produces highly reflective calcite coccoliths (observations from
ships report the sea becoming milky white during intense blooms of this species). (b) A smaller area of the same bloom
in more detail. Here the visible reflectance is shown using a grey scale and compared with the sea surface temperature
(the infra-red band), with a close correspondence between the two. The swirling patterns are very suggestive of eddy
motion. An animated sequence of satellite pictures has shown the eddies pulling out filaments of high and low con-
centration to produce the streaky appearance. The impression is that ocean eddies are efficient stirrers and mixers of
properties within the ocean. However, the sharp distinction between the various water masses is maintained for sever-
al days, and even enhanced by the eddies. This suggests that the final mixing of properties may take some time. (It
must also be remembered that the property being measured is related to a living organism that is growing and dying. It
is not simply a passive tracer.) The study of the effect of eddies on biological production in the ocean is an on-going
research issue. This particular bloom was found to have a significant impact on the levels of CO2 in the ocean and on
the production of the gas dimethyl sulphide, a possible agent in the making of rain4. (Courtesy of Mr S. Groom,
Plymouth Marine Laboratory, Plymouth, UK.)

and Box 5.1, Figures 5.12 and 5.13, for further And at smaller scales … ?
examples), giving us a large spacial coverage, very
impressive pictures of the eddying patterns and And at smaller scales of around 10 km or less there
their statistical properties, and certainly more detail are a host of structures. ‘Meddies’, ‘smeddies’, and
than can be gained using ships. However, the story ‘Beaufort eddies’ are a few of the names given to a
does not end there. The relationship between the class of animals in the eddy zoo known as sub-
sea surface fields and the eddy structure at depth is meso-scale coherent vortices, or SCVs.
a subject of on-going research. Through the use of They are lens-like structures that are small in
mathematical models we know that the behaviour horizontal extent (a few kilometres), have limited
of the eddies is very dependent on the way the flow vertical extent (typically a few hundred metres),
and density vary in the vertical. We know through and, in the vast majority of cases, have a circula-
experiments such as Polymode that interactions tion in an anticyclonic (clockwise in the northern
between the eddies can produce rapid changes to hemisphere) sense. SCVs are identified by a bulging
the flow conditions, to form intense jets or more of density surfaces and a core of water with anom-
quiescent plumes. However, subsurface measure- alous properties compared with its surroundings,
ments are limited in time and space. We are at an which may have originated several thousand kilo-
early stage in characterizing the eddy structures in metres away.
the open ocean. The SCVs that have been studied the most are

65
 K.J. Richards and W.J. Gould

35˚
4.12 N
MEDDY 2 4.13
Nov 85 m
00
40 MADEIRA
IS.

2000 m Oct 84
4000 m MEDDY 1 2000 m
First Cruise

Hyères
July 86 Day 20
Smt. m
00
2000 m 40
30˚
4000 m
June 85 0m
Second Cruise 200

CANARY
IS.
Oct 85
Third Cruise

m
00
20
25˚
Oct 85

MEDDY 3

July 86
Oct 86
Fourth Cruise Day 30

20˚ Mar 87

30˚W 25˚ 20˚ 15˚

Figure 4.12 Intense ‘blobs’ of Mediterranean water are

formed near Cape St Vincent and keep their rapid rota-
tion and distinct temperature and salinity characteristics
for years as they propagate into the ocean interior11. The
tracks of SOFAR floats that were seeded into ‘meddies’
demonstrate both the rapid rotation and the southward Day 60
drift over 2.5 years.

the meddies (Figure 4.12). These are lenses of water

that originate in the outflow of the Mediterranean
and propagate into the North Atlantic, carrying the
relatively warm and salty Mediterranean water Figure 4.13 Example of an unstable flow. The three pan-
els show the evolution of a jet set-up in a channel. The
with them. They can last for a long time. One has channel is 500 km wide, 1000 km long, and 2000 m
been tracked11 for a period of 2 years. deep and oriented in an east–west direction. The results
SCVs are abundant in the ocean. Long-lived shown are from a numerical model which simulates flow
lenses of water transporting their chemical compo- in the ocean. On the right is a typical cross-section of the
sition over large distances may be responsible for a surface flow down the channel. On the left is the depth
significant flux of water properties. of a density surface. Initially, the jet is in the centre of the
channel. Because the flow is strongly constrained by the
How do Eddies Originate? earth’s rotation, there is a corresponding slope in the sur-
A wide range of fluid flows cannot remain smooth or faces of constant density, as seen by the shallowing of the
particular density surface shown from 900 m on the
unchanging in form and are intrinsically ‘unstable’; southern side of the channel to 100 m on the northern
witness the break-up into drops of a jet of water from side. At day 0 the jet is perturbed by a random distur-
a tap or the turbulent, chaotic nature of the flow in a bance with a wide range of horizontal scales. These per-
stream or river. Small disturbances to the flow are turbations grow with a growth rate dependent on the
amplified, interact, and produce very irregular scale of the perturbation. By day 20 the fastest growing
motions which totally change the nature of the flow. disturbance, which has a wavelength of 125 km,
The flows at larger scales in the atmosphere and becomes apparent in the depth of the density surface. By
ocean are no exception. Currents that are sheared in day 30 there has been significant growth. The velocity
either the vertical or horizontal may be susceptible to cross-section shows a tightening up of the core of the jet.
small disturbances that trigger an explosive growth By day 60 the density surface has become very distorted
and the jet has been replaced by a chaotic eddy field. As
of eddying motions. An example of such an instabili- the instability grows there is a general slumping of the
ty is shown in Figure 4.13. For a full description of density surface. The potential energy associated with the
flow instability, and a good guide to ocean dynamics initial displacement of the density surface has been con-
in general, see Gill3 and Pedlosky6. verted into the kinetic energy of the eddies.

66
 4: Ocean Weather – Eddies in the Sea

In an unstable flow, the energy for the eddying able forcing by fluctuating winds. Eddy energy may
motions can be extracted from either the potential be exported from a region where it is generated by
or kinetic energy in the system, or from both. In the transport, by the flow, or by radiative mechanisms.
case of the large-scale flows of the atmosphere and Eddies interact with each other, waves, and the mean
ocean, if the source of the eddy energy is from the flow. There are several proposed generation mecha-
(time) mean potential energy the instability is nisms for sub-meso-scale coherent vortices – current
referred to as a baroclinic instability; if the eddy instability, flow over topographic features, vortex
energy comes from the mean kinetic energy the concentration in geostrophic turbulence, and mixing
instability is known as a barotropic instability. and adjustment of the density field.
Although there are some well-studied particular The mechanism for the ultimate demise of eddies
examples of flows that are baroclinically or is not well-understood, but is probably a conse-
barotropically unstable, and some known necessary quence of bottom friction (if the eddy extends to
conditions for instability, the stability of a given the bottom), loss of energy by the radiation of
flow regime, in general, is not known and we have large-scale waves, or dissipation by smaller scale
to resort to experiments with numerical models motions. Isolated eddies may decay through hori-
(such as that shown in Figure 4.13) to investigate zontal interleaving with the surrounding water or
the evolution of the flow. may be scattered and broken up over topographic
The growth rate of an unstable perturbation is features. We still have much to learn about eddies
dependent on the horizontal and vertical structure in the ocean, their dynamics, and the controlling
of the current, as well as on the vertical gradient of factors for their growth and decay.
density. In regions of highly sheared flow, such as
the Gulf Stream, the time-scale for the development Are Eddies Important?
of meanders and the pinching off of warm and cold At any one location in the ocean the day-to-day
core rings is a few days. In the centre of ocean variation in current speed, temperature, and salini-
basins, away from swift currents and fronts, the ty is very much dictated by the evolving eddy field.
time-scale for the development of eddies may be Predicting the state of the ocean weather is useful
around 100 days. The disturbance with the fastest for some activities, such as commercial fishing
growth rate has a wavelength of around 4–6 times a where particular fish stocks are known to favour
length-scale called the Rossby radius of deformation certain types of eddy and their associated fronts.
(named after the famous Swedish oceanographer, However, to an oceanographer trying to measure
Carl-Gustaf Rossby). From numerical experiments large-scale gradients in the ocean, or the flow aver-
it is found that this length-scale dominates in a fully aged over a long period, eddies are a nuisance. The
developed eddying flow. The Rossby radius is a fun- overwhelming effect of the eddy field masks the
damental length-scale of flows on our rotating plan- weaker signal of the larger-scale circulation of the
et, and is the scale at which rotational forces acting ocean. Time series of currents at many locations in
on the fluid motion become comparable to buoyan- the ocean have to be of several years duration in
cy forces. It depends on the rotation rate of the order to obtain a reliable measure of the mean
earth, the latitude of the motion, the vertical density flow.
gradient, and the depth of the fluid. The latitudinal But are eddies more than just a nuisance? Are
dependence is such that the length-scale decreases they important in shaping the circulation of the
with distance away from the equator. In the ocean, ocean or in the oceanic transport of heat? The
the equatorial Rossby radius is around 200 km. At answer to many such questions is that we simply
mid-latitudes the Rossby radius is around 30 km, do not know. For the atmosphere, the intimate
decreasing to less than 10 km in Arctic and links between weather and longer-term changes in
Antarctic waters. Meso-scale eddies in the ocean are the state of the atmosphere, or climate, are well-
found to have a similar scale to the Rossby radius, established. We have yet to establish the link
with the same latitudinal dependence. In the atmos- between ocean weather and the climate of the
phere, the Rossby radius is closer to 1000 km, the ocean.
scale of the atmospheric weather systems. The The major difficulty in assessing the role of
eddies in the ocean and atmosphere, with their very eddies is the problem of sampling the eddy field.
different scales of motion, not only have a similar We do know that eddies are effective at dispersing
appearance, but also are dynamically equivalent. and mixing properties in the ocean (for instance,
Thus, the term ‘ocean weather’ is a particularly apt see Figures 4.6 and 4.11), or that they do affect the
description of ocean eddies. distributions of plants and animals in the ocean
Not all meso-scale eddies in the ocean originate (Figure 4.10) and the exchanges between the
from unstable currents. Other generating mechanisms atmosphere and the ocean. An important aspect of
include shedding from topographic features and vari- eddy mixing is the homogenisation of potential

67
K.J. Richards and W.J. Gould

vorticity, a dynamical property which dictates the References

behaviour of ocean currents12. Eddies also transport 1. Dickson, R.R. (1983), Global summaries and intercompar-
heat (a striking example is the Agulhas Current isons: flow statistics from long-term current meter
eddies which transport heat from the Indian Ocean moorings, in Eddies in Marine Science, Robinson,
to the Atlantic around the tip of South Africa; see A.R. (ed.), Springer Verlag, Berlin.
2. Fofonoff, N.P. and Webster, F. (1971), Current measure-
Chapter 3, Figures 3.19 and 3.20). ments in the Western Atlantic, Phil. Trans. R. Soc. A,
Quantifying the effect of eddies is very difficult. 270, 423–436.
Often we have to resort to mathematical models of 3. Gill, A.E. (1982), Atmosphere–Ocean Dynamics, Academic
the ocean to discover what eddy processes may be Press, New York, 662 pp.
4. Holligan, P.M., et al. (1993), A biogeochemical study of the
important. From these models we have learnt much coccolithophore, Emiliania huxleyi, in the North
about how eddies interact and affect the mean Atlantic, Global Biogeochem. Cycles, 7, 879–900.
flow. For instance, eddy–eddy interactions can pro- 5. Kort, V.G. and Samoilenko, V.S. (1983), Atlantic
duce intense vortices; eddy–mean flow interactions Hydrophysical Polygon-70: Meteorological and
can tighten a jet and split a jet into a number of Hydrophysical Investigations, Amerind Publishing
Company, New Delhi, for NSF/IDOE, 398 pp.
multiple jets. It is only very recently that computer 6. Pedlosky, J. (1987), Geophysical Fluid Dynamics, Springer-
technology has advanced sufficiently that meso- Verlag, New York, 710 pp.
scale eddies can be represented explicitly in models 7. Pollard, R.T. and Reiger, L. (1990), Large potential vorticity
of the circulation in ocean basins or in the entire variations at small scales in the upper ocean, Nature,
348, 227–229.
world’s ocean (see, for example, Semtner and 8. Richardson, P.L. (1983), Gulf stream rings, in Eddies in
Chervin16). Whether or not eddies play a major role Marine Science, Robinson, A.R. (ed.), Springer
in the oceanic conveyor belt (Chapter 3) is still Verlag, Berlin, pp 19–45.
unclear. 9. Richardson, P.L. (1985), Drifting derelicts in the North
Atlantic 1883–1902, Prog. Oceanogr., 14, 463–483.
Epilogue 10. Richardson, P.L. (1993), A census of eddies observed in the
North Atlantic SOFAR float data, Prog. Oceanogr.,
The future will bring an increase in the power of 31(1), 1–50.
computers, allowing more detailed and more realis- 11. Richardson, P.L., Armi, L., Price, J.F., Walsh, D., and
tic models of the world’s oceans to be developed. Schroter, M. (1989), Tracking 3 meddies with
SOFAR floats, J. Phys. Oceanogr., 19(3), 371–383.
We are becoming more skilled in the use of mea- 12. Rhines, P.B. and Young, W.R. (1982), Homogenization of
surements from space to tell us about the dynamics potential vorticity in planetary gyres, J. Fluid
of the oceans. New instruments are being devel- Mechan., 122, 347–367.
oped to measure not only the physical properties of 13. Robinson, A.R. (ed.) (1983), Eddies in Marine Science,
Springer Verlag, Berlin, 609 pp.
the ocean, but also biological and chemical proper- 14. The MODE-1 Atlas Group (1977), Atlas of the Mid-Ocean
ties in greater detail. With new technology and Dynamics Experiment (MODE-1), MIT, Boston, 274
insight we will learn more about ocean eddies and pp.
their effect on the ocean’s climate, biology, and 15. The MODE Group (1978), The Mid-Ocean Dynamics
chemistry. The study of the weather of the oceans Experiment, Deep-Sea Res., 25, 859–910.
16. Semtner, A.J. and Chervin, R.M. (1988), A simulation of the
has brought together theoreticians, satellite global ocean circulation with resolved eddies, J.
oceanography, and those who make measurements Geophys. Res., 93, 15502–15522.
at sea. This collaboration will continue well into 17. Stommel, H. and Arons, A.B. (1960), On the abyssal circu-
the future. lation of the world ocean, Deep-Sea Res., 7, 140–154;
217–233.
18. Swallow, J.C. (1955), A neutral-buoyancy float for measur-
ing deep currents, Deep-Sea Res., 3, 74–81.
General References 19. Swallow, J.C. (1971), The Aries current measurements in
Gill, A.E. (1982), Atmosphere–Ocean Dynamics, Academic the Western North Atlantic, Phil. Trans. R. Soc. A,
Press, New York, 662 pp. 270, 460–470.
Pedlosky, J. (1987), Geophysical Fluid Dynamics, Springer- 20. Zenk, W. and Muller, T.J. (1988), Seven-year current meter
Verlag, New York, 710 pp. record in the eastern North Atlantic, Deep-Sea Res.,
Robinson, A.R. (ed.) (1983), Eddies in Marine Science, Springer 35A(8), 1259–1268.
Verlag, Berlin, 609 pp.

68
CHAPTER 5:

Observing Oceans from Space

I.S. Robinson and T. Guymer

Introduction – A New Way of The scientific understanding of ocean processes

Viewing the Ocean depends upon being able to measure the wide vari-
At first consideration it may seem surprising to find ety of variable parameters which describe the sea.
a chapter on space in a book about oceanography. Until recently, oceanographers could make mea-
But it is not so surprising when we recall that it surements only from ships or buoys (see Chapter
was the lunar astronauts’ first pictures of Earth, the 19), and were consequently limited by the sampling
beautiful blue and green planet seen from a barren, constraints of an Earth-bound perspective. The
dead moonscape, which awakened a new aware- vantage point from aloft, which Earth-orbiting
ness of the fragility of the Earth’s natural environ- satellites provide, has enabled spatially detailed
ment. It took a visit to our nearest neighbour in measurements (using, e.g., the electro-magnetic
space to emphasise how important it is to under- spectrum, Figure 5.1) to be made almost instanta-
stand the global oceanic and atmospheric environ- neously over wide areas and provide a novel per-
ment on which we depend for our continued exis- spective of the ocean. It has also made possible the
tence. That global view and wide spatial perspec- regular, repeated monitoring of the ocean on a
tive eventually led to a new branch of oceanogra- global scale, and the detection of some of the
phy. In this chapter we examine the basic principles changes of climate discussed in Chapter 3.
of how the oceans can be measured remotely from ‘Satellite oceanography’ had its birth as a scien-
satellites (and also from aircraft), explain some of tific endeavour in the 1960s, when the first astro-
the techniques which are used by particular sen- nauts in orbit around the Earth noticed features in
sors, and present some illustrative examples of the sea which were so much easier to discern from
image data and their oceanographic interpretation. above than from ground level. Soon satellites were
1 GHz 1 MHz 1 kHz 5.1

1018 1015 1012 109 106 103

Frequency (Hz) Radio and TV transmission
Data telemetry
Active microwave
(radars)
Spectral windows
used for remote
sensing
Microwave emission

Thermal infra-red
Near infra-red
Figure 5.1 The electromagnetic
Visible
spectrum showing the variation 100%
with wavelength of atmospheric
transmission and the spectral Transmittance
windows used for remote sens- through the
atmosphere
ing. Note that microwaves are
usually defined in terms of fre- 0%
quency (upper scale), whereas 1 pm 1 nm 1 +m 1 mm 1m 1 km
visible and infra-red radiation are
generally referred to by wave- 10–12 10–9 10–6 10–3 1 103
length (lower scale). Wavelength (m)

69
I.S. Robinson and T. Guymer

5.2
PASSIVE SENSORS ACTIVE

Visible Infra-red Microwave

Sensor waveband Radar
class sensors sensors instruments
sensors

Sensor Multi-spectral
scanners Infra-red Scanning Scatterometer
type imaging microwave Imaging radar
Imaging radiometers radiometers
spectrometers Altimeter

Primary Ocean Sea surface Surface Surface

measure colour temperature roughness height/slope

Chlorophyll Surface winds Geostrophic

Mixed-layer
Derived temperature Wave height currents
Suspended
parameters particulates Wave spectra Ocean geoid
Skin
temperature Internal waves
Bathymetry Sea-floor
Surface slicks
bathymetry

Figure 5.2 Schematic showing the different classes of sensors used in satellite
oceanography, the different types of instrument, the primary measurement which each
class can make, and examples of the ocean parameters which can be derived from
these measurements.

flown specifically for the purpose of Earth observa- sor types designed with the object of measuring
tion, and by 1978 sensor technology had developed one or more particular ocean parameter (see
to the stage where oceanographers could begin to Figure 5.2). Sensors may be either passive or
make scientific measurements of the sea. In that active. Passive sensors measure naturally occurring
year three satellites (Seasat, Tiros-N, and Nimbus- radiation, using either the Sun’s rays scattered
7) were launched with sensors capable of recording from below the sea surface, or energy emitted by
a variety of ocean processes, and since then the use the sea itself in the infra-red or microwave parts of
of satellite data has become widely accepted in the spectrum. Active sensors produce their own
marine science. source of electromagnetic energy which is emitted
toward the sea and measured after reflection.
Remote Sensing Methods Sensors that rely on solar illumination, like the
human eye, can operate only in the daytime, when
How can we observe the ocean from space? the Sun is not too close to the horizon, but the
Remote sensing relies on electromagnetic radiation other sensors can operate both day and night.
to convey information about the sea to a sensor on Active sensors deployed on satellites are restricted
a satellite or aircraft. Most electromagnetic wave- to microwaves, although visible-wavelength lasers
lengths are absorbed or scattered by the atmos- have been used from aircraft.
phere, and so are of no use for this purpose, but The instrument designer can choose the particu-
there are three distinct wavebands, ‘spectral win- lar electromagnetic property which a sensor will
dows’, at which rays can penetrate the atmosphere measure. This may simply be the magnitude of
with less interference (see Figure 5.1). These are the radiation at a particular wavelength, or the relative
visible waveband (400–700 nm), parts of the infra- magnitude at several wavebands. It may also relate
red (the regions around 3.7 mm and 10–13 mm are to a particular polarisation state of the radiation.
used), and microwaves, which include radar, longer Active sensors provide more options because they
than about 10 mm. Visible and infra-red light can- can measure not only the magnitude of the
not penetrate clouds without being scattered or returned pulse relative to that emitted, but also its
absorbed, but microwaves are much less affected timing, the shape of the pulse, and any frequency
and permit all-weather monitoring of the sea. shift. For a sensor to be useful for a particular
As well as selecting the spectral window, the application requires that the measured radiation be
remote sensing scientist can use a number of sen- influenced in some way by the ocean parameter

70
 5: Observing Oceans from Space

(a) (b) 5.3

Infra-red
Visible waveband radiometer Microwave
multi-spectral radiometer
radiometer

sunlight scattered thermal emission

from below the surface (infra-red and
microwave)

(c) (d)

Imaging
Satellite orbit radar
Radar
altimeter

backscattered
microwaves
reflected
microwaves

Figure 5.3 Depiction of the four primary measurements of the sea which can be made in
satellite oceanography, and an indication of how they are made. (a) Ocean colour; (b) sea sur-
face temperature; (c) sea surface slope (length scale of tens of kilometres); (d) sea surface
roughness (length scale of 10 mm to 10 m).

being monitored. The influence can be direct, as surface roughness can be interpreted in terms of
with infra-red radiation emitted by the sea surface wind speed. Some more details about how oceano-
according to its temperature. It can be more com- graphic information is extracted from these prima-
plex, as in the way the spectral composition, or ry measurements are given in the examples in
colour, of sunlight scattered from below the surface Boxes 5.1–5.6 which describe different applications
is influenced by both the chlorophyll and particu- of remote sensing.
late content of the water (see Chapter 14). It may Remote sensing has many limitations. There are
be more subtle, as in the way the sea state changes many important aspects of the ocean that cannot
the shape of the reflected pulse of a nadir-viewing be detected remotely, such as salinity, at least with
radar. After two decades of experience a number of the present technology and knowledge. Remote
useful oceanographic instruments have emerged sensing cannot penetrate far below the sea surface
from the wide range of potential design options. and cannot tell us much about how the ocean prop-
What can we measure from space? erties and composition change with depth. Like any
measurement tool, we should not misapply it, but
Despite the variety of instruments, remote sensing use it for what it can do well, and be aware of its
is able to measure just four basic properties of the particular characteristics.
sea; the surface temperature, the colour of the near-
surface waters, the surface roughness (at short
length-scales), and the slope of the surface averaged
Sampling Capabilities of Remote Sensing
over tens of kilometres (Figure 5.3). It is from these The greatest advantage of remote sensing lies in its
primary measurements that a range of other prop- ability to provide measurements of ocean parameters
erties can be derived, depending on the detailed over a wide sea area. The size of the area that can be
design of the sensor. For example, given measure- viewed almost simultaneously, the spatial resolution
ments of the appropriate wavebands, colour can of the resulting image data, and the frequency with
yield an estimate of the chlorophyll content and which it can be revisited depend on the type of sen-
hence the primary production in the upper ocean sor and the platform carrying the instrument.
(see Chapter 6). From the detection of surface slope Sensors record individual point measurements of
can be derived a measure of ocean currents, while a property of the ocean corresponding to an aver-

71
 I.S. Robinson and T. Guymer

5.4 Figure 5.4 Earth viewing by a non-

Orbit path
scanning sensor. The IFOV is the
area of ground observed at any
instant by the sensor, but the record-
ed measurement is representative of
Ground track the footprint covered by the IFOV
during the instrument integration
time.

Instantaneous field
of view (IFOV)
Footprint

Distance moved by IFOV along

ground track during sensor
integration time

age over the instantaneous field of view (IFOV), samples along the track illustrated in Figure 5.5.
which may be as small as a few metres or as Over many days, given a suitable orbit, the gaps
large as hundreds of kilometres, depending on between the single-day paths can be filled, provid-
the sensor. Some sensors, such as an altimeter, ing a fully global picture of the oceans.
simply view in a single direction, following the If the sensor is one which scans several times per
ground-track beneath the moving satellite or air- second across the direction of travel, however, an
craft (Figure 5.4). Mounted on a polar orbiting array of data is collected almost simultaneously
satellite such sensors gradually build up coverage which corresponds to an image (Figure 5.6) of the
of the Earth’s surface. There are typically 15 ocean surface, composed of individual picture ele-
orbits per day during which time the Earth ments, or pixels, each of which is an independent
rotates underneath the sensor, which therefore measurement. Careful specification of the sampling

5.5

Figure 5.5 The Earth coverage of a typical near-polar satellite orbit during one day. Orbit one is shown in blue,
descending in local daytime and ascending in local night. This leads on to orbit two (shown in red). For this and
the rest of the orbits (in black) that day, only the daytime pass is shown. Note how the spacing between ground
tracks is closer toward the poles.

72
 5: Observing Oceans from Space

rate and scan rate relative to the satellite speed can slowly so that the orbit plane remains fixed relative
ensure that each IFOV is contiguous with its neigh- to the Sun, despite the seasonal changes of the
bour, and then sampling is at its most efficient. The Sun’s position relative to the fixed stars. Such a
swath width depends on the spatial resolution ‘Sun-synchronous’ orbit ensures that the local
required. For a small IFOV to give high spatial res- (solar) time is always the same whenever the satel-
olution, the swath must be narrower than if the res- lite crosses a given latitude, no matter what the lon-
olution is to be coarser. If the swath is wide gitude. This ensures a uniformity of solar illumina-
enough, the coverage of successive orbits overlaps tion for ocean colour scanners, and helps to elimi-
and coverage of the whole Earth can be achieved nate problems of the diurnal heating cycle. The
daily. Thus, there is a trade-off between frequency drawback of the low polar orbit is the difficulty of
of coverage and spatial resolution. Colour scanners covering the Earth more frequently than once per
and imaging radars with a resolution of about day (or twice if night-time images can be used), and
30 m may take 15–25 days to revisit every point on the necessity of viewing the sea surface very
the Earth, whereas the medium resolution sensors obliquely at the swath extremities if daily coverage
of about 1 km resolution can view the Earth once is required.
every one or two days. The sampling capability of The only other type of satellite used for remote
the sensor–satellite combination must be matched sensing of the sea is that in geostationary orbit.
as carefully as possible to the characteristics of the Parked at a height of about 36000 km above the
oceanographic phenomenon to be observed. Earth, this orbits once per day, so that if it is placed
Earth observation satellites in near-polar orbit over the equator it remains stationary relative to an
fly at an altitude of about 700–1000 km above the observer rotating with the Earth. From this vantage
Earth. The orbit is normally arranged to precess point a sensor can scan the whole of the visible disc

5.6
Orbit

Ground track

Scan lines

IFOV

Swath
width

Figure 5.6 An imaging sensor scans across the direction of travel. If the scan rate is matched to the IFOV
and synchronised with the speed of the satellite over the ground, the IFOVs are contiguous and the
ground area is covered efficiently.

73
 I.S. Robinson and T. Guymer

5.7

Figure 5.7 The Earth coverage achieved by geostationary meteorological satellites. Locations of the five
satellites which span the equator are shown as plus signs. Although the horizon of the Earth-view from
geostationary satellites is about 60°, at latitudes greater than 50° the viewing angle is too oblique to be
useful, as indicated by the limit lines.

Figure 5.8 Space–time sam- 5.8

pling characteristics of major
classes of sensor. The capability 10000 Meteosat 108
AVHRR
of each sensor is defined as a
box on the diagram. The left 1000 106
CZCS
boundary represents the sam-
pling frequency; it defines the ATSR
shortest revisit period (assuming 100 104
no cloud cover) and depends on
Length scale

Area scale
a combination of the orbit path 10 102

(km2)
and swath width. The bottom
(km)

boundary represents the spatial

resolution of the sensor and 1 1
defines the length (left scale) or
area (right scale) of a picture 0.1 Landsat 10–2
element (pixel). The top bound- TM
ary indicates the largest length-
scale that can be observed near- 0.01 ERS-1 SAR 10–4
instantaneously, as controlled Airborne
by the swath width. The right scanners
0.001
boundary depicts the span of
time over which image 0.01 0.1 1 10 100 103 104
sequences are obtained, limited Time scale (days)
only by the lifetime of the sensor 1 10 years
or series of sensors. ATSR, The
Along-Track Scanning Radiometer, an infra-red sensor, flown on the ERS-1 satellite, to measure sea surface temperature.
AVHRR, The Advanced Very High Resolution Radiometer, flown on the NOAA meteorological satellite series, to
observe clouds and land surfaces (visible channels), and the temperature of the sea and cloud tops (infra-red channels).
CZCS, The Coastal Zone Colour Scanner, flown in 1978–1986 by NASA to measure ocean colour in four narrow-band
spectral channels. ERS-1 SAR, The Synthetic Aperture Radar on the European Space Agency’s first remote sensing satel-
lite (ERS-1), launched in 1991. Landsat TM, The Thematic Mapper, a high-resolution multichannel visible wavelength
sensor mainly for land applications, flown on the US Landsat satellite series. Meteosat, The European geostationary
meteorological satellite (one of five which by international agreement should be in place around the equator) observes
clouds and land, sea, and cloud temperatures using visible and infra-red radiometers.

74
 5: Observing Oceans from Space

Figure 5.9 Out- Required input Procedure Explanation 5.9

line of the proce-
dures required
for processing Digital data
data received from satellite
from a satellite
into oceano- Converts the numbers transmitted
graphically useful
Pre-launch calibration data Sensor from the satellite into a measure of
parameters and In-flight calibration calibration the e/m radiation reaching the sensor.
image arrays.

Multi-channel data
Makes allowance for the effect of the
Atmospheric atmosphere on the remote sensing
signal and estimates the e/m radiation
Atmospheric models correction which actually left the sea surface.

Multi-channel data
Converts the water-leaving e/m radiation
Calibration algorithms
Geophysical into one of the primary ocean variables
and then, if possible, into a derived
calibration oceanographic parameter.
In-situ validation data

Satellite orbit ephemeris The geographical location of each

Geolocation pixel on the image is defined.
If required the image is re-sampled
Ground-control points onto a standard map base.

Calibrated image
OUTPUT of ocean parameter

of the Earth (Figure 5.7), and can sample at any fre- the atmosphere on the transmission of the ocean
quency. Present meteorological sensors scan the full information from sea to satellite. The atmospheri-
view once in 30 minutes, but spatial resolution is no cally corrected value is an estimate of the radiation
better than 5 km, and generally worse because of the which the sensor would have detected leaving the
oblique view. Such sensors cannot view high latitudes surface of the sea if there were no intervening
above 55º because of the curvature of the Earth. atmosphere. A variety of approaches are used for
The sampling characteristics of a number of this, including modelling of the atmospheric effects,
oceanographically useful sensors are depicted in and using multispectral methods to correct for vari-
Figure 5.8. able factors which cannot easily be modelled. For
data derived in infra-red and visible wavebands,
Procedures for Analysing Satellite Data screening for cloud detection is applied, in which
Before the raw data transmitted from the satellite all pixels corrupted by cloud are rejected, leaving
to the ground station can be of use to the marine blanks in the final image.
scientist, a number of operations have to be per- While for some applications the value obtained
formed. Figure 5.9 outlines the various procedures after atmospheric correction may be a useful para-
involved in acquiring satellite data and processing meter, e.g., the sea surface temperature (SST), in
them to yield oceanographic information. Data are many cases it is appropriate to apply what is
recovered from the satellite as a sequence of num- termed a geophysical calibration to convert the
bers corresponding to the digitized signal every measured properties into an oceanographically use-
time it is sampled. Given a knowledge of the detec- ful parameter, e.g., converting the radiance mea-
tor characteristics, it is possible to apply a sensor sured in several bands within the visible spectrum
calibration to the digital signal in terms of the into an estimate of chlorophyll concentration, or
property of radiation being measured at the satel- converting radar back-scatter into wind speed and
lite. The digitization capacity, i.e., whether the val- direction. Calibration algorithms are generally
ues are stored as 6-, 8-, 10-, or 16-bit numbers, derived empirically using in situ data and calibra-
determines the resolution and therefore can strong- tion or validation experiments.
ly affect the usefulness of the data. Another essential requirement is to identify the
An atmospheric correction also has to be applied geographical position of every data value in a
in most cases. This attempts to remove the effect of geolocation procedure. If image data are to be pre-

75
I.S. Robinson and T. Guymer

Box 5.1 Sea Surface Temperature Maps from Infra-Red Sensors

Figure 5.10 Infra-red sen- 5.10
sors measure the radiation
emitted from the sea
surface in the wavebands Satellite track
3.5–3.9 μm, 10.3–11.3 μm,
Forward view has twice
and 11.5–12.5 μm. The the atmospheric path Nadir view has the
emitted radiation increases length of the shortest path through
with the sea surface temper- nadir view the atmosphere
ature (SST), but is reduced
by passing through the
atmosphere, so the detected Absorption Atmosphere
(brightness) temperature is
not the true value. The
atmospheric effect varies Thermal emission from sea surface
with wavelength, and so the
difference between bright-
ness temperatures at differ-
ent wavebands can be used as the basis for an atmospheric correction. The ATSR sensor also observes each sea
area twice, once from above and once obliquely through twice the atmospheric path length. As the difference in
temperature between the two views gives a direct measure of the atmospheric effect this approach gives a high
absolute accuracy of SST recovery.

Figure 5.11 This image 5.11

depicts SST in the Bay of
Biscay and English Channel,
derived from measurements
made by the AVHRR sensor
on the NOAA-9 satellite on
6 August, 1988. The tem-
perature has been colour-
coded according to the
scale bar displayed. Black
corresponds to the land and
white indicates that cloud
has been detected. There is
a wealth of oceanographic
information to be found in
this image, including the
overall distribution of tem-
perature, showing a drop of
9°C between northern Spain
and the English south coast,
the discontinuous nature of
the thermal structure rather
than smooth gradients, and
the evidence of meso-scale
eddies in the patterns drawn
out in the temperature field.
There appears to be cool
coastal upwelling along the
French west coast. Off the
Brittany peninsula, the dis-
tinctive cold front, on which small-scale eddies have developed, is probably the boundary between offshore strat-
ified water with a warmer surface layer and the tidally well-mixed water inshore. Further offshore from Brittany,
and disappearing under the cloud, is evidence of cool surface temperatures above the shelf edge, caused by
breaking internal waves stirring cooler water into the mixed layer.

76
 5: Observing Oceans from Space

Figure 5.12 This brightness temperature image of 5.12

a region 500 km2 in the South Atlantic, just north
of the Falkland Islands, was obtained from the
ATSR 11 μm channel on 13 October, 1992. It
illustrates the richness of the dynamical structures
which can be found in the ocean and demon-
strates the unique capability of remote sensing to
capture them. The image is from a region centred
about 55°W 45°S. The coldest water comes from
the Falkland current, which flows from the south-
west. The warmest water derives from the Brazil
current flowing from the north. As the two currents
interact, large eddying motions occur, entraining
warm water into the cold and vice versa. What is
striking from the resolution achieved with this sen-
sor is the narrowness of some of the warm and
cool filaments drawn out in a complex interleav-
ing, leading in some places to the growth of small
eddies. Eventually, these will diffuse into a uni-
form body of intermediate temperature, but this
particular instant within a rapidly evolving pattern
has been captured by the satellite in a way that
could not have been possible using ships. Remote
sensing is providing new challenges for the study
of the dynamical processes which control these
temperature patterns.

Figure 5.13 Individual high-resolution 5.13a

images are used to contribute to global
compilations of sea surface temperature.
(a) The contribution of just one day’s over-
passes of the ATSR. The regular gaps show
the limits of daily coverage; the gaps with-
in the swath are due to cloud being
encountered. Over a month the swath
covers most of the globe, and very few
locations remain persistently cloud-cov-
ered (black on the image), leading to the
monthly average SST map (b). This is
based on a resolution cell of 0.5° latitude
and longitude. It reveals the global struc-
ture of SST, warmest at low latitudes and
coldest toward the poles. However, the 5.13b
east–west asymmetry in the oceans is
apparent, particularly in the Atlantic. Also
evident is the characteristic cool zone
along the equator in the Pacific, and the
cool upwelling off the east coasts of Africa
and South America. Despite its coarse res-
olution, an image like this still shows eddy
activity (see Chapter 4), particularly in the
Atlantic sector of the Antarctic
Circumpolar Current.

77
I.S. Robinson and T. Guymer

sented in the form of a regular map projection, geo- on the scale of ocean basins to reveal clearly how
metric correction has to be applied. This may the primary production in the upper ocean varies
require resampling of the image onto a different spatially and seasonally. At the much higher spatial
grid base; since this degrades the data slightly it is resolutions achievable from aircraft, it is possible to
usually performed only when all the other correc- detect small-scale dynamical processes occurring in
tions have been applied. For many purposes it is coastal seas and estuaries, and to use this to under-
convenient to combine the data from several differ- stand the processes which control coastal pollution
ent satellite overpasses into a single map represent- and the dispersion of river discharges.
ing average conditions over a given time span, a Images derived from synthetic aperture radar or
procedure known as compositing. SAR (Box 5.3) represent the variation of radar back-
For example, a global map of sea surface temper- scatter over the sea. They are not so readily inter-
ature is built up from the cloud-free pixels derived preted as colour or thermal images, but the spatial
from all the overpasses on several days (see Box patterns often provide clues for interpretation. On
5.1). This enables gaps due to cloud on some days to some images, swell waves are clearly shown by the
be filled from clear overpasses on others, although way in which back-scatter varies between trough,
care has to be taken if the ocean variable is likely to crest, and leading and trailing face. Sometimes the
vary rapidly within the compositing period. refraction of wave patterns is imaged. Surprisingly,
since the radar is influenced only by the surface
Ocean Applications of Satellite Data roughness, patterns appear which are associated
The routine processing outlined above generally with phenomena beneath the water surface, such as
leads to images or maps of a particular ocean vari- sea-bed bathymetry, or the presence of undulations
able. Although much challenging scientific work of density layers known as internal waves. In other
has been done developing these procedures to the images the radar is strongly influenced by the pres-
point where satellite measurements are scientifically ence of surface films which generate slicks of
reliable and have a specified accuracy, this aspect smoother water having low back-scatter. The spatial
of satellite oceanography is just the prelude to the coverage of the satellite SAR over 100 km can reveal
main task which is to analyse and interpret the the coupling between local roughness phenomena
data. Here, the exciting scientific opportunity is to and meso-scale dynamical features.
use the unique capabilities of remote sensing to Altimeters do not generate images like scanning
extend our understanding of ocean processes. sensors, although over time they can build up evi-
Satellite data potentially have a contribution to dence of sea surface topography which can be
make to all aspects of marine science, but here we interpreted in terms of ocean currents (Box 5.4).
have room only to look at a few examples. Until there is an independent measure of the shape
In Box 5.1 the measurement of SST using infra- of the geoid due to the variation of gravity, only
red sensors is described. One of the primary objec- deviations of slope from the mean can be obtained,
tives in developing these sensors has been to but these can reveal the variability in ocean cur-
acquire regular global maps of SST distribution and rents and map the global distribution of eddy kinet-
its seasonal variation. Once a long-enough time ic energy (EKE, see Chapter 4). The altimeter’s
sequence has been established (useful records com- ability to estimate wave height from the shape of
menced in 1980), it will be possible to identify the reflected radar pulse (Box 5.5) gives it an
anomalous trends which may herald changes in the important operational role in sea state forecasting.
patterns of global climate. Already such data pro- The scatterometer (Box 5.6) measures the aver-
vide early indicators of the onset of irregular cli- age roughness over large areas of the sea surface
mate phenomena, such as the El Niño–Southern and enables the local wind speed and direction to
Oscillation (ENSO) events (see Chapter 3). From be derived. While the direct application of this is in
the same sensors, but analysing data at higher spa- meteorology, it is valuable for indicating the way in
tial resolution, maps of SST provide synoptic views which the dynamical forcing of the ocean by the
of the ocean ‘weather’, the meso-scale eddies atmosphere (Chapter 2) can vary spatially. It can
(Chapter 4) which help to distribute heat and also be used operationally in improving wave fore-
ensure that the ocean is far from homogeneous. casting models.
Such images can be used by fishing fleets to locate Although microwave radiometry cannot match
regions, such as frontal zones, which are favourable the fine spatial and radiometric resolution of infra-
for finding certain species of fish. red sensors, its reduced sensitivity to cloud cover
A completely different technique, using the mea- enables it to play a valuable role in cloudy regions.
surement of ocean colour, is illustrated in Box 5.2. Microwave radiometers can yield not only a mea-
Interestingly, this method can also reveal similar sure of SST, but also can measure wind stress and
meso-scale dynamical processes to those imaged by detect the presence of sea ice. They therefore have a
their temperature signature. It can also be applied valuable operational role to play in polar seas.

78
 5: Observing Oceans from Space

Box 5.2 Ocean Colour Images

Figure 5.14 The apparent colour of 5.14
the water is affected by the combi-
nation of scattering and absorption
of the Sun’s light, as illustrated. The
red end of the spectrum is preferen-
✹ The ratio of green/blue in the
Scattering by the water-leaving light is a measure
atmosphere changes
tially absorbed by sea water, and the apparent colour
of the chlorophyll content
of the water-leaving
light scattered from deep in the sea light
consequently appears blue. How-
ever, chlorophyll present in the
water absorbs blue light, so that the
water-leaving light appears greener
in the presence of phytoplankton,
provided there is enough particulate Scattering from shallow
particles (top 1 m):
material to scatter the green light all visible wavelengths
reflected to sensor
before it is absorbed by the water
itself. Other substances can also Scattering from deeper
particles (1 to 10 m):
Scattering from
phytoplankton:
influence the water colour, such as red light already absorbed,
green and blue reflected
blue light absorbed by
chlorophyll, green light
Scattering from deep particles:
red and green already
dissolved organic material (known reflected if not too deep absorbed, only blue reflected

as yellow substance) and exception-

al ‘red tide’ plankton blooms (see
Chapter 6). The brightness of scat-
tered light also gives a qualitative indication of the amount of suspended particulates in the water.
Consequently, the remote sensing of ocean colour can be used to measure the concentration of phytoplankton
in the upper layers and to monitor the distribution of suspended particulates, as well as to act as a tracer for the
patterns of dynamical features.

Figure 5.15 An image defining the distribution of chlorophyll pig- 5.15

ment concentration in the North Atlantic off the Iberian peninsula on
7th April, 1980, derived from CZCS observations of ocean colour.
Images such as this illustrate how variable the ocean can be, and Mean pigment (mg/m3
demonstrate the value of remote sensing methods for instantaneously
capturing the spatial distribution of phytoplankton. The heterogeneity
derives from the patchiness of primary production in spring, which
depends on the distribution of nutrients and of overwintering seed
populations. Once ‘hot spots’ of production have commenced, the
variable currents associated with meso-scale dynamical features move Lisbon

the patches by advection, stretching and twisting them into complex

patterns. One of the most interesting features in this image is the
strong filament which extends several hundred kilometres offshore
from Lisbon. This, and other less distinct filaments further north, are
important for the transport of continental shelf water into the deep
ocean. The image suggests, however, that between the offshore flow-
ing filaments are zones of less productive water flowing toward the
coast to form an onshore–offshore circulation cell.

Figure 5.16 An enhanced colour image of part of Southampton 5.16

Water, UK, generated from Airborne Thematic Mapper data. The
image has been constructed from measurements in the blue, green,
and red parts of the spectrum to give a near-real colour image, but
because the contrast is enhanced the colours appear brighter than
they would to the naked eye. The most apparent feature on this image
is the plume in the wake of the large container vessel. Whereas the
surrounding water appears dark blue–green, the plume appears yel-
lower, indicating that red as well as green and blue light is being
back-scattered. This occurs when there are high suspended-sediment
concentrations close to the surface and the image indicates that the
passage of the ship has stirred up material from the sea bed. Brighter
patches elsewhere also indicate increased suspended sediment, possi-
bly caused by the earlier passage of other ships.

79
I.S. Robinson and T. Guymer

Box 5.3 Synthetic Aperture Radar Imaging of

Small-Scale Ocean Processes
Figure 5.17 It is not possible with a satellite 5.17
sensor to focus microwaves in the same Azimuthal resolution achieved by
way as visible or infra-red radiation. To synthesizing a long aperture
from repeated pulses
achieve a high-resolution image in the along the track

range direction (the radar pointing direc-

tion), the radar uses the timing of the return
pulse to determine the precise distance to Radar
the patch of sea surface being viewed. To antenna

achieve comparably high resolution in the Satellite

orbital path Azimuth
azimuth (along track) direction would direction
require a very large aperture antenna. This
Ground range
cannot be constructed physically, but its direction
effect is synthesized digitally using the Satellite
ground track
Swath
recorded pulse returns from many positions
of the satellite along its orbit, hence the
Range resolution achieved
name Synthetic Aperture Radar. SARs on by timing the return pulse
satellites are capable of resolving down to
20 m. A typical SAR image consists of up to
4000 x 4000 picture elements (pixels). A SAR produces an image corresponding to the magnitude of the radar
energy returned from the sea surface. Since the radar views the sea at an oblique angle, the signal relies on
detecting back-scattered radiation caused by the interaction between the incident radar waves and the rough-
ness of the sea surface. The microwaves do not penetrate the surface at all, and so the patterns on the radar
image are due to variations of the sea surface roughness.

Figure 5.18 An image of the southern North 5.18

Sea adjacent to the Dover Straits acquired by
SAR on Seasat, 19th August, 1978. The origi-
nal data have been averaged to a resolution
cell of 168 m, eliminating the speckle effect
and resulting in a very smooth image. The
width of the image represents about 90 km
and the direction of North is 26.1° clockwise
from the top of the image. The radar back-
scatter patterns appear to reveal the bathym-
etry of the sea bed – the image looks rather
like a photograph of the sea bed laid bare
and illuminated from an oblique angle.
Images like this took many oceanographers
by surprise when they were first obtained. It
must be emphasised that the radar signal
does not penetrate through the sea to the sea
bed. The imaging mechanism is this: the
tidal currents fluctuate in magnitude as they
flow over shallow or deep regions, causing
horizontal convergence and divergence
which concentrates or reduces the surface
wave energy, and hence the roughness con-
trolling the radar return. Although such a
mechanism seems rather complex, the clarity of bathymetric features in the image, which correlate very well
with bathymetric charts, demonstrates that this is an effective method for mapping sandbanks and sandwaves.
Current research is determining whether quantitative as well as qualitative information can be recovered, and
examining the conditions of tide and wind which are necessary to produce such clear images of the sea bed.

80
 5: Observing Oceans from Space

Figure 5.19 This ERS-1 SAR image of the English 5.19

Channel off the Isle of Wight was recorded on 2
July, 1993, and the width of the image represents
80 km. Slicks, regions in which the sea surface is
smoother than normal, appear as dark regions on
the image and may be caused either by the pres-
ence at the surface of material such as organic
films, or by dynamical features which produce
local divergence of the surface current. Either
process reduces the amplitude of the short surface
waves which influence the radar back-scatter. In
this image the slicks provide a way of detecting
other dynamical processes. In some of the coastal
embayments narrow slicks appear to be aligned
with the local tidal circulation. In other areas the
slicks relate to old ship wakes which have left a
trail of smoother water behind. Further offshore the
larger slicks are probably patches of surface film
discharged from ships. The corrugated shape of
these patches is due to the shear associated with
the strong tidal streams which flow parallel to the
coast. (The data from which this image was derived
were supplied by the European Space Agency.)

Figure 5.20 This ERS-1 SAR image of the Atlantic 5.20

coast off Portugal contains a lot of wave-like phe-
nomena consisting of clusters of between four and
eight slightly curved, concentric crests. Since the
size of the whole image represents 100 km across,
the wavelengths of these phenomena, greater than
1 km, imply that they are the surface manifestation
of internal waves. These are subsurface waves pro-
ducing undulations of the interface between the
upper mixed layer of the sea and deeper, cooler,
denser layers, and are generated by tidal flow
encountering the shelf edge. Packets of several
internal waves are produced by each tide and the
surface currents associated with them act to com-
press or stretch the wind waves and thus generate
zones of rougher and smoother sea surface.
Observations at sea coincident with SAR overpass-
es confirm that the zones of rough and smooth sea
are in phase with the internal waves, which thus
acquire a signature in the SAR image. Because
these waves propagate quickly, it is not possible to
use ship measurements to map their spatial extent.
SAR images provide the only systematic way of
detecting their occurrence and defining their spa-
tial distribution. By making assumptions about their
tidal origin, it is possible to estimate their propaga-
tion speed from the spacing between distinct wave
packets. (The data from which this image was
derived were supplied by the European Space
Agency.)

81
I.S. Robinson and T. Guymer

Box 5.4 Sea Surface Topography

As implied by its name, the altimeter primarily measures height – in this case, that of sea level. The
method is conceptually very simple. A sharp pulse is transmitted vertically toward the Earth’s surface
from an antenna on the satellite. Some of the signal is reflected back to the same antenna from those
parts of the surface within the 7 km footprint which are aligned perpendicular to the beam. The key to
the technique is the accurate measurement of the time delay between transmission and reception of
the pulse by the antenna. If the signal’s propagation speed is known then the distance between the
satellite and the surface can be obtained. The height of the satellite orbit can be determined by track-
ing stations and the difference between the two provides an estimate of sea level. As the satellite cir-
cles the Earth, so sea level changes in space and time can be monitored.
In practice, it is very difficult to achieve the accuracy required for oceanography (better than 10 cm).
Corrections must be made for the effects of the atmosphere on the propagation speed and, because
the reflected pulses are very noisy, sophisticated processing techniques are needed to time the arrival
of such pulses with high precision. Even the task of precisely determining the position of the satellite
presses orbital dynamics computations to their limit, because large satellites in low altitude orbits are
influenced significantly by such processes as air drag and solar radiation pressure. It is, indeed,
remarkable that the present generation of altimeters flying at altitudes of 1000 km can measure sea
level to 4 cm – a precision of 1 in 25 million!
What can we learn from the resulting sea level data? The largest variations by far, amounting to
more than 200 m, are those of the geoid, a surface connecting points of equal gravitational potential.
The Earth’s gravity field is rather uneven because of the way in which mass is distributed, and this is
manifested in sea level. Although of great interest to solid Earth geophysicists – the influence of ocean
trenches and seamounts is often readily discernible – it is a nuisance to oceanographers wanting to
use altimetry because it prevents the calculation of mean currents. What is required is global mapping
of the geoid, independently of altimetry. This surface must be subtracted from the altimeter data to
yield the oceanographic signal, referred to as the dynamic topography. An example is shown in Figure
5.21. Dominant large-scale oceanographic features, such as the Antarctic Circumpolar Current and the
Subtropical Gyre, can be identified, but most of the structure represents uncertainties in the geoid
which mask the relatively weak signals produced by ocean dynamics. We must await the launch of
satellites dedicated to mapping the Earth’s gravity field at high spatial resolution before we can signifi-
cantly improve our ability to separate the ocean’s dynamic topography from the geoid.
Happily, the need to know the geoid can be eliminated by studying time-varying altimeter signals
of the ocean, since the geoid is constant on the time-scales of interest to us. One of the most obvious
changes in sea level with time is that due to tides. Altimeters now provide the most accurate means of
mapping open-ocean tides on a global basis. However, after allowing for these regular rises and falls
there remain the irregular changes due to ocean currents. Their speed and direction are related to sea
level in the same way as winds in the atmosphere are associated with the horizontal distribution of air
pressure. Thus, as an oceanic eddy or meander passes through an observation point there is a change
in sea level (depressed for a cyclonic eddy and raised for an anticylonic one). These sea level changes
can be expressed as departures from a long-term mean – we call them anomalies. Figure 5.22 shows
sea level anomalies in the South Atlantic obtained from ten days of altimeter data. Over most of the
region the signal is less than 5 cm, but two areas (off South Africa and in the Southwest Atlantic)
show positive and negative anomalies of 40 cm magnitude. The former corresponds to the Agulhas
Retroflection (where the Agulhas Current turns and flows toward the east, south of South Africa) and
the latter to the exit region of the Brazil–Falklands Confluence – see also Figure 5.12; both are known
to be highly energetic and to spawn meso-scale eddies (Chapter 4). Eddies are important in contribut-
ing to horizontal heat transports and, through upwelling of nutrients, help to determine biological vari-
ability. Altimeter data offers the possibility of monitoring the movement and development of such fea-
tures and, by assimilation into models, should provide a key element in a future ocean-forecasting
system.

82
 5: Observing Oceans from Space

5.21

Figure 5.21 Mean sea surface height of the South Atlantic in metres above the
geoid. The high in the centre corresponds to the Subtropical Gyre; to the south the
Antarctic Circumpolar Current shows up as a region of increased north–south gradi-
ent. (Courtesy of Matthew Jones, James Rennell Division for Ocean Circulation,
SOC, and Mullard Space Science Laboratory, UCL.)

5.22

Figure 5.22 Sea surface height anomaly field derived from ten days of TOPEX/POSEI-
DON data, calculated with respect to a 2-year mean. The colour scale shows heights
of the anomalies in metres. (Courtesy of Matthew Jones, James Rennell Division for
Ocean Circulation, SOC, and Mullard Space Science Laboratory, UCL.)

83
I.S. Robinson and T. Guymer

Box 5.5 Wave Height

Contained within the shape of the return pulse of the radar altimeter is information on significant
wave height. This quantity corresponds closely to the mean of the highest one-third of the waves
being observed at any one time and is similar to the wave height estimated from visual observations.
In contrast to the calm sea case, reflection takes place earlier from the wave crests and later from their
troughs, which leads to a smearing of the leading edge of the received pulse. The higher the waves
the greater is the slope of this leading edge; it is this behaviour that is used to calculate values of wave
height. Comparisons with buoys have shown that accuracy of wave height estimates may be better
than 10%.
Our knowledge of global wave climate from conventional data is rather poor: buoys equipped with
wave sensors are confined to coastal regions and most ship values come from visual observations,
which can be very subjective and are biased to fair-weather conditions. Altimeters can provide uni-
form coverage regardless of weather conditions (except in those areas where sea ice is present).
Figure 5.23 portrays the seasonal variation in significant wave height. Significant differences between
winter and summer occur, especially in the northern hemisphere. The high wave heights in the Indian
Ocean in June are a consequence of the southwest monsoon.
Less obvious, but more intriguing, is the variation from year-to-year [a selection of cases is shown
in Figures 5.23(b) and 5.24] compiled from three altimeter missions – Geosat, ERS-1, and
TOPEX/POSEIDON. The plots are for winter in the northern hemisphere. Some features are common
to all years – highest waves occur in the mid-latitudes of both hemispheres, corresponding to the
stormy westerly winds. The maxima are displaced to the eastern side of ocean basins, where the fetch
is greatest. Despite it being summer, wave heights in the Southern Ocean are as large as their north-
ern counterparts, although the high wave height zones do not extend so far equatorward. Wave
heights in the tropics are always small when averaged over a month or so. However, some differences
can be detected. In some years (e.g., 1992–1993) the North Atlantic and Norwegian Sea are rougher;
when this happens the North Pacific is calmer. Likewise, the South Atlantic shows large changes with
a minimum in 1987–1988 and a maximum in 1993–1994. Why does this interannual variability occur?
Since it presumably reflects changes in the wind forcing, there are important implications for ocean
circulation. We are exploring these links as part of the contribution to the World Ocean Circulation
Experiment (see Chapter 3).

84
 5: Observing Oceans from Space

5.23a 5.23b

Figure 5.23 Mean significant wave height (m) obtained from TOPEX/POSEIDON altimeter data: (a) June 1994;
(b) December 1993–January 1994. (Courtesy of David Cotton, James Rennell Division for Ocean Circulation,
SOC.)

5.24a 5.24b

5.24c 5.24d

Figure 5.24 Mean significant wave height (m) for four northern hemisphere winters: (a) 1986–1987 from
Geosat; (b) 1987–1988 from Geosat; (c) 1991–1992 from ERS-1; (d) 1992–1993 from TOPEX/POSEIDON.
Wave heights vary from less than 0.5 m to greater than 5 m as shown in the colour scale. (Courtesy of David
Cotton, James Rennell Division for Ocean Circulation, SOC.)

85
I.S. Robinson and T. Guymer

Box 5.6 Scatterometer

Like the altimeter, the scatterometer is also a radar sensor, but instead of measuring time delay or the
distortion of pulse shape, it relies on using the amount of back-scattered power to infer an oceano-
graphic quantity. From the early days of radar it has been known that when the sea surface is viewed
obliquely ‘ground clutter’ exists, which increases with sea state. In glassy calms there is no detectable
return because the incident radiation is reflected away by the mirror-like surface, with no return along
the beam. As the wind speed increases small waves are set up which scatter some energy back
toward the antenna. The ‘cat’s paws’ often seen on the sea surface when a breeze springs up are an
example. Microwave radars operating at wavelengths corresponding to this ripple-scale (ca 2 cm) are
found to be most sensitive to wind variations. Furthermore, the back-scatter is not the same in all
directions, being a maximum in the direction of the wind and a minimum across it. This is due to the
effect of longer wind waves which are aligned more or less across the local wind.
Such knowledge is exploited in scatterometry in that broad swaths are illuminated with microwave
radiation at incidence angles of 20–55º and with spatial resolution cells of side 50 km. The effect of
individual waves on the back-scatter is averaged out on this scale. Several antennae are deployed at
various azimuth angles so that the same area of ocean can be viewed from different directions over a
period of a few minutes, allowing wind direction to be inferred. There is no accepted model of the
wind dependence of the measured back-scatter and so the retrieval algorithm is empirically deter-
mined from simultaneous radar back-scatter and in situ wind measurements (see Figure 19.13). In the-
ory this model function can be achieved by comprehensive airborne measurements prior to the
launch of the satellite, but in practice it has proved necessary to carry out considerable modifications
using the satellite data themselves.
Despite the lack of fundamental understanding, comparisons of scatterometer wind velocities have
shown agreement to 2 m/s and 20º with independent in situ data. The accuracy, spatial resolution, and
all-weather coverage of scatterometer data make them very useful for studying individual meteorolog-
ical events and for obtaining the time-averaged wind fields needed to calculate surface momentum
and heat and water vapour fluxes for large-scale ocean circulation studies. Figure 5.25 is an example
of the detailed winds in a hurricane, generated from scatterometer data and superimposed

5.25

Figure 5.25 ERS-1 scatterometer surface wind vectors across Hurricane Emily on
30 August, 1993, superimposed on a coincident Meteosat cloud image. The
length of the vectors is proportional to wind speed and those in red near the eye
of the hurricane correspond to winds greater than 15 m/s. (Courtesy of the
European Space Agency.)

86
 5: Observing Oceans from Space

on an infra-red cloud image obtained by a geostationary weather satellite. The realistic distribution of
wind speed relative to the hurricane and the counterclockwise flow centred on the eye cloud add cred-
ibility to the scatterometer data. In some cases tropical storms have been detected in ERS-1 scat-
terometer data before they have appeared on analysis charts, presumably because of the sparseness
of the routine weather observations in relation to the relatively small scale of these tropical systems.
For some oceanographic purposes, e.g., study of localised coastal upwelling or deepening of the sur-
face mixed layer, the scatterometer wind data should prove extremely valuable, especially in remote
ocean areas where other forms of data may not be available. Global wind fields can be produced by
combining data from consecutive passes of the satellite, as in Figure 5.26. The Trade Winds can be
observed converging on the Equator. At higher latitudes the winds are stronger and more variable in
direction, being influenced by the passage of depressions. The largest areas of strong westerlies are
found in the Southern Ocean.

5.26

Figure 5.26 Average global wind field for April 1993 derived from ERS-1 scatterometer data. The strength of
the wind is indicated by the length of the arrows (lightest winds are in blue, strongest are in red). (Courtesy of
the European Space Agency.)

87
I.S. Robinson and T. Guymer

The Future for Satellite Oceanography However, the increased use of satellites is unlike-
From these few examples of the applications of ly to reduce the requirement for in situ measure-
satellite data, the versatility and wide scope of ments. In fact, the reverse is true. The calibration
marine remote sensing should be apparent. New and validation of satellite sensors requires regular
sensors with higher performance are being designed, comparisons with data acquired at sea, so there is a
and new methods for processing the data are being demand for new ocean instrumentation to measure
developed to improve the accuracy with which parameters, such as surface roughness and surface
ocean parameters can be measured. The space agen- radiation temperature, which have not been readily
cies of Europe, North America, and Japan have a measured from ships before.
continuing programme of Earth Observation Remote sensing methods, by offering a new per-
Satellites planned, with a number of sensors dedi- spective on the ocean, are stimulating new oceano-
cated to oceanographic applications, and there is graphic endeavours which will embrace all types of
scope for new ideas to be injected into the develop- oceanographic techniques. The challenge for the
ing technology. next generation of oceanographers is to combine
An even more challenging task is to improve the satellite data with more traditional methods, and to
way in which remotely sensed data are integrated use an integrated approach to answer questions
with more conventional measurements from ships about global ocean processes that could otherwise
and buoy-mounted instruments. The goal should be not even be asked.
to enable the different types of measurements to
complement each other in global ocean-monitoring
programmes. An important integrating factor will
be the use of computer models. Because of their General References
pixellated nature, satellite data are well adapted for Maull, G. (1985), Introduction to Satellite Oceanography,
Martinus Nijhoff, Dordrecht, 606 pp.
assimilation into numerical models or for valida- Robinson, I.S. (1985), Satellite Oceanography, Ellis Horwood,
tion of model predictions. The rapidly improving Chichester, 460 pp.
performance of global telecommunications will Saltzman, B. (ed.) (1985), Satellite Oceanic Remote Sensing,
ensure that the vast quantities of numbers being Advances in Geophysics, Volume 27, Academic Press,
London, 511 pp.
transmitted from Earth Observation Satellites can Stewart, R.W. (1985), Methods of Satellite Oceanography,
readily be used to improve operational forecasts, as University of California Press, Berkeley.
well as contributing to an improved understanding
of oceanic processes.

88
CHAPTER 6:

Marine Phytoplankton
Blooms
D.A. Purdie

Introduction very dense blooms are often termed ‘exceptional

Marine phytoplankton are often referred to as the blooms’; they are recurring features in some sea
‘grass of the ocean’, since they form the productive areas and lead to a variety of notable consequences,
basis on which most animal life in the oceans is as described below.
ultimately dependent, be it a herbivorous zooplank- In estuaries or coastal waters, phytoplankton
ton consumer, a top carnivore, or a benthic deposit blooms may be small localised features, a square
feeder. ‘Phytoplankton’ is a collective term to kilometre or so in extent, but they are often much
describe the single-celled microscopic algae (they more extensive in deep water, sometimes covering
may also form chains or clumps of cells), that float hundreds of square kilometres. The discolouration
in the surface, well-illuminated waters of the sea. of the sea that they cause can be detected by air-
When conditions are optimum for their rapid craft or satellites, and photographed1 or measured
growth, or if more dispersed cells are aggregated by by specialised narrow-range spectral sensors
some mechanism, such as the fronts described later, mounted on airborne5 or space platforms6,7 (see
planktonic micro-algae become sufficiently abun- Chapter 5). An upwelling bloom event along a
dant to colour the sea; this phenomenon is often 60 km front in the central mid Pacific has been
referred to as a plankton ‘bloom’. photographed from satellites, ships, and the space
The term bloom was originally used by analogy shuttle Atlantis16. An extensive bloom of the coc-
with terrestrial plants, to describe the spring flow- colithophore Emiliania huxleyi, detected in the
ering of the diatoms which characterises most tem- North Atlantic from a series of satellite images7 (see
perate waters. Blooms normally consist of one par- Figure 4.11), covered an area of ca 250,000 km2.
ticular species that dominates the plankton in the The mechanisms responsible for changes in the
surface waters affected. The term ‘Red Tide’ is also micro-algal carbon biomass with time in the near-
often used to describe blooms, as many of the surface waters of the sea, known as the euphotic
organisms cause a red colouration to the sea – zone (i.e., the zone in which there is sufficient light
although green, brown, and yellow colourations for growth and cell division of phytoplankton), are
are also known. These characteristic colourations controlled by a balance between the primary (algal)
of the water are a direct result of high concentra- production rate and losses from the plankton com-
tions of photosynthetic compounds (i.e., chloro- munity through respiration, grazing by herbivores,
phylls) and accessory pigments which are found in and sedimentation of intact cells. During non-
the micro-algae. The pigments absorb certain wave- bloom periods, micro-algal biomass in the sea
lengths of visible light, and algal cells scatter a sub- remains quite constant – the production of biomass
stantial fraction of the submarine light, causing the is approximately balanced by respiration, grazing,
apparent change in the colour of the water (see also and sedimentation losses. Phytoplankton blooms
Box 5.2 and Chapter 14). result when production exceeds loss. The environ-
Blooms of micro-algae represent a large increase mental conditions required for a phytoplankton
in biomass and generally last for relatively short bloom to occur are:
periods, a week or so. They are not exclusive to the
• Availability of the inorganic plant nutrients
sea; they also occur in freshwater lakes and rivers.
nitrate and/or ammonium, phosphate, and sili-
Blooms become visible as a discolouration of sea
cate (in the case of diatoms).
water when chlorophyll concentrations exceed
10 mg/m3 and cell concentrations in the nanoplank- • Sufficient irradiance levels within the surface
ton size-range exceed 0.5 x 109 cells/m3. Typically, waters.
however, chlorophyll levels and cell densities may • An imbalance or lag between the production of
exceed 100 mg/m3 and 30 x 109 cells/m3, respec- micro-algae and their grazing by herbivores,
tively, and may range up to 2000 mg/m3 and 100 x advection, or sedimentation out of the produc-
109 cells/m3 in some localised coastal regions. These tion zone.

89
 D.A. Purdie

6.1 Figure 6.1 Colonies of Phaeo-

cystis sp showing small dot-like
cells embedded in a mucous
colony. The spherical colony is
about 1 mm in diameter.

Once nutrients become exhausted in surface colonies of cells embedded in a mucous envelope
waters, cells begin to die and the phytoplankton (Figure 6.1). Dense blooms of Phaeocystis occur in
sediment out of the water column. This may result many coastal regions, including the southern North
in production of rapidly sinking amorphous aggre- Sea. A related group of marine phytoplankton, the
gates of dead and decaying phytodetritus, known coccolithophores, produce calcite plates that cover
as ‘marine snow’. This material may descend great the outside of the cell [Figure 6.2(a)] and, during
depths through the water column in a few days, bloom conditions, cause the water to take on a
and result in a thick flocculant covering of the sur- milky turquoise colour [due to the scattering of
face sediment (see Chapter 7). These processes of light by the cells; Figure 6.2(b)].
bloom growth and decay can be expressed in math- The two most frequently observed ‘protozoa’
ematical terms and represented in biophysical that cause discolouration to the sea when present in
models15. high abundance are Noctiluca and Mesodinium
Marine phytoplankton blooms are natural phe- rubrum, neither of which are known to posses any
nomena. There is, however, increasing evidence toxic effects.
and consequent concern that anthropogenic activity Noctiluca is a heterotrophic (i.e., organism
(e.g., the supply of nitrates by rivers) is fertilising deriving nutrition from assimilation of organic
coastal waters, resulting in an increase in the inten- matter) dinoflagellate and a voracious phagotroph-
sity and frequency of bloom events, as well as ic (i.e., particle ingesting) feeder. It is pigmented by
affecting the natural balance of phytoplankton orange carotenoids (i.e., non-photosynthetic pig-
species diversity (see also Chapter 15). These ments), possesses the ability to control its buoyancy
‘unnatural’ blooms have a number of both direct in sea water, and is often noticed as a tomato-red
and indirect affects on the marine biota, many of colouration on the surface during calm conditions.
which are deleterious and result in severe environ- In enclosed shallow harbours blooms of this organ-
mental problems. ism can cause severe oxygen depletion, leading to
mortality of marine animals.
Common Examples of Bloom-Forming The highly motile ciliate Mesodinium rubrum
Marine Phytoplankton (Figure 6.3) contains a photosynthetic cryptomon-
Diatoms are the dominant marine phytoplankton ad symbiont (i.e., an algal cell living within the pro-
in most marine waters, particularly during the tozoan host) and can be functionally considered a
spring bloom. The cells are enclosed in a micro- member of the phytoplankton3. It is extremely pro-
scopic porous shell or frustule constructed from sil- ductive in some regions, reaching levels of chloro-
ica assimilated as dissolved silicate from sea water. phyll concentration of 1000 mg/m3, with a highest
Dinoflagellates are spirally swimming organisms recorded production rate of >2000 mgC/m3/h. M.
propelled by flagella; they are the major primary rubrum causes red-water events in coastal, estuar-
producers in many marine regions during summer. ine, and upwelling ecosystems in many parts of the
Phaeocystis sp is a prymnesiophyte that forms large world 3 . It has been intensively studied in the

90
 6: Marine Phytoplankton Blooms

6.2a 6.2b

Figure 6.2 The coccolithophore Emiliania huxleyi: (a) an electron micrograph of cells (about 6 μm in diameter)
showing calcite plates covering the cell (courtesy of D. Harbour, Plymouth Marine Laboratory, Plymouth, England);
(b) milky turquoise colouration to surface waters (picture taken from RV Charles Darwin during June 1991 in the
northern North Atlantic).

Southampton Water estuary in southern England, ance. Subsequently, it was realised that the depth of
where it forms red tides almost annually (Figure mixing also plays an important role. A Norwegian
6.4). During the development of these blooms, the scientist, H.H. Gran, first established that, in win-
inorganic nutrient levels (ammonia and nitrate) in ter, phytoplankton are mixed vertically to the bot-
the estuary are considerably reduced; also, bottom- tom of the mixed layer and spend considerable
water oxygen levels are depleted by the organism periods at depths where there is insufficient light
due to its downward displacement at night4. The for photosynthesis; with the onset of the spring
unusually high swimming speed of the organism warming of surface waters and a decrease in storm
reduces its chances of being flushed from the estu- mixing, the mixed-layer depth and therefore the
ary, since it avoids increased surface turbulence depth to which phytoplankton cells are regularly
caused by the ebb tide by swimming to deeper mixed, decreases; they are consequently exposed to
depths. increasingly higher average light levels, allowing
blooms to occur. The classic quantitative explana-
The Spring Bloom tion of the spring bloom, known as the ‘critical
The best-known type of phytoplankton bloom depth theory’, was first proposed by Sverdrup14.
occurs in spring in temperate and subpolar waters. We now know that the spring bloom is initiated
It was once thought that onset of the spring bloom when the carbon fixed by phytoplankton photosyn-
was triggered by the spring increase in solar irradi- thesis in the mixed layer exceeds the respiration

6.3a 6.3b

Figure 6.3 (a) Photomicrograph of the photosynthetic ciliate Mesodinium rubrum (about 100 μm diameter) (Courtesy
of Dr D. Crawford, Department of Oceanography, SOC, Southampton, England.) (b) Red tides caused by Mesodinium
rubrum in Southampton Water estuary, UK.

91
 D.A. Purdie

losses per unit area, or when phytoplankton are bloom in the North Atlantic12 is to follow the water
confined above a ‘critical depth’. Riley11 showed movement rather than to stay in a fixed (e.g.,
from empirical evidence that phytoplankton cannot mooring) position; this has highlighted the com-
bloom before the average irradiance in the mixed plexity of spatial and temporal variability in an off-
layer exceeds 3.25 E/m2/day. This critical irradiance shore region, where mesoscale eddies complicate
is reached as a consequence of seasonal increases in the picture of temporal chlorophyll changes (see
both density stratification (shallowing of the mixed also Figure 4.10).
layer) and increasing solar irradiance. In some well- The poleward migration of the spring bloom in
mixed coastal regions and estuaries, a decrease in the North Atlantic has been reported by Strass and
turbidity may also result in a deepening of the criti- Woods13 from data collected using a towed vehicle,
cal depth9. the SeaSoar (see Figure 19.6). Chlorophyll distribu-
Diatoms are almost without exception the group tion throughout the surface 200 m between the
of micro-algae that dominate the spring bloom in Azores (38°N) and Greenland at 58°N was mea-
temperate waters. This is ascribed to their rapid sured from chlorophyll fluorescence using a sub-
growth rates, which are sustained by the high con- mersible fluorometer attached to SeaSoar. The
centrations of available nutrients in early spring near-surface chlorophyll fluorescence showed a
and the tolerance of these organisms to the turbu- patchy bloom development between 40–49°N in
lent conditions which prevail during this period. An April, with a bloom centred on 49°N in June–July
example of a spring bloom in an estuary is shown and low chlorophyll levels throughout most of the
in Figure 6.4. transection in August–September, with bloom con-
Whereas many studies have been reported of the ditions existing only north of the polar front at
temporal development of phytoplankton biomass 52°N.
and the species succession associated with the At a particular location, the spring bloom can be
development of spring blooms in coastal waters, a very transient feature and is easily missed if water
few detailed studies of the spring bloom are avail- samples are not collected frequently, a particular
able for offshore regions. One approach to study- problem in offshore waters. This is well-demon-
ing the development of the phytoplankton spring strated in the data obtained during a study in the

6.4 (a) 500 4000

3500
Total diatoms (103 cells/l 1)

400
3000
Mesodinium rubrum
(103 cells/l 1)

300 2500
2000
200 1500
1000
100
500
0 0
Figure 6.4 Seasonal changes in the
abundance of diatoms (blue) and the
(b) 50 1500 photosynthetic ciliate Mesodinium
rubrum (red) in the mid-estuary region
of Southampton Water, UK, during
40 1200 1988. The initial peak in diatoms dur-
ing April was dominated by chains of
Chlorophyll a (+g/l 1)

NO –N (+g/l 1)

30 900 the small-celled Skeletonema costa-

tum, with the second peak dominated
by the larger celled Rhizosolenia deli-
20 600 catula. Diatom numbers remain rela-
3

tively low during the rest of the year

due to limited silicate availability in
10 300 the estuary. Note the spring bloom
development of diatoms and the sum-
mer bloom of the photosynthetic
0 0 ciliate.

92
 6: Marine Phytoplankton Blooms

Figure 6.5 Seasonal changes in depth (m) distribution of Temperature (ºC) 6.5
(a) water temperature, (b) dissolved oxygen saturation, SEP JAN APR AUG
and (c) chlorophyll (derived from fluorescence) at a sta-
tion in the central North Sea, sampled once a month. (a)
These waters become thermally stratified in the spring
due to surface heating; in autumn, storms cause complete
mixing of the water column. Phytoplankton blooms were
detected in late summer in surface waters; however, the
transient surface spring bloom chlorophyll peak in April ABOVE
17.5
19.0
– 19.0

was not detected from the research ship. Settled bloom 16.0
14.5
–
–
17.5
16.0
material in near-bottom waters was observed in late 13.0
11.5
–
–
14.5
13.0
April. Dissolved oxygen is supersaturated during surface 10.0 – 11.5
8.5 – 10.0
blooms and the development of undersaturated oxygen 7.0
BELOW
– 8.5
7.0
conditions is evident in bottom waters in late summer
and autumn. Oxygen saturation (%)
SEP JAN APR AUG
North Sea, in which water samples and water col- (b)
umn data were collected by a research ship that vis-
ited a particular site in the central North Sea every
four weeks and failed to detect the peak chloro-
phyll levels. Innovative new autonomous instru-
mentation to monitor chlorophyll fluorescence at ABOVE 135.0
125.0 – 135.0
intervals of up to four times per hour over a 115.0 – 125.0
105.0 – 115.0
month’s deployment on a mooring at this site did 95.0
85.0
– 105.0
– 95.0

detect a transient peak in chlorophyll levels 75.0

65.0
– 85.0
– 75.0

between ship surveys15. The levels were up to three 55.0

BELOW
– 65.0
55.0

times higher than those measured from the ship.

Chlorophyll (mg/m3)
Deleterious Effects of Blooms SEP JAN APR AUG

Toxic effects and ‘Red Tides’ (c)

The term ‘Red Tide’ is often used in an emotive
way to describe a poisoning of water. Only a few
dozen of the thousands of marine phytoplankton
species are known to be toxic, and most of these
are dinoflagellates, prymnesiophytes, or ABOVE
3.5 – 4.0
4.0

chloromonads. Human life can be affected if shell- 3.0

2.5
–
–
3.5
3.0

fish, such as clams, mussels, oysters, or scallops, 2.0

1.5
–
–
2.5
2.0

ingest the algae as food and retain and concentrate 1.0

0.5
–
–
1.5
1.0

the toxins in their tissues. Although the shellfish are 0.0 –

BELOW 0.0
0.5

not greatly affected themselves, a single clam can

accumulate sufficient toxin to kill a human being.
Shellfish-poisoning syndrome is known as paralytic,
diarrhoetic, and neurotoxic shellfish poisoning, or the marine environment. However, the immense
PSP, DSP, and NSP, respectively. A new type of increase in organic matter associated with blooms
shellfish poisoning was reported in 1990 on the can cause other indirect deleterious effects on local
eastern seaboard of Canada, termed amnesiac shell- waters and their adjacent environment.
fish poisoning or ASP, in which infected individuals During the day micro-algae photosynthesise and
suffered short-term memory loss. The toxins produce new organic matter and oxygen. Rapidly
responsible for these poisonings are from families growing phytoplankton populations cause oxygen
of compounds such as the saxitoxins, which cause supersaturation of surface waters, with values of
PSP by disrupting nerve impulses to the muscles greater than 160% having been recorded; this may
and nervous tissue. It is unclear why certain algae cause some motile macrofauna to avoid such
produce toxins, but it is possible they serve as a regions, since oxygen bubbles may form in their
deterrent to grazers and may give the algae an gills. A more frequent problem is that caused by the
unpleasant taste. severe undersaturation of oxygen, or hypoxic con-
ditions, that occur during the latter stages of phyto-
Other nuisance effects plankton blooms (Figure 6.5). Respiration by the
Many phytoplankton bloom-forming species have algae themselves, as well as by animals and bacte-
no apparent directly toxic effect on organisms in ria, consumes oxygen. In some estuaries, such as

93
D.A. Purdie

Chesapeake Bay, the bottom waters may become although not apparently directly toxic. The extent
deficient in oxygen after a phytoplankton bloom of bottom-oxygen deficiency in the central northern
and, in extreme cases, anoxic (i.e., totally devoid of Adriatic has also been shown to have increased sig-
oxygen), causing mass mortalities of benthic fauna. nificantly over the past 20 years, as a consequence
This process is exacerbated if the water column of this material sinking to deeper depths following
becomes stratified, either by freshwater inputs to its wind-driven advection offshore.
estuaries or by seasonal heating of surface waters Phytoplankton blooms can cause a reduction in
(causing a surface mixed-layer above a thermo- benthic plant communities due to an increased tur-
cline). Such conditions occur in the German Bight bidity of the water, thus reducing the penetration
region of the North Sea; during the early 1980s of light to bottom-dwelling plants. This has caused
massive fish kills occurred, which were thought to a reduction in sea-grass beds in some coastal tropi-
be in part caused by severe undersaturation of cal regions, and has also affected coral-reef devel-
near-bottom waters. Subsequent data, however, opment.
showed that the waters below the thermocline Many of the marine bloom-forming phytoplank-
become reduced in oxygen following the spring ton are known to produce the volatile sulphur com-
bloom; this is caused by the algal material sedi- pound dimethyl sulphide (DMS)10. DMS undergoes
menting through the thermocline and decomposing photochemical oxidation into sulphur dioxide in
in deeper waters8 (Figure 6.5). the atmosphere and contributes to the production
Phytoplankton blooms may influence recreation- of acid rain. Bloom-forming phytoplankton species,
al use of coastal regions by causing foam and nox- such as Phaeocystis and Emiliania huxleyi, are
ious slime deposits on the beach. Many beaches in known to be important producers of DMS; in
Belgium and the Netherlands become covered in coastal regions they may add significantly to
thick foam, washed up following the breakdown of localised atmospheric sulphur levels10.
the Phaeocystis blooms in adjacent coastal waters.
The occurrence of Phaeocystis in coastal regions of
the North Sea has been linked to enhanced inor-
ganic nutrient enrichment (eutrophication) of these
waters. A rare time-series of measurements was General References
conducted over a period of 14 years, from Anderson, D.M. (1994), Red tides, Sci. Amer., 271(2), 52–58.
1976–1988, by Cadee2 at one point in the Dutch Cosper, E.M., Bricelj, J.M., and Carpenter, E.J. (eds) (1989),
Novel Phytoplankton Blooms: Causes and Impacts of
Wadden Sea. Water samples were collected weekly Recurrent Brown Tides and Other Common Blooms,
in spring and summer, and less frequently, during Springer Verlag, Berlin.
the rest of the year from the pier at the Netherlands Hallegraeff, G.M. (1993), A review of harmful algal blooms and
Institute for Sea Research on the island of Texel. their apparent global increase, Phycologia, 32, 79–99.
Lancelot, C., Billen, G., and Barth, H. (eds) (1990),
Results showed that both the intensity and dura- Eutrophication and Algal Blooms in North Sea
tion of Phaeocystis blooms had significantly Coastal Zones, the Baltic and Adjacent Areas, Water
increased over the 14-year period, which is proba- Pollution Research Reports, Commission of the
bly related to the eutrophication of this region. European Communities, Brussels, 281 pp.
Many coastal regions of the Adriatic Sea have Parker, M. and Tett, L.P. (eds) (1987), Exceptional phytoplank-
ton blooms, Rapports et proces-verbaux des reunions,
been affected in recent years by the production of 187, 9–18.
an extracelluar excretion of mucus or slime caused, Smayda, T.J. and Shimizu, Y. (eds) (1993), Toxic
apparently, by dense, mixed phytoplankton popu- Phytoplankton Blooms in the Sea, Elsevier,
lations. It appears that no single organism is Amsterdam, 952 pp.
Vollenweider, R.A., Marchetti, R., and Vivani, R. (eds) (1992),
responsible for the production of this mucous Marine Coastal Eutrophication, Supplement to Sci.
material, but a number of diatoms and dinoflagel- Total Environ., 1310 pp.
late species have been collected within the affected
areas, on both the Italian and Yugoslavian (as was) References
sides of the Adriatic. Eutrophication of these 1. Berge, G. (1962), Discolouration of the sea due to
coastal waters, particularly caused by large inor- Coccolithus huxleyi ‘bloom’, Sarsia, 6, 27–40.
ganic nutrient inputs from the river Po outflow, has 2. Cadee, G.C. (1990), Increase of Phaeocystis blooms in the
westernmost inlet of the Wadden Sea, The Marsdiep,
been blamed for these recent events, with chloro- since 1973, in Eutrophication and Algal Blooms in
phyll levels rising to more than 800 mg/m3 in some North Sea Coastal Zones, the Baltic and Adjacent
coastal parts of the Emilia Romagna region. The Areas, Lancelot, C., Billen, G., and Barth, H. (eds),
occurrence in recent years of these events each sum- Water Pollution Research Report 12, Commission of
mer has greatly affected the tourist trade in many the European Communities, Brussels, pp 105–112.
3. Crawford, D.W. (1989), Mesodinium rubrum: the phyto-
regions, since the slime occurs on beaches, yields a plankton that wasn’t, Marine Ecol. Progr. Ser., 58,
noxious smell, and is unpleasant to swim in, 161–174.

94
 6: Marine Phytoplankton Blooms

4. Daneri, D., Crawford, D.W., and Purdie, D.A. (1992), Algal 10. Malin, G., Turner, S.M., and Liss, P.S. (1992), Sulfur: The
blooms in coastal waters: a comparison between two plankton/climate connection, J. Phycol., 28, 590–597.
adaptable members of the phytoplankton, Phaeocystis 11. Riley, G.A. (1942), The relationship of vertical turbulence
sp. and Mesodinium rubrum, in Marine Coastal and spring diatom flowering, J. Marine Res., 5,
Eutrophication, Vollenweider, R.A., Marchetti, R., 67–87.
and Vivani, R. (eds), Supplement to Sci. Total 12. Savidge, G., Turner, R.D., Burkill, P.H., Watson, A.J.,
Environ., pp 879–890. Angel, M.V., Pingree, R.D., Leach, H., and Richards,
5. Garcia, C.A.E., Purdie, D.A., and Robinson, I.S. (1993), K.J. (1992), The BOFS 1990 spring bloom experi-
Mapping of the photosynthetic ciliate Mesodinium ment: temporal evolution and spatial variability of the
rubrum in an estuary from Airborne Thematic hydrographic field, Progr Oceanogr., 29, 235–281.
Mapper data, Estuar. Coastal Shelf Sci., 37, 287–298. 13. Strass, V and Woods, J.D. (1988), Horizontal and seasonal
6. Holligan, P.M., Aarup, T., and Groom, S.B. (1989), The variation of density of chlorophyll profiles between
North Sea satellite atlas, Continent. Shelf Res., 9, the Azores and Greenland, in Towards a Theory on
667–765. Biological–Physical Interactions in the World Ocean,
7. Holligan, P.M., Fernandez, E., Aiken, J., Balch, W.M., Boyd, Rothschild, B.J. (ed.), Kluwer, Dordrecht, pp
P., Burkill, P.H., Finch, M., Groom, S.B., Malin, G., 113–136.
Muller, K., Purdie, D.A., Robinson, C., Trees, C.C., 14. Sverdrup, H.U. (1953), On conditions for the vernal bloom-
Turner, S.M., van der Wal, P. (1993), A biogeochemi- ing of phytoplankton, J. Conseil Perm. Int. Exp.
cal study of the coccolithophore, Emiliania huxleyi, in Mer., 18, 287–295.
the North Atlantic, Global Biogeochem. Cycl., 7(4), 15. Tett, P.B., Joint, I.R, Purdie, D.A., Baars, M., Oosterhuis, S.,
879–900. Daneri, G., Hannah, F., Mills, D.K., Plummer, D.,
8. Howarth, J.M., Dyer, K.R., Joint, I.R., Hydes, D.J., Purdie, Pomroy, A.J., Walne, A.W., and Witte, H.J. (1993),
D.A., Edmunds, H., Jones, J.E., Lowry, R.K., Moffat, Biological consequences of tidal stirring gradients in
T.J., Pomroy, A.J., and Proctor, R. (1993), Seasonal the North Sea, Phil. Trans. Roy. Soc. Lond., A., 343,
cycles and their spatial variability, Phil. Trans. Roy. 493-508.
Soc. Lond., A., 343, 383–403. 16. Yoder, J.A., Ackleson, S.G., Barber, R.T., Flament, P., and
9. Kifle, D. and Purdie, D.A. (1993), The seasonal abundance Balch, W.M. (1994), A line in the sea, Nature, 371,
of the phototrophic ciliate Mesodinium rubrum in 689–692.
Southampton Water, England, J. Plankt. Res., 15,
823–833.

95
CHAPTER 7:

Snow Falls in the

Open Ocean
R.S. Lampitt

Historical Developments of the Concept

When I think of the floor of the deep sea, the single, over- We should not think that interest in the way mater-
whelming fact that possesses my imagination is the accu- ial is transported and modified has only recently
mulation of sediments. I see always the steady, unremitting, been kindled. In the second half of the ninteenth
downward drift of materials from above, flake upon flake, century, there was a vigorous debate among the
layer upon layer – a drift that has continued for hundreds giants of old (Jeffreys, Thomson, and Lohmann)
of millions of years, that will go on for as long as there are about the way in which the sediments of the oceans
seas and continents ... For the sediments are the materials were formed and if, as some believed, there was life
of the most stupendous snowfall the earth has ever seen ... in the deep sea, how it was sustained. While
Rachael Carson5 Wallich25 thought that such life was supported by
what is currently referred to as chemo-autotrophy,
Thomson22 thought particulate transport of shal-
Introduction low water macrophytes (seaweeds) and terrestrial
It is hard to improve on this inspiring prose, run-off were the keys to the food supply of the
describing so graphically as it does a process which deep-living animals. Jeffreys10, however, thought
continues to excite and tantalise marine scientists. that the dead remains of surface dwelling organ-
It is widely accepted that the means by which mate- isms would be an important source of their food.
rial is transported down to the abyssal depths and Lohmann14 made some surprisingly modern calcu-
the rates at which this occurs present a range of lations about the rate at which the sediments of the
fundamental questions. These are of direct interest ocean were formed by the deposition of planktonic
to a wide variety of scientists, from those interested material. He commented on the fact that near-bot-
in global biogeochemical cycling (see Chapter 11) tom water above the abyssal sea bed sometimes
to, for instance, those keen to understand the popu- contained a surprising range of thin-shelled phyto-
lation dynamics of the deep sea benthic fauna (see plankton species, some still in chains and with their
Chapter 13). fine spines well preserved. He deduced that they
In this brief review I acknowledge the technical must have been transported there very quickly from
and philosophical developments which have their near-surface habitat, and thought that the fae-
occurred over the past 120 years, but focus primar- cal pellets from some larger members of the plank-
ily on the progress which has been made during the ton (Doliolids, Salps, and Pteropods) were the like-
last 10, progress which has transformed our under- ly vehicles. His deductions may, in many cases,
standing of the processes involved in material flux. have been entirely correct and it remains a poor
The means by which small particles at the top of comment on our science that these observations
the ocean, with almost negligible sinking rates, are were largely ignored during the next few decades.
transformed into larger, rapidly sinking aggregates The descriptions of Rachael Carson5 soon found
which reach the deep-sea floor in only a few weeks their mark in the minds of a group of Japanese
continues to provide exciting insights into the com- oceanographers using the submersible observation
plex processes of material cycling in the oceans. chamber ‘Kuroshio’, suspended from a fisheries
In the language of the travel agent, the open training ship (Figure 7.1). The amorphous particles
ocean has become a smaller place, and the linkages they could see through the portholes were clearly
between the top sunlit zone and the dark, cold, not living and they coined the term ‘marine snow’
deep sea interior are now thought to be far closer to describe them21. The term is still only loosely
than previously imagined. Rachael Carson’s stu- defined, but is generally recognised to encompass
pendous snowfall 5 is no less awe-inspiring now immotile particles of diameter greater than 0.5 mm.
than it was to her in 1951, and without doubt the In the open ocean these are all biogenic and are
questions surrounding its elucidation are as thought to be the main vehicles by which material
demanding. sinks to the sea floor.

96
 7: Snow Falls in the Open Ocean

The submersible used by the Japanese oceanog- 7.1

raphers was a cumbersome device and did not per-
mit anything but the simplest of observations to be
made. They did, however, manage to collect some
of the material and reported that its main compo-
nents were the remains of diatoms, although with
terrestrial material present that provided nuclei for
formation.
In spite of these observations and the outstand-
ing questions surrounding material cycles in the
oceans, it was, until the late 1970s, a widely held
belief that the deep-sea environment received mate-
rial as a fine ‘rain’ of small particles. These, it was
assumed, would take many months or even years to
reach their ultimate destination on the sea floor.
The separation of a few kilometres between the top
and bottom of the ocean was thought sufficient to
decouple the two ecosystems in a substantial way,
such that any seasonal variation in particle produc-
tion at the surface would be lost by the time the
settling particles reached the sea bed. This now Figure 7.1 The submersible observation chamber
seems to have been a fundamental misconception. ‘Kuroshio’ as used to make observations on the distribu-
Part of the reason for this is the lack of understand- tion and characteristics of marine snow aggregates in the
ing of the role of marine snow aggregates. 1950s. The chamber was lowered on a cable from the
We now ask the most basic of questions about mother ship, providing a rather cramped view of the sea’s
interior and sea bottom through the small portholes. Two
this important class of material: What is marine
or three investigators were able to fit into the chamber,
snow, how is it distributed in time and space, and descending to depths of up to 200 m for several hours.
why is it of such significance? (Courtesy of Dr Masahiro Kajihara, Hokkaido University.)
A Vertical Profile throughout the upper mixed layer, but are located
Marine snow is found throughout the world’s at its base, a feature which is directly related to the
oceans in all parts of the water column. It is not rates of production and loss of the marine snow
uniformly distributed, either in space or time, but is particles in this highly dynamic part of the water
usually found in higher concentrations in the upper column.
water column and in the more productive regions
of the oceans. Although it had been suspected since The upper mixed layer (UML) factory
the early observations of Suzuki and Kato21 that The mixed layer at the top of the oceanic water col-
marine snow concentration decreased with increas- umn varies in thickness from up to a few hundred
ing depth, this has been confirmed only recently. metres in winter to a few tens of metres in the
The profiles now becoming available do not, how- spring and summer. It is subject to rapid changes in
ever, suggest a simple decrease. There is consider- light, heat, turbulence, nutrient concentration, and
able structure, undoubtedly related to the processes depth of mixing on the scale of hours, as well as
of production, destruction, and sinking. These are having distinct seasonal variations. This physical
all related to the physics, biology, and chemistry of forcing creates changes in the biological processes
the water column and of the particles themselves. which depend on them. Furthermore, it is highly
Figure 7.2 shows some examples of profiles variable in the spatial sense, producing a rapidly
from different parts of the world and using a vari- changing mosaic of physical, chemical, and biologi-
ety of techniques (Box 7.1). Bearing in mind the cal properties not found elsewhere in the water col-
strong seasonal variation which can occur even umn (Chapters 4 and 5).
well-below the upper mixed layer (see below) and It is here in the UML that the primary produc-
the different techniques employed to obtain these tion of material occurs as a result of phytoplankton
profiles, a common story seems to be emerging. growth (see Chapter 6). The cells thus produced
Apart from profiles near the continental slope, (mainly in the range 1–50 μm diameter) are imme-
where snow concentrations tend to increase near diately subject to attack from many other elements
the sea bed due to resuspension, there is generally a in the plankton community, but principally from
rapid fall in concentration over the top 100 m. the microplankton (20–200 μm diameter) and the
Peak concentrations are not, however, found mesozooplankton (0.2–20 mm body length). The

97
R.S. Lampitt

7.2 Volume concentration (p.p.m.)

0 200 400 0 200 400 0 200
0▼ 0 xx 0
▼
x ▼x▼x
▼ x▼x x
▼ ▼▼ xx
x ▼ x ▼xx x▼▼
x ▼ xxxxx
▼
▼▼▼
▼ ▼
▼▼▼ ▼ xx ▼ x x xxxxx
x ▼ ▼▼ x
▼ x x x
x xxxxxxx x
▼ ▼x x x
50 50 ▼ ▼▼ ▼
▼
▼
x x
x x 50 x xxxxxxx x
▼ x ▼
▼ x x xx
▼ xx
▼ ▼x▼ x xx ▼
x xxxx
x xx
x xx ▼▼▼
▼
x x x▼ xxxxxxx
x▼▼x x▼ x▼ xxx x
▼x ▼ x▼ x x
x ▼x x xxxx
▼ x x ▼x▼ x ▼ ▼ xxxxxx xx
100 x
100 xx ▼x 100 x xx x
xx xx x
x
▼ x x xxx x x
▼ xxx x x
xxxx x
x xxx x x
Depth (m)

▼ xx xxxx
xx xxx ▼ xxxx
x x
x xx xx
▼ x
xx x x [Chl-a] xxx
150 [Chl-a] 150 150 xxx x
x xx xx ▼ x
xxx
▼ x x
xxx
x xx x x▼ x
▼ xxx x x
x Atten. xxx x
x Atten. x
▼ x x x
200 xx x
200 x xxxx
x x▼
▼ 200 x
▼ xx
x
Sigma t
x xx
x xxx Sigma t x
x
Sigma t
▼ xx x
x x x▼ x
xxxx x
▼ xxxx
250 x 250 x x▼xx 250 x
x xx
x xx xx
▼
▼ xxx x
xx
x xx xx
▼ xx xx
x xx x
300 ▼x▼x 300 xxx xx
300

(a) (b) (c)

Volume concentration (p.p.m.)
0 2 4 6 8 0 10 20 30
0

500
Depth (m)

1000

1500
(d) (e)

Figure 7.2 Examples of the distribution of marine snow particles greater than 0.6 mm diameter, expressed as volume
concentration (p.p.m.). (a) 19 May 1990 (MSP 11), (b) 22 May 1990 (MSP 15), and (c) 26 May 1990 (MSP 19) from the
Northeast Atlantic at a water depth of 4800 m (Lampitt et al.12). Sigma t is a measurement of water density, [Chl a]
(Chlorophyll a) is the concentration of phytoplankton pigment, and Atten (attenuation) is a measure of the concentra-
tion of the smaller particles as determined by their effect on light transmission through the water. In terms of abun-
dance, the maximum concentration of particles at around 50 m depth is about 200/l, decreasing to a deep-water mini-
mum of approximately 30/l. (d) 22 November 1987 and (e) 28 January 1988 from Northwestern Gulf of Mexico at a
water depth of 1500 m on two occasions; the lines are 9-point running means26. In most instances there is a peak in
concentration near the surface, but not necessarily at the surface. Near the sea bed there are also elevated levels.
There appears to be a very wide range in concentrations and, although some of this may be related to differences in
technique, there can be considerable temporal variability at any one site (Figure 7.6); there are also likely to be large
regional differences reflecting the structure and dynamics of the biological communities in the different environments.

98
 7: Snow Falls in the Open Ocean

Box 7.1 How to Examine the Distribution in Time and Space

of Marine Snow Aggregates
In situ studies: enumeration, measurement, and collection
Marine snow is a highly variable and ephemeral commodity (Figure 7.2). It varies dramatically, not
only in its abundance, size, and sinking rate, but also in its origins, composition, and value as a food
source. One of the greatest and yet apparently simplest current requirements is to obtain good data
on its variability in time and space. Most attempts have used in situ photographic techniques. A vari-
ety of other sensors are now often attached to the cameras in order to measure other environmental
variables, such as water density, fluorescence (a measure of phytoplankton concentration), turbidity
(determined primarily by the smaller particles), and oxygen. Subaqua divers have also been used to
count particles, but because of the difficulty in seeing the smaller classes of marine snow and the bias-
ing influence of different ambient lighting conditions, divers are mainly reserved for collection and in
situ experimentation. Manned and unmanned submersibles are also used to observe and capture
marine snow. Figure 7.3 shows examples of some of the photographic and diver-operated techniques.
Both video and emulsion-based photography are used; a light source produces a collimated beam
which can then be viewed at right angles by the camera. This produces an image of a known water
volume with the particles displayed on a dark background. For long-term observations of temporal
changes, such devices have been deployed on moorings 11. Profiling instruments are generally
deployed from research ships20,26.
7.3b
7.3a 7.3c

F
A D B
C

C
A
E
B
E

B
D

7.3d
G Figure 7.3 A variety of photographic devices used to record the spatial and
temporal variation in marine snow aggregates in situ. (a) The Large Aggregate
H Profiling System (LAPS)8 with collimated light beam: 35 mm camera (A),
flashlight (B), CTD (C), fresnel lens (D), and transmissometer (E). A volume of
about 20 l is photographed and particles binned into six separate size cate-
gories from 0.5 mm to >3 mm. (Courtesy of Drs Gardner and Walsh, Texas A
A
& M University.) (b) The Atlantic Geosciences Centre Floc Camera Assembly
(FCA)20. This carries three cameras (A) and flash (B), and records particles
I >0.25 mm in a volume of 0.2 l. (Courtesy of Dr Syvitski, Bedford Institute of
Oceanography.) (c) In this case the device is associated with a variety of
D other instruments, such as a fluorometer to measure the concentration of phy-
toplankton, conductivity and temperature sensors, an echo-sounder (E), and a
B transmissometer to measure the concentration of the smaller particles. The
grey vertical tubes (F) on top of the device are sampling bottles activated by
the conducting wire on which the device is suspended. Particles greater than
0.5 mm are recorded in a volume of 40 l using 35 mm film. These are later
analysed on an image analyser (see Figure 7.6). (d) In contrast to (a)–(c), this
instrument is attached to a mooring to record temporal changes (Figure 7.6).
It has the same arrangement as in Figure 7.3(c) and the frames are similarly
analysed. Also shown are a buoy (G), vane (H), and battery pack (I).

99
 R.S. Lampitt

7.4a 7.4b

Figure 7.4 Examples of Types A and B marine snow aggregates. (a) Type A aggregate comprising an abandoned filter
net or ‘house’ of an appendicularian. It is identifiable by the presence of two in-current filters and a large U-shaped
internal filter surrounded by an envelope of particle-studded mucus. These are typically several centimetres in diame-
ter. (Courtesy of Dr Alldredge, University of California.) (b) Type B aggregate comprising living chain-forming diatoms
(scale bar = 1 cm). (Courtesy of Dr Gotschalk, University of California.)

complex food web which thus develops is being the base of the UML where internal waves, wind
intensively researched through experimentation and driven sheer, and tidal sheer are pronounced15. In
numerical modelling (e.g., Fasham et al. 7 ). all cases, significant concentrations of snow can
However, the focus of this chapter is the material only develop if there are sufficiently high concen-
lost from this UML community, or at least in the trations of the component particles. Until very
process of being lost. recently there seemed to be not enough particulate
Various parts of the food web produce material matter in the water column for this to occur.
which is less attractive to consumers. Some of this However, the problem may have been related to an
material is easily considered as a waste product, underestimate of the abundance of the component
such as faeces, senescent phytoplankton cells, particles. Recent observations show that at least
gelatinous feeding webs, and zooplankton moults. one class of particles are naturally transparent
Marine snow can be divided conveniently into two unless stained in the laboratory, and yet they are
main classes based on the origin of the compo- intimately connected to the development of aggre-
nents: Type A is derived from the gelatinous houses gates. These are the transparent exopolymer parti-
constructed by some zooplankton species and cles (TEP), which are formed from the polysaccha-
mucous feeding webs used by others, and Type B is rides excreted by living phytoplankton and bacteria
derived from small component particles such as the (Figure 7.5d)16. A further requirement for aggrega-
waste products mentioned above (Figures 7.4 and tion is that the particles must be sufficiently sticky,
7.5). The mechanisms by which the snow is formed as are the TEPs, to aggregate when they do collide3.
clearly depend on its class, Type A starting as rela- The material produced in the UML reaches the
tively large particles which remain the dominant bottom of this layer both by mixing processes and
component. They are probably not much affected by gravitation settling; once there, the physical and
by the physics of the water column in their creation biological processes conspire to aggregate the mate-
and growth. rial into yet larger particles to produce the charac-
Type B snow particles are likely to develop in teristic peak at this depth.
response to specific biological, chemical, and physi- In the case of Type A snow particles, the mecha-
cal conditions; it is the nature of these conditions nisms of production and, indeed, loss are quite dif-
which we are only now starting to appreciate. ferent and depend on the presence of the particular
Aggregates can be formed by a variety of processes zooplankton species which produce these struc-
which cause collisions, such as differential settling tures.
during which one particle sinks at a faster rate than As mentioned above, the distributions shown in
the one below and collides with it. Turbulent mix- Figure 7.2 do not, of course, simply reflect the rates
ing also increases collision rates between particles, of production, but also the rates of loss either by
a situation which may be particularly important at consumption or sinking. These are discussed below.

100
 7: Snow Falls in the Open Ocean

7.5a 7.5b

7.5c 7.5d

The long descent

Once it has left the upper reaches of the ocean and,
in particular, has passed through the seasonal ther-
mocline at the base of the mixed layer, settling
material experiences much shallower gradients in
its physical, chemical, and biological environment.
Differential settling still occurs and some de novo
production of Type A snow particles still proceeds.
One might think that transformations would take
place at a leisurely pace. This is not, however, like-
ly as long as there are significant numbers of zoo-
plankton and nekton migrating from the surface at
night and to depths as great as 1000 m by day (see
Chapters 14 and 15). This phenomenal diel migra-
Figure 7.5 (a) Miscellaneous aggregate containing a vari- tion, demanding large energy expenditure by the
ety of planktonic particles. (Courtesy of Dr Alldredge, migrating organisms, may have a significant effect
University of California.) (b) Scanning electron micro- on the vertical transport of sinking material as the
graph of snow particles dominated by diatoms. In this migrants consume food primarily in the upper
case they are primarily of the genus Chaetoceros. The water, where it is most abundant, but excrete,
spherical cell body on the right hand side is about 10 μm respire, defecate, and die throughout the water col-
in diameter. (Courtesy of Dr Gotschalk, University of umn. Furthermore, from the perspective of marine
California.) (c) In situ marine snow aggregate dominated snow, these relatively large organisms are the ones
by diatoms; the aggregate is about 1 mm in diameter.
(Courtesy of Dr Gotschalk, University of California.) (d)
most likely to feed on the sinking material.
Diatom aggregate after staining with Alcian blue to show Recent observations of long-term changes in
transparent exopolymeric material gluing cells together marine snow at 270 m in the northeast Atlantic
(scale bar = 100 mm). (Courtesy of Dr Alldredge, show dramatic changes in their abundance and vol-
University of California.) ume concentration. Peaks correspond to elevated

101
 R.S. Lampitt

7.6a April May June July August Sept 7.6b

Volume concentration (p.p.m.)

11th June–
Volume concentration (p.p.m.)

30th June
Size
5 pt running classes (mm)
means
(3.4 days) 6.3–9.9
3.9–6.3
2.5–3.9
1.6–2.5
1.0–1.6
0.6–1.0

→
Total particulate volume

Day number 1990 Time of day

Figure 7.6 (a) Temporal variation in marine snow volume concentration (night values only to remove diel periodicity)
at 270 m depth in the northeast Atlantic(48°N 20°W), derived from photographs taken using the marine snow camera
system, Figure 7.3(d). The photographs obtained with this instrument are analysed by computer to give a maximum
and minimum dimension of each particle in the frame. The volume of each particle is calculated and assigned to one
of six size categories. Dramatic peaks in concentration are caused primarily by the larger size classes and occur soon
after pulses in productivity of the phytoplankton in the surface water layer above11. (b) Accumulated data for volume
concentration of marine snow at 270 m depth during the period 11–30 June 1990. During this period concentration
levels were significantly higher between the hours of 05.00 and 14.00, identified by the dotted vertical lines.

7.7 (a): 1/5 (b): 15/6

(c): 22/6 (d): 29/6

Figure 7.7 Examples of time-

lapse photographs of the sea
bed at 4025 m depth off the
European continental slope,
taken using the ‘Bathysnap’
instrument. Between 1 May
and 15 June there is little
change in its appearance, but
during the rest of the summer
there is a progressive increase
(e): 14/7 (f): 10/8
in the amount of material cov-
ering the sea bed, visible as
dark patches obscuring the
underlying sediment. Between
14 July and 10 August there is
a progressive decrease in this
covering. The mound in the
centre of the frame is 18 cm
across.

102
 7: Snow Falls in the Open Ocean

Figure 7.8 The photographs in Figure 7.7 were used to

7.8a
derive a semi-quantitative measure of the material lying

Quantity of phytodetritus on sea bed

on the sea bed on each frame (a); the green band is to
highlight the trend over the year. Also shown in (a) (verti-
cal bars) is a semi-quantitative estimate of the degree of
resuspension in each frame, which reduces visibility of
the sea bed. It can be seen that resuspension only occurs

Resuspension
after the deposition of phytodetritus and is not a constant
feature. The letters A–F indicate the times at which the
photographs in Figure 7.7 were taken. (b) The current
speed which shows that it is only when currents exceed May June July August 7.8b
about 7.5 cm/s that a significant resuspension occurs. It is
thought that the material is not resuspended very high in
the water column as, from other Bathysnap deployments,
resuspension causes an initial loss of material from the

Current speed (cm/s)

sea bed, which is followed within only a few tens of min-
utes by its redeposition. The current meter rotor stalls at
speeds less than 2 cm/s, as indicated. The tidal cycle is
clearly evident, but during late July and early August the
minimum of the tidal cycle is above the stall speed of the
rotor, indicated by red shading. These are particularly
energetic periods. May June July August

levels of primary productivity in the overlying the established view of weak linkages between the
water and precede peaks in material flux deter- top and bottom of the oceans. Repeated sampling
mined by sediment traps 3000 m below. As can be of the deep-sea sediment, particularly the loose sur-
seen in Figure 7.6(a), the values of volume concen- face layer, was technically very difficult and very
tration in the spring are more than ten times those demanding on ship time. It was therefore fortunate
later in the year; this is mainly due to the contribu- that in the 1980s, time-lapse photographic tech-
tion of the larger-size categories. An additional fea- niques using an instrument called the ‘Bathysnap’
ture of this data set is that there is a distinct diel became available for use in an area which was sub-
variability; in Figure 7.6(b) data from one period ject to very strong seasonal depositions of phytode-
are presented to demonstrate this signal. It is trital material. In Figure 7.7, recently deposited
tempting to relate this signal to the diel migration material can be readily seen on the sea bed as dark
of most of the larger members of the planktonic patches of loose fluffy material.
biosphere, but at present the mechanism is The image of the material above the sediment
unknown. surface is a reflection of supply and demand. It is
Although much of the material sinks at rates of the difference between the supply of settling parti-
a few tens to hundreds of metres per day, some of cles from above and the demand by benthic com-
it is modified or remineralised in support of the munity members, which either ingest or bury these.
mid-water community of bacteria and zooplank- Under less productive parts of the ocean, such as
ton. Other components dissolve during the descent, the Madeira Abyssal Plain, no detrital layer is visi-
especially calcite particles when below the calcium ble on benthic photos; this not only reflects the
compensation depth (see Chapter 13), but it seems lower overall level of productivity, but also the
that in spite of this long descent, some of even the reduced variability in particle flux. Here the supply
delicate structures of phytoplankton spines are still is more constant and the benthic community con-
apparent when the deep-sea floor is reached, just as sumes or buries it at the same rate as it arrives on
found by Lohmann14 back in 1908. the sea floor.
Although difficult to quantify on the pho-
The benthic experience tographs, an approximation can be made by mea-
The ultimate repository for much of the material suring the density of the film at one particular loca-
which sinks into the deeper parts of the water col- tion on each frame of the time series (Figure 7.8).
umn is the deep-sea floor; it is here we now look We can now collect the material using a remark-
for evidence of a close link with the surface of the able coring device, the Scottish Marine Biological
ocean and indications about the role of marine Association (SMBA) multiple corer. This has the
snow. ability to collect sediment samples with virtually no
Rapid changes in material flux in deep mid- disturbance to the light interfacial layer, and we
water and apparent seasonal reproduction by some can state with certainty that the dark patches seen
of the larger benthic fauna posed a serious threat to on the photographs are primarily of phytoplankton

103
 R.S. Lampitt

7.9 Figure 7.9 Plastic core tube from the SMBA multiple
corer with sediment taken in May 1981. The core tube is
56 cm in diameter and a layer of phytodetritus can be
clearly seen lying above the less granular sediment. This
core was taken on the European continental slope
(Porcupine Seabight) at a depth of 2000 m.

the phytodetrital layer is moved about and resus-

pended by the near-bottom currents (Figure 7.8),
fed upon by the benthic biota and buried by some
of the larger members of this community. In Figure
7.11, visual examples of the benthic fauna’s res-
ponse are shown. Here, at last, was an explanation
for the apparent seasonal reproduction of some of
the larger benthic species24, a response that has
now been found in the macrofaunal and meiofau-
nal communities.

origin (Figure 7.9). We also know that the phyto- Properties and Characteristics of
plankton species composition is oceanic, demon- Marine Snow
strating that it does not slump off the adjacent con- The physical properties of marine snow particles are
tinental shelf. Finally, in a water depth of 4000 m as diverse as the origins of the material from which
the peak deposition occurs only 4 weeks after the they are derived. Some are sticky, some are fragile,
expected phytoplankton bloom in the overlying some are porous, and some sink fast. Their biologi-
water. This all gives strong evidence that this mater- cal properties are also diverse, reflecting their ori-
ial is transported vertically and rapidly to the deep- gins as mentioned above, but also displaying widely
sea floor. On closer inspection of some of the benth- varying rates of biological processes; some have
ic time-series photographs, one can even discern the unusually active bacteria within them and others
nature of the particles arriving to form this carpet. have rapidly growing phytoplankton. In all these
Figure 7.10 shows an enlargement of one part of a cases there are aggregates which display entirely the
pair of photographs of a series taken at 2600 m opposite property, and generalisations are hard to
depth on the European continental slope. They are make. The methods used to collect the particles are
separated in time by only 16 minutes, but during also far from simple (Box 7.2).
this interval an aggregate of several millimetres
diameter arrived on the sea bed. It remained there Physical and biological properties
and, in common with many other similar particles, As indicated above, the physical characteristics of
gradually degraded over the subsequent few days. stickiness and fragility determine whether marine
In those productive regions, once on the sea bed snow aggregates tend to grow in size, and their

7.10a 7.10b

Figure 7.10 Enlargements of consecutive ‘Bathysnap’ frames of the sea bed in the Porcupine Seabight (northeast
Atlantic continental slope) at a depth of 2600 m in May 1981. (a) The arrow shows the spot where 16 minutes later (b)
a 10 mm marine snow aggregate (grey) arrived between frames. Over the subsequent few days this particle, and many
others like it, degraded and coalesced to form a continuous layer.

104
 7: Snow Falls in the Open Ocean

7.11a 7.11b

7.11c 7.11d

Figure 7.11 Examples of the response of the large benthic fauna to a layer of
phytodetritus. (a) The holothurian Benthogone rosea of body length 15 cm at a
depth of 2008 m in the Porcupine Seabight in June 1982. (b) 30 minutes later
the specimen has left a faecal cast to demonstrate its satisfaction with the phy-
todetrital meal. The echinoid Echinus affinis tends to move into depressions in
the sediment at times when phytodetritus is present, as in (c), one-third up in the
centre, and (d), three-quarters to the right, both found in the Porcupine Seabight
at a depth of 2000 m in May 1982.

105
R.S. Lampitt

Box 7.2 How to Collect and Study Marine Snow Aggregates

Sediment traps can be used to collect settling particles, but it is unlikely that the integrity of the indi-
vidual particles so-collected is maintained, even when they are immersed in a preservative as is usual-
ly the case. Because of the fragility of some types of marine snow, and because of their low abun-
dance and high sinking rates, traditional water-bottle techniques are usually of little value. In spite of
the inherent difficulties of subaqua diving (limitations of depth, sea state, and personnel), it remains
one of the best methods to obtain undamaged snow particles for experimentation (Figure 7.12).
Large bottles have been developed by some groups with the specific aim of collecting marine
snow, Figure 7.13(a). This 100-litre water bottle is closed at depth by sliding a weight, the ‘messenger’,
down the supporting wire. On recovery it is left on deck for several hours to allow the snow aggre-
gates to settle to the bottom of the vessel. The top 95 litres are drained off and the bottom chamber
removed for collection of the particles in the laboratory, Figure 7.13(b). This has proved to be a useful
and successful method where diving is not possible.

7.12

Figure 7.12 Subaqua diver behind a very large marine snow aggregate in the
northern Adriatic sea. (Courtesy of Dr Stachowitsch, University of Vienna.)

7.13a 7.13b

Figure 7.13 The large-volume marine snow

catcher, the ‘snatcher’, for collecting undamaged
samples of the aggregates. (a) The entire device
just prior to deployment from RRS Discovery. (b)
The lower chamber of the ‘snatcher’ in the cold
laboratory on board ship. Aggregates are being
collected from the base of the chamber using a
wide-bore pipette to avoid damage. The lights
are directed through the transparent sides of the
chamber for ease of collection.

106
 7: Snow Falls in the Open Ocean

7.14a 7.14b

7.14c 7.14d

Figure 7.14 Scanning electron micrographs of marine snow particles collected off Baffin Island, showing a wide range
of morphologies and composition. (a) Large mucoid aggregate collected at 30 m depth. (b) ‘Stringer’ collected at 1 m
depth. (c) Mixed agglomerate dominated by biogenic material collected at 100 m depth. (d) Aggregate dominated by
mineral matter collected at 5 m depth. (Courtesy of Dr Azetzu-Scott, Bedford Institute of Oceanography.)

excess density over that of the surrounding water during this stage that the concentration of TEP
controls the speed with which they descend increases16 and the particles probably become more
through the water column. The proportion of free sticky. With increasing size of aggregate, excess
water in an aggregate, its porosity, determines how density tends to decrease1, so that the increase in
fast its internal environment changes in response to sinking rate is not as pronounced as might be
varying external conditions; for example, during expected. Simultaneously, and when still in the
sinking, porosity controls the rate of exchange of euphotic zone, primary production may be
water within the aggregate with that outside. enhanced within the aggregate due to an efficient
Porosity also influences the rate at which small par- use of the ammonia released within them9. In fact,
ticles, such as clays, accumulate on the aggregate production may proceed so fast that bubbles of free
and the rate at which surface-active elements, such oxygen are created which, in turn, cause the snow
as thorium, are adsorbed onto the snow. Rates of to rise in the water17. There is also good evidence
adsorbtion and desorbtion are of considerable rele- that the bacteria19 and protozoa13 find the micro-
vance to particle-cycling models, which use adsorb- environment within the aggregates attractive, some-
tive radioisotopes, such as 234Th, as proxies of solid times producing anoxic microzones18. This may
material (see later). reduce sinking rates if free gases are produced. The
The physical properties of an aggregate are activity of the microbiota may change, however,
clearly closely tied to its chemical and biological during the descent of the aggregate; it now seems
components. However, as a general pattern, Type B that this activity may be inhibited by increasing
aggregates develop as bloom conditions are reached pressure23.
and nutrients become limiting (see Chapter 6). It is The composition (Figure 7.14) of Type B aggre-

107
 R.S. Lampitt

7.15

Figure 7.15 Marine snow aggregates collected using the ‘Snatcher’ (Figure 7.13) during May 1990 in the northeast
Atlantic over the Porcupine Abyssal Plain from depths of either 45 m (A, C, and D) or 300 m (B). The composition of
the aggregates can best be observed under the microscope using ultraviolet excitation, either after staining with acri-
dine orange (A and B) or relying on the autofluorescent properties of the material (C and D). The aggregates contain a
wide range of component particles (d, lorica of the tintiniid Dictyosysta elegans; p, pennate diatom; z, zooplankton
carapace; c, small chlorophyte; m, bacterial matrix). In this instance, the tintiniid lorica tended to be physically dam-
aged when found further down in the water column [scale bars = 100 μm (A and B) and 50 μm (C and D)]. (Courtesy
of Dr C Turley, Plymouth Marine Laboratory.)

Quantitative Study of Particulate Flux

gates generally reflects that of the suspended parti- In order to understand the cycling of biogeochemi-
cles in the euphotic zone. As such, the aggregates cal components of the oceans, data must be
present at any one time at a particular location are obtained about the time-varying fluxes of the prin-
of a similar composition2, whether they be domi- cipal compounds, elements, and particle types at
nated, for instance, by diatoms, faecal material, different depths in the sea. The concentrations and
coccolithophores, or flagellates (Figure 7.15). size distributions of particles can only give a gener-
Whatever the origin of the particles, and however al indication of fluxes, and then only after some
porous the aggregates, it is safe to assume that the major assumptions on sinking rates. There are few
environment of a free-living organism or particle ways in which particle flux can be measured; these
changes dramatically when it becomes incorporated can be divided broadly into direct measurements
into an aggregate. Not only will its chemical envi- using sediment traps and numerical modelling
ronment change, but so too will its sinking rate and, approaches, frequently based on the distributions
in the case of an organism, its ability to control its of various chemicals and particles.
depth by, for instance, buoyancy modification. An The particle interceptor trap, or sediment trap, is
organism associated with an aggregate will also be a device akin to the rain gauge, having a funnel
subject to quite different predator pressures from its into which falling particles are collected. In the case
free-living counterparts, as discussed below. of modern time-series traps (Figure 7.16), a rotat-

108
 7: Snow Falls in the Open Ocean

7.16a 7.16b

7-day collection periods

except where indicated

April May June July Aug

ing carousel moves a fresh collecting jar under the

bottom of the trap at predetermined times, such
that data may be obtained over several months’
duration with a resolution of a week or so. There Figure 7.16 (a) Time-series sediment trap photographed
are several problems related to the accuracy with just after recovery. Settling material enters the yellow
which sediment traps measure particle flux, but cone and from there into the white collecting cups
these devices provide the only means of determin- below. The trap had been at a depth of 3200 m for the
ing directly the flux of material at a particular previous 6 months, in a water depth of 4800 m, and the
collected material can be clearly seen in the cups (previ-
depth; furthermore, they are the only way in which
ously filled with a preservative, formaldehyde). Such
the material responsible for the downward flux can devices are suspended on a supporting wire, with a bal-
be collected, described, and analysed. last weight on the sea bed attached to the wire by an
In 1978, soon after the sediment trap had acoustically operated release mechanism. Buoyancy
become a reliable oceanographic technique, a long spheres at the top of the wire carry the entire mooring to
time-series of measurements was initiated off the surface after the release has been activated. (b)
Bermuda at a depth of 3200 m (Figure 7.17). This Sample cups from deep-water sediment traps deployed at
remarkable series, still continuing, demonstrates a two locations in the northeast Atlantic (7 day collection
strong seasonal signal of deposition with a peak in periods except where indicated). The increased flux from
the early part of the year. This is certainly not a the end of May resulted from elevated productivity in the
surface waters about 4 weeks previously. (Courtesy of Dr
universal pattern, due to the large physical and bio-
Williamson, University of East Anglia.)
logical differences between oceanographic regimes.
In this case, the depositional peak lags behind a

7.17
Dry mass flux (mg/m2 d)

Figure 7.17 Variation in parti-

cle flux over an 8-year period
at 3200 m in the Sargasso Sea.
This is the longest such record
from any region and shows the
seasonal cycle in particle flux,
but with some distinct irregu-
larities during the period
1981–1983. At this location,
elevated productivity in the
surface is caused by a deepen-
ing of the mixed layer, which
introduces new nutrients.
About 3 weeks later particle
flux at 3200 m became
enhanced (from Deuser6). Year

109
R.S. Lampitt

peak in the UML by about 6 weeks; the reason is material pool which is influenced by this micro-
that the deepening of this layer introduces nutrients environment, a major challenge to our understand-
into the upper ocean to produce a phytoplankton ing of material cycling in the oceans.
bloom. The data also demonstrate significant inter-
annual variability, a feature which has been found Role as a food source
in most studies for which there are sufficient data. Considering now the next spatial scale, that of the
Explanations are, however, not always readily planktonic feeder, if, as stated above, small parti-
available. cles such as cyanobacteria become incorporated
In the upper ocean, zooplankton tend to swim into aggregates, they will be available to quite dif-
into the sampling cups, thus contaminating the ferent predators than when they are free living.
material by defecating and/or dying in them. This is Observations by divers have, on several occasions,
particularly unfortunate as it is here where the commented on the physical proximity of some zoo-
sharpest gradients in flux occur and where there is plankton species and marine snow, which suggests
the greatest requirement, from the modelling per- that the zooplankton are feeding on the snow parti-
spective, for good quality data. Most of the organic cles or a component of them. Preliminary experi-
carbon lost from the UML in spring and summer is mental evidence is now giving support to this con-
remineralised before it reaches a depth of 1000 m, clusion, extending the range and diversity of species
but at present it is not clear how much is lost from which feed on snow aggregates (Figure 7.18). One
the UML. of the important conclusions from this with regard
Indirect methods of monitoring particle flux to biogeochemical cycling is that it represents a
usually demand measurements of the vertical distri- food-chain short-cut in which small particles can be
bution of the dissolved and particulate phases of consumed by species normally restricted to a much
certain radionucliides which are particle reactive. larger size of food. Once again, if models of bio-
Most promising in this regard is 234Th, which is geochemical cycling are to be realistic, they must
produced from dissolved 238U at a known rate. not only consider the bulk properties of the sea
Using a box model, uptake and removal rates of water, but also the micro-environments within it.
234
Th can be calculated and carbon fluxes derived4.
These methods do not, of course, provide material Role as a transport vehicle
for examination and at present frequently give The vast majority of material produced in the UML
rather different conclusions from those using the is recycled there. As described above, there is a
direct methods. sharp reduction in the vertical flux over the top few
hundred metres, reflecting the activity of the organ-
Significance of Marine Snow isms which feed on the sinking material. Particles
in Biogeochemical Cycling which sink slowly are particularly susceptible to
There is an ever-increasing interest in the material this recycling, as their residence times are long.
cycles of the oceans and the relevance of these to Although some marine snow particles have low or
global cycles. The significance of marine snow in even negative sinking rates, those which leave the
these cycles may be considered in several ways, euphotic zone sink at high rates, between several
depending on the physical dimensions of interest. metres per day and several hundred metres per day.
Marine snow may be considered as an environment The close temporal coupling between surface
in which biogeochemical processes occur (scales processes and the deep sea, the undegraded nature
ranging from micrometres to centimetres), as a of material collected in deep sediment traps, and
resource on which the plankton and nekton can the visual observations of snow-sized particles
feed (scales from millimetres to metres), and as a arriving on the deep-sea floor all give strong sup-
vehicle by which material is transported down port to the contention that they are the principle
through the water column (scales from decimetres vehicles by which material is transferred to depth.
to kilometres). This is important in terms of both the transport of
The micro-environment mass and of key compounds, such as those contain-
ing organic carbon, trace metals, and radioisotopes
The micro-environment of a marine snow aggre-
adsorbed onto the surfaces of particulate material.
gate has already been described from the perspec-
tive of its properties. With regard to biogeochemi-
cal cycling, the conclusion is that the bulk proper- Conclusion
ties of sea water (nutrient, oxygen, etc.) may have Marine snow is a class of large inanimate particles
little bearing on the way material is modified if the found throughout the world’s oceans. Fragile, and
micro-environment of the marine snow aggregate occurring at low abundances, these particles have
contains a significant proportion of sea water. It is, proved to be difficult to record and collect, but are
however, still an open question as to the size of the now thought to be the principal vehicles by which

110
 7: Snow Falls in the Open Ocean

7.18a 7.18b

Figure 7.18 Living zooplankton of the genus Oncaea feeding on marine snow aggregates. (a) This specimen, of body
length 3 mm, was studied in the laboratory for 3 days during which it made repeated visits to an aggregate comprising
cyanobacteria, phytoplankton, and microzooplankton remains (Type B). During this period it produced a large number
of its own faecal pellets. (b) In situ specimen of about 1.5 mm body length associated with an aggregate derived from a
discarded larvacean house (Type A; courtesy of Dr J. King, University of California).

material is transported from the site of primary 4. Buesseler, K.O., Bacon, M.P., Cochran, K., and Livingston,
production in the upper sunlit zone of the water H.D. (1992), Carbon and nitrogen export during the
JGOFS North Atlantic Bloom Experiment estimated
column to the deep sea. Due to their rapid sinking from 234Th: 238U disequilibria, Deep-Sea Res., 39,
rate, these large aggregated particles link the top 1115–1137.
and bottom of the ocean in a much closer temporal 5. Carson, R. (1951), The Sea around Us, Oxford University
sense than had been thought the case; they provide Press, Oxford, 230 pp.
6. Deuser, W.G. (1987), Variability of hydrography and parti-
an explanation for apparent seasonal changes in cle flux: Transient and long term relationships, in
the appearance of the deep-sea floor and for the Particle Flux in the Ocean, Degens et al. (eds),
seasonal reproduction of some species of deep-sea Mitt.Geol.-Palaont., Inst.U Hamb., pp 179–193.
benthic animals. It now seems that the rates of pro- 7. Fasham, M.J.R., Ducklow, H.W., and McKelvie, S.M.
duction and destruction of marine snow are high so (1990), A nitrogen-based model of plankton dynam-
ics in the oceanic mixed layer, J. Marine Res., 4,
as to facilitate diurnal changes in the particle pool 591–639.
well below the UML of the ocean, a feature which 8. Gardner, W.D. and Walsh, I.D. (1990), Distribution of
will have far-reaching implications for our under- macroaggregates and fine grained particles across a
standing of particle cycling. continental margin and their potential role in fluxes,
Deep-Sea Res., 37, 401–411.
There are still many unanswered questions sur- 9. Gotschalk, C.C. and Alldredge, A.L. (1989), Enhanced pri-
rounding the mechanisms of production and mary production and nutrient regeneration within
destruction of marine snow particles, but it is aggregated marine diatoms, Marine Biol., 103,
expected that, with the techniques now available, 119–129.
major advances will be made over the next decade. 10. Jeffreys, J.G. (1869), The deep-sea dredging expedition of
H.M.S. ’Porcupine’, Nature, 1, 135–137.
11. Lampitt, R.S., Hillier, W.R., and Challenor, P.G. (1993a),
Seasonal and diel variation in the open ocean concen-
References tration of marine snow aggregates, Nature, 362,
1. Alldredge, A.L. and Gotschalk, C. (1988), In situ settling 737–739.
behaviour of marine snow, Limnol. Oceanogr., 33, 12. Lampitt, R.S., Wishner, K.F., Turley, C.M., and Angel,
339–351. M.V. (1993b), Marine snow studies in the Northeast
2. Alldredge, A.L. and Gotschalk, C. (1990), The relative con- Atlantic: distribution, composition and role as a food
tribution of marine snow of different origins to bio- source for migrating plankton, Marine Biol., 116,
logical processes in coastal waters, Contint. Shelf 689–702.
Res., 10(1), 41–58. 13. Lochte, K. (1991), Protozoa as makers and breakers of
3. Alldredge, A.L. and McGillivary, P. (1991), The attachment marine aggregates, in Protozoa and Their Role in
probabilities of marine snow and their implications Marine Processes, Reid, P.C., Turley, C.M., and
for particle coagulation in the ocean, Deep-Sea Res., Burkill, P.H. (eds) (NATO ASI Series G: vol 25),
38, 431–443. Springer-Verlag, Berlin, 506 pp.

111
R.S. Lampitt

14. Lohmann, H. (1908), On the relationship between pelagic 21. Suzuki, N. and Kato, K. (1953), Studies on suspended mate-
deposits and marine plankton, Int. Rev. Ges. rials (marine snow) in the sea. Part 1. Sources of
Hydrobiol. Hydrogr., 1(3), 309–323, in German. marine snow, Bulletin of the Faculty of Fisheries of
15. Lueck, R.G. and Osborn, T.R. (1985), Turbulence measure- Hokkaido Univ., 4, 132–135.
ments with a submarine, J. Phys. Oceanogr., 15, 22. Thomson, C.W. (1873), The Depths of the Sea. An account
1502–1520. of the general results of the dredging cruises of
16. Passow, U., Alldredge, A.L., and Logan, B.E. (1994), The H.M.S. Porcupine and Lightning during the summers
role of particulate carbohydrate exudates in the floc- of 1868, 1869 and 1870 under the scientific direction
culation of diatom blooms, Deep-Sea Res., 41, of Dr Carpenter, F.R.S., J.Gwyn Jeffreys F.R.S. and
335–357. Dr Wyville Thomson, F.R.S., Macmillan, London.
17. Riebesell, U. (1992), The formation of large marine snow 23. Turley, C.M. (1993), The effect of pressure on leucine and
and its sustained residence time in surface waters, thymidine incorporation by free-living bacteria and
Limnol. Oceanogr., 37(1), 63–76. by bacteria attached to sinking oceanic particles,
18. Shanks, A.L. and Reeder, M.L. (1993), Reducing micro- Deep-Sea Res., 40, 2193–2206.
zones and sulphide production in marine snow, 24. Tyler, P.A. (1988), Seasonality in the deep sea Oceanogr.
Marine Ecol. - Prog. Ser., 96, 43–47. Mar. Biol. Annu. Rev., 26, 227–258.
19. Simon, M., Alldredge, A.L., and Azam, F. (1990), Bacterial 25. Wallich, G.C. (1862), The North Atlantic Seabed: compris-
carbon dynamics on marine snow, MEPS, 65(3), ing a diary of the voyage on board H.M.S. Bulldog, in
205–211. 1860; and observations on the presence of animal life,
20. Syvitski, J.P.M., Asprey, K.W., and Heffler, D.E. (1991), and the formation and nature of organic deposits, at
The floc camera: A three-dimensional imaging system great depths in the ocean, Van Voorst, London.
of suspended particulate matter, in Microstructure of 26. Walsh, I.D. and Gardner, W.D. (1992), A comparison of
Fine-Grained Sediments, Bennet, R.H., Bryant, W.R., aggregate profiles with sediment trap fluxes, Deep-
and Hulbert, M.H., Springer-Verlag, New York, pp Sea Res., 39A, 1817–1834.
281–289.

112
CHAPTER 8:

The Evolution and Structure

of Ocean Basins
R.B. Whitmarsh, J.M. Bull, R.G. Rothwell, and J. Thomson

Plate Tectonics Convergent plate boundaries are often marked by

Plate tectonics is the paradigm, developed during oceanic trenches, where the oceanic lithosphere is
the 1960s, by which the majority of large-scale subducted into the asthenosphere (when two conti-
topographic features on the Earth’s surface can be nents collide a mountain range results instead). At
explained. Plate tectonics also explains the kinemat- transform faults, adjacent plates slide past each
ic and dynamic behaviour of the outer parts of the other and material is neither created nor destroyed.
Earth in terms of rigid lithospheric plates. These There are a dozen major plates, most of which
plates, which comprise the uppermost mantle and in contain both oceanic and continental lithosphere,
most cases both continental and oceanic crust, ride which move relative to each other at
on a relatively weak viscous asthenosphere and 10–200 mm/yr (Figure 8.3).
move relative to each other over the Earth’s surface The movement of the lithospheric plates over a
(Figure 8.1). The plate motion is driven by mantle spherical Earth can be described by Euler’s
convection, the main mechanism by which heat, Theorem. This states that the relative motion
derived from radioactive decay, is transferred from between two plates can be defined by a rotation
the Earth’s deep interior to the surface. about an axis passing through the centre of the
One of the axioms of plate tectonics is that the Earth and intersecting the surface at the Euler
plates are strong and rigid, with deformation poles. The most accurate technique for determining
occurring only at their margins, where the plates the current Euler pole for two adjacent plates uses
interact. This is consistent with the fact that most the locus of the oceanic transform faults between
of the Earth’s major topographic features occur at them (Figure 8.4). Such faults follow the traces of
the past or present edges of plates and most seismic small circles centred upon the Euler pole; they are
(Figure 8.2) and volcanic activity is found at active mapped using bathymetry, side-scan sonar images,
plate boundaries. There are three types of plate and maps from satellite altimetry (see Box 8.1).
boundary (Figure 8.1). Accretionary plate bound- Thus, the Euler pole can be determined by the
aries (Chapter 10) are elongate ridges where new intersection of great circles drawn normal to the
oceanic crust is formed; subsequently, the crust is transform faults. Statistical techniques are used to
carried away from the ridge axis by sea-floor constrain the likely position of the Euler pole and
spreading and subsides as the lithosphere cools. to give an estimate of its uncertainty.

Figure 8.1 The principal fea- (c) (a) (b) 8.1

tures of plate tectonics, includ-
ing (a) accretionary (divergent),
(b) convergent, and (c) trans-
form plate boundaries. Arrows
on the lithosphere represent the
relative motion between pairs Lithosphere Lithosphere
of plates and arrows in the
asthenosphere represent com-
Asthenosphere
plementary flow in the mantle.
The lithosphere is typically
100 km thick; the horizontal Mesosphere
extent of the features shown is
often at least tens or hundreds
of kilometres (after Isacks et
al.19; © American Geophysical
Union).

113
 R.B. Whitmarsh, J.M. Bull, R.G. Rothwell, and J. Thomson

8.2

Figure 8.2 Global earthquake activity superimposed on a topographic map of the world (based on a cylindrical equidis-
tant map projection). The epicentres of earthquakes with body-wave magnitudes greater than five are shown as coloured
dots, colour-coded according to the focal depth of each earthquake (<50 km, red; 50–100 km, yellow; 100–300 km,
green; >300 km, blue). Note how most epicentres are concentrated in bands which delineate the plate boundaries (cour-
tesy of the National Oceanographic and Atmospheric Administration/National Geophysical Data Center).

8.3 0˚ 90˚ 180˚ –90˚ 0˚

60˚ 23 60˚

Eurasian 72 North American

11
87 28
25
95 Yellowstone
10 32 49 57
30˚ Ar 30˚
ine

ab Hawaii
ian 34 26
Philipp

73
18 91 Caribbean
African o C
17 110 cos
78
0˚ Pacific 0˚
66 62
South
37 72 80 161 American 38
Reunion 94
23 Indian MacDonald Nazca
92 106
–30˚ 94 185 25 –30˚
40
74 65 80 Martin
Vaz
74 37
19 99
6 66
Kerguelen
88
–60˚ –60˚
Antarctic 66 Antarctic

0˚ 90˚ 180˚ –90˚ 0˚

Figure 8.3 The Earth’s major plates and their relative and absolute motions. The plates are bounded by
oceanic ridges (dotted lines), oceanic trenches, mountain ranges, and transform faults (all as dashed lines).
Thin arrows show the directions and rates (in mm/yr) of relative motion at selected points on the plate
boundaries [after Bott6; reproduced by permission of Edward Arnold (Publishers) Ltd., and based on data in
Chase11; © Martin H.P. Bott, 1982].

114
 8: The Evolution and Structure of Ocean Basins

Figure 8.4 Method to determine the Euler pole for a spread- 8.4
ing ridge system. Transform faults (thick lines with double
arrows), which offset the ridge segments (double lines) Euler
describe small circles about the pole. Hence lines drawn pole
normal to the transform faults intersect at the Euler pole.

Rates of contemporary sea-floor spreading at

accretionary plate boundaries are also used to con-
strain the angular rotation rate about the Euler
pole, which is required to describe current plate
movements. However, active convergent plate
boundaries are more problematic and relative
velocities have to be determined indirectly. For
example, if the relative divergent motions between
plates A and B and between plates A and C are
known, the relative convergent motion between
plates B and C can be determined by simple vector
algebra. The above approach can be extended to
include all known plate boundaries, with the addi-
tional constraint that, globally, all the vectors must
form a single self-consistent set (Figure 8.3).
DeMets et al.14 have completed the most recent

Box 8.1 Satellite Gravity Fields

Short-wavelength (<400 km) gravity anomalies are highly correlated with small-scale topography; because of
this it is possible to use high-resolution gravity fields computed from satellite data to map the bathymetry and
tectonic features of the sea floor. This is particularly useful in parts of the Southern Ocean where ship-collected
data are sparse (Figure 8.5).

8.5a 180˚ 190˚ 200˚ 210˚ 220˚ 230˚ 240˚ 8.5b

–45˚ –45˚

–50˚ –50˚

–55˚ –55˚

–60˚ –60˚

–65˚ –65˚

mGals
180˚ 190˚ 200˚ 210˚ 220˚ 230˚ 240˚

Figure 8.5 (a) An ‘illuminated’ image of high-resolution gravity anomalies over the Pacific–Antarctic Ridge.
This accretionary ridge, outlined in orange and yellow, runs ENE–WSW across the figure, and is offset by
numerous transform faults (dark blue and purple; courtesy of W.H.F. Smith, first published in Sandwell and
Smith32). (b) An interpretation of the gravity image, which reveals features as small as a few tens of kilometres
and the complex relationships between spreading segments (thick lines) and small and large offset transform
faults and fracture zones (thin lines).

115
 R.B. Whitmarsh, J.M. Bull, R.G. Rothwell, and J. Thomson

global analysis of current plate motions. The power depth of a few kilometres, at the base of the crust.
and use of this technique is that, provided the rele- Oceanic lithosphere thickens with time as it cools
vant observations are available, it can be applied to and reaches about 100 km in the oldest parts of the
deduce past plate motions; by working backward ocean basins. The continental lithosphere thickness
from the present the relative positions of the plates varies between 100–200 km. The thermal defini-
in the past can be computed. tion of the base of the lithosphere is in good gener-
The definition of the boundary between the al agreement with observations of a zone of low
lithosphere and asthenosphere depends on the par- seismic velocity and high shear-wave attenuation at
ticular physical property under consideration (e.g., similar depths. These observations, in turn, are
temperature, seismic velocity, or flexural rigidity). consistent with the existence of a weak zone mark-
One of the main factors affecting the strength of ing the lithosphere–asthenosphere boundary.
sub-surface materials is temperature. Melting
occurs where the temperature–depth curve inter- Models of Lithospheric Evolution
sects the melting curve or ‘solidus’ (see Figure 8.12 An important observation first made in the 1960s
later). A thermal definition of lithospheric thickness was that heat flow is highest at mid-ocean ridges
is given by the depth at which the mantle is closest and decreases with distance from the ridge-axis, as
to melting. In fact, only a small fraction of the the mean depth of the ocean increases. These two
asthenosphere is believed to be molten, because it observations have provided the main constraints on
can transmit seismic shear waves which, if it was models of the thermal evolution of the oceanic
largely fluid, it could not do. Clearly, the depth of lithosphere. Two models explain these observa-
the base of the lithosphere depends on the tempera- tions. In the first model, the lithosphere behaves as
ture gradient within the Earth (or geothermal gra- the cold, upper boundary of a cooling half-space,
dient), the melting temperature of mantle minerals, such that the depth varies as (age)0.5 and heat flow
and the relative abundance of such minerals. varies as (age)–0.5. In the second model, the lithos-
Therefore, under ocean ridges, where the geother- phere is treated as a cooling plate with an isother-
mal gradient is high, the asthenosphere occurs at a mal lower boundary. The lithosphere thus behaves
as a cooling boundary layer until such time as the
lithospheric temperature gradient, constrained by
8.6 (a) the lower boundary condition, causes the curves
describing the variation of depth and heat flow
250 with age to flatten and change more slowly with
GDH1 time, as shown in Figure 8.6. In this model old
PSM
HS
lithosphere approaches the asymptotic value of
200 thermal plate thickness. Parsons and Sclater29 origi-
nally found that a 125 km thick lithosphere with a
Heat flow (mW/m2)

basal temperature of 1350°C fitted the data then

150 available. However, more recently Stein and
Stein35, using improved heat flow data, predicted a
hotter and thinner lithosphere (1450°C at the base
100 of a 95 km thick plate; Figure 8.6).
Another property of the lithosphere is its
response to vertical loading, or flexure. The conti-
50 nental lithosphere is commonly loaded by ice

0 Figure 8.6 Observations and models for (a) heat flow

and (b) oceanic depth as a function of crustal age. Depths
(b) are average values for the North Pacific and Northwest
3 Atlantic Oceans; heat-flow measurements are from sites
in these regions. Depths and heat-flow values (dots) have
been averaged over 2 Myr intervals. Also shown are
Depth (km)

4
functions computed from the plate model of Parsons and
5 Sclater29 (PSM), a cooling half-space model with the same
GDH1 thermal parameters (HS), and the Stein and Stein 35
6 PSM (GDH1), plate model. In (a) the HS and PSM curves over-
7
HS lap for ages younger than ca 120 million years (Myr; from
0 50 100 150 Stein and Stein35; reprinted with permission from Nature,
Age (Myr) © 1992 Macmillan Magazines Ltd).

116
 8: The Evolution and Structure of Ocean Basins

Age (Myr) 8.7

9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9

Gilbert Gauss Matuyama Brunhes Matuyama Gauss Gilbert

reverse normal reverse normal reverse normal reverse
C Layer 2

x
xBx Layer 3
10–60 mm/yr 10–60 mm/yr Moho
x
x x x
Oceanic x A x
lithosphere x x x x x Upper mantle
x x x x x
x x x
x x x
x x x
x
Asthenosphere x

Figure 8.7 Sea-floor spreading and the generation of magnetic lineations by the Vine–Matthews hypothesis. A marks
the rise of basaltic magma to form the oceanic crust, and the residual material which is left to form the mantle part of
the oceanic lithosphere. B marks the magma chamber, which eventually cools to form the widespread Layer 3 in the
crust. C marks the rapid cooling of basaltic magma to form the pillow lavas and dykes of crustal Layer 2. In Layer 2,
blocks of normal magnetic polarity are shown black and blocks of reverse polarity are shown white. ‘Moho’ indicates
the seismic discontinuity which marks the base of the crust. Vertical exaggeration is about 10:1 for a half-spreading
rate of 30 mm/yr [after Bott6; reproduced by permission of Edward Arnold (Publishers) Ltd; © Martin H.P. Bott, 1982].

sheets, sedimentary basins, or mountain ranges, years or more, and a compass needle would have
while the oceanic lithosphere is loaded by pointed to the South Magnetic Pole. Vine and
seamounts, sediments, and aseismic (volcanic) Matthews38 were the first to combine the notion of
ridges. If the lithosphere behaves perfectly rigidly, sea-floor spreading with the phenomenon of polari-
then it can be treated as an elastic plate; thus, the ty reversals of the geomagnetic field. They envis-
amount of flexure depends on its flexural rigidity, aged that new oceanic crust acquires a stable ther-
the magnitude of the load, and the elastic thickness mal remanent magnetisation after cooling through
(defined as the thickness of an equivalent uniform the Curie temperature (the temperature at which
sheet that responds elastically in the same way as magnetic minerals acquire the magnetisation vector
the real lithosphere). Within the ocean basins, elas- of the ambient geomagnetic field). They proposed
tic thickness increases with the age of the lithos- that the formation of oceanic crust is symmetric
phere39. However, in general it is found that elastic and continuous, on geological time-scales, as the
thickness is much less than thicknesses based on lithosphere moves away on either side of the axis of
thermal or seismological criteria. This is not sur- an accretionary ridge. Thus, the stripy magnetic
prising, given the simplified rheology inherent in anomalies characteristic of the oceanic crust are
the elastic plate model. The upper part of the produced by alternating crustal blocks of opposite
lithosphere actually behaves in a brittle manner, polarity that have recorded the polarity reversals
while the lower lithosphere behaves plastically. The (Figure 8.7).
duration of the loading also needs to be considered. Sequences of polarity reversals observed on land
In summary, therefore, the lithosphere can be enable magnetic anomaly patterns to be predicted
defined in several ways dependent on the processes at sea. The use of the geomagnetic polarity reversal
being considered. time-scale to date the oceanic lithosphere is based
on the identification of characteristic magnetic
Sea-Floor Spreading anomaly profiles and their relation to the indepen-
The magnetic field of the Earth is approximately dently established reversal chronology (see Box
that of a dipole, with an axis close to, but not nor- 8.2). The Vine–Matthews hypothesis was verified
mally coincident with, the Earth’s axis of rotation. by confirming the age of the oceanic crust predicted
At present, the polarity of the geomagnetic field is from the magnetic anomaly pattern by drilling, and
such that a compass needle points to the North dating with microfossils, the sediments immediately
Magnetic Pole, but sometimes in the past the geo- overlying the oceanic basement. The magnetic
magnetic field had the opposite polarity for 105 anomalies preserved in the oceanic crust have

117
 R.B. Whitmarsh, J.M. Bull, R.G. Rothwell, and J. Thomson

8.8 (a) (b) Figure 8.8 A sequence of maps, recon-

structed using sea-floor spreading and
palaeomagnetic data, to illustrate the
break-up of the Pangaea supercontinent.
(a) The configuration of land about
200 Myr ago; the dark blue on the east
side of the supercontinent is the Tethys
Ocean. The configurations for 160 Myr
(b) and 80 Myr ago (c), and the present
day (d) show that the dispersal of the
continents occurred by the opening of
the Atlantic and Indian Oceans, and the
resultant closure of the Tethys Ocean to
form the Alpine–Himalayan mountains
(c) (d) (dark blue, Atlantic-type – interior –
ocean; medium blue, Pacific-type – exte-
rior – ocean; light blue, flooded conti-
nent; from Nance et al. 27; © 1988 by
Scientific American, Inc. All rights
reserved).

allowed the sea-floor spreading history of about the causing uplift or because of the geometry of the plate
past 200 Myr to be reconstructed in detail. For boundaries surrounding the continent. This tension
example, Figure 8.8 shows reconstructed positions causes the development of normal (extensional)
of the continents following the break-up of faults in the brittle crust and ductile flow at depth,
Pangaea, a supercontinent that existed about 200 leading to significant crustal thinning. The East
Myr ago. African Rift Valley and the Rio Grande Rift in New
Mexico are examples of this phase, Figure 8.9(a).
The Life Cycle of Ocean Basins With continued extension the crust is eventually
The creation, evolution, and eventual destruction of thinned to such an extent that hot mantle material
the ocean basins is cyclical and has been called the rises to the surface, oceanic crust begins to form,
Wilson Cycle [Figure 8.9(a)–(e)] after the famous and an ocean basin is born (see later for a fuller
Canadian Earth scientist Tuzo Wilson. The cycle explanation). The uplifted continental flanks are
starts when a continental area is thrown into ten- gradually eroded (see, e.g., Chapter 9) and sedi-
sion, either because of a sub-crustal heat source ments are deposited in the new ocean basin. The

Box 8.2 Sediment and Rock Dating

Fossiliferous sediments and rocks younger than 600 Myr may be dated by a variety of techniques.
These involve the recognition of individual fossil species, or the content of assemblages of such
species, within a biostratigraphic framework which describes the sequence of appearances and disap-
pearances of species over geological time. The marine oxygen-isotope record (see Box 8.3) also pro-
vides a well-constrained means for dating Late Cenozoic biogenic carbonate sections. Physical meth-
ods, such as geomagnetic polarity reversal stratigraphy (see text) and radiometric techniques, may
also be used for dating; such methods are especially useful when applied to nonfossiliferous sedi-
ments. Radiometric dating methods depend on the high-precision measurement of the parent–daugh-
ter pairs of isotopes by mass spectrometry; knowing the natural radioactive decay chains and decay
constants of the individual isotopes in such chains, it is possible to compute ages from these mea-
surements. Sediment dates tend to be based on the decay of shorter-lived natural or cosmogenic
radionuclides; rock dates are generally based on the predictable rates of conversion of long-lived nat-
ural parent radionuclides into daughter products by radioactive decay over time.

118
 8: The Evolution and Structure of Ocean Basins

(a) (b) 8.9

Fault-block mountains Sea level

Normal faults Rift valley

Sea level
Oceanic
crust
Continental crust

Magma Magma
Sediment

(c)
Ocean
Spreading axis Transform fault

Sediment
Oceanic crust
Magma

Lithosphere
Continental crust
(d) Ocean

India Tibet

Sediment on Fore-arc basin

continental
margin Fore-arc ridge
Trench

Lithosphere
(e) Oceanic crust
Asthenosphere
Ganges Plain Tibetan Plateau
Ganges Plain Figure 8.9 Schematic block
sediments Himalaya diagrams showing the Wilson
cycle of plate tectonic activity.
Suture
zone (a), (b), and (c) illustrate the rift-
ing of a continent to form a
new ocean basin bounded by
rifted continental margins. (d)
and (e) illustrate the destruction
of an ocean basin, initially by
subduction of the oceanic
lithosphere (d), followed by
full-scale continental collision
(e). See the text for more dis-
cussion.

119
 R.B. Whitmarsh, J.M. Bull, R.G. Rothwell, and J. Thomson

Red Sea is an example of such a young oceanic rift, Intraplate Deformation

Figure 8.9(b). The ocean basin then expands by Although most volcanic and seismic activity is con-
sea-floor spreading. The Atlantic Ocean is an centrated at plate boundaries, some activity occurs
example of a mature ocean basin surrounded by within plates. For example, the ocean basins are
rifted continental margins, Figure 8.9(c). marked by the frequent occurrence of seamount
Cooling of the oceanic lithosphere continues chains and aseismic ridges associated with one or
until its density exceeds that of the asthenosphere more active volcanoes. Much of this intraplate vol-
and a potentially unstable situation arises. A sub- canic activity is related to ‘hot spots’ caused by
duction zone may form. The machanism for this is thermal plumes rising from the lower mantle. Over
presently controversial; it is the one part of the time, the motion of the plates over the near-station-
cycle which is poorly understood. The main charac- ary hot spots leads to elongate chains of volcanoes
teristics of a subduction zone are shown in Figure and aseismic ridges. Two conspicuous examples are
8.9(d), which illustrates the incipient collision of the Hawaiian–Emperor seamount chain in the
India and Tibet. The subduction of oceanic lithos- Pacific Ocean and the Ninetyeast Ridge in the
phere may cause the ocean basin to contract. The Indian Ocean.
Pacific Ocean is a present-day example of a con- Bergman4 recognised that oceanic intraplate seis-
tracting ocean delimited by active plate margins mic activity could be divided into two categories
(subduction zones and transform faults). based on the focal mechanisms of the earthquakes
Eventually, if all the oceanic lithosphere has (the focal mechanism is the form of rupture at the
been subducted, continent–continent collision takes seat of the earthquake inferred from the distribu-
place. The collision of India with Asia, which led to tion of initial compressional or dilatational
the formation of the Himalayan mountains and the motions, generated by the earthquake, at the
high Tibetan Plateau, is a dramatic example of the Earth’s surface). A large proportion of the seismic
closure of an ocean basin (the Tethys Ocean – see activity occurs in oceanic lithosphere less than
Figure 8.8) and continent–continent collision, 35 Myr old; focal mechanisms of these events are a
Figure 8.9(e). Present-day Asia represents a com- poor indicator of the state of regional stress,
plex assembly of a large number of small continen- because they are dominated by the early thermal
tal blocks that have collided to form a large conti- evolution of the lithosphere. The second type of
nent, or supercontinent. Eventually, this continent seismic activity occurs in lithosphere older than
will be rifted, a new ocean basin will form, and the 35 Myr and is characterised by thrust or strike–slip
Wilson Cycle will begin again. focal mechanisms. These indicate an intraplate
stress field dominated by maximum horizontal
compression oriented normal to the ridge crest. The
focal mechanisms associated with this type accu-
8.10 (a) rately reflect the regional state of stress.
Seismicity in old oceanic lithosphere is not entire-
ly randomly distributed throughout the oceans and
may be associated with a few areas of intraplate
(b)

Figure 8.10 A simplified model for the evolution of

transform and rifted continental margins based on the
example of the Côte d’Ivoire and Ghana margins off West
Africa. (a) Onset of continent-to-continent active trans-
(c) form contact. (b) Continent-to-continent and continent-
to-rifted-margin contacts. Shearing produces a lateral
marginal ridge and associated tectonic deformation. (c)
Progressive drift of the end of the hot accretionary ridge
along the transform margin. Thermal exchange occurs
between continental and oceanic lithosphere, leading to
vertical adjustments. (d) Mature stage of rifting; evolution
of the margin includes thermal subsidence. 1, Direction
of relative plate motion; 2, transform motion between
(d) continental crusts; 3, transform motion between oceanic
crusts (transform faults); 4, thick continental crust; 5,
thinned continental crust; 6, oceanic crust; 7, axis of
accretionary plate boundary; 8, marginal ridge and
region of related tectonic deformation (courtesy of J.
Mascle, with permission).

120
 8: The Evolution and Structure of Ocean Basins

1300 km
0 1300 km 0
8.11

1 Myr 30 Myr

(+4 km) (+120 km)

125 km 125 km

1300 km 0 1300 km
0

10 Myr 40 Myr

(+40 km) (+160 km)

125 km
125 km

1300 km
1300 km 0
0

50 Myr
20 Myr

(+260 km)
(+80 km)

125 km 125 km

Figure 8.11 Numerical modelling of strain rate within a continental lithosphere which is stretched under tension by
200 km over 50 Myr relative to the fixed left side of the model. The intensity of deformation is shown by the colours;
greater redness indicates more intense deformation and emphasises the loci of extension. The colour scale is logarith-
mic from 10–15/s (darkest blue) to 10–11/s (red). The crust–mantle boundary is shown in white (horizontal line at 1 Myr).
From the start the model has an in-built mid-crustal zone of weakness (white lines outlining the enclosed region at
centre top) and an upper mantle weakness caused by the thickened crust to the right of the upper crustal weakness.
Note how the locus of extension begins in the upper crust, but shifts to the right and into the lower lithosphere after
20 Myr. Models such as this explain why crustal rifting precedes the appearance of volcanic products at the surface
and why the rifting and volcanism occur in different places. If the rifting illustrated here continues beyond 50 Myr,
eventually new oceanic crust will be formed in the centre of the model (courtesy of D.S. Sawyer, with permission).

deformation. These areas have also been called dif- Rifted margins are more common; for example,
fuse plate boundaries42. The best-developed exam- they border the Atlantic Ocean from 50°S to the
ple is in the northeastern Indian Ocean, where a Norwegian Sea. They are of particular interest for
1500-km-wide zone of intense tectonic deformation economic reasons, since the same processes which
separates the Indian and Australian plates9,14. lead to rifted margins can also form major sedi-
mentary basins in which oil and gas may accumu-
The Margins of the Ocean Basins late (see Chapter 21). Transform margins are often
An ocean basin is created by the rifting and break- shorter and have been less widely studied.
ing apart of a continental lithospheric plate and is Continental rifting, like most other major Earth
bounded by rifted or transform continental margins processes, is driven by heat from within the Earth.
(Figure 8.10). Rifted margins extend from the edge The temperature of the asthenosphere largely deter-
of the continental shelf into oceanic depths; conse- mines the amount of melt produced during rifting;
quently, such margins are major bathymetric fea- the temperature of the lithosphere determines
tures on the Earth’s surface. They contain evidence when, and how much, uplift and/or subsidence is
of the processes which accompanied the initiation experienced by the margin. The rifting and break-
of sea-floor spreading, as well as the poorly under- up of continents is often thought to be a response
stood transition from continental to oceanic crust. to tensional forces resulting from the sub-lithos-

121
 R.B. Whitmarsh, J.M. Bull, R.G. Rothwell, and J. Thomson

8.12 (a) (b)

Depth (km) 50
(c)
0 100 200
m
20 70 k

1900 du s
ui 40 km 1480 ˚C
L iq 100
0
0.7 km
5 130

Melt thickness (km)

45 0.1idus 1580 ˚C 15
38 0.2 So l
1700 0.4 30

Melt thickness (km)

Temperature (˚C)

1580 ˚C
1480 ˚C

1380 ˚C 1480 ˚C 10 70 km
1500 1330 ˚C 20 km
100 1380 ˚C
1280 ˚C
0 km
13
1380 ˚C
1300 10 5
1330 ˚C 70 km
100 km 1280 ˚C
1280 ˚C 130 km
1100 0 0
0 2 4 6 8 0 40 80 120 160 1 2 3 4 5
Pressure (GPa) Depth (km) Stretching factor `

Figure 8.12 (a) Adiabatic decompression paths in temperature–pressure space (asthenospheric potential temperature is
given on the right). Melting begins when material rising along an adiabatic decompression curve (arrowed lines) meets
the solidus curve. The melt fraction is indicated on the curves between the solidus and the liquidus. (b) The higher the
temperature of the adiabatic decompression curve, the greater the thickness of melt produced. The total thickness of
melt present below a given depth, calculated by integrating the volume of melt implied by the curves in (a), is plotted
as a function of depth. (c) Predicted thicknesses of melt produced as a result of different lithospheric stretching factors
ß. The numbers against the curves give the thickness of the mechanical boundary layer and the temperatures on the
right are the interior potential temperatures, i.e., the temperatures the rocks would have if brought to the surface with-
out loss of heat (Figures 7 and 22 in McKenzie and Bickle24; by permission of Oxford University Press).

pheric loading associated with regions of reduced adiabatically (i.e., heat is conserved), the tempera-
density in the mantle. Numerical dynamic model- ture rises, and eventually the more volatile compo-
ling of rifting suggests that a continental lithos- nents of the rock begin to melt (Figure 8.12). The
pheric plate subjected to prolonged tension begins higher the initial temperature and the greater the
to stretch and thin. At the top it behaves brittly, so ascent of the asthenosphere, the more magma is
the upper crust is dissected by normal faults; deep- produced. The magma separates from the host rock
er, the deformation is viscous (Figure 8.11). The and rises by its own buoyancy. Eventually, after
asthenosphere responds passively to this thinning several million years, following the surface faulting,
in a very important way. Material rises to fill the magma may reach the surface. Even later, the con-
space vacated by the lithosphere and consequently tinental lithosphere ruptures and the asthenosphere
experiences lower pressure. Usually this happens itself essentially reaches the surface. This is when

8.13 T JURASSIC CRETACEOUS P EOCENE O MIOCENE

Sea level
0
Wate
rd epth

1000
Depth (m)

Ther
mal
subs 2000
iden
ce
Figure 8.13 Subsidence history of
Well 62/7-1 in the Goban Spur Basin,
3000 Southwestern Approaches to the
T = TRIASSIC English Channel. The total subsidence
P = PALAEOCENE
O = OLIGOCENE
Tota
l subs is the sum of the thermal subsidence
iden
ce 4000 and the isostatic subsidence due to
sediment loading (Figure 4 in Colin et
150 100 50 0 al.12; reproduced by permission of ©
Time (Myr BP) Butterworth Heinemann Ltd).

122
 8: The Evolution and Structure of Ocean Basins

Figure 8.14 Models of gravity and surface and (a) 8.14

deep-towed magnetometer profiles across the West East
1000 1000

magnetometer (km) anomaly (nT)

3D model
West Iberia margin (see inset for location; OC

Magnetic
= oceanic crust; TC = transitional crust; CC = 0
observed
0
2D model
continental crust). (a) Deep-tow and (b) sur-
face magnetic profiles – the magnetic models 0
898 900
0

k
k
(intensities in A/m, black blocks are normally J M3 PR

Depth below
M0

magnetised oceanic crust) indicate that sea- 4 1 –1 1 –1 1 2 0.8 3 0.8 .3.81.2 0.8 1.2A/m

floor spreading began at about Chron M3

OC TC
(130 Myr ago) at 9 mm/yr; extrapolating the 10 10

reversal time-scale further back in time does 400

(b) 400

Magnetic anomaly
not yield computed anomalies that fit the +
++
+ +
200 200
observations (crosses). Hence, east of the peri- +
+
++

(nT)
++
+ ++
dotite ridge normally magnetised blocks with 0
+
+
++ +++
++ +
+
+ ++
++++++
+++
+++ ++ +++
++ ++ +++++++++++++++++++++++++ ++++++++++++++++++++++++++++++++
++++ ++
++++++++
0
+ ++ +++
variable intensities of magnetisation were +
+ +
+
++ +
+ + ++ +
+ +
used, such as might be produced by continen- –200 ++ –200

tal crust. Note how the deep-towed magne- 5 5

Depth b.s.l.
M0 J M3
tometer observations provide better resolution
(km)
10 6 .75 1.25 .25 .45 .15 .9 .3 .5 .1 10
of the sea-floor spreading anomalies. (c)
OC TC CC
Gravity profile – the model used to fit the 15 15
anomaly (mGal)

50
observed profile (crosses) is constrained by 50 (c)
+
++ +++
+
+
Gravity

+
++ + ++ + + + ++ + + + ++
++++++++++++++++++++++++++++++ + + + + +
layer thicknesses and velocities calculated –50
+++ ++++ ++
+ + + + + ++ + ++ ++ +++++
++ +
+
++
++ ++ ++++++
++
+
–50
from seismic refraction lines L1 to L3 (aligned L4
0 0
normal to the profile) and L4; densities are in L3 L2
900 901
L1

Mg/m 3; PR indicates a peridotite basement

k
k
PR
Depth (km)

2.0
2.5 2.15 2.25
ridge drilled during Ocean Drilling Program 10 2.7
2.9
2.55 10
3.0 2.8
Leg 149 and numbered arrows indicate other 3.26

OC TC CC
Leg 149 boreholes (work in progress, R.B. 20
3.4

20
Whitmarsh and P.R. Miles). 100 200 300
Distance (km)

the steady-state process of magma production, and gin has been studied using seismic refraction and
its intrusion and extrusion at the ocean floor, reflection profiles, magnetic and gravity anomalies,
which we know as sea-floor spreading, begins. An heat flow measurements, and cores from scientific
ocean basin has been born. drilling41 (Figure 8.14). The first-formed oceanic
As well as influencing horizontal movements, crust is abnormally thin (only about 3–4 km), indi-
the above events also determine the vertical cating a poor magma supply, and is bounded on
motion of the margin. If the asthenosphere is hot the landward side by a basement ridge of serpen-
enough, then initially the lithosphere expands suf- tinized peridotite (peridotite is an iron- and magne-
ficiently to cause uplift, even well-above sea-level, sium-rich silicate rock originating in the mantle;
of the continent adjacent to the rifting. However, this rock can become altered by a process known as
in all cases following break-up the lithosphere serpentinization, which occurs when peridotite
cools, and therefore subsides, with a time constant comes into contact with water at low tempera-
of about 60 Myr. Erosion of the adjacent conti- tures). Between this ridge and the thinned continen-
nents frequently leads to substantial offshore sedi- tal crust, typified by gently landward-tilted fault
ment accumulations, which load the margin; these blocks and half graben, there is a recently discov-
can flex the margin and cause further isostatic sub- ered 130 km wide transitional zone. This zone has
sidence. It is very difficult to measure accurately linear magnetic anomalies parallel to the isochrons
the subsidence history of a margin; palaeodepth (i.e., lines of constant crustal age) of sea-floor
estimates based on benthic microfossils from deep spreading, yet does not have the magnetic or seis-
wells are often crude. However, given the uncer- mic character of oceanic crust. It is probably either
tainties, numerical cooling and isostatic models a mixture of continental crust and igneous intru-
can usually be fitted to such curves quite well sives–extrusives or unroofed mantle. A similar tran-
(Figure 8.13). sitional zone may be exposed today on the Arabian
The West Iberia margin is a good example of margin of the central Red Sea.
one which rifted over a cool asthenosphere, as indi- The western edge of the Rockall Plateau in the
cated by the apparent absence of syn-rift volcanism northeast Atlantic is a good example of a margin
(rifting associated with a relatively hot asthenos- which rifted over a hot asthenosphere. Here, the
phere is accompanied by the production of large uppermost crust overlying the continent–ocean
quantities of volcanic lavas). The crust of the mar- transition is marked by seaward-dipping reflectors

123
 R.B. Whitmarsh, J.M. Bull, R.G. Rothwell, and J. Thomson

8.15 D Figure 8.15 Seismic velocity

H G F E DD C B A model from NW (left) to SE
NW SE
k

k
k

k
k
0 (right) of the northwest margin
2.0 4.0 of Rockall Plateau, northeast
2.0 3.0 5.5 Atlantic Ocean. A–H and DD
5.5 indicate locations of expanding-
6.5 6.0
spread profiles aligned normal
6.5 to the northwest–southeast sec-
10 7.3
tion. VE is the vertical exaggera-
Depth (km)

tion. The diagonal shading indi-

8.0 High-velocity layer 7.0 cates the estimated extent of the
seaward-dipping reflectors and
20 ‘late stage volcanics’. Moho rep-
Mo resents the Mohoroviçic discon-
ho

7.3
V.E. = 2.3 tinuity, the depth at which seis-
mic velocity exceeds 7.8 km/s
8.0
and where the mantle begins
30 (Figure 4 in Morgan and
50 km Barton 25 ; by permission of
Elsevier Science).

(Figure 8.15). Drilling of similar reflectors off The Structure of the Oceanic Crust
Norway has shown that they represent a thick A vertical section through the Earth beneath the
sequence of dipping lava flows. Seismic refraction ocean basins generally consists of hundreds or
measurements show that the crust off the western thousands of metres of sediment overlying 7 km of
edge of the Rockall Plateau is unusually thick, with igneous crust, formed by sea-floor spreading, which
up to 15 km of material with a velocity over in turn overlies the upper part of the mantle. Our
7.3 km/s at the base of the crust (Figure 8.15). Both knowledge of the structure and composition of the
the dipping volcanic reflectors and the thick crust igneous crust comes from direct sampling and from
can be explained by a high asthenospheric tempera- remote sensing by geophysical measurements.
ture and extensive melt production at the time of Sampling by dredging, where rocks are exposed
break-up. The lavas denote magma which reached on the sea floor, reveals a wide variety of rock
the surface and the high-velocity lower crust signi- types from basalts and gabbros of the crust to
fies denser magma ‘underplated’ at the base of the ultramafics (iron- and magnesium-rich rocks)
crust. The precise reason for the prodigious melt derived from the uppermost mantle. However, it is
production at this margin is debatable. It may have frequently difficult to relate even well-located
been due to the proximity of the Iceland plume or dredged rocks to an unambiguous layering of the
hot spot. An alternative hypothesis is that the melt crust. Drilling is a better way to do this (e.g., see
was produced by vigorous local convection within Chapter 19). The international Ocean Drilling
the asthenosphere. Program (ODP) has succeeded in drilling hundreds
World-wide, few transform margins have been of metres into the crust at several holes around the
studied. Such margins are believed to develop very world. The deepest is the 2111 m deep Hole 504B
differently from rifted margins. Principally, they in the eastern Pacific Ocean2 (Figure 8.17). In spite
experience shear motion, perhaps with a minor of this unique achievement the hole has so far
component of extension or compression, along one probably penetrated only basaltic pillow lavas and
or more deep crustal faults (faults which are very sheeted dikes of the uppermost crust. Down-hole
difficult to detect on reflection profiles). Every physical measurements demonstrate that the crust
point on the margin also experiences a temporary is vertically variable on a scale of metres or tens of
rise in temperature, and consequent uplift, as the metres. Lateral variability is also expected from
end of the adjacent spreading ridge traverses along models of crustal creation, which include a varying
the margin (Figure 8.10). Figure 8.16 is a section magma supply and magmatic, as well as purely tec-
across the southern Exmouth Plateau margin off tonic, extension. A simpler picture emerges if we
northwest Australia. Here there is a relatively use remote geophysical measurements to obtain a
abrupt transition from 26 km thick underplated more averaged view of the crust.
continental crust to 8 km thick oceanic crust. The best way to investigate remotely the Earth
Between the two there is a 50 km wide zone of beyond the reach of the drill is through seismic
complicated structure which may include a zone of energy. Using the wide-angle refraction technique,
intrusion and, seaward, a tilted fault block. where the seismic source (usually towed at the sea

124
 8: The Evolution and Structure of Ocean Basins

SW 1 3 5 7 9 1 3 5 7 9 1 3 5 7 9 1 3 5 7 9 NE 8.16a
0 ESP C3 ESP C2A ESP C2 POSTRIFT SYNRIFT SECTION ESP C1

2 1 3 5 7 9 km/s
Core detachment
4 ESP C4 d
Depth (km)

6 PERMIAN PRERIFT d2

8 OCEANIC CRUST TRIASSIC d1

10 detachments

12
14 0 km 25
16.8 km
16 Underplated layer ?
26.8 km

9000 8000 7000 6000 5000 4000 3000 2000 1000 CDP 200

Zone of
intrusion ESP E5 8.16b
2.06 ESP C3 ESP C2A ESP C2 1.83 ESP C1
0 1.83 1.0 0
ESP C4
2.06
2.38

2.55 Detachment 2.55 Complexes

10 2.82 10
2.65
Depth (km)

2.95 2.72 Crystalline crust

3.29 20
20 3.04

0 km 50 3.29
Strike-slip deformation zone
30 30
SW NE

8000 6000 4000 2000 CDP 200

Figure 8.16 (a) Multichannel seismic reflection profile after migration and depth conversion, from southwest (left) to
northeast (right) across the Exmouth Plateau transform margin (vertical exaggeration = 3). Near vertical lines denote
features interpreted as faults; other possible near-horizontal faults are labelled detachment or d, d1, d2. The positions of
five intersecting expanding-spread profiles (ESPs), labelled C1, C2, etc., and the velocity–depth profiles computed from
them, are also shown. (b) A two-dimensional gravity model (densities in Mg/m3), based on the seismic observations,
which incorporates transitional crust (grey) between common depth point (CDP) gathers 6500–8000 (see horizontal
axes) and underplated crust (black) landward of this zone. The transitional and underplated crusts may have been
intruded as the adjacent rift axis migrated along the transform margin (Figures 2 and 3 in Lorenzo et al.23; by permis-
sion of the authors).

surface) and receivers (usually placed on the ocean

floor) are many kilometres apart, it is possible to Table 8.1. Mean Oceanic Crustal
compute the increase of seismic velocity with depth Structure (from Table 8 in White et al.40)
and to measure the thickness of the crust (Table
Velocity (km/s) Thickness (km)
8.1; see Figure 19.16). Although the crustal struc-
ture is often reported in terms of two or more dis- Layer 2 2.5–6.6 2.11±0.55
tinct layers, in practice, velocity increases more or
less steadily with depth. We know from sampling Layer 3 6.6–7.6 4.97±0.90
that the uppermost layers (at least 2.5 km/s) repre- Mantle >7.6
sent basaltic pillow lavas. Although solid basalt has
a velocity of about 6.3 km/s, cracks, fissures, and Total
widespread voids in the lavas lead to lower veloci- igneous 7.08±0.78
ties. In time, many of these spaces are filled by crust
hydrothermally deposited minerals; consequently,
the upper crustal velocity is greater in older crust.
The lava flows are underlain by dikes, subvertical

125
 R.B. Whitmarsh, J.M. Bull, R.G. Rothwell, and J. Thomson

Figure 8.17 Down-hole logs from Ocean Drilling

Lithostratigraphy
8.17
Program Site 504B (as at the end of Leg 148, February,
Seismic layer
1993). Note the broad changes in velocity (Vp) and resis-
Vp (km/s) Resistivity (1m) tivity between the volcanic and sheeted-dike layers, and
3 4 5 6 7 10 1000 the short wavelength (ca. 1–10 m) variations in these
properties (profiles begin at 250 mbsf; mbsf = metres
2A
below sea floor; Figure 2 in ODP Leg 148 Shipboard
Scientific Party28; © American Geophysical Union).
Volcanics

500

2B

Fault density of rocks is strongly correlated with their

Sub-bottom depth (m.b.s.f.)

Transition

seismic velocity (Figure 8.18); this useful property

1000
means that we can infer density from velocity and
check our seismic models by computing their gravi-
tational effect and comparing this with independent
gravity observations (e.g., Figure 8.14).
Another important property of the igneous crust
Sheeted dikes

2C
1500 is its magnetisation. When magma, containing a few
percent of certain iron oxides, cools it acquires a
remanent magnetisation, in the direction of the con-
temporary Earth’s field, which is stable over mil-
lions of years (see above). This magnetisation pro-
vides the ‘memory’ in the rocks, whereby sea-floor
2000 spreading records reversals of the Earth’s magnetic
3? Layer 3? Fault
field. The remanent magnetisation slowly decreases
with time as some iron oxides undergo further oxi-
dation, a process accelerated by hydrothermal circu-
sheets that intrude parallel to the ridge axis during lation. Older basalts acquire a stronger, secondary,
sea-floor spreading. Below the dikes there is a layer possibly chemical, magnetisation (Figure 8.19).
of 6.7–7.2 km/s material which forms the greater Hence a minimum magnetisation occurs in rocks
part of the crust. This oceanic layer, or Layer 3, has which are 8–20 Myr old. The magnetic susceptibili-
rarely been drilled in situ. From samples, and by ty of the crustal rocks is usually relatively insignifi-
analogy with ophiolites (sequences of mainly cant, so that it does not have to be included when
igneous rocks which contain the same rock types,
and in the same order, as are found within the
oceanic crust) exposed on land, most geologists Compressional wave velocity (km/s) 8.18
think it largely consists of gabbro, a coarser 3.0 4.0 5.0 6.0 7.0 8.0
3.2
grained rock representing the frozen melt which
was the source of the basaltic flows and dikes. At
the base of the crust, velocity increases, often
3.0
abruptly, to around 8.0 km/s, comparable to the ▲▲
▲
Mohoroviçic discontinuity at the base of the conti- ▲ ▲
▲ ▲ ▲▲
▲
▲
nental crust; this marks the top of the Earth’s man- ▲ ▲
▲
Wet-bulk density (Mg/m3)

▲
tle. The velocity–depth structure in Table 8.1 is an 2.8
▲
▲▲ ▲
▲
▲

average for normal crust; in fracture zones the crust ▲ ▲

is often thinner and may be underlain by velocities ▲

thought to represent serpentinized peridotite; and 2.6

near hot spots, such as Iceland, it is thicker. The

2.4
Figure 8.18 Wet-bulk density of samples of oceanic crust
and ophiolites plotted against the inverse of compression-
al-wave velocity. The solid line represents a least squares
2.2 Basalt
fit, with the standard error indicated by the dashed lines ▲ Diabase
(Figure 1 in Carlson and Raskin10; reprinted with permis- Gabbros
sion from Nature, © 1984 Macmillan Magazines Miscellaneous
Limited). 2.0
0.35 0.30 0.25 0.20 0.15 0.10
Inverse velocity (s/km)

126
 8: The Evolution and Structure of Ocean Basins

Figure 8.19 Summaries of crustal magnetisation against (a) 8.19

crustal age. (a) Obtained from the inversion of marine mag-
5 Cretaceous magnetic normal period
netic anomalies. (b) Derived from measurements of normal

.
remanent magnetisation made on drilled basalt samples.

Intensity (A/m)
Both curves show the decrease in magnetisation caused by
the oxidation of magnetic minerals in the first 20 Myr and 1
the subsequent increase indicating the acquisition of a
0.5
remanent (chemical?) magnetisation (Figure 4 in Sayanagi
and Tamaki33; © the American Geophysical Union).

computing the magnetic effect of these rocks. 0.1

0 20 40 60 80 100 120 140 160 180
Other important physical properties of crustal Age (Myr)
rocks include electrical resistivity, shear wave (b)
Cretaceous magnetic normal period
velocity, permeability, and thermal conductivity; 20

.
many are anisotropic.
10
Recently, our perception of the igneous crust has

Intensity (A/m)
5
been improved by seismic-reflection profiling 26.
Using the same equipment and ships as in the
search for oil and gas, but with specially designed
configurations, intriguing reflecting surfaces have 1
been detected below the Atlantic Ocean, deep with- 0.5
in the crust (Figure 8.20). On profiles acquired 0 20 40 60 80 100 120 140 160 180
along isochrons, these surfaces tend to have low Age (Myr)

8.20

Figure 8.20 Multichannel seismic-reflection profiles of oceanic crust produced at slow-spreading rates in the North
Atlantic Ocean, illustrating a variety of forms of reflectivity (the vertical scale in seconds is the time required for sound
to be reflected back to the sea surface). A, B, D, E, and F are from the western North Atlantic; C is from the eastern
North Atlantic. A and B are from flow-line profiles, C is oblique to the spreading direction, D and E are isochron lines,
and F is along the trough of the small-offset Blake Spur fracture zone. Typically, the shallow crust contains distinct
sub-horizontal reflections and the middle crust is almost reflection-free. The lower crust exhibits the strongest and
most diverse reflectivity, including banded patterns of dipping reflectors. The dipping reflectors may have a tectonic or
igneous origin. A distinct Moho reflection is seldom seen; the reflective lower crust typically merges downward to a
reflection-free upper mantle (Figure 4 in Mutter and Karson26; © 1992 by the AAAS).
127
 R.B. Whitmarsh, J.M. Bull, R.G. Rothwell, and J. Thomson

8.21

Figure 8.21 Chart of the global distribution of sediment in the ocean (green, calcareous sediments; yellow, siliceous
sediments; brown, terrigenous sediments; blue, glaciogenic sediments; pink, deep-sea clay; white, margin sediments;
drawn by R.G. Rothwell).

dips, be highly reflective throughout the igneous material is transported to the ocean floor by a
crust, and dip bi-directionally; on profiles parallel number of mechanisms and processes (Figure 8.22;
to the spreading direction they are steeper, usually also see Chapter 9).
dip toward the spreading centre, and offset the Rivers form the main pathway of terrigenous
basement surface. The former may represent either sediment to the oceans, although wind transport is
contrasts developed during the igneous creation of particularly important for fine-grained detrital
the crust or faults; the latter are probably ridge- material. Glacier input is important at high lati-
parallel normal faults, mostly active in the early tudes. The main factors controlling the flux of sedi-
development of the crust. ment derived from continental erosion are climate,
precipitation, type of weathering, character of the
Sediment Provenance and coarse-grained material, topography, and land area
Transport Processes in the source regions. When sea-level was low, such
Sediments cover most of the ocean floor. Our as during glacial periods, deep-sea terrigenous sedi-
knowledge of these sediments, and the Earth histo- mentation was especially dynamic. At such times,
ry they record, has increased markedly in the past the mechanical erosion of continents and the sedi-
three decades, through gravity and piston coring ment loads of rivers were much greater. Much of
(see Figure 19.4), and deep-sea drilling. The sedi- the continental shelves were exposed as coastal
ments comprise, in varying amounts, detrital mate- plains, resulting in rivers that transported their
rial derived from the weathering of the continents, loads to the outer edge of the continental shelf for
biogenic debris derived from planktonic organisms, more rapid deposition into the deep-sea basins.
and clay-size material. Sediment types are distin- Pelagic sediments are typically dominated by
guished by particular constituents; different sedi- biogenic material, but vary considerably with lati-
ments show well-defined global distributions tude and water depth20. They include the deep-sea
(Figure 8.21; see also Figure 11.8). Sedimentary calcareous and siliceous oozes, composed largely

128
 8: The Evolution and Structure of Ocean Basins

8.22

Figure 8.22 Sediment transport

routes to the deep ocean (1, Aeolian
transport; 2, fluvial transport; 3, coastal ero-
sion; 4, fallout from volcanic ash clouds; 5, bio-
genic debris; 6, authigenesis; 7, ice-rafting; 8, mass
gravity flows; 9, hydrothermal activity; 10, submarine vol-
canism; 11, fallout from high-altitude jetstreams; 12, micro-
meteorites; drawn by R.G. Rothwell).

of the remains of planktonic organisms, and deep- tion between the near-surface ocean and the deep
water clays (Figure 8.23). Four main processes ocean. Locally, these two distinct environments
control the character of biogenic oozes: the supply interact through wind-driven upwelling of deep
of biogenic material, its dissolution in the water water, and the consequent downwelling of near-
column, its dilution by nonbiogenic material, and surface water, and through the constant ‘rain’ of
subsequent diagenetic alteration (see later). Pelagic skeletons from dead planktonic organisms (Figure
sedimentation can be viewed as a form of interac- 8.23). This ‘rain’ of biogenic particulate matter
forms the primary sink in the ocean basins
(Chapter 7).
8.23 Cosmic The spatial distribution of calcareous oozes is
source Windblown
Volcanic
source Ash Meteorites,
dust
Land
source
controlled by depth due to dissolution (Chapter
dust 11). Calcite, which forms the main skeletal material
At mos p here of many planktonic organisms (such as
Ice-rafting

foraminifera and coccolithophores), shows increas-

ing solubility with water depth; this is related to
Ocean increased hydrostatic pressure, increasing CO2 con-
CaCO3 SiO2
Pteropods Foraminifera Coccolithophores Radiolaria Diatoms
tent within the water, and decreasing temperature.
T R O P I C A L TROPICAL BOREAL Therefore, there is a depth, called the lysocline, that
T E M P E R A T E HIGH PRODUCTIV ITY separates well-preserved from poorly preserved,
PTEROPOD
solution-etched foraminifera and coccolithophores
OOZE
RADIOLARIAN OOZE

GLACIAL MARINE

Aragonite compensation depth

DIATOM OOZE

SEDIMENTS

GLOBIGERINA
OOZE

'Lysocline'
COCCOLITH
OOZE

Calcite compensation depth

RED CLAY Figure 8.23 Sources and pathways of pelagic sedimenta-

tion in the oceans (after Hay17).

129
 R.B. Whitmarsh, J.M. Bull, R.G. Rothwell, and J. Thomson

8.24 Rate of supply of carbonate Figure 8.24 Conceptual model for calcite dissolution
0 rate in the ocean, showing the relationship between the
CARBONATE PRECIPITATION
calcite compensation depth (CCD) and the lysocline.
Increased carbonate supply at the equator depresses the

SUBTROPICAL SUPPLY RATE

EQUATORIAL SUPPLY RATE

1 DISSOLUTION RATE CCD (Figure 11 from Berger et al. 3 ; © American
Geophysical Union).

2 Terrigenous sediments are composed largely of

Depth (km)

detrital material derived from the weathering of

continents. They include turbidite muds, which
cover the abyssal plains, and glacial material,
3
which covers substantial parts of the sea floor in
Hydrographic polar regions. Such sediments are characterised by
lysocline
high rates of deposition, usually contain small
4 Southern quantities of biogenic material, and are usually
subtropical
CCD
transported to the deep sea by some form of sedi-
ment gravity flow, such as debris flows or turbidity
5
Central equatorial
calcite compensation depth
currents31,36 (Figure 8.25 and Chapter 9).

Rate of dissolution Sediment Diagenesis

The constituents which comprise a newly deposited
(Figure 8.24). The lysocline varies, but generally marine sediment include relatively unreactive detri-
lies at depths of 3000–5000 m. At some greater tal materials, introduced to the ocean from the con-
depth, called the calcite compensation depth tinents by winds and rivers (and by ice at high lati-
(CCD), the rate of supply of biogenic calcite equals tudes), and more reactive biogenic materials, such
its rate of dissolution. Below the CCD, only car- as CaCO 3 , opal (an amorphous form of silica
bonate-free sediments accumulate. Regionally, the which contains water) and organic matter supplied
depth of the CCD is a function of a number of vari- by ocean-surface productivity. The concentration
ables which reflect oceanic productivity patterns of a particular component subsequently observed in
and the shoaling of the lysocline near continental
margins. The CCD varies between 3500 and
5500 m in the Atlantic and Pacific Oceans, but has Cosmic 8.25
source
a mean depth of around 4500 m.
Siliceous oozes are the typical pelagic sediments Meteorites, dust

found beneath regions of high productivity (the Windblown

dust At mos p here Ash
equatorial and polar belts and areas of coastal
upwelling), especially where the sea floor is deeper Land Volcanic
source source
than the CCD. The dissolution of opaline skeletons
Rivers
within the surface sediments releases silica to deep
waters; its dissolution within the sediments forms
silica-rich interstitial solutions which migrate along SHELF
SEDIMENTS Turbidity currents, PYROCLASTIC
SEDIMENTS
bedding planes and fractures to precipitate in near- slumps

by permeable lenses and layers, as deep-sea cherts. SEDIMENT

RIDGES
Such deposits seem to occur more frequently at DEEP SEA
FANS
particular times in the geological record (e.g., in the Contour
currents
late Eocene), which may reflect changes in the glob-
al supply of silica to the ocean or changes in ocean- Ocean CONTINENTAL
RISE
ic productivity. RED
CLAY
Pelagic clays are generally found only in deep Turbidity
areas far from land. They generally contain less currents
ABYSSAL
PLAINS
than 10% biogenic material and are mainly com- Bottom currents

posed of clay minerals and fine-grained quartz, the

Bottom currents
bulk of which have been derived from aeolian fall- Turbidity currents
TRENCHES

out. They are commonly reddish brown to choco-

late brown and have accumulated slowly, at gener-
ally less than 1 mm per 1000 years, compared to
the 10–30 mm per 1000 years typical of calcareous Figure 8.25 Processes of clastic sedimentation in the
and siliceous oozes. oceans (after Hay17).

130
 8: The Evolution and Structure of Ocean Basins

Figure 8.26 Schematic representation of trends of pore-water concentration profiles of Concentration 8.26
oxidants (O2 and NO3–) and oxidation–reduction products (Mn2+ and Fe2+) against sub-bot-
tom depth in deep-sea sediments as a result of bacterially mediated early diagenesis. –
Oxygen and nitrate from sea water are consumed during oxidation of organic matter [for- NO3
O2
mally (CH2O)106(NH3)16(H3PO4)] according to reactions (a) and (b) below. Mn2+ and Fe2+ do
not co-exist with high pore-water oxygen concentrations, but are present in pore waters as Mn2+
a consequence of similar reactions of organic-matter oxidation deeper in the sediments,

Depth
with Mn and Fe oxyhydroxides initially present in the solid phase. These reactions may
take place within a few centimetres or over several metres in deep-sea sediments, depen-
dent mainly on the relative fluxes of organic matter and oxygen (reprinted from Froelich et
al.15; © 1979 Elsevier Science Ltd, with kind permission).

(CH2O)106(NH3)16(H3PO4) + 138O2 = 106CO2 + 16HNO3 + H3PO4 + 122H2O (a)

Fe2+
(CH2O)106(NH3)16(H3PO4) + 94.4HNO3 = 106CO2 + 55.2N2 + H3PO4 + 177.2H2O (b)

a sediment of a particular age, however, depends reduction of water content and porosity with burial
not only on the original relative rates of supply of depth and time, and concurrent chemical changes,
all the sedimentary constituents, but also on the which result in cementation and chemical com-
preservation of each constituent after burial. The paction. It is these diagenetic processes of lithifica-
physical, chemical, and biological processes respon- tion which convert carbonate sediments into lime-
sible for converting an original, water-rich, uncon- stones, siliceous oozes into cherts, clays into clay-
solidated sediment into solid rock are collectively stones, and so on. Organic-rich marine sediments
termed diagenesis. which have undergone these later diagenetic modi-
Deep-sea sediments contain about 0.2–0.5% fications are also the potential source rocks for
organic carbon, while shelf sediments contain up to most of the world’s oil reserves (see Chapter 21).
5%. Although organic matter is therefore a minor What conditions are required for the initial forma-
component, its importance is out of all proportion tion of organic-rich sediments remains controver-
to its abundance. As organic matter is the only sial. The traditional interpretation that black (i.e.,
reductant supplied to sediments in any quantity, its organic-rich) shales develop under anoxic water
microbially mediated degradation drives early dia- columns13 has been challenged by the alternative
genetic reactions, which progressively consume the contention that high primary productivity provides
oxidants (electron acceptors) available in sediment the first-order control30. Regardless of the relative
pore-waters (e.g., oxygen, nitrate, and sulphate) and importance of preservation and productivity, how-
as coatings on sediment grains (e.g., manganese and ever, organic matter undergoes progressive com-
iron oxyhydroxides). This gives rise to a vertical plex diagenesis as it is buried and experiences
succession of geochemical environments or zones in increases in pressure and temperature over a pro-
the sediments15 (Figure 8.26) as the electron accep- longed period (Figure 8.27). This diagenesis causes
tors are consumed in the order of decreasing ther- a loss of hydrogen and oxygen relative to carbon in
modynamic advantage (i.e., in the order of decreas- the residual sediment, compared with the organic
ing energy produced by each mole of carbon that is matter originally deposited.
oxidised). One classification of these successive The low temperature, bacterially mediated, early
early diagenetic environments is in terms of the con- diagenesis discussed above is referred to as eogenesis,
centrations of oxygen and total dissolved sulphide while the later reactions deeper in the sedimentary
of the pore water; it recognises oxic, post-oxic, sul- column are referred to as catagenesis or metagenesis.
phidic, and methanic environments with increasing Catagenesis takes place at moderate temperatures
depth. Other redox-sensitive elements in the sedi- (50–150°C) and pressures (30–150 MPa), and
ments also respond to the changes in geochemical includes the ‘oil window’, which generates liquid
conditions experienced within these different envi- hydrocarbons of medium-to-low molecular weight.
ronments. In particular, the abundant metals iron Metagenesis occurs at higher temperatures and pres-
and manganese form a variety of authigenic miner- sures, and generates methane in the ‘dry gas’ zone1,7.
als (i.e., new minerals formed from constituents pre-
existing within the sediment), which can also impart The Oceans in the Past
characteristic colours to the sediments5. Deep-sea drilling has shown that conditions in
As sediments are progressively buried, they ancient oceans were sometimes different from those
experience both physical changes, which result in a of today16,18,22. Mid-Cretaceous sediments, cored by

131
 R.B. Whitmarsh, J.M. Bull, R.G. Rothwell, and J. Thomson

8.27 Figure 8.27 Calculating the volume of hydrocarbons

3 6 3 generated from a given source-rock unit. The yield of
Hydrocarbon 8 10 12 14 16 18 m x 10 gas/km
generation hydrocarbons generated per 1% total organic carbon
from a 1% TOC
Shale with m3 x 106 oil/km3 (TOC) in the source rock is indicated on the horizontal
Type II keragen 2 4 6 8 10 12
1000 3000 5000 p.p.m. (wt/wt) axes in p.p.m. (wt/wt) and other units. As shown, C8–C15
Biogenic methane and C15+ hydrocarbons (the major components of a typi-
cal North Sea oil) are generated in large quantities from
100 80–130°C, while over this temperature range the light
Immature hydrocarbons (C2–C7 and methane) are present in rela-
50 tively small quantities. Once heavy hydrocarbon genera-
tion has ceased, at about 130°C, a presumed cracking
Early reaction takes over, increasing the yield of the C2–C7 frac-
200
mature
for
tion and CH4 at the expense of heavier hydrocarbons
oil 100 (from Brooks et al.8, reproduced with permission).
Temperature (˚F)

Temperature (˚C)
Mid
C15+ C8 –C15 mature
(oil) for oil
(cond-
ensate) Late
the Deep Sea Drilling Project and the Ocean
mature
for oil 150
Drilling Program from the Atlantic and some parts
300
of the Pacific, commonly contain organic-rich
black shales, testifying to periods of probably quite
C2 –C7
(wet gas)
brief, but widespread, anoxia. The black shales
200 may have been caused by the lack of a regular sup-
Methane
400 CH4 ply of cold, dense, well-oxygenated water to the
(dry gas)
deep oceans, due to the absence of ice-caps, and by
abundant biological production encouraged by the
250 warm conditions and extensive continental shelves
of the time. Recently, it has been suggested that
0 100 200 300 400 500 widespread volcanism during the Cretaceous may
TOC (mg/g) have played a role in causing contemporary deep-

Box 8.3 Oxygen Isotope Stratigraphy

Oxygen has three stable isotopes (16O, 17O, and 18O) with atomic mass numbers of 16, 17, and 18:

• 16O makes up 99.763% of natural oxygen.

• 17O makes up 0.033% of natural oxygen.
• 18O makes up 0.204% of natural oxygen.

Oxygen makes up 90% of water by weight; the 16O isotope is lighter than the 18O isotope. Therefore,
16
O is preferentially evaporated relative to 18O.
During glacial periods, 16O-enriched water vapour is precipitated as snow which builds up to form
glacier ice and ice caps. This ice is relatively depleted in 18O. The oceans, however, become relatively
enriched in 18O, because of evaporation of 16O-enriched water vapour. The larger the ice caps, the
larger the proportion of 16O removed from seawater, and the more the 18O:16O ratio of the sea water
increases.
Marine organisms, such as foraminifera, which form skeletons or tests of calcium carbonate, incor-
porate different proportions of 16O and 18O from the water, according to the temperature; but, more
importantly, according to the background ratio of 18O:16O in the sea water, which reflects global ice vol-
umes. Measurements of the small differences in the 18O:16O ratio in different samples using a mass
spectrometer allow the sequence and age of warm and cold conditions to be determined.
The 18O:16O ratio of foraminifera, especially benthic species which live in low-temperature bottom
water (and hence are not affected by temperature changes), can therefore be taken as a measure of
the amount of water held in ice sheets at any given time, and hence also as an indicator of global sea
level.

132
 8: The Evolution and Structure of Ocean Basins

Figure 8.28 Combined plot of global production of 5

8.28
oceanic crust, high latitude sea-surface palaeotempera-
tures, long-term eustatic sea-level, black-shale deposi- 250

Sea level (m)

OIL
tion, and rate of production of the world’s oil resources 4
against geological time (from Late Jurassic to Pleistocene; 200

Oil resources
Figure 1 from Larson21, by permission of the author).

(109 t/Myr)
SEA LEVEL
150
3 25

Palaeotemp (˚C, ice free)

sea anoxia. The resulting increase in outgassing of PALAEOTEMP 20

mantle volatiles and CO2 and increased ocean crust 2 15

production (Figure 8.28) at this time may have 10

resulted in an enlarged supply of nutrients and car-
1 5
bon to the ocean. The ‘greenhouse’ effect of
35
increased CO2 in the atmosphere led to relatively

Ocean-crust production
high sea-surface temperatures. The resultant explo-

(106 km3 /Myr)

30
sion in productivity led to vast increases of organic
carbon in the marine system.
25
During the Cenozoic, the CCD fluctuated widely OCEAN
CRUST
(Figure 8.29), which may have been related partly to
20
changes in sea level. Continental shelves, being shal-
low, are favourable places for carbonate accumula- BLACK SHALES

tion in times of high sea-level and may, therefore, act

OLIG

TITH
EOC
MIO

PAL
CRETACEOUS
as carbonate traps, thereby removing CaCO3 from PLI
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
the oceanic chemical cycle. However, changes in
Millions of years
productivity also lead to changes in the CCD, since
biogenic production of CaCO3 skeletons lowers sat-
uration. Therefore, fluctuations in the CCD may was particularly pronounced toward the end of the
possibly reflect productivity fluctuations too. Eocene (Figure 8.28), presumably due to the ther-
Oxygen isotope studies (Box 8.3) of the skele- mal isolation of the Arctic and Antarctic oceans
tons of benthic foraminifera have shown a general from the rest of the world ocean, but also perhaps
cooling trend in the oceans since the Cretaceous. due to albedo changes resulting from changing veg-
Palaeoceanographic changes, caused by plate etation and snow cover. This led to the equator-
motions over the same period, led to greater parti- ward shifting of climatic belts and, on high latitude
tioning of the world ocean system with time and shelves, to the cooling of water, which became cold
played a major role in causing this trend. and dense enough to sink and fill the deep ocean
Palaeoceanographic studies show that polar cooling basins. The late Eocene cooling therefore resulted
in a new type of world ocean – one characterised
8.29
by the development of marked contrasts between
3 high and low latitudes, between different oceans,
and between the deep sea and the ocean margins.
Large amounts of ice-rafted debris in sediments
Non-equatorial Pacific around Antarctica since the middle Miocene testify
to the build up of the Antarctic ice cap. The micro-
Depth (km)

fossil record, particularly of siliceous types, indi-

4 cates that fundamental changes in deep-water cir-
culation were occurring in the Miocene, concur-
rently with the build up of Antarctic ice and the
Equatorial Pacific world-wide cooling of abyssal waters. Subsequent
changes in palaeogeography, particularly the north-
ward drift of land masses and possibly the closing
of the Panama seaway (about the middle Pliocene),
5
reinforced by other mechanisms, such as mountain
building, resulted in the onset of northern hemi-
0 10 20 30 40 50
sphere glaciation 2.5–3.5 Myr ago and in the
Time (Myr BP) oceans we know today.
Figure 8.29 Reconstructions of past fluctuations in cal-
cite compensation depth in the Pacific Ocean (recon-
struction from Van Andel37; after Seibold and Berger34).

133
R.B. Whitmarsh, J.M. Bull, R.G. Rothwell, and J. Thomson

General References the oceanic lithosphere, Nature, 344, 855–858.

10. Carlson, R.L. and Raskin, G.S. (1984), Density of the ocean
Barnes, M.A., Barnes, W.C., and Bustin, R.M. (1984), crust, Nature, 311, 555–558.
Diagenesis 8: Chemistry and evolution of organic 11. Chase, C.G. (1978), Plate kinematics: The Americas, East
matter, Geosci. Canada, 11, 103–114. Africa, and the rest of the world, Earth Planet. Sci.
Brooks, J. and Fleet, A.J. (eds) (1987), Marine Petroleum Source Lett., 37, 355–368.
Rocks, Blackwell Scientific Publications, Oxford, for 12. Colin, J.P., Ionnides, N.S., and Vining, B. (1992), Mesozoic
The Geological Society. stratigraphy of the Goban Spur, south-west Ireland,
DeMets, C., Gordon, R.G., Argus, D.F., and Stein, S. (1990), Marine Petrol. Geol., 9, 527–541.
Current plate motions, Geophys. J., 10, 425–478. 13. Demaison, G.J. and Moore, G.T. (1980), Anoxic environ-
Kennett, J.P. (1982), Marine Geology, Prentice-Hall, Eaglewood ments and oil source bed genesis, American
Cliffs, NJ, 813 pp. Association of Petroleum Geologists Bulletin, 64,
McKenzie, D., (1978), Some remarks on the development of 1179–1180.
sedimentary basins, Earth Planet. Sci. Lett., 40, 14. DeMets, C., Gordon, R.G., Argus, D.F., and Stein, S.
25–32. (1990), Current plate motions, Geophys. J., 10,
Mutter, J.C., Buck, W.R., and Zehnder, C.M. (1988), 425–478.
Convective partial melting: I. A model for the forma- 15. Froelich, P.N., Klinkhammer, G.P., Bender, M.L., Luedtke,
tion of thick basaltic sequences during the initiation N.A., Heath, G.R., Cullen, D., Dauphin, P.,
of spreading, J. Geophys. Res., 93, 1031–1048. Hammond, D., and Hartman, B. (1979), Early oxida-
Open University Course Team (1989), Ocean Chemistry and tion of organic matter in pelagic sediments of the
Deep Sea Sediments, Pergamon Press, Oxford, 134 eastern equatorial Atlantic: suboxic diagenesis,
pp. Geochim. Cosmochim. Acta, 43, 1075–1090.
Parsons, B. and Sclater, J.G. (1977), An analysis of the variation 16. Haq, B.U. (1984), Palaeoceanography: A synoptic overview
of ocean floor bathymetry and heat flow with age, J. of 200 million years of ocean history, in Marine
Geophys. Res., 82, 803–827. Geology and Oceanography of Arabian Sea and
Seibold, E. and Berger, W. (1993) The Sea Floor, An Coastal Pakistan, Haq, B.U. and Milliman, J.D. (eds),
Introduction to Marine Geology, Springer Verlag, Van Norstrand Reinhold Co., New York, pp
Berlin, 356 pp. 201–231.
Vine, F.J. and Matthews, D.H. (1963), Magnetic anomalies over 17. Hay, W.W. (1974), Studies in Paleooceanography, Society
oceanic ridges, Nature, 199, 947–949. of Economic Palaeontologists and Mineralogists,
White, R.S. and McKenzie, D.P. (1989), Magmatism at rift Tulsa, Oklahoma, p. 2.
zones: the generation of volcanic continental margins 18. Hsu, K.J. (1986), Mesozoic and Cenozoic Oceans, American
and flood basalts, J. Geophys. Res., 94, 7685–7730. Geophysical Union, Geological Society of America
White, R.S., McKenzie, D.P., and O’Nions, R.K. (1992), (Geodynamics Series Volume 15), 153 pp.
Oceanic crustal thickness from seismic measurements 19. Isacks, B., Oliver, J, and Sykes, L.R. (1968) Seismology and
and rare earth element inversions, J. Geophys. Res., the new global tectonics, J. Geophys. Res., 73,
97, 19683–19715. 5855–5899.
20. Jenkyns, H.C. (1986), Pelagic environments, in Sedimentary
Environments and Facies, 2nd edn, Reading, H.G.
References (ed.), Blackwell Scientific Publications, Oxford, pp
1. Barnes, M.A., Barnes, W.C., and Bustin, R.M. (1984), 343–397.
Diagenesis 8: Chemistry and evolution of organic 21. Larson, R.L. (1991), Geological consequences of super-
matter, Geosci. Canada, 11, 103–114. plumes, Geology, 19, 963–966.
2. Becker, K. et al. (1989), Drilling deep into young oceanic 22. Leggett, J.K. (1985), Deep-sea pelagic sediments and palaeo-
crust at Hole 504B, Costa Rica Rift, Rev. Geophys., oceanography: a review of recent progress, in
27, 79–102. Sedimentology: Recent Developments and Applied
3. Berger, W.H., Adelseck, C.G., and Mayer, L.A. (1976), Aspects, Brenchley, P.J. and Williams, B.P.J. (eds),
Distribution of carbonate in surface sediments of the Geological Society Special Publication No. 18,
Pacific Ocean, J. Geophys. Res., 81, 2617–2627. Blackwell Scientific Publications, Oxford, pp 95–121.
4. Bergman, E.A. (1986), Intraplate earthquakes and the state 23. Lorenzo, J.M. et al. (1991), Development of the
of stress in the oceanic lithosphere, Tectonophys., continent–ocean transform boundary of the southern
132, 1–35. Exmouth Plateau, Geology, 19, 843–846.
5. Berner, R.A. (1981), A new geochemical classification of sed- 24. McKenzie, D.P. and Bickle, M.J. (1988), The volume and
imentary environments, J. Sediment. Petrol., 51, composition of melt generated by extension of the
359–365. lithosphere, J. Petrol., 29, 625–679.
6. Bott, M.H.P. (1982), The Interior of the Earth, 2nd edn, 25. Morgan, J.V. and Barton, P.J. (1990), A geophysical study
Edward Arnold (Publishers) Ltd, London. of the Hatton Bank volcanic margin: a summary of
7. Brooks, J. and Fleet, A.J. (eds) (1987), Marine Petroleum the results from a combined seismic, gravity and mag-
Source Rocks, Geological Society Special Publication netic experiment, Tectonophys., 173, 517–526.
No.26, Blackwell Scientific Publications, Oxford, for 26. Mutter, J.C. and Karson, J.A. (1992), Structural processes at
The Geological Society. slow-spreading ridges, Science, 257, 627–634.
8. Brooks, J. et al., (1987), The role of hydrocarbon source 27. Nance, R.R., Worsley, T.R., and Moody, J.B. (1988), The
rocks in petroleum exploration, in Marine Petroleum supercontinent cycle, in Shaping the Earth Tectonics
Source Rocks, Brooks, J. and Fleet, A.J. (eds), of Continents and Oceans, Moores, E.D. (ed.),
Geological Society Special Publication No.26, Readings from Scientific American, W.H. Freeman
Blackwell Scientific Publications, Oxford, for The and Co., New York, pp 177–188.
Geological Society, pp 17–46. 28. ODP Leg 148 Shipboard Scientific Party (1993), ODP Leg
9. Bull, J.M. and Scrutton, R.A. (1990), Fault reactivation in 148 barely misses deepest layer, EOS Trans. Am.
the Central Indian Ocean Basin, and the rheology of Geophys. Un., 74(43), 489–494.

134
 8: The Evolution and Structure of Ocean Basins

29. Parsons, B. and Sclater, J.G. (1977), An analysis of the vari- 36. Stow, D.A.V. (1986), Deep clastic seas, in Sedimentary
ation of ocean floor bathymetry and heat flow with Environments and Facies, 2nd edn, Reading, H.G.
age, J. Geophys. Res., 82, 803–827. (ed.), Blackwell Scientific Publications, Oxford, pp
30. Pedersen, T.F. and Calvert, S.E. (1990), Anoxia versus pro- 399–444.
ductivity: What controls the formation of organic-car- 37. Van Andel, Tj.H., Thiecke, J., Sclater, J.G., and Hay, W.W.
bon-rich sediments and sedimentary rocks?, Amer. (1977), Depositional history of the South Atlantic
Assoc. Petrol. Geol. Bull., 74, 454–466. Ocean during the last 125 million years, J. Geol., 85,
31. Pickering, K.T., Hiscott, R.N., and Hein, F.J. (1989), Deep 651–698.
Marine Environments, Clastic Sedimentation and 38. Vine, F.J. and Matthews, D.H. (1963), Magnetic anomalies
Tectonics, Unwin Hyman, London, 416 pp. over oceanic ridges, Nature, 199, 947–949.
32. Sandwell, D.T. and Smith, W.H.F. (1994), New global 39. Watts, A.B., Bodine, J.H., and Steckler, M.S. (1980),
marine gravity map/grid based on stacked ERS-1, Observations of flexure and the state of stress in the
Geosat and Topex Altimetry, EOS Trans. Am. oceanic lithosphere J. Geophys. Res., 8, 6369–6376.
Geophys. Un., 75(16), 321. 40. White, R.S., McKenzie, D.P., and O’Nions, R.K. (1992),
33. Sayanagi, K. and Tamaki, K. (1992), Long-term variations Oceanic crustal thickness from seismic measurements
in magnetization intensity with crustal age in the and rare earth element inversions, J. Geophys. Res.,
northeast Pacific, Atlantic and southeast Indian 97, 19683–19715.
Oceans, Geophys. Res. Lett., 19, 2369–2372. 41. Whitmarsh, R.B., Pinheiro, L.M., Miles, P.R., Recq, M., and
34. Seibold, E. and Berger, W. (1993) The Sea Floor, an intro- Sibuet, J.-C. (1993), Thin crust at the western Iberia
duction to marine geology, Springer Verlag, Berlin, ocean–continent transition and ophiolites, Tectonics,
356 pp. 12, 1230–1239.
35. Stein, C.A. and Stein, S. (1992), A model for the global vari- 42. Wiens, D.A., et al. (1985), A diffuse plate boundary model
ation in oceanic depth and heat flow with lithospheric for Indian Ocean tectonics, Geophys. Res. Lett., 12,
age, Nature, 359, 123–129. 429–432.

135
 CHAPTER 9:

Slides, Debris Flows,

and Turbidity Currents
D.G. Masson, N.H. Kenyon, and P.P.E. Weaver

Introduction flow) associated with the 1929 Grand Banks earth-

Gravity-driven flows, in a variety of forms ranging quake, from which the first velocity estimate for a
from turbulent suspensions to coherent sliding turbidity current was produced5 (Figure 9.1); this is
masses, are the major agents of downslope sedi- discussed later. Understanding of downslope sedi-
ment transport in the deep sea. They sculpt the ment transport processes has developed, not only
continental slopes into complex shapes, carry land- through studies in the modern ocean, but also
derived sediment into the deep ocean basins, and through studies of ancient marine sequences now
redistribute biogenic sediment on a vast scale. Slope exposed on land, and through experimental work
failures and resultant flows are often near-instanta- in the laboratory. Important contributions include
neous events, capable of the destruction of marine the concept of sequential deposition of fining-
installations and submarine telecommunications upward sediment in individual turbidites (the sedi-
cables and, in some extreme cases, of generating mentary layers or deposits laid down by turbidity
deadly tsunamis. In ancient rocks, sand bodies once currents 1 ), and the comprehensive theoretical
deposited by gravity flows, such as the sands found analysis of turbidity current flow and turbidite
in submarine sediment fans, are a major reservoir deposition12–14. The development of seismic profil-
facies for oil and gas, and have considerable eco- ing and side-scan sonar equipment over the past 40
nomic importance. years (see Chapter 20) has revolutionised the way
The study of gravity flows in the deep ocean has in which we analyse the sea floor, leading to a new
progressed rapidly since the early 1950s. Perhaps appreciation of the extent and importance of gravi-
the best-known study is the analysis of cable breaks ty flows and their deposits. The discovery of huge
caused by the turbidity current (a sediment-laden sediment slides on the flanks of the Hawaiian

9.1
0
30 Velocity
(m/s) 1000
20 Instantaneous
breaks near
10 epicentre 2000
Depth
0
(m)
800 600 Distance (km) 400 3000
59

797 618 541 183 4000

5000

800 600 400 200 0

Distance (km)

Figure 9.1 Cross-section through the continental slope and rise south of Newfoundland in the area affected by the
1929 Grand Banks earthquake and turbidity current. Green arrows mark the positions of cable breaks, with the time
after the earthquake in minutes. Inset shows the turbidity current velocity as deduced from the timing of cable breaks
(redrawn from Heezen and Ewing5).

136
 9: Slides, Debris Flows, and Turbidity Currents

Table 9.1. Statistics of some major slides, debris flows, and turbidity currents.a
Name/Location Waterdepth (m) Area (km2) Length (km) Thickness (m) Volume (km3) Slope

Nuuanu Slide (Hawaii)b 0–4600 23,000 230 up to 2000 5000 ?5→–0.1°

Storegga Slide/Debris 150–3000 112,500 850 up to 430 5580 ?1.5→0.05°

Flowc

Saharan Slide/Debris 1700–4800 48,000 700 5–40 600 1.5→0.1°

Flow

Canary Debris Flow 4000–5400 40,000 600 up to 20 400 1→0°

f turbidite (Madeira ?–5400 >60,000 1000+ up to 5 190 ?0.2° average

Abyssal Plain)

1929 Grand Banks 600–6000 160,000 1100 ? up to 3 185 ?→0.01°

Turbidite

a
Turbidite areas are those covered by deposit only, slide and debris flow areas include scar and deposit.
b
Nuuanu Slide flowed uphill for final 140 km.
c
total of three slide events.

Islands using the GLORIA long-range side-scan Size and Scale

sonar (see Chapter 19) is one example of the appli- The largest slope failures on earth occur around the
cation of these technical advances15. margins of and in the ocean basins. This is a conse-
quence of the relief and shape of ocean basins (see
Classification of Gravity-Driven Chapter 8), as well as the huge quantities of uncon-
Sediment Flows solidated or partially consolidated sediment which
Gravity-driven sediment transport includes a wide occur on smoothly sloping continental margins,
variety of processes, such as slumping, sliding, often under geotechnical conditions only marginal-
debris flow, grain flow, and turbidity currents. ly in favour of slope stability. Should failure occur,
However, sediment slides, debris flows, and turbid- the ocean floor offers unimpeded slopes and flat-
ity currents are the three major gravity-driven floored basins hundreds of kilometres in length,
processes which transport significant volumes of allowing flow over enormous distances.
sediment over large distances in the deep ocean. Individual sediment slides and debris flows can
A slide is defined as the movement of an upper involve many thousands of cubic kilometres of
layer on a basal failure surface. It can result in the material (Table 9.1). The largest of the huge slides
downslope transport of large coherent blocks of on the flanks of the Hawaiian Islands is up to 2 km
material, with internal deformation ranging from in thickness, with a volume of 5000 km3. A volume
negligible to severe. as great as 20,000 km3 has been ascribed to the
Debris flow has been described as the movement Agulhas Slide, off South Africa, but the available
of granular solids, sometimes mixed with minor evidence perhaps suggests that this is a complex of
amounts of entrained water (or, on land, air) on a failures rather than a single gigantic event. The
low slope. A common and effective analogue is Storegga Slide2, off Norway, and the Canary and
with the movement of wet concrete. Saharan Debris Flows11, off West Africa, all have
A turbidity current is a type of gravity or density runout distances of 600–800 km, much of this on
current driven by gravitational buoyancy forces slopes less than 0.5°.
resulting from the difference in density between two The volume of sediment carried by the largest
fluids. To the geologist, it is the downslope flow, known turbidity currents is an order of magnitude
under the influence of gravity, of a suspension of less than that of the largest sediment slides and
sediment in water. The sediment particles, kept in debris flows. Individual turbidites of 100–200 km3
suspension primarily by turbulence, provide the are known from several abyssal plain basins in the
excess density which drives the flow. Atlantic. However, transport distances can be

137
 D.G. Masson, N.H. Kenyon, and P.P.E. Weaver

9.2 Figure 9.2 TOBI 30 kHz side-scan sonar image showing

1 km a slope failure in sediments on the flank of an abyssal hill
(A), almost certainly caused by undercutting of the slope
by erosion of the channel floor (B). Note the progressive
disintegration of the large slabs of sediment as they slide
toward the channel floor. The topographic relief between
the channel floor and the crest of the abyssal hill is about
80 m.

rupted matrix. Deposits of this type are widespread

A in the geological record (Figure 9.3). For example,
Macdonald et al.9 have described a sediment slide
B in Mesozoic sediments in Antarctica which covers
an area of at least 20 km x 6 km. This slide is made
up of coherent blocks, some in excess of 1 km
across, in a mudstone matrix, although the propor-
tion of matrix to blocks is small. Some blocks are
completely undeformed, some show minor defor-
mation only at their edges, while a few are folded.
Almost all observations of debris flows have
spectacular, frequently exceeding 1000 km. It has been made in the subaerial environment (Figure
been suggested, but not proven, that single turbidi- 9.4), but there is no reason to believe that subma-
ty currents may travel up to 4000 km in the rine debris flow processes differ significantly. The
Northwest Atlantic Mid-Ocean Channel, which classic model of debris flow is of an upper raft of
extends from the northern Labrador Sea to the semi-rigid material carried along on a basal layer
Sohm Abyssal Plain south of Newfoundland. undergoing intense shearing. The flow is predomi-
It is important, however, to realise that the low nantly laminar, although some internal mixing is
resolution of the tools commonly used to survey clearly required to account for the observed chaotic
the ocean floor tends to lead to overemphasis of the clast structure of most debris-flow deposits. Debris
role of large-scale failures. In addition to most flows differ from simple viscous fluid flows in that
small failures, the role played by slow creep goes they have a finite strength. This manifests itself in
unrecognised in the modern submarine environ-
ment, even though well-represented in ancient 9.3
rocks.
Sediment Slides and Debris Flows
Processes
Sediment sliding and debris flow are clearly closely
related, with the principal difference being in the
degree of fluidity, deformation, and clast (i.e., frag-
ments of broken sediment layers) mixing. It is clear
that a slide can be transformed into a debris flow
as downslope movement leads to its progressive
disintegration. Many of the larger ‘sediment slides’
described in the literature (e.g., the Saharan Slide)
appear to be complex events involving elements of
both sliding and debris flow.
Most, if not all, submarine slides in which trans-
port over a significant distance is known to have
occurred show evidence for deformation and dis-
ruption of the slide material. Although a slide may
begin as a single displaced block (Figure 9.2), most
slides rapidly disintegrate under the stresses
imposed during transportation. The typical end-
product, when imaged from the sea surface, is an Figure 9.3 Section through a debris flow exposed in a
area of hummocky topography indicative of a mass road cutting in N.W. Ecuador (photograph courtesy of
of displaced blocks embedded in a more highly dis- C.D. Evans, British Geological Survey).

138
 9: Slides, Debris Flows, and Turbidity Currents

9.4 9.5

Figure 9.4. Man-made debris flows of muddy sand pro- Figure 9.5 A large, rafted block in a subaerial debris
duced by a gravel washing plant, showing the steep avalanche on the flanks of Mount Rainier, in the north-
snout and flanks typical of all debris flows. The flows in western US. The block is 50 m long, 40 m wide, and
the foreground are a few centimetres thick (photograph 18 m high (note the person on top of the block for scale;
courtesy of Professor J.R.L. Allen, University of Reading, reprinted from Rockslides and Avalanches, 1, Natural
UK). Phenomena 20, with permission of Elsevier Scientific
Publishing).

the typical steep margins of debris flow deposits port may include a component of sliding or rolling,
and contributes to the ability of debris flows to with the matrix giving some buoyancy and acting
support large clasts. as a lubricant. In submarine debris flows, it seems
Subaerial debris flows are renowned for their likely that sliding is important in the emplacement
ability to carry seemingly impossibly large boulders of the largest rafted blocks. Off Northwest Africa,
(Figure 9.5). In theory, clasts in debris flows are for example, rafted blocks in both the Saharan and
supported primarily by a combination of clast Canary Debris Flows show evidence of having
buoyancy and the cohesive strength of the matrix. moved more slowly than the bulk of the flow, sug-
However, in many flows the largest clasts may not gesting some frictional drag on the underlying sea
be totally supported by the matrix, and their trans- bed (Figure 9.6).

9.6

1 km

Figure 9.6 TOBI 30 kHz side-scan sonar image of part of the Saharan debris flow deposit (see Table 9.1), illustrating a
rafted block (A) with a streamlined ‘halo’ of chaotic debris. The block is about 1 km in length, no more than 50 m
thick, and rises only a few metres above the general level of the flow. The block appears as a dark feature, because of
its relatively flat and smooth surface which back-scatters little sound. The flow is from right to left. The halo, and pres-
sure ridges (B) upstream of the block, suggest that the block moved more slowly than the bulk of the flow, possibly
because it was sliding on the underlying substrate.

139
 D.G. Masson, N.H. Kenyon, and P.P.E. Weaver

9.7 Figure 9.7 Mosaic of GLORIA 6.5 kHz side-scan sonar

images showing the blocky surface of the combined
N Nuuanu and Wailau debris avalanches north of the
25 km Hawaiian Islands of Oahu (bottom, left) and Molokai
(bottom, right). The largest transported block (arrow, cen-
tre) is 20 km long and stands up to 2000 m above the
surrounding sea floor.

Hawaiian Island slides occur within poorly bed-

ded volcanic rock sequences. Slide deposits range
from relatively coherent masses up to 10 km in
thickness (called slumps in the literature) to thinner
(up to 2 km thick), more disaggregated masses
(called debris avalanches). The former are probably
emplaced by slow, intermittent movement, the lat-
ter by individual catastrophic failures. Many of the
mapped slides may be intermediate between these
two end members. The best-known slow moving
failure is the Hilnia Slump on the southern flank of
Kilauea volcano on the main island of Hawaii. This
affects an area of about 5200 km 2, some three
times greater than that of the subaerially exposed
volcano. Bathymetric evidence, showing a stepped
Case studies submarine slope, suggests that the slump consists of
Hawaii two or more enormous rotational slide blocks, each
The idea that submarine landslides were of funda- up to 60 km long and 20 km wide. Alternatively,
mental importance in shaping the Hawaiian Islands the steeper areas of sea floor may represent the
has provoked controversy since the late 1890s, but fronts of enormous slump folds. In 1975, a large
was only confirmed following extensive GLORIA earthquake located beneath the head of the slump
surveys in the late 1980s; submarine landslides was associated with 3.5 m of subsidence and sever-
have now been recognised on the flanks of every al metres of seaward movement along much of the
Hawaiian Island15 (Figure 9.7). Similar failures are south coast of Hawaii. Two similar events occurred
now recognised on the flanks of many other ocean- in the nineteenth century. The earthquake epicen-
ic volcanic islands 6 (Figure 9.8). Landsliding tre, at a depth of around 10 km and approximately
appears to begin even before a submarine volcano coincident with the pre-volcanic sea floor, indicates
grows enough to reach sea level, peaks in frequency that the slump probably involves the entire 10 km
during the shield-building stage of volcanic con- thick volcanic pile, with lateral movement concen-
struction, and continues at a decreasing rate long trated within the pre-volcanic pelagic sediment
after volcanic dormancy. layer which covers the oceanic crust.
9.8
5 km

Figure 9.8 Mosaic of TOBI 30 kHz side-

scan sonar images showing part of a debris
avalanche off Hierro in the Canary Islands
(the arrow shows the flow direction). The
largest blocks are up to 1 km across and
200–300 m high. The lack of any obvious
flow fabric seems to be typical of this type
of flow (see also Figure 9.7).

140
 9: Slides, Debris Flows, and Turbidity Currents

35° 9.9

4
2
Madeira
Abyssal 4
Plain Madeira

Canary Canary Is
Debris Flow

Saharan
Debris Flow KEY
Terrigenous Sediment
Volcaniclastic Sediment 25°
4

Calcareous Sediment
200 km Debris Flow
Debris Avalanche
2

Abyssal Plain

20°W 10°W
Figure 9.9 Map of the northwest African continental margin around the Canary Islands, showing the
locations of known, major, debris flow deposits and turbidity current pathways. Solid arrows show
mapped turbidity current channels; dashed arrows are more generalised pathways where detailed infor-
mation is not available. The debris flows have volumes of several hundred cubic kilometres and transport
material for up to 600 km. Individual turbidity currents, carrying up to 200 km3 of sediment, can travel
distances in excess of 1000 km. Contours are in kilometres.

The largest Hawaiian slide mapped to date is the astonishing example, which occurred around
Nuuanu Debris avalanche, originating on the north 105,000 years ago, reaching over 300 m above
flank of Oahu Island. This avalanche is 230 km sealevel on the island of Lanai. This same tsunami
long, has a maximum thickness of 2 km, covers may even have affected the east coast of Australia,
23,000 km 2 , and has a volume of perhaps some 7000 km distant.
5000 km3. The surface of the avalanche has a dis-
tinctive blocky texture, with a clear downslope West Africa
decrease in the size of the slide blocks (i.e., toward The most common submarine slope failures typical-
the top of Figure 9.7). The largest block, originally ly involve only the upper few metres to tens of
mapped as a seamount, is 30 km long by 17 km metres of relatively unconsolidated sediment. These
wide, with a flat summit about 1800 m above the failures often produce a thin, narrow tongue of
adjacent avalanche surface. The Nuuanu Avalanche debris extending downslope from a distinct failure
flowed across the Hawaiian Deep, the moat-like scar. They can occur on very low slopes (<1–5°)
feature surrounding the islands, and climbed at and have enormous runout distances. The Saharan
least 300 m up the flank of the Hawaiian Arch to and Canary Debris Flows (often referred to as
the north. This indicates that the slide had consid- slides) on the northwest African margin near the
erable momentum. There is evidence that some Canary Islands, are large examples of this type of
debris avalanches give rise to associated turbidity failure11 (Figure 9.9).
currents which can climb at least 500 m, to overtop The Saharan Debris Flow originated at about
the Hawaiian Arch, and which may travel up to 2000 m water depth on the African continental
1000 km from the islands. These catastrophic slope south of the Canary Islands, and flowed
events can also generate huge tsunamis, with one northwest and then west for 700 km across the

141
 D.G. Masson, N.H. Kenyon, and P.P.E. Weaver

9.10

2 km 75 m
A

Figure 9.10 3.5 kHz high-resolution profile across the Saharan debris flow deposit (A), immediately southwest of the
western Canary Islands (see Figure 9.9), showing the typical expression of debris flow deposits (for survey methods see
Chapter 20). In this location, the 25 m thick debris flow deposit sits within and partially fills a broad channel across
which the profile has been taken.

slope and upper continental rise. The sea-floor gra- ern slopes of the Canary Islands, at about 4000 m
dient decreases downslope from about 1.5° in the water depth. It produced a relatively broad
source area to as little as 0.1° near the end of the (60–100 km wide), but thin (usually <20 m thick)
flow. The debris flow incorporated around 600 km3 debris sheet, which extends for 600 km from the
of sediment. Its failure scar is bounded by a com- source area to the edge of the Madeira Abyssal
plex scarp 20–80 m in relief. Southwest of the Plain. This sheet has an average thickness of 10 m,
Canaries, on the continental slope below 4000 m a volume of about 400 km3, and covers an area of
water depth, the debris flow deposit forms a narrow 40,000 km 2. Gradients decrease from 1° in the
tongue about 25 km wide, ranging in thickness source area to effectively 0° at the edge of the
from 5–40 m (Figure 9.10). In this area, high-reso- abyssal plain. The Canary Debris Flow has a com-
lution side-scan sonar data show spectacular images plex outline, which appears to have been strongly
of flow banding, longitudinal shears, lateral ridges, influenced by even the gentlest topography, partic-
and transported blocks (Figures 9.6 and 9.11). ularly at its distal end, where very subtle topo-
The Canary Debris Flow originated on the west- graphic lows (e.g., pre-existing shallow channels)

9.11
1 km
A

Figure 9.11 False-colour TOBI 30 kHz side-scan sonar image showing the edge of the Saharan debris flow deposit
southwest of the western Canary Islands (see Figure 9.9). Blue is low back-scatter, and yellow is high back-scatter. The
lateral ridge (A), which has a relief of about 5 m, is believed to comprise chaotic rubble deposited along the edge of the
flow (B). It is separated from the main part of the flow deposit by a distinct longitudinal shear. The main flow has a char-
acteristic ‘woodgrain’ fabric (C), which may be evidence for drawing out of the debris into a flow-parallel banding.

142
 9: Slides, Debris Flows, and Turbidity Currents

clearly control the path of narrow tongues of Turbidity Currents

debris. The head of the Canary Debris Flow is
somewhat unusual because no clear headwall scarp Processes
is present. Instead, there is a broad zone of appar- Turbidity currents may be generated directly from
ent shallow rotational faults some 30 km in width. sediment suspensions delivered to the shelf edge by
The sediment surface between the faults shows dis- agencies such as tidal currents, rivers, or storms.
ruption increasing downslope until a featureless, Many others appear to have their origins in sedi-
apparently completely homogenised facies is ment failures on the continental slope, although the
reached. Within this facies, rafted blocks of undis- actual initiation mechanisms remain poorly under-
turbed sediments up to 5 km across are seen. stood. One possibility is that they evolve from
Shears within the debris surrounding the blocks, debris flows, in effect continuing the process of dis-
extensional depressions adjacent to their downs- integration which first led from slide to debris flow.
lope margins, and trails of fragments behind some This evolution requires the dilution of the flow by
blocks all suggest that the blocks moved more incorporation of water and a transition from lami-
slowly than the bulk of the flow, presumably nar to turbulent flow. One elegant mechanism,
because they were in contact with, and dragging which has been demonstrated in flume experi-
on, the underlying sea floor. ments, is the generation of turbidity currents by
erosion of the steep snout of a debris flow as it
Storegga Slides moves downslope 4 . An alternative mechanism
Some of the world’s largest known catastrophic involves mixing of water into the body of the flow,
earth movements are found beneath the sea off the perhaps due to internal flow turbulence. This latter
heavily populated coasts of northwestern Europe2. process is attractive because it offers a mechanism
They are in an area of the mid-Norwegian margin for transforming entire debris flows into turbidity
known as the Storegga (‘great edge’), because the currents. However, it has proved difficult to repro-
290 km wide headwall of a submarine slide forms duce in experiments, and its occurrence in nature
the top of the continental slope. remains hypothetical.
The last major slide event occurred about 7000 Proving turbidite–debris flow relationships in the
years ago. It involved erosion of up to 300 m thick- modern ocean basins is difficult, the main problem
ness of slope sediments and displacement of about being the collection of appropriate samples from
1700 km3 of material. The deposits include very deposits which may be spread over thousands of
blocky slide deposits, containing some huge, largely square kilometres. One set of cores, collected across
intact slabs up to 10 km x 30 km in plan view and the snout of the Canary Debris Flow deposit at the
200 m thick. The blocky nature is due to the failure edge of the Madeira Abyssal Plain, does, however,
having cut down into more consolidated sediments. conclusively demonstrate such a relationship
A very thick (up to 20 m) homogeneous fine- (Figures 9.12 and 9.13). Here, the debris flow actu-
grained turbidite covers the entire deep basin of the ally occurs within the turbidite, interrupting the lat-
Norwegian Sea. It is believed to be related to this ter’s fining-upward depositional sequence. It appears
latest slide event. Probable tsunami deposits on the that the faster moving turbidity current began
coasts of Scotland and Norway are also believed to depositing sediment at the edge of the plain, perhaps
be related to the latest slide. They confirm the age a few hours before the arrival of the debris flow,
of about 7000 years. which then buried the lower turbidite layers.
An earlier slide affected an even wider area and Deposition of the finer fraction of the turbidite then
displaced about 4000 km3 of sediment. It involved continued on top of the debris deposit. A wider
shallower, less consolidated sediments and the study of this debris flow–turbidity current pairing
resulting deposits are less blocky. This event indicates that the two phases have markedly differ-
occurred about 30,000 years ago. Two older slide ent (although overlapping) depositional patterns,
events are now known to have occurred in the and that they clearly had divergent paths as they
region and there were probably others whose scars crossed the lower continental slope and rise10. This
have been removed by later erosion, but whose strongly suggests that the debris flow did not contin-
deposits may be present in the deep basin. uously spawn a turbidity current as it moved downs-
The slides are believed to have been triggered by lope, but that the two phases evolved in or near the
earthquakes and the decomposition of gas hydrates source area and then travelled independently.
(see later). Gas hydrates have been recognised on The flow characteristics of large turbidity cur-
seismic profiles from the slope just to the north of rents in the modern ocean have not been directly
the slide. observed. Flow models are based on indirect obser-

143
 D.G. Masson, N.H. Kenyon, and P.P.E. Weaver

vations of both the flow (e.g., cable breaks, Figure

9.12 W
CD56-4 CD56-5
E
CD56-9
9.1) and the structures it leaves along its path (e.g.,
CD56-2
0 channels, levees; see later), geological evidence



@
a a
a a a1
from ancient rock sequences (Figure 9.14), and
CD56-3 flume experiments. The available evidence suggests

@


1 b
a1 a1 that flows range widely in concentration, thickness,
b
turbulence, and velocity, although some of these

@@ @

2 parameters are obviously linked. For example, one
- - - - - - - - -
model for the emplacement of the thick, ungraded
b

@


b b mud turbidites found on abyssal plains is based on
3 thin (<20 m thick), high concentration
(50–100 kg/m3) and low velocity (<1 m/s) flows,

@@
@

which, as they approach their point of deposition,
4
are almost nonturbulent. In contrast, a model for

@


Depth
(m) thin, fine-grained turbidites, such as those which
5
d1 characterise channel-overbank sequences, indicates
thick (up to several hundred metres), low density (5
d 2 kg/m3) flows, but with similar velocities. Overall,
6
d1
e
b velocities are known to vary from <1 m/s to at least
25 m/s.
d The classic turbidite, described in minute detail
7
from ancient rocks, is a fining-upward sequence
with grain size ranging from sand to mud.
e e Turbidites are usually described in terms of the

@
8
CORE KEY Bouma sequence1, a five-fold division based on sed-
Clay imentary structures and grain size. Using flume
f Marl
9
Ooze experiments, the progression of sedimentary struc-
Silt and sand tures has been shown to relate to the progression of
f Debris
b Mud turbidite bedforms seen under a decelerating flow. In nature,
10
few turbidites exhibit a complete Bouma sequence.
For example, the thick fine-grained turbidites
which characterise many abyssal plains may consist
Figure 9.12 Diagrammatic logs of a transect of cores (for entirely of only the upper two Bouma divisions.
a description of coring devices, see Chapter 19) across
the snout of the Canary debris flow deposit, showing the Case studies
relationship between the debris flow and a coeval tur- The 1929 Grand Banks turbidity current
bidite, identified as ‘b’ in the Madeira Abyssal Plain tur- In November 1929, a magnitude 7.2 earthquake
bidite sequence (see Figure 9.15). Lettered units are tur-
occurred beneath the upper continental slope just
bidites which can be correlated across the Madeira
Abyssal Plain. The feather edge of the debris flow deposit south of the Grand Banks of Newfoundland. Six
occurs within ‘b’, indicating that the two flow phases submarine telephone cables in the immediate vicini-
must be part of the same event (see text for a more ty of the epicentre were broken instantaneously and
detailed explanation). a further six, in an orderly downslope sequence,
over the next 13 hours 20 minutes (Figure 9.1). In
attempts to explain these observations, various the-
ories, based on sea-floor faulting or movement of

9.13 Figure 9.13 Gloria 6.5 kHz

side-scan sonar image of the
Canary debris flow snout. The
chaotic nature of the debris-flow
A material (A) generates a high
level of back-scatter (light tones)
and ensures a strong acoustic
B 5 9 contrast with the flat (low back-
scatter) abyssal plain (B).
4 2 Numbers show locations of the
CD56 cores shown in Figure
3 9.12.

144
 9: Slides, Debris Flows, and Turbidity Currents

Figure 9.14 Cyclical turbidite 9.14

sequences believed to be char-
acteristic of sandy fan lobes
(reproduced from E. Mutti’s
Turbidite Sandstones16, by per-
mission of AGIP, Milan, Italy).

sediment leaving sections of cable unsupported, Canary Basin off northwest Africa, are probably
were proposed during the next 20 years. None, the best-studied in the modern ocean basins21. This
however, satisfactorily explained the orderly turbidite sequence has been created by enormous
sequence of cable breaks, the fact that substantial turbidity currents which carry sediment from the
sections of cables were buried (but only on the African continental margin, often over distances in
deeper, less steep, area of the continental slope), or excess of 1000 km. On the plain, these currents
the lack of damage to cables on the continental become ‘ponded’ by the surrounding higher topog-
shelf. The hypothesis that the cables were broken raphy and deposit their sediment load. Indeed, the
by a turbidity current is the only explanation for flat plain results from the stacking of numerous
this combination of observations, as was realised turbidites on top of each other, forming a 350 m
by Heezen and Ewing 5 in their classic paper. thick layer which has levelled off the otherwise
Subsequent sampling in the Sohm Abyssal Plain to irregular topography. The upper 35 m of this
the south of the cable break area proved the exis- sequence, corresponding to the past 750,000 years,
tence of a basin-wide turbidite, underlain by has been sampled, allowing its depositional history
Holocene sediments, at the sea floor. This turbidite to be determined.
covers an area of some 160,000 km2, has a maxi- During that time, a turbidity current reached the
mum thickness in excess of 3 m, and a volume of Madeira Abyssal Plain, on average, once every
about 185 km3. 30,000 years. Most occurred during periods of
From the timing of the cable breaks, Heezen and rapidly changing sealevel (both rises and falls) asso-
Ewing realised that it was possible to calculate the ciated with Pleistocene glacial cycles (Figure 9.15).
velocity of the turbidity current. They estimated a The turbidity currents derive from the northwest
maximum velocity of about 25 m/s on the conti- African continental slope, from the flanks of the
nental slope, decreasing, over a distance of about Canary Islands, and, occasionally, from seamounts
500 km, to about 6 m/s at the edge of the abyssal to the west of the plain.
plain. Lack of precise knowledge of the point of The turbidites deposited on the abyssal plain
origin of the turbidity current, within the general range in volume from a few cubic kilometres to
source area, casts some doubt on their maximum almost 200 km 3, with those over a few tens of
value. Most authorities, however, agree that veloci- cubic kilometres forming layers over the whole
ties of at least 15–20 m/s were attained. Similar plain. They consist predominantly of fine-grained
velocities have since been calculated for turbidity mud; any sand present tends to be deposited at the
currents associated with both the 1954 Orleansville break of slope at the edge of the plain (Figure
(Algeria) earthquake and the 1979 slope failure off 9.16). Much of the coarser sediment is transported
Nice (southern France). across the continental rise through deep-sea chan-
nels, while, in contrast, the finer material appears
The Madeira Abyssal Plain to move as an unconfined sheet flow. The incoming
The turbidites of the Madeira Abyssal Plain (for turbidity currents cause no significant erosion of
location, see Figure 9.9), the deepest part of the the underlying sediments, and successive turbidites

145
 D.G. Masson, N.H. Kenyon, and P.P.E. Weaver

9.15 OXYGEN
OXYGEN TURBIDITE GROUP
TURBIDITE GROUP are separated by sediment layers built up by a slow
ISOTOPE
ISOTOPE organic volcanic
organic volcanic calcareous
calcareous
STAGES
STAGES pelagic rain of biogenic carbonate and wind-blown
aa dust from the Sahara Desert. Fossils in this pelagic
00 record allow us to date the turbidite sequence, and
11
b
b even to assign approximate ages to individual tur-
33
d
d d11
d bidity currents.
ee Turbidites such as those sampled on the Madeira
100
100 55 Abyssal Plain (Figures 9.15 and 9.16), which origi-
ff nate as failures on sedimented slopes, contain a
mixture of sediments (and microfossils) with an age
66
range corresponding to that of the failed sediment
200
g
g mass. In theory, if the pelagic fossil record in the
200
77 h
h source area is well-known, it should be possible to
estimate the age range and thus thickness of the
ii
original sediment failure. If, in addition, the volume
88 of the resultant turbidite is known, then the area
300
300 kk jj eroded to form the corresponding turbidity current
(kyr)

99
AGE (kyr)

can also be calculated. In the case of the Madeira

10
ll Abyssal Plain turbidites, this theory has been put
10
llll
AGE

into practice, allowing a typical sediment failure on

400
400 11
11 the northwest African margin, 1000 km away from
the study area, to be described – this failure is a few
12
ll22
12 ll33 9.16
n
n
500
500 13
13 o
o
p
p
14
14
q
q Mud
600
600
15
15
rr
10 cm

16 Turbidite
16 ss
17 tt
17 uu
700
700
18 vv
18 w
w
20
20 Silt or
Sand
Figure 9.15 Summary of turbidite emplacement on the
Madeira Abyssal Plain, showing the ages of individual
turbidites (each identified by a letter) and turbidite groups
based on source area (organic turbidites – green arrows –
from the northwest African continental margin, Glacial
volcanic – red arrows – from the Canary Islands and
Madeira, and calcareous – blue arrows – from seamounts
Pelagic
to the west of the plain). These groups can be further sub-
divided into turbidites from north (filled arrows) and sediment
south (open arrows) of the Canaries. Ages of turbidites
Interglacial
have been determined relative to the oxygen isotope
time-scale18 (blue and even numbers are glacial periods,
brown and odd numbers are interglacials). Note the
strong correlation of turbidite emplacement with oxygen
isotope stage boundaries, suggesting a relationship
between turbidity currents and changing sea level. Figure 9.16 Core photograph of some typical Madeira
Abyssal Plain sediments. The upper part consists of a tur-
bidite with a black volcanic sand/silt basal unit and a
brown mud top. The lower part consists of pelagic sedi-
ment, with the brown sediment marking deposition dur-
ing a cold glacial climate and the white that during a
warmer interglacial.

146
 9: Slides, Debris Flows, and Turbidity Currents

Figure 9.17 (a) The distribution, through time, of five key %

Isotope
Stage
9.17a
coccolith species occurring in pelagic sediment in the 0
0 50 100
1
Madeira Abyssal Plain area. For sediments deposited dur-
3
ing the most recent half million years, variation in the
ratio of these species gives an age accurate to within one KEY
100 5
or two oxygen isotope stages, i.e. a few tens of thousands
of years. (b) Erosion of sediment representing more than a E.huxleyi
6
few tens of thousand years produces coccolith mixtures G. muelleri 200
not seen in the pelagic record, but dependent upon the 7
Age
age range of sediments which were eroded. Using the G. aperta (kyr) 8
distribution of coccolith species through time (a), it is rel- 300
atively simple to calculate what age range any observed G. caribbeanica 9

coccolith mixture represents. In the example shown, for P. lacunosa

10

turbidite ‘f’ of the Madeira Abyssal Plain sequence 400 11

(emplaced at the end of oxygen isotope stage 6), synthet-
ic mixtures can be created for the erosion of stage 6 sedi- 12

ments only, 6 + 7, 6–8, and so on. By comparing these 500 13

with the actual mixture found in ‘f’ (inset), it can be seen
14
that ‘f’ contains sediments originally deposited during
isotope stages 6–12, i.e. between 130,000 and 480,000
years ago. This corresponds to erosion of about 50 m of
60
sediment from the source region of ‘f’ on the African mar- Turbidite f
9.17b
gin south of the Canaries. 40
100 %
20
80
tens of metres thick, covers an area as great as 0
6000 km2, and incorporates sediment with an age 60
%
range of 50,000–500,000 years (Figure 9.17). An 40
important additional observation based on this
20
study is that none of the turbidites examined con-
tains a significant excess of surface sediment, sug- 0

6 - 11

6 - 12

6 - 13
6-8

6 - 10

6 - 14
stage 6

6-9
6+7

gesting that, once formed, the turbidity currents

which transported them were virtually nonerosion-
al, and that they travelled many hundreds of kilo-
metres in this state.
Turbidites and sediment fans
Turbidites and related deposits, such as debris
flows and debris avalanches, have the greatest vol-
ume of any types of sediment in basin fills. The
processes that form them concentrate the sands
into bodies that are potential reservoirs for hydro- 9.18
carbons. It is necessary for effective hydrocarbon
exploration that the size, shape, and relationships
of the bodies are known. These deposits usually lie
in front of a subaerial and shelf-feeder system, such
as a river drainage basin or a glacially carved cross-
shelf trough. The largest, high-input, feeder sys-
tems, such as the Indus, Amazon, Mississippi, and
Nile, can supply enough material to the deep sea to A
form bodies of sediment (sediment fans) that are
A
10 km or more thick. Glacial-fed submarine fans, A
like that in front of the Barents Sea, can be of an A
equivalent size. However, small fans fed through A
A
the almost ubiquitous submarine canyons which
dissect the continental slopes are much more com-
mon. Submarine canyons can be thousands of Figure 9.18 A hierarchy of tributary gullies feeding into
metres deep, and are usually fed by submarine trib- the main, flat-floored Var Canyon (A), off Nice in the
utary systems with a hierarchy of gullies (Figure northwest Mediterranean Sea. The main canyon is up to
9.18). The fans at the mouths of these canyons are 300 m deep; depth contours are in metres.

147
 D.G. Masson, N.H. Kenyon, and P.P.E. Weaver

9.19 Figure 9.19 High resolution

acoustic profiles showing the
100 m down-fan changes in the mor-
phology of the Indus Fan. (a)
(a) 5 km Large channel–levee system
about 300 m high; (b) small
channel–levee systems nested
above each other; (c) and (d)
(b)
the flat sandy lobe where the
sand content is preventing
much sound penetration
(c) (arrows show locations of
channels).
(d)

difficult to detect as positive morphologic features, nels, which decrease in depth down fan, are seen
often merging together with the neighbouring sys- on swath mapping systems, such as GLORIA, to
tems to form a continental rise or slope apron. have a remarkable resemblance to river channels
Paradoxically, these more common fan types have in their sinuosity. They tend to be highly sinuous
been relatively little-studied because of the difficul- or meandering in their middle reaches 8 (Figure
ty of detecting them and because they are usually 9.20).
sand rich, which makes them difficult to sample Migration of meanders can leave riverine fea-
with conventional corers. tures, such as abandoned reaches reminiscent of
The high-input fans have extensive distributary oxbow lakes. Just as sands in rivers tend to deposit
channel–levee systems, only one of which seems to on the inside of meander bends, in what are called
be active at any one time. This implies that there point bars, so too do sands in submarine channels.
are many instances where there has been a switch Submarine point bars can be as wide as the mean-
to a new channel. The gradients of these large der belts (typically 2–10 km). Sands can also be
fans are low. The levees (banks on either side of deposited within channels, in what are called chan-
the channel) can be hundreds of metres high, as nel lag deposits, causing build-up of the channel
on the upper Indus Fan (Figure 9.19), in contrast floor. This accumulation of sand can be imaged on
to the levees of rivers, which reach a maximum geophysical records, where it is seen as strong
height of only 10 m or so. The submarine chan- acoustic reflectors beneath the channel floor. Sand

9.20

50 km

Figure 9.20 A mosaic of GLORIA images of meandering channels on the middle Indus Fan. The channels are about
1 km across; water depth varies from 3550–3650 m, from left to right. The stronger backscatter is purple.

148
 9: Slides, Debris Flows, and Turbidity Currents

90° 88° 86° 84°

Figure 9.21 Interpre- Mississippi 9.21
Delta Debris flows
tation of the youngest
Sandy depositional lobes
major sequence of the
Major channels
Mississippi Fan based
Limit of GLORIA coverage
on GLORIA survey Lineations on levees
data. The sediment 500
slides have a swirly pat- m
100
tern, whereas the exten- 0m Florida
Escarpment 28°
sive sandy lobes, fed by 0m
De Soto
a single channel, have a 100 Channel
Levee 100 km
distributary, frond-like
pattern (small grey box 0m
at the right shows area 200
of Figure 9.22).

3000 Mississippi
m Levee
Channel

26°
Levee

Campeche
Escarpment

bodies associated with channels are thickest in the sandy lobes which are only a few metres thick,
upper or proximal fan (typically tens to hundreds reaching a maximum of about 20 m. When seen in
of metres thick). Sandy lobes which occur beyond high resolution, individual sandy flow deposits dis-
the ends of channels (i.e., on the distal fan) have play a remarkable frond-like pattern 19 (Figure
only recently been detected on large fans. The 9.22). Each of these deposits was laid down over a
Mississippi fan (Figure 9.21) has a hierarchy of relatively short period, probably at an early stage in

9.22

Figure 9.22 Frond-like

pattern at the end of one
of the sandy lobes of the
Mississippi Fan, seen on
high-resolution side-
scan sonar. The relative-
ly high level of acoustic
back-scatter from the
sandy lobes is shown by
dark colours. The diago-
nal line (top left to mid-
1 km dle right) across the
image lies directly
beneath the sonar vehi-
cle path (courtesy of D.
Twichell, U.S. Geolog-
ical Survey).

149
 D.G. Masson, N.H. Kenyon, and P.P.E. Weaver

9.23 Average gradient : 1 in 500

Basin

Plain
Average gradient : 1 in 30
Le
v ee

Longitudinal bedforms
(erosional?)
Decreasing m
Crescentic scours peak 200
current
Braid bars / channels speed

Sand sheets (?) 100 km

(3-10 m thick)

Figure 9.23 The Umnak turbidite system in the Bering Sea basin, mapped using GLORIA. The subdivisions of the
extensive sandy lobe are shown in blue, green, and yellow; islands of the Aleutian chain are shown in brown (from
Kenyon and Millington7).

a phase of rising sealevel. Hence the flows that gave may dominate the sediment sequences in deep-sea
rise to them must have been frequent. In recent basins, in historical terms they are infrequent
times, flows have occurred approximately once a events, except on the largest river-fed submarine
year on the active Zaire Fan. fans. Since they also occur beneath an opaque blan-
The more common low-input types of fan gener- ket of sea water, it is not surprising that little is
ally have small subaerial drainage basins and well- known about the circumstances in which they are
developed tributary canyons that feed poorly devel- generated. Among the many factors which may
oped, usually single, channel–levee systems. A good contribute to the triggering of slope failures are:
example has been mapped using GLORIA side-scan
sonar on the northern flanks of the volcanic islands • Earthquakes.
of the Aleutian Chain7 (Figure 9.23). It is thought • Loading and oversteepening of slopes.
that this fan is particularly well-imaged by the • Underconsolidation (i.e., when the fluid pressure
sonar because it is young and fresh, being in an within the sediment exceeds hydrostatic -
area of very active earthquakes. The single, pressure), usually due to rapid sedimentation, to
straight leveed channel has a trumpet-shaped gas build-up, or to gas hydrate decomposition
mouth fed from an elaborate distributary channel within the sediment.
system, beyond which is an enormous spread of • Sealevel change.
flat and strongly patterned ground that is thought • The occurrence of slope-parallel weak layers
to be a sandy lobe. The patterns are arranged in within bedded sequences.
zones that indicate a decreasing energy of flow
away from the channel mouth. Maximum channel The majority of historically recorded slope fail-
gradients are usually greater than for the high- ures, recognised because they broke telephone
input fans and channels are fairly straight. Sands cables or caused tsunamis, have been triggered by
should be found mainly in the channel mouth earthquakes. Other historical failures, such as the
lobes (typically tens of kilometres across). Flows in 1979 Nice Slide, may have been caused by man-
low-input fans are believed to be infrequent. For made slope loading. One example of earthquake-
instance, flows on the low-input Var system, off induced failure, from off northern California,
Nice in southern France (Figure 9.18), have occurred during a magnitude 6.5 earthquake in
occurred about once every 1000 years throughout 1980. Parts of the area had fortuitously been sur-
the Holocene. veyed with a high-resolution profiler in the late
1970s, when no deformation was found; repeat
Triggering Mechanisms surveys immediately after the earthquake found
Although individual slope failures may be cata- evidence3 for failure on slopes as low as 0.25°. The
strophic in nature and, cumulatively, their products common occurrence of failures in areas of very

150
 9: Slides, Debris Flows, and Turbidity Currents

rapid sedimentation, such as major river delta 4. Hampton, M.A. (1972), The role of subaqueous debris flow
fronts, is persuasive evidence that underconsolida- in generating turbidity currents, J. Sedimen. Petrol.,
42, 775–793.
tion contributes to the failure process. Such fail- 5. Heezen, B.C. and Ewing, M. (1952), Turbidity currents and
ures have been implicated in the damage, and even submarine slumps, and the 1929 Grand Banks
loss, of oil-drilling platforms in the Mississippi Earthquake, Am. J. Sci., 250, 849–873.
Delta area. In this area, failures have occurred pri- 6. Holcomb, R.T. and Searle, R.C. (1991), Large landslides
from oceanic volcanoes, Marine Geotech., 10, 19–32.
marily during hurricanes, suggesting that loading 7. Kenyon, N.H. and Millington, J. (1995), Contrasting deep-
by wave action may be a triggering factor. sea depositional systems in the Bering Sea, in Atlas of
Many slope failures, however, have no obvious Deep Water Environments: Architectural style in tur-
trigger. This is particularly true on passive continen- bidite systems, Pickering, K.T., Hiscott, R.N.,
tal margins around the North Atlantic, for example Kenyon, N.H., Ricci Lucci, F., and Smith, R.D.A.
(eds), Chapman and Hall, London, pp 196–202.
off the east coast of the US. In these areas slopes are 8. Kenyon, N.H., Amir, A., and Cramp, A. (1995), Geometry
low, as is seismicity. Sedimentation rates are low-to- of the younger sediment bodies of the Indus Fan, in
moderate, and there is no evidence for gas in the Atlas of Deep Water Environments: Architectural
upper sediment column. Failure on slope-parallel style in turbidite systems, Pickering, K.T., Hiscott,
R.N., Kenyon, N.H., Ricci Lucci, F., and Smith,
bedding planes is the common type of failure in this R.D.A. (eds), Chapman and Hall, London, pp 89–93.
situation17. One suggestion is that these failures are 9. Macdonald, D.I.M., Moncreiff, A.C.M., and Butterworth,
the end product of slow deformation under the P.J. (1993), Giant slide deposits from a Mesozoic
action of gravity or under the repeated loading effects fore-arc basin, Alexander Island, Antarctica, Geology,
of earthquakes, none of which, individually, are 21, 1047–1050.
10. Masson, D.G. (1994), Late Quaternary turbidity current
capable of causing failure. Eventually, one or more pathways to the Madeira Abyssal Plain and some
weak layers fail under the cumulative strain effects. constraints on turbidity current mechanisms, Basin
The prediction of submarine slope failures is not a Res., 6, 17–33.
simple task. Nevertheless, the ability to forecast geo- 11. Masson, D.G., Kidd, R.B., Gardner, J.V., Huggett, Q.J., and
Weaver, P.P.E. (1992), Saharan continental rise:
logical hazards, or at least areas prone to such haz- Facies distribution and sediment slides, in Geologic
ards, is of fundamental importance to man’s activi- Evolution of Atlantic Continental Rises, C.W. Poag
ties, both offshore and in low-lying coastal areas. The and P.C. de Graciansky (eds), Van Nostrand
rewards for a better understanding of submarine Reinhold, New York, pp 327–343.
landslides and their causes could be very large indeed. 12. Middleton, G.V. (1966a), Experiments on density and tur-
bidity currents: I. Motion of the head, Can. J. Earth
Sci., 3, 523–546.
Acknowledgement 13. Middleton, G.V. (1966b), Experiments on density and tur-
We would like to acknowledge financial support bidity currents: II. Uniform flow of density currents,
through EC MAST II contracts MAS2-CT94-0083 Can. J. Earth Sci., 3, 627-637.
14. Middleton, G.V. 1967. Experiments on density and turbidi-
and MAS2-CT93-0064. ty currents: III. Deposition of sediment. Can. J. Earth
Sci., 4, 475–505.
15. Moore, J.G., Clague, D.A., Holcomb, R.T., Lipman, P.W.,
Normark, W.R., and Torresan, M.E. (1989),
General References Prodigious submarine landslides on the Hawaiian
Johnson, A.M. (1970), Physical Processes in Geology, Freeman, Ridge, J. Geophys. Res., 94, 17465–17484.
Cooper and Co., San Francisco. 577 pp. 16. Muti, E. (1992), Turbidite Sandstones, Agip and Istituto di
Pickering, K.T., Hiscott, R.N., and Hein, F.J. (1989), Deep Geologica, Universita di Parma, p. 176.
Marine Environments, Unwin Hyman, London, 416 17. O’Leary, D.W. (1991), Structure and morphology of subma-
pp. rine slab slides: clues to origin and behaviour, Marine
Pickering, K.T., Hiscott, R.N., Kenyon, N.H., Ricci Lucci, F., Geotech., 10, 53–69.
and Smith, R.D.A. (1995), Atlas of Deep Water 18. Shackleton, N.J. and Opdyke, N.D. (1973), Oxygen isotope
Environments: Architectural style in turbidite and palaeomagnetic stratigraphy of equatorial Pacific
systems, Chapman and Hall, London, 333 pp. core V28-238: Oxygen isotope temperatures and ice
Simpson, J.E. (1987), Gravity Currents in the Environment and volumes on a 105 year and 106 year scale, Quatern.
the Laboratory, Ellis Horwood, Chichester, 244 pp. Res., 3, 39–55.
19. Twichell, D.W., Schwab, W.C., Nelson, C.H., Kenyon,
N.H., and Lee, H.J. (1992), Characteristics of a sandy
References depositional lobe on the outer Mississippi Fan from
1. Bouma, A.H. (1962) Sedimentology of some Flysch Deposits, SeaMARC 1A side-scan sonar images, Geology, 20,
Elsevier, Amsterdam, 168 pp. 689–692.
2. Bugge, T., Belderson, R.H., and Kenyon, N.H. (1988), The 20. Voight, B. (1978), Rockslides and Sandstones, 1, Natural
Storegga Slide, Phil. Trans. Roy. Soc. London, Ser. A, Phenomenon, Elsevier Scientific, Amsterdam, p. 186.
325, 357–388. 21. Weaver, P.P.E., Rothwell, R.G., Ebbing, J., Gunn, D.E., and
3. Field, M.E., Gardner, J.V., Jennings, A.E., and Edwards, B.E. Hunter, P.M. (1992), Correlation, frequency of
(1982), Earthquake induced sediment failures on a emplacement and source directions of megaturbidites
0.25° slope, Klamath River Delta, California, on the Madeira Abyssal Plain, Marine Geology, 109,
Geology, 10, 542–546. 1–20.

151
 CHAPTER 10:

Mid-Ocean Ridges and

Hydrothermal Activity
C.R. German, L.M. Parson, and R.A. Mills

Mid-Ocean Ridges volcanically active continuous zone on the Earth’s

The mid-ocean ridge system is the largest continu- surface, where hot, soft mantle rises from a depth
ous topographic feature on the Earth’s surface, of several tens of kilometres in the Earth’s interior,
comprising a mountain range of volcanoes and to melt and form abundant basalt magmas (semi-
faulted blocks that, in places, rise several kilome- molten rock) in ponds or reservoirs beneath the sea
tres from the surrounding sea floor. Sections of the floor. The basalt seeps through the overlying rock
ridge system extend throughout all of the world’s in complex plumbing systems and is extruded onto
oceans (Figure 10.1) and total more than the sea-floor surface. Flows become broken and
50,000 km in length, some four times the diameter fragmented as they are shunted off-axis by succes-
of the globe. sive additions of new sea floor. The lava flows, vol-
The ridge marks the zone along which the tec- canoes, and dyke systems combine to create some
tonic plates that comprise the Earth’s surface are 20 km3 of new crust per year in a layer 6 km thick
separating, allowing new sea floor to be created15 (Figure 10.2).
(see also Chapter 8). The new sea floor is generated Continuous separation of the two tectonic plates
at the ridge axes in the form of a dense igneous at a spreading centre induces extensional stress on
rock called basalt. The mid-ocean ridge is the most those sections of ridge where no new sea floor is

10.1
dg es
Ri kjan
e

▲
y

▲ ▲
Re

▲
▲

▲ ▲ ▲
▲

▲
▲ ▲

▲
▲
▲ ▲ ▲
▲

▲
▲

▲
▲
▲ ▲
▲
▲ ▲
▲

▲
▲

▲
Red
▲ ▲

▲
▲

▲ ▲ ▲

▲
▲
▲ Sea
M

▲ ▲ ▲ d
▲

▲ Atla
i

▲
▲

ntic
▲
▲

▲
▲

▲
Ridge

▲ Central
▲

▲
ise

▲
▲

▲
▲ ▲ ▲ ▲ Indian
t Pacific R

▲
▲ ▲

Ridge
▲

▲
▲

▲ ▲
▲

Eas

e
Chi idg
▲

R
ian
le

R is
▲

Southeast Indian R e Ind

▲

idg st
we
▲
▲

e ▲
uth
▲

So
▲
ge
▲

id

R
▲

ic
rct
Pacific Anta

Figure 10.1 Distribution of the principal plate tectonic boundaries and the locus of the global mid-ocean ridge system
is shown in blue. Convergent margins are shown in black, the triangles indicating the direction of subduction.

152
 10: Mid-Ocean Ridges and Hydrothermal Activity

/ Axis
/

!
10.2

6 km

Figure 10.2 The likely complexities of magmatic plumbing in the shallow crust beneath a slow-spreading
ridge system. Basaltic sheet flows and axial volcanoes (brown) are fed by a complex plumbing arrangement
(red). Steeply inclined, sheeted intrusives (purple) carry complex sill/dyke fabrics (orange) and overlie gab-
bros in the lower crust (green). As the new crust moves away from the ridge axis, it cools and large faults
and fissures develop (black).

being generated. These stresses cause thinning and marked by inward-facing fault scarps which may
eventual cracking of the crust. The degree of fissur- have more than 500 m of throw. The faults bound
ing and faulting of the crust varies with the rate of an axial floor, commonly 10–12 km wide, within
spreading; hence, to a first approximation, the which a range of neovolcanic and neotectonic
shape of the ridge system can be used as a clue to activity takes place. Volcanoes, which are common-
the rate of spreading activity26. ly distributed across the axial floor, are fed through
The rate at which different tectonic plates sepa- a complex system of pipes and cracks which con-
rate varies considerably. While the American and nects them to magma chambers. The magma has
European–African plates are moving apart at rates risen and ponded in shallow levels of the crust, in
of between 20 and 60 mm/yr (roughly the speed at what are believed to be discrete supply systems fol-
which fingernails grow), the American and Pacific lowing the melting of crustal rock during the uprise
plates are distancing themselves at nearly ten times of heat from the upper mantle (Figure 10.3).
that speed. The type of sea floor generated at these Inevitably, as the plates continue to move apart,
different speeds is usually very different, in terms of sediments accumulate on the surface of the crust,
the topography and the composition of the basaltic burying the volcanic and tectonic features which
rocks generated to make the crust. In general, characterise the ridge.
faster-spreading ridges are characterised by an ele- The technology which has become available to
vated crestal region, while slower spreading sys- marine scientists during the past two decades has
tems have a well-defined axial valley or trough, allowed the mid-ocean ridge system to be studied

10.3

Fracture
zone

Figure 10.3 The deeper magma

supply system to the mid-ocean
ridge. Regularly spaced diapirs of
magma rise from the mantle, pond
as reservoirs, and feed the central
portions of ridge segments. Here,
the ridge segments are separated
by a fracture-zone offset, which
perturbs the magma supply.

153
C.R. German, L.M. Parson, and R.A. Mills

10.4

Figure 10.4 Enhanced Lansat TM satellite image of the southern branch of the Afar rift triple junc-
tion, northeast Africa, showing an area of 120 km by 120 km. Marked NNE linear fabrics indicate
faulting, which locally cuts and is cross-cut by fresh volcanic features, such as volcanoes and
basaltic lava flows (© C. Oppenheimer and J.P. Rogers, Open University).

with a degree of detail hitherto impossible. On the Figure 10.4, can be elucidated using computer pro-
broadest scale, satellite imagery (see Chapter 5) has grammes that generate high-resolution structural
been used to increase the understanding of mid- maps, which are then interpreted by geologists.
ocean ridge processes10. Although most of the glob- From satellite data, faults can be mapped over
al ridge system is submerged, a few sections are many hundreds of kilometres; their relative
subaerially exposed (e.g., Iceland and the East chronology of activity can be estimated, along with
African Rift). These provide the opportunity to cal- their relationship with volcanic activity, from inter-
ibrate our interpretations of satellite data by direct secting or cross-cutting features. On the ground, as
examination of the outcrop. The developing in Iceland (Figure 10.5), where the Mid-Atlantic
rift–ridge systems of the East African Rift, as illus- Ridge is exposed, even finer details are recorded;
trated in the LANDSAT thematic mapper images in features of ridge systems can be measured, such as

154
 10: Mid-Ocean Ridges and Hydrothermal Activity

Figure 10.5 One of the steep- 10.5

walled rift valleys exposed on
land in Iceland, where the
Mid-Atlantic Ridge passes
from its submarine to a sub-
aerial setting. A volcanic cone
can be seen in the distance.
This section of the rift displays
intense hydrothermal venting
(hot springs, geysers, etc.).

valley width, slope angles of faults, and volumes of ity or topography (see Chapter 20); these new
lava flows. methods have provided the greatest recent advances
Satellite altimetry has been used offshore to map in understanding the origins of the ridge system
the mid-ocean ridge in detail, in places where ships (Figure 10.6).
seldom venture – like the Southern Ocean – and to On the scale of a few metres, molten magma is
demonstrate the extent to which the ridge system is extruded at the sea floor and rapidly cooled to
cut by closely spaced fracture zones. Otherwise, form sheet-like lava flows, rounded pillow basalts,
offshore surveying of the ridge relies on the acquisi- hummocky pillowed topography, or volcanoes of
tion of remotely sensed marine geophysical data varied shapes (Figure 10.7). Piles of pillows accu-
obtained from ships. Methods include conventional mulate to form pillow mounds which, with contin-
seismic-reflection profiling and magnetic and gravi- ued supply, build into volcanoes.
ty surveying, but also new methods of mapping The factors which control the shapes of volca-
whole swaths of sea floor in terms of their reflectiv- noes, their size, and their number along the mid-

10.6 10.7

Figure 10.7 Photograph taken from a submersible at

3080 m water depth, showing pillow basalts outcropping
on the floor of the Mid-Atlantic Ridge at 29°N (the view
is ca 6 m across; © Deep Submergence Laboratory,
Woods Hole Oceanographic Institution).

Figure 10.6 TOBI deeptow side-scan image of the central portion of the Mid-Atlantic Ridge at 29°N (the image width is
ca 5 km). The lighter tones represent areas which are rough unsedimented fresh basalt, or are orientated normal to the
acoustic beam and have higher levels of back-scattered energy. Darker tones mark surfaces with higher acoustic absorp-
tion or shadows. Clear examples of volcanoes and hummocky sea-floor surfaces are recognisable.

155
C.R. German, L.M. Parson, and R.A. Mills

ocean ridge system are not yet well-understood27. ic crust and, in some cases, close to molten magma.
Variations in the composition of volcanic material, In these hot rocks, the sea water is first heated
speed of plate motion, and rate at which magma is before it reacts chemically with the surrounding
fed to the ridge are certainly contributory. More host basalt. As it is heated, the water expands and
significantly, we recognise that the localised coales- its viscosity reduces. If these processes occurred on
cence of pillow basalt mounds and seamounts into land, at atmospheric pressure, catastrophic explo-
linear ridges several kilometres in length is a funda- sions would result as temperatures would rise
mental mechanism in the generation of new crust. above 100°C and the water would turn into steam.
It is a major thrust of marine geoscientific endeav- But, because mid-ocean ridges lie under
our in the 1990s to understand the relationship of 2000–4000 m of sea water, at pressures 200–400
these volcanic features, and their subsurface plumb- times greater than atmospheric pressure, the react-
ing, to the occurrences of hot-fluid circulation sys- ing sea water reaches temperatures up to
tems (hydrothermal activity), and to predict their 350–400°C without boiling. At these temperatures,
location along the world’s spreading ridges. the altered fluids do become extremely buoyant,
however, with densities only about two-thirds that
Hydrothermal Circulation of the downwelling sea water; thus, they rise rapid-
One of the most exciting developments since the ly back to the surface as hydrothermal fluids. The
beginning of the study of oceanography was the dis- movement of the fluid through the rock is such
covery, in the late 1970s, that submarine hydrother- that, while the downward flow proceeds by gradual
mal vents were associated with the volcanically percolation over a wide area, the consequent
active zone at the crest of the mid-ocean ridge sys- upflow is often much more rapid and tends to be
tem (Figure 10.1). The base of the oceanic crust is focussed into natural channels emerging at ‘vents’
extremely hot (>1000°C), yet its upper boundary is on the sea floor.
in contact with sediments and sea water at tempera- Beneath the sea floor, the reactions between sea
tures that are only a few degrees above 0°C. Since water and fresh basalt remove the dissolved Mg2+
the earliest recognition of mid-ocean ridges and and SO42– ions that are typically abundant in sea
their significance to plate tectonics in the 1960s, water, resulting in the precipitation of a number of
geophysicists have known that conductive heat flow sulphate and clay minerals. As the water seeps
measured through young ocean crust could not lower into the crust and the temperature rises, met-
account for all the heat lost at these tectonic spread- als, silica, and sulphide are all leached from the
ing centres. So, they predicted that some alternative rock to replace the original Mg2+ and SO42– ions.
convective heat-transfer process must also occur The hot and, by now, metal-rich and sulphide-bear-
near the crests of mid-ocean ridges33. ing fluids then ascend rapidly through the ocean
It was not until the late 1970s that these predic- crust to the sea floor1. As soon as they begin to mix
tions were proved correct, with the discovery of, with the ambient, cold, alkaline, well-oxygenated
first, low-temperature submarine hydrothermal deep-ocean waters there is an instantaneous precip-
activity (10–30°C) at the Galapagos Spreading itation of a cloud of tiny metal-rich sulphide and
Centre 3 and, second, high-temperature (350°C) oxide mineral grains6. These rise within the ascend-
hydrothermal activity on the East Pacific Rise30. The ing columns of hot water, giving the impression of
importance of these discoveries went much further smoke. Precipitation around the mouths of the
than simply proving the geophysicists’ theories cor- vents over time builds chimneys through which the
rect. To marine geochemists, hydrothermal activity smoke pours, hence the term ‘black smokers’
represented a new supply of chemicals to the oceans, (Figure 10.8); hot water gushes out of these tall
comparable in importance to the influence of rivers chimney-like sulphide spires at temperatures of ca
flowing from the land4. For marine biologists, the 350°C and at velocities of 1–5 m/s. Upon eruption,
discoveries were, perhaps, even more outstanding. this hydrothermal fluid continues to rise several
Here, fed by the chemicals emanating from the hundred metres above the sea bed, mixing with
vents, living in total darkness, and isolated from ordinary sea water all the time, in a buoyant turbu-
almost all other strands of evolution, were extraordi- lent plume (see later).
nary, undescribed species. These animals were previ- At slightly lower temperatures (below 330°C),
ously unsuspected and certainly not looked for, yet the fluid may cool and mix with sea water suffi-
were soon recognised to be living in an entirely novel ciently to deposit some metal-rich precipitates in
ecosystem whose origins could be traced back to the the walls of the channels up which the fluids rise,
earliest life on Earth (see also Chapter 13). before reaching the surface. In such cases, the par-
The pattern of hydrothermal circulation is one ticulate material formed when the ascending hot
in which sea water percolates downward through water finally emerges at the sea bed is made up pre-
fractured ocean crust toward the base of the ocean- dominantly of amorphous silica and various sul-

156
 10: Mid-Ocean Ridges and Hydrothermal Activity

Figure 10.8 Black smokers on the sea floor of the East 10.8
Pacific Rise at approximately 2600 m depth, near 21°N.
Individual chimneys measure approximately 30–40 cm
tall and 10 cm across, and are seen venting hydrothermal
fluids at temperatures of 350–400°C and velocities of
1–5 m/s. The fluid within the chimneys is extremely clear,
but rich in dissolved metals and hydrogen sulphide. As
soon as this fluid mixes with cold oxygen-rich sea water,
at the very mouths of the vents, precipitation of a range of
sulphide and oxide minerals occurs, giving rise to the
clouds of tiny black ‘smoke’ particles, which are seen bil-
lowing upward above the chimneys into the overlying sea
water. A previously active vent chimney, which has been
cemented solid by mineral precipitation, is at the extreme
right. The light, angular area to the top of the structure
represents the relatively fresh internal composition of the
dark grey, almost cylindrical, chimney structure, where it
has been broken open using the robot arm of the sub-
mersible (photograph courtesy of Woods Hole
Oceanographic Institution and the American Geophysical
Union17).

phate and oxide minerals, yielding a white cloud of studied to date is that the dominant species of ani-
mineral precipitates; the common name for these mals at any site often appear to be extremely large.
slightly cooler vents is thus ‘white smokers’5,31. This has raised the question: “Where do the ani-
Individual high-temperature vents at mid-ocean mals get food to grow at all, let alone enough to
ridges may only be ca 10 cm in diameter at their reach such a size?” The answer is that hydrother-
mouth, yet over time, growing like stalagmites from mal vent animals derive their food from a chain
the sea floor, they can form chimneys anywhere that is driven by geothermal (terrestrial) energy,
from 1 m to 30 m tall. A typical vent field might unlike all the other organisms that rely, directly or
comprise several such chimney structures spread indirectly, upon sunlight for their survival. In
over a circular area ca 100 m in diameter. hydrothermal-vent communities, free-living bacte-
Throughout this area there may also be a number of ria, which are anchored on the sea bed or float free
lower temperature vents emitting hot, shimmering in the water column, coexist with symbiotic sul-
water from the sea bed. Even these vents are as hot phide-oxidising bacteria, which live within the larg-
as 10–30°C, which is notably warmer than typical er vent-specific organisms; these exploit the free
deep-ocean water (2–3°C)18,25. It is in the vicinity of energy of reaction released when hydrogen sulphide
these warm and more diffuse emissions that the present in the vent fluids interacts with dissolved
majority of vent-specific biota are most abundant. carbon dioxide and oxygen in ordinary sea water
to form organic matter (equation 10.1), where
Life at Hydrothermal Vents (CH2O)n is a carbohydrate. Because it is a chemical,
Hydrothermal fluids are enriched in dissolved hydrogen sulphide, which plays the role that sun-
hydrogen sulphide, a substance which is toxic to light plays in the more familiar process of photo-
most forms of life; elsewhere in the oceans it is synthesis in the warm surface waters of the oceans
found only in lifeless, stagnant anoxic basins such and on land, this unique deep-ocean, sunlight-
as the Black Sea. It is quite remarkable, therefore, starved process has been given the name chemosyn-
that sites of active hydrothermal venting are not thesis2.
barren wastelands where no life can exist. Instead, n(CO2+ H2S + O2 + H2O) = (CH2O)n + n(H2SO4) (10.1)
scientists diving at the very first vent site to be dis-
covered, on the Galapagos rift in the eastern equa- Approximately 95% of all animals discovered at
torial Pacific, were surprised to find dense concen- hydrothermal vent sites are previously unknown
trations of benthic (sea-bed dwelling) megafauna species. So far, over 300 new species have been
(large animals) living within the vent-field area. identified and, for many of these, the differences
Since that time a number of new vent sites have from previously known fauna are so great that new
been discovered around the world – and most have taxonomic families have had to be established in
remarkably high concentrations of animals around order to classify them satisfactorily32. Some of the
them17,32. The reason why these areas attract such most exciting examples of hydrothermal vent
abundant life is even more remarkable. A common species discovered include the spectacular tube-
observation at all the vent fields that have been worms found along the East Pacific Rise vent sites,

157
 C.R. German, L.M. Parson, and R.A. Mills

10.9 Figure 10.9 Dense populations of tube

worms (Riftia pachyptila) and brachyu-
ran crabs inhabiting the Genesis
hydrothermal vent site, 13°N, East
Pacific Rise (depth, 2636 m). The tube-
worms, which can grow up to 2–3 m
in length and may be 2–5 cm across,
derive their nutrition from symbiotic
sulphide-oxidising bacteria which live
within their gut. The crabs, which may
only grow to a modest 5–10 cm across,
feed by scavenging on the red haemo-
globin-filled ‘plumes’ of the tube-
worms; these can extend up to 10 cm
beyond the end of their protective
chitin tubes to draw in the hydrogen
sulphide rich waters necessary for
chemosynthesis, but can be withdrawn
rapidly back into their tubes for protec-
tion (photograph by R.A. Lutz, Institute
of Marine and Coastal Studies, Rutgers
University, and the American
Geophysical Union17).

which can measure 2–3 m or more long, and which deep-sea floor (Chapter 13). In contrast, biodiversi-
typically appear in thick clusters as shown in Figure ty at individual vent sites (i.e., the total range of
10.9. Also common along the East Pacific Rise are different species present; see Chapter 15) is surpris-
giant clams and mussels (Figures 10.10 and 10.11), ingly low. Not only that, but the species present at
which can often reach the size of a large dinner vent sites in the different oceans show remarkably
plate. The total biomass at any one hydrothermal little similarity. For example, no giant tube-worms
site is typically very high. Indeed, hydrothermal or giant clams have been found at any of the five
vent fields have been likened to submarine oases known hydrothermal fields discovered so far in the
which punctuate the deserted barren plains of the North Atlantic Ocean. Instead, for example, the

10.10 10.11

Figure 10.10 Giant white clams (Calyptogena magnifica) Figure 10.11 Dense beds of mussels (Bathymodiolus
living in clusters within the crevices between basalt pil- thermophiolus) and clams (Calyptogena magnifica) at the
lows, in an area known as Clam Acres at 21°N on the Rose Garden vent site along the Galapagos Rift. Again,
East Pacific Rise. Other vent fauna include a clump of individual clams and mussels are typically 20–30 cm in
tube worms (Riftia pachyptila), galatheid crabs length. Other vent fauna seen include galatheid crabs
(Munidopsis susquamosa), and limpets. Note the robot and snails (photograph by R.A. Lutz, Institute of Marine
arm of the submersible ‘Alvin’ collecting a clam approxi- and Coastal Studies, Rutgers University, and the
mately 25 cm in length (photograph by R.A. Lutz, American Geophysical Union17).
Institute of Marine and Coastal Studies, Rutgers
University, and the American Geophysical Union17).

158
 10: Mid-Ocean Ridges and Hydrothermal Activity

Figure 10.12 A schematic cross-section of the 10.12

TAG hydrothermal mound, 26°N Mid-Atlantic
Ridge (the co-ordinates give the dive site). Hot
vent-fluid (shown in pale blue) flows up an open
fissure in the deeply faulted oceanic crust, and
then percolates out through the entire sulphide
mound along a tortuous network of interconnect-
ed channels, giving rise to the highest-tempera-
ture (350–365°C) black smoker fluids at the apex
(50 m across and 50 m above the sea floor) and
to the lower temperature (270–300°C), partly
diluted white smokers around the outer section of
the mound (200 m across and 20–30 m above the
sea floor). Extinct chimneys are also seen across
much of the outer mound, where earlier fluid
flow has ceased because subsurface mineral pre-
cipitation has choked the flow-channels solid.
Toward the flanks of the mounds, rubble deposits
occur where oxidised and altered material from
Dive site
the hydrothermal mound has been broken up by
mass-wasting (‘landslide’) events and carried out
across the sea floor, to be deposited upon more
typical volcanic basement and a thin veneer of
more typical pelagic sediments (courtesy of Pierre Minon, © National Geographic Society).

southernmost three of these sites are characterised Therefore, only if these species give rise to plank-
by abundant small shrimp, which cluster in their tonic larval stages is there any prospect of them
millions around the black-smoker chimneys. being able to migrate along mid-ocean ridge axes
Such completely isolated biological communities and colonise new vent fields as and when they
indicate separate paths of evolution over many gen- occur32.
erations – perhaps stretching back millions of
years. However, we know that individual vent Hydrothermal Deposits
fields and chimneys may only remain active for Hydrothermal deposits on the sea floor range from
periods of perhaps 100–1000 years at a time, single chimneys to large, sprawling mounds topped
before the chimneys themselves become choked by clusters of chimneys20. In general, the size of the
with minerals and the flow of warm fluids becomes hydrothermal deposits appears to be related to the
blocked. For any one species of hydrothermal length of time for which active venting persisted at
organism to have survived down through the gener- those sites. Pacific vent fields, with very localised
ations, therefore, we know that the ability to chimney deposits, have life-spans of just tens of
migrate from one vent site to another must be vital- years to perhaps a hundred years or so before min-
ly important. Thus, an important question – partic- eral precipitation cements their chimneys solid11. By
ularly given the globe-encircling nature of mid- contrast, certain Atlantic sites are characterised by
ocean ridges – is “How do animals move from one mounds stretching in excess of 200 m across and
hydrothermal field to colonise another on an intra- reaching 30–50 m high, and which dating reveals
basinal scale, yet live within communities which, have remained active for thousands and even tens
on an inter-basinal scale, are evidently quite isolat- of thousands of years13.
ed?” One key area of research currently underway One such site that has been studied extensively
involves mapping the entire mid-ocean ridge system by the international community is called TAG
in sufficient detail to determine the total number of (named after a North Atlantic basin study in the
hydrothermal vent sites world-wide, their average 1970s – the Trans-Atlantic Geotraverse) 22. The
spacing one from another, and how that spacing is TAG site comprises one large active mound and
controlled by the tectonic and volcanic nature of several mounds that are now extinct. The deposit is
the mid-ocean ridges which host them. The second one of the largest active sites known and is approx-
key issue to explain the biodiversity and separate imately 200 m in diameter and 50 m high. There is
evolution problems is to understand how animals a vast range in the style of fluid venting and type of
reproduce and migrate along the lengths of these mineral deposit at different sites on the mound.
ridges. It is proposed that the processes of repro- Fluid flow through the mound is pervasive, fed by a
duction and migration must be intimately related to complex network of channels through which high-
each other, because a large majority of vent-specific temperature fluid flows, rising through the mound
organisms exhibit quite sessile lifestyles as adults. from the underlying basement (Figure 10.12). Hot

159
C.R. German, L.M. Parson, and R.A. Mills

10.13

Figure 10.13 A cluster of inactive chimneys from the outer portion of the TAG hydrothermal
mound. The chimneys each measure several metres in height and 10–30 cm in width. They have
become clogged with minerals, as the supply of hydrothermal fluid from the underlying network
of channels has waned – presumably due to further mineral precipitation subsurface, within the
mound. Oxidation of the iron-, copper-, and zinc-sulphide minerals gives rise to the alteration of
the chimneys’ colour from blue–grey to brown. In time, these chimneys will become sufficiently
altered and weakened to collapse, adding to the unconsolidated sulphide rubble which makes
up much of the TAG hydrothermal mound. Evidence that this particular site has not long been
inactive is also provided, however, by the presence of the luminous pale-blue mineral seen plat-
ed to the exterior of many of the chimneys. This mineral is anhydrite (a form of calcium sul-
phate), which only precipitates in the oceans at high temperatures and will continue to dissolve
as the chimneys age and cool (courtesy of Woods Hole Oceanographic Institution).

(360°C) acidic fluids gush from the summit at Three different species of shrimp have been discov-
speeds of several metres per second; the whole ered at TAG; the most abundant is Rimicaris exoc-
mound apex is shrouded in black smoke which is ulata, which swarms over the black-smoker edifice.
rapidly entrained upward into a buoyant The lower-temperature diffuse flow areas host a
hydrothermal plume. Those chimneys that have community of anemones and crabs. The distribu-
been sampled are 1–10 cm in diameter and consist tion of organisms can give semi-quantitative infor-
predominantly of copper- and iron-sulphides (chal- mation as to the fluid flow regime32.
copyrite, marcasite, bornite, pyrite) and calcium Hydrothermal activity is intermittent, even at
sulphate (anhydrite)31. Lower temperature, ‘white individual sites. Inactive chimneys have been dis-
smokers’ vent at rates of centimetres per second to covered on the outer portions of the main mound
the southeast of the main black smoker complex. and are shown in Figure 10.13. The chimneys have
Here the temperatures are up to 300°C, and the been oxidised to orange iron oxides and are begin-
chimneys are bulbous and zinc sulphide (sphalerite) ning to crumble and collapse. Eventually, they will
rich. The white ‘smoke’ consists of amorphous sili- be completely weathered down to the mound sur-
ca mixed with zinc- and iron-sulphides. Much of face. The mound itself is unstable and subject to
the rest of the mound surface is covered with red collapse and mass wasting events, in which subma-
and orange iron oxides, through which diffuse, rine equivalents of landslides, perhaps triggered by
low-temperature fluids percolate. Areas of diffuse minor earthquakes, sweep altered chimney material
flow are delineated by clusters of white anemones down off the slopes of the mound and out onto the
and shimmering water20. surrounding sea-floor sediments. As a result, the
The range in fluid-venting styles is evident from sediments immediately adjacent to the flanks of the
the distribution of fauna over the mound surface. mound are also full of metal-rich sulphide and

160
 10: Mid-Ocean Ridges and Hydrothermal Activity

oxide minerals, just like the hydrothermal chimneys 10.14

from which they are derived7,19. Warm fluids can
continue to percolate up through these flanking
sediments, over time altering their mineral assem-
blage to clays. Despite this mineral alteration, the
metal concentrations preserved within these sedi-
ments remain extraordinarily high – up to 45%
iron and 34% copper, and as much as 10–15%
zinc. The potential for mining such deposits in the
future is discussed in Chapter 21.
Eventually, hydrothermal sediments are buried
beneath the normal background pelagic sediments
and transported away from the ridge axis by plate-
spreading processes (see earlier). Deep-sea sedi-
ments have been drilled by the Ocean Drilling
Program at over 700 sites. On those occasions
where drilling has reached the basement rock, a
metal-rich layer has often been observed at the bot-
tom of the sediment pile, representing ancient met-
alliferous sediments 14. Such basal metal enrich-
ments are often the only evidence that hydrother-
mal activity has been extensive, not only spatially,
but also throughout geological time.

Hydrothermal Activity, Ocean

Circulation, and Ocean Composition
The effects of hydrothermal activity are not
restricted to the immediate vicinity of black-smoker
vent sites – although this is where their most visual-
ly spectacular impact is best observed. As
hydrothermal fluids erupt from the sea floor, they
remain buoyant as they mix with sea water, and
rise, carrying their mix of particles and fluid
upward in a conical expanding plume (Figure
10.14). As this turbulent, continuously mixing
plume rises it eventually reaches a stage where it is
no longer more buoyant than the surrounding Figure 10.14 A three-dimensional acoustic image of the
water column, and so ceases to rise28. This particle- lower 40 m of two buoyant hydrothermal plumes dis-
rich fluid is then dispersed as an approximately charging from adjacent black smoker vents at 2635 m
depth on the East Pacific Rise, near 21°N. This image,
horizontal layer (often referred to as the neutrally obtained using a sonar (echo-sounder) system mounted
buoyant plume), flowing along isopycnal (constant on the sail of the US submersible ‘Alvin’, shows coales-
density) surfaces. Dispersion of this material cence of the two plumes as they rise, as well as a ‘bend-
through the oceans is then driven, primarily, by ing-over’ of the uppermost portion of their merged plume
large-scale ocean circulation patterns16. in the prevailing ocean-current direction. (courtesy of
It takes the fluid, solutes, and particles in a P.A. Rona, Institute of Marine and Coastal Studies,
buoyant hydrothermal plume less than an hour to Rutgers University, and the American Geophysical
rise 100–300 m above the sea bed, before being Union23).
carried away by the prevailing deep-ocean currents.
During this time, mixing is so turbulent that the
initial vent fluid is typically diluted approximately venting has greatly increased the ability of geo-
10,000 fold by ordinary sea water24. Because the chemists to locate new sites of hydrothermal activi-
initial vent fluid is enriched by factors of up to a ty8,9,12,21. Instead of needing to photograph every
million in certain key chemical tracers (e.g., dis- square metre of sea bed to look for individual vent
solved methane, dissolved manganese, and total fields, it has become sufficient simply to sample the
suspended particulate matter), strong enrichments overlying water column, perhaps just once every
can still be detected after emplacement into and few kilometres, to detect the presence of any chemi-
dispersion within neutrally buoyant hydrothermal cal or physical anomalies characteristic of
plumes. Recognition of this characteristic of active hydrothermal discharge in any particular area.

161
 C.R. German, L.M. Parson, and R.A. Mills

1800
10.15

1850

1900

1950
Depth (m)

2000

2050

2100

2150

2200
0 1000 2000 3000 4000 5000

Distance (m)

Figure 10.15 A two-dimensional cross-section of suspended particle enrichment along a 5 km segment of the
‘Rainbow’ hydrothermal plume at 36°15’N, Mid-Atlantic Ridge. Particle enrichment is caused by the precipitation of
thousands of tiny (1–10 μm across) metal oxide and sulphide particles from black smoker chimneys, which are then
swept up into the water column, up to 300 m above the sea bed, before being dispersed by prevailing ocean currents.
Because the deep ocean is typically very clear and contains very little suspended particulate matter, optical devices to
measure sea-water transparency (in this case, a nephelometer) can readily detect the presence of hydrothermal plumes
and, hence, deduce the location of active hydrothermal vent fields. In this case, suspended particle concentrations
increase approximately 10-fold from the background (dark blue; )35 μg of suspended particles per litre of sea water) to
maximum plume concentrations directly above the vent site (shown in pink and purple) of 300–350 μg/l.

Figure 10.15 shows an example of a section of buoyant hydrothermal plumes are dispersed pas-
ridge crest some 5 km long that was surveyed in sively by the prevailing deep-ocean currents, recent
this way in September 1994. A nephelometer, an work suggests that this may not be the case. New
optical device which measures relative particle con- research by physical oceanographers indicates that
centrations in sea water, was towed through the a process known as ‘vortex shedding’ may occur, in
water column, and detected and mapped the pres- which the forces of the Earth’s rotation combine
ence of a particle-rich plume of water approximate- with the upward motion of a buoyant hydrother-
ly 300 m above the sea bed over the Mid-Atlantic mal plume to break off hydrothermal eddies (essen-
Ridge near 36°N – the ‘Rainbow’ hydrothermal tially, spiralling rings of neutrally buoyant plume
field9. material); these then migrate through the water col-
A new study of hydrothermal vents at umn, away from their source at an active vent
Steinahóll, close to Iceland, has thrown up a new site29. This idea, however, has only recently been
technique for studying hydrothermal activity. At predicted, following laboratory-based tank experi-
this shallow site, the confining pressures (from only ments and theoretical calculations. No such fea-
300 m of water) are insufficient to prevent dis- tures have been positively identified in the real
solved gases – including carbon dioxide, methane, oceans, yet, but the possibility that they might exist
and hydrogen – from bubbling out of solution. raises some very important questions. If hydrother-
These bubbles rise all the way up the water column mal eddies are formed regularly above vent sites,
and can be imaged acoustically by an echo-sounder could they represent ‘incubating’ parcels of chemi-
as they travel from the sea bed (Figure 10.16). In cally enriched sea water, in which young larval
extremely calm weather, they can also be seen by stages of vent fauna might survive for hundreds of
the naked eye, ‘popping’ at the sea-surface, but in days as they are carried tens of kilometres up and
more typical, rougher, weather the sea surface is down the world’s mid-ocean ridge axes? Such a
too disturbed and only sonar can reveal the mechanism might go a long way toward resolving
hydrothermal activity below. the key problem of how vent fauna colonise newly
Although it has long been believed that neutrally formed hydrothermal vent sites. In the extreme

162
 10: Mid-Ocean Ridges and Hydrothermal Activity

10.16

Sea bed

Depth
(m)

Time (minutes)

Figure 10.16 A high-frequency (38 kHz) echo-sounder trace of gas-rich venting along a 1000 m section of the
Reykjanes Ridge through the Steinahóll vent field at 63°06’N, immediately southwest of the Reykjanes Peninsula,
Iceland. Echoes from the solid sea-bed produce a strong (bright red) acoustic reflection, which is overlain by a more
diffuse (blue) absorption–reflection pattern caused by plumes of bubbles rising from the sea bed. (At a ship’s speed of
1.0–1.5 knots, 5 minutes of survey time, as indicated along the x-axis, is approximately equivalent to 200 m survey
coverage over the sea bed.)

case, on a fast-spreading section of mid-ocean ridge General References

crest, where abundant hydrothermal activity occurs Humphris, S.E, Zierenberg, R.A., Mullineaux, L.S., and
at very close spacing, could the cumulative heat Thomson, R.E. (eds) (1995), Physical, Chemical,
flux and consequent numerous hydrothermal Biological and Geological Interactions within
eddies be of sufficient magnitude (acting in concert) Hydrothermal Systems, Geophysical Monograph,
American Geophysical Union, Washington, DC.
to exert some control on the general physical circu- Parson, L.M., Walker, C.L., and Dixon, D. (eds) (1995),
lation of the deep-water column over the length- Hydrothermal Vents and Processes, Special
scale of entire ocean basins? Publication, The Geological Society, London, 396 pp.
Perhaps unsurprisingly for a field that is scarcely Sinton, J.M. (ed.) (1989), Evolution of Mid-Ocean Ridges,
Geophysical Monograph 57, American Geophysical
more than 15 years old, the questions currently Union, Washington, DC, 77 pp.
arising from hydrothermal study greatly outweigh
the answers that have been discovered, so far. References
Indeed, current lines of research seem to suggest 1. Bowers, T.S., Von Damm, K.L., and Edmond, J.M. (1985),
that the full significance of hydrothermal activity Chemical evolution of mid-ocean ridge hot springs,
and its impact upon numerous other aspects of Geochim. Cosmochim. Acta, 49, 2239–2252.
oceanography have yet to be fully appreciated. It 2. Childress, J.J. and Fisher, C.R. (1992), The biology of
seems certain that hydrothermal circulation will hydrothermal vent animals: physiology, biochemistry
and autotrophic symbioses, Oceanogr. Marine Biol.
continue to be a focus of research for many years Annu. Rev., 30, 337–441.
to come. 3. Corliss, J.B., Dymond, J., Gordon, L.I., Edmond, J.M., von
Herzen, R.P., Ballard, R.D., Green, K., Williams, D.,
Bainbridge, A., Crane, K. and van Andel, Tj.H.
(1979), Submarine thermal springs on the Galapagos
Rift, Science, 203, 1073–1083.
4. Edmond, J.M., von Damm, K.L., McDuff, R.E., and
Measures, C.I. (1982), Chemistry of hot springs on
the East Pacific Rise and their effluent dispersal,
Nature, 297, 187–191.

163
C.R. German, L.M. Parson, and R.A. Mills

5. Edmond, J.M., Campbell, A.C., Palmer, M.R., 19. Mills, R.A., Thomson, J., and Elderfield, H. (1993b), A dual
Klinkhammer, G.P., German, C.R., Edmonds, H.N., origin for the hydrothermal component in a metallif-
Elderfield, H., Thompson, G., and Rona, P. (1995), erous sediment core from the Mid-Atlantic Ridge, J.
Time series studies of vent fluids from the TAG and Geophys. Res., 98, 9671–9681.
MARK sites (1986, 1990) Mid-Atlantic Ridge: a new 20. Mills, R.A. (1995), Hydrothermal deposits and metalliferous
solution chemistry model and a mechanism for Cu/Zn sediments from TAG, 26˚N Mid-Atlantic Ridge, in
zonation in massive sulphide orebodies, in Hydrothermal Vents and Processes, Parson, L.M.,
Hydrothermal Vents and Processes, Parson, L.M., Walker, C.L., and Dixon, D. (eds), Special
Walker, C.L., and Dixon, D. (eds), Special Publication, The Geological Society, London, pp.
Publication, The Geological Society, London, pp 121–132.
77–86. 21. Murton, B.J., Klinkhammer, G., Becker, K., Briais, A., Edge,
6. Feely, R.A., Lewison, M., Massoth, G.J., Robert-Baldo, G., D., Hayward, N., Millard, N., Mitchell, I., Rouse, I.,
Lavelle, J.W., Byrne, R.H., von Damm, K.L., and Rudnicki, M., Sayanagi, K., Sloan, H., and Parson, L.
Curl, H.C. J. (1987), Composition and dissolution of (1994), Direct evidence for the distribution and
black smoker particulates from active vents on the occurrence of hydrothermal activity between
Juan de Fuca Ridge, J. Geophys. Res., 92, 27˚N–30˚N on the Mid-Atlantic Ridge, Earth Planet.
11347–11363. Sci. Lett., 125, 119–128.
7. German, C.R., Higgs, N.C., Thomson, J., Mills, R.A., 22. Rona, P.A., Klinkhammer, G., Nelsen, T.A., Trefry, J.H.,
Elderfield, H., Blustajn, J., Fleer, A.P., and Bacon, and Elderfield, H. (1986), Black smokers, massive sul-
M.P. (1993), A geochemical study of metalliferous phides and vent biota at the Mid-Atlantic Ridge,
sediment from the TAG hydrothermal mound, Nature, 321, 33–37.
26˚08’N, Mar, J. Geophys. Res., 98, 9683–9692. 23. Rona, P.A., Palmer, D.R., Jones, C., Chayes, D.A.,
8. German, C.R., Briem, J., Chin, C., Danielsen, M., Holland, Czarnecki, M., Carey, E.W., and Geurrero, J.C.
S., James, R., Jónsodottir, A., Ludford, E., Moser, C., (1991), Acoustic imaging of hydrothermal plumes,
Ólafsson, J., Palmer, M.R., and Rudnicki, M.D. East Pacific Rise, 21˚N, 109˚W, Geophys Res. Lett.,
(1994a), Hydrothermal activity on the Reykjanes 18(12), 2233–2236.
Ridge: the Steinahóll vent-field at 63˚06’N, Earth 24. Rudnicki, M.D. and Elderfield, H. (1993), A chemical
Planet. Sci. Lett., 121, 647–654. model of the buoyant and neutrally buoyant plume
9. German, C.R., Parson, L.M., and Scientific Party of RRS above the TAG vent field, 26˚N, Mid-Atlantic Ridge,
Charles Darwin cruise CD89’HEAT’ (1994b), Geochim. Cosmochim. Acta, 57, 2939–2957.
Hydrothermal exploration at the Azores triple-junc- 25. Schultz, A., Delaney, J.R., and McDuff, R.E. (1992), On the
tion, EOS, Trans. Am Geophys. Union., 75, 308. partitioning of heat flux between diffuse and point-
10. Haxby, W.F. (1987), Gravity Field of the World’s Oceans: source seafloor venting, J. Geophys. Res., 97,
A Portrayal of Gridded Geophysical Data Derived 12229–12314.
from SEASAT Radar Altimeter Measurements of the 26. Searle, R.C. (1992), The volcano-tectonic setting of oceanic
Shape of the Ocean Surface (Scale 1:40,000 at the lithosphere generation, in Ophiolites and their mod-
Equator), World Data Centre for Marine Geology ern oceanic analogues, Parson, L.M., Murton, B.J.,
and Geophysics. and Browning, P. (eds), Special Publication 60, The
11. Haymon, R.M. and Kastner, M. (1981), Hot spring deposits Geological Society, London, 65–80.
on the East Pacific Rise at 21˚N: preliminary descrip- 27. Smith, D.K. and Cann, J.R. (1993), Building the crust at the
tion of mineralogy and genesis, Earth Planet. Sci. Mid-Atlantic Ridge, Nature, 365, 707–715.
Lett., 53, 363–381. 28. Speer, K.G. and Rona, P.A. (1989), A model of an Atlantic
12. Klinkhammer, G., Rona, P., Greaves, M.J., and Elderfield, and Pacific hydrothermal plume, J. Geophys. Res.,
H. (1985), Hydrothermal manganese plumes over the 94, 6213–6220.
Mid-Atlantic Ridge rift valley, Nature, 314, 727–731. 29. Speer, K.G. and Helfrich, K.R. (1995), Hydrothermal
13. Lalou, C., Thompson, G., Arnold, M., Brichet, E., Druffel, plumes: a review of flow and fluxes, in Hydrothermal
E., and Rona, P.A. (1990), Geochronology of TAG Vents and Processes, Parson, L.M., Walker, C.L., and
and Snakepit hydrothermal fields, Mid-Atlantic Dixon, D. (eds), Special Publication, The Geological
Ridge: witness to a long and complex hydrothermal Society, London, pp 373–385.
history, Earth Planet. Sci. Lett., 97, 113–128. 30. Spiess, F.N., MacDonald, K.C., Atwater, T., Ballard, R.,
14. Leinen, M. (1981), Metal-rich basal sediments from north- Carranza, A., Cordoba, D., Cox, C., Diaz Garcia,
eastern Pacific Deep Sea Drilling Project sites, in V.M., Francheteau, J., Guerrero, J., Hawkins, J.,
Initial Reports of the Deep Sea Drilling Programme Haymon, R., Hessler, R., Juteau, T., Kastner, M.,
63, Yeats, R.S. and Haq, B.U. (eds), US Govt. Larson, R., Luyendyke, B., Macdougall, J.D., Miller,
Printing Office, Washington, pp 667–676. S., Normark, W., Orcutt, J., and Rangin, C. (1980),
15. Le Pichon, X. (1968), Sea floor spreading and continental East Pacific Rise; hotsprings and geophysical experi-
drift, J. Geohpys. Res., 73, 3661–3697. ments, Science, 207, 1421–1433.
16. Lupton, J.E. and Craig, H. (1981), A major 3He source at 31. Thomson, G., Humphris, S.E., Schroeder, B., Sulanowska,
15˚S on the East Pacific Rise, Science, 214, 13–18. M., and Rona, P. (1988), Active vents and massive
17. Lutz, R.A. and Kennish, M.J.(1993), Ecology of deep-sea sulfides at 26˚N (TAG) and 23˚N (Snakepit) on the
hydrothermal vent communities: a review, Rev. Mid-Atlantic Ridge, Canad. Mineralog., 26, 697–711.
Geophys., 31(3), 211–242. 32. Tunnicliffe, V. (1991), The biology of hydrothermal vents:
18. Mills, R.A., Thomson, J., Elderfield, H., and Rona, P.A. ecology and evolution, Oceanogr. Marine Biol. Ann.
(1993a), Pore-water geochemistry of metalliferous Rev. 29, 319–407.
sediments from the Mid-Atlantic Ridge: diagenesis 33. Wolery, T.J. and Sleep, N.H. (1976), Hydrothermal circula-
and low-temperature fluxes, EOS, Trans. Am. tion and geochemical flux at mid-ocean ridges, J.
Geophys. Union, 74, 101. Geophys. Res., 84, 249–276.

164
CHAPTER 11:

The Ocean: A Global

Geochemical System
J.D. Burton

Introduction
When the first systematic measurements of the pheric testing of nuclear weapons in the late 1950s
chemical composition of sea water were made as and early 1960s, and some of which are due to dis-
part of the work of the Challenger Expedition of charges of low-level waste from nuclear fuel repro-
1872–1876, the constituents which could be cessing. Examples of the kind of information
analysed with any accuracy were very limited in obtained by the use of such tracers are given in
number. They comprised some of the more abun- Figure 11.2. Constituents such as nitrate, phos-
dant dissolved salt components (Figure 11.1) and phate, and dissolved silicon have been studied very
certain dissolved gases. Subsequent advances in intensively because they are important plant nutri-
analytical chemical techniques, allied with ents, with an influence on the photosynthetic pro-
improved methods of sampling, gradually enabled duction of organic matter. Others, including trace
a fuller knowledge of the concentrations of a wider metals (such as cadmium, mercury, and lead) and
range of constituents to be acquired. These organic micropollutants (such as pesticides and
advances accelerated remarkably from about the polynuclear aromatic hydrocarbons) are of concern
mid 1970s, so that there is now information not because of their potential impact on marine life in
only on the concentrations, but also on the patterns estuaries and coastal waters, which now receive
of oceanic distribution for the majority of the ele- increased inputs of these substances as a result of
ments. Information on the concentrations of some agricultural uses and the disposal of industrial and
important dissolved elements in sea water is given domestic wastes.
in Table 11.1. In a solution with so complicated a composition
Marine scientists need information on chemical as sea water, many elements are present in a variety
constituents for various purposes. Some con- of physicochemical forms. For example, in the pres-
stituents are useful as tracers, providing informa- ence of the various inorganic, negatively charged
tion on the circulation and mixing of water bodies ions (anions) in sea water, the dissolved element
in the ocean. They include the chlorofluorocarbons copper occurs to only a small extent as the positive-
(freons), added to the atmosphere and thence to the ly charged divalent ion, the cupric Cu2+ cation, and
ocean by man’s use of them from about 1950 is largely associated (complexed) with the anions.
onward, and anthropogenic radioactive materials, The distribution of copper between the simple
some of which entered the ocean from the atmos- cation and the complexes formed with various

Figure 11.1 Relative concentrations of dissolved major

constituents in sea water. The bulk content of dissolved
material is expressed by salinity (S), a defined dimension- Mg2+ 11.1
less quantity which approximates to the total concentra- Ca2+
tion (in g/kg) of the substances present in solution. K+ Br –
Na+ F–
Salinity can be measured at best to 0.001; only the few
B+
major constituents in the diagram contribute significantly
– HCO3– Sr 2+
to it. The actual mean concentrations of the constituents SO 24
are given in Table 11.1. The mean concentration of dis-
solved silicon exceeds 1 mg/kg (approximately 0.001 in
salinity), but its distribution is much more variable than
those of the elements shown, and it is in a chemical form Cl –
largely undetected by the widely employed conductimet-
ric determination of salinity. It is therefore grouped with
the minor or trace constituents, many of which occur at
concentrations between 1 μg (10–6 g) and 1 ng (10–9 g)/kg,
and others at still lower concentrations.

165
J.D. Burton

Table 11.1 Concentrations and most important chemical species of some dissolved elements in sea water.

Element Average concentration Range Units (per kg) Most abundant chemical speciesa

Lithium 174 b μg Li+

Boron 4.5 b mg H3BO3
Carbon 27.6 24–30 mg HCO3–, CO32–
Nitrogenc 420 <1–630 μg NO3–
Fluorine 1.3 b mg F–, MgF+
Sodium 10.77 b g Na+
Magnesium 1.29 b g Mg2+
Aluminium 540 <10–1200 ng Al(OH)4–, Al(OH)30
Silicon 2.8 <0.02–5 mg H4SiO40
Phosphorus 70 <0.1–110 μg HPO42–, NaHPO4–, MgHPO40
Sulphur 0.904 b g SO42–, NaSO4–, MgSO40
Chlorine 19.354 b g Cl–
Potassium 0.399 b g K+
Calcium 0.412 b g Ca2+
Manganese 14 5–200 ng Mn2+, MnCl+
Iron 55 5–140 ng Fe(OH)30
Nickel 0.50 0.10–0.70 μg Ni+, NiCO30, NiCl+
Copper 0.25 0.03–0.40 μg CuCO30, CuOH+, Cu2+
Zinc 0.40 <0.01–0.60 μg Zn2+, ZnOH+, ZnCO30, ZnCl+
Arsenic 1.7 1.1–1.9 μg HAsO42–
Bromine 67 b mg Br–
Rubidium 120 b μg Rb+
Strontium 7.9 b mg Sr2+
Cadmium 80 0.1–120 ng CdCl20
Iodine 50 25–65 μg IO3–
Caesium 0.29 b μg Cs+
Barium 14 4–20 μg Ba2+
Mercury 1 0.4–2 ng HgCl42–
Lead 2 1–35d ng PbCO30, Pb(CO3)22–, PbCl+
Uranium 3.3 b μg [UO2(CO3)3]4–

a Refers to the inorganic speciation in oxygenated waters.

b Variations are determined entirely or largely by those in salinity, i.e., the element is essentially conservative (see text for definition). For these
elements, average concentration given is for sea water of salinity 35.
c Concentrations refer to combined nitrogen; element occurs also as dissolved nitrogen (N ) gas. Species other than NO – are often important in
2 3
the upper ocean (e.g., NO2–, NH4+).
d Concentrations are affected by inputs to surface ocean of atmospherically transported lead from combustion of leaded petroleum.

Note:
Based mainly upon information in Bruland4. Usual ranges for oceanic waters are shown; concentrations of certain elements can be higher in
some coastal waters.
Chemical oceanographers often employ molar units instead of the mass units shown here. For sodium (atomic weight 22.99) the concentration
given above of 10.77 g/kg can alternatively be expressed as 0.449 mol/kg.

anions is shown in Figure 11.3. Copper is one of a by compounds which are described as ‘marine
number of metals which are also strongly com- humic material’, because they show some analogies
plexed by the organic matter present in sea water, to the humic and fulvic acids produced in soils by
and so the speciation shown in Figure 11.3 actually the decomposition of terrestrial vegetation. These
applies to only a small fraction of the dissolved marine humic substances have not been fully char-
copper. The dissolved organic material in sea water acterised as regards their molecular structures, but
comprises a great diversity of molecules dominated show a range of capacities to form complexes with

166
 11: The Ocean: A Global Geochemical System

Figure 11.2 (a) Distribution of tritium (3H) in a longitudi-

nal section in the western North Atlantic Ocean in the 7 11.2a
1970s (Broecker and Peng3; with permission from the 6 5
authors). Values are given in tritium units (TU; 1 TU = 1 x 4
10–18 atoms of tritium per atom of hydrogen). Surface 3
1.0
concentrations are greatly enhanced above natural back- 2
ground values by the input of tritium from the atmos- 1
phere, due to thermonuclear weapon testing, particularly
2.0
in the late 1950s and early 1960s. The penetration of tri-
tium into deeper waters at high latitudes reflects the for-

(km)
mation of North Atlantic Deep Water. Because tritium is

Depth (km)
0.2
an isotope of hydrogen and enters the ocean largely as 3.0

Depth
tritium-labelled water molecules, it is an ideal tracer for
the advection and diffusion of water from the surface to
the deeper ocean. (b) Distribution of a chlorinated fluoro-
carbon (freon), CFC 11, in a section across the South 4.0
Atlantic Ocean in the 1990s. The section runs along 45°S
from the western margin to the mid-Atlantic Ridge and
then in a northeastern direction to the eastern margin at 5.0
30°S. The higher concentrations in surface waters reflect
exchange with the atmosphere and transfer downward by
advection and diffusion. The increase toward the sea bed
reflects the inputs to the deepest ocean by transport of 0 10 20 30 40 50 60 70 N
Antarctic Bottom Water of more recent surface origin Latitude
Latitude (˚)(°)
than the overlying North Atlantic Deep Water. This pene-
trates the Argentine Basin, west of the mid-Atlantic Ridge, cfc-11 pmol/l
pmol/l
more effectively than the Cape Basin to the east (unpub- 00
11.2b
4.0000
4.000
lished data provided by D. Smythe-Wright and S. 600
dbar

3.0000
3.000
Boswell). 1
1200 2.0000
2.000
1.0000
1.000
1800
0.4500
0.450
press

2
0.3500
Depth (km)

2400 0.350
0.2500
0.250
3000
3 0.1800
0.180

3600 0.1400
0.140
0.1000
0.100
4
4200 0.0600
0.060

4800 0.0400
0.040
5 0.0250
0.025
5400 0.0150
0.015
0.0050
0.005
6
6000
-60 -45 -30 -15 0 15
–60
lon –45
degrees –30 –15 0 15
W Longitude (˚) E

Figure 11.3 The distribution of copper between the free 11.3

cation and the complexes formed with various inorganic
anions in sea water (chloride, Cl–; sulphate, SO42–; carbon-
ate, CO32–; and hydroxyl, OH–). For trace constituents,
information of this kind cannot be obtained directly by –
measurements on sea water, but is derived from models CO 23
which assume equilibrium between the different species
and use data on the stabilities of the various complexes.
The nature of the organic complexes that account for a
–
large part of the actual chemical speciation of copper in
sea water cannot be similarly specified (see text).
SO 24
Cl – OH –

Free Cu2+

167
 J.D. Burton

metals. In contrast with copper, the abundant ele- interference of one in the analysis of another must
ment sodium has a very simple chemistry; it occurs be avoided. For these reasons, analytical measure-
in sea water almost entirely as the Na+ ion. ments on sea water often involve quite complicated
The probable main inorganic chemical species chemical manipulations to concentrate and sepa-
for some important elements are shown in Table rate the required chemical forms. Very few chemi-
11.1. Understanding chemical speciation is impor- cal constituents can be measured by probing the
tant because it affects the reactions of the elements ocean with a sensor, in the way that salinity can be
in sea water. For example, the toxicity of copper to measured (see Chapter 19). Much of the picture of
organisms such as phytoplankton (see Chapter 6), their oceanic distributions is, therefore, coarsely
is related to the concentration (or, more strictly, the resolved, but the great progress made in the acqui-
equivalent thermodynamic quantity, the activity) of sition of reliable data over the past two decades
the cupric ion rather than that of total copper. In has, nevertheless, underpinned comparable
chemical analyses, however, the chemical specia- advances in understanding the chemical processes
tion is rarely resolved, so dissolved concentrations that determine these distributions.
generally refer here to the total element or ion.
Knowledge of the chemistry of the ocean relies The Geochemical Context
heavily on the analytical chemical measurements The oceanographic study of chemical properties
which reveal the spatial distributions of chemical and processes is undertaken, as already indicated,
constituents and their variations in time. Because from a variety of standpoints. The most central,
many constituents of interest are present in sea however, and the one which provides the most
water at very low concentrations, they can be valuable unifying insights, is that of geochemistry –
determined only by sensitive methods; special pre- the science which addresses questions concerning
cautions are necessary to avoid contamination dur- the occurrence and distribution of the chemical ele-
ing sampling and analysis. Given the complexity of ments within Earth as a whole and within its major
the mixture and the different levels of concentra- components, in space and time. These major com-
tion at which constituents occur, the potential ponents include the crust (lithosphere), the atmos-

11.4 Total H2O in atmosphere,

0.105 x 1020 g
Precipitation on land from
ocean, 0.40 x 1020 g/yr

Evaporation from
land, 0.71 x 1020 g/yr
Precipitation on Evaporation from
oceans, 3.85 x 1020 g/yr oceans, 4.25 x 1020 g/yr
Stream discharge,
Total precipitation on 0.36 x 1020 g/yr
land, 1.11 x 1020 g

Total H2O in oceans, Total H2O in ice,

13,500 x1020 g 165 x 1020 g
Total H2O in lakes,
0.34 x 1020 g
Total H2O in pores Ground water flow
of rocks, 3200 x 1020 g to ocean, 0.04 x 1020 g/yr

Figure 11.4 The global water cycle – the numbers show the total amounts of water held in the major reservoirs and
the fluxes of water between reservoirs. In effect, the cycle acts not only as a transport system for weathered continental
material, but also as a liquid extraction system for continental solid phases. Some of the water distilled by evaporation
from the ocean, containing only a small proportion of sea-water salts as aerosol (see text), is deposited as precipitation
on the continents and returned to the ocean containing dissolved material leached from continental rocks. Repeated
cycles create a continuous net transfer of dissolved constituents to the ocean (based on data from Baumgartner and
Reichel2).

168
 11: The Ocean: A Global Geochemical System

11.5
N
60˚ 30

▲
100
444

▲
66 133
17 31 131

▲
40˚ ▲

▲
256 1738
▲ 110
930
▲

20˚
210 1438 ▲
3000
▲

▲
232 67 286
0˚ 290 3228
▲
▲

28

▲
150 113
20˚
18 154
40˚ 17 62
▲

60˚
S

80˚W 60˚ 40˚ 20˚ 0˚ 20˚ 40˚ 60˚ 80˚ 100˚ 120˚ 140˚ 160˚ 180˚ 160˚ 140˚ 120˚E

Figure 11.5 Fluxes of river-borne particulate material discharged from the major drainage basins. The widths of the
arrows are proportional to the discharge; numerical values are in units of 106 tonne/yr. Fluxes to the Arctic Sea,
totalling 226 x 106 tonnes/yr, are not shown. The directions of the arrows are arbitrary. Continental erosion in South
East Asia and adjacent islands accounts for much of the discharge. The Amazon carries about 20% of the total water
discharge to the ocean, but is not proportionately important as a source of either particulate or dissolved material. In
some catchments, weathered particulate material is efficiently trapped in lakes or artificial reservoirs so that relatively
little reaches the coast, e.g., the St. Lawrence (after Milliman and Meade15; with permission from the University of
Chicago Press).

phere, and the aquatic environments collectively products, of which the clay minerals are particular-
described as the hydrosphere, of which the ocean ly important.
comprises the major mass, accounting for more The global water cycle (Figure 11.4; see also
than 98% of the free water at Earth’s surface. The Figure 2.2) thus drives a continuous flux of dis-
geochemical perspective emphasises the fact that solved and particulate material to the seas, globally
the ocean is a dynamic system, chemically as well amounting to estimated annual inputs of about
as physically, continuously exchanging material at 3.7 x 1015 g of dissolved material and about 15 x
its boundaries and also distributing it internally 1015 g of particles. Estimates of global fluxes are
among its major water masses. essential for understanding the ocean’s material
budget, but these values are averages of discharges,
The major sedimentary cycle which show large differences regionally, as illus-
The role of the ocean is central in the major sedi- trated for particulate material in Figure 11.5.
mentary cycle. Continental rocks undergo physical About 10% of the riverine flux of dissolved materi-
weathering, which produces particles (mineral dust) al is recycled from the ocean to the continents,
small enough to be lifted from the ground by wind mainly through the agency of bubbles of air: these
action and transported, particularly from arid bubbles are produced largely by breaking waves,
regions, through the atmosphere (aeolian trans- are then carried below the sea surface, and, on
port); particles are also carried by surface water returning to the surface, burst to create sea-salt
run-off. Water, containing carbon dioxide dis- aerosols, a fraction of which is transported through
solved from the atmosphere, is a powerful agent of the atmosphere and deposited on land.
chemical weathering, reacting with rock minerals Much of the particulate material discharged
to produce dissolved constituents and solid-phase from rivers settles out to form bottom sediments in

169
J.D. Burton

estuaries, deltas, and the deeper-water submarine ous process; direct evidence for this is provided by
fans formed by some major tropical rivers. It is inputs of 3He of mantle origin at mid-ocean ridge
estimated that, globally, under 10% of the input of spreading centres. It is reasonably certain, however,
particles from rivers reaches the open ocean. that there was substantial degassing in the early
Aeolian transport carries to the ocean an estimated stages of the formation of igneous rocks from
0.9 x 10 15 g of mineral dust annually. As this magma, the released gases subsequently reacting
mechanism is capable of distributing material over with the crystallised rock minerals at low
great distances before deposition on the ocean sur- temperatures.
face, mainly within the latitude belt of injection The major stages in the evolution of the ocean
into the atmosphere, it represents a source of parti- and the atmosphere were largely completed by the
cles for the open ocean comparable in magnitude beginning of the Phanerozoic Eon (about 6 x 108
with the rivers. Transport of the products of weath- years ago); subsequently, the mass of sedimentary
ering by water and wind is not the only process by material has changed little. During the past 2.5–3 x
which material enters the ocean, but it is certainly 10 9 years, the mass of sedimentary rock that
the best understood and is probably the dominant formed is about five times as much as that existing
input for most elements. It is therefore a logical now7, material having been recycled by reweather-
starting point for considering the way in which the ing of sedimentary rocks or destroyed by high-tem-
ocean fits into the larger framework of geological perature metamorphism, after subduction of sedi-
processes. ments into the mantle. The present ocean can be
The primary source for weathered material is regarded broadly as being close to a steady-state
igneous crustal rock, because the sedimentary system, with the production of sediments by weath-
rocks, such as limestones, sandstones, and shales, ering largely balanced by their high-temperature
are themselves products of the major sedimentary metamorphism, leading to a recycling of the
cycle. The present-day supply of weathered materi- volatiles consumed in weathering. The only one-
al to the ocean, however, contains a large contribu- way process is the continuing release of primary
tion, approximately 75% of the total, derived from volatiles from the mantle. Carbon dioxide is the
reweathering of sedimentary rocks formed from main volatile involved in present-day weathering.
sediments deposited earlier in geological time and While many features of oceanic composition can
emplaced on the continents tectonically (see be understood in terms of the interaction of man-
Chapter 8). Early geochemical calculations indicat- tle-derived volatiles and continental rocks, reac-
ed that the total amount of some constituents pre- tions between sea water and the basaltic oceanic
sent in the ocean, sediments, and sedimentary rocks crust must also be considered. These reactions
could not be accounted for by the weathering of occur under a wide range of temperatures and
the same quantity of igneous rock that is sufficient ratios of rock-to-water mass21. The most investigat-
to supply the amounts of the major rock-forming ed are the hydrothermal reactions involving sea
elements, such as sodium and calcium, found in water that has penetrated deep into newly formed
these reservoirs. These ‘excess’ constituents in the basalt at mid-ocean ridge spreading centres, and
sedimentary and oceanic domains share the proper- which is returned to the ocean by convective circu-
ty that they are all volatile compounds, or can be lation in the crust (Chapter 8). For some con-
derived from volatile compounds; they include the stituents, such as magnesium and sulphate, these
major anions present in sea water (chloride, sul- reactions lead to removal from sea water; for oth-
phate, bicarbonate, and bromide), water itself, and ers, such as calcium and important trace metals,
nitrogen, which is the dominant constituent of the including manganese, they create an additional
atmosphere. input to the ocean. The rate at which sea water cir-
culates through hydrothermal reaction zones is a
The chemical evolution of the ocean matter of much controversy. This is reflected in
On this basis the evolution of the ocean, atmos- uncertainties as to the global-scale impact of the
phere, sediments, and sedimentary rocks can be primary hydrothermal reactions and also the reac-
considered as a global-scale reaction, over geologi- tions which occur when hydrothermal fluids mix
cal time, between the minerals (mainly silicates and with waters of normal composition in the depths of
oxides) contributed by the igneous rocks and acidic the ocean.
volatiles, particularly hydrochloric acid, sulphur- A major aspect of the evolution of the ocean is
containing gases, and carbon dioxide. The primary the establishment of oxygen as a permanent con-
supply of the volatiles is by outgassing of the man- stituent of the atmosphere, since this enabled the
tle. There are several models for this process and its change to occur from a chemically reducing Earth-
time course, and for the rate of deposition of the surface environment to one which is oxidising.
sedimentary rocks. Mantle outgassing is a continu- Oxygen is produced by photosynthesis and by dis-

170
 11: The Ocean: A Global Geochemical System

sociation of water vapour in the upper atmosphere, extreme difference in perspective occurs when con-
with loss to space of some of the hydrogen released sidering those reservoirs of carbon (atmosphere,
in the process. While photosynthesis heavily domi- ocean, biota, humus), which constitute a very
nates the present-day production of oxygen, the rel- short-term exchange system in which carbon is
ative importance of the processes in earlier periods turned over on time-scales up to about 1000 years.
of geological time is uncertain. In the initial stages, In this system, the ocean is a major storage
the oxygen produced would have been consumed in reservoir.
reactions with the abundant reducing materials at Evidence, particularly from salt deposits formed
the Earth’s surface. The build-up of atmospheric by crystallisation from sea water under conditions
oxygen from the photodissociation of water vapour of high evaporation, suggests that over the last 108
requires that the production rate exceeds the con- years, and perhaps for most of the Phanerozoic
sumption rate in such reactions. For photosynthesis Eon, the average composition of sea water, with
to contribute to the build-up of atmospheric oxy- regard to major constituents, has not undergone a
gen, some of the organic matter must be buried and major evolutionary change. Some constituents may
not oxidised in respiration, otherwise all the oxy- have varied in concentration by less than a factor
gen produced would be consumed. The interven- of two, but constraints on the likely concentration
tion of man, acting as a geochemical agent on a of calcium do not preclude variations during the
global scale, has led to the combustion, in less than Phanerozoic Eon in a range 25–400% of the pre-
two centuries, of a significant amount of this sent value8. The present-day river input of dissolved
buried carbon. This has produced the fossil-fuel calcium maintained over the period would, howev-
effect of significantly rising atmospheric concentra- er, have introduced about 80 times the amount of
tions of carbon dioxide (e.g., see Figure 12.3). This calcium now present in the ocean. Such considera-
combustion has removed some oxygen from the tions support the assumption made by geochemists
atmosphere, but because the atmosphere contains that, in many respects, the ocean can be modelled
so much higher a concentration of oxygen (21%) as a steady-state system in which supply of material
compared with that of carbon dioxide (0.03%), the is balanced by its removal. As discussed later, the
change in concentration of oxygen is very difficult turnover time for the major constituents in the
to detect. ocean is long. The steady-state assumption, there-
At the present time the atmosphere can be fore, is not incompatible with evidence that supply
regarded as close to a steady-state, in which oxygen has exceeded removal, or vice versa, over shorter
production is balanced by respiration and oxygen periods, such as during periods of rising or falling
participation in weathering reactions with reducing sea level. For constituents with more rapid rates of
materials, such as sulphides and ferrous iron. turnover in the ocean, significant changes in ocean-
Probably, the partial pressure of oxygen was still ic concentration can occur over shorter periods and
rising in the early Phanerozoic Eon, but has there is considerable evidence, discussed below, of
changed relatively little during the Cenozoic Era. changes in the ocean’s regulation of atmospheric
Although ocean surface waters typically contain concentrations of carbon dioxide over the glacial to
dissolved oxygen at saturation concentrations, or inter-glacial time-scale (see also Chapter 12).
somewhat above, respiratory processes of organ-
isms, including bacteria, can lead to low concentra- The marine sedimentary sink
tions in some subsurface waters. In environments Incorporation of material into bottom sediments is
where the deeper waters are isolated from mixing a removal mechanism from the ocean, so that the
with surface waters, such as the Black Sea, anoxic sediments function dominantly as an ocean sink.
conditions can develop. The extent of these envi- The particulate material settling down the water
ronments is very limited at the present time, but column and depositing as sediments has two domi-
because of their sensitivity to changes in circulation nant sources. The input to the ocean of particles
the extent has varied episodically within the derived from the lithosphere (lithogenous material)
Phanerozoic Eon. has already been described. In the remote ocean,
The role of the ocean in relation to Earth surface this material accretes slowly, typically of the order
conditions appears differently from the standpoints of 1 mm in a thousand years. The other major
of different time-scales. Over the whole period of source is the production within the ocean of
the evolution of the sedimentary system major biogenous particles, mainly from surface-living
shifts have occurred in, for example, the distribu- organisms within the ocean. This includes a small
tion of sulphur between sulphides and sulphates, fraction of the organic matter which escapes oxida-
and that of carbon between carbonates and organic tion by respiration in the water column or at the
carbon7,8. In terms of these transfers, the ocean acts sediment surface (see Chapter 7 for an account of
as a medium of transfer rather than of storage. An the deposition of ‘marine snow’). The major

171
 J.D. Burton

11.6 11.7a

b
a

11.7b
d
c

e f
5 cm

Figure 11.6 Skeletons of planktonic organisms constitute Figure 11.7 Cross sections of ferromanganese nodular
a major part of the sediments over extensive areas of the concretions (‘manganese nodules’) from the sea floor. (a)
deep sea. The photographs, obtained from oceanic oozes Polished radial section of a nodule from the equatorial
with a scanning electron microscope, show calcareous North Pacific Ocean, showing asymmetrical growth of
[(a) foraminiferan, x 53; (b) coccolith, x 4000; (c) dis- layers around a nucleus (dashed white line) of consoli-
coaster, x 2700] and siliceous [(d) radiolarian, x 440; (e) dated sediment (note the discontinuity in the dark layer,
diatom, x 440; (f) silicoflagellate, x 970] planktonic indicated by white arrows, which shows that the orienta-
remains (Figure 7.7 from Kempe9, with permission from tion of the nodule on the sea bed has not remained con-
the Natural History Museum, London). stant; from von Stackelberg23; with permission from the
author and D. Reidel Publishing Company). (b) Polished
radial section of a nodule from the Blake Plateau
(Atlantic Ocean). The light-coloured veins consist largely
of calcite and clay, which have accumulated in cracks in
the growing nodule. The nodule accreted around a piece
of phosphorite, which is no longer clearly distinguishable
as a nucleus, having been partly replaced by ferroman-
ganese material (photograph by J. Mallinson).

biogenous components, however, are the minerals are the sediments deposited by hydrothermal
produced as skeletal material by organisms (Figure plumes (see Chapter 10) and the ferromanganese
11.6). Where these are important sedimentary com- concretions (Figure 11.7), which are abundant on
ponents, they accumulate typically at rates of about the sea bed in some regions and have potential eco-
1 cm every thousand years. nomic value (see Chapter 21). Other notable authi-
Of less importance in terms of sediment mass, genic phases include phosphorite (carbonate fluora-
but of great geochemical interest, are the solid patite), which occurs particularly in eastern bound-
phases formed by chemical reactions in sea water; ary regions of the ocean, where coastal upwelling is
these sediments are sometimes termed hydrogenous frequent, and glauconite (a potassium-rich alumi-
(formed in the hydrosphere), but more commonly nosilicate). Glassy basalts reacting with sea water
are described as authigenic. Important examples are so extensively altered that the minerals formed

172
 11: The Ocean: A Global Geochemical System

11.8

A B C
D E

Figure 11.8 Dominant sediment types, classified by origin, in different areas of the deep-sea bed. Low concentrations
of calcareous sediments in the North Pacific Ocean reflect a shallower calcite compensation depth (see text) in this
region; a general tendency can be seen for high concentrations of calcium carbonate to occur on the mid-ocean
ridges, where the depth of water is usually less than the calcite compensation depth (A, calcareous sediment; B, deep-
sea clay; C, glacial sediment; D, siliceous sediment; E, terrigenous sediment; ocean margin sediments occur in the
areas left blank; from Davies and Gorsline6, with permission from the authors and Academic Press).

should also be regarded as authigenic; these reac- operate around the world, they imprint their mater-
tions account for the abundant zeolite mineral, ial burdens on accumulating sediments.”
phillipsite, and the clay mineral, montmorillonite, The particle conspiracy involves many processes
in sediments of the South Pacific basin. During and reactions. Dissolved ions undergo electrostatic
some geological periods, large masses of evaporite exchanges at charged sites on particle surfaces.
sediments have been produced by the crystallisation Potentially stronger associations arise by specific
of salts from sea water that has become concentrat- binding with chemical groupings, especially
ed by evaporation in shallow marginal areas. Such hydroxyl (OH–) and carboxyl (COO–), a process
deposits are the starting materials for continental strongly encouraged by the ubiquity of coatings of
salt beds of economic importance. The major fea- organic matter on particle surfaces. Incorporation
tures in the distribution of sediment types in the into organic material, and into skeletal minerals
deep sea are shown in Figure 11.8 (see also Figure produced by organisms, involves the major struc-
8.21). tural elements, such as carbon, calcium, and sili-
con. Also, many of the trace elements become
The Oceanic Particle Conspiracy caught up in the formation processes, either
The dissolved constituents in the ocean can be because they have essential biochemical roles, like
removed, apart from those trapped in accumulating the metals which occur in enzymes, or adventitious-
sediments as pore waters (see also Figure 8.26), by ly, because of chemical similarities to the major ele-
conversion into particles or by binding to particle ments or a tendency to form complexes with organ-
surfaces. The solution composition is controlled by ic substances. As a result of these various processes
the balance between the input processes and the the concentrations of many elements are regulated
particle–solution interactions that occur in the at levels greatly below those at which actual precip-
water column. Turekian22 summed up these inter- itation occurs. The various processes by which ele-
actions as follows: “The great particle conspiracy is ments in solution at low concentrations become
active from land to sea to dominate the behaviour associated with particles already present in the sys-
of dissolved species ... As these scrubbing agents tem are collectively referred to as ‘scavenging’.
173
 J.D. Burton

11.9 Figure 11.9 Relative concentrations of the principal dis-

+ Mg2+ K+ solved constituents in the global river supply to the
Na ocean. This overview for the global average river water
(total salt concentration, ca 0.1 g/kg) is obtained from
data on the composition of many individual rivers, repre-
senting the range of lithologic, geographic, and climatic
SiO2 Cl– conditions. Individual rivers may show very wide varia-
tions from the mean (based on data of Meybeck14).
Ca2+
ocean with this drainage reside there for much
–
SO 24 longer times, on average, than do water molecules,
so sea water becomes more concentrated in both
elements (i.e., these accumulate in the ocean).
Sodium, however, accumulates to a greater extent
than does calcium. Elements of high geochemical
reactivity, with MORTs less than that of the water
HCO–
3 supplied by rivers, are depleted in sea water relative
to the concentration in which they are supplied in
river water. The contrasting abundances of a wide
range of elements in river and sea waters are shown
in Figure 11.11.
The vertical and horizontal distributions
Elements differ widely in their tendency to of chemical constituents
undergo transfer from solution to solid phases; The particle conspiracy not only affects the rates of
their ability to do so is often described as their ‘par- turnover of the elements in the ocean, it also pro-
ticle reactivity’ or ‘geochemical reactivity’. This ten- duces considerable variations in the concentration
dency is correlated with certain physicochemical of some elements within the ocean, both vertically
parameters, such as ionic potential (the ratio of and horizontally. The ocean mixes internally on a
charge-to-radius of the element cation), which pro- time-scale of about 1000 years. This is short
vide a rationalisation of geochemical behaviour in enough for the inputs of elements with long
terms of fundamental chemical properties, such as MORTs to become very thoroughly mixed within
periodicity. Differences in geochemical reactivity the ocean. Variations in concentration for these
are apparent from a comparison of the average constituents do occur near their sources and where
composition of the dissolved material in rivers rainfall and evaporation change the salt content
entering the ocean (Figure 11.9) and that of sea significantly. When normalised for salinity, howev-
water (Figure 11.1). Some differences (e.g., for er, the concentrations are uniform, within the limits
chloride) are mainly because the constituent is
dominantly supplied as a volatile rather than as the
product of rock weathering. Features such as the
much lower ratio of calcium-to-sodium in sea 11.10
water than in river water, however, reflect the fact Input flux OCEAN Output flux
that constituents introduced by rivers accumulate TOTAL MASS
to different extents in sea water; calcium is much #
of constituent
OF CONSTITUENT #
of constituent
more efficiently removed from the ocean than is (X tonne)
(Y tonne/yr) (Y tonne/yr)
sodium, because, as already seen, it enters into the
formation of abundant sediments containing
biogenous calcium carbonate.
An extremely useful, albeit crude, index of geo- Figure 11.10 The mean oceanic residence time (MORT)
chemical reactivity is the mean oceanic residence of a constituent is derived by making the steady-state
time (MORT) (Figure 11. 10). The MORTs for assumption that input and removal are balanced (see
sodium and calcium are of the order of 108 years text) and is equal to X/Y years. Estimates of MORT are
commonly based on the river input to the ocean. The
and 106 years, respectively. The MORT of water model also assumes that a constituent becomes well
molecules in the ocean reservoir is dependent on mixed within the ocean reservoir prior to removal. This is
their passage through the atmosphere and return not the case with the most particle-reactive metals (such
through continental drainage and can be derived as aluminium, iron, and manganese). For these elements,
from the information in Figure 11.4; it is about 4 x the calculated MORTs are of the order of 100–1000
104 years. Sodium and calcium ions carried into the years, which must be treated as nominal values.

174
 11: The Ocean: A Global Geochemical System

Figure 11.11 Comparison of the average concentrations 11.11

Cl Na
of various elements in sea water (cSW) with those in the
average river-water (cRW) entering the ocean. The solid Mg
S
line represents equal concentrations of an element in sea
water and in river water (the dashed lines show limits of 2 Ca K
C
one order of magnitude above and below the 1:1 rela- Br
tionship). Elements above the line accumulate in sea B
water to concentrations above those in the river water, in F
4 Sr N Si
some cases to a very substantial degree (e.g., Na). Those Li
considerably below the line are particle-reactive ele-

–log csw
ments (e.g., Pb), which are markedly depleted in sea Rb P
water relative to the concentration in the average river- 6 I
Ba
water supply to the ocean (after Whitfield and Turner24; Mo Al
Zn
with permission from the authors and Macmillan Ni As Y Fe
Magazines Ltd). U
8 Se Cr Ti Cu
Cs Sb Mn
W Ag Co
of analytical measurement, over almost the whole
Ga Zr
of the ocean. For example, the ratio of the concen- 10
Hg
Au Sn Pb
tration of sulphate to salinity in the Irish Sea does Th
Sc La
not differ detectably from that in the Pacific Ocean.
Such constituents are termed ‘accumulated’ [Figure
11.12(a)]; like salinity, their concentrations are
10 8 6 4
conserved in ocean mixing and they are, therefore,
also referred to as conservative constituents. The –log crw
major constituents tend to be conservative. The
near-constancy of the relative proportions of the
major constituents in sea water provides the basis mal), they tend to decrease markedly in concentra-
for practical measurements of salinity for both the tion away from the lateral and surface boundaries.
present-day approach, based on conductivity, and These constituents are termed ‘scavenged’ [Figure
the main previous approach based on chlorinity (a 11.12(b)].
defined quantity related to the concentrations of There is a very important further group of con-
chloride and bromide). stituents with intermediate residence times that are
Some minor constituents also show essentially long enough, on the face of it, for them to become
conservative behaviour, but their low abundance in quite uniformly distributed by the water-mixing
the supply to the ocean limits the oceanic concen- processes in the ocean [Figure 11.12(c)]. Their par-
trations that they can attain. These elements ticle reactivity is largely accounted for by associa-
include lithium, caesium, and rubidium, which tions with biogenous particles; they include the
have chemical properties similar to sodium and main micronutrients – phosphate, nitrate, and dis-
potassium, and uranium, which has a low particle solved silicon. These constituents tend to be effi-
reactivity because it forms a strong complex with ciently stripped from the surface waters, where
dissolved carbonate ions. most biogenous particles are produced.
Elements with short MORTs are removed, how- Decomposition of organic matter and dissolution
ever, too rapidly to become well-mixed. Their high of mineral skeletons deeper in the water column,
geochemical reactivity means that, in the absence of or at the sedimentary interface, lead, however, to
local sources at the deep sea-bed (e.g., hydrother- release of these elements from sinking particles.

11.12 Figure 11.12 Schematic repre-

(a) (b) (c) sentation of idealised vertical
profiles for elements represent-
Concentration Concentration Concentration ing the three general types of
0 behaviour in the ocean: (a)
accumulated (normalised to
Depth (km)

salinity); (b) scavenged; and (c)

recycled. The factors which
underlie these types of distribu-
tion are explained in the text
and actual examples of types
(b) and (c) are given in Figures
4 11.14 and 11.15, respectively.

175
 J.D. Burton

Where the permanent pycnocline forms a stratified particle reactivity and undergo recycling and scav-
water column, the return of the dissolved material enging, but for some the consequences, in terms of
to the surface by water transport is much less distribution, are less marked; in some cases they
effective than is the down-column transport when may be difficult or impossible to detect analytically.
the element is linked to a particle. There is, there- This may be because the down-column flux is low
fore, an efficient ‘biological pump’ for these con- in absolute terms or because it is low relative to the
stituents from surface to deep water, so they dissolved concentration in the ocean reservoir. Only
increase in concentration with depth over much of a slight vertical concentration gradient is detected
the ocean, an increase which may be very pro- for calcium in the ocean, despite the large flux of
nounced [Figure 11.12(c)]; they are termed ‘recy- calcium carbonate from surface waters to the deep
cled elements’. A proportion of the sinking parti- ocean, because of the high concentration of calcium
cles undergo deposition into the sediment, but the in sea water. Few constituents of sea water are as
average atom of a recycled element travels with a abundant as calcium, however, and generally a
particle into the deeper water and is released and major involvement in the particle conspiracy is
returned to the surface layer many times before it reflected in clearly nonconservative behaviour.
is finally removed from the water column. The two main types of nonconservative behav-
The organic matter produced by photosynthesiz- iour are reflected in systematic horizontal variation
ing organisms in the upper ocean is a complex mix- in concentrations in the intermediate and deep
ture of compounds. Nevertheless, in the deeper waters of the major ocean basins3. A major part of
waters, where recycling occurs, the increases in the the deep water in the ocean basins has a common
concentrations of nitrate and phosphate and the origin in surface waters of the North Atlantic
decrease in the concentration of dissolved oxygen, Ocean, particularly in the Norwegian Sea, where
which is used to oxidise the organic matter, occur in dense waters formed by surface cooling sink and
essentially constant proportions. This reflects the fact flow south in the western Atlantic (Figure 11.13).
that, when averaged over time and space, the uptake The deep water, modified by additions from the
of the main elements (carbon, nitrogen, and phos- Southern Ocean, flows northward in the Indian
phorus) used to synthesize cellular material occurs in Ocean and in the Pacific Ocean. During this transit,
a common relationship, which reflects the mean com- the composition of the water is influenced by the
position of the mixture of compounds produced, particles which enter it from the surface. As the
such as carbohydrates, lipids, proteins, and nucleic water becomes older, in the sense that more time
acids. The atomic ratios in which carbon, nitrogen, has elapsed since it sank from the surface, more
and phosphorus are utilised are taken as 106:16:1; particles passing through on their way to the sedi-
oxygen is utilised in the oxidation of organic matter ments scavenge the strongly particle-reactive con-
in the ratio of 138 molecules per 106 atoms of car- stituents. The concentration of these constituents
bon. These are known as Redfield ratios18. therefore decreases along the direction of the deep-
The types of distribution shown in Figure 11.12 water flow (Figure 11.14). In contrast, the recycled
are idealised, but they also match quite closely the constituents increase in concentration (Figure
actual vertical profiles of constituents which show 11.15), because the older waters have been subject-
these behaviours to a marked extent (see Figures ed to more releases from biogenous particles sink-
11.14 and 11.15). All constituents show some ing from the waters that overlie their route.

11.13
ATLANTIC

Source
PACIFIC

▲
▲
▲

▲ INDIAN
Figure 11.13 Schematic representation of the circu-
▲
▲

lation of deep water in the major ocean basins

▲
▲

(Broecker and Peng 3; with permission from the

authors). The solid lines show the flow of deep
▲
▲

water, originating largely by sinking in the far North

Recooling Atlantic Ocean. The dashed lines show the flow of
surface water and the filled circles represent areas
ANTARCTICA of localised upwelling.

176
 11: The Ocean: A Global Geochemical System

Figure 11.14 Vertical profiles of some (a) (b) (c) 11.14

scavenged dissolved elements in the
Dissolved manganese (nmol/l) Dissolved aluminium (nmol/l) Dissolved aluminium (nmol/kg)
oceanic water column: (a) dissolved 0 0.5 1.0 1.5 2.0 0 20 40 0 2 4 6
manganese and (b) dissolved alumini- 0 0 0
um in the northeastern Atlantic
Ocean; (c) dissolved aluminium in the 1 1
central North Pacific Ocean. ( )
2
Dissolved aluminium has a MORT of 2

Depth (km)
Depth (km)
2

Depth (km)
about 1000 years, and is therefore
removed from the ocean on a time- 3
3 4
scale similar to the transit time of deep
water from the North Atlantic Ocean 4
4
to the North Pacific Ocean. There is 6
thus a marked decrease in concentra- 5
( ) 5
tion between the two profiles. A
strong sedimentary source shows up
in the bottom waters in the North Pacific profile against the low deep-water concentrations. Surface input is dominated by
dissolution of aeolian-transported mineral dust in profiles (a) and (c); this also adds significantly to the off-shelf transport of
dissolved aluminium, which is important in the surface waters in profile (b). Dissolved manganese has a shorter MORT
than aluminium and varies less in concentration in the deep ocean, except where in situ sources (e.g., hydrothermal) are
significant. Compare these examples with the schematic representation in Figure 11.12(b). Data are from (a) Statham et
al.19 (with permission from Macmillan Magazines Ltd), (b) Measures et al.13 (with permission from the authors and Elsevier
Science Ltd), and (c) Orians and Bruland17 (with permission from the authors and Elsevier Science Ltd).

Figure 11.15 (a) Vertical profiles of some recycled dissolved (i) Si (+mol/l) (iii) Cd (nmol/l) 11.15a
constituents in three major ocean basins (squares, North Atlantic4; 0 100 200 0 0.5 1.0 1.5
16 4
circles, southwestern Indian ; diamonds, central North Pacific ):
the micronutrients are (i) silicon and (ii) phosphate, and the trace
metals are (iii) cadmium and (iv) nickel. Within the overall feature
Depth (km)

2
of increasing concentration with increasing ‘age’ of the deep
water, detailed differences in the profiles can be seen; in particu-
4
lar, nickel is less depleted in the surface waters than are the other
constituents. A very close similarity in distribution is shown by
phosphate and cadmium, reflecting the association of the metal, 6
which has no known biological function, with organic matter (ii) 0 PO43– (+mol/l) (iv) Ni (nmol/l)
sinking from the surface to deeper waters. The more gradual 2 4 0 5 10 15
increase in concentration with depth for silicon, compared with
phosphate, occurs because the skeletal material which carries sili-
2
Depth (km)

con down the water column is recycled deeper in the ocean than
are the soft organic tissues. Compare these examples with the
schematic representation in Figure 11.12(c). Figure from Morley et 4
al.16 (with permission from Elsevier Science Ltd). (b) The vertical
distribution of concentrations of nitrate (μmol/kg) along a section
6
through the western Atlantic, and Southern and Central Pacific
20
Oceans (after Sharp , based on GEOSECS data; with permission
from the author and Academic Press). The section illustrates more fully the pronounced increase in concentration with
depth and the increase along the direction of the major deep-water circulation.

Central Pacific Antarctic Western Atlantic 11.15b

15 1
1015 5
10 15
20 20 25
35 33 34 35
1 30 25 34 30
32 31 32
44 41 34 33
45 43 31 30 19
42 40 35 17 17
44 31 18
2 43 41
36
25 19
39 32
42
Depth (km)

20 16
20
38
3 32
33
37

6
54˚N 32˚ 16˚ 0 16˚ 32˚ 48˚ 68˚S 58˚S 48˚ 32˚ 16˚ 0 16˚ 32˚ 48˚ 64˚N

177
J.D. Burton

Marine Biogeochemical Processes and aquatic concentrations of nitrate should lead to the
Global Environmental Conditions production of more organic matter, the respiration
It is clear from the foregoing that biological of which by normal aerobic processes utilises oxy-
processes play a crucial role in the major sedimen- gen. The maintenance by organisms of non-equilib-
tary cycle and in the internal cycling of material rium conditions by feedback mechanisms, illustrat-
within the ocean. Indeed, many geochemical ed in a simplified form by this example, is a key
processes are termed biogeochemical because of element in the Gaia hypothesis10,11 that the bios-
this role. Biological oceanographic processes, dri- phere controls its environment in such a way that it
ven essentially by solar radiation which provides is favourable for life processes (see also Box 12.7).
the energy for photosynthetic primary production Further consequences of biological processes are
of organic matter (see Chapter 6), have three conse- that the distribution of constituents between the
quences of the greatest importance for surface- major reservoirs of material at the Earth’s surface is
Earth conditions. very different from what it would be in a similar
One of these consequences is that many elements abiotic system, and that the distribution of many
can occur in chemical forms which are thermody- constituents within the oceanic reservoir is likewise
namically unstable; their concentrations reflect the different. These aspects are well-illustrated by refer-
balance between their continuous biological produc- ence to the marine carbonate system (Chapter 12).
tion and their conversion into more stable chemical Most of the carbon dioxide released as a magmatic
species. An example of this is the element nitrogen, volatile during Earth’s chemical evolution (see ear-
which is supplied to the surface Earth as nitrogen lier) has been bound up as carbonates in sediments
(N2) gas, a volatile which accounts for about 80% and sedimentary rocks, with a considerable fraction
by volume of the atmosphere. Under the oxidising also locked away in sedimentary deposits as
conditions which generally prevail at the Earth’s sur- unrespired organic carbon compounds. If all the
face, dinitrogen is thermodynamically unstable and, carbon dioxide was present in the gaseous form,
on that basis, should be oxidised to provide, ulti- the atmosphere would be fairly similar to that of
mately, nitrate ions in the hydrosphere. Oxidation of Venus, where the surface pressure is about 75
nitrogen by ordinary chemical reactions occurs, atmospheres and carbon dioxide constitutes 95%
however, at a negligible rate, because the bond of the atmosphere, compared with 0.03% of
between the two nitrogen atoms is exceptionally Earth’s atmosphere. The radiation balance for
strong, representing an energy barrier to the process. Venus, in which the ‘greenhouse’ gases carbon
This energy barrier is overcome during high-energy dioxide and water vapour play a major role, leads
events in the atmosphere. Also, some organisms have to a surface temperature of about 450°C, com-
the ability to fix nitrogen enzymatically, that is to pared with the mean value for Earth of 15°C.
convert the nitrogen molecule into ammonia, in The ocean is the major reservoir of carbon in the
which form it can be utilised biologically; this nitro- exchangeable carbon system, containing about 60
gen eventually is converted into nitrate in the nitro- times as much carbon as the atmosphere. The way
gen cycle. These conversions are slow in relation to in which this reservoir exerts a major role in deter-
the magnitude of the atmospheric mass of dinitro- mining the steady-state partial pressure of carbon
gen, so the gas has a mean atmospheric residence dioxide in the atmosphere, and the manner in
time of about 106 years. Nevertheless, if biological which it has responded to increases in partial pres-
processes did not reverse these reactions, the reser- sure as a result of human activities over the past
voir of atmospheric nitrogen would run down to a 200 years, are among the most important questions
very low concentration over a period of some 107 in chemical oceanography today. The physicochem-
years, with a correspondingly massive increase in the ical and geochemical factors involved are compli-
concentration of nitrate in the ocean. Yet, the major cated by the fact that inorganic carbon in the ocean
composition of the atmosphere has remained rela- is present in different chemical forms: as dissolved
tively constant over a significantly longer period8. carbon dioxide, entering and leaving by gas
The reversal through reduction of nitrate to nitrogen exchange across the atmosphere–ocean interface,
by organic matter, a process termed denitrification, and as bicarbonate (HCO3–) and carbonate (CO32–)
occurs through bacterial respiration in certain aquat- ions, entering with river water as a result of the
ic and sedimentary environments, which are defi- weathering reactions of carbon dioxide with conti-
cient in free oxygen. Marine environments of this nental rocks.
kind, while they form only a limited part of the Dissolved carbon dioxide in the ocean reacts
ocean system, account for a significant part of global with water to form carbonic acid (H2CO3), which
denitrification. dissociates to produce bicarbonate and carbonate
Denitrification can act as a negative feedback ions. It also reacts with carbonate ions derived
mechanism in the nitrogen cycle, because increased from the river-water input to form bicarbonate

178
 11: The Ocean: A Global Geochemical System

ions. As a consequence, sea water can absorb more temperature, both of which lead to an increased
carbon dioxide than would be the case if the salts solubility of calcium carbonate. The concentration
consisted entirely of, say, sodium chloride; with the of calcium in sea water varies to only a small extent
present-day composition of the ocean, this factor throughout the ocean, its behaviour being nearly
increases the absorptive capacity of the ocean for conservative. In contrast, the concentration of dis-
carbon dioxide by a factor of about eight. solved carbonate ions shows a marked general
A further consequence is that the carbon dioxide decrease with depth and also decreases broadly
dissolved in the mixed layer of the ocean equili- with the ‘age’ of the deep water as it is transported
brates with atmospheric carbon dioxide rather through the Atlantic, Indian, and Pacific Oceans.
slowly; virtually complete equilibration takes about These features reflect both the operation of the bio-
1.5 years. As a result, marked long-term disequilib- logical pump, transporting carbon dioxide to the
ria can exist regionally between the partial pres- deeper waters, and the high concentrations of car-
sures in the atmosphere and in surface waters, par- bon dioxide, derived from atmospheric exchange,
ticularly in areas of major upwelling and down- in cold waters. The overall consequence of the
welling of water, although the exchange of gas into increase in concentration of carbon dioxide is to
and out of the surface layer is in approximate bal- convert some carbonate ions into bicarbonate ions.
ance for the whole ocean. Thus, in some parts of the ocean, the concentration
The biological pump exerts a powerful influ- of carbonate ions in the deeper waters falls below
ence, drawing down carbon dioxide into the deeper that corresponding to saturation of the water with
ocean. This occurs through the uptake of carbon respect to solid-phase calcium carbonate.
dioxide and its conversion into organic carbon There are two horizons in the water column
compounds by photosynthesizing organisms in the which are significant in relation to the preservation
surface waters, and the transfer of a fraction of the of calcium carbonate particles in the bottom sedi-
organic material into deeper waters, where the car- ments. At the lysocline, the degree of undersatura-
bon dioxide is released by respiratory oxidation, tion with respect to calcium carbonate is sufficient
mainly bacterial. The factors controlling the verti- for dissolution of the particles to become signifi-
cal distribution of the inorganic species in the cant (see Figure 12.6). If the total depth of water is
ocean are particularly complex. In addition to the less than the depth of the lysocline, particles accu-
recycling processes, waters at different depths mulate without loss by dissolution. If it is greater
acquired different concentrations of carbon dioxide than that of the lysocline, some dissolution will
by exchange with the atmosphere when they were occur, and if the sea bed is at a depth much below
at the surface; also, the distribution of the total the lysocline, calcium carbonate can undergo disso-
inorganic carbon among its principal species is lution to such an extent that it ceases to be a signif-
dependent upon the in situ temperature and pres- icant sedimentary component. This latter horizon is
sure, which affect the equilibrium constants for the termed the calcite compensation depth, calcite
system (Chapter 12). The importance of the biolog- being the dominant mineral form of calcium car-
ical pump can be seen, however, from calculations3 bonate produced by the surface-living organisms in
which show that if deep water of average composi- the ocean. The difference in the depths of the lyso-
tion were brought to the surface and warmed to cline and the calcite compensation depth is not con-
about 20°C, then in the absence of biological activ- stant; it depends primarily upon the gradient in
ity its partial pressure of carbon dioxide would be concentration of carbonate ions in the water col-
about 1000 parts per million (p.p.m.), compared umn overlying the sediment. On average, the depth
with the present atmospheric partial pressure of at which the concentration of calcite in sediments
about 350 p.p.m. falls to only a few per cent is about 700 m deeper
The distribution of inorganic carbon among its than the lysocline.
principal species in deep water also exerts a geo- The depths at which these horizons occur are
chemical control upon the accumulation of calcium typically some thousands of metres, but are not the
carbonate in sediments. In the surface ocean, the same throughout the oceans, since they depend
concentrations of dissolved calcium and carbonate essentially upon the concentration of carbonate
ions correspond to a condition of supersaturation ions in the bottom waters (they are deeper in the
with respect to solid-phase calcium carbonate. The Atlantic Ocean than in the Pacific Ocean). Such dif-
precipitation of calcium carbonate is, however, ferences are reflected in the distribution of calcare-
dominantly a consequence of the production of ous sediments below the world ocean (see Figure
skeletal material by organisms, rather than a 11.8). Since a change in the concentration of car-
physicochemical phenomenon. As particles of calci- bon dioxide in sea water alters the concentration of
um carbonate sink down the water column, they carbonate ions, such a change also leads to greater
enter environments with higher pressure and lower dissolution or preservation of calcium carbonate in

179
 J.D. Burton

11.16 Number concentration Scattering of solar Cloud Figure 11.16 Postulated feedback mechanism linking
of cloud droplets radiation by droplets albedo global climate and production of dimethylsulphide
(fixed LWP) +
(DMS) by phytoplankton in the surface layers of the
ocean (from Charlson et al.5; with permission from the
Cloud nucleation authors and Macmillan Magazines Ltd). The processes
+ shown in the ovals affect the quantities shown in the rec-
Loss of solar tangles; the sign in the oval indicates whether an increase
CCN radiation to space in the preceding quantity leads to an increase (+) or
–
decrease (–) in the next quantity. The mechanism can act
Formation of as a regulator of climate if increased surface-Earth tem-
water-soluble particles
+
perature and solar irradiance lead to an increased pro-
duction of DMS by marine phytoplankton; this depends
NSS SO24
– on a complex of factors, especially since production of
Surface Solar
temperature irradiance DMS varies with the species composition of phytoplank-
Oxidation
of Earth below clouds ton populations (NSS SO 42– = non-sea-salt sulphate;
CCN = cloud-condensation nuclei; LWP = liquid water
+
path).
DMS(gas)

Sea-to-air
the carbon dioxide added to the atmosphere by fos-
transport sil fuel combustion, about 35% has been taken into
+ the ocean (see Chapter 12). Understanding the role
Atmosphere
of the ocean in the carbon system is clearly a vital
Ocean
DMS(aq)
Production of DMS component, alongside knowledge of the role of the
by marine phytoplankton
+/–?
terrestrial biosphere, of our ability to model and
predict global chemical and climatic change (see
Chapter 3). Ultimately, however, the direction and
extent of climatic change depend upon many inter-
sediments. This occurs, however, on longer time- linked factors, involving the cycles not only of car-
scales than those applicable to the exchangeable bon, but also of other major biogeochemical
carbon system. elements.
It is apparent that the global cycle of carbon is Two examples serve to illustrate this point. First,
crucially dependent upon the role of the ocean, in the volatile compound, dimethylsulphide, is released
which the balance of the biogeochemical processes to the atmosphere as a result of its production in
is finely tuned to the preservation of carbonate sed- surface waters by some kinds of phytoplankton (see
iments over extensive areas of ocean floor and to Chapter 6); it is the most abundant of a number of
the storage of carbon dioxide in the deep waters – sulphur-containing gases which are released from
to a much greater extent than would be the case for the surface ocean as a result, directly or indirectly,
an abiotic ocean. Consequently, changes in the of biological activity1. In the atmosphere, it is con-
complex of interacting processes involved in ocean verted into sulphur dioxide, which forms a sul-
circulation and biological productivity may be phuric acid aerosol that can act as a nucleus for
expected to perturb the system significantly. cloud condensation. It has been postulated5 that
Evidence from ice cores [see Figure 12.1(a)] shows changes in the flux of dimethylsulphide as a result
that during the 160,000 years prior to 1800, when of changes in the temperature of the upper ocean
human activities began to increase the concentra- could feed back to modify the temperature through
tion of carbon dioxide in the atmosphere, the a change in cloud cover (Figure 11.16); another
atmospheric pressure of carbon dioxide varied example of the Gaia hypothesis of homoeostatic
within the range of about 200–300 p.p.m. regulation of environmental conditions.
Minimum values occurred around the most recent Second, another hypothesis links the aeolian
glacial maximum, about 18,000 years ago. These transport of mineral dust to the recycling of carbon
are highly significant changes, but when it is dioxide. At high latitudes, surface waters, unlike
remembered that there is so much more carbon in those of the stratified ocean, do not become strong-
the ocean than in the atmosphere, the comparative- ly depleted in the principal micronutrients (nitrate,
ly narrow limits of variation provide striking testi- phosphate, dissolved silicon) needed for phyto-
mony to the efficiency of the ocean as a regulator plankton growth. Experimental work has shown
of Earth’s surface conditions. that the addition of dissolved iron to such waters
Since man became a significant agent in the car- promotes an increased utilisation of these relatively
bon cycle, the atmospheric partial pressure of car- abundant micronutrients and of carbon dioxide in
bon dioxide has increased to about 350 p.p.m. Of photosynthesis (see Chapter 12). There is evidence

180
 11: The Ocean: A Global Geochemical System

from ice cores in Antarctica that during the most Chester, R. (eds), Academic Press, London, pp
recent glacial period a much higher deposition of 157–220.
5. Charlson, R.J., Lovelock, J.E., Andreae, M.O., and Warren,
mineral dust (which is a potential source of dis- S.G. (1987), Oceanic phytoplankton, atmospheric sul-
solved iron for surface waters) occurred than in the phur, cloud albedo and climate, Nature, 326,
interglacial periods. These findings have led to the 655–661.
hypothesis12 that an increased efficiency of the bio- 6. Davies, T.A. and Gorsline, D.S. (1976), The geochemistry of
deep-sea sediments, in Chemical Oceanography,
logical pump in the Southern Ocean played an Volume 5, Riley, J.P. and Chester, R. (eds), Academic
important role in the reduction of the atmospheric Press, London, pp 1–80.
partial pressure of carbon dioxide during the 7. Garrels, R.M. (1985), Sediment cycling during Earth history,
glacial period. in Physical and Chemical Weathering in Geochemical
Cycles, Lerman, A. and Meybeck, M. (eds), Kluwer,
Concluding Comments Dordrecht, pp 341–355.
8. Holland, H.D. (1984), The Chemical Evolution of the
In this chapter emphasis has been placed mainly on Atmosphere and Oceans, Princeton University Press,
large-scale features in the distributions of chemical Princeton, 582 pp.
species and the role of the ocean in the context of 9. Kempe, D.R.C. (1981), Deep ocean sediments, in The
the global environment. Knowledge in these areas Evolving Earth, Cocks, L.R.M. (ed.), British Museum
has advanced rapidly during the past two decades, (Natural History) and Cambridge University Press,
London and Cambridge, 264 pp.
following the systematic measurements of con- 10. Lovelock, J.E. (1979), Gaia: A New Look at Life on Earth,
stituents in the major ocean basins that was begun Oxford University Press, Oxford, 157 pp.
in the late 1960s under the Geochemical Ocean 11. Lovelock, J.E. and Margulis, L. (1974), Atmospheric home-
Sections (GEOSECS) programme. As explained in ostasis by and for the biosphere: the Gaia hypothesis,
Tellus, 26, 2–10.
the Introduction, these advances have depended 12. Martin, J.H. (1990), Glacial–interglacial CO2 change: the
critically upon improved analytical methods. The iron hypothesis, Paleoceanogr., 5, 1–13.
ocean system is chemically driven by processes that 13. Measures, C.I., Edmond, J.M., and Jickells, T.D. (1986),
occur on small scales, such as dissolved chemical Aluminium in the northwest Atlantic, Geochim.
species being scavenged at particle surfaces, or Cosmochim. Acta, 50, 1423–1429.
14. Meybeck, M. (1988), How to establish and use world bud-
taken up and transformed by microalgae; the role gets of riverine materials, in Physical and Chemical
of organic material in such processes is recognised, Weathering in Geochemical Cycles, Lerman, A. and
but little understood at a molecular level. Meybeck, M. (eds), Kluwer, Dordrecht, pp 247–272.
Developments in ocean chemistry depend upon 15. Milliman, J.D. and Meade, R.H. (1983), World-wide deliv-
ery of river sediment to the oceans, J. Geol., 91, 1–21.
advances in these complementary approaches, and 16. Morley, N.H., Statham, P.J., and Burton, J.D. (1993),
upon the new insights developing, for example, Dissolved trace metals in the southwestern Indian
through the application of photochemistry and col- Ocean, Deep-Sea Res. I, 40, 1043–1062.
loid science to marine environmental systems. 17. Orians, K.J. and Bruland, K.W. (1986), The biogeochem-
istry of aluminium in the Pacific Ocean, Earth Planet.
Sci. Lett., 78, 397–410.
18. Redfield, A.C., Ketchum, B.H., and Richards, F.A. (1963),
The influence of organisms on the composition of sea
General References water, in The Sea, Vol. 2, Hill, M.N. (ed.),
In addition to the texts by Broecker and Peng3 and Holland8 list- Interscience, New York, pp 26–77.
ed below, several other books provide useful background: 19. Statham, P.J., Burton, J.D., and Hydes, D.J. (1985), Cd and
Chester, R. (1990), Marine Geochemistry, Unwin Hyman, Mn in the Alboran Sea and adjacent North Atlantic:
London, 698 pp. geochemical implications for the Mediterranean,
Libes, S.M. (1992), An Introduction to Marine Biogeo- Nature, 313, 565–567.
chemistry, Wiley, New York, 734 pp. 20. Sharp, J.H. (1983), The distributions of inorganic nitrogen
Volumes in the series Chemical Oceanography [Vols 1–4 and dissolved and particulate organic nitrogen in the
(1975), Riley, J.P. and Skirrow G. (eds), and Vols 5–10 sea, in Nitrogen in the Marine Environment,
(1975–1989), Riley, J.P. and R. Chester (eds), Academic Press, Carpenter, E.J. and Capone, D.G. (eds), Academic
London] provide accounts in depth of topics across the whole Press, New York, pp 1–35.
range of marine chemical interests. 21. Thompson, G. (1983), Hydrothermal fluxes in the ocean, in
Chemical Oceanography, Volume 8, Riley, J.P. and
Chester, R. (eds), Academic Press, London, pp
271–337.
References 22. Turekian, K.K. (1977), The fate of metals in the oceans,
1. Andreae, M.O. (1990), Ocean–atmosphere interactions in Geochim. Cosmochim. Acta, 41, 1139–1144.
the global biogeochemical sulfur cycle, Marine 23. von Stackelberg, U. (1987), General history and variability
Chem., 30 1–29. of manganese nodules of the equatorial North Pacific,
2. Baumgartner, A. and Reichel, E. (1976), The World Water in Marine Minerals, Teleki, P.G., Dobson, M.R.,
Balance, Elsevier, Amsterdam, 176 pp. Moore, J.R., and von Stackelberg, U. (eds), Reidel,
3. Broecker, W.S. and Peng, T.-H. (1982), Tracers in the Sea, Dordrecht, pp 189–204.
Eldigio Press, Palisades, 690 pp. 24. Whitfield, M. and Turner, D.R. (1979), Water–rock parti-
4. Bruland, K.W. (1983), Trace elements in sea-water, in tion coefficients and the composition of sea water and
Chemical Oceanography, Vol. 8, Riley, J.P. and river water, Nature, 278, 132–137.

181
CHAPTER 12:

The Marine Carbonate

System
M. Varney

Introduction
0.4
Global warming is more than just a topic of scien- 12.1
370
tific interest. It has come to dominate the environ-

Carbon dioxide concentration (parts per million)

Mean temperature change from 1950 – 1979 (˚C)
mental agenda, to tax the minds of political leaders 0.2 360
around the world, to affect the way energy compa-
350
nies do their business, and to concern the man in
0
the street. 6T 340
The main cause of global warming is the rise
330
since 1900 in the content of the so-called green- –0.2
house gases, the main culprit being carbon dioxide, 320
CO2. The current rise in atmospheric levels of CO2
310
(Figure 12.1) is of great concern because CO 2 –0.4
pCO2
absorbs spectral components of solar radiation – it 300
is a strong short- and long-wavelength, infra-red
–0.6 290
absorber (Figure 12.2). The present day concentra-
tions of CO2 are the highest in the recorded history 280
of the Earth (compare Figures 12.3 and 12.4). The –0.8
1860 1880 1900 1920 1940 1960 1980 2000
absorbed energy is partly dissipated as heat and
Year
warms the air, land, and sea. When CO2 concentra-
tions rise, the rate of heat loss to space is exceeded Figure 12.1 The relationship between temperature and
by the absorption of incident solar radiation by the CO2 levels has been quite variable over the past 130
atmosphere, which increases the amount of trapped years. The true relationship is more complex than this
heat at the Earth’s surface – causing a rise in sur- illustration implies, and there has been considerable
face temperatures and in sea level (from the melting debate about whether changes in the atmospheric [CO2]
of glaciers and the thermal expansion of water). are the response to temperature fluctuations, or vice
Rainfall patterns and storm activity are likely to versa. Several hypotheses have been proposed, relating
change as a direct result. Our climate is becoming changes in [CO2] to changes in biological productivity, in
steadily hotter and we cannot predict what will sea level, and in the circulation of surface and deep
happen in the future because we have no accurate waters (including the relative importance of deep water-
mass formation in polar latitudes, and of upwelling
comparison from the past. Predictions of changes regions, where biological productivity is high). As these
to the Earth’s climate are extremely limited due to factors are all interrelated, the resulting models are com-
our current lack of knowledge about the various plex and no definitive answer is yet available. It is rea-
interactions of CO2 with land vegetation and the sonably certain that varying levels of CO2 are not the
oceans. main cause of temperature fluctuations – but once estab-
The oceans play an important role in the climate lished they are likely to reinforce climatic changes, not
system by regulating the amount of CO 2 in the initiate them. The concentrations of other atmospheric
atmosphere. This is one reason why it is important gases, such as methane, ozone, and freons, are also ris-
to understand thoroughly the workings of the ing. We are therefore in the middle of a global experi-
marine carbonate system, which controls the move- ment in which several geochemical cycles are being per-
turbed. Insights into the mechanisms of these cycles can
ment of CO2 between the ocean and the atmos- be obtained from palaeoceanographic research (see also
phere. The marine carbonate system is also one of Figure 12.3). (Adapted from Schneider9.)
the key ingredients in the story of life in the sea,
dissolved CO2 being the primary source of carbon
for marine plant life and the primary component
(along with water) in the process of photosynthesis.
CO2 is the third most abundant dissolved gas in

182
 12: The Marine Carbonate System

Figure 12.2 The 100

12.2

atmospheric gases (%)

absorption of eoretical radiation
s th
incoming solar 80 rth' CO2 H2O

Absorption by
H2O Ea
(infra-red) radia- 60
tion by the atmos- O3
pheric gases H2O, 40
CO2, and O3. The 20
amount of radia-
tion emitted by the 0
4 6 8 10 12 14 16 18 20 22 24 26
Earth’s surface is
shown by the Wavelength (+m)
dashed line.
(Adapted from
Coulson2.)

Figure 12.3 In polar regions, partial melting of surface

Carbon dioxide concentration (parts per million)

5.0 300 12.3
snow and ice occurs during the summer months; refreez-
ing in winter produces an ice layer on top of the old

Temperature change from present (˚C)

2.5 280
snow. Since snow is porous, small pockets of air are
trapped in its pores when the surface layer of ice freezes. pCO2
These ‘bubbles’ reflect the atmospheric composition at 0 260
the time of entrapment. These annual markers, like tree
rings, of ice and snow can be used to pinpoint the time of –2.5 240
their deposition (see also Box 8.3). Observations from ice
and sediment cores demonstrate that large-scale changes 6T
in atmospheric CO2 levels occurred over geological time –5.0 220
and are correlated with atmospheric temperature changes
over the past 160,000 years. The correlation between –7.5 200
atmospheric [CO2] and temperature change (6T) is clear.
Cold temperatures (during glacial periods) appear to be –10.0 180
associated with low levels of atmospheric CO 2. The
deglaciation events of about 140,000 years and 15,000
years ago are particularly obvious. These fluctuations are 160 140 120 100 80 60 40 20 0
as expected – CO2 is more soluble in cold than in warm Thousands of years ago
water, and its atmospheric concentration should there-
fore be less during glacial (lower mean temperatures)
than interglacial periods. (Adapted from Schneider9.)

Figure 12.4 The annually-aver- 332 12.4

xxxx
aged atmospheric concentra- x xx
xxxxxxxx
x
tions of CO 2 at Hawaii (dots) xxx
x
xxxxxxxxxxxxxxx
and the South Pole (crosses), 328 x
xxx
xx
showing a global increase of xxx xx
x xxx
x x
approximately 20 p.p.m. x xxxxx
CO2 (p.p.m.)

x x
xxx xx
between 1958 and 1976. The 324 xx
xxx
x
x
x xx
curves are smoothed, piecewise xxx x xxxxx
xxxxxx xxx
fits to the data. (Adapted from xxxxxxxxxx
x
320 x xxxx
Bacastow and Keeling1.) xxx x
x x
x xxxx xxx
x xx x
xxxxxxxx x
x
xx
316 x
x xxx
xx
x x xx
x
x

312

1958 1960 1962 1964 1966 1968 1970 1972 1974 1976
Year

183
 M. Varney

12.5 pCO2 (millibars) (at in situ temperature and 1 bar total pressure) sea water, after nitrogen and oxygen (see Chapter
0
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 11). Most gases dissolve in sea water in proportion
to their atmospheric partial pressures, but CO2 is
an exception. It is an extremely reactive gas and
has an elevated aquatic concentration relative to its
atmospheric partial pressure. CO 2 is intimately
1
involved in biological processes, being consumed
by plankton, which photosynthesise in the surface
waters, so leading to the production of organic
Depth (km)

matter tissue (equation 12.1):

2
CO2 + H2O →‘CH2O’ + O2 (12.1)

where ‘CH2O’ is a general term for organic plant

3 material. A consequence of primary production (see
Chapter 6) is that the upper waters of the ocean are
generally undersaturated in CO2 over large areas.
In contrast, upwelling of deep waters in the equato-
rial region and along the west coast of the
4
American continent, for instance (see Chapter 3),
brings water supersaturated with CO2 to the sur-
face. These waters are rich in CO2 because sinking
Figure 12.5 The typical variation of CO2 concentration organic matter has decomposed in the deeper
with depth. The profile shows a very sharp sub-surface waters of the ocean. The amount of the gas that is
increase to a maximum at approximately 500 m, due to dissolved in sea water is determined by an interplay
planktonic respiration processes which release the previ- of chemical, physical, and biological factors. In
ously fixed CO2 back to the water column. The profile turn, CO2 helps to maintain the acidity of sea water
was taken in the Eastern Pacific (28˚20’N 121˚41’W) dur- in the range of pH 8.0±0.5 (see Box 12.1).
ing the National Science Foundation GEOSECS The capacity of the ocean to absorb CO2 from
(Geochemical Ocean Sections) program in 1969. the atmosphere appears great (Figures 12.5–12.7),
(Adapted from Takahashi et al.10)
and ultimately it can dissolve orders of magnitude
more than is already present. Estimating the true
capacity is, however, difficult since nothing is at
equilibrium and the system is highly dynamic. CO2
forms carbonic acid with water, which then dissoci-
Box 12.1 The buffering ates to form hydrogen carbonate (HCO3–) and car-
capacity of sea water bonate (CO32–), which are the main forms of dis-
The pH of sea water varies over a surprisingly solved carbon in sea water. A simplistic view of the
narrow range, centred at pH 8±0.5. The most dissolution of CO2 into sea water is given in equa-
important point is that the dissociation of car- tion (12.4) [details are discussed later – for exam-
bonic acid (H2CO3) forms a buffering system, ple, see equations (12.6)–(12.9)].
which may be summarised by the general
weak acid–conjugate base equilibrium, equa- CO2(g) + H2O = H2CO3 = H+ + HCO3– = 2H+ + CO32–
tion (12.2), which results in a solution pH (12.4)
given by equation (12.3), where Ka is the equi-
librium constant for the dissociation reaction. Reactions (12.1) and (12.4) illustrate the buffer-
Small additions of acids or bases alter the
ing ability of the marine carbonate system. An
ratio of anion (HCO 3–) to acid (H 2CO 3) only
slightly, and have little effect on the pH of the increase in atmospheric CO 2 increases the total
solution. The buffering capacity is the extent amount of inorganic carbon within the sea. While
to which the pH is changed by a given addi- this increases the buffering capacity, it also induces
tion of acid or base. The higher the concentra- a slight increase in the ocean’s acidity and thus acts
tion of carbonic acid, the greater the buffering to oppose further entry of the gas.
capacity. CO2 concentrations increase with depth because
CO2 is used during photosynthesis and released
H2CO3 = H+ + HCO3– (12.2) again during respiration, and because the solubility
of CO2 increases with pressure (Figure 12.5). When
pH = pKa + log([HCO3–]/[H2CO3]) (12.3)

184
 12: The Marine Carbonate System

Figure 12.6 Total alkalinity, Saturation with calcite (%) 12.6

A, total dissolved inorganic
30 50 70 90 110 130 150 170
carbon (DIC) concentration, 0 ▲
and degree of calcite satura- ▲
tion as a function of depth in ▲

the equatorial Atlantic. Also ▲

indicated are the lysocline 1

(see Chapter 11) at approxi-
mately 4600 m depth, where ▲

there is a perceptible ▲
2
amount of calcium carbon- ▲
ate dissolution, and the com-
pensation depth at approxi- ▲
Depth (km)

mately 4900 m, below 3 Saturation depth DIC

which all calcium carbonate ▲

should be dissolved. Since A

▲
calcite dissolves at deeper
depths than it should 4 ▲
according to the calcite satu-
ration index, this implies that ▲

plant and animal remains Lysocline depth

▲
(which contain calcite) sink 5 Compensation depth ▲
faster than they can be dis- ▲
▲
solved. (Adapted from
Edmond and Gieskes3.)
6
2.1 2.2 2.3 2.4
A or DIC (milliequivalents/kg)

Figure 12.7 The biogeo- 0.048–0.18 12.7

chemical cycle of carbon. Primary
magmatic gases
The numbers show the cur- Atmosphere
rent estimates of the major 748
l

l
(mostly CO2)
reservoirs (in units of l
1015 gC) and fluxes (in units
l

l
l

of 1015 gC/yr). BP = transport Land use and Decay Solution Solution

of carbon to the deep sea by deforestation Photosynthesis
0.6–2.6 100–120 Respiration 50–60 70 22
the ‘biological pump’; PS = 40–60
Degassing Degassing
l

conversion of dissolved inor- 80 35

420–830 Terrestrial
ganic carbon (DIC) into par- biota
ticulate organic carbon
l

(POC) by photosynthesis; DIC Warm surface Cold surface

1200–1600 Soil and l
water DIC
l
water DIC
DOC = dissolved organic detritus rivers 600 Surface 300
Fossil fuel 0.5 circulation
carbon; RESP = conversion 5–6 18 RESP l 9 RESP
of organic carbon into DIC
l

l
Marine 20 PS l Marine
10 PS biota 1
by respiration. (Adapted biota 2 l
l
l

from Post et al.8 and Moore 2 BP 37 Eddy 1 Eddy 37 BP and

diffusion diffusion formation
and Bolin7.)
l

of deep
Intermediate and deep water DIC water
33,000–37,000 masses
Uplift l DOC 1000
and l
erosion Sinking and Detrital POC 30 Sinking and
remineralization remineralization
0.5 sedimentation
l

0.048–0.18 ?
Sediments
l
l

50,000,000 10,000,000
Organic
Limestone
and
dolomite 5000
Fossil fuels
Volcanoes

185
 M. Varney

12.8 350 various industrial processes (e.g., cement manufac-

Atmospheric CO2 concentration

ture), forest fires, and changes in agricultural and

300
land use patterns (Figure 12.8).
Experimental Measurement
(p.p.m.)

250 of the Carbonate Species

It would be useful to predict what the capacity of
South Pole
200 Siple (Arctic) sea water might be to absorb some of the CO 2
Mauna Loa
increase in the atmosphere, and the likely conse-
150
quences for terrestrial and marine life. It would
0 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 also be useful to quantify all elements within the
Year carbonate system using ship-board measurements
at a minimum of experimental effort and cost.
Figure 12.8 Extrapolated annually-averaged atmospheric Curiously, despite the crucial importance of these
levels of CO2 at locations away from major terrestrial measurements, no instruments have yet been devel-
influences. Concentrations prior to 1960 are estimated
from ice-core samples; post-1960, they are derived from
oped to make them in situ. To interpret the marine
data such as that in Figure 12.3. (Adapted from Post et carbonate system quantitatively we must first cal-
al.8) culate the contributions of the individual compo-
nents to the carbonate cycle.

CO2(aq) + CO32– = 2HCO3– (12.5)

photosynthesis occurs, CO2 is stripped from the CO2(g) = CO2(aq) (12.6)
water – the pH then increases so the carbonate con- H2O + CO2(aq) = H2CO3 (12.7)
centration ([CO32–]) increases. CO2 is generated by
H2CO3 = H+ + HCO3– (12.8)
the oxidation of decaying organic matter in sub-
surface waters [i.e., equation (12.1) moves to the HCO3– = H+ + CO32– (12.9)
left], this production being accompanied by a fall in
pH, the loss of dissolved oxygen, and the liberation Each reaction in the carbonate system can be
of nutrients by the material. The death and conse- characterised by an equilibrium constant (which
quent sinking of phyto- and zooplankton detritus varies as a function of salinity, pressure, and tem-
(see Chapter 7), the dissolution or precipitation of perature). Carbonate, hydrogen carbonate, and dis-
calcium carbonate (see Figure 12.6), and the forma- solved CO2 show considerable variations in con-
tion of limestone are now, and have been in the centration (spatially and temporally) which can be
past, inherently linked in the sea water carbonate related to biological, chemical, and physical
system. oceanographic processes. The solubility of the gas
The global cycle of carbon is complex; it is diffi- in its aqueous phase, CO2(aq), is enhanced by reac-
cult to explain adequately the processes, scales, and tion with carbonate ions, equation (12.5). The
magnitudes involved. Figure 12.7 shows some of position of equilibrium for this reaction (the equi-
the more important reservoirs and pathways, of librium constant, K) is so far to the right of this
which little is fully understood. Biogenic limestone equation that most of the CO2 entering the ocean is
is the single largest crustal reservoir of inorganic rapidly converted into hydrogen carbonate, HCO3–.
carbon. Sedimentary organic matter is the second In fact, the above chemical reaction proceeds in the
largest, and is composed primarily of inorganic car- laboratory as fast as the solutions can be mixed.
bon fixed by marine photosynthesis. However, The major steps in the sea water carbonate equilib-
while these two reservoirs may be the ultimate sink ria are given in equations (12.6)–(12.9), where
for the atmospheric increase in CO2, they have an equation (12.6) is the air–sea exchange of atmos-
effect only on geological time-scales. Geological pheric (g) to oceanic (aq) CO2, equation (12.7) is
uplift will eventually expose these ancient marine hydration, equation (12.8) is the first ionisation,
deposits to the atmosphere. Chemical weathering and equation (12.9) is the second ionisation.
will oxidise the carbon back to gaseous CO2, which
will be taken up by plants, thereby closing the glob- K0 = [H2CO3]/[CO2(aq)] (12.10)
al biogeochemical cycle of carbon. The concentra- K1 = [H+][HCO3–]/[CO2(aq)] (12.11)
tion of CO 2 in the atmosphere is about 0.03% K2 = [H+][CO32–]/[HCO3–] (12.12)
(dependent on sampling height and locality). A
long-term increase is evident, due to the release of Unfortunately, it is not possible to measure
(juvenile) CO2 from the Earth, the domestic and directly the concentrations of the two most impor-
industrial burning of fossil fuels (carbon previously tant species, [HCO3–] and [CO32–], without inadver-
fixed by marine phytoplankton and land plants), tently shifting the above equilibria to one side or

186
 12: The Marine Carbonate System

Box 12.2 Definitions of alkalinity and dissolved inorganic carbon

The alkalinity, At, of sea water is the combined negative charge due to hydrogen carbonate (HCO3–)
and carbonate ions (CO32–), expressed in molal concentrations. Alkalinity is based on sea water being
electrically neutral and is determined by titration (and therefore given the subscript t). In practice, the
alkalinity is the amount of acid (hydrogen ions) needed to convert all the anions back into their respec-
tive un-ionised acids. The alkalinity equation expresses electroneutrality between the positively and
negatively charged ions, equation (12.13), which can be rearranged to give equation (12.14), where
[SA], the surplus alkalinity, is the concentration of all weak acids other than boric and carbonic acids.
The second and third ionisation constants of boric acid are very low, so that boric acid is the major
contributor to At from the borate system. In sea water, ([OH–] – [H+]) and the amount of SA are small,
so equation (12.14) may be simplified, to give equation (12.15). SA cannot be ignored in anoxic
regions, when sulphide, ammonia, and phosphate concentrations may be very high.
Typical values for the alkalinity of sea water are given Table 12.1. The carbonate alkalinity, CA, is
the contribution to alkalinity from carbonate and hydrogen carbonate, equation (12.16), and is normal-
ly obtained by subtracting the borate contribution from At. The borate contribution, [B(OH)4–], is
obtained from the expressions for total boron, BT, and the equilibrium constant for boric acid dissocia-
tion, KB, equations (12.17)–(12.19). These equations are also a function of salinity, pressure, and tem-
perature. The borate contribution is usually 3–5% of CA.
The total dissolved inorganic carbon concentration, DIC, is defined as the concentration of all of the
dissolved forms, and is given by equation (12.20). Note that the terms ([HCO3–] and [CO32–]) refer to the
total equilibrium concentrations and include contributions from ion pairing with Na+, Ca2+, Mg2+, etc.
(see Box 12.3). There are other weak acids and bases that are potential contributors toward the
alkalinity of sea water, but under typical conditions their concentrations are extremely small; thus,
their contributions are usually much less than 1% of At and can be ignored in the above calculations.
(In the oceanographic literature, total alkalinity is sometimes referred to as TA, or simply A; carbonate
alkalinity as AC; total carbonate as TCO2, TIC, or -CO2.)
At the pH of most ocean waters, less than 1% of the inorganic carbon exists as ([CO2(aq)] + [H2CO3]),
and the H2CO3 concentration is only about 0.2% of that of CO2(aq).
It is important to remember the definition of alkalinity as the concentration of hydrogen carbonate
and carbonate ions. It is not a measure of the pH or how alkaline sea water is. Once this is recognised,
it is easy to appreciate that sea water alkalinity and acidity change in the same direction: where the
DIC is high, so are alkalinity and acidity (low pH); conversely, where the DIC is low, so are alkalinity
and acidity.
One of the few places where inorganic precipitation of calcium carbonate occurs is on the Bahamas
Banks, where the sea is shallow and warm, and salinity is high (exceeding a salinity of 37). The
warmer and more saline the water, the lower the solubility of gases, including CO2. The concentration
of CO32– is also large, and often rises sufficiently for the water to be supersaturated with respect to
CaCO3, so that the inhibiting effect of the MgCO3 ion-pair is overcome, and small crystals of calcium
carbonate (in the form of aragonite) are precipitated. In these conditions, the term (At – DIC) is large.
This helps to explain the apparent contradiction that when the DIC is high, calcium carbonate is more
likely to dissolve, and vice versa.

At + [H+] = [HCO3–] + 2[CO32–] + [OH–] + [B(OH)4–] + [SA] (12.13)

Table 12.1. Typical concen-
At = [HCO3–] + 2[CO32–] + ([OH–] – [H+]) + [B(OH)4–] + [SA] (12.14) trations of carbonate
parameters in the ocean.
At = [HCO3–] + 2[CO32–] + [B(OH)4–] (12.15)
Parameter Typical oceanic
CA = [HCO3–] + 2[CO32–] (12.16) concentration

BT = [H3BO3] + [B(OH)4–] (12.17) pH 7.8–8.4

PCO2 2.0–13.0 x 10–6 mol/kg
KB = {[H+][B(OH)4–]}/[H3BO3] (12.18) AT 2.3–2.6 x 10–3 mol/kg

[B(OH)4–] = KBBT/(KB + [H+]) (12.19)

DIC = [HCO3–] + [CO32–] + [CO2] (12.20)

187
M. Varney

Box 12.3 Effect of ion-pair formation on the carbonate system

Oceanic surface waters are supersaturated almost everywhere with respect to calcium carbonate –
otherwise many marine molluscs and crustaceans would not be able to form shells or hard exoskele-
tons. The possibility arises that the anthropogenic CO2 being pumped into the atmosphere can be
bound into the carbonate coral growths. This also raises the question of why spontaneous inorganic
precipitation of calcium carbonate happens only infrequently. The reason lies in the inhibiting pres-
ence of Mg2+ ions. Much of the carbonate is in the form of MgCO3 ion pairs. It requires the intervention
of marine organisms to precipitate calcite or aragonite, which have the same chemical formula but dif-
ferent chemical structures. Aragonite is thermodynamically less stable than calcite, so aragonite dis-
solves more readily.
Ion-pair formation is usually neglected in the formulation of carbonate equilibria. It occurs princi-
pally between the cations Na+, Ca2+, and Mg2+ (the ‘hard’ cations with an inert gas electron structure),
and anions such as F–, HCO3–, CO32–, and SO42–. We do not consider the pairs NaSO4–, MgSO4, and
CaSO4, although their formation is a considerable percentage of the total metal concentration (Table
12.2).
Borate forms ion pairs with NaB(OH)4, CaB(OH)4–, and MgB(OH)4–; approximately 44% of the total
borate concentration can be associated in this way. Although most of the carbonate and hydrogen car-
bonate are associated with Na+, Ca2+, and Mg2+, they also associate with other ions that occur at lower
concentrations in sea water (such as Pb2+, Cu2+, and Zn2+).
The principal effect is to consume an amount of the ‘free’ carbonate and hydrogen carbonate in
solution, i.e. equations (12.21)–(12.23). This implies there are a number of ion pairs, and each ion will
associate differently. The existence of ion-pair formation is another reason why the apparent ionisa-
tion constants are dependent on ionic composition (not just ionic strength). Many estimates for the
various formation constants are based either on thermodynamic theory (and extrapolation from a few
measurements), or on actual potentiometric titration measurements using, e.g., ion-selective
electrodes and pressure rigs.

[Total species] = [free ions] + [ion pairs] (12.21)

Table 12.2. Typical proportions of ion
[HCO3–]T = [HCO3–]f + [M.HCO3] (12.22)
pairs that various chemical species form
[CO32–]T = [CO32–]f + [M.CO3] (12.23) in sea water.
Chemical species Percentage as ‘free’ ion

Free SO42– 39
Na2SO4 37
MgSO4 20
CaSO4 4

the other. Instead, indirect measurements are made include pH, total CO2, alkalinity, or dissolved inor-
for which we either have existing methods or ganic carbon (see Box 12.2), and the partial pres-
instruments (such as pH electrodes, or infra-red gas sure of CO2 in equilibrium with the solution. In the
analysers). [HCO 3 – ] and [CO 3 2– ] can only be carbonate system, it is not possible to infer the
inferred through algebraic rearrangement of equa- component concentrations from one measurement
tions (12.6)–(12.9). Steps (12.7)–(12.9) have associ- alone, such as pH. Instead, pairs of measurements
ated balance points (equilibrium constants) in their of different variables are made and the calculations
reactions, equations (12.10)–(12.12). The equilibri- manipulated to derive the appropriate concentra-
um constants, K0, K1, and K2 are functions of salin- tions. Extreme care must be taken to choose the
ity, temperature, and pressure, and increase with all correct methods and values, the experimental pro-
three. Analytically, the equilibrium constants are cedures used for the various definitions, their style
the heart of an algebraic definition of the carbonate of calibration, and other practicalities such as ionic
system. composition, to avoid sources of error and possible
The usual (carbonate-related) quantities mea- confusion when handling the complex algebra asso-
sured in samples of sea water collected at sea ciated with the carbonate system.

188
 12: The Marine Carbonate System

The theoretical development of the determina- 6 12.9

Fossil-fuel emissions (Gt/yr)

tion of carbonate species has been problematic,
much of it because the measurement methods are 5
sensitive to apparent changes in concentration
(chemists refer to this as the activity of species). In 4
general, the activities of dissolved ions change due
to the differing composition of sea water around 3
the globe and with depth. The carbonate equilibri-
um constants almost always need conversion for 2
the salinity of the sample. In addition, all compo-
nents in the carbonate system change their distribu- 1
tion (their position of equilibrium) with change in
depth, due to the changes in the partial molar vol- 0
umes (sea water is compressed by approximately 1860 1890 1920 1950 1980
4% at 10 km depth). There is (unfortunately) little Year
reliable partial molar volume data in the literature
to enable theoretical prediction of the effects. Figure 12.9 The estimated fossil-fuel emissions of carbon
Alkalinity and CO2 can be regarded as conservative (in Gt/yr, 1015 gC/yr) from 1860 to recent. (Adapted from
properties when the pressure changes. Alkalinity Marland et al.6)
therefore varies directly with salinity, and is some-
times expressed as specific alkalinity when correct-
ed to a constant salinity (often to a salinity of 35). absorbed by the oceans, but the fate of the remain-
The approach taken is a practical one: experimen- ing 30% is not known. An increase in the terrestri-
tal methods have been developed that can be used al biomass (plants, trees, etc.) is thought to be the
on board ships, with no or few theoretical correc- missing sink, but the estimates in the flux calcula-
tions necessary. tions are subject to large errors. Figure 12.10 sug-
gests that, overall, the ocean absorbs about 1.6
Can we Predict the Absorption of gigatonnes (Gt) of the 5.3 Gt of fossil fuel derived
Atmospheric CO2 by the Ocean? carbon emitted each year. As the annual atmos-
Based on recorded changes in atmospheric compo- pheric increase is 3.0 Gt, this leaves 0.7 Gt unac-
sitions (Figure 12.8) and the output from fossil-fuel counted for. This figure is about 4% of the net
emissions (Figure 12.9), it appears that 50% of all annual terrestrial primary production, or approxi-
anthropogenic emissions of CO 2 remain in the mately 3% of the total CO2 transferred from the
atmosphere. About 70% of the rest has been atmosphere to the ocean.

–0.08
80˚N 80˚N 12.10
–0.28
60˚ 60˚

+0.12 +0.03
–0.16
40˚ 40˚

–0.13 +0.16 –0.23

20˚ 20˚
+0.06

0˚ +0.06 +0.09 +1.00 +0.28 0˚

20˚ 20˚
–0.24 –0.56 –0.11
40˚ 40˚
–1.14
–0.07
60˚ 60˚
–0.35
80˚S 80˚S

Figure 12.10 Averaged annual air–sea exchange of CO2 across the sea surface (in units of 1012 gC/yr; orange shades
indicate where CO2 evades seawater, blue shade indicates net sinks). (Adapted from Takahashi10.)

189
M. Varney

Box 12.4 Normal atmospheric equilibrium concentration

(NAEC) model
When the rates of gaseous molecular transfer into and out of the sea surface are equal, the gases are
at equilibrium; that is, the air and water concentrations are constant over time, but the gas molecules
are freely and reversibly transported backward and forward across the air–sea interface. If one side of
the interface has a higher or lower concentration of the gas, so that there is no longer an equilibrium,
there will be a net transfer of gas molecules to counteract that difference. For example, the average
partial pressure of CO2 in surface water at 0°C is 15.2 Pa (150 μatm). Since this is less than its atmos-
pheric partial pressure, 35.2 Pa (348 μatm), surface sea water at 0°C experiences an average influx of
CO2 from the atmosphere.
The degree to which any gas exerts a ‘pressure’ in water can be derived through the solubility rela-
tionships, based upon Henry’s Law. The gaseous exchange can be represented by equilibrium (12.24),
where CO2(g) and CO2(aq) represent the gaseous and aqueous concentrations of CO2 molecules. The
position of the equilibrium between the gaseous and aqueous phases is given by the equilibrium con-
stant K, equation (12.25).
All gas concentrations are more usually expressed by their partial pressure, P, related to the gas
law, PV = nRT, where V is volume, n is the number of gas molecules in that volume, R is the gas con-
stant, and T is the temperature in degrees Kelvin. Therefore, the atmospheric concentration of CO2,
[CO2(g)], is given by equation (12.26), which is substituted into equation (12.25), to give the concentra-
tion of CO2 in water, [CO2(aq)], equation (12.27).
K/RT is known as the Henry’s law constant, KH. As KH is inversely related to temperature, gas solu-
bility decreases with increasing temperature. Gas solubility also decreases with increasing salinity and
hydrostatic pressure. KH changes nonlinearly with salinity and temperature, and therefore a number of
multiparametric empirical formulae have been developed to compute the concentration of the gas in
sea water.
The equilibrium concentration is the gas concentration that would be attained if the water body
were allowed to come to equilibrium at its in-situ temperature and salinity. This is termed the Normal
Atmospheric Equilibrium Concentration (NAEC) model.

CO2(g) = CO2(aq) (12.24)

K = [CO2(aq)]/[CO2(g)] (12.25)

[CO2(g)] = nV = PCO2/RT (12.26)

[CO2(aq)] = KPCO2/RT = KHPTCO2 (12.27)

The surface of the ocean undergoes gaseous and the dissolved gases are carried to deeper levels
exchange with the atmosphere. The direction of mainly by turbulent mixing. The seasonal or main
the exchange depends on the relative temperatures, thermocline that separates the upper and lower
and the difference in [CO2], between the water and water columns provides an impediment to vertical
the overlying air mass. If the sea surface waters mixing. The upper water column can freely
have a lower partial pressure of CO2 than that in exchange gases with the atmosphere. The equili-
the atmosphere, then the gas moves from the bration steps involve a number of processes, most-
atmosphere into the sea. This gaseous invasion ly physical mixing of water masses, and some oth-
continues until either the partial pressures equalise ers, such as diffusion across the air–sea interface
(the position of normal atmospheric equilibrium and hydration of the gas, which are slow and pos-
concentration, NAEC – see Box 12.4) or the sibly rate-determining.
watermass sinks below the mixed layer before Air–sea exchange tends to be a kinetically limit-
reaching equilibrium. The rates of invasion and ed process compared to other processes that affect
escape (evasion) are influenced by meteorological the partial pressure of CO2 in the surface waters.
conditions, surface waves and films, and contami- The upper 100 m mixed layer of the ocean reaches
nants in the sea surface. Diffusion of gases across equilibrium with the atmosphere in approximately
the air–sea interface increases in stormy weather, one year. At high latitudes, warm water from the

190
 12: The Marine Carbonate System

80˚ 80˚N 12.11

60˚ –90 L 60˚
–60
–30
40˚ –45 –60 L 0 –30 40˚
–30 L
0 –15 0 –60
15 0 –15
–30
20˚ –30 L 30
20˚
0 H
H 45 15 0 0
30 30 30 15 0
15 30 60 30 –15 L 0
0˚ 15 75 0˚
30 60 90 H 90 15 30 H
15 0 30
0 –15 0 60
0 15 30
20˚ –15 –15 H 15 15 20˚
30
–30 L –30 –30 –15
L
40˚ –30 –15 L 0 –45
–15 15 0 15 –15 –30 40˚
0 –30 –30
0
60˚ 0 15 60˚

80˚S 80˚S

Figure 12.11 The averaged annual differences in the partial pressure of CO2 between surface sea water and the over-
lying atmosphere (p.p.m.). Positive values indicate that sea water is supersaturated and negative values indicate that it
is undersaturated with respect to CO2. (Adapted from Keeling4.)

equator carried north by the oceanic circulation (POM) and calcium carbonate into deeper waters
becomes undersaturated in CO2 upon cooling, and (see Chapter 7). As a result of this ‘biological
absorbs CO 2 from the atmosphere. The rate of pump’, which moves carbon from surface waters to
invasion is enhanced through photosynthetic reduc- deep waters, the bottom waters contain so much
tion of CO2, especially in sub-polar waters, where CO2 that they are supersaturated on average by
nutrients are generally high and productivity is about 30% relative to the NAEC. As the dissolved
stimulated. It is therefore rare for surface waters to inorganic carbon (DIC) content of the oceans con-
be at equilibrium with the atmosphere (Figure tinues to rise, water masses that are supersaturated
12.11); equatorial waters tend to be supersaturat- with CaCO3 (e.g., calcite, aragonite, etc.) become
ed, and polar waters undersaturated. Indeed, it is undersaturated [equation (12.5)] – then the CaCO3
even possible to stimulate photosynthetic activity to in the sediments begins to dissolve. A doubling of
increase the ‘draw-down’ of CO2 (see Box 12.5, the present atmospheric CO 2 (g) concentration
and Figure 12.12). would increase the DIC by about 5–6%, and dou-
Most of the CO2 absorbed from the atmosphere ble the [H+] in surface waters (the effect would be
is probably confined within the upper water col- smaller in deeper waters due to mixing). A possible
umn or mixed layer near the ocean surface. The consequence is that the production of calcite and
limited mixing into deeper waters is the result of aragonite by planktonic organisms in near-surface
strong vertical stratification at high latitudes, where waters might be diminished as the water becomes
most of the anthropogenic CO2 enters the ocean. less supersaturated – plankton communities may
The exceptions are the deep water formation areas change.
(see Chapter 3) that receive wind-borne CO2. This In summary, although it is possible to measure
process is likely to have absorbed only a small the important components within the marine car-
amount of anthropogenic CO2 so far. There are bonate system, such measurements are analytically
clear differences in the atmospheric levels and sea- complicated and the theory is fraught with difficul-
sonal trends in CO2 between the hemispheres. The ties. Measurements are restricted to ‘spot’ samples
‘draw down’ of CO2 by the ocean is limited by that are made on board ship (or back in the labora-
wind patterns and is therefore probably confined to tory). Ships and measurement systems are very
some sea areas only, and may be accentuated dur- expensive, so the cost of even a simple experiment
ing certain periods of the year (such as during the can be prohibitively large. It follows that there are
winter when the upper water column is cooled and very few accurate records of geographical, spatial,
mixes to greater depths). or temporal changes in CO 2 in the ocean.
The capacity of the ocean to absorb CO 2 is Consequently, predictions are based on historical
greatly augmented by the downward mixing and evidence and have a high level of uncertainty. Under
sinking of biogenic particulate organic matter these circumstances, any data that we do collect

191
M. Varney

Box 12.5 The removal of atmospheric CO2 by iron enrichment of

surface waters
Little effort has been made to alter the rate of fossil-fuel consumption or land use – despite the social,
economic, and political impacts. This inactivity is partly due to the level of uncertainty regarding the
current predictions of environmental change that will be induced by the increasing atmospheric con-
centrations of CO2. It is also due to the perceived prohibitive costs of developing alternative energy
sources and systems. As a result, attention has focused on removing the excess of CO2 from the
atmosphere, rather than cutting its production at source.
This has led to a proposal that marine phytoplankton growth might be stimulated by fertilising sur-
face waters of the North East Pacific, the equatorial Pacific, and the Southern Ocean (an area of over
10% of the world’s ocean). These waters contain abundant nitrate and phosphate, but support an
unusually low biomass. The late John Martin believed that primary production in these nutrient-rich
waters is limited by iron. Laboratory (microscale) experiments had already convincingly shown that
nanomolar iron enrichments of high-nitrate, low-chlorophyll waters do, indeed, stimulate phytoplank-
ton growth and biomass12. The ‘bottle experiments’ have also been successfully repeated in waters
south of the Galapagos Islands (project ‘IronEx’ – see Figure12.12), although the results differed from
those expected. The observed changes in the partial pressure of CO2 and in nitrate, fluorescence, and
chlorophyll levels were considerably less than theory predicted, presumably due to various unquanti-
fied loss terms (the grazing by zooplankton fortunate enough to benefit from the increase in phyto-
plankton numbers, the export of organic carbon, etc.). However, mesoscale field experiments are
notoriously difficult to undertake and interpret, and contain many unexpected and unquantifiable
factors.
If the ‘iron hypothesis’ is correct, adding sufficient soluble iron to these waters should stimulate
enough primary productivity to consume one-third to one-half of the anthropogenic CO 2 flux.
Supporters of this hypothesis have described this as a rapid method for recreating sedimentary organ-
ic matter, to counterbalance the rate of fossil fuel destruction. However, in the extreme, it is also pos-
sible that sustained elevated levels of photosynthesis might also remove sufficient CO2 from the
atmosphere to induce periods of glaciation.

12.12
(a) (b)
Chl-a (mg/m3) PP (mgC/m3/day)
0 20 40 60 0 20 40 60
0

20
Depth (m)

40

Patch
60 In
Out
Plume
Downstream
Upstream
80

Figure 12.12 During project IronEx in November 1992, a patch of sea water was injected with iron, continu-
ously tracked, and various chemical and biological parameters recorded. This was a particularly tricky opera-
tion involving aircraft overflights, radio-drogue buoys, and special chemical tracers (SF6, for instance). The
ship spent a total of 10 days sampling the patch of ‘injected’ iron. Here are shown comparisons of (a) chloro-
phyll concentration (Chl-a) and (b) primary production (PP), both in and out of the patch, and both upstream
and downstream of the Galapagos Islands. There is a clear increase in both the chlorophyll concentrations
(standing biomass) and the primary production (current activity of the plankton). (Data courtesy of the US
JGOFS Steering Committee.)

192
 12: The Marine Carbonate System

Box 12.6 Dynamic or stable – a question of scale!

What is the time taken for a mass of water to re-establish carbonate equilibrium if the temperature
were to suddenly decrease 1°C from 21 to 20°C? At this temperature, the concentration of dissolved
CO2 drops by 4%. The original CO2 in solution associates with CO3– to form hydrogen carbonate ions in
order to accommodate the shift in the equilibrium concentrations of all species. CO2 has to invade sur-
face waters because the water is out of equilibrium with the atmospheric CO2 concentration. Using the
basic equations for CO2 equilibria, it is possible to calculate that the total inorganic carbon concentra-
tion would have to increase by 0.4%, or 8 x 10–6 mol/kg, assuming the original CO2 concentration was
2 x 10–3 mol/kg. If we assume that the thermocline (the main barrier to mixing with deeper waters) is at
100 m, then the immediate effect is to cause a CO2 deficiency of approximately 0.8 x 10–3 mol/kg total
inorganic carbon. Because of the physics of gas exchange and other meteorological factors, the time
taken to re-establish equilibrium with the atmosphere will be slightly longer than 1 year.
Quite a long time! In reality, as CO2 enters the ocean the deficit between the air and the sea
becomes less and the net flux across the interface falls exponentially, so it is difficult to calculate
exactly how long it would all take. The oceans change temperature seasonally (this ‘drop’ takes place
every six months). The mixing of waters between polar and equatorial regions happens over a longer
time scale (e.g., years), so the equilibration of air and sea never quite reaches completion – hence the
term dynamic!

have to be fully utilised, and we must make use of can find out more about the marine environment.
developments elsewhere to help in our research. This is a highly political and controversial area (see
Future research in this area must concentrate on the Figure 12.13, for example). Although the Earth may
development of sensors that can be deployed (over seem to be in some sort of homeostatic state of reg-
monthly and annual periods of time) on buoys or ulation (Gaia – Box 12.7) this should not lead us
submersible platforms. The present evidence sug- into a false sense of security. Short-term variations
gests that the marine environment interacts strongly in atmospheric CO 2 may still create significant
with the atmosphere in such a way that the climatic changes in meteorological conditions and global cli-
consequences of increasing levels of atmospheric mate. This is an area that we have only just begun
CO2 may be accurately modelled – provided that we to research, and in which we have much to learn.

12.13
Global temperature change
4 (5 year running mean)

Observed
3 Scenario A
Scenario B
Scenario C
6T (˚C)

Figure 12.13 Predicted temper-

ature increases, 6T, for three
different scenarios. In scenario
A, the current rate of CO2 emis- 1
sions is continued. In scenario
B, the emissions of CO 2 are 0.5
held at their present values.
Finally, in scenario C, the out- 0
put emissions of CO 2 are cut
such that atmospheric levels
stabilise by the year 2000 AD. 1960 1980 2000 2020 2040 2060
(Adapted from Libes5.) Year

193
M. Varney

Box 12.7 The Gaia hypothesis

The biosphere appears to counteract naturally the artificial increase in atmospheric CO2 by acting as a
sink for it, and so buffering the greenhouse effect. The principal cause of the progressive fall in the
ratio [CO 2]/[O 2] is biological activity, removing CO 2 and releasing O 2 during photosynthesis.
Relationships of this kind have led to the novel concept that the surface of our planet is actively main-
tained as a life-supporting environment by biological activity which acts as a feedback mechanism.
This is James Lovelock’s concept of the Gaia hypothesis, first proposed in the 1970s:
... without life’s interference, CO2 would accumulate in the air until dangerous levels might be reached.
It is the strong interaction of the geochemical cycles of the elements and the biosphere that regulates
the environment, and Gaia’s policy is always to turn existing conditions to its advantage. The bios-
phere actively maintains and controls the composition of the atmosphere so as to maintain an opti-
mum environment for life and self-perpetuation. The geological fact that atmospheric conditions have
remained practically the same over several tens or hundreds of thousands of years suggests that this
dynamic system has been really quite stable!

General References
Broecker, W.S. and Peng, T.H. (1982), Tracers in the Sea, 4. Keeling, C.D. (1968), J. Geophys. Res., 73, 4547.
Eldigio Press, New York, 690 pp. 5. Libes, S. (1991), Marine Biogeochemistry, Academic Press,
Butler, J.N. (1982), Carbon Dioxide Equilibria and Their New York, 734 pp.
Applications, Addison-Wesley, Reading, MA, 259 pp. 6. Marland, G.T., Boden, T.A., Griffin, R.C., Huan, S.F.,
Lovelock, J. (1991), Healing Gaia, Harmony Books, New York, Kancircuk, P., and Nelson, T.R. (1989), Historical
1992 pp. and Predicted Emmissions of Greenhouse Gases,
Skirrow, G. (1975), The dissolved gases – carbon dioxide, in ORNL/CDIAC-25, Oak Ridge National Laboratory,
Chemical Oceanography, Vol. 2, Riley, J.P. and Oak Ridge, TN.
Skirrow, G. (eds), Academic Press, London, Ch 9. 7. Moore, B. and Bolin, B. (1987), The marine carbonate cycle,
Oceanus, 29, 11.
References 8. Post, W.M., Peng, T.H., Emmanuel, W.R., King, A.W., Dale,
H., and DeAngelis, D.L. (1990), The biogeochemical
1. Bacastow, R.B. and Keeling, C.D. (1979), Models to predict cycle of carbon, Amer. Sci., 78, 314.
atmospheric CO2, in Workshop on the Global Effects 9. Schneider, S.H. (1989), The changing climate, Sci. Amer.,
of CO2 from Fossil Fuels, Elliot, W.P. and Machta, L. 261, 38–47.
(eds), Report CONF-770385, US Department of 10. Takahashi, T. (1989), The effect of the marine carbonate
Energy, Washington, DC. system on climate, Oceanus, 32, 29.
2. Coulson, K.L. (1975), Solar and Terrestrial Radiation, 11. Takahashi, T., Weiss, R.F., Culberson, C.H., Edmond, J.M.,
Academic Press, London. Hammond, D.E., Wong, C.S., Li, Y.-H., and
3. Edmond, J.M. and Gieskes, J.M.T.M. (1970), On the calcu- Bainbridge, A.E. (1970), Global effects of CO2 from
lation of the degree of saturation of sea water with fossil fuels, J. Geophys. Res., 75, 7648–7666.
respect to calcium carbonate under in situ conditions, 12. Wells, M. (1994), Pumping iron in the Pacific, Nature, 368,
Geochim. Cosmochim. Acta, 34, 1261–1291. 295.

194
CHAPTER 13:

A Walk on the Deep Side:

Animals in the Deep Sea
P.A. Tyler, A.L. Rice, C.M. Young, and A. Gebruk
And now, how can I retrace the impression left upon me by mals measuring millimetres at most and hidden
that walk under the waters? Words are impotent to relate from our eyes within the bottom muds beneath our
such wonders. feet. The richness of this tiny fauna became appar-
Professor Pierre Aronnax, in Jules Verne’s Twenty Thousand ent only in the 1960s, when the Americans,
Leagues under the Sea Howard Sanders and Robert Hessler, started sam-
pling the deep-sea floor with much finer nets and
Introduction recovered an amazing variety of small
Water is a totally alien environment for us humans. invertebrates5. A consequence of studies using sam-
Few of us do more than enter the sea and, as we pling apparatus to collect the smaller organisms is
swim, dip our heads briefly beneath the surface to the recent suggestion2 that the deep sea may have a
take a momentary glimpse of a blurred and strange species diversity equivalent to, if not greater than,
world. Even the best-equipped Self-Contained that of tropical rain forests (see also Chapter 15).
Underwater Breathing Apparatus (SCUBA) divers With this proviso in mind, let us plan our expedi-
can penetrate no more than a few tens of metres, tion. Although nets collect animals from the sea
for to go deeper requires highly specialised equip- floor, the use of the deep-sea camera has allowed
ment or very sophisticated and expensive sub- us to view this environment undisturbed. The pho-
mersibles. But imagine that, like Captain Nemo tographs presented in this chapter come either from
and the crew of Jules Verne’s Nautilus, we could cameras mounted on trawls, from ‘Bathysnap’ (see
walk freely across the ocean floor and let us review Chapter 7), or from cameras mounted on sub-
what we might experience if we took the long jour- mersibles, the last of which allows the operator to
ney across the Atlantic Ocean from Britain to the select the target.
Bahamas.
If an extra-terrestrial visitor to the Earth wanted Route Planning
to take back a collection of animals from its most Our route takes us from the intertidal (see Chapter
typical environment, he could do no better than 16), across the continental shelf to the west of the
sample the abyssal deep-sea floor across which our British Isles (Figure 13.1). As we cross the edge of
route takes us. For the world ocean covers 70% of the shelf at about 200 m depth, the slope of the sea
the planet’s surface and over 80% of it is more floor steepens perceptibly and the already dim
than 3000 m deep. Yet our own knowledge of downwelling light diminishes rapidly. By the time
what lives there is very recent. Although a few sam- we reach a depth of 300–400 m we are no longer
ples had been taken in relatively deep water in the able to detect any daylight from the surface and we
first half of the nineteenth century, serious study of enter a vast zone of perpetual darkness.
the deep-sea fauna began with the cruises of HMS Our path takes us down the gentle northeastern
Porcupine in 1868 and 18696, and received a fur- slope of the Porcupine Seabight, named after HMS
ther boost from the circumnavigation of HMS Porcupine, for this was one of the very first areas
Challenger from 1872–1876 (see Chapter 1). The of deep sea to be sampled. We head first south and
reports on the data collected from HMS Challenger then west where, at a depth of about 3500 m, the
filled 50 large volumes and included descriptions of Seabight opens out onto the Porcupine Abyssal
hundreds of new species. But because of the coarse Plain. A more direct route would have been down
nets routinely used from HMS Challenger, and for the eastern flank of the Seabight, but this area is
the next 80 years or so, most of the animals collect- riven with deep channels known as submarine
ed from the sea floor were relatively large and lived canyons. The fauna here is different from that on
on or just beneath the sediment surface, just the our route, but the difficulty of the terrain makes
sort that we are likely to see. But we should be sampling with surface-deployed gear hazardous so
aware that this is just a fraction of the deep-sea that this area, together with similar rough topogra-
fauna, for the vast majority is made up of tiny ani- phy in the world’s ocean, is best sampled from sub-

195
 P.A. Tyler, A.L. Rice, C.M. Young, and A. Gebruk

13.1 80˚ 70˚ 60˚ 50˚ 40˚ 30˚ 20˚ 10˚W 0˚

2 Porcupine
2
Bank 60˚
2

CANADA NORTH
GREAT
4 BRITAIN
Porcupine
50˚
Seabight
Porcupine
ATLANTIC Abyssal

E
DG
Plain
2

RI
40˚

C
OCEAN

TI
UNITED SPAIN

AN
STATES
4
L
Hatteras
AT
Abyssal 4
4 30˚
MID

Plain

2
N.W. AFRICA
4

Tongue of 20˚N
the Ocean 4

Figure 13.1 Chart of the route from the southwest of the British Isles to the Bahamas (compare with Figure 1.9).
Contours are shown at 2 km and 4 km depth.

mersibles. An even more precipitous route, which ocean ridge which now descends steadily to the
might be taken by a party of more adventurous Hatteras Abyssal Plain with a general depth, like
individuals, would be from the Porcupine Bank the Porcupine Abyssal Plain, of between 4000 and
westward into rapidly deepening water. This 5000 m. The ‘landscape’ here is also not very dif-
descent to the floor of the abyssal plain would be ferent, the monotony being broken only by the
almost like absailing down a cliff face, for this par- Bermuda Plateau, which just breaks through the
ticularly steep continental slope is typical of some surface at Bermuda itself.
thousands of kilometres off northwestern Europe The Hatteras Abyssal Plain extends all the way
and is one of the most spectacular geophysical fea- to the Bahamas and, as we reach the base of these
tures on the planet. islands, we enter a blind-ended channel, known as
Whichever route is taken, the Porcupine Abyssal the Tongue of the Ocean, running between New
Plain (Figure 13.1) must be crossed westward over Providence Island and Andros Island. We ascend a
a vast and almost flat expanse, gradually deepening steep slope and, because of the clarity of the water,
to almost 6000 m, but interrupted occasionally by we begin to perceive daylight at about 500 m depth,
small abyssal hills rising a few tens to a hundred or considerably deeper than where we last saw day-
two metres above the general sea floor. light on the other side of the ocean. There is no real
As we approach the middle of the ocean the sea continental shelf in this region and our path takes
bed starts to rise on the eastern flank of the mid- us steeply upslope to the coral reefs fringing the
Atlantic Ridge to reach the crest at about 2000 m. Bahama Islands, where we finally reach dry land.
Now we climb down into the narrow graben, or
axial valley, and walk westward past lava flows The View on the Route
and occasional hydrothermal vents (see Chapter The descent into the abyss
10). A second steep climb brings us out of the As we enter the sea our last view of dry land is the
graben and onto the western flank of the mid- rocky shore common on the west side of the British

196
 13: A Walk on the Deep Side: Animals in the Deep Sea

Figure 13.2 A rocky shore to 13.2

the west of the British Isles.
This is the last view of land as
we enter the sea on our route
to the west.

Isles (Figure 13.2). Our route takes us into steadily 21), and much richer than we find in the deep
deepening water. The sea bed is composed of sands ocean.
of varying grades that are commonly stirred up by Beyond the shelf edge, the upper part of the con-
the surge of storm waves breaking overhead or tinental slope on our main path descends gently
sculptured into ripples by the strong currents. The into the Porcupine Seabight. With increasing depth
fauna contains a wide variety of species, many of the soft sediment becomes finer, reflecting the
them recognisable to the amateur naturalist1. At decreasing influence of disturbance from the sur-
almost every step we are likely to see a fairly large face and the reduced strength of the bottom cur-
bottom-dwelling animal, perhaps a scallop or one rents. Large animals become less abundant, for the
of its molluscan relatives, a starfish or brittlestar bottom is further away from the primary source of
(sometimes as many as 300/m2), or a crab scuttling food, the thin surface layer which is lit well-enough
away to find shelter. We also disturb the occasional to support the growth of the phytoplankton.
flat-fish, such as plaice, sole, or halibut, while Nevertheless, there is plenty of evidence in the form
around us we quite often see shoals of cod, whit- of tracks, trails, burrow openings, and mounds that
ing, and herrings. For, despite the ravages of over- large animals are here. And as we start to lose the
fishing, the continental shelf still supports a fish animals with which we became familiar as we
resource exploited by local fishermen (see Chapter crossed the shelf, we begin to see new ones, the first
of the true deep-sea fauna.
13.3
One of the most obvious is the red crab, Geryon
trispinosus (Figure 13.3), a member of a family with
many representatives occupying much the same
depth zone in different parts of the ocean, some of
them abundant enough to support commercial fish-
eries. Geryon is clearly responsible for many of the
large burrows and mounds on this part of the slope.
At the upper slope depths, around 500–1000 m
deep, we also see strange plant-like colonies of the
sea-pens Kophobelemnon stelliferum and Pennatula

Figure 13.3 The Atlantic red crab Geryon trispinosus,

adopting an aggressive posture at 500 m depth in the
Porcupine Seabight.

197
 P.A. Tyler, A.L. Rice, C.M. Young, and A. Gebruk

13.4 Figure 13.4 The sea-pens

Kophobelemnon stelliferum
(right) and Pennatula aculeata
(left) anchored in soft sediment
at 500 m depth in the Porcupine
Seabight. The anemone-like
polyps of the 30 cm high K. stel-
liferum are clearly visible, but
on P. aculeata they occur as
small polyps on each leaf.

aculeata (Figure 13.4), colonial octocorals with rapidly until we are standing amongst large patch-
numerous anemone-like polyps. Many similar ani- es, tens or hundreds of metres across and contain-
mals occur in shallow water, and these outliers of ing many sponges on each square metre, like a
the deep-sea community represent a life-style, that great prickly carpet. As we carry on down the slope
of suspension feeding, which becomes less and less this amazing population rapidly becomes more
common with increasing depth, to be replaced by sparse until, by the time we reach a depth of
deposit feeding. Curiously, however, we are about 1300 m or so, it disappears completely, just as
to come across a remarkable abundance of animals abruptly as it had appeared at the upper limit of
by deep sea standards – and all suspension feeders. the sponges’ depth range. We do not know exactly
For as we approach the 1000 m isobath, we sud- why these sponges should be so abundant here, or
denly see patches of the hexactinellid sponge why they should be so restricted in their depth dis-
Pheronema carpenteri (Figure 13.5), spherical tribution, for in other areas they occur over a much
blobs the size of a large grapefruit, ‘rooted’ into the wider depth range. It may be a result of the local
bottom by a mass of long glass spicules and with a hydrography, a combination of internal waves and
large exhalant aperture in the top. Initially, we see the bottom slope resulting in an enhancement of
the odd Pheronema here and there, but as we move the near-bottom currents which, in turn, keep food
on a few hundreds of metres, increasing our depth material in suspension and therefore available to
by only 10–20 m, the population density builds up the sponges. The sponges also harbour a whole

13.5

Figure 13.5 A cluster of the sponge

Pheronema carpenteri. On the sponge
on the left can be seen the small squat
lobster Munida tenuimana. Phero-
nema occur over a wide depth range
in the Atlantic in general, but in the
Porcupine Seabight they are restricted
to a narrow band round 1200 m,
probably because of the local hydro-
graphic conditions (see text).

198
 13: A Walk on the Deep Side: Animals in the Deep Sea

Figure 13.6 The sea cucumber 13.6

Paelopatides gigantea, swim-
ming over the sea bed by gentle
undulations of its dorso-ventral-
ly flattened body at about
2000 m depth in the Porcupine
Seabight.

Figure 13.7 The sea cucumber 13.7

Benthogone rosea, walking
over the sediment surface at a
depth of about 1500 m on the
Goban Spur. The mouth is at
the right, and the small tenta-
cles can be seen collecting food
particles from the sea bed.

community of other animals, ranging from a wide elements, particularly the carcasses or moulted
range of small organisms living in or among the skins of small planktonic animals and their faecal
sponge spicules to the small squat lobster Munida pellets (see Chapter 7). Phytodetritus sinks sudden-
tenuimana (Figure 13.5) and the curious small spi- ly and rapidly to the sea floor over much of the
der crab Dorhynchus thomsoni. north Atlantic, following the spring phytoplankton
As we move down toward the 2000 m level we bloom, and thus introduces a seasonality to an oth-
come across occasional specimens of the holothuri- erwise monotonously stable environment. Like the
ans (sea cucumbers) Paelopatides gigantea and discovery of the richness of the small infaunal ani-
Benthogone rosea (Figures 13.6, 13.7). These ani- mals of the deep-sea floor, this seasonal deposition
mals, typically 15–30 cm long, are deposit feeders, of phytodetritus is a very recent discovery. Its exis-
ingesting sediment as they move across the sea tence was undreamt of prior to about 1980 and its
floor and extracting from it small particles of discovery proved that the deep sea is a much less
organic matter. They are representatives of one of constant environment than had been thought for
the most typical deep-sea groups, for holothurians more than a century4. Its arrival, and subsequent
are found throughout the deep ocean and many dif- very patchy distribution, may be one of the factors
ferent species are to be found. which results in the surprisingly high species diver-
As we pass from the Seabight and out onto the sity of the benthic fauna (see Chapter 15).
abyssal plain the general appearance of the sea bed The alternative route onto the abyssal plain,
depends upon the time of year. At most times it down the precipitous slope to the west of the
simply appears as a plain brownish-grey mud. But Porcupine Bank, has a quite different fauna. Here,
in the late spring or early summer, say during May in contrast to the soft sediments of the Seabight
or June, it looks markedly different. For then the proper, the slope off the Bank is mainly exposed
sea bed is covered in a flocculent brownish green rock, black with a manganese coating (see Chapter
‘gunge’, moved around by the near-bottom currents 21), and with sediment accumulating only in the
and therefore patchily distributed, being most con- occasional areas where the slope is gentle. The
centrated in depressions and on the down-current fauna on the rock surfaces is quite unlike that in
sides of mounds (Figures 13.8, 13.9). This material normal sediment-covered regions, for it is bathed
is phytodetritus, a complex cocktail based mainly by a northward-flowing current of 3–10 cm/s
on the dead cells of normally surface-dwelling phy- which is laden with food particles that support a
toplankton species, but also containing many other rich suspension-feeding community. Typical fauna

199
 P.A. Tyler, A.L. Rice, C.M. Young, and A. Gebruk

13.8 13.9

Figure 13.8 Fresh phytodetritus collecting in depressions Figure 13.9 ‘Old’ phytodetritus concentrated in biogenic
(about 20 cm in diameter), some of them feeding marks depressions (about 50 cm diameter) on the Porcupine
left by sea stars, at a depth of about 2000 m in the Abyssal Plain at a depth of 4850 m in September 1989.
Porcupine Seabight in May 1982. Phytodetritus consists The patchiness produced by this phenomenon may at
largely of dead and dying phytoplankton cells which least partly explain the high biodiversity of the deep-sea
picks up a variety of other particles as it sinks through the floor communities.
water column (Chapter 7). Much of it is recycled during
its downward journey, but a significant proportion reach-
es the deep-sea bed as a seasonal pulse and forms an
important food source for the bottom-living animals, in
this case the sea-urchin Echinus affinis.

include sponges such as Euplectella sp and the Porphyrocrinus thallassae are the most spectacular
pompom-shaped sponge Crateromorpha (Figure examples. A similar life-style is adopted by the
13.10). The cnidarians are richly represented here brisingid sea stars, such as Brisingella multicostata,
by gorgonians, the most dramatic being the bottle- which sit atop rocks in this zone to benefit from the
brush Thouarella sp., Iridiogorgia sp., and the fan- accelerated flow over topographical highs (Figure
shaped Paramuriceidae (Figure 13.11). 13.13).
The last of these orientates normal to the pre- The fauna of this part of the slope can be used
vailing current, allowing the maximum cross-sec- to determine current direction (Figure 13.14). Non-
tion for particle filtration. Filter feeding is also random orientation of the crinoid Porphyrocrinus
found in the crinoids, the dominant echinoderm thallassae and the downstream winnowing of sedi-
group in this habitat, of which the bright yellow ment behind a glacial erratic indicate the flow of
Anachalyspicrinus nefertiti (Figure 13.12) and Northeast Atlantic Deep Water along this slope8.

13.10

Figure 13.10 The sponge Crateromorpha at

about 2120 m depth to the west of the
Porcupine Bank. The main filtering apparatus
is the pompom at the distal end of the stalk,
some 20 cm above the rock surface.

200
 13: A Walk on the Deep Side: Animals in the Deep Sea

13.11 13.12

Figure 13.11 A branched gorgonian (ca. 1 m high) Figure 13.12 The brilliant yellow crinoid
belonging to the family Paramureicidae at a depth of Anachalypsicrinus nefertiti at 2480 m depth to the west
2800 m on the western slope of the Porcupine Bank. The of the Porcupine Bank. The star-shaped part of the animal
small anemone-like polyps found along each branch fil- filters particles from the water column, the stalk keeping
ter particles from the water column. These colonies, this filtering apparatus well above the sea bed. The ani-
which may be a metre high, are orientated at right angles mal can be up to 25 cm high.
to the current so the broadest face of the colony faces the
current, thus maximising particle capture.

13.13 13.14

Figure 13.13 Three individuals of the multi-armed sea Figure 13.14 Individuals of the bright red crinoid
star Brisingella multicostata resting on a small rocky Porphyrocrinus thallassae attached to the side of a block
hillock at 2890 m depth to the west of the Porcupine of rock at 2310 m depth to the west of the Porcupine
Bank. The arms are extended into the water column to Bank. The crinoids are all orientated in the same way,
trap small animals and particles brought past the rock by showing that the current is from the right.
the current.

201
 P.A. Tyler, A.L. Rice, C.M. Young, and A. Gebruk

13.15 Figure 13.15 Small macrofauna

from the abyssal plain. This is
the residue left on the sieve after
most of the fine sediment has
been removed. The main
macrofauna seen in this sample
are bivalve mollusca (B) and the
cumacean crustaceans (C). The
B small spherical structures are
protozoans. (Courtesy of Dr J.D.
Gage, Scottish Association for
Marine Science, Oban,
Scotland.)

C
1mm

The abyssal plain short of descending into one of the deep ocean
At the foot of this steep rocky slope there is an trenches. All the incoming organic matter, includ-
abrupt transition to a much flatter bottom, typical ing phytodetritus (see Chapter 7), has had to run
of the edge of the abyssal plain, with soft sediment the gauntlet of the mid-water animals in the
similar to that crossed on the main path at the 4–5 km of water above. Much of it has been recy-
mouth of the Seabight. The depth is about 4000 m cled in those communities, with the inevitable loss
here; although our path does take us deeper, we are of material at every step in the food web. So, above
no longer able to detect any significant slope. For all, the abyss is an environment where food is at a
we are now truly on the abyssal plain, with gradi- premium and nothing is wasted.
ents typically of the order of 1:1000 or less. Apart An obvious indicator of this state of affairs is the
from the occasional ghostly flashes of biolumines- relative paucity of the fauna. The hidden small ani-
cence (see Chapter 14), it is, of course, totally dark. mals living in the sediment (Figure 13.15) are about
It is also very cold, with the temperature constantly 100 times less abundant here than on the continen-
hovering around 2˚C. Finally, it is almost deathly tal shelf. For the bigger ones, the difference is near-
still compared with the shallow coastal waters. For, er to 1000! Curiously, despite the low animal
although the currents still have a clear tidal compo- abundances, we see more evidence of their activities
nent, they are relatively sluggish, rarely exceeding here than in shallower regions. A combination of
10 cm/s. Here, we are just about as far from the high current speeds, storm disturbance, and many
source of food as it is possible to be in the ocean, animal ‘feet’ trampling over the sea bed rapidly

13.16

Figure 13.16 An area of about 20 m2 of the

Porcupine Abyssal Plain at a depth of 4850 m,
which contains many small burrow openings,
probably made by worms and crustaceans.
The star-shaped marks (A) are the feeding
traces of echiuran worms, while the meander-
B
ing trail (B) was made by a sea cucumber.

202
 13: A Walk on the Deep Side: Animals in the Deep Sea

13.17 13.18a

13.18b

Figure 13.17 A pair of the hermaphroditic sea cucumber

Paroriza pallens at 1670 m depth on the Goban Spur;
such pairs are probably formed to ensure successful fertil-
ization between individuals of species with low popula-
tion densities.

destroys trails, faecal casts, and unoccupied bur-

rows and mounds in coastal and upper slope
waters. Here, in the tranquil abyss, these features
last much longer. So there are many burrow open-
ings in the mud, often with clear feeding lines radi-
ating from them like spokes in a wheel (Figure Figure 13.18 The sea cucumber Oneirophanta mutabilis
13.16). Many of these are made by echiuran at a depth of 4800 m on the Porcupine Abyssal Plain in
worms, clearly much more abundant than their September 1992. This is from a time-series (5 h intervals
occurrence in trawl and sledge catches would indi- between frames) obtained with the deployed camera
cate. Some of them are sizeable beasts, for the feed- ‘Bathysnap’ and also shows a sea anemone and the
ing traces made by their probosces are 2 m or more remains of that season’s phytodetritus.
across. We will also see many trails, a few centime-
tres across and meandering over the bottom for
tens of metres (Figure 13.16). There are several 13.19
quite distinct types, some being smooth grooves,
some consisting of parallel rows of separate pits,
and some much less regular, as if something has
simply ploughed through the bottom mud more or
less at random. Most of these trails are made by
holothurians, for the sea cucumbers are without
doubt the dominant invertebrate megafaunal group
in the deep sea, crawling slowly across the sea floor
and hoovering up the sediment to extract its small
organic content. The basic cucumber shape familiar
in shallow-water members of the group is repre-
sented here by species such as Paroriza pallens
(Figure 13.17), but other deep-sea representatives
modify the sediment in a bewildering variety of
ways. These include the elegant white
Oneirophanta mutabilis, with its short, peg-like
Figure 13.19 From the ‘Bathysnap’ sequence, at 4850 m
‘legs’ and long dorsal tentacles (Figure 13.18), and on the Porcupine Abyssal Plain, this time showing the
the large, purple Psychropotes longicauda, with its strange sailed sea cucumber Psychropotes longicauda.
strange caudal ‘sail’ (Figure 13.19). The function of the sail is unknown.

203
 P.A. Tyler, A.L. Rice, C.M. Young, and A. Gebruk

13.20a 13.21a

13.20b 13.21b

Figure 13.20 A Bathysnap sequence of the galatheid Figure 13.21 The penaeid shrimp Plesiopenaeus
crab Munidopsis crassa, moving away from the camera armatus, attracted to bait, in this case a mackerel
in a northerly direction at 4850 m depth on the wrapped in muslin, at a depth of about 4600 m on the
Porcupine Abyssal Plain. Maderira Abyssal Plain.

Although the holothurians are by far the most during our abyssal walk is the carcass of a dead
common large organisms likely to be seen, we also animal – and this is mostly because of the activities
occasionally come across other echinoderms, such of another crustacean group, the amphipods, the
as the starfish Styracaster or one of the stalked group to which sand-hoppers and freshwater
crinoids. We may also see the occasional large crus- shrimp belong. They are represented throughout
tacean, though true crabs are never found as deep the oceans, both on the bottom and in mid-water,
as this, few of them penetrating beyond a depth of where they adopt a wide range of life-styles. But on
about 2000 m. Here, in the abyss, their place is the deep-ocean floor they are the scavengers par
taken by species of the genus Munidopsis (Figure excellence. Food is at such a premium in the deep
13.20), which is well-represented in all the deep sea that any large lump, such as the carcass of a
oceans of the world. They lack a common name, fish, squid, or whale, arriving on the bottom is
but are related to the squat lobsters of shallow attacked very rapidly. The first to arrive are the
waters. But the most spectacular crustacean found amphipods, mostly only a few millimetres long, but
here is Plesiopenaeus armatus, an impressive bril- with some species reaching a respectable 10 cm or
liant red shrimp reaching a length of 30 cm or more. These voracious scavengers can reduce the
more (Figure 13.21), living generally on, or very carcass of a large fish to a well-picked skeleton in a
close to, the sea floor, but also an accomplished matter of hours. Little wonder, then, that we are
swimmer. Plesiopenaeus must be good to eat, but unlikely to see such a carcass. But if we did come
the difficulties of fishing it at these depths would across one before it had been completely destroyed
make its price prohibitive! it would be the centre of a frenzy of activity – or, at
One thing we almost certainly would not see least, what passes for a frenzy in this curious envi-

204
 13: A Walk on the Deep Side: Animals in the Deep Sea

ronment. For many animals will want their share of 13.22a

the spoils. Most, including echinoderms, molluscs,
and crawling crustaceans, move in relatively slowly
to pick up any morsels left by the amphipods. The
bonanza also attracts faster moving animals, on the
Porcupine Abyssal Plain particularly the rat-tail or
grenadier Nematonurus armatus, very widely dis-
tributed in the Atlantic at depths below about
2000 m (Figure 13.22). Curiously, if a fish carcass
arrives on the deep-sea floor much further south in
the eastern Atlantic, on the Madeira Abyssal Plain
(see Figure 9.9), it is likely to attract large numbers
of Plesiopenaeus (Figure 13.21) and rather few
Nematonurus, partly because the relative abun-
dance of the two species is reversed in the two
localities.
If carcasses are short-lived and therefore rare in 13.22b
the deep sea, evidence of human activity is both
abundant and long lasting. From the results of
trawling in the abyssal north Atlantic, it is clear
that the sea floor is littered with all manner of rub-
bish discarded from ships overhead. Such dumping
is now banned by all the main maritime nations
(see Chapter 22), but for hundreds of years it was
the accepted way of disposing of waste at sea. As a
result, bottles, cans, bits of pottery, and synthetic
materials of all sorts, ranging from polythene sheets
and bags, through netting and rope of man-made
fibres, to cocktail sticks from the great days of pas-
senger liners, are present. Although we might be
lucky enough to come across a fascinating artefact
from the days of sail, most of this was jettisoned Figure 13.22 Rat-tail fish, Nematonurus armatus, attract-
during the past 30–40 years. The most common ed to bait at a depth of 4850 m on the Porcupine Abyssal
sea-floor artefact of all is almost entirely attribut- Plain.
able to the century between about 1850 and 1950.
During this period, when most ocean-going vessels
were powered by coal-fired steam engines, the
unburned residues from the fireboxes were routine- Ridge, rock again protrudes through the mud. In
ly dumped over the side as clinker, ranging from contrast to the steep rocky slope off the Porcupine
small fragments no larger than the tip of your fin- Bank, these rocks have little or no fauna, for there
ger to great crusty lumps a metre or more across. is almost no suitable food in this region. The bar-
Like the other ship-borne rubbish, clinker is natu- renness of the sea bed continues as we climb over
rally concentrated beneath the main shipping the crest of rock into the median valley of the ridge
routes. But deep-sea biologists hardly ever take a system, contrasting with the almost unbelievable
trawl or dredge sample, even in the remotest parts luxuriance of the rich populations in the few hun-
of the ocean, that does not contain at least a few dreds of square metres surrounding the occasional
small fragments of clinker. So it seems as though, hydrothermal vents that we may pass. Unlike all
in little more than 100 years, mankind managed to the animals seen so far, these amazing vent commu-
affect, albeit only slightly, virtually every square nities are not dependent upon food material pro-
kilometre of the largest and most remote environ- duced by photosynthesis in the overlying surface
ment on Earth – a sobering thought as we continue layers, but are supported instead by a quite differ-
our way toward the very centre of the ocean. ent system based on the vents themselves. The
superheated water gushing from the vents is laden
The Mid-Ocean Ridge with reduced inorganic compounds, particularly
As the rocks underlying the sea floor are younger sulphides, that are used as an energy source by spe-
nearer the mid-ocean ridge (see Chapter 8), the sed- cialised anaerobic bacteria to build organic carbon
imentary veneer becomes thinner as we travel west- molecules on which the rest of the vent community
ward. Eventually, on the flanks of the Mid-Atlantic is dependent (see Chapter 10). These oases in the

205
 P.A. Tyler, A.L. Rice, C.M. Young, and A. Gebruk

13.23 Figure 13.23 A swarm of the shrimp Rimicaris exoculata

at a hydrothermal vent at 3680 m on the Mid-Atlantic
Ridge. Interspersed within the swarm of Rimicaris are the
vent crab Segonzacia and another, less abundant, but
large and yellow shrimp, Chorocaris chacei.

ly laden ‘smokers’ often deposit their chemical bur-

den as they emerge, to form enormous natural
‘chimneys’ many metres high (see Chapter 10).
These sights would be amazing enough, but it is
the biology that puts the final touch to the experi-
ence. Sometimes it is difficult to see the vent water
because of the seething mass of life surrounding it.
The shrimps here are similar to those seen earlier,
but probably belong to a different species. If they
seemed like individual bees before, now we have an
enormous swarm. Hundreds of thousands of
shrimps mill around the vent (Figure 13.23), so
close-packed that it is almost impossible to follow
the movements of any individual. What on Earth
are they doing? Well, recent research by American,
Russian, and French scientists has pieced together
an amazing story of adaptation to this remarkable
environment. First, in at least one of the shrimp
species the eyes seem to have become modified to
‘see’ not light, but heat, for they appear to be sensi-
tive to the infra-red emissions from the hot vent-
water (see Chapter 14). So they can perhaps detect
sea have been found in all the major oceans, sur- a vent from some distance away. This is important,
rounded by communities that are hundreds of times because although the vent water emerges at very
richer in terms of biomass than those on the neigh- high temperatures, frequently in excess of 200˚C, it
bouring sea floor. Many vent species, ranging from cools rapidly in contact with the cold abyssal water
bacteria to fish, are found nowhere else, but each so that, no more than a metre or two from a vent
individual vent community has a relatively low opening, it is virtually impossible to locate it from
species diversity, being dominated by only one or temperature alone. Next, the mouthparts of the
two species which may occur in enormous shrimp are highly modified for a very un-shrimp-
numbers. like diet of bacteria, which grow on the mouthparts
So what are we likely to see as we approach a themselves, on the surface of the carapace, and on
mid-Atlantic vent? Well, long before the vent itself, the lining of the gill chamber; and the mouthparts
the animals with which we have already become can reach and scrape off the bacteria from all of
familiar become much more abundant. Then, as we these areas. Finally, the shrimps’ behaviour keeps
get to within 20–30 m of the vent, we begin to see them in just the right conditions of temperature
new and unfamiliar animals, including snails, and water chemistry for their bacterial ‘gardens’ to
decapods, and anemones. The decapods include grow most rapidly. That is what all the jostling is
orange–white shrimps, 5 cm or so long, that swim about.
around like bees on a summer’s day (Figure 13.23). As we wonder at this incredible sight, we should
Although the general abundance of life is dramati- remember that the vent shrimps, and all the other
cally greater than seen so far, it does not prepare us animals which are associated specifically with
for life at the vents themselves. As our lights finally vents, have a serious problem that most animals in
pick out the vents a few metres ahead, there is a ‘normal’ environments never have to face. For the
sudden magnificent abundance of life. At the centre vents are ephemeral, apparently lasting for no more
is the vent water, emerging from fissures in the than a few decades before the flow of life-support-
rocks as columns rising far above our heads, either ing, hot, chemical-laden water suddenly slows
glass-clear and shimmering as the temperature con- down and stops. The adult animals around the vent
trast refracts the light from our torches, or belching site at this time are doomed. So all the time they
forth like white or black smoke because of the must be making arrangements for the survival of
heavy load of particles carried with it. These heavi- their species by ensuring that their eggs or larvae

206
 13: A Walk on the Deep Side: Animals in the Deep Sea

are being cast to the watery winds so that they can 13.24
colonise any new vents which suddenly start up in
the same general area. This is one of the big puzzles
of vent biology, and one which oceanographers are
actively pursuing7.
Return to the surface
Left with this, and other fascinating questions
about vents in our minds, we climb out of the
median valley, cross the western rim of the Mid-
Atlantic Ridge and descend along the rocky west-
ern flanks. The climb across the Mid-Atlantic
Ridge has been very long and rugged (approximate-
ly equivalent to a walk from Edinburgh to Rome,
but with Alpine terrain the whole way) so the flat
terrain and soft sediment we encounter on the Figure 13.24 The giant isopod Bathynomus giganteus at
desert-like Hatteras Abyssal Plain is a welcome 840 m depth off the Bahamas. In contrast to its terrestrial
change. The fauna here, like that of the Porcupine cousins, this species can grow up to about 15 cm long.
Abyssal Plain crossed earlier, is composed mainly
of small animals buried in the mud, with only occa- turns south and becomes a dead-end ‘box’ canyon,
sional large megafaunal animals crawling across surrounded completely by islands and shallow
the surface (see Heezen and Hollister 3 ). The banks, and known as the Tongue of the Ocean.
Hatteras Abyssal Plain is the largest spatial entity Here we begin our very steep ascent toward the
in the north Atlantic. Its monotony is broken only coral reefs and sea-grass meadows of the tropical
by the pinnacle that rises to become the Island of western Atlantic.
Bermuda and by some slight undulations that pass As we hike up the slope, an isopod crawls by.
for abyssal hills. If we stroll to the far southwest Small isopods are so common and diverse in the
corner of this giant plain, we encounter a deep deep sea that a single individual would not normal-
canyon-like extension that cuts into the shelf sup- ly catch our eye. But this one is not like the gribble
porting the Bahamas Archipelago. At the point of that consume our pier pilings or even like the pill-
entry, between Abaco Island on the north and bugs or woodlice that we find behind our couches
Eleuthera on the south, this canyon is known as and in our gardens. This lumbering giant,
Northeast Providence Channel, but it eventually Bathynomus giganteus (Figure 13.24) is some
15 cm long. We do not know why a few isopods
from the Antarctic and the deep sea display gigan-
13.25 tism, but we do know that these giants have been
around since before the dinosaurs. Indeed, during
the Cretaceous age, when the White Cliffs of Dover
were being deposited, giant isopods would proba-
bly have been encountered anywhere along our
trans-Atlantic stroll; today, they are limited to a
narrow depth range in the Gulf of Mexico and
Caribbean region, indicating that the sea bottom
has changed substantially over the intervening
aeons.
At a depth of 900 m there is what appears to be
a bright reddish-purple jellyfish, pulsating just
above the bottom. As we approach, however, we
find this beast to be very un-jellyfish-like in its
anatomy. Whereas a jellyfish would have its mouth
on the underside and surrounded by long stinging
tentacles, this animal has its mouth on the top and
surround by tentacles with the appearance of tiny
florets of a cauliflower. The strange animal is a sea
Figure 13.25 The swimming holothurian Enypniastes cucumber, Enypniastes eximia (Figure 13.25). Like
eximia. This brilliant red species swims above the sea its more sedentary relatives, it feeds by shovelling
bed, descending to the bottom to feed using the cauli- sediment into its mouth on the sea floor, but unlike
flower-like tentacles. other cucumbers, its body wall is not thick and

207
 P.A. Tyler, A.L. Rice, C.M. Young, and A. Gebruk

13.26 Figure 13.26 The multi-armed

sea star Novodinea antillensis in
a typical feeding posture at
660 m depth off the Bahamas.
Food, such as small crustaceans,
is trapped by tiny pincers on the
arms and moved to the mouth
by the tube feet. The brilliant
colour is typical of brisingid sea
stars (see also Figure 13.13).

leathery. If we reach out and touch this animal, we form a velcro-like surface capable of entrapping
find the body is soft and delicate. It can be dam- any small objects unfortunate enough to encounter
aged beyond recognition by only a moderate poke it. When we brush the starfish arm with our own, it
of the finger. immediately grasps our hairs in hundreds of places.
On a nearby rock outcrop there is a brilliant red We have no trouble escaping, but a small crus-
starfish resembling Brisingella, which we saw on tacean whose legs are trapped is in a much worse
the slope west of Ireland. Novodinea antillensis position. Once captured, the prey is enclosed in a
(Figure 13.26), the Bahamian version, is perched on loop of the arm, grasped with the tube feet, and
the highest available portion of the bottom, where moved directly to the gaping mouth on the central
it has access to the currents that pass by. Its posture underside of the disk.
is that of a filter-feeder, but close examination of Novodinea is not the only animal here that gives
its long arms reveals it to be a formidable predator us a feeling of déjà vu. As we look at the stalked
of euphausiids and other swimming crustaceans. crinoids, the gorgonians, the hexactinellid sponges,
The entire body is covered with tiny pincers which and the other animals populating this rock outcrop,

13.27

Figure 13.27 The crowns of six indi-

viduals of the crinoid Endoxocrinus
at 640 m depth off Egg Island in the
Bahamas. Some of the arms are
raised to be flicked rapidly down-
ward, a mechanism which apparent-
ly they use to dislodge annoying
crustaceans. Compare the number of
arms seen on these individuals with
the five seen on the species in Figure
13.12.

208
 13: A Walk on the Deep Side: Animals in the Deep Sea

it is apparent that, given the same substratum and 13.28

depth, faunas may be very similar in widely sepa-
rated parts of the world ocean. The names of the
players may be different on the two sides of the
Atlantic, but the roles they play are very much the
same.
Like flowers struggling toward the light in a
shaded garden, stalked crinoids cluster along the
ridges where currents are faster, thereby bringing
more food to the waiting arms and pinnules (Figure
13.27). Each crinoid bends with the currents in
exactly the same way as its neighbours; all have
arms poised in a parabolic fan to entrap tiny parti-
cles in thick mucus. Occasionally, we see one of
these arms abandon its normal posture to move up
and down rapidly. It appears that the crinoid is
waving to us, but a closer look reveals the fallacy
of this interpretation. Small crustaceans occasional-
ly dart toward the feeding groove on the oral side
Figure 13.28 The sea-urchin Stylocidaris lineata on a
of the arm, possibly to steal some of the concen- rocky outcrop at 500 m depth off the Bahamas. During
trated food that the crinoids have accumulated in their reproductive season individuals of this species come
their mucus. The crinoid responds to this irritation together in pairs for spawning to aid their fertilisation
by flicking the offender away, much like a horse success.
swishing flies with its tail.
On the soft sediment near this rock outcrop is caribaeum) carries odd bits of broken shell and
one of the most common sea-urchins of this region, other refuse around the margin of its shell.
Stylocidaris lineata (Figure 13.28). If it is spring- By 450 m depth, our dark-adapted eyes start to
time, probably each individual will be paired with perceive a greyish glow of light from the surface.
another of its kind. Sea-urchins do not mate, but Two large sea-urchins catch our eye. The first,
instead cast their eggs and sperm into the sea, Calocidaris micans [Figure 13.29(a)], is the largest
where fertilisation occurs. Normally, these deep-sea sea-urchin we have seen anywhere on the trip and
animals are so far apart that egg–sperm encounters the second, Coelopleurus floridanus [Figure
are unlikely. They pair in the spring to increase the 13.29(b)], is the most beautiful. The spines of
odds that sperm will find eggs to fertilise. A pure Calocidaris are perfectly straight and have the
white snail living on this same slope (Tugurium consistency and lustre of ivory elephant tusks.

13.29a 13.29b

Figure 13.29 Protection in the deep sea. (a) The sea-urchin Calocidaris micans at 260 m on sediment off the Bahamas;
the straight spines can be up to 20 cm long. (b) The striped sea-urchin Coelopleurus floridanus, at a depth of 450 m on
sediment in the Tongue of the Ocean. The black, white, and red shell, and the elegant curve of the spines, make this a
most striking animal.

209
 P.A. Tyler, A.L. Rice, C.M. Young, and A. Gebruk

13.30 Figure 13.30 The slit shell Perotrochus midas, on the

underside of an overhang at 650 m depth off the
Bahamas. The slit along the side of the shell gives the
name, and the species is much prized by collectors.

Spines of Coelopleurus, by contrast, are tapering, leave the deep sea and enter the much better under-
red and white in colour, and flare upward from the stood world of the tropical coral reef (Figure 13.31).
colourful body in graceful arcs. Our journey has taken us thousands of kilome-
The terrain becomes ever steeper and, in most tres horizontally and through 5 km of depth. It has
places, climbing equipment is required to scale the been equivalent to a march across North America,
coral cliffs. As the light increases, sponges become complete with the Rocky Mountains and Great
larger, more colourful, and more diverse. There are Plains in the middle. During our stroll, we have
also more and more soft corals, antipatharians, and focused on those animals that live out in the open
gorgonians. Comatulid crinoids and brittle stars and are large enough to see easily with the naked
perch on large glass sponges to elevate their feeding eye. We may have formed the impression that the
structures in the flow. Basket stars use large gor- deep sea contains relatively few species compared
gonians for the same purposes. Here, on the steep- with a rain forest or a coral reef.
est portions of the cliff, there is a large snail with a However, our impression is biased by a failure
shell that appears to be made of pure gold. This slit to examine the little creatures dwelling below the
shell, Perotrochus midas (Figure 13.30), is prized surface of the mud and the diverse creatures occu-
by collectors, but finds refuge in this rugged envi- pying the deep water column itself. If we were to
ronment where divers cannot reach and dredges count up all of the species, both seen and unseen,
become hopelessly entangled. that we passed on our journey, our list would rival
As we climb up the last piece of the cliff, plants the species diversity of a tropical rain forest (see
appear and the light becomes strong enough to Chapter 15). The deep sea is at once the largest,
support symbiotic algae that provide much of the most diverse, and least understood environment on
nourishment for hard corals. Colours explode as we Earth.

13.31

Figure 13.31 Typical coral reefs

in the shallow sunlit waters of
the Bahamas. A coral knoll is
surrounded by gorgonians relat-
ed to, but distinct from, those in
Figure 13.11.

210
 13: A Walk on the Deep Side: Animals in the Deep Sea

General References 2. Grassle, J.F. and Maciolek, N. (1990), Deep-sea species rich-
ness: regional and local diversity estimates from quan-
Campbell, A.C. (1994), Guide to the Fauna of Intertidal and titative bottom samples, Amer. Natural., 139,
Shallow Seas, Hamlyn, London, 320 pp. 313–341.
Gage, J.D. and Tyler, P.A. (1991), Deep Sea Biology: a natural 3. Heezen, B.C. and Hollister, C.D. (1971), The Face of the
history of organisms living at the deep-sea floor, Deep, Oxford University Press, New York, 659 pp.
Cambridge University Press, Cambridge, 504 pp. 4. Rice, A.L. (in press), Changing views on the biology of the
Grassle, J.F. and Maciolek, N. (1990), Deep-sea species rich- deep sea floor: from sterility to unprecedented biodi-
ness: regional and local diversity estimates from quan- versity, Port Erin Marine Laboratory Centenary
titative bottom samples, Amer. Natural., 139, Volume.
313–341. 5. Sanders, H.L., Hessler, R.R., and Hampson, G.R. (1965), An
Heezen, B.C. and Hollister, C.D. (1971), The Face of the Deep, introduction to the study of the deep-sea benthic fau-
Oxford University Press, New York, 659 pp. nal assemblages along the Gay Head-Bermuda tran-
Thomson, C.W. (1873), The Depths of the Sea, MacMillan, sect, Deep-Sea Res., 12, 845–867.
London, 527 pp. 6. Thomson, C.W. (1873), The Depths of the Sea, MacMillan,
London, 527 pp.
7. Tunnicliffe, V. (1991), The biology of hydrothermal vents:
ecology and evolution, Oceanogr. Mar. Biol.: Ann.
Rev., 29, 319–407.
8. Tyler, P.A. and Zibrowius, H. (1992), Submersible observa-
References tions of the invertebrate fauna on the continental
1. Campbell, A.C. (1994), Guide to the Fauna of Intertidal and slope southwest of Ireland (NE Atlantic Ocean),
Shallow Seas, Hamlyn, London, 320 pp Oceanol. Acta, 15, 211–226.

211
CHAPTER 14:

Light, Colour, and Vision

in the Ocean
P.J. Herring

Introduction noisy, and disconcertingly mobile space, but with

The reflected surface colour of the ocean, whether the advantage of immediate access to captured ani-
seen from the deck of a ship or measured by an mals. It may be on shore, if the technical difficulty
orbiting satellite, provides information about the of maintaining the particularly delicate deep-sea
contents, and processes, immediately below. Many species can be overcome. Such studies have shown
of these near-surface processes, ranging from the that the eyes of deep-sea animals are irreparably
absorption of red light by chlorophyll to the multi- damaged by brief exposure to daylight (or a sub-
hued visual signalling of reef fish, are dependent on mersible’s floodlight) – so special cold and dark
colour. Interactions deep below the surface are also trawl buckets (or cod-ends) have had to be devel-
affected by colour; indeed, the colours of the ani- oped for experiments on the vision of these
mals caught in a research trawl give a good indica- animals.
tion of the depth from which they have come. The brief account below draws on all these
‘Colour’ itself is a rather subjective concept; its per- methods for its data, but previous experience indi-
ception is critically dependent upon both the light cates that our present interpretations will not do
conditions of the environment and the visual sys- full justice to the extraordinary abilities of the
tems of the observer. A more objective assessment deep-sea fauna.
requires knowledge of the spectral distributions of
ambient, reflected, and absorbed light11. Characteristics of Light in the Sea
Like most research tasks, understanding the role Light in the sea derives from two sources, the Sun
of light and colour in the sea has not been straight- and the organisms. Sunlight (and reflected moon-
forward. Ironically, many of the life activities of light) illuminates at most only the upper kilometre
deep-sea animals have been assumed from the of the ocean. Bioluminescence occurs at all depths.
anatomy of dead specimens. This can make for The intensity distribution of sunlight at the sea sur-
some spectacular mistakes (such as earlier interpre- face varies by only a factor of about two over the
tations of light-emitting organs as eyes or ears), but visible range of 400–700 nm (Figure 14.1).
in most cases the conclusions are probably reason- Moonlight and starlight intensities are, respectively,
ably close to the truth. However, it is salutary to about 6 and 9 log units lower than sunlight, while
remember that illustrations of the deep-sea fauna, the intensities of bioluminescent sources can
portraying fish and other animals oriented horizon- approach that of moonlight2.
tally, do so more by convention than by informa- Light below the surface differs from that above
tion. One of the great advances has been the recent in both quality and quantity. A major change
use of remote vehicles, cameras, and manned sub- occurs at the air–sea interface; here, the combined
mersibles to see what life in the deep sea really consequences of reflection from the surface (at
looks like (or, at least, looks like when floodlit!). A glancing angles of incidence), and refraction into it,
surprising number of fish, for example, routinely reduce the angular distribution of light entering
stand on their heads or their tails. Swimming brittle calm water to a narrow cone of solid angle 97˚. A
stars, luminous swimming sea cucumbers, metres- fish looking upward sees the world above the sur-
long siphonophores coiled in huge spirals, squids face through this limited window (Snell’s window).
that shed their tails; these are but a few of the All it sees outside this window is back-scattered
many unexpected behaviours that have been light from deeper water (by total internal reflection
observed. at the air–sea interface11). Any initial asymmetry in
Much more about the interactions of light and the radiance distribution within the window (due
animals can be learned only by careful experiments to the Sun’s angle, for example) is rapidly degraded
in the laboratory on perfect specimens. The labora- with depth, through the effects of scattering by
tory may be on a ship, with all the limitations and both water molecules and suspended particles. Ice
discomforts attendant on working in a cramped, cover acts as a spectrally neutral diffuser and reflec-

212
 14: Light, Colour, and Vision in the Ocean

Figure 14.1 Light (b) 14.1

at depth differs (a)
markedly from that
at the surface. (a) 120
The narrow spec- s
tral distribution of 100 0

light at a depth of
about 500 m in 80
Intensity (%)
clear ocean water
(d), compared with 60 d
that of sun- and
sky-light above the 40
surface (s). (b) The
symmetrical angu- 20
lar distribution of
light in the ocean 0
about the vertical 400 450 500 550 600 650 700
axis. The length of Wavelength (nm)
the arrows indi-
cates the relative
radiance (intensity per unit solid angle) from point O in the respective directions. The radiance is maximum vertically
downward and halves at 35˚ to the vertical. The vertically downward:upward ratio (D:U) in oceanic waters is about
200:1 (after Denton2).

tor to produce a uniform, but dimmer, light field at Colour

the surface. Even in open water the light field soon
becomes symmetrical about the vertical axis, with a Shallow water
downward radiance some 200 times the back-scat- Life in the upper layers of the ocean is dominated
tered upward radiance2 (Figure 14.1). At increasing by daylight. In the near surface layers the spectral
depths this angular distribution of radiance remains consequences of absorption and scattering are rela-
unchanged, but scattering and absorption reduce tively minor and the animals are exposed to a
the light intensity in an exponential way. Both bright light containing all wavelengths. At the mar-
these processes are wavelength selective. Water gins of the oceans this is associated with a struc-
absorption preferentially removes both long (red) turally complex background of shore or reef within
and short (ultraviolet) wavelengths, rapidly result- (or against which) individuals can hide. The bright
ing in near-monochromatic blue light, which even colours of many reef fish are responses to this com-
in the clearest oceanic waters is reduced by 90% plex optical (and biological) environment. They
for every 70 m of depth; this means that moonlight allow their owners to display and signal to each
makes the same contribution to ambient light at other with the relative immunity of a safe haven of
400 m as does sunlight at 800 m. retreat. The body colours are effective only because
In coastal waters dissolved yellow material the ambient light contains all wavelengths at suffi-
derived from plant decay (both marine and from cient intensity for different colours to be selectively
river input) may absorb additional short wave- reflected. They also presuppose that the observers
lengths and result in a greener hue to both the have sufficient variety of visual pigments to allow
downwelling and back-scattered light. Scattering by them to discriminate the colours, or at least to
particles of very small size is inversely proportional recognise their contrasts.
to the fourth power of the wavelength of the light
(1/h4); blue light at 470 nm is scattered five times Open ocean
more than red light at 700 nm. Larger particles In the open ocean the animals are faced with the
scatter all wavelengths more evenly (1/h). These same illumination, but with the radical difference
characteristics determine the light conditions in the that the background is uniform. There is no com-
sea and the resulting adaptations of the organisms. plexity of pattern or topography within which to
The biological imperative for every marine animal hide. Bright colours could be used to send the same
is to survive and reproduce; this requires the ability messages as in shallow waters, but, with nowhere to
to find food and (usually) a mate, and to avoid or escape to, these beacons would become a dangerous
deter predators. Sensory systems such as vision, liability. Disappearance is possible only by matching
and effector systems such as bioluminescence and the background radiance. Animals living right at the
colour, provide some of the means whereby this is air–sea interface, and vulnerable to predators from
achieved. above, are frequently blue in colour, matching the

213
 P.J. Herring

14.2 14.3

Figure 14.2 The 40 mm diameter coelenterate Porpita Figure 14.3 This 10 mm long surface living shrimp,
floats at the surface, where it is camouflaged by a blue Hippolyte coerulescens, is camouflaged by a blue
carotenoprotein pigment. carotenoprotein pigment and white reflective dorsal
chromatophores. (Courtesy of the Southampton
Oceanography Centre, England.)

14.4 upward scattered radiance of clear ocean waters.

The colour is achieved in different ways by different
species. Blue carotenoprotein or biliprotein pig-
ments are commonly used (Figures 14.2, 14.3). Blue
structural colours (selective diffuse or specular
reflection) are other means of achieving the same
result (Figure 14.4). The deep blue colour with
which many upper ocean fish camouflage their dor-
sal surfaces has a similar structural basis.

Transparency and silvering

For smaller organisms in the upper waters, camou-
flage can be achieved by transparency. Many
planktonic species, particularly the gelatinous
forms, rely on this for their protection (Figure
14.5). For larger animals the complexity of the
body tissues renders transparency impracticable.
However, in the particular light environment of
open water (with the brightest light from vertically
above and a symmetrical radiance distribution
about this axis), a mirror stood vertically in the
water becomes invisible from any angle of side
view. Many animals have taken advantage of this
to mimic transparency by turning themselves into
the equivalent of vertical mirrors. This is best

Figure 14.4 Blue colour can be achieved without pig-

ment. In the isopod Idothea metallica (12 mm long), tiny
reflective particles in the upper structure reflect blue light
much better than red. The transmitted red is then absorbed
by a dark pigment beneath the reflective layer. (Courtesy of
the Southampton Oceanography Centre, England.)

214
 14: Light, Colour, and Vision in the Ocean

14.5a 14.5b

Figure 14.5 Many jellyfish are often almost invisible by virtue of their transparency. Typical examples are (a) the
comb jelly Beroe (60 mm diameter) and (b) the medusa Aequorea (25 mm diameter). (Courtesy of Image-
Quest 3-D.)

achieved by flattening the body, so that the flanks 100% reflectance is achieved with only 5–10 crys-
are vertical, and covering these with reflective tals. Without such spacing, five crystals would have
material. Fish are consummate examples of this a reflectance of only 20%.
strategy, none more so than the hatchet fish (Figure One potential drawback of the system is that the
14.6). For their reflective material they use tiny best-reflected wavelengths shift toward the blue
crystals of guanine (an excretory product derived end of the spectrum as the angle of viewing
from nucleic acids), which are aligned parallel to becomes more oblique.
the body surface. The crystals are arranged in mul- Effective silveriness requires reflection of all
tiple stacks, alternating crystal and cytoplasm (the wavelengths, at all angles of view. To achieve this,
watery matrix of the cell), whose spacing is such as the spacing of the stacks is adjusted so that either
to achieve constructive interference reflection. For a different colours are reflected from different stacks
particular wavelength of light, h, viewed at right (which may be adjacent or superimposed on one
angles to the stack of crystals, ideal interference another), or the spacing within the stacks varies in
reflection occurs 1 when the optical thickness of a regular way. Vertical flanks are not compatible
each layer is 0.25h. With such a system, almost with a muscular stream-lined body, so most fish

14.6

Figure 14.6 In the radially sym-

metrical light distribution in the
ocean (Figure 14.1), a vertical mir-
ror is invisible from the side. These
hatchet fish (Argyropelecus,
25–50 mm long) have turned
themselves into mirrors by flatten-
ing their sides and silvering them,
using stacks of guanine platelets as
interference reflectors.

215
 P.J. Herring

14.7 Figure 14.7 Where some organs remain opaque they can
still be camouflaged by silvering them separately, as is
the case for the liver and eyes of the 40 mm long squid
Cranchia scabra.

Even when much of the body is transparent,

there may be particular tissues that remain opaque
(e.g., eyes, red muscle, or digestive organs). These
organs can still be silvered individually to achieve
effective camouflage. Squid use the same strategy,
but employ reflective platelets of protein rather
than guanine crystals (Figure 14.7).
The concealment value of vertical reflective sur-
faces rapidly disappears as the surfaces are tilted.
This property can be used to good effect to distract
predators (e.g., the flashing of a twisting school of
fleeing sardines or the eponymous silversides).
Changes in body orientation can also be used to send
optical signals to nearby members of the school.
The dramatic colour changes visible in some
oceanic fish (e.g., the coruscating colours of a cap-
tured dolphin fish, Coryphaena) are brought about
by very rapid changes in the spacing of the crystals
in the reflecting cells. Each stack of crystals behaves
like venetian blinds as contractile elements in the
cells tilt the individual crystals, altering their dis-
tance apart and hence their reflected colour. Rapid
colour changes, under similar control, take place in
the reflective stripe of the freshwater neon tetra and
in some damselfish.

Deep-water colours
If reflection is the saviour of animals in reasonably
have a more elliptical cross-section. If the reflecting well-lit open water, it spells potential disaster for
stacks in the skin or scales remain vertically orient- those in the dark of the ocean depths. Here, the
ed, despite the curved body surface, the effect of a reflection of a bioluminescent flash or glow could
vertical mirror is retained1. break the cover of an animal hitherto invisible

14.8 14.9

Figure 14.8 The uniform scarlet colour of the 70 mm long Figure 14.9 The black melanin pigment of this typical
deep-living shrimp Acanthephyra purpurea is due to a deep-sea fish (Gonostoma bathyphilum, 110 mm long)
carotenoid pigment that absorbs any incident blue light. plays the same camouflage role as does the scarlet pig-
We see it as scarlet in the white light of the camera flash; it ment of the shrimp. (Courtesy of the Southampton
is invisible in the deep sea, where blue bioluminescence is Oceanography Centre, England.)
the norm and most predators only have blue-sensitive
eyes. (Courtesy of the Southampton Oceanography Centre,
England.)

216
 14: Light, Colour, and Vision in the Ocean

14.10 Figure 14.10 In midwater, where light

from the surface is still important and
day–night changes are substantial, the
limited colouring of the 50 mm long
shrimp Sergestes is mostly distributed in
large dorsal chromatophores. This
allows the animal to change its appear-
ance according to the light environment.

against the black background. Thus, silvering is an are now grey, pale, or even white. There is no obvi-
anathema in deeper water, and is replaced by uni- ous rationale for this change, except that in any
form matt colours of brown, purple, black, or scar- bioluminescent light the paler species present a less-
let (Figures 14.8, 14.9). This is a particularly clear er contrast when seen against the lighter sediment
example of the functional equivalence of subjective- than do their heavily pigmented pelagic relatives
ly very different colours. Dramatically different as only a few tens of metres above (Figure 14.11).
these colours may appear to us in the sunlight on Many of the animals on the bottom have large sen-
deck, they are all equally effective at preventing the sitive eyes, so vision clearly still plays an important
reflection of any residual dim blue light filtering role in this environment.
down from the surface, or of stray flashes of blue Many oceanic animals undertake substantial
bioluminescent light. As the bioluminescence may vertical migrations during their lifetimes. In gener-
come from any direction, the colouring is spread al, the larvae and juveniles live at shallower depths
over the whole of the animal. than the adults, and thus experience gradual
At intermediate depths, where daylight from the changes in the light conditions as they move deeper
surface is dim but still significant, and day–night in the water column. Their appearance at any given
intensity changes are still important, a compromise stage of development reflects the depth at which
is reached in which animals have some element of they are living. Early shrimp larvae near the surface
colour and some of silvering or transparency. This may be transparent, the juveniles at mid-depths
enables them to adjust their colouring quite half-red, and the adults at depth uniformly scarlet.
markedly in response to changes in light intensity. Colour is clearly a key feature in the life of oceanic
Fish silvered in daylight have mobile dark pigment animals, but its appearance to the denizens of the
cells which disguise their silveriness at night. deep is not always quite what it seems to the
Shrimps have a ‘half-red’ appearance, in which the human eye in sunlight.
red pigment is present in large pigment cells, and
are able to disperse or aggregate the colour as
appropriate to the light conditions (Figure 14.10).
In both cases, the pigment distribution is primarily 14.11
dorsal, in response to the continued dominance of
downwelling light.
As the bottom is reached, quite marked changes
occur in the colours of the animals. Many animals

Figure 14.11 Many bottom and near-bottom animals are

very pale, like this 220 mm long rat-tail Nezumia, at a
depth of 1100 m off southwest Ireland. In the light of the
flash these animals are not easily distinguishable from the
pale sediment on the bottom; this may also be the case in
whatever dim bioluminescence exists on the sea floor.
Also here (just above and to the left of the fish) is a shrimp
with a reflective eye (see also Figure 14.18) and (above
and to the right) two large pot-like glass sponges.
(Courtesy of the Southampton Oceanography Centre,
England.)

217
P.J. Herring

14.12 125

100
Number of rods

75

50

25

450 460 470 480 490 500 510 520 530 540
Wavelength of maximum absorbance (nm)
Figure 14.12 The distribution of 1370 measurements of the visual pigment maxima in the
rods of 57 species of deep-sea fish shows a good match with the blue–green light in their
environment (Figure 14.1). Most pigments are rhodopsins; the few longer wavelength por-
phyropsins are shown as open blocks (from Partridge et al.13).

Vision bright reflective object is viewed horizontally, max-

The light environment of the ocean is matched by imum contrast can be achieved by exploiting the
the visual adaptations of its inhabitants. Vision spectral differences between the reflected and back-
depends on the absorption of photon energy by the ground light, using a visual pigment whose absorp-
visual pigments and its transduction into a neural tion maximum is offset from that of the back-
signal. The spectral sensitivity of the eye is deter- ground.
mined by the absorption characteristics of the visu- Visual pigments are formed by linking a protein,
al pigments in the retinal receptors. In vertebrates one of the opsins, to a vitamin A1 or A2 derivative
these receptors are single cells (rods and cones); in (forming a rhodopsin or a porphyropsin, respec-
invertebrates, they are units (rhabdoms) formed tively). Porphyropsins absorb at longer wavelengths
from several cells. than do their rhodopsin partners. Although a few
marine fish do have this pair of pigments, many of
Visual pigments them lack the porphyropsin, but have more than
In general, the absorption maximum of the main one rhodopsin (i.e., vitamin A 1 with different
visual pigment is a good match to the spectral char- opsins) and thus retain the potential for colour
acteristics of the environment. Thus, deep-sea fish vision. Additional visual pigments may also be pre-
usually have rod visual pigments with absorption sent, usually in different types of cone cell, and
maxima in the blue wavelengths around 480 nm coloured filters or oil droplets may further differen-
(Figure 14.12), while shallow coastal species have tiate the spectral sensitivity of individual receptors.
maxima at longer wavelengths11,13. Near the surface In the most extreme cases (some mantis shrimps)
the high intensity and broad spectral range of there may be up to eight kinds of receptors, each
ambient light provide the opportunity for both with different spectral sensitivities. Recent work
colour vision and high acuity. The dominant visual has shown that the shrimp Systellaspis debilis has a
task is to maximise the contrast present in the tar- visual pigment which is sensitive to near-ultraviolet
get area. When a dark object, or silhouette, is seen light, as well as one sensitive to blue–green light4
against the background of downwelling light, or (Figure 14.13).
horizontally against an infinite background of scat- Colour vision is also theoretically possible for
tered light, the contrast is maximised by having a fish with only one visual pigment, but with a retina
visual pigment which matches the spectral trans- containing multiple banks of rods. Each layer mod-
mission of the water. When, on the other hand, a ifies the spectral nature of the light transmitted to

218
 14: Light, Colour, and Vision in the Ocean

14.13 Figure 14.13 The shrimp Systellaspis debilis

(60 mm long) has two visual pigments, one
absorbing in the blue–green, the other in the
near ultraviolet. Since these wavelengths
have different transmission characteristics,
their ratio could give the animal an indica-
tion of its depth. The dark spots on the tho-
rax and abdomen are light-emitting organs.

the next layer, giving them, in effect, different spec- few hundred metres are likely to sight prey or detri-
tral sensitivities. tal particles within an angle of 35˚ from the verti-
Major changes in the light environment of fish cal; here, the downward radiance does not drop
occur during the lifetime of those, like the eel, below 50% of its maximum value (Figure 14.1).
which have a marine and a freshwater phase or, This limited, but brighter cone of view (70˚), domi-
like the pollock, migrate into deeper water as an nates the visual environment. Many animals at
adult. These changes are compensated by visual these depths have responded by evolving upwardly
pigment changes, either between rhodopsin and pointing eyes9. Every stage between fully lateral
porphyropsin pairs or by opsin shifts between dif- and fully upward eyes can be found in one or other
ferent rhodopsins. In either case the new suite of species of mesopelagic fish, culminating in the
pigments is more appropriate to the visual tasks of extreme cases of Opisthoproctus, Benthalbella
the new environment. (Figure 14.14), and the hatchet fish Argyropelecus.
Visual acuity (resolution) needs to be maximised in
The upward view this direction and both amphipods and euphausiids
Light in the deep-water environment is dimmer, have a gradation of forms whose eyes range from
bluer, and highly directional. Animals in the upper round, with a uniform acuity over the whole visual

14.14

Figure 14.14 The tubular

eyes of the 100 mm long
fish Benthalbella provide
a binocular overlap,
allowing it to determine
the range of prey as well
as providing a large aper-
ture for high sensitivity.
(Courtesy of the South-
ampton Oceanography
Centre, England.)

219
 P.J. Herring

14.15 (a)
Figure 14.15 Some examples of tubular eyes
are shown here. (a) The fish Benthalbella
(seen from the front, lens diameter about
3 mm) showing the fields of view of the two
eyes and their binocular overlap (dark
(b)
hatched area). The lens pad [arrow, here and
in (b)] extends the field by collecting light
l
from the side and below. (b) A transverse sec-
tion of the left eye superimposed on an out-
line of a ‘normal’ fish eye with the same size
lr lens (l). A focused image is formed on the
ventral retina (vr), and unfocused light from
vr
the lens pad is detected by the lateral retina
(lr). (c) Sections of the eye (3 mm diameter) of
the deep-sea octopod Amphitretus and (d) of
the divided eye (2 mm diameter) of the
euphausiid shrimp Stylocheiron suhmii, one
(c) (d)
l ir portion of which looks upward, the other
cc downward and to the side (cc, crystalline
cones; ir, iris; og, optic ganglion; rh, rhab-
doms). The fish eye is effectively the central
part of a normal eye, and the octopod eye is
optically similar. The euphausiid eye pro-
vides an upwardly directed (‘tubular’) region
of high acuity (but narrow visual field); the
rh lower region covers a wider visual field, but
at lower acuity, in the much dimmer light
vr
from the side and below (after Land9, from
cc Marshall12 and Locket10).
og

field (Meganyctiphanes), through double eyes with eyes have the individual units (ommatidia, each
a narrow upper visual field of high acuity and a containing one rhabdom) optically isolated from
more extensive lower visual field of low acuity each other (usually by pigment). The eye has good
(Phronima, Stylocheiron suhmii, Figure 14.15), to resolution, but the effective aperture is that of a
those with effectively only an upward eye of high single facet. In superposition eyes, the units (omma-
resolution and narrow field (Cystisoma). tidia) are not optically isolated and the light from a
large number of ommatidia can be focused on a
Maximum sensitivity single receptor. The aperture is now the facet area
At greater depths (and, of course, at night nearer of all these ommatidia, giving a huge increase in
the surface) the light levels become marginal for sensitivity (Figure 14.16). Deep-sea shrimps of all
vision and the adaptations of the eyes of the ani- kinds have such superposition eyes; often, their lar-
mals are largely dedicated to maximising their sen- vae, at shallower depths, have apposition eyes,
sitivity. This is determined by two factors: first, the which change into the superposition form as the
illuminance of the retina and, second, the probabil- larvae descend into deeper and darker water during
ity of photon capture by the receptors. their development.
Image brightness The aperture of the fish eye is filled by the spher-
The illuminance (or image brightness) is a function ical lens (see Figure 14.15). Only in near-surface
of the aperture (pupil) and focal length of the eye. species is there an iris diaphragm to stop down the
The aperture determines how much light enters. In aperture in bright light. A larger aperture requires a
the compound eyes of crustaceans, the aperture can larger lens, but the construction of the fish (and
be greatly increased by changing from an apposi- cephalopod) lens is such that the focal length is a
tion type of eye to a superposition one9. Apposition constant ratio of the lens radius (2.5:1,

220
 14: Light, Colour, and Vision in the Ocean

14.16 (a) Figure 14.16 (a) A crustacean apposition eye, in

Light Light which each unit (ommatidium) is optically isolated,
with the receptor (rhabdom, rh) joined to the crys-
talline cone (cc). Only a very narrow beam of near-
axial light reaches each receptor through its own
cone; off-axis light is absorbed by the dark pigment
p
(p). (b) A superposition eye, in which the receptors are
cc
separated from the cones by a clear zone (cz); the
rh aperture is now formed by a large group of facets,
greatly increasing the receptor illuminance. This kind
of eye can be constructed using refractive cones, as in
(b)
(b) (e.g., in euphausiid shrimps), or by reflective mir-
Light
Light
rors (m), as in (c) (e.g., decapod shrimps). (From
Herring and Roe8.)

cz

cc

rh

(c)

Light Light

Matthiessen’s ratio). This means that any increase ment is normally present in the rods of deep-water
in aperture necessitates an equivalent increase in fish. Cones are usually absent. The density of the
focal length, i.e., a larger eye. Large eyes are a fea- visual pigment varies little between species, but the
ture of many deep-sea fish – but there is a practical length of the light path through the receptors can
limit to how large an eye can be, yet still fit on the be increased by lengthening the rods or having mul-
head. A tubular eye overcomes much of this prob- tiple banks of short rods10. A simple means of dou-
lem. A very large aperture is possible because only bling the light path is to place a specular or diffuse
a tubular portion of the equivalently sized normal reflector (or retinal tapetum) at the back of the eye.
eye is retained. Fish such as Dolichopteryx have
huge lenses, each almost half the width of the head;
this is achieved by having a narrow tubular eye 14.17
with a very limited field of view (Figure 14.17).
Many fish with tubular eyes (including
Dolichopteryx) have evolved secondary methods of
increasing their visual field. Light guides, lens pads,
and accessory reflectors capture some light from
below the main visual field and convey it to an
accessory retina. It is not focused, so no image is
formed, but it does offer additional information to
that available solely through the lens10.
Photon capture
The probability of photon capture by the receptors
is a function of the visual pigment density, its
Figure 14.17 An extreme development of a tubular eye
absorption maximum, the length of the light path, occurs in the 220 mm long fish Dolichopteryx (seen from
and the diameter of the receptors. Maximum sensi- above) in which the huge lenses filling the aperture pro-
tivity requires that the visual pigment match the vide the maximum possible retinal illuminance, while
ambient light very closely. Only a single visual pig- still just fitting on the head.

221
 P.J. Herring

14.18 Figure 14.18 Many deep-sea animals have reflective lay-

ers in the eye behind the retina, increasing the chances of
photon capture. In the shrimp Plesiopenaeus (130 mm
long, attracted to a baited camera) this is visible as a
bright eyeshine. (Courtesy of the Southampton
Oceanography Centre, England.)

there may just be able to detect the black body

radiation of the hot submarine vents where this
animal lives14 (see also Chapter 13).
Bioluminescence
The presence of functional eyes in many abyssal ani-
Many seals, fish, and crustaceans have such layers, mals, well below the maximum possible penetration
which give the animal an intense eyeshine in bright of daylight, testifies to the importance of biological
light (Figure 14.18). In sharks and their relatives, light, or bioluminescence, in the ecology of the deep
black pigment covers the tapetum during the day sea7. Fireflies and glow-worms are the well-known
and withdraws at night. The tapetum may be terrestrial examples of this phenomenon, but there
formed of layers of crystals, arranged for construc- are few others on land and only one (a New Zealand
tive interference of the ambient blue wavelengths, limpet) in fresh water. In contrast, the sea contains
or of uniform granules or lipid spheres which give a an immense variety of luminous organisms spread
more diffuse reflectance. Receptor diameter is effec- across 16 different major groups, or phyla. The
tively increased by pooling the output from a group organisms range from bacteria and dinoflagellates to
of rods or rhabdoms into a single ganglion cell. squid and fish, and their distribution extends from
This increases sensitivity, but at the expense of acu- the surface waters to the greatest depths. This rich
ity, unless the focal length (i.e., eye size) is also variety of organism is matched by the astonishing
increased to retain the same angular separation of range of their bioluminescent capabilities and func-
the enlarged receptor units. tions. The conclusion to be drawn from the variety
of luminous groups, and the different chemistries
Eye reduction involved, is that the ability to emit light has evolved
In the depths of the ocean, where the only light is independently in many different organisms5.
bioluminescence, some animals retain large sensi-
tive eyes, but others lose much of the optical com- Mechanism
plexity present in shallower species and/or reduce Bioluminescence is the harnessing of a chemilumi-
the eye size to almost rudimentary proportions. In nescent reaction by a living organism, with a suffi-
shrimp such as Hymenodora, all the focusing ele- cient quantum (photon) yield for the light to be
ments are lost and the eye retains only very com- visible. The overall reaction involves the oxidation
plex rhabdoms and reflective material. In the fish of a substrate (‘luciferin’) catalysed by an enzyme
Ipnops, the lens is also lost and the eye is little (‘luciferase’), so that some of the energy released by
more than a flattened sheet of rod receptors. the reaction is emitted as light rather than heat.
The deep-water shrimp Acanthephyra cur- The energy can be transferred into a second, fluo-
tirostris has tiny reflective eyes, compared with the rescent, compound which then emits light at its
larger pigmented ones of shallower relatives, such own characteristic wavelength (always longer than
as A. purpurea and S. debilis (Figure 14.19; see also that produced by the original reaction). There are
Figures 14.8, 14.13). The deepest-living euphausiid
shrimps have equally tiny eyes, as do abyssal rat-
14.19
tail fish. One remarkable shrimp (Alvinocaris) has
completely lost its compound eyes, but has devel-
oped what appears to be a lens-less accessory visual
system in the thorax. The visual pigment present

Figure 14.19 In very deep waters many animals reduce

the size and organisation of their eyes. Acanthephyra cur-
tirostris (65 mm long) has a much smaller eye than its
shallower-living relatives. The eye no longer has any dark
pigment and appears diffusely reflective (compare with
Figure 14.13). (Courtesy of the Southampton
Oceanography Centre, England.)

222
 14: Light, Colour, and Vision in the Ocean

14.20a 14.20b

Figure 14.20 (a) The fish Malacosteus has two light 1.0 14.20c
organs near the eye, one (appearing white) behind the
eye and a larger dark brown one under the 5 mm diame-
ter eye (courtesy of the Southampton Oceanography Relative intensity
Centre, England). (b) When illuminated with ultraviolet
light the contents of the dark brown light organ fluoresce 0.5
an intense red. The white organ emits a blue biolumines-
cence and the brown a deep-red bioluminescence. Their
respective emission spectra are shown in (c). It is likely
that the red light is produced by transferring the energy of
the bioluminescence reaction to the red fluor within it
(see text), and then further filtering the light through the 400 500 600 700 800
deep brown surface of the light organ. Wavelength (nm)

many different kinds of luciferin, some restricted to maximising their effective signalling range. Some
particular groups of organisms, others spread animals, particularly hydrozoan jellyfish, add a
across several different phyla. Relatively few have green fluorescent protein to the luminous cells and
yet been identified. Most organisms make their emit green light. It is not clear whether this is a
own luciferin (although some may acquire it in response to a more coastal environment, with its
their diet), but some fish, squid, and tunicates use greener water transmission, or a means of increas-
symbiotic luminous bacteria as their source of light. ing the quantum efficiency (i.e., the light yield) of
The brightest bioluminescence can only match the reaction.
the intensity of moonlight, so bioluminescence It may be more important (for maximising the
makes no significant contribution to the light of effective range) to have a bright green light than a
near-surface oceanic waters during the day. In tur- dimmer blue one, if there is little difference in an
bid coastal waters, however, it may be the only observer’s spectral sensitivity to the two wave-
light visible as shallow as 30 m, even during the lengths. There are a few other colours of biolumines-
day. At all other times and depths it is a major, and cent light; the worm Tomopteris has a yellow lumi-
often the only, source of light in the sea. nescence and the fish Malacosteus (and a few others)
has a deep red light, as well as a blue one. In the lat-
Colours of bioluminescence ter case, the red organ (under the eye) has a narrow
If the eyes of marine animals have adapted primari- bandwidth emission with a maximum at about
ly to the characteristics of submarine daylight, with 700 nm; this light is invisible to the dark-adapted
its blue transmission maximum in clear waters, it is human eye. A blue organ just behind it emits at
likely that bioluminescence has followed this visual 480 nm. The red light is probably produced by the
lead. The colours of marine bioluminescence are, same chemistry as the blue light, but with a combi-
indeed, very largely confined to blue–green wave- nation of an additional red fluorescer in the light
lengths6. These, of course, are not only best per- organ and a deep-red filter on its surface (Figure
ceived by other marine organisms, whether or not 14.20). The red organ cannot be very efficient; about
they are themselves bioluminescent, but also will be 80% of the light it produces is absorbed by the filter
best transmitted through the water, both features in order to achieve the narrow bandwidth.

223
 P.J. Herring

14.21 Figure 14.21 Malacosteus (175 mm

long) has both blue- and red-sensi-
tive visual pigments. The red sensi-
tivity is further enhanced by a red
reflector (tapetum) behind the retina,
providing the red colour to the eye
visible here in daylight.(Courtesy of
Dr N.A. Locket, Adelaide University,
Australia.)

14.22a 14.22b

Figure 14.22 (a) The hatchet fish Argyropelecus (Figure 14.6), seen from below, showing the photophores arranged
along its ventral margin, each containing a magenta-coloured filter. This results in the emitted bioluminescence being
a clear blue, exactly matching the colour of light in the sea (courtesy of P.M. David, Southampton, England). (b) A
luminescing specimen by its own light.

The long wavelength light must be very impor- coelenterate, a sea anemone-like zoanthid, has
tant to the fish, but it would not normally be able colonies in which some individuals have green light
to see it if it had only a typical blue–green sensitive and some yellow.
visual pigment. Malacosteus turns out to have a
red-sensitive visual pigment as well. It also has a Luminous camouflage
scarlet tapetum (see earlier) to maximise its sensi- The ecological value of most of the exceptions noted
tivity to these long wavelengths (Figure 14.21). It above is not yet clear, but the value of precisely con-
thus has a ‘private’ wavelength, which could be trolling the colour has been well-established in the
used either as a secure communication with others hatchet fish. This fish lives at depths where daylight
of the same species or to break the camouflage of is still important and, as already described, has mir-
red shrimps, whose colour only works in blue illu- ror-like camouflage and tubular eyes. Despite being
mination and whose blue-sensitive eyes would not very laterally flattened, it cannot altogether avoid
detect the red illumination. Red light of these wave- being seen in silhouette from below. Like many
lengths is rapidly absorbed by sea water, so it can other fish at mid-depths, it eliminates this silhouette
only be effective over short visual ranges3. by having rows of light organs along its ventral sur-
A few other marine animals produce light of face [Figure 14.22(a)]. All its light organs point
more than one colour. The stalked sea-pen downward, except one, which points into the eye.
Umbellula, for example, has green luminescence on By adjusting the luminescence shining into the eye to
its stalk, but blue emission from the polyps at the match the downwelling daylight, it simultaneously
top. Some squid produce green and blue light from matches all its ventral lights to the surrounding
different light organs, and can also change the emis- light – and vanishes. However, the broad bandwidth
sion spectrum from a single light organ. Another light produced within the light organ is not quite the

224
 14: Light, Colour, and Vision in the Ocean

Figure 14.23 Mature females of the pelagic octopod 14.23

Japetella (100 mm long) develop a yellow-coloured bio-
luminescent oral ring which degenerates after they
spawn. Males have no luminous organ, so it is assumed
that this provides a sexual signal.

same colour as light in the sea. The aperture of each as an active defence, which can take the form of a
light organ contains a purple filter pigment which single, short, bright flash (dinoflagellate), a volley of
corrects this spectral mismatch. After passage flashes (some fish), a wave of flashes moving over
through the filter the luminescence has a narrow the body (sea-pens and medusae), or the discarding
bandwidth blue emission with a maximum at about of sacrificial parts of the body which flash indepen-
475 nm, corresponding exactly to the spectrum of dently to distract a predator (scale worms and brit-
light in the sea [Figure 14.22(b)]. To complete the tle stars). Another widely employed active defence is
camouflage, the design of the reflectors in the light that of a squirted luminescence. Many shrimps,
organs ensures that the angular distribution of the worms, a few squid and fish, and some medusae
luminescence also matches that of submarine day- and ctenophores produce copious amounts of light
light3, as illustrated in Figure 14.1(b). in this form. Particularly among the ctenophores
and medusae, it is not simply a cloud, but is com-
Other luminous defences
posed of separately scintillating particles. Many of
Camouflage is an example of a passive defensive these gelatinous animals hang passively in the water
function. The most common use of light in the sea is fishing for prey; for any larger animal in the area,
they collectively form a luminous minefield.
14.24
Sexual light signals
Small crustaceans, such as copepods and ostracods,
have luminous glands whose secretions the animals
kick away defensively as they swim. They can also
be used for sexual displays. The pattern, timing,
and trajectory of the luminous gobbets produced in
the mating displays of males of the Vargula group
of ostracods identify them to the waiting females,
who swim up to join them. Syllid worms have anal-
ogous displays in many parts of the world (e.g., the
Bermuda fireworm). Sexual differences in the size
or position of the light organs of male and female
lantern fish, stomiatoid fish, and some cephalopods
also suggest that they have a sexual function
(Figure 14.23). Female angler-fish have lures con-
Figure 14.24 Female angler-fish, such as this 45 mm long taining luminous bacterial symbionts, but the males
Chaenophryne, maintain a culture of luminous bacteria in do not. It is assumed that the lure attracts prey, but
the very elaborate lure. Although regarded primarily as a
means of attracting prey, it may also be a means of identi-
perhaps it also sends a specific signal to the males
fying the female to the non-luminous male. (Figure 14.24).

225
 P.J. Herring

14.25 Figure 14.25 The flashlight fish

Photoblepharon is a shallow-
water fish, some 90 mm long,
which rises from dark crevices
to feed on the reef at night. The
light of the bacteria in the pho-
tophore beneath the eye is used
to illuminate its plankton prey.
The light can be turned off by
pulling a shutter up over the
light organ.

Luminous bacterial symbionts glow continuously; the host has to develop a means
Those animals utilising luminous bacteria have to of turning them off when necessary. Usually, this is
constrain them in a particular gland or tissue, pro- achieved by pulling a shutter across the aperture of
vide them with the right environment to thrive, dis- the organ, rotating the organ to face inward, or dis-
pose of dead ones, and ensure that the right bacte- persing dark pigment over its surface, but control
ria are transferred to the next generation. Each host of the oxygen (blood) supply may sometimes be
requires a particular species of bacterium, though involved.
many share the same species. Many hosts have light
organs which are connected to the gut lumen and Photophore structure
from which the symbionts can easily be cultured in The possible optical complexities of light organs
the laboratory. As the natural (but largely acciden- fully match those of eyes7. The light-emitting cells
tal) bacterial flora of the gut contains some lumi- sit within a pigment cup which limits the aperture,
nous species, acquisition of the right bacterium for while a specular interference reflector of guanine or
the light organ is probably easy. The same bacterial protein platelets, or a diffuse granular reflector,
species, and a few additional ones, are found ‘free- greatly increases the output efficiency. A lens focus-
living’ on the surfaces of animals, on marine snow es the light, a lamellar ring collimates it, and inter-
(see Chapter 7), and possibly free in the water. In ference or pigmentary filters change its spectrum.
the angler-fish and flashlight fish (Figure 14.25), Light guides spread the output over a larger area
there is no light organ connection to the gut and and light pipes even allow the light to be emitted
the bacteria are unculturable, so their origin is not some distance from the light organ. Many hundred
known. Genetic data indicate that they are new light organs are present on some fish and squid
species and very host-specific. Luminous bacteria (Figure 14.26); individual species may have three or

14.26

Figure 14.26 Some animals have hun-

dreds of separate light organs. The
45 mm long Japanese firefly squid,
Watasenia scintillans (seen here from
below by its own light), has most of
them arranged over its ventral surface
as a camouflage, similar in principle to
that of the hatchet fish. (Courtesy of
Prof. Y. Kito, Osaka University, Japan.)

226
 14: Light, Colour, and Vision in the Ocean

four structurally quite different types at different 3. Denton, E.J., Herring, P.J., Widder, E.A., Latz, M.I., and
sites on or in the body. We do not yet know what Case, J.F. (1990), The roles of filters in the pho-
tophores of oceanic animals and their relation to
kind of light many of these organs produce, nor vision in the oceanic environment, Proc. R. Soc.,
what functions they serve. If light, colour, and Lond. B, 225, 63–97.
vision in the ocean are regarded as an interlocking 4. Frank, T.M. and Widder, E.A. (1994), Evidence for behav-
jig-saw puzzle, these are some of the pieces which ioral sensitivity to near-UV light in the deep-sea crus-
tacean Systellaspis debilis, Marine Biol., 118,
have not yet been fitted in. 279–284.
5. Hastings, J.W. and Morin, J.G. (1991), Bioluminescence, in
Neural and Integrative Animal Physiology, Prosser,
C.L. (ed.), Wiley-Liss, New York, pp 131-170.
General References 6. Herring, P.J. (1983), The spectral characteristics of luminous
marine organisms, Proc R. Soc., Lond., B, 220,
Hastings, J.W. and Morin, J.G. (1991), Bioluminescence, in
183–217.
Neural and Integrative Animal Physiology, Prosser,
7. Herring, P.J. (1985), How to survive in the dark: biolumines-
C.L. (ed.), Wiley-Liss, New York, pp 131-170.
cence in the deep sea, in Physiological Adaptations of
Herring, P.J. (1985), How to survive in the dark: biolumines-
Marine Animals, Laverack, M.S. (ed.), Symp. Soc.
cence in the deep sea, in Physiological Adaptations of
Exp. Biol., 39, 323–350.
Marine Animals, Laverack, M.S. (ed.), Symp. Soc.
8. Herring, P.J. and Roe, H.S.J. (1990), The photoecology of
Exp. Biol., 39, 323–350.
pelagic oceanic decapods, Symp. Zool. Soc. Lond.,
Locket, N.A. (1977), Adaptations to the deep-sea environment,
59, 263–290.
in Handbook of Sensory Physiology, Vol VII/5,
9. Land, M.F. (1990), Optics of the eyes of marine animals, in
Crescitelli, F. (ed.), Springer-Verlag, Berlin, pp
Light and Life in the Sea, Herring, P.J., Campbell,
67–192.
A.K., Whitfield, M., and Maddock, L. (eds),
Lythgoe, J. (1979), The Ecology of Vision, Clarendon Press,
Cambridge University Press, Cambridge, pp 149–166.
Oxford, 244 pp.
10. Locket, N.A. (1977), Adaptations to the deep-sea environ-
Marshall, N.B. (1979), Developments in Deep-Sea Biology,
ment, in Handbook of Sensory Physiology, Vol VII/5,
Blandford Press, Poole, 566 pp.
Crescitelli, F. (ed.), Springer-Verlag, Berlin, pp
67–192.
11. Lythgoe, J. (1979), The Ecology of Vision, Clarendon Press,
Oxford.
12. Marshall, N.B. (1979), Developments in Deep-Sea Biology,
References Blandford Press, Poole.
1. Denton, E.J. (1970), On the organization of reflecting sur- 13. Partridge, J.C., Shand, J., Archer, S.N., Lythgoe, J.N., and
faces in some marine animals, Phil. Trans. Roy. Soc. van Groningen-Luyben, W.A.H.M. (1989),
B, 258, 285–313. Interspecific variation in the visual pigments of deep-
2. Denton, E.J. (1990), Light and vision at depths greater than sea fish, J. Comp. Physiol. A, 164, 513–529.
200 metres, in Light and Life in the Sea, Herring, P.J., 14. Van Dover, C.L., Szuts, E.Z., Chamberlain, S.C., and Cann,
Campbell, A.K., Whitfield, M., and Maddock, L. J.R. (1989), A novel eye in ‘eyeless’ shrimp from
(eds), Cambridge University Press, Cambridge, pp hydrothermal vents of the Mid-Atlantic Ridge,
127–148. Nature, 337, 458–460.

227
 CHAPTER 15:

Ocean Diversity
M.V. Angel

Biodiversity is the term used to describe the rich (based on the fossil record), although several mass
variety of life found on Earth. It was the subject of extinctions have interrupted this trend (Figure
the UNCED conference in Rio in 1992, which 15.1). The causes of these mass extinctions are still
resulted in the signing of the International a subject of vigorous debate17. However, since the
Convention on Biodiversity and the adoption of beginning of the Mesozoic era the increase has been
Agenda 21, which lays down the guidelines under almost linear. If we are to understand the present
which the Convention will operate. However, bio- patterns of biodiversity in our seas, we need to
diversity is used to express this variety over a great appreciate, first, how and why evolution has given
range of levels of organisation. It can be applied to the world such a rich compendium of species and,
different types of ecosystems, or to express the second, how ecological processes continue to main-
number of species found locally or globally, or even tain this richness.
to the amount of genetic variation within individ-
ual species. This can lead to confused thinking, The Origins of Disparity
unless the term’s precise meaning is explicitly stated The earliest traces of life that have been found so
in each context. Our discussion concentrates on far are imprints of single-celled micro-organisms in
ecosystem diversity, and species richness and domi- sedimentary rocks laid down in the prototype
nance within local and regional areas. ocean about three and a half billion years ago. The
Diversity at these levels of organisation is the first to appear were prokaryotic cells or bacteria,
product of evolution. New species almost always which have no nucleus and are termed prokaryotes.
evolve as a result of a subpopulation becoming iso- Since they have the most ancient lineage of any liv-
lated from its parent population for thousands of ing thing so far, these apparently simple cells show
years, and being subjected to different selective remarkable variations in their physiology, cell
pressures. Generally, the stock of species of the chemistry, and genetics. As we learn more about
world has increased linearly over geological time them, there are proving to be greater differences

15.1 Figure 15.1 Diversity of marine families

through the geological record since metazoan
organisms first appeared. Note the steady
800 increase until the end of the Ordovician (O in
the lower scale), when the first of the mass
extinctions occurred. Then the numbers of
k

families remained roughly constant until the

600 mass extinctions of the Triassic (T) and the
k

beginnning of the fragmentation of the super-

k

k
k
Families

continent Pangea (see Figure 15.4). Compare

the steady increase in numbers of families
400 since the Triassic with the patterns of continen-
tal drift shown in Figure 15.4. The one inter-
ruption was the mass extinction at the end of
the Cretaceous (Cr), which more-or-less coin-
200
cided with the split developing between
Australasia and Antarctica, and resulted in the
start of circumpolar circulation in the Southern
Ocean (redrawn from Sepkowski 17 ). (V =
600 400 200 0 Vendian; C = Cambrian; O = Ordovician; S =
Time (106 y) Silurian; D = Devonian; Ca = Carboniferous; P
= Permian; T = Triassic; J = Jurassic; Cr =
V C O S D Ca P T J Cr Cz Cretaceous; Cz = Cenozoic.)

228
 15: Ocean Diversity

Figure 15.2 Specimens of the large 15.2

deep-living copepod species
Megacalanus princeps. Copepods
numerically dominate the vast majori-
ty of plankton samples, no matter at
what depth they are collected. They
outnumber all other animals of com-
parable size in any other ecosystem,
including all insects, and yet there are
only just over 1900 known species
compared with perhaps a million
insects. (© Heather Angel.)

within the prokaryotes than within all other living includes several hundreds of thousands of known
organisms9. species, and whose true global richness is estimated
It took another two billion years before the first to be in excess of a million. Much the same applies
appearance of single-celled organisms with nuclei – to the plants; there are an estimated 250,000
the eukaryotes. Both the nucleus, which contains species of terrestrial green plants, but merely
the genetic information, and other organelles with- 3500–4500 in the oceans (Figure 15.3). Why are
in the cells – the mitochondria, in which the reac- there such major differences in species richness?
tions occur that provide the cell with energy – are
thought to have originated as a result of different 15.3
prokaryotes forming symbiotic associations, which
through time became a permanent and obligate
relationship.
It was almost another billion years before the
first multicelled organisms made an appearance,
just before the beginning of the Cambrian era some
670 million years ago, and it was yet another 155
million years before multicellular forms began to
invade the land during the late Silurian.
Considering its much greater geological age, it is
hardly surprising that the fauna of the oceans is far
more disparate than the terrestrial fauna, being
made up of 28 phyla – the name given to the basic
types of animals (Table 15.1), compared with just
the 11 phyla represented in terrestrial faunas.
However, is this greater disparity also repeated in a
greater species richness in the oceans compared to
the richness on land?
Terrestrial Versus Marine Species Richness
For pelagic faunas the answer appears to be no; ter-
restrial faunas appear to be far richer in species.
For example, the most abundant and species-rich Figure 15.3 Giant phytoplankton cells (1–2 mm in diame-
ter) of the monad Halosphaera viridis. Apart from the large
group of plankton are the copepods (Figure 15.2), algae that grow in shallow coastal waters, virtually all life
of which there are just over 1900 species known in the ocean is dependent on the photosynthetic produc-
from oceanic and brackish waters. Compare this tion of the 3500–4500 species of phytoplankton. Which
with the most diverse inhabitants of terrestrial should receive the greatest priority for conservation, one of
habitats – the insects (a group virtually absent from the few phytoplankton species or one of the quarter of a
the oceans), in which just one order, the beetles, million terrestrial green plants? (© Heather Angel.)

229
M.V. Angel

Table 15.1 Distribution of phyla in major habitats12.

Phyla Marine Freshwater Terrestrial Symbiotic

Subphyla Benthic Pelagic Benthic Pelagic Moist Xeric Ecto Endo

Porifera +++ + +
Placozoa +
Orthonectida +
Dicyemida +
Cnidaria +++ ++ + + +
Ctenophora + +
Platyhelminthes +++ + +++ ++ + ++++
Gnathostomulida++
Nemertea ++ + + + +

Nematoda +++ + +++ + +++ + +++ +++

Nematomorpha ++
Acanthocephala ++
Rotifera + + ++ ++ + + +
Gastrotricha ++ ++
Kinorhyncha ++
Loricifera +
Tardigrada + ++ +
Priapula +

Mollusca +++++ + +++ +++ + + +

Kamptozoa + + +
Pogonophora ++
Sipuncula ++ +
Echiura ++
Annelida ++++ + ++ +++ ++
Onychophora +
Arthropoda
Crustacea ++++ +++ +++ ++ ++ ++ ++
Chelicerata ++ + ++ ++ ++++ +++ ++ +
Uniramia + + +++ ++ +++++ +++ ++ ++

Chaetognatha + +
Phoronida +
Brachiopoda ++
Bryozoa +++ +
Echinodermata +++ +
Hemichordata +
Chordata
Urochordata +++ +
Cephalochordata +
Vertebrata +++ +++ ++ +++ +++ +++ + +

‘Pluses’ indicate approximate abundance of living described species:

+ = 1–102
++ = 102–103
+++ = 103–104
++++ = 104–105

230
 15: Ocean Diversity

15.4

175 BP 31 BP

100 BP 16 BP

59 BP 5 BP

Figure 15.4 Reconstruction of the distribution of the continents and the main surface circulation (red arrows) of the
oceans through the past 175 million years, following the fragmentation of the supercontinent Pangea (the blue arrows
indicate where deep water forms). The distribution of old genera and families still retains the imprint of the ancient cir-
culation patterns. Events like the interruption of circumequatorial and circumpolar circulation in the Southern Ocean
have had a major impact on patterns of speciation and redistribution (modified from Parrish and Curtis11).

Factors Leading to Speciation basin morphology, other events have had major
New species tend to evolve when populations are impacts on the distribution patterns of species.
split and isolated. Isolation in both marine and ter- During the Miocene, for example, the Straits of
restrial environments can arise from tectonic Gibraltar repeatedly closed and opened. Each time
events, resulting from the drift of the continents the straits closed the sea dried out, leaving behind
altering the shape and morphology of ocean basins, vast deposits of salt that today still underlie a drape
and forming new islands and mountain ridges. of pelagic sediments beneath the Mediterranean
Over the past 175 million years, plate tectonic Sea. When the Straits re-opened the sea catastroph-
movements have resulted in major redistributions ically reflooded, lowering sea levels globally by tens
of the land masses (see Chapter 20). Since the pat- of metres within relatively few years, maybe cen-
tern of large-scale thermohaline circulation of turies. Such catastrophic lowering of sea level
ocean currents is determined by the interaction caused the deaths of many shallow-living species,
between the shape of the oceans and wind patterns and probably led to many extinctions. Another
(Chapter 2), ocean circulations during the geologi- such event was the opening of the Isthmus of
cal past can be reconstructed with confidence. In Panama about ten million years ago, which allowed
the modern ocean, the distributions of major pelag- the exchange of shallow-living faunas between the
ic ecosystems match the patterns of the major Pacific and Atlantic Oceans. As a result, many shal-
ocean current gyres, and are likely to have done so low-living tropical families, genera, and even
in the past (Figure 15.4). Analyses of the present- species have circumequatorial distributions.
day distributions of some of the more ancient fami- Throughout the Quaternary sea levels fluctuated
lies and genera of pelagic organisms reveal persis- by around 100 m as a result of the alternating
tent echoes of these ancient circulations18. glacial and interglacial stages, which appear to
In addition to these major changes to ocean- have been a response in the Earth’s climate to

231
 M.V. Angel

15.5 Figure 15.5 Changes in the numbers of pelagic species

A sampled across the shelf, shelf-break, and continental
slope off Florida in the Gulf of Mexico: A, copepoda; B,
mid-water fish; C, decapod crustaceans; and D,
40 B
euphausiids. For these four pelagic taxa the numbers of
Species (numbers)

species increase offshore and often peak at or close to the

shelf-break (modified from Hopkins et al.6).

Fine-Scale Habitat Variability

▲
20 ▲▲ Such large-scale events play an equally important
▲ ▲ C role in speciation on land, but in addition there are
▲
▲
▲ much finer-scale processes at work. These result
▲
▲ ▲
▲ D from terrestrial ecosystems being more finely struc-
▲ tured, with much greater development of microcli-
mates, and dispersal between the different patches
▲
▲ ▲
▲
▲
0 being much more restricted. The physical structure
and aspect of the landscape, the types of soils, and
the availability of water and nutrients all contribute
to the creation of fine-scale habitat mosaics in ter-
1
restrial ecosystems. These patches are dynamic; for
Depth (km)

example, forest clearings are created by trees dying

or being razed to the ground by violent storms.
2 Animals and plants vary in their abilities to dis-
Depth
perse, so populations of a given species may
become isolated as the patches become too far
3 apart for individuals to move between them, or a
0 100 200 300 phenomenon such as a landslide or lava flow
Offshore distance (km) creates an impassable barrier. So, even within a
mosaic of habitats, similar patches may be support-
eccentricities in its orbit around the Sun, the so- ing slightly different assemblages of plant and
called Milankovitch cycles. Around the East Indies, animal species.
in particular, fluctuating sea levels resulted in con- Since the dominant terrestrial plants are larger
nections between the islands appearing and disap- and longer-lived than their marine counterparts,
pearing, so that the deep basins between the islands many have co-evolved with invertebrates, especially
were isolated and then reconnected. Consequently, with insects; for example, in Britain the oak tree
populations both on land and in the coastal seas alone has about 600 insect associates. Moreover,
were divided and then recombined. terrestrial plants by their physical presences, as well
This led to a series of biogeographical divisions as their physiological activity, modify the habitats
between the islands’ faunas and floras, such as the they inhabit; the environment inside a wood is very
Wallace’s, Weber’s, and Lydekker’s Lines, which different from that immmediately outside.
are reflected, to some extent, in the marine faunas There is some similar modification of habitats
as well4. Similarly, in the Mediterranean there are by the biota in the sea (for example, around coral
glacial relict species which entered the sea during reefs, in kelp forests, and over mussel beds), but the
the most recent glaciation and are now not only creation of such three-dimensional habitat struc-
totally isolated from their parent stock in the North tures by the organisms themselves is almost entirely
Atlantic, but are also living in very different limited to shallow coastal waters. The hermatypic
conditions. corals, which create much of the structure of coral
During the past 22,000 years, since the height of reefs, are limited to depths where their symbiotic
the latest glaciation (note that recent calibrations of algae can obtain enough light to photosynthesise.
14
C dating methods have resulted in this time inter- In shallow seas, even where there is no biologi-
val being increased from 18,000 years), it is proba- cal structuring of the environment, the complex
ble that at least some of these relicts have diverged interactions between tides, waves, and the coastal
sufficiently to become genetically incompatible. morphology and geology create much finer-scaled
Thus, the isolation between populations in the mosaics of different habitats than occur in open-
ocean, that in time results in their diverging and ocean habitats, particularly pelagic ones (see
evolving into new species, tends to be related to Chapter 16). A sampling transect across a conti-
large-scale and persistent events. nental-shelf sea and into the open ocean reveals rel-

232
 15: Ocean Diversity

atively few species at a given location over the shelf exponentially and depth profiles of biomass show
and a sharp increase in the number of species just that the pelagic biomass at a depth of 1000 m is a
beyond the shelf-break (Figure 15.5). However, if tenth that at the surface, and declines by another
the sampling is extended to become regional in its tenth at 4000 m (Figure 15.6). The only place that
coverage, the number of species caught inshore these vertical gradients combine to be steep enough
soon begins to exceed those caught in the open to act as a barrier to the dispersion of a few species
ocean; oceanic habitats are large-scale, whereas is across the seasonal thermocline. But even there
inshore and coastal habitats (which range from some of the vertically migrating species are unin-
muddy estuaries to sandy shores and rocky beach- hibited in their movements, as they commute daily
es) are much finer in scale. from the safety of day-time depths of 500 m or
Another feature of coastal habitats is that they more, to spend the night feeding in the surface
are geographically linear, so any break caused by a waters. During these migrations they may experi-
large estuary, a strait, or even a major piece of ence greater changes in temperature than occur
coastal engineering can act as a barrier to the dis- throughout the year in the surface waters.
persal of some taxa. Most (but not all) of the Horizontal gradients are almost always weaker,
species which live between the tides or in associa- even at boundaries between different water masses,
tion with the sea bed have dispersal phases as such as at the Antarctic Convergence. The one pos-
planktonic larvae, which may last from a few days sible exception is in upwelling areas, where boluses
to several months. Those with short planktonic of cold subthermocline water that have upwelled to
phases, or none at all, may not be able to cross the surface can create very sharp sea-surface tem-
these barriers because of the prevailing currents. As perature discontinuities. However, such fronts are
a result, taxa which inhabit coastal waters are far too ephemeral to have a long-term influence on
much more restricted in their distributions and are diversity. Otherwise, the continual stirring of the
globally more diverse than similar taxa inhabiting water by currents and eddies blurs all such bound-
the open ocean. This trend is illustrated below. aries. As a result, there are almost no clearly
Recently, some human activities have begun to defined ecosystems in pelagic environments; the
break down some of these barriers to distribution, assemblages of animals which merge almost seam-
with serious implications for the maintenance of lessly across the gradients are described as eco-
marine biodiversity. For example, the building of tones. Locally, patchiness and variability is high,
the Suez Canal has led to increasingly large num-
bers of species exchanging between the Red Sea
and the Mediterranean13. More recently, evidence Log1 Biomass 15.6
has emerged that the transoceanic transportion of –1 0 1 2
coastal species in the ballast waters of fast cargo H G F E DB A
vessels is disturbing local faunas3.

Gradients and Ecotones in the

Oceanic Water Column C
H G
In the water column of the deep ocean, there are
important vertical gradients in many chemical and
physical properties, including temperature, salinity,
nutrient and oxygen concentrations, light, and
2000
hydrostatic pressure. The primary production on
which all life depends is almost entirely restricted
to the surface 100 m or so, where there is sufficient
sunlight for photosynthesis (the chemosynthesis
Depth (m)

which occurs in the immediate vicinity of

hydrothermal events is the main exception, see
Chapter 10). Only about 1–3% of the organic mat-
ter produced by photosynthesis at the surface
reaches the sea bed at depths of 4000 m. So, the 4000 A 20˚N 21˚W m
general availability of food declines approximately B 42˚N 17˚W p
C 49˚N 14˚W p
D 42˚N 17˚W m
E 49˚N 14˚W m
Figure 15.6 Regression lines of biomass profiles of F NE Atlantic p
macroplankton (p) and micronekton (m) against depth at G 31˚N 25˚W p
H 31˚N 25˚W m
four locations in the northeast Atlantic (redrawn from
Angel and Baker2).

233
 M.V. Angel

15.7 Figure 15.7 An example of a

specialist species in the oceanic
plankton, the shell of a large
(3 cm long) species of pteropod,
Clio recurva, is covered with a
colony of commensal hydroid
Campaniclava clionis. (©
Heather Angel.)

but this is not translated into large-scale variability. ing to be based more on size of ‘particle’ than on
Over the time-scales that are important to the quality. Thus, an individual may occupy very dif-
creation of biodiversity, ocean waters mix rapidly; ferent niches throughout its life cycle, changing not
in the Atlantic and the Indian Oceans the waters only what it feeds on, but also the depth at which it
exchange every 250 years and in the Pacific about lives. Terrestrial species tend to be much more con-
every 500 years (see Chapter 3), times which are servative in the role they play in food-webs and in
extremely short for evolutionary changes. So, per- the space they occupy.
haps a pertinent question is ‘Are all pelagic taxa in The contrast in species richness between oceanic
the open ocean completely cosmopolitan?’ It and freshwater communities is indicative.
should be no surprise to find that they are, indeed, Freshwater environments are more fragmented and,
far more cosmopolitan than their benthic or coastal over geological time-scales, ephemeral. But fresh-
counterparts. water microhabitats are distinct and favour special-
isation; in the great lakes of Africa some of the fish
Specialists Versus Generalists are believed to have speciated within the past 500
Another factor that may help to keep the numbers years. Freshwater biomes contain a tiny fraction of
of pelagic species low is the lack of opportunity to the Earth’s water and occupy a minute area of the
specialise, to become adapted to a narrow range of land surface, and yet global species-richness of
environmental factors or a specific source of food freshwater taxa is much greater than that of marine
(for example, many insect species are associated taxa. Moreover, given long-enough geological con-
with a single food plant in terrestrial habitats). tinuity, local species-richness can be exceptionally
Palaeontologists have observed in the fossil record high. For example, in the catchment of the River
that specialist genera have much higher rates of Zaire there are over 690 species of fish, of which
turnover (extinction and speciation) than do gener- 84% are endemic16, not so very many fewer than
alist genera. In the pelagic realm, habitat variability are known from the world’s deep oceans. These
is considerable over all spatial scales, being domi- freshwater fish are often highly specialised in their
nated by the changing conditions induced by ecological requirements and have been reported to
oceanic turbulence and meso-scale eddies (see speciate within a few hundreds of years.
Chapter 4); this seems to have maintained a strong In oceanic fish, the main adaptations appear to
selective pressure for pelagic species to remain tol- be related to predator avoidance and finding
erant of these changes and hence remain as enough food within specific depth zones; in an
generalists. environment where hydrostatic pressure increases,
Similarly, species inhabiting latitudes where the the light environment changes systematically
production cycle is more seasonal have less oppor- (Chapter 14), and availability of food decreases
tunity to specialise than do those inhabiting envi- with increasing depth. In the tropics, water temper-
ronments where the supply of food is more uniform ature may change more in 1 km vertically than in
and continuous. As always, there are some excep- 2000–3000 km horizontally. So, as depth increases,
tions to every generalisation (Figure 15.7 illustrates there are systematic changes in both pelagic and
one). In addition, in the oceans food-web relation- benthic species composition, and in their physiolog-
ships appear to be very differently structured, tend- ical and morphological adaptations. If horizontal

234
 15: Ocean Diversity

Figure 15.8 Histograms showing the number of new

15.8 (a) species of (a) euphausiids and (b) mysids described in
each decade. These data suggest that the total of 86
20 known species of euphausiid is likely to be very close to
the global total, whereas the 983 named mysid species
(up until 1994) is probably well short of the global total.
The difference between the two groups is that the
(b) 140 euphausiids are all pelagic (and most are oceanic),
whereas the mysids are predominantly found in shallow,
inshore waters and are often associated with the sea bed
120 (updated from Mauchline and Murano10).
Number of new species

100 Some Examples of Pelagic

Species-Richness
80
Figure 15.8 illustrates the numbers of new
60 euphausiids and mysid species described each
decade. The euphausiid data clearly show that, since
40
the period of the great exploratory expeditions
around 1900, very few new species have come to
20 light. Euphausiids are all pelagic crustaceans (Figure
15.9), the most famous of which is krill, Euphausia
superba – the staple diet of many whales, seals, and
1780 1820 1860 1900 1940 1980 penguins in the Southern Ocean (see also Chapter
21) – and they are predominantly oceanic. Mysids,
Year
which are also shrimp-like crustaceans (Figure
gradients in the ocean were as strong as those in 15.10), are mostly either inhabitants of shallow,
the vertical, then diversity in the ocean might be
expected to rival that of the land. 15.9

Where is Species Diversity Highest

in Marine Communities?
This is not a straightforward question to answer,
partly because very few surveys attempt to identify
all the taxa present, even within the limited size Figure 15.9 A typical euphausiid, Meganyctiphanes
range of organisms collected by a single sampler. norvegica, from the North Atlantic. This species
Moreover, for certain size ranges known to be (35–40 mm in length) is known as the northern krill
highly speciose, such as benthic macrofauna and because it is a staple component in the diets of baleen
meiofauna, the systematics of most of the compo- whales in the North Atlantic. (© Heather Angel.)
nent taxonomic groups are far too sketchy for
15.10
inter-regional comparisons to be meaningful.
Another problem which militates against making
sensible inter-regional comparisons is the lack of
standardisation of sampling methods and protocols
of analysis. With the added problems created by
the absence of any international databases incorpo-
rating taxonomic information, it is well-nigh
impossible to compile all the relevant data; the true
difficulty of making even rough attempts at
answering this very basic question is all too appar-
ent. For limited areas in which standardised sam-
pling and analytical procedures have been used,
and for those groups for which the numbers of
species appears to be well-known and for which the
systematics have remained quite stable, compar-
isons are possible.
Figure 15.10 A typical pelagic mysid, Meterythrops picta
(15 mm in length), caught at a depth of 1000 m in the
northeast Atlantic. (© Heather Angel.)

235
 M.V. Angel

15.11

75;13 118;14
39; ?
70;11
270;
42 137;42
80;45 87;50
111;46
110;44
116;
44

11;11 42;11 44;16 11;11

Figure 15.11 Zoogeographical regions used by Mauchline and Murano10 in their review of mysids, together with the
numbers of mysid:euphausiid species presently known from each. There is some suggestion of a latitudinal increase in
euphausiid species toward the equator. The data for mysids are more ambivalent, but it is uncertain whether this is
representative or an artefact created by the bias with which the group has been researched.

coastal, or inshore waters, or are associated with undersampling of the Southern Ocean regions, or
the sea bed. There are a few open-ocean pelagic because there are fewer and less varied coastal
mysid species, which, as might be expected, are far areas around the Southern Ocean than at similar
more cosmopolitan than their shallow-water rela- latitudes in the northern hemisphere. Similarly, the
tives. The total of euphausiid species described up high numbers of species around Japan and the
to 1994 is 86, and this is likely to be close to the South China Sea may, indeed, be a true reflection a
actual total number in the world ocean. In contrast, richer mysid fauna in this region relative to all the
the number of known mysids species (983) has con- others – a richness that could be a result of either
tinued to increase steadily as more coastal faunas the Pleistocene oscillations of sea level or the rich
around the world are studied, and the list must still variety of coastal environments in the island archi-
be far from complete. This, of course, assumes that pelagos of the region. However, it may also be an
the presence of numerous cryptic species (those artefact resulting from the particular fascination
which cannot be separated using traditional mor- the group has had for Japanese taxonomists. Figure
phological characters, but which are distinguished 15.12 shows the numbers of species of both mysids
using molecular biological techniques) will not be and euphausiids inhabiting one or more of the
revealed when the genetics of euphausiid species regions shown in Figure 15.11. It is the pelagic,
are investigated. open-ocean species that are trans- and inter-oceanic
There is also an interesting contrast between the in their distributions and so occur in several of the
pattern of distribution of the mysids and that of the zones, whereas most of the coastal, inshore, and lit-
euphausiids. Figure 15.11 shows one way of divid- toral species are restricted to a single region. In the
ing up the world’s seas into biogeographical euphausiids, only 11 of the 86 species are neritic
regions, together with the numbers of mysid species (10 of these 11 are restricted to a single region),
known from each. There is a suggestion in the whereas the vast majority of the 75 oceanic species
southern hemisphere of there being a poleward lati- occur in more than one region. Compare this with
tudinal decrease in the numbers of species present. the 720 species (73%) of mysids which in habit
However, this may result either from the relative coastal inshore waters and occur in just one region.

236
 15: Ocean Diversity

Figure 15.12 Histograms of the number of (a) 15.12

regions occupied by (a) euphausiids (N =
10
86) and (b) mysids (N = 983). The typically
oceanic euphausiids tend to be much more
widespread than the more coastal mysids.
0

(b) 720

100

Number of species

50

0
1 2 3 4 5 6 7 8 9 10 11 12 13
Regions

Seasonality Such blooms can be readily detected in cloud-free

conditions by colour scanners in satellites (Chapter
The latitudinal limits at 40° for the regions in 5). Using satellites to record the timing of the
Figure 15.11 represent a significant change in the blooms has shown that toward subpolar and polar
ecology of the oceans. This is the approximate latitudes, the spring and autumn blooms converge
poleward limit of oceanic waters in which the near- to form a single summer peak. Following close
surface 100 m remain thermally stratified through- after these peaks in production are peaks in sedi-
out the year. During winter at higher latitudes, the mentary fluxes, which transmit the products of sea-
surface layers are cooled sufficiently for deep mix- sonal events in the surface waters to the benthic
ing to break-up the stratification and to renew sup- communities (Chapter 7). Generally, on the equato-
plies of nutrients in the surface waters. As a result, rial side of the 40° latitude line, production is not
at temperate and subpolar latitudes not only is only far less variable seasonally, but is lower,
there a marked seasonal cycle in the primary pro- except in upwelling regions where the physical con-
duction, but also the mean annual primary produc- ditions result in deep nutrient-rich water being
tion is higher. At temperate latitudes, the produc- mixed or upwelled toward the surface.
tion cycle is marked by a sharp increase in produc- The annual amount of secondary production
tion in the spring, which occurs as the seasonal (i.e., how much animal growth occurs) is related to
thermocline begins to be re-established. In most the amount of primary production, although the
years there is often another smaller peak in produc- ecological efficiency with which material and ener-
tion in the autumn, after the equinoctial storms gy is transferred from plants to herbivores and on
have started to erode the thermocline but not to carnivores and the consumers of detritus varies
destroyed it. between different ecosystems. Likewise, the stand-
These increases in production may or may not ing crop of biomass also shows some relationship
be accompanied by a sharp increase in the standing to primary productivity, but tends to be higher
crop of phytoplankton, i.e., a bloom (Chapter 6). where the production cycle is markedly seasonal.

237
 M.V. Angel

15.13 These shifts in the annual levels and seasonality of

production must play some role in the maintenance
of diversity in the communities.
Latitudinal Trends in Species Richness
Species lists often indicate that tropical and sub-
tropical faunas are richer in species than are tem-
perate and boreal faunas. However, the crude dis-
tributional data for the euphausiids and the mysids
presented in Figure 15.11 give little indication of
how these environmental changes may affect their
diversity.
More detailed data are available for planktonic
ostracods (Figure 15.13) which, while having the
advantage of being sampled using the same tech-
Figure 15.13 Specimens of the planktonic ostracod nique and also analysed with the same protocols,
Conchoecia valdiviae, which inhabits depths of are limited to the North Atlantic. Plots of the num-
600–1500 m in tropical and subtropical waters. bers of ostracod species caught in the top 2000 m
Planktonic ostracods are often the second most abundant within contiguous depth strata (100 m in the top
group of planktonic organisms in plankton samples,
occurring from the surface to the greatest depths. (©
1000 m and coarser between 1000–2000 m) at sta-
Heather Angel.) tions between 10–60°N along 20°W (Figure 15.14)
show that relatively few species are caught near the
surface, and the numbers increase with depth to
reach a maximum at around 1000 m. This is quite
different to the trend for biomass (Figure 15.6).
15.14 Species (number)
When these profiles of species number are superim-
0 10 20 30 40
posed a clear latitudinal trend emerges. At all the
0 depths sampled, the ostracod communities are rich-
▲
▲ ▲ ▲ er in species at latitudes <40°N, and are apprecia-
bly poorer at higher latitudes.
▲ ▲
▲ ▲
Analysis of the total numbers of species caught
▲
▲ ▲ ▲ in three other pelagic groups in the same set of
samples shows a similar latitudinal trend (Figure
▲ ▲ ▲ ▲
15.15), although from these data it is impossible to
400 determine whether the decrease across 40°N is
▲
▲
▲ smooth or stepped. This latitudinal trend in species
▲

▲
▲
richness runs counter to the trends in total annual
▲
▲
productivity and integrated pelagic biomass along
▲
▲ ▲ ▲ this transect. The maximum in numbers of species
occurred near 18°N. At the time of the sampling,
Depth (m)

▲
▲ ▲ ▲
the boundary between the South and North
800
▲
▲
▲ Atlantic central waters was very close to this locali-
ty, and the communities included a mixture of both
▲
▲ ▲
▲ southern and northern faunal elements. Since this
boundary moves latitudinally by 10° or so, it seems
likely that this maximum in species richness will
move with it.
▲
▲
▲
1200 ˚N
▲ ▲ 11
▲ ▲
18
30 ▲
▲
40 Figure 15.14 Profiles of the numbers of planktonic ostra-
44 cod species throughout the surface 2000 m along 20°W
▲ ▲ 53
▲ ▲ in the northeast Atlantic, showing how at latitudes <40°N
60
there are more species at all depths than at higher lati-
1600 tudes. The change in average species richness appears to
occur at the southern boundary of seasonal turnover in
the near-surface waters, and in the regions where there is
▲ ▲ ▲
a spring peak in phytoplankton production1.

238
 15: Ocean Diversity

Figure 15.15 Total numbers of 15.15

species of four pelagic taxa caught at
six stations along along 20°W in the
northeast Atlantic; at each station 14
day and 14 night samples were col-
lected systematically from the top 120 Fish
2000 m of water column1. ▲ Ostracoda
Decapoda
✦ Euphausiids

Species (number)
80 ▲

▲ ▲
▲

▲
40 ▲
✦
✦ ✦
✦
✦ ✦ ▲
0
10 20 30 40 50 60 70
Latitude (˚N)

Species Richness in Benthic Communities (a) 60 15.16

The depth profiles of species numbers for four tax-
onomic groups of benthic megafauna from the 40
Porcupine Seabight region off southwest Eire
(Figure 15.16) show much the same trends as for 20
the pelagic ostracods, with peaks in species richness
at around 1000 m in three of the groups. However, 0
the data for decapod crustaceans appear to be (b) 40
somewhat anomalous because the most species-rich
group, the brachyuran crabs, is almost entirely lim-
20
ited to shelf and upper-slope depths.
Earlier data from the eastern seaboard of the US
Species (number)

for four groups of macrobenthos (i.e., animals pass- 0

ing through 4 mm but retained by 1 mm sieves) – (c) 20
polychaetes, gastropods, protobranchs, and
cumaceans – implied (on the basis of a statistical 0
method for comparing the numbers of species in
samples of different sizes) that the maximum (d) 20
occurred in these groups rather deeper, at
0
(e) 100
Figure 15.16 Depth profiles of the numbers of species of
four abundant megabenthic taxa: (a) fish, (b) decapod
crustaceans, (c) holothurians, and (d) asteroids (starfish);
(e) is the sum of these four. These were sampled down-
slope during a major sampling programme covering 5
0
years in the Porcupine Seabight to the southwest of Eire.
All but the decapods have maxima in species richness at 0 1 2 3 4 5
about 1–2 km depth (IOSDL database). Depth (km)

239
 M.V. Angel

15.17 Figure 15.17 Depth profiles of the expect-

(a) E (S100) (b) E (S50)
ed numbers of species found in a sample of
0 5 10 15 10 15 20 25 S specimens, randomly selected from
0 0 dredge and sledge samples down-slope in
the northwestern Atlantic, of (a) proto-
1 1 branch molluscs, (b) polychaetes, (c)
cumaceans (crustaceans), and (d) gastro-
pod molluscs (redrawn from Rex14).
Depth (km)

2 2

3 3

4 4

5 5

(c) E (S150) (d) E (S68)

0 5 10 15 20 0 5 10 15 20 25
0 0

1 1
Depth (km)

2 2

3 3

4 4

5 5

2000–3000 m (Figure 15.17). As for pelagic com- equivalent to the area in square kilometres of the
munities, both benthic biomass and the density of ocean floor – over 300 million. Moreover, this
organisms decrease almost exponentially with depth study did not include the meiobenthos, which con-
(Figure 15.18), with a 10% reduction in biomass tains the two most speciose groups in the oceans –
between 200 m and 2000 m; so, once again, the the nematodes and foraminifers. Over 100 species
link between productivity and diversity is not of each of these groups were found in a cubic cen-
straightforward, since the peak in species richness timetre of mud scraped off the surface of the sea
occurs where the biomass of the benthic communi- bed at a depth of just over 1000 m off the south-
ty has fallen to nearly a tenth of that at the shelf- west of Ireland.
break. No one really knows how to extrapolate from
Some recent evidence suggests that the species the results of a few very localised samples to give a
richness of benthic assemblages may match the rich credible estimate of how many species occur on
disparity and may even be on a par with that of Earth, or even if this is possible. Many benthic
tropical rain forests. An intensive programme of species appear to have very extensive geographical
sampling the benthic communities living on the sea distributions, so estimates that they exceed a mil-
bed at depths of 2000 m off the east coast of the lion might seem excessive, but are they? It is worth
US, involving the analysis of over 200 grab sam- noting that the huge estimates of the numbers of
ples, revealed the presence of nearly 800 different species inhabiting tropical rain forests are based on
species, of which over 50% were new to science5. similar extrapolations from data obtained by mist-
When the sampling was extended both northward ing just six trees with insecticides, and so are equal-
and southward along the slope, an extra species ly lacking in credibility! But perhaps, in this case,
was added for each extra square kilometre that the the extrapolations are not so wildly excessive given
sampling programme covered. If that rate is main- the very large numbers of plants, especially trees,
tained throughout the global ocean, then the num- that occur in rain forests and the strong ecological
bers of benthic species would be approximately links between the plants and specialist insects.

240
 15: Ocean Diversity

Effects of the Seasonality Log10 gAFDW/m2 15.18

on Benthic Diversity –3 –2 –1 0
0
Are changes in the production cycle reflected in
changes in benthic diversity? We know that at tem-
perate latitudes, where the production cycle is high-
ly seasonally pulsed, the amounts of detrital materi-
al reaching the sea bed vary seasonally (Chapter 7).
So, once again, we expect there to be substantial 1
change in the benthic communities at latitudes of
around 40°. Each year the deep-sea bed to the west
of Britain, at depths of 4000 m, becomes carpeted
with detrital material 6–8 weeks after the onset of

Depth (km)
the spring bloom. This provides a food bonanza for 2
the bottom-living animals, some of whom appear
to specialise in exploiting this detritus. No such
bonanza is seen further south, either at 30°N on
the Madeiran Abyssal Plain or at 20°N on the
Cape Verde Rise. The animals which specifically
3
feed on these detrital falls are missing at these lati-
tudes, and there appears to be insufficient food
available to support some of the larger animals.
So each region is inhabited by a different assort-
ment of animals, with much fewer large species at
the more southerly sites. For example, sea cucum- 4
bers – the holothurians – which are animals that
feed on the enriched surficial sediment layer,
become rare where the sedimentary flux is too low.
Figure 15.18 Profile of benthic biomass expressed as
So their species richness runs counter to the expect- ash-free dry weight (AFDW) per square metre versus
ed latitudinal trend seen, to some extent, in other depth in the Porcupine Seabight region off southwest
groups (Figure 15.19). Fish abundances are also Eire8. The scatter is the result of the patchiness of the ben-
sharply reduced at the lower latitudes, but the spec- thic communities, especially at depths of around 1250 m,
imens there belong to an unexpectedly rich assort- where some samples contained several specimens of the
ment of species. The numbers of species taken in sponge Pheronema.
the limited sampling that has been done are lower
than those taken at temperate latitudes. However,
forcing. For example, in such muddy environments,
comparisons of the species counts in samples of a
is there always an unlimited amount of organic
standard size indicate that sample diversity is as
matter around, irrespective of the production cycle?
high as, if not higher than, those at temperate
latitudes.
The difficulties of drawing generalisations are Diversity and Productivity
further illustrated by recent results comparing mac- There is an apparent paradox in the trends in
robenthic communities from undisturbed soft pelagic diversity with depth and geographically,
muddy sediments at depths of 30–80 m off which run counter to the trends in productivity and
Spitzsbergen (78°N), in the North Sea (55°N), and biomass (standing crop) of animals, certainly in the
off Java (7°N), all sampled and analysed using water column and maybe on the sea floor.
exactly the same methods7. The results show that Populations in low-productivity regions seem to be
the diversity profiles at all three sites were indistin- characterised by being rich in species-richness with-
guishable, and gave no indication of a latitudinal out any one or two species being overwhelmingly
trend (Figure 15.20). Is this observation merely a dominant. In contrast, where productivity is high
quirk of history? In other words, were the investi- (and often more variable seasonally) far fewer
gators unfortunate in choosing sites where, by species occur and one or two species tend to be
chance, evolutionary history had resulted in the numerically dominant. Does this explain why,
diversity being identical. This solution seems far- when we fertilise the seas with our sewage and with
fetched, but is testable by repeating the comparison the agricultural run-off of dissolved nitrates and
elsewhere. An alternative explanation might be that phosphates in our rivers, the local productivity goes
the processes that maintain the diversity at these up, but the numbers of species goes down (a
three locations are not influenced by latitudinal process called eutrophication)?

241
 M.V. Angel

15.19 60 Figure 15.19 Numbers of

(a) three macrobenthic taxa
▼
expected to be caught in
▼
▼ samples of S specimens

▼
40 ▼▼
▼
▼
from deep-ocean basins in
▼
the Atlantic Ocean: (a)
E (S200)

▼▼
▼
▼ ▼
isopoda, (b) gastropoda, and
▼
(c) bivalva (modified from
Rex et al.15).

▼
20
▼
▼
▼
▼

0
(b) ▼
▼ Argentine Basin
▼▼
Cape Basin
20 ▼ ▼
▼ Angola Basin
▼ Brazil Basin
▼ ▼
▼▼
Equator
▼
▼ Guayana Basin
E (S50)

▼ ▼
▼▼
▼
▼ ▼
▼ Gambia Basin
▼ North American Basin
▼

10 ▼ West European Basin

▼
▼
Norwegian Sea
▼▼

0
20
(c)

▼
▼
▼

▼
▼
▼
E (S75)

▼
▼
10 ▼
▼▼ ▼▼▼
▼▼ ▼▼ ▼

▼
▼

▼
▼
▼ ▼
▼ ▼

0
60 40 20 0 20 40 60 80
˚S ˚N
Postscript – Genetic Diversity
Figure 22.3 shows the effects of several decades One important type of diversity, genetic diversity,
of dumping sewage sludge off Garroch Head in the has received no mention in this chapter.
Clyde on the biomass, number of species, and num- Populations living in stable environments are
ber of specimens of benthic animals across the expected to have low genetic diversity, in that they
dump site, with the central position showing a mas- show relatively little variation in their DNA and
sive increase in specimens, a smaller increase in bio- hence in the structure of their proteins and
mass, and a sharp reduction in numbers of species. enzymes. When observations were first made on
Is such an impact the consequence of ecological dis- oceanic species the results were expected to reveal
turbance regardless of its nature? It warns us that if very low levels of genetic heterogeneity, but the
we concentrate all our efforts on conserving those outcome has been quite the opposite. The genetic
regions richest in species, then we will run the dan- variability within pelagic oceanic species is proving
ger of allowing degradation to go unchecked in to be among the highest. There are numerous
those regions where ecological processes are most hypotheses to explain this high variability, but little
important in keeping the Earth habitable. in the way of proof. For example, are the meso-
scale eddies so prevalent in the ocean persistent
enough to select different genetic strains, but not to

242
 15: Ocean Diversity

15.20 100 Norse, E.A. (1993), Global Marine Diversity; A Strategy for
Building Conservation into Decision Making, Island
Press, Washington DC, 383 pp.

References
Cumulative percentage dominance

1. Angel, M.V. (1994), Biodiversity of the pelagic ocean,

Conserv. Biol., 7, 760–772.
2. Angel, M.V. and Baker, A. de C. (1982), Vertical standing
crop of plankton and micronekton at three stations in
the North-east Atlantic, Biolog. Oceanogr., 2, 1–30.
50 3. Carlton, J.T. (1989), Man’s role in changing the face of the
ocean: Biological invasions and implications for con-
servation of near-shore environments, Conserv. Biol.,
Java 3, 265–273.
North Sea 4. Flemminger, A. (1986), The Pleistocene equatorial barrier
Spitzbergen
between the Indian and Pacific Oceans and a likely
cause for Wallace’s Line, UNESCO Tech. Papers
Marine Sci., 49, 84–97.
5. Grassle, J.F. and Maciolek, N.J. (1992), Deep-sea richness:
regional and local diversity estimates from quantita-
0 tive bottom samples, Amer. Natural., 139, 313–341.
6. Hopkins, T.L., Milliken, D.M., Bell, L.M., McMichael, E.J.,
1 10 Hefferman, J.J., and Cano, R.V. (1981), The land-
Rank ward distribution of oceanic plankton and micronek-
ton over the west Florida continental shelf as related
Figure 15.20 Plots of the relative abundances of species to their vertical distribution, J. Plank. Res., 3,
in soft-bottom communities (expressed as cumulative per- 645–659.
centage dominance) at three sites, Java, the North Sea off 7. Kendall, M.A. and Aschan, M. (1993), Latitudinal gradients
Northumberland, and in a fjord in Spitzbergen. The simi- in the structure of macrobenthic communities: A com-
larity of these curves implies that the observed diversities, parison of Arctic, temperate and tropical sites, J. Exp.
expressed as the relative abundances of species, in the Marine Biol. Ecol., 172, 157–169.
8. Lampitt, R.S., Billett, R.S.M., and Rice, A.L. (1986),
bottom-living communities at these very different lati- Biomass of the invertebrate megabenthos from 500 to
tudes are virtually indistinguishable (modified from 4100 m in the Northeast Atlantic Ocean, Marine
Kendall and Aschan7). Biol., 93, 69–81.
9. Margulis, L. and Sagan, D. (1987), Microcosmos: Four
Billion Years of Evolution from our Microbial
keep them isolated? Is this why pelagic species, Ancestors, Allen & Unwin, London, 301 pp.
despite having wide geographical ranges, do not 10. Mauchline, J. and Murano, M. (1977), World list of the
show a greater tendency to split into races and sub- Mysidacea, Crustacea, J. Tokyo Univ. Fish, 64,
species? Or should the species concept be modified 39–88.
in some way to account for how oceanic popula- 11. Parrish, J. and Curtis, R.J. (1982), Atmospheric circulation,
upwelling and organic-rich rocks in the Mesozoic and
tions diverge and recombine? Cenozoic eras, Paleogeogr., Paleoclim., Paleoecol.,
On land, within time-scales of millenia, geo- 40, 31–66.
graphical localities are fixed, but in the ocean 12. Pearse, J. and Buchsbaum, A. (1987), Living Invertebrates,
‘place’ is dynamic and so has much less signifi- Blackwell Scientific Publications, Oxford, 848 pp.
13. Por, F.D. (1990), Lessepsian migration. An appraisal and
cance, even for the many benthic species which new data, Bull. l’Instit. Oceanogr., Monaco, 7, 1–10.
have a dispersive phase to their life history. Genetic 14. Rex, M.A. (1983), Geographic patterns of species diversity
studies have hardly begun to scratch the surface of in deep sea benthos, in Deep-Sea Biology. The Sea,
these problems. The removal of barriers to disper- Vol 8, Rowe, G.T. (ed.), Wiley Interscience
sion through human introduction, either purposeful Publications, New York and Chichester, pp 453–472.
15. Rex, M.A., Stuart, C.T., Hessler, R.R., Allen, J.A., Sanders,
for mariculture or by accident in the ballast waters H.L., and Wilson, G.D.F. (1993), Global scale latitu-
of large bulk-carrying vessels, may well have a far dinal patterns of species diversity in the deep-sea ben-
greater impact at this level of organisation than we thos, Nature, 365, 636–639.
are prepared for, but will it matter? There is still so 16. Ribbink, A.J. (1994), Biodiversity and speciation of freshwa-
ter fishes with particular reference to African cichlids,
much to understand before we can evaluate in Aquatic Ecology: Scale, Pattern and Process, Giller,
whether maintenance of biodiversity in the oceans P.S., Hildrew, A.G., and Raffaelli, D.G. (eds), 34th
is a cause for real concern or a red herring. Symposium of the British Ecological Society,
Blackwell Scientific Publications, Oxford, 649 pp.
17. Sepkoski, J.J. (1992), Phylogenetic and ecologic patterns in
the Phanerozoic history of marine biodiversity, in
General References Systematics, Ecology and the Biodiversity Crisis,
Angel, M.V. (1994), Biodiversity of the pelagic ocean, Conserv. Eldredge, N. (ed), Columbia University Press, 220 pp.
Biol., 7, 760–772. 18. White, B.N. (1994), Variance biogeography of the open-
Eldredge, N. (ed.) (1992), Sytematics, Ecology, and the ocean Pacific, Progr. Oceanogr., 34, 257–284.
Biodiversity Crisis, Columbia University Press, New
York, 220 pp.

243
 CHAPTER 16:

Life in Estuaries, Salt Marshes,

Lagoons, and Coastal Waters
A.P.M. Lockwood, M. Sheader, and J.A. Williams
Introduction these effects in the water column the diversity of
Any one location offshore generally presents rela- types of substratum available for attachment or
tively little change in its physical and chemical con- burrowing (Figure 16.5) and the impact of man
ditions over a short time-scale. The same cannot be (Figures 16.1, 16.3–16.5) and it is apparent that
said of near-shore waters, estuaries, and coastal such waters can be inhabited only by species capa-
lagoons, which provide some of the more varied ble of responding to physical, mechanical, and
and unstable environments on Earth. The organ- physiological challenges.
isms inhabiting such areas may experience wave Response to environmental fluctuations cannot,
action to differing degrees, tidal change, salinity, however, be restricted to mere passive tolerance;
temperature, and oxygen variations, high sediment adaptive features must incorporate a whole gamut
loads, and tidal currents of varying velocities on a of positive measures, including biochemical, physi-
daily or seasonal basis. Locations such as salt- ological, and morphological features.
marsh pools (Figure 16.1), estuarine creeks (Figure Reproductive responses, too, must be involved;
16.2), and shallow sloping beaches (Figure 16.3), it is not much use having adult stages adapted to a
can experience quite wide ranges in one or more of particular environment if the motile young lack the
their physical characteristics. Tidal progression facility to locate and colonise suitable habitats.
(Figure 16.4) may further compound the difficulties The conditions, then, are harsh, but few of the
for colonisers by imposing rapidity in the changes challenges presented by coastal waters have not
of factors such as salinity and temperature. Add to been met. True, sites such as mobile shingle banks

16.1 16.2

16.3 Figure 16.1 Estuarine raised marsh on the River Test,

Hampshire, England, showing tidal creeks and saline
pools. The creeks experience substantial salinity variation
during the tidal cycle, while the pools are exposed to
sudden changes when the surrounding grass is
submerged on high spring tides.

Figure 16.2 Closer view of the tidal creek shown in the

middle distance in Figure 16.1. This is the site of the
salinity measurements illustrated in Figure 16.8.

Figure 16.3 Estuarine beach areas are often impacted by

anthropogenic influences, as well as experiencing
physical changes.

244
 16: Life in Estuaries, Salt Marshes, Lagoons, and Coastal Waters

16.4a 16.4b

Figure 16.4 Views of an estuarine site (a) at low water and (b) at high water on a neap tide. On high water spring
tides, the raised marsh to the left would generally be submerged.

16.5 16.6

1600

Laminaria (kelp)
1400

Spartina (cord grass)

Approximate annual production [ gC/(m2/y) ]

1200

Thalassia (sea grass)

forest
Macrocystis (kelp)

Rain
1000

800

Temperate
woodland
Figure 16.5 The variable habitats presented by a tidal 600

Upwelling zone
creek. Rocks with fucoid algae provide refuge at low

Estuarine sub-littoral
water for non-burrowing forms, such as the gammarid

Continental shelf sea

amphipods; the mud banks are burrowed by the annelid 400
Nereis diversicolor, the amphipod Cyathura carinata, and
by the non-feeding adults of the isopod Paragnathia
Open sea
formica. The prawn Palaemonetes varians and the mysid 200
Neomysis integer may be found where deeper pools are
left at low water.

and salt-marsh pools have a restricted fauna, but Figure 16.6 Estimates of primary production (gC/m2/yr)
no coastal environment is without any colonisers. for various marine and terrestrial ecotypes. Note that val-
So successful, indeed, are some of the inhabitants ues vary widely with both conditions and plant density.
that shore and near-shore regions have the highest The figures indicate, first, that the productivity of sea-
primary productivity per square metre of any grasses and the fixed algae of the coastal edge can be as
high as that of tropical rain forest and, second, that the
marine area (Figure 16.6). unit area productivity of offshore waters, even in the
There is, of course, no single answer to the prob- nutrient-rich upwelling zones, tends to be considerably
lems posed for organisms in the coastal and shelf lower.
seas, nor is there any single simple lifestyle. In this
chapter, therefore, we select three diverse topics to
illustrate the amazing versatility of estuarine and
coastal forms. These include:
• The physiology of osmoregulation – the regula- importance of the geological and hydrodynami-
tion of water and ion balance by organisms – to cal regime for benthic animals.
indicate the variety of solutions to the problems • Patterns in community structure of zooplank-
generated for living forms by variation in the ton, to interrelate the changing populations
chemistry (salinity) of the habitat. with features of the physical environment and
• Animal–sediment interactions, to outline the biotic factors.

245
 A.P.M. Lockwood, M. Sheader, and J.A. Williams

16.7 2 Figure 16.7 Waters

Surface 3
isohalines with a salinity less
250 (‰)
4 than that of sea
5 water tend to show
Finland
3 2 depletion of faunal
5 4 variety. In part this is
200 Sweden 6 for physiological
6
30 reasons (see text), in
20 part because of habi-
Number of species

15 7 tat restrictions. The

150 figure illustrates such
8
10 a decline in the
number of species of
four groups of mol-
100 Snails lusc in relation to
Sea slugs the salinity in the
Bivalves
Squid Baltic (derived from
numerical data by
50 Jaeckel, see Remane
and Schlieper16). The
inset shows the
salinity at different
locations, measured
35 30 25 20 15 10 5 0 in parts per thou-
North Kattegat Belt Outer Gulf of sand.
Sea Sea Baltic Bothnia
Salinity (‰)

The Physiology of Osmoregulation

The body surface of most marine organisms is terms of the penetration of animals into estuaries.
rather permeable to water and inorganic ions. Just how far various marine species can extend their
Consequently, few species have the capacity to range depends not only on the physiology and
maintain the extracellular fluids at a concentration lifestyle of the organism, but also on the morpho-
significantly different from that of the waters they logical and hydrological characteristics of the estu-
inhabit. As there is also generally rather little toler- ary. Where the estuarine shore is shallow sand or
ance to dilution by the cells of most marine forms, mud flats and the river flow is small by comparison
the consequence is that few species have the capaci- with the tidal incursion, then the mid-to-upper
ty to extend their range into the waters of dilute shore generally only experiences undiluted water
seas, estuaries, and coastal lagoons. when immersed. Such areas may be colonised by
The point is well-illustrated with reference to the animals with little tolerance of dilution, provided
Phylum Mollusca. This group is represented by they can survive or avoid aerial exposure at low
numerous species in the saline waters of the North tide. By contrast, the channels carrying the river at
Sea, but the dilute waters of the Baltic create prob- low water may represent a greater challenge to
lems resulting in the elimination of all but the colonisers by virtue of the rapid and profound
hardy few (Figure 16.7). Even those forms which changes in salinity that occur over the tidal cycle.
can penetrate into the middle Baltic, where salinity This point is well-illustrated by data from a creek
is only about 20% of that in the North Sea, are on the River Test in the UK, where salinity in the
often but a shadow of their relatives from more channel may vary from virtually fresh water to
saline waters. The common blue mussel Mytilus some 65% that of North Sea water and back again
edulis in the waters south of Stockholm only grows over a single tidal cycle (Figure 16.8).
to about half the size of those from the North Sea. The colonisers of such variable waters have but
Also, its shell, instead of being blue, robust, and three choices:
opaque, is often fragile, partially translucent, and
• Balance passive osmotic uptake of water at low
brownish. The Baltic Herring, the strömming,
salinities with passive osmotic withdrawal of
(beloved by Swedish epicures after it has matured
water at high tide; altogether a rather high-risk
packed in leaves below ground for some weeks) is
strategy that can only work over a restricted
similarly but a poor cousin, in terms of size, to the
part of an estuary. Nevertheless, it appears that
North Sea herring16.
the little flatworm Procerodes (Gunda) ulvae,
Osmotic and ionic levels also pose problems in
which typically lives around the neap tide high-
246
 16: Life in Estuaries, Salt Marshes, Lagoons, and Coastal Waters

16.8 (a)
22 (b) Low water, 230 m from high water mark
40
140 m from high water mark
20 + High water mark

18

16 30

14
Salinity (‰)

Salinity (‰)
12
20
10

8
+
6
10
4

2
+
+ +
0
15 10 5 0
0 1 2 3 4 5 6 7 8 9 10
Depth in sand (cm)
Time (h)

Figure 16.8 (a) Changes in the salinity of a salt-marsh creek at Totton, UK, during a spring-tide cycle (based on
Ralph14). Note that the salinity varies from about fresh water to some two-thirds sea water within the cycle, imposing
harsh physiological conditions on exposed fauna (Figure 16.2 illustrates the actual site). (b) Salinity of the interstitial
water of the sediment underlying a stream crossing the intertidal region of a beach (drawn from data in Reid15). The
distances are measured from the high water mark. Note the decrease in salinity variation with depth in the sediment.

water mark in small streams, is largely such an of the region from mid-estuary to fresh water
osmotic yo-yo. use some form of regulatory mechanism(s). The
• Avoid the problem by burrowing. This option systems involved are considered shortly, but
takes advantage of the much slower rates of first mention must be made of the effects on
change of salinity in the interstitial water than colonisation of the configuration of estuaries.
in the overlying medium, Figure 16.8(b). The Where the geomorphology is such that the estuary
use of this strategy enables some burrowing basin contains saline water at low tide, as is the
annelids – segmented worms – to penetrate con- case in the seaward reaches of most estuaries, then
siderably further along estuaries than can one of two situations arises. Either the water col-
epibenthic species exposed to the full vicissi- umn is well mixed so that there is effectively no
tudes of tidal salinity change. vertical stratification of salinity or, alternatively,
• The third option is to regulate the body fluids especially where freshwater input is relatively large,
independently, or at least partially independent- a stratified water column occurs in which a more
ly, of the concentration of the medium. saline salt wedge penetrates upstream with fresher
Virtually all the benthic and pelagic colonisers water above (Figure 16.9). The latter situation

16.9

Salinity (‰)
1 5 7 10 15 20 25 30
Water surface
River Mixing zone
Less than 1‰
Figure 16.9 Stylised repre-
sentation of a stratified estu-
&
ary, showing the wedge of Sea
35 ‰
more saline water penetrating
upstream under outflowing
dilute water.

247
 A.P.M. Lockwood, M. Sheader, and J.A. Williams

16.10 which inorganic ions can participate in effector

reduction; a substantial part of any change is there-
fore generally accomplished through a decrease in
100 the concentration of organic molecules. Much of
the variation relates to the removal of non-essential
amino acids, such as glycine, glutamic acid, ala-
[Haemolymph] (% sea water)

nine, and proline. This removal is effected either by

conversion into organic acids, which can then be
metabolised, or by allowing diffusion across the
cell membrane.
The process is reversible, so when the medium
becomes more saline inorganic ions play an impor-
tant role in directing the activity of the enzymes
involved in the conversion of organic acids into
Hypo–hypertonic amino acids. Amino acids are also then taken up at
Hyper–isotonic
Conformers an increased rate across the cell surface (see Gilles,
Hypertonic 1979).
The majority of osmoconformers have a limited
tolerance range.
0 100 Osmoregulators
[Medium] (% sea water)
In contrast to osmoconformers, osmoregulators
Figure 16.10 Comparison of the haemolymph–medium maintain body fluids at a level differing from that
relationship for various types of regulation. Where the of the medium over at least part of their tolerance
haemolymph concentration is above or below the isos- range. Figure 16.10 illustrates the three known
motic line, the animal actively regulates the concentra- types of relationship between concentration of the
tion of its extracellular fluids. medium and concentration of body fluids.
Hypotonic refers to something less concentrated
than the reference medium, in contrast to hyperton-
tends to favour the upstream penetration of marine
ic – more concentrated. Hyper-isotonic regulators,
forms.
as the name implies, are conformers at high salini-
Osmoregulatory adaptations ties, but hold the body fluids at a higher concentra-
tion than that of the medium at lower salinities.
All marine forms which colonise areas subject to
The shore crab Carcinus maenas is an example of
dilution tend to have a relatively wider tolerance of
such a regulator. Hypo-hypertonic regulators have
salinity change than do their fully marine relatives
body fluids which are maintained at a lower con-
(i.e., they are ‘euryhaline’ rather than ‘stenoha-
centration than the medium at high salinities and at
line’). Tolerance is found both in species where the
a higher concentration when the external concen-
body fluid concentration tends to follow that of the
tration is low. They thereby keep the blood and cell
surrounding medium as the latter changes (osmo-
concentrations within a relatively limited range,
conformers) and in those capable of at least some
thus placing less of a requirement for osmotic con-
maintenance of body fluid concentration at levels
trol on the body cells themselves. Typical examples
differing from that of the medium (osmoregula-
of such hypo-hypertonic regulators include the
tors).
euryhaline bony fish and some species of prawn,
Osmoconformers isopods, and mysids. However, it is the brine
Soft-bodied forms, such as worms and molluscs, shrimp Artemia salina which is the undisputed
tend to have a high surface permeability and most prima donna in this role, since it still maintains its
are incapable of maintaining body fluids more con- internal concentration at a low level even when in
centrated than the medium. Those with any appre- saturated brine. The third, and final, class of regu-
ciable tolerance of dilution must therefore regulate lation involves maintenance of hypertonicity to the
at the cellular level. The problem is simple. The medium over the entire tolerance range. So far this
cells contain osmotic effectors, mostly inorganic category is solely represented by the larva of the
ions and small organic molecules, and the cell walls isopod Paragnathia formica. This larva is parasitic
are highly permeable to water. Consequently, on on euryhaline fish, such as young flounder, and
dilution of the extracellular fluid, water moves into potentially has an osmotic problem should the fish
and expands the cells, unless steps are taken to swim seaward. However, by being always signifi-
reduce the amount of internal osmotic effectors. cantly hypertonic to any medium to which it is
The need to retain the electrical excitability of acclimated, it gives itself at least some buffer before
cells such as nerves and muscles limits the degree to the risk of dehydration.
248
 16: Life in Estuaries, Salt Marshes, Lagoons, and Coastal Waters

16.11 nal salinity changes. The amphipod crustacean

Gammarus duebeni is a case in point. When in 2%
18
sea water the permeability to water is less than half
Half-time (min) for exchange of total body water

16
that when it is in sea water (Figure 16.11).
Even dynamic changes in permeability are not
14 alone sufficient to maintain an osmotic gradient
across the body surface. Additional processes are
12 also required, which include:
10 • Conservation of ions in the body by the produc-
tion of urine less concentrated than the body
8 fluids.
• The active uptake of ions across the body
6 surface4.
4
Hypotonic urine
2 The representatives of many animal groups inhabit-
ing fresh water have developed the capacity to
excrete urine which is more dilute than the blood.
20 40 60 80 100 120 140 By this means they can remove water brought into
[Medium] (% sea water) the body by osmosis while conserving valuable
ions. By contrast, in brackish waters only fish and a
Figure 16.11 Variation in permeability to water with few species of amphipod are known to have a
salinity in Gammarus duebeni (insert; after Lockwood et
al.10). The graph shows the time for half the body water
significant capability in this respect (Figure 16.12).
volume to be exchanged (using tritiated water as a mark- Active transport of inorganic ions
er), with the animals acclimated to different salinities. The active regulation of internal ion levels is a fea-
The half-time for exchange is taken as a reciprocal func- ture of animal cells; derivations from this system
tion of the permeability. The error bars represent ± one
standard deviation.
have allowed transport of ion species, particularly
the major ions sodium and chloride, across surface
membranes4. In marine forms isosmotic with their
Inevitably, hypertonic regulation carries with it medium such transport may be associated with the
the twin penalties of uptake of water by osmosis maintenance of some ions at levels of concentration
and loss of inorganic ions across the body surface differing from those in sea water. Additionally, ion
and in the urine. Maintenance of the status quo of transport sometimes appears to be associated with
the body fluids therefore demands the expenditure the need to bring water into the body for urine
of energy to counteract these effects. The larger the production or replacement of fluid lost by haemor-
gradients and the greater the permeability the more rhage. Similarly, the expansion of some amphipods
substantial is the energy need, so it is advantageous after moulting in sea water seems to be associated
to decrease the permeability. Unsurprisingly, there with increased ion uptake.
is a general tendency for coastal and estuarine To adapt an ion-transport system, originally
forms to be less permeable than offshore marine designed to transport water, to the new role of
species. In some cases the permeability to water can maintaining the body fluids hypertonic to the medi-
be varied at the level of the individual if the exter- um would seem to involve rather little further

400 16.12
Blood and urine concentrations

350
300
(mM / l. NaCl)

Figure 16.12 Production of urine hypotonic to the 250

haemolymph at low salinities in Gammarus duebeni 200
(after Lockwood9). The solid dots represent individual ne
c li
urine samples taken directly after micturition from the 150 toni
urinary papilla. The hollow circles are blood samples. Iso
100
The osmotic pressure of the tiny samples (volumes as
small as 10 –3 mm 3) are measured using the Ramsay 50
cryoscope. Production of hypotonic urine, when the
haemolymph is more concentrated than the medium, 0 50 100 150 200 250 300
may be expected to assist in the conservation of inorgan- Sea water concentration
ic ions in the body. (mM / l. NaCl)

249
A.P.M. Lockwood, M. Sheader, and J.A. Williams

change; perhaps no more than a variation in the coarse sediments, it follows that the available
relative permeability of the body surface to ions organic resource, organics plus bacteria, is likely to
and water. A small step in evolution, perhaps, but be greater in fine rather than coarse sediments.
no less important for that. Had such a capability Suspension feeders require a sediment which will
never evolved it is improbable that life as we know provide stability and support, but are also depen-
it in fresh water and on the land would ever have dent on particulate food resources in the near-
developed. bottom water. Both requirements are in turn depen-
dent upon the local hydrodynamics and sedimen-
Animal–Sediment Interactions in Coastal tary transport processes.
and Shelf Environments The trophic structure (the relative importance of
Although rocky substrata are not uncommon in lit- different feeding types) and species composition are
toral and near-shore areas, most of the sea floor in thus affected to a substantial degree by the sedi-
shelf and coastal environments is composed of sedi- ment structure and dynamics. However, benthic
ments. Unlike the deep sea, where the substratum is organisms are not merely passive responders to
predominantly fine mud, coastal areas are charac- environmental influences, but are themselves able
terised by sediments that are extremely variable to actively exert effects on the sediments.
(clay, silt, sand, shingle, cobbles, and boulders or Modification of the sediment can result from a
various admixtures of these), reflecting their prox- number of biologically mediated processes. Of
imity to sediment sources and the more dynamic these, bioturbation, biodeposition, and biosecre-
sediment transport processes that occur in the shal- tion, as described below, are the most important.
low waters of the coastal and shelf regions. Each process acts to change the structure and prop-
Sediments, especially in shallow waters, are gen- erties of the sediment and this, in turn, may affect
erally nutrient rich and teem with life. Benthic the range of potential colonising species.
organisms (i.e., those living in or on the sea floor) Bioturbative processes lead to the movement of
are a diverse group, and species can be found in sediment as a direct or indirect result of the activity
association with all types of sediment. While the of organisms. Animals may select particles of a par-
greatest variety of species is usually found in sedi- ticular size to line their burrows or build tubes in
ments with a broad range of particle size, the great- which to live, and burrow walls may be compacted.
est density of organisms and highest productivity Both processes may locally change sediment prop-
tend to be associated with littoral, or near-shore, erties around the burrow. Animals burrowing,
sediments in sheltered environments. Most estuar- moving through the substratum, or simply grow-
ies fall into this category; the high animal produc- ing, also displace sediment. However, the greatest
tivity reflects the wide range of food resources rates of bioturbation are caused by large deposit
available, both from within and imported from feeders feeding on and passing sediment through
outside the system. their guts. Examples are found in most of the major
The organisms found in sediments range in size benthic groups, notably the polychaete worms, the
from bacteria to megafauna. Bacteria and smaller thalassid shrimps, the sea cucumbers (holo-
protozoa are usually associated with organic mater- thuroids), the heart urchins (spatangoids), and the
ial and particle surfaces, and represent a food bivalves. As a result of the activity of these deposit
source for many of the larger organisms. Bacteria feeders, rates of sediment reworking may be as high
have great enzymatic versatility and are able to as several kilograms per square metre per year, and
digest a wide range of organic materials which, the whole of the top few centimetres of sediment
though not directly available to these larger species, may be reworked several times over the same time-
become available as microbial biomass. scale.
Most of the animals living in sediment feed Intensive reworking changes the sediment struc-
either directly on sediment (deposit feeders) or on ture. Many of the finer particles are ingested and
particles collected from the water overlying the sed- become bound together in faecal pellets, thereby
iment surface (suspension feeders). On the conti- effectively increasing the particle size (Figure
nental shelf, deposit-feeding species reach their 16.13). The resulting sediments have a higher water
greatest diversity in fine, muddy sediments and content and a lower bulk density, and so are more
their greatest abundance in sheltered inshore muds, easily eroded by relatively weak currents 17. An
such as those of estuaries, lagoons, salt marshes, example of biologically mediated change in sedi-
and mangrove swamps. One reason for this relates ment structure is given in Figure 16.14, which
to availability of food resource. Organic matter and shows seasonal changes both in sediment shear
bacteria within sediments are both often associated strength and in bioturbation rate by the polychaete
with particle surfaces and, since the surface area of worm Arenicola marina.
fine particles is greater per unit volume than that of The development of graded bedding is another

250
 16: Life in Estuaries, Salt Marshes, Lagoons, and Coastal Waters

Figure 16.13 A sample of 16.13

surface sediment taken
from a productive coastal
lagoon in the UK. Almost
all the surface sediment
has been pelletised (indi-
vidual faecal pellets are
about 300 μm in length)
by the feeding activities
of the fauna.

frequent consequence of bioturbation2. This fol- tion of organisms. Indeed, the effect of such
lows from deposit feeders, such as Arenicola, enhanced flow can be up to two orders of magni-
selecting finer particles to eat and expelling the fae- tude greater by volume than that of particle biotur-
cal waste at the surface. The result is a surface layer bation. Together, the two processes alter the sedi-
of finer particles overlying coarser materials accu- ment chemistry and influence the depth of the
mulating at the maximum feeding depth. redox layer (the boundary between the oxic and
Fluid movement through sediments may also be anoxic sediment).
enhanced as a result of the excretion and respira- Biodeposition results from the activity of suspen-
sion feeders that remove particles from the water
16.14 100 6 column and usually expel faecal pellets onto the
sediment surface, thereby increasing the rate of
5
deposition of particles to the sediment. Many sus-
pension feeders, and some deposit feeders, build
Sediment shear strength (k Pa)

75
Bioturbation rate (ml/m2/d)

tubes that extend above the sediment surface. For

4 suspension feeders, raising the feeding structure
above the surface in this way is advantageous in
50 3 that both the water flow around the feeding
appendages and the rate of encounter with food
particles increase.
2
Tube structures have a marked effect on the sta-
25 bility of sediments, though the extent of the impact
1 is dependent on their size and abundance3. Low
tube densities may result in increased turbulence in
0 the lee of each tube, with a consequent sediment
J J A S O N D J F M A M J J scour and erosion at low or moderate current
Month velocities. By contrast, high tube densities can
result in a skimming flow over the top of the tubes,
Figure 16.14 The seasonal pattern of bioturbation (blue which reduces both flow and erosion at the sedi-
line) by the polychaete worm Arenicola marina in a UK ment surface. Fine particles are able to settle in the
coastal lagoon. Changes in the shear strength (red line) of low-flow regime between the tubes, often resulting
the top 5 cm of sediment mirror bioturbative activity. in sediment accumulation and a fall in the mean

251
 A.P.M. Lockwood, M. Sheader, and J.A. Williams

16.15 Figure 16.15 Tube beds of the poly-

chaete worm Pygospio elegans in the
Baie de Somme, France. The edge of the
raised tube bed is being eroded to form a
‘cliff’ about 15 cm high. The densely
packed tubes constructed by the worms
are 0.5–1.0 mm in diameter.

sediment grain size. Figure 16.15 shows such a high species, ranging from viruses, bacteria, and single-
abundance of tubes, in this case the polychaete celled plants and animals to jellyfish, some crus-
worm Pygospio elegans, on an estuarine sand flat taceans (Figure 16.18) and molluscs, and the early
at a density in excess of 100,000 tubes per square life stages of fish. Some are relatively non-motile,
metre. At such high densities the whole bed while others are capable of reasonably sustained
becomes raised above that of the surrounding sand movement. What connects the members of the
as a result of the accumulating fine sediment, plankton, however, is that their capacity to propel
together with the continual extension of the worm themselves tends to be small relative to the water
tubes. The end result is a change in the structure currents and turbulence they experience. To some
and properties of the sediment (Figure 16.16). extent, therefore, their distribution is related to the
Biosecretion by benthic organisms is the release mass movements of the water in which they live.
of organic materials, either as waste products or as Together, the plankton members make up an
functional secretions. These organic materials pro-
vide both a food resource for micro-organisms and
Shear strength (k Pa) 16.16
a means for altering sediment chemistry. Of great-
est importance with regard to change in sediment 5 10 15 20 25
properties is the secretion of mucopolysaccharides, 0
or mucus. Mucus can bind sediment particles
together and thereby increase the stability of sedi-
ments. Many species produce mucus, but particu-
Depth below sediment surface (cm)

larly important contributors are micro-organisms, 5 Pygospio bed

especially bacteria and single-celled organisms. In
littoral and shallow sublittoral areas, micro-algae,
such as diatoms, are also contributors to the
mucus-binding of sediment particles (Figure 16.17). 10
In polar, temperate, and certain tropical shelf Sandflat
and coastal environments, bioturbation, biodeposi-
tion, and biosecretion show clear seasonal patterns,
which are reflected in annual cycles in the structure 15
and properties of sediments (Figure 16.14). Also
influenced by seasonal cycles are the planktonic
communities, from which much of the organic mat-
20
ter providing the energy input for the benthic com-
munity ultimately derives.
Figure 16.16 Changes in shear strength with sediment
Estuarine Plankton: Patterns of depth in Pygospio beds and in adjacent sediment con-
Community Structure taining low numbers of the polychaete worm. The dense
The planktonic community of the water column aggregation of tubes (more than 100,000/m 2) greatly
consists of an amazing spectrum of microscopic alters the surface sediment characteristics and properties.

252
 16: Life in Estuaries, Salt Marshes, Lagoons, and Coastal Waters

Figure 16.17 Diatom blooms on the 16.17

surface of an estuarine mud flat. Such
blooms are important in binding sedi-
ment particles by producing mucus.

intricate food web which transfers the photosyn- The plankton of estuaries and coastal water
thetic primary production of the phytoplankton in require similar environmental conditions for sur-
the upper water column through a number of levels vival and reproductive success as do those in open
within the zooplankton to the top pelagic water. In addition, however, they must have the
predators13. Transfer of energy also occurs down capacity to respond to the much greater degree of
the water column through the food web to sustain environmental physicochemical variability of the
the benthic communities, in the ways already dis- near-shore waters, which imposes daily and season-
cussed. al patterns that in turn influence their population
dynamics6. Adaptational responses to the environ-
16.18
mental changes are likely to involve complex ener-
3
➞ gy-demanding processes, but off-setting this disad-
vantage, as far as the phytoplankton are concerned,
is the high availability of nutrients in estuaries,
which leads to high productivity (Figure 16.6).
Similarly, the many environmental and biologi-
cal factors described earlier all combine to restrict
the variety of zooplankton species that occupy
➞
1 these waters by comparison with the varieties of
the open sea. By contrast, the higher levels of pri-
mary production make for high levels of abundance
of those zooplankton species which are able to
respond to the physicochemical variability of estu-
aries. High overall levels must not be taken to
mean that there is a uniformly high population
throughout estuaries. Patchiness is a common fea-
ture, with rapid differences in population densities
occurring over short distances. Great care has to be
2 taken when sampling with standard nets to ensure
the correct interpretation of the distribution of the
various elements in the plankton.
In temperate latitudes there are dramatic season-
al changes in both productivity – the rate of
increase of the weight of biological material per
unit area per day (mg/m2/d) – and standing crop –
Figure 16.18 Typical sample of crustacean zooplankton
illustrating the relative size differences between (1)
the weight of biological material per unit area
calanoid copepods, (2) decapod zoeae larvae, and (3) (mg/m 2). The classic pattern for coastal waters
nauplii larvae – total body length of the copepod (1) is involves a low phyto- and zoo-plankton abundance
approximately 1.5 mm. and production during the winter. In spring, there
is a very rapid production of phytoplankton, the

253
 A.P.M. Lockwood, M. Sheader, and J.A. Williams

16.19 Figure 16.19 Idealised relation-

ship between phytoplankton (P)
and meso-zooplankton (Z) sea-
sonal abundances in temperate
coastal waters. The temporal
pattern does not represent the
relative abundances of phyto-
and zoo-plankton.
P Z

Winter Spring Summer Autumn

spring bloom, which is followed by a summer Also, often observed coincident with peak chloro-
decline in the standing crop of phytoplankton as phyll levels in early summer are blooms of the pho-
the population of zooplankton grazers rises (Figure tosynthetic ciliate Mesodinium rubrum.
16.19; see also Chapters 6 and 15). These grazing The high levels of primary production through-
zooplankton decline in turn during the summer as a out the spring–summer period mean that the herbi-
result of food limitation, known as bottom-up con- vore-dominated zooplankton community is proba-
trol, and/or the development of predator popula- bly not significantly food-limited during this time.
tions, known as top-down control. Often, both Nevertheless, these zooplankton populations do
phytoplankton and zooplankton exhibit a smaller show both spatial and seasonal variations.
peak of abundance in coastal waters in the autumn. Two elements contribute to the zooplankton
As previously indicated, the overall situation in
estuaries tends to be more complex, as can be illus- (a) Total diatoms 16.20
2500
trated with reference to Southampton Water in 0
England.
Schroederella delicatula

Number of cells
Rhizosolenia delicatula
Here, riverine and other inputs ensure that nutri-
Number of cells (x107/m2)

650 650

(x107/m2)
ent supply is less limiting than in offshore waters. 0 0

In consequence, the spring phytoplankton bloom is 2500

Skeletonema costatum Thallassiosira cf fallax
500

prolonged and is represented by a succession of dif- 0 0

ferent species; typically, there is an initial succes- 300

Mesodinium rubrum
sion of diatom species, which are numerically dom- 0

inant in March–April. Following on is an increase Microflagellates

in the dinoflagellate and microflagellate communi- 1250

0
ty, which continues at high densities throughout
J F M A M J J A S O N D
the summer until September–October [Figure
16.20(a)]. Overall, therefore, despite fluctuations in (b)
abundance associated with the species succession, Julian days
there is an essentially prolonged bloom of primary 0 100 200 300

400 400
producers through from spring to autumn. This
Total phytoplankton
extended bloom is well-illustrated by the levels of
Total phytoplankton cells (108/m2)

▲ ▲ Chl-a
▲

photosynthetic pigment chlorophyll-a in the water 300 300

column over the annual cycle [Figure 16.20(b)].
Chl-a (mg/m2)

▲
▲
200 200
▲

Figure 16.20 (a) Phytoplankton species succession in the ▲

middle part of Southampton Water, England, indicating 100 100

the dominant Dinoflagellate species and total diatom, ▲
▲

▲
▲

microflagellate, and Mesodinium rubrum temporal sig- ▲ ▲

▲
▲
▲
▲

▲ ▲ ▲

nals. The cell abundances are integrated throughout the 0 ▲ ▲

▲ ▲ ▲ ▲▲ ▲ ▲ ▲ ▲ ▲
0

upper 6 m of the water column. (b) Seasonal pattern of J F M A M J J A S O N D

chlorophyll-a and total phytoplankton cell density. Months

254
 16: Life in Estuaries, Salt Marshes, Lagoons, and Coastal Waters

Figure 16.21 Seasonal patterns of (a) 16.21

meso-zooplankton in the mid-estuary 10000

(numbers/m3)
area of Southampton Water, England,

Abundance
7500
for 1990–1991. The community is pre-
sented in terms of (a) abundance and (b) 5000
the percentage composition of the most 2500
significant elements of the meso-zoo- 0
plankton.
(b) 100

90

Composition of 'total' zooplankton (%)

80

70

60

50

40

30

20

10

0
A M J J A S O N D J F M A M J J A S O N D J

Time (months)

Calanoids Nauplii Rest

larvae

grazers, the ‘holoplankton’, which spend their meso-zooplankton is made up of meroplanktonic

whole life cycle in the plankton, and the ‘mero- larvae of benthic polychaetes, molluscs, and crus-
plankton’, which are the larval stages of benthic taceans, and of pelagic gelatinous species, all of
forms. Overall, in Southampton Water the holo- which appear and disappear sporadically during
plankton, and in particular the copepod crus- the spring through to autumn.
taceans, dominate for much of the year. While this basic temporal pattern is presented
Following low winter densities of zooplankton throughout the estuary, there is also a spatial com-
there is a rapid rise in abundance throughout the ponent to the overall community structure, with
estuary in spring coinciding with increasing temper- the lower part of the estuary exhibiting a more het-
atures and abundance of phytoplankton. The zoo- erogeneous species composition than the upper.
plankton population then declines in June through Changes in abundance, as well as in species compo-
to August, often by as much as 75–90%. A second sition, are also apparent11. Thus, while copepods
and smaller peak of abundance sometimes occurs dominate the upper estuary, their numbers decline
during autumn, between September and late toward the mouth where they represent only 50%
October. The calanoid copepods, which typically of the total zooplankton population.
constitute 80–95% of the meso-zooplankton, mir- Considering, therefore, that the sustained phyto-
ror the pattern of the overall population change18 plankton bloom should provide sufficient food for
(Figure 16.21). The nauplii (copepod and barnacle the meso-zooplankton, it is probable that factors
crustacean) larvae population are the second most such as food limitation do not play a major part in
dominant component of the zooplankton of regulating the meso-zooplankton community. What
Southampton Water. Their numbers are relatively can be deduced about other factors which might
high throughout the spring and summer, but do not lead to the summer decline of the calanoid cope-
exhibit the summer decline shown by the adult pods, in particular within the estuary?
copepods. The nauplii, therefore, can make up Two factors which could contribute to this sum-
45–80% of total zooplankton abundance during mer decline are, first, the Mesodinium red tide and,
the summer (Figure 16.21). The remainder of the second, predation by gelatinous species.

255
A.P.M. Lockwood, M. Sheader, and J.A. Williams

16.22

5 km

Figure 16.22 LANDSAT image of Mesodinium red-tide bloom in the upper

reaches of Southampton Water, England; the sub-surface bloom is represented by
a red–purple signal. Sites 1 and 2 refer to typical upper and mid-estuary sampling
stations. The estuary runs from NW–SE, with the Solent toward the bottom right.

Coincident with the decline in copepod numbers cantly exclude either zooplankton or juvenile fish
is the appearance in the estuary, in high numbers, from the vicinity of blooms5. There is, indeed, no
of the photosynthetic ciliate Mesodinium rubrum, obvious inverse relationship between zooplankton
which forms a non-toxic red tide bloom, particular- numbers and either Mesodinium cell density or the
ly in the upper estuary7 (Figure 16.22). During late reduced oxygen profile in the water column, as can
May to early August, Mesodinium is patchily dis- be seen by comparing Figures 16.23(a) and 16.23(b).
tributed throughout the estuary. In bloom condi- Any cause-and-effect relationship between
tions, particularly in June and July, cell densities Mesodinium blooms and summer decline in zoo-
can rise to more than 5000 cells/ml, and in these plankton can therefore be questioned. Red tides,
circumstances zooplankton density and distribution caused by other species, have, however, produced
may be influenced by the effect of Mesodinium on more severe effects elsewhere (see Chapter 22).
the water column. Mesodinium undergoes diurnal Also coincident with the period of reduced
vertical migrations. During the day it is typically meso-zooplankton abundance during the summer is
located in a layer just below the surface and its the appearance of populations of gelatinous preda-
photosynthetic activity can supersaturate the upper tors, jellyfish, and ctenophores, which have distinct
water column with oxygen. Downward dispersal of peaks of abundance between March and late
the bloom at night, coupled with respiration, can, September11. These organisms are acknowledged as
in contrast, result in the deeper waters becoming important predators in the marine plankton ecosys-
sub-oxic [Figure 16.23(a)]. There is substantial anec- tem 1,8 . All the gelatinous predators found in
dotal evidence that these red tides result in reduction Southampton Water are generalist feeders, in that
of fish numbers. Critical analysis of the situation they catch and consume a wide range of both
suggests, however, that Mesodinium does not signifi- species and prey size.

256
 16: Life in Estuaries, Salt Marshes, Lagoons, and Coastal Waters

0 1 2 3 4 5 ✕ Zooplankton density (x103/m3 10

16.23 0 3 6 9 12 15 Mesodinium cells (x102/ml1) Aurelia aurita 16.24
40 60 80 100 120 140 02 Saturation (%) 8
24 26 28 30 32 34 Salinity (‰)
12 14 16 18 20 22 Temperature (˚C)
(a) 6
0
1 ✕ 4
2
2
3 ✕

4 0
Depth (m)

5 ✕ 2.0 Pleurobrachia pileus

6

7 ✕
1.5
8
1.0

Abundance (numbers/m3)
9 ✕

10 0.5

0.0
(b) 12 Clytia hemispherica
0
1 ✕
10
2 8
3 ✕ 6
4 4
Depth (m)

5 ✕
2
6
0
7 ✕
4 Sagitta setosa
8
9 ✕
3

2
Figure 16.23 Vertical profile of the water column during
(a) Mesodinium bloom conditions and (b) non-bloom 1
conditions. The nature of the distribution pattern and
abundance of meso-zooplankton is similar in both 0
situations. O N D J F M AM J J A S O N D J F M AM J J A S O N D J

Figure 16.24 Typical seasonal abundance pattern of Months

gelatinous meso-zooplankton predators in mid and upper
Upper Southampton Water
Southampton Water, England. Species are not drawn to
Mid Southampton Water
scale.

The jellyfish Aurelia aurita is invariably the first on a normal basis and up to as much as 70% when
to appear in the plankton, being seen first as the there are predator swarms12. The evidence therefore
young ephyra stage in early March and growing on suggests that the gelatinous predators play a signifi-
until the end of June12 (Figure 16.24). This species cant top-down role in structuring the decline in
has a large body size when mature and typically meso-zooplankton abundance in the summer.
accounts for a large proportion of the gelatinous
biomass of the upper estuary (85–90%). Following Conclusion
Aurelia, the ctenophore Pleurobrachia and the At the start of this chapter it was claimed that estu-
jellyfish Clytia appear in the water column between aries, coastal zone, and shelf seas present some of
late April and mid-September. The chaetognath the most demanding habitats on Earth. Touching,
Sagitta is also present sporadically throughout as we have, on only one aspect of the physiological
much of the year, but is really an autumn–winter problems, one aspect of benthic existence, and the
predator in the estuary. population sequencing of one estuary does but
Throughout the estuary the greatest period of poor justice to the theme. Every neritic (coastal and
gelatinous production is between April and July with shelf-sea) region has its unique conditions and spa-
a peak in May. At this time it has been calculated tial and temporal biotic responses. The elucidation
that the impact of these species is such that they of the processes that enable the biota to adapt and
remove some 5–25% of the copepod zooplankton respond to such a range of conditions continues to

257
A.P.M. Lockwood, M. Sheader, and J.A. Williams

provide the excitement, challenge, and stimulus for 3. Ekman, J.E., Nowell, A.R.M., and Jumars, P.A. (1981),
further study of coastal waters in all their aspects, Sediment destabilisation by animal tubes, J. Marine
Res., 39, 361–374.
both now and in the foreseeable future. 4. Gupta, B.L., Moreton, R.B., Oschman, J.L., and Wall, B.J.
(eds) (1977), Transport of Ions and Water in
Animals, Academic Press, London, 817 pp.
5. Hayes, G., Purdie, D.A., and Williams, J.A. (1989), The dis-
tribution of Ichthyoplankton in Southampton Water
General References in response to low oxygen levels produced by a
Gilles, R. (ed.) (1979), Mechanisms of Osmoregulation in Mesodinium rubrum bloom, J. Fish. Biol., 34,
Animals: Maintenance of Cell Volume, John Wiley 811–813.
and Sons, Chichester, 667 pp. 6. Jerling, H.L. and Wooldridge, T.H. (1991), Population
Gray, J.S. (1974), Animal–sediment relationships, Ann. Rev. dynamics and estimates of production for the
Oceanogr. Marine Biol., 12, 223–261. Calanoid Copepod Pseudodiaptomus hessei in a
Gupta, B.L., Moreton, R.B., Oschman, J.L., and Wall, B.J. (eds) warm temperate estuary, Estuar. Coast. Shelf Sci., 33,
(1977), Transport of Ions and Water in Animals, 121–135.
Academic Press, London, 817 pp. 7. Kifle, D. and Purdie, D.A. (1993), The seasonal abundance
Lopez, G., Taghon, G., and Levinton, J. (eds) (1989), Ecology of the phototrophic ciliate Mesodinium rubrum in
of Marine Deposit Feeders, Lecture Notes on Coastal Southampton Water, England, J. Plank. Res., 15,
and Estuarine Studies, 31, Springer-Verlag, Berlin, 823–833.
322 pp. 8. Larson, R.J. (1986), Seasonal changes in the standing stocks,
Parsons, T., Takahashi, M., and Hargrave, B. (1984), Biological growth rates and production rates of gelatinous
Oceanographic Processes, 3rd edn, Pergamon Press, predators in Saanich Inlet, British Columbia, Marine
Oxford, 330 pp. Ecol. Prog. Ser., 33, 89–98.
Pequeux, A., Giles, R., and Bolis, L. (1984), Osmoregulation in 9. Lockwood, A.P.M. (1961), The urine of Gammarus duebeni
Estuarine and Marine Animals, Springer-Verlag, and G. pulex, J. Exp. Biol., 38, 647–658.
Berlin, 221 pp. 10. Lockwood, A.P.M., Inman, C.B.E., and Courtenay, T.H.
Raymont, J.E.G. (1980), Plankton and Productivity in the (1973), The influence of environmental salinity on the
Oceans, Vol. 1, Phytoplankton, Pergamon Press, water fluxes of the amphipod crustacean Gammarus
Oxford, 489 pp. duebeni, J. Exp. Biol., 58, 137–148.
Raymont, J.E.G. (1983), Plankton and Productivity in the 11. Lucas, C.H. and Williams, J.A. (1992), A preliminary exam-
Oceans, Vol. 2, Zooplankton, Pergamon Press, ination of the seasonal succession of gelatinous preda-
Oxford, 824 pp. tors within the zooplankton community of
Remane, A. and Schlieper, C. (1971), The Biology of Brackish Southampton Water, Porcupine Newsletter, 5, 77–83.
Water, John Wiley and Sons, New York, 372 pp. 12. Lucas, C.H. and Williams, J.A. (1994), Population dynamics
Sommer, U. (ed.) (1989), Plankton Ecology – Succession in of the Scyphomedusae Aurelia aurita in Southampton
Plankton Communities, Springer-Verlag, Berlin, Water, J. Plank. Res., 16, 879–895.
369 pp. 13. Olsson, P., Graneli, E., Carlsson, P., and Abreu, P. (1992),
Williams, R. (1984), An overview of secondary production in Structuring of a post-spring phytoplankton communi-
pelagic ecosystems, in Flows of Energy and Materials ty by manipulation of trophic interactions, J. Exp.
in Marine Ecosystems, Fasham, M.J. (ed.), NATO Marine Biol. Ecol., 158, 249–266.
Conference Series IV, Vol. 13, Plenum Press, New 14. Ralph, R.R. (1965), Some aspects of the Ecology and
York, pp 361–405. Osmoregulation of Neomysis integer, Leach, PhD
Yingst, J.Y. and Rhoads, D.C. (1980), The role of bioturbation Thesis, University of Southampton, UK.
in the enhancement of microbial turnover rates in 15. Reid, D.M. (1930), Salinity interchange between sea water
marine sediments, in Marine Benthic Dynamics, in sand and overflowing fresh water at low tide, J.
Tenore, K.R. and Coull, B.C. (eds), Univ. S. Carolina Marine Biol. Assoc. UK., 16, 609–614.
Press, Columbia, pp 407–422. 16. Remane, A. and Schlieper, C. (1971), The Biology of
Brackish Water, John Wiley and Sons, New York,
372 pp.
References 17. Rhoads, J.M. and Young, D.K. (1970), The influence of
1. Alldredge, A.L. (1989), The quantitative significance of deposit-feeding organisms on sediment stability, J.
gelatinous zooplankton as pelagic consumers, in Marine Res., 28, 150–178.
Flows of Energy and Materials in Marine Ecosystems, 18. Williams, J.A. and Reubold, J. (1990), Observations on
Fasham, M.J. (ed.), NATO Conference Series IV, Vol. Aurelia aurita within the zooplankton of
13, Plenum Press, New York, pp 407–433. Southampton Water with particular reference to its
2. Cadee, G.C. (1976), Sediment reworking by Arenicola mari- ‘Absence’ in Summer 1988, Porcupine Newsletter, 4,
na on tidal flats in the Dutch Wadden Sea, Neth. J. 192–197.
Res., 10, 440–460.

258
CHAPTER 17:

Artificial Reefs
K.J. Collins and A.C. Jensen

tions of fish being attracted to objects placed in the

Introduction water. In Sardinia, tuna have been caught for hun-
An artificial reef is a structure placed in the sea, dreds of years in complex floating net traps weight-
either to protect or enhance the existing habitat or ed with stones. At the end of each season the stones
to create a new type of habitat for marine animals are cut loose and fall to the sea bed. Fishermen
and plants. This may lead to the enhancement of noticed how many fish species were attracted to
local fisheries. The reef design, material used, and these accumulating piles of weights. Similarly, acci-
the site chosen are all influenced by the required dental shipwrecks have long been popular fishing
function of the reef. sites. Such structures (Figures 17.1–17.3) provide
Natural reefs niches for marine life, attracting many fish species,
which take advantage of the ship’s structure to
Many people’s image of the marine environment is shelter from currents and predators and to find
that of a beach extending underwater, the sea bed food.
being formed from large expanses of sand. Reality is
different; sea floors range from fine muds through 17.1
sand and gravel to boulders and rock. While on
land, rocks and cliffs are among the more barren
areas, underwater they are covered in a profusion of
marine organisms. Fixed seaweeds need light to
grow and hence are restricted to shallow water.
More widely distributed are sedentary or encrusting
animals, such as sponges, hydroids, anemones, and
bryozoans. By living on elevated hard surfaces,
exposed to currents, these animals increase their
opportunities to catch plankton while minimising
the smothering effect of settling sedi-ments.
Most animals with sessile adult stages produce
pelagic eggs and larvae that live in the water col-
umn in order to distribute their offspring to areas
suitable for colonisation. Consequently, any new
hard surfaces are quickly occupied. Mid-water fish
are often attracted to rocks or wrecks that give
some shelter from tidal currents and predators, and
increase feeding potential. Where such hard sub-
stratum contains crevices, these can also provide
shelters for crabs and lobsters. Rocks and reefs
therefore offer valuable settlement sites for sea-
weeds and invertebrates, and shelter for mobile
bottom-dwelling forms and fish, so providing addi-
tional feeding opportunities for many animals. Any
extension of such a habitat, say by the deployment
of an artificial reef, increases the quantity of ani-
mals able to live within a given area of sea bed.
Artificial reefs Figure 17.1 Fish are attracted to shipwrecks, which provide
The use of artificial reefs as fishing sites has a long shelter from currents and predators. Here, shoals of sardines
history, presumably arising from chance observa- swarm on the wreck of a freighter, near Sardinia, Italy.

259
 K.J. Collins and A.C. Jensen

17.2 Figure 17.2 Wrecks also provide settlement substrate for

epibiota, as seen by a diver looking at growth on the rail-
ings of a wrecked freighter, near Sardinia, Italy.

Artificial reefs are incidentally created by engi-

neering works, such as harbour breakwaters
(Figures 17.4–17.6) and supports for bridges
(which are often built on muddy or sandy sea beds).
Such structures provide habitats for species that
could not live on the open sea bed, and attract
mobile benthic species, such as cryptic fish, lob-
sters, and crabs. Oil production platforms also
attract numerous fish species and provide a good
settlement surface for animals and plants. In the
Gulf of Mexico, the states of Louisiana and Texas
are involved in an active artificial-reef creation pro-
gramme utilising obsolete platforms. These are
either relocated to shallow water and sunk, or top-
pled in situ. The money saved by the oil companies
in disposal costs is ploughed back into the manage-
ment of these structures for recreational fishing –
over 4000 structures are available to anglers. In the
UK there is interest in using North Sea oil produc-
tion platforms as artificial reefs, now that they are
beginning to reach the end of their useful working
life (25–30 years).
17.3

Figure 17.3 The excavation of the historic wreck of the Mary Rose (sunk 1545)
provided a range of new habitats in the muds and silts of the Solent (southern
England). Large shoals of pouting congregated around the excavated timbers.

260
 17: Artificial Reefs

17.4 17.5

Figure 17.4 Coastal defence structures, such as these Figure 17.5 This breakwater in Monterey, California,
concrete tetrapods off Funchal, Madeira, provide new sur- provides a convenient spot for Californian sea-lions to
faces for colonisation by algae and encrusting animals. bask in the sun and groom. (Courtesy of Jane Jensen.)

17.6

Figure 17.6 Below water the breakwater boulders provide good anchorage for a
giant kelp forest (Monterey, California). The boulders are also settled by a profu-
sion of sessile animals, such as these colourful solitary corals.

261
 K.J. Collins and A.C. Jensen

17.7 reinforced plastic, although some stabilised ash is

also now used. The engineering and design aspects
of Japanese artificial reefs are well refined (Figure
17.7). Quality standards regarding building materi-
als, design, location, and construction have been
produced, which must be met if structures are to
qualify for government certification and therefore a
subsidy toward deployment costs. However, the
biological appraisal of artificial reef performance is
not so well advanced. Some have concluded that
there are insufficient biological and economic data
to enable judgement of the cost-effectiveness of
many of the operations.
In the US, the artificial reef programmes of
many Maritime States are run for the benefit of
recreational sports fishing, SCUBA (Self-Contained
Underwater Breathing Apparatus) diving, commer-
Figure 17.7 Extensive government support for artificial cial fishing, waste disposal, and environmental mit-
reef construction in Japan has led to the development of igation. An example of the latter is seen in the con-
complex artificial reef structures. Favoured materials are struction of the Pendleton artificial reef by the
concrete, steel, and plastics. This unit is made of con-
crete containing coal fly ash. (Courtesy of Tatsuo Suzuki,
Southern California Edison company. This 120
Hazama Corporation, Japan.) hectare reef was built offshore to compensate for
the possible loss of a giant kelp (Macrocystis) habi-
tat due to power station construction.
American experience of reef construction dates
back over 100 years, in which time a variety of
Fisheries (mostly waste) materials have been used, including
Artificial reefs around the world concrete, rock, construction rubble, scrap tyres,
Probably the majority of artificial reefs have been cars, railway carriages, and ships. The US has a
built by artisan fishermen in tropical countries. The national artificial-reef plan, but no government
purpose of such a reef is to increase catches in local funding commitment. Funding has come from the
fishing grounds using simple, readily available Federal Aid in Sport Fish Restoration Program,
materials, such as rocks, trees, bamboo, and scrap which may provide up to 75% of reef construction
tyres. For example, 1600 pyramid bamboo mod- costs, with individual States providing the rest (in
ules have been made and deployed in clusters of 50 1987, more than US$140 million was provided by
by local fishermen in the central Visayan Islands of the Federal Aid Program). Besides the government
the Philippines. The catches of a wide range of fish plan, individual states have local artificial reef pro-
from these reefs exceeded construction costs within grammes and Interstate agreements promote a
the first year, providing an annual harvest of rational approach to the deployment of artificial
8 kg/m2. reefs. Florida, with over 200 sites, has the greatest
The Japanese are the world leaders in artificial number of artificial reefs in the US.
reef technology for commercial fishery enhance- Elsewhere, artificial reefs have been developed
ment and have been creating artificial reefs since at according to local requirements, using materials
least the eighteenth century. Currently, Japan is in judged to be suitable. Only Japan and the US have
the third phase of artificial reef development, that a national development plan. Countries such as
of creating entire fishing grounds where there had Malaysia and the Philippines use waste tyres to
been none before. This programme commenced in build many of their artificial reefs. This is also the
1974 with the goal of diverting Japanese fishing case in Australia, where frequently reefs have been
effort from distant water fishing to mariculture and built from ‘materials of opportunity’, such as tyres
resource management in Japanese waters. and redundant ships. These reefs are used primarily
Government investment has been substantial; for as a focus for recreational angling, with some
example, in 1988 US$150 million was allocated to SCUBA diving. In Taiwan many fishing vessels
subsidise the construction of 2.2 x 106 m3 of fishing (made obsolete by government policy to reduce the
reefs. Over 17 x 10 6 m 3 of reefs have been size of the fishing fleet) were sunk to provide new
deployed, covering almost 10% of the sea bed shal- habitats.
lower than 200 m. Materials used are generally In Europe artificial reefs were pioneered along
‘prime materials’, i.e., concrete, steel, and glass- the Mediterranean coast in the late 1960s. At pre-

262
 17: Artificial Reefs

sent, most reefs are still associated with scientific Fisheries enhancement, especially of fin-fish fish-
research. Italy, France, and Spain have been the eries, is probably the area in which artificial reefs
most active reef-building countries since 1970. meet their greatest challenge. The attractiveness of
Currently, Spain is placing more artificial reefs into a reef to many, although not all, species of fish,
its coastal waters than any other European country both bottom dwelling and free swimming, is in lit-
(about 100 reef complexes are in place or planned). tle doubt. However, the contribution to the overall
Reef building has, until recently, been carried out biological productivity of free-swimming fish
nationally, with little cross-border cooperation. species is difficult to assess. The use of artificial
This is changing; in 1991, Italian artificial reef sci- reefs to attract and concentrate a population of
entists formed an Italian reef group to encourage fish, so facilitating exploitation, is a powerful tool
liaison between research groups, and an association in fisheries management; one that can be used to
of Mediterranean artificial reef scientists now increase exploitation or, by placing a reef in a pro-
exists. Current initiatives include a network of tected area, provide sanctuary.
European artificial-reef research scientists to estab- Artificial reefs essentially provide habitat. If the
lish a coordinated direction for reef research within fish attracted to the reef gain some positive advan-
the European Union. tage from the reef structure, then this can be set
against the negative impact that fishing has on fish
Artificial reefs for fisheries enhancement numbers. The greater the benefit from the provi-
Many fish species are attracted to objects in the sion of a reef (e.g., if a fish uses the reef as a feed-
water, which provide an orientation point and, ing or spawning site), the more likely it is that the
possibly, some shelter from sunlight or currents. reef will provide some net gain to the fish popula-
This behaviour has been exploited by fishermen tion. If this translates into an increase in the num-
using fish-attracting devices (FADs). Fish attracted bers of fish, and therefore an increase in the total
to the FAD are more easily caught because the fish- number available for exploitation, the fishery pop-
ing effort can be concentrated in one place. For ulation can be said to be enhanced. Due to the
example, artisan fishermen may use simple bamboo mobility of many commercially valuable fish
rafts to support bunches of coconut palm leaves in species this is very difficult to quantify. With terri-
the water below. More modern and complex FADs torial species, which spend part or all their life
are often reminiscent of kites that are ‘flown’ in cycle on a reef, the value of the structure is more
mid-water. Such devices make existing fish stocks easily defined.
easier to exploit; in the Philippines, increases in Examples of non-commercial fish species that
tuna catches from under 10,000 tonnes in 1971 to apparently benefit from association with an artifi-
266,000 tonnes in 1986 are attributed to the use of cial reef can be seen on one in England (in Poole
FADs. While increasing catches, FADs make no Bay, central south coast). Large shoals of pouting
contribution to biological productivity and so must (Trisopterus luscus), a species which spawns off-
be used with caution if overexploitation is to be shore, congregate around the reef during the day
avoided. (Figure 17.8). They disperse at night to feed on

17.8

Figure 17.8 Shoals of

pouting congregate
around units of the Poole
Bay artificial reef, UK,
during the day, gaining
shelter from tidal currents.

263
 K.J. Collins and A.C. Jensen

17.9 17.10

Figure 17.9 In early summer, male corkwing wrasse Figure 17.10 The corkwing wrasse constructs a complex
build nests of seaweed between the blocks forming the nest from seaweeds, seen here in its mouth, taken from
Poole Bay artificial reef. They defend a territory around algae growing on the reef and also drifting past in the
the nest and are not easily intimidated by observing current.
scientists.

nocturnally active small crustaceans which inhabit bers of larvae, the artificial reefs are not being
the surrounding sea bed. The main advantage that colonised at the expense of other rocky areas, but
they gain from the reef is shelter from tidal cur- in addition to them. In the Adriatic, this fisheries
rents, reducing the amount of energy needed for enhancement case appears to be well-proven; mus-
swimming, and so allowing increased growth. sel harvests boost the commercial catches from arti-
Another fish species, the corkwing wrasse ficial reefs, giving a three-fold return on deploy-
Crenilabrus melops, is territorial, staying within, or ment costs over 7 years. Net proceeds for fishermen
close to, the reef structure. In May–June each year operating within the artificial reefs have been
since 1990 males have built seaweed nests between shown to be 2.5 times that of activity outside the
the reef blocks (Figures 17.9 and 17.10). Females reefs. Elsewhere in Europe, attention has focused
lay eggs within these nests, which are maintained on lobsters as a commercial species under increas-
and guarded by the male until the eggs hatch some ing fishing pressure that may benefit from habitat
3–4 weeks later. Wrasse have been observed feed- provision in the form of artificial reefs.
ing on the reef epifauna (animals living on hard
surfaces), such as barnacles. For this species the Case study: artificial reefs and lobsters
reef provides food, shelter, and a site for Man-made shelters (‘pesqueros’ in Cuba, ‘casitas
reproduction. cubanas’ in Mexico), have been used to provide
The potential fishery value of artificial reefs is small, temporary refuges for spiny lobsters
more easily demonstrated with less mobile, but (Panulirus argus). These function as a focus for
valuable commercial species, such as molluscs and fishing effort, rather like a FAD for fin fish.
crustacea. Artificial reefs in the Adriatic Sea pro- Canada, Israel, and the UK, interested in other
vide settlement sites for large numbers of mussels, species of lobsters, have focused attention on artifi-
an important commercial species throughout cial reefs as a specific lobster habitat. Canada built
Europe. Because mussels produce such large num- the first artificial reef specifically for lobster

264
 17: Artificial Reefs

research in 1965. Over the following 8 years the mercial species found off the Mediterranean coast.
lobster population of the artificial reef was moni- Research showed that slipper lobsters preferred
tored by diving scientists. The reef was initially horizontal shelters with two narrow entrances on
colonised by large specimens of the American the lower portion of the reef. Using shelters to hide
clawed lobster (Homarus americanus), which were from predators is believed to be an important
thought to have outgrown their burrows, so being defense mechanism for these animals, so the pres-
forced to roam to seek new shelter. By 1973 the ence of the artificial reef provided a new and suit-
size-frequency distribution of the artificial reef pop- able habitat for colonisation. Slipper lobsters
ulation was similar to that on natural reefs in the migrate into deeper water as the inshore water tem-
area. It was concluded that the standing crop on perature rises in summer, but tagged individuals
the reef might be increased by a different pattern of were seen to return to a coastal tyre reef during
rocks. However, a cheaper source of reef material spring (Figures 17.11 and 17.12) over the project
or a multiple-use reef was required before an artifi- period of 3 years. In the long-term, populations of
cial reef could be considered an economically viable these heavily exploited animals could be protected
proposition, so the research was halted. against fishing effort by building appropriately
In Israel, efforts focused on the non-clawed slip- designed artificial reefs in protected areas, such as
per lobster, Scyllarides latus, an important com- underwater parks and reserves.

17.11

Figure 17.11 Scrap tyres provide a

resilient material for constructing artifi-
cial reefs. However the tyres must be
securely bound together and weighted,
as in this example off the Israeli
Mediterranean coast at Haifa.
(Courtesy of Ehud Spanier, Centre for
Martime Studies, University of Haifa,
Israel.)

17.12

Figure 17.12 The spaces within tyres

provide ideal shelter for fish such as
these squirrel fish. This Red Sea
species has migrated through the Suez
canal into the Mediterranean Sea.
(Courtesy of Ehud Spanier, Centre for
Martime Studies, University of Haifa,
Israel.)

265
 K.J. Collins and A.C. Jensen

Work continued since 1989 on the experimental then returning. Only the larval stage of the lobsters
reef in Poole Bay, England, found that lobsters life cycle is planktonic; the juvenile and adult lob-
(Homarus gammarus; Figure 17.13) appeared on ster live either in burrows in the sediment or in
the reef within 3 weeks of its deployment. Tagging shelters in rocky sea bed for the rest of their lives.
studies were initiated in 1990, and data to June Diver observations and evidence from pot-caught
1994 show that lobsters have found the artificial lobsters suggests that the reef can support all
reef a suitable long-term habitat, the longest period aspects of the benthic life cycle; berried (egg carry-
of residence standing at 4 years. This can be com- ing) females utilise the shelters and release their lar-
pared to a maximum age in the range of 10–20 vae from them, some reproducing more than once
years, sexual maturity being reached at 4–5 years on the reef. Lobster larvae have been taken from
of age. Conventional tagging of lobsters below the the waters above the artificial reef and a wide size-
fishery minimum landing size of 85 mm carapace range of juvenile and adult animals has been cap-
length (ca 250 mm total length) in a nearby fishery tured and/or observed by diving scientists.
revealed that these lobsters do not undertake any Artificial reefs have been shown to support effec-
seasonal migration, and the range of most move- tively three species of commercially important lob-
ments is less than 4 km in magnitude. The use of a ster. Research in the UK has shown good survival of
novel electromagnetic telemetry system has started hatchery-reared juvenile lobsters released into the
to reveal complex local movement behaviour. wild and subsequently recruited into the fishery. It
These data reveal that lobsters are mostly active at seems feasible that an artificial reef could be ‘seeded’
night and frequently change their daytime shelter. with hatchery-reared juveniles and that they would
The internal galleries and tunnels of the conical reef live to become part of the fishery. At present, the
units (1 m high, 4 m diameter) made from random- maximum densities of lobster population that can be
ly stacked, cement-stabilised, pulverised fuel ash achieved are not established, but data for H. ameri-
blocks (40x40x20 cm) were often occupied by canus suggests that the Canadian quarry rock reef
more than one lobster. An animal was also moni- supported one lobster per 6 m2 while the Poole Bay
tored leaving the reef site for up to 3 weeks and reef is thought to hold one H. gammarus per 2 m2.
Since neither structure was designed to maximise
17.13 lobster habitat there is a potential for improvement.
The artificial reef densities can be favourably com-
pared to results (in the order of one lobster per
30 m2) of diver surveys of natural reefs. Animal den-
sity is strongly correlated to the number of suitable
shelters available. Lobster territorial behaviour also
influences the usage of habitat. It seems more than
possible that in the future a reef could be designed to
provide lobster shelters in various sizes to minimise
‘off-reef’ movement, caused by the need to seek a
new shelter after increasing in size following moult-
ing. Already predictions have been made of the num-
ber and size of shelters in a reef made up of spherical
boulders (a starting point calculation for more realis-
tic material shapes), which can be linked to results
describing the habitat requirements of lobsters.
Aquaculture
Artificial reefs can serve several roles in fisheries
management, from the enhancement of wild fish-
eries to more intensive aquaculture systems.
Settlement of mussels on reef structures in the
Adriatic is described above. Stock density can be
intensified by adapting the well-established ‘sus-
pended rope’ system of mussel cultivation, prac-
tised throughout the world, stringing seeded ropes
Figure 17.13 The randomly stacked blocks of the Poole of mussels between artificial reef units. Suspended
Bay artificial reef provide a wide variety of crevices and rope is also used for kelp culture in Japan.
tunnels. Here a lobster emerges from the entrance to a In Japan, specially designed artificial reefs are
gallery within the reef unit where it spends most of the day. used for the culture of kelp, urchins, and abalone.

266
 17: Artificial Reefs

Combined systems for all three have been construct- 17.14

ed where the kelp supplies food for the urchins and
abalone. There is interest in aquaculture systems to
develop abalone as a cultured species in Europe;
artificial reefs may play a part in this.
Many salmon farms in northwest Europe have
started to utilise the ‘cleaning’ capabilities of species
of wrasse, such as the corkwing and goldsinny, to
remove sea lice from the bodies of the captive
salmon as an ‘environmentally friendly’ alternative
to treating the fish with chemicals. Currently, most
of the wrasse are caught in the wild, so localised
populations are depleted. As wrasse are territorial
rock-fish and spawn and nest in rocky habitats,
there is a potential to artificially enhance wrasse
habitat near salmon cages. This would provide a
self-sustaining source of ‘cleaning’ fish near the site
of salmon farms. The potential of artificial reefs and
their epifaunal filter-feeding community to act as
‘biofilters’ to remove waste (uneaten food and fae-
cal material) close to aquaculture facilities in the
Baltic Sea is being investigated. Mussels, acting as
filters, could play an important role in lessening the Figure 17.14 After 4 years underwater, the Poole Bay
impact of aquaculture on the environment. artificial reef density and variety of colonising animals
and plants closely resemble the biological communities
Biodiversity Management seen on local natural reefs.
The contribution that artificial reefs can make in
biodiversity (see Chapter 15) management is that of 17.15
habitat manipulation. This has the potential to
increase the number of species in an area and pro-
vide purpose-designed habitats for target species.
Provision of new habitat
Artificial reefs are usually constructed to provide
elevated hard substrata where formerly there was
none. Artificial reefs mimic natural reefs, but can
be built to provide greater surface area, elevation,
current shadow/disturbance, or niches/crevices to
favour target species.
An early European artificial reef study was that
by the Association Monégasque pour la Protection
de la Nature off Monaco in the 1970s. Artificial Figure 17.15 Colonisation of the Poole Bay artificial reef
reefs (2 m 3), made from hollow blocks or tiles was rapid. Within 1 year the surfaces were entirely cov-
cemented together, were laid on a muddy sea bed ered by a variety of hydroids, bryozoans, and sponges.
within a marine reserve. These attracted a good set-
tlement of epifauna and provided a habitat and the reef were unaffected by reef deployment. The
refuge for spiny lobster. Following on from this lowest estimates of epifaunal biomass per unit area
work, specifically designed cave habitats for rare were equivalent to that of the previous infaunal
red coral have been successfully developed by the biomass. However, the greater surface area avail-
Institut Océanographique at Monaco. able on the reef (2.5–3.0 times ground area lost)
The placing of artificial reefs on a mud–sand sea gave a higher biomass estimate than from the sea
bed smothers and kills infauna (animals living in bed that the reef covered. This new habitat led to a
the sediment) directly under the reef (see also rapid increase in species numbers and diversity.
Chapter 22, references to the Garrock Head waste- After 4 years the epibiota (Figures 17.14 and
dumping ground). Some 100 infaunal species were 17.15) provided a food source for molluscs, lob-
replaced by more than 250 epifaunal species within sters, crabs, and fish; these, in turn, provided food
two years in the Poole Bay artificial reef. The infau- for cuttlefish and predatory fish. The reef is provid-
nal populations in undisturbed sediment around ing a valuable site for reproduction, such as nesting

267
 K.J. Collins and A.C. Jensen

17.16 17.17

Figure 17.16 Artisan fishing communities in Kerala, Figure 17.17 Smaller concrete slabs, cast in the sand on
Southern India, have used locally available materials, the beach, have been assembled into artificial reef units
such as bamboo, to construct artificial reefs which are and are being paddled out to sea balanced on canoes
deployed on their village fishing grounds immediately made from three tree trunks lashed together (CWARP,
offshore. (Courtesy of Steve Creech, Hampshire, Coal Waste Artificial Reef Program). (Courtesy of Steve
England.) Creech, Hampshire, England.)

by corkwing wrasse or the laying of egg masses by It seems ironic that concrete should be used to
whelks (Buccinum undatum). restore the reef destroyed in the quest for building
materials.
Restoration of damaged habitat Around the UK coast there are several offshore
Modern techniques have increased the environmen- sites where waste could be dumped under licence
tal impacts of fishing; overexploitation of stocks by (see also Chapter 22). Off Blythe (northeast coast
modern, powerful vessels and heavier fishing gears of England), coal fly-ash from coal-burning power
(such as ‘rockhopper’ trawls; it is estimated that stations has been dumped for many years. This has
every square metre of the North Sea is trawled 3–5 resulted in the smothering of rocky outcrops with a
times annually), have led to the destruction of some sterile blanket of ash, covering ideal lobster and
sea-bed habitats. As an example, in southwest India crab habitat, and fish-feeding grounds. While
development aid was provided to equip fishermen dumping is legislated to cease in 1995, it will be
with trawlers to increase fish harvests. The coast- many years before this area will return to full pro-
line of Kerala province is one of the most densely ductivity. One restoration method would be to con-
populated areas of India. Artisan fishing communi- tinue dumping coal ash as stabilised ash blocks
ties depend on catches immediately off the beach (tested in the Poole Bay reef) to restore habitat
on which they live, up to 3 km offshore, the range diversity and accelerate the recolonisation process.
of their log canoes. The trawlers, based in the
northern part of the province, decimated the arti- Protection of existing habitat
san fisheries further south, which, as catches In the Mediterranean Sea, most artificial reefs have
declined, were no longer economic to operate. In been placed as nature conservation and/or habitat
an attempt to restore damaged fishing grounds the protection structures. At least 150 artificial reefs
coastal communities have used a variety of local have been deployed for habitat protection by Israel,
materials (stones, cast concrete and bamboo, Italy, France, and Spain. Reef complexes range in
Figures 17.16 and 17.17), with some success, to size from a few hundreds of square metres to sever-
restore bottom-habitat diversity and fish catches. al square kilometres. Their prime role is to prevent
In the Maldives, the lack of construction materi- the destruction of sea-grass meadows by trawling
als has led to the use of coral (as blocks and aggre- (Figure 17.18) – sea-grass is a valuable habitat for
gate) for building on the low-lying islands. many commercial species of fish.
However, the loss of the coral reefs that provide a One example of such a protective reef or barrier
living coastal defence barrier ultimately threatens is found off the new port of Loano (northwest
the entire archipelago. In an experiment, concrete Italy). Here, Posidonia sea-grass has been protected
structures normally used for coastal defence in by an artificial barrier some 3 km 2 in area,
Europe have been laid, and living corals transplant- deployed in 1986. Most of the anti-trawling barrier
ed onto them in an attempt to re-establish coral consists of 350 1.2 m3 cubes placed on the sea bed.
reefs. Corals have grown and fish have been attract- In the centre of the artificial barrier a number of
ed back to what had become a coral rubble desert. concrete block ‘pyramids’ (Figures 17.19–17.21),

268
 17: Artificial Reefs

17.18 17.19

Figure 17.18 Illegal trawling of sensitive marine areas in

Spain is being counteracted by these anti-trawling units.
The steel railway lines protruding from the solid concrete
blocks snag and destroy nets. (Courtesy of Technologia
Ambiental, SA, Madrid, Spain.)

Figure 17.19 Another anti-trawling reef, off Loano,

northwest Italy, uses pyramids of 2 m concrete cubes
with holes to encourage cryptic animals.

17.20 17.21

Figure 17.20 The Loano artificial reef was deployed to Figure 17.21 The Loano artificial reef units provide a
protect an area of sea-grasses, which are important nurs- variety of habitat types. This massive bryozoan colony
ery grounds for juvenile fish. being inspected by the diver is typical of animals found
on shaded overhangs on natural rock outcrops.

269
 K.J. Collins and A.C. Jensen

17.22 17.23

Figure 17.22 In Taiwan, high structural-strength coal-ash Figure 17.23 The open structure of the Taiwanese reef
concrete has been developed to construct complex artifi- modules attracts shoals of fish. (Courtesy of Kwang-Tsao
cial reef units. These modules are about to be deployed Shao, Institute of Zoology, Academica Sinica, Taipei,
from a barge. (Courtesy of Kwang-Tsao Shao, Institute of Taiwan, ROC.)
Zoology, Academica Sinica, Taipei, Taiwan, ROC.)

17.24 ‘Stabilisation’ is one technique of ensuring that a

finely divided material does not release potentially
harmful compounds into the environment. The
majority of stabilisation studies have involved
cement stabilisation of coal ash, including pul-
verised fuel ash, and flue gas desulphurisation
sludge containing gypsum. Initial research was pio-
neered in New York in the late 1970s, leading to
the Coal Waste Artificial Reef Program. Besides
addressing the engineering problems posed by the
effect of sea water on solidified ash in sea water,
this group laid the foundations for full environmen-
tal impact studies. These investigate the block
Figure 17.24 The surfaces of the coal-ash concrete mod- chemistry, monitor colonisation of the material,
ules are colonised with encrusting animals. (Courtesy of and look for evidence of bio-accumulation in the
Kwang-Tsao Shao, Institute of Zoology, Academica reef-associated biota. Similar studies have followed
Sinica, Taipei, Taiwan, ROC.) in the UK, Italy, and most recently in Hong Kong.
Coal ash combined with calcium carbonate
made from 2 m3 blocks, provide additional height residues has been studied in the Bohai Sea, China.
to the structure and are a focus of scientific The largest scale experiments have taken place
research. The whole barrier construction is a in Japan and Taiwan (Figures 17.22–17.24), where
marine reserve created by the Comune di Loano high structural-strength, stabilised coal-ash was
(local government), the regulations of which are developed. This enabled the construction of com-
supported by active policing of the reserve zone. plex structures, as opposed to the simple blocks
Research scientists have documented the colonisa- used in the other experimental ash reefs. There are
tion of the barrier structures by a variety of epifau- ambitious plans in both countries to scale up the
na and epiflora and the arrival of shoals of fish experiments to the full-scale utilisation of coal-fired
attracted to the barrier. The reintroduction of spiny power station outputs. The most exciting is the sea
lobster and grouper into the area is planned. mount concept from Japan, which envisages the
construction of structures in about 1000 m of
Waste Materials water to divert nutrient-rich oceanic currents by
The use of waste materials in the construction of placing a barrier in their path and so causing an
artificial reefs has economic advantages over the upwelling of productive sea water (Figure 17.25).
use of rock, concrete, and steel, but carries a poten- Experiments with oil ash (from oil-fired power
tial environmental risk; there should be stringent stations) stabilisation were undertaken in Florida
environmental compatibility assessments before (Figure 17.26), along similar lines to the coal ash
any material, including wastes, is allowed to be studies. In New York, work has now been extend-
used for reef construction. ed to the use of stabilised municipal waste incinera-

270
 17: Artificial Reefs

Figure 17.25 Artist’s impression 17.25

of a planned ‘Super Ridge’, to be
deployed in deep waters off
Japan to create an upwelling of
nutrient-rich water, which will
provide the basis for aquaculture.
This will be constructed from the
cement-stabilised coal-ash output
of a coastal power station.
(Courtesy of Tatsuo Suzuki,
Hazama Corporation, Japan.)

tor ash. This latter material is also the subject of of large numbers of this extremely durable product,
research in Florida, Bermuda, and the UK. but the enthusiasm for solving a serious waste dis-
Ash from power stations is used in terrestrial posal problem must be tempered with the caution
building applications (road bases, as filler in of environmental impact analysis.
cement, and in blocks), but a considerable quantity The principal international legislation covering
(about 50% in the UK) is still dumped (a typical the deposition of waste and other matter in the
2000 MW coal-fired power station produces about ocean is the London Convention, 1992 (formerly
800,000 tonnes of ash per year). The research into the London Dumping Convention – see also
construction of stabilised-ash artificial reefs shows Chapter 22). Placement of material for the con-
that it would be environmentally acceptable to use struction of artificial reefs is not covered by the
this waste material in the sea, contributing positive- Convention. However, aware of the range of mate-
ly to the resolution of the land-dumping problem. rials that have been used for such purposes, the
While research is continuing into the use of ash London Convention Scientific Group has recom-
waste, relatively little attention is being given to the mended that the guidance prepared for the inter-
effects of compounds leached from other waste pretation of the Annexes to the Convention in rela-
materials, such as tyres, that have been used in arti- tion to dumping at sea contains all the considera-
ficial reef construction for several decades. Reef tions needed to assess the placement of an artificial
construction is an obvious solution to the disposal reef or structure. The newly revised regional

17.26

Figure 17.26 In this successor to

the American CWARP pro-
gramme, oil ash and coal ash
were combined with cement to
produce artificial reef blocks
which were deployed off the
eastern Florida coast. The sur-
faces were initially colonised by
barnacles and the units attracted
shoals of fish. (Courtesy of Fred
Vose, Department of Fisheries
and Aquatic Sciences, University
of Florida, US.)

271
K.J. Collins and A.C. Jensen

Oslo–Paris Convention, covering the northeast Artificial reefs can be seen as a positive way of
Atlantic area, has included placement of matter, managing and/or contributing to the marine envi-
such as for the construction of artificial reefs, with- ronment. We are likely to see an increased demand
in its purview and is establishing a set of technical for reef construction to provide artificial reefs for
guidelines for the practice. recreational angling and SCUBA diving, commer-
cial fishery enhancement and/or protection, aqua-
The Future culture structures, and breakwaters for coastal
Artificial reefs have a long history and are used protection.
world-wide to promote fisheries, but the basic
understanding of their role in marine ecosystems is
limited. There still is a fundamental argument as to
whether artificial reefs increase the total productivi-
ty or simply attract stock from surrounding areas. It
General References
D’Itri, F.M. (ed.) (1986), Artificial Reefs – Marine and
is likely that the answer is species specific; in Japan, Freshwater Applications, Lewis Publications, Chelsea,
no increase in flat-fish catches was caused by the Michigan, USA, 588 pp.
deployment of reefs, just a redistribution of the fish Duedall, I.W., Kester, D.R., and Park, P.K. (eds) (1985), Wastes
population. On the other hand, octopus landings in the Oceans, Vol. 4: Energy Wastes in the Ocean,
John Wiley and Sons, New York, 818 pp.
and catch rates were increased by 1.8 kg/m3 of reef Seaman, Jr., W. and Sprague, L.M. (1991), Artificial Habitats
per year after 50,000 m3 of artificial reefs were for Marine and Freshwater Fisheries, Academic Press,
deployed. This was considered a net increase in San Diego, 285 pp.
total stock, as there was no evidence of redistribu- Stone, R.B. (1985), National Artificial Reef Plan, US
tion. As octopus are often habitat-limited, the cre- Department of Commerce, National Oceanic and
Atmospheric Administration, National Marine
ation of new habitat with suitable dens was thought Fisheries Service, NOAA Tech. Memorandum NMFS
to be the major influence behind the stock increase. OF.6, 39 pp + Appendices 49 pp.
As is apparent from our discussion, much of the
work on artificial reefs has been largely descriptive,
with some quantification of fish numbers to com- Those wishing to learn more of the development
pare with natural reefs. On a world-wide basis, and present state of knowledge about artificial reefs
greater fish densities are reported from artificial might also look at the proceedings of the following
rather than natural reefs. The species assemblages conferences:
are usually similar for a given location. There is a
need to improve on the quality of ecological data, Third International Artificial Reef Conference, 3–5 November
measure the productivity of reefs, study the inter- 1983, Newport Beach, California, Bull. Marine Sci.,
37(1), 1–402, 1985.
relationship of different types of organisms, and Fourth International Conference on Artificial Habitats for
thus, eventually, determine the energetics of an Fisheries, 2–6 November 1987, Miami, Florida, Bull.
artificial reef system. Marine Sci., 44(2), 527–1082, 1989.
We have described how artificial reefs can serve Fifth International Conference on Artificial Habitats for
a number of roles in fisheries management, protect- Fisheries, 3–7 November, 1991, Long Beach,
California, Bull. Marine Sci., 55(2+3), 265–1359,
ing or providing suitable habitat for particular 1994.
stages in the life cycle of fishery species. There is General Fisheries Council for the Mediterranean (1990), Report
still a need to understand the behaviour of com- of the First Session of the Working Group on
mercially exploited species to find out how artifi- Artificial Reefs and Mariculture, 27–30 November
1989, Ancona, Italy, FAO Fisheries Report, No.428,
cial reefs can best be constructed to supply their FAO, Rome, 162 pp.
requirements. The utilisation of reefs in fisheries Indo-Pacific Fishery Commission (1991), Papers Presented at
management will rely not only on scientific data, the Symposium on Artificial Reefs and Fish
but also on the perceived social benefit of maintain- Aggregating Devices as Tools for the Management of
ing artisan fisheries, the cost-effectiveness of con- Marine Fishery Resources, 14–17 May 1990,
Colombo, Sri Lanka, RAPA Report 1991/11, 435 pp.
struction against value of catch, and national legis- Waste material artificial reefs session, Second International
lation (e.g., that allowing an organisation to own Ocean Pollution Symposium, 4–8 October, 1993,
and manage a reef and sea bed). Beijing, China, Chem. Ecol., 10(1), 1–189 (1995).

272
CHAPTER 18:

Scientific Diving
J.J. Mallinson, A.C. Jensen, N.C. Flemming, and K.J. Collins

Introduction laboratory or on land, making decisions and judge-

A team of competent divers can offer field-work ments in real time in response to events that occur
assistance to virtually all branches of oceanograph- and opportunities that arise.
ic science, whether for pure or applied research, sci- The development and widespread availability of
entific archaeology, or fisheries management. SCUBA (Self Contained Underwater Breathing
Underwater, they measure and observe a wide vari- Apparatus) since the late 1950s has expanded the
ety of physical, geological, and biological processes, scope of scientific activity underwater. Marine sci-
such as the microhabitat that exists on the under- entists use SCUBA as a means of reaching their
side of the polar ice (Figure 18.1), submarine fresh- research sites. The freedom to work untethered is
water springs billowing out from the sea bed, ice- of enormous value; the diver in situ is able to think
age stalactites in drowned caves, and the most and adapt to changing conditions, conducting
secret and intimate behaviour of marine animals. work that is more complex and detailed than any-
These processes, objects, and events may be diffi- thing that can be achieved by the present genera-
cult, more expensive, or impossible to observe oth-
erwise. Scientific divers have worked in many parts 18.1
of the world’s oceans, including Rockall in the
open Atlantic, the coral reefs of Belize, and even in
the highest lakes in the world on the slopes of the
Himalayas.
Scientific divers may themselves be scientists
who have been trained to dive, or divers who have
received training in scientific techniques. Divers
are able to position accurately, deploy, and set-
up, maintain in situ, and later recover delicate
instrumentation on the sea bed; such instruments
can remain in place for extended periods, moni-
toring natural processes in the least invasive man-
ner. Some tasks may require specific skills, such as
marine-life identification, recognising and map-
ping small-scale underwater topographic features,
measuring artefacts in relation to a fixed datum
point, or the underwater construction of
apparatus.
The ideal of the scientific diver is to conduct
observational and experimental work under the sea
with the same accuracy as would be achieved in the

Figure 18.1 At the British Antarctic Survey Signy

Research Station, samples are collected from the unique
environment beneath the ice. In order to dive under the
winter sea-ice in Antarctica, a hole is cut using a chain
saw and the diver descends on a lifeline to ensure a safe
return to the access hole. The diving equipment is adjust-
ed to safeguard against freezing up in the cold water.
(Courtesy of G. Wilkinson, British Antarctic Survey,
Cambridge, England.)

273
 J.J. Mallinson, A.C. Jensen, N.C. Flemming, and K.J. Collins

18.2 Figure 18.2 Divers sup-

port remotely operated
equipment – each have
their uses. While the ROV
is not adversely affected
by cold or pressure and
can work indefinitely, the
diver is far more manoeu-
vrable, has more dexterity,
and, above all, can think.

tion of underwater Remotely Operated Vehicles fied, the diver can become involved in a variety of
(ROVs) (Figure 18.2). scientific activities, from primary observational sur-
The International Scientific Committee of the veys to very detailed tasks designed and carried out
World Federation of Underwater Activities by research scientists. The level of training required
(CMAS) encourages the highest standards of diving and the need for supplementary training varies
safety and promotes innovative research techniques from country to country, but all scientific diving
for all extremes of diving conditions around the follows well-established codes of practice and safe-
world. These standards are published by ty procedures to ensure that the task in hand is
UNESCO1. Most scientific divers initially train as undertaken in safety. An estimated 150,000 scien-
recreational divers with one of the international tific dives take place world-wide each year.
diver-training organisations, such as the British
Sub-Aqua Club (BSAC), the Professional Observation and Recording
Association of Diving Instructors (PADI), or the Quantitative observation and recording requires
National Association of Underwater Instructors accurate position fixing (see Chapter 20) and divers
(NAUI). All three are used world-wide, with differ- to make estimates of the occurrence and frequency
ent countries favouring one or another. Once quali- of different phenomena. The divers use quadrats,

18.3

Figure 18.3 The small-scale

spatial distribution of oys-
ters is determined by count-
ing individuals in each of
the four 0.5 m2 sections of a
1 m quadrat. The quadrat is
moved across the sea bed
until an area of 10 x 10 m
has been covered, giving a
statistically acceptable
matrix of 400 units.

274
 18: Scientific Diving

Figure 18.4 A diver’s sketch of 18.4

the sea bed off southern
California shows the variation in
biota from the beach of La Jolla

?
Shores to the edge of the Scipps
Canyon. There is little life in the
surf zone of the big Pacific
breakers, flat-fish and echino-
derms extend across the sand
flats, and macro-algae with asso-
ciated cryptic fish and crustacea
colonise the rocky outcrops at
the start of the drop-off into the
deep water of the canyon.

usually of 0.5 x 0.5 m, or 1.0 x 1.0 m, to concen- 18.5

trate their study and estimates into small manage-
able areas (Figure 18.3). Once the animal or feature
has been identified and counted in several quadrats,
it may be assumed that this density of occurrence
can be extrapolated over a larger area of similar
type. This method, in practice, requires the divers
to measure many quadrats in a systematic pattern,
or deliberately random pattern, in order to check
that they have obtained a truly representative mea-
sure of the organism’s or feature’s variability.
On a slightly larger scale, communities on differ-
ent substrata underwater can be recorded and
mapped as they exist in nature (Figure 18.4) – a far
cry from the mangled heaps of biota that are tipped
Figure 18.5 Frontier Tanzania, a joint programme
out onto the deck of a survey vessel from a bottom
between the Society for Environmental Expeditions and
trawling net. Divers can quantify their observations the University of Dar es Salaam, has combined scientific
by counting individuals or applying a semi-quanti- expeditions involving young people with extensive map-
tative scale to a range of organisms. ping of the marine environment around Mafia Island (off
Divers can map extensive areas of sea bed by the east coast of Tanzania). Expedition members are
recording biological and topographical features as trained in survey techniques, as well as in fish and other
they swim along a tape measure laid across the sea marine-life identification, to produce data that directly
bed. Such mapping has been used in coastal-zone contributes to the formulation of management plans for
management to ensure that uses, such as commer- the proposed Mafia Island Marine Park. This park will be
cial fishing, tourism, and resource extraction (oil, a multi-user, multi-zone park with areas designated for
limited resource use, tourist activity, fishing, scientific
aggregate, etc.), do not conflict with each other or
research, and conservation. These last two types of area,
cause excessive damage. This is of particular to be left undisturbed, are often ecologically important
importance in ecologically sensitive areas; destruc- sites for fish and coral reproduction and/or recruitment.
tion by any activity cannot be assessed unless a (Courtesy of M.A. Baldwin, Frontier Tanzania.)
baseline has been established (Figure 18.5).
The movement of the water itself is also a sub-
ject for direct examination in situ. Intense thermo- centimetres, or even a few centimetres. Sometimes,
clines develop at moderate depths in many parts of in the Mediterranean, a diver can descend through
the world where the water temperature decreases a thermocline that is so sharp that it is like dipping
rapidly through a vertical interval of a few tens of one’s foot or finger into a cold bath. These rapid

275
 J.J. Mallinson, A.C. Jensen, N.C. Flemming, and K.J. Collins

18.6

T
I

B
L

Figure 18.6 In the 1960s, a series of experiments off Malta by John Woods in which dye was injected into the seasonal
thermocline revealed a number of unexpected features of small-scale flow. Persistent streakiness provided proof that the
flow is mainly laminar, whereas previously it had been assumed to be continuously turbulent. Undulations revealed
high-frequency internal waves trapped on ocean fine-structure sheets of enhanced stability frequency. The orbital
motion of water displaced by these waves created shear across the sheet, occasionally provoking Kelvin–Helholtz insta-
bility, which produces billows that are typically 0.2 m high. The billows in turn generate a patch of turbulence which
persists for about 10 minutes. Here all four phenomena are shown: laminar flow streaks, L; an internal wave, I; a billow,
B; and turbulent mixing, T. (Courtesy of J.D. Woods, Imperial College, London University, England.)

transitions are occasionally visible in nature naval operations, climatology, and plankton ecolo-
because films of dust or dead organic particles set- gy of the upper ocean.
tle on the layer where the density changes (cold
water is denser than warm water). Divers have Scientific Marine Archaeology
studied the internal waves which travel along these Many scientific techniques are used in archaeology.
surfaces by releasing fluorescene dye which spreads Underwater archaeology takes observation and
along the surface in undulating sheets (Figure mapping to its extreme, as valuable information
18.6). about the context and position of the artefacts and
These experiments showed that the flow in the
seasonal thermocline is overwhelmingly laminar 18.7
rather than turbulent, which led to a paradigm shift
in oceanography with profound implications for

Figure 18.7 In 1967, a remote archaeological site at

Asopos in the south of Greece was surveyed using
snorkelling equipment and an underwater camera.
Standing stone walls were photographed at a depth 3.5 m
underwater, in a style and pattern which could not be
identified at the time. It was 12 years later before the
authorities granted a permit for the site to be surveyed in
detail by a large team of divers; most of the underwater
city was surveyed, but by 1979 the strange walls in deep
water had completely disappeared, presumably destroyed
by winter storm waves. The photographs had been pub-
lished in 1969 so the walls’ existence is indisputable.

276
 18: Scientific Diving

features uncovered can be lost if not accurately 18.8

recorded quickly. Changing sea levels have caused
the inundation of many coastal cities and settle-
ments through the ages. Once underwater, struc-
tures eventually deteriorate, some through sudden
catastrophe, while others gradually erode away
(Figure 18.7).
Well-preserved shipwrecks have revealed a mass
of historical information, unknown from land
excavations. Working long hours, often in difficult
conditions of poor visibility, darkness, and cold,
diagrams are produced that can be used to recon-
struct events or items that cannot be brought to the
surface, such as impressions in the sediment of
long-since corroded iron weapons or structural fea-
tures of a fragile ship (Figure 18.8). Study of the Figure 18.8 The Mary Rose, a Tudor warship sunk on the
preservation of submerged timbers provides infor- South coast of England in 1547, was buried in mud with
mation (often all that is available) about the long- one side virtually missing. From measurements and draw-
term presence and effects of marine organisms. The ings made by divers excavating inside the hull and bur-
rowing underneath it, a cradle was constructed to exactly
depth to which artefacts of known dates are buried fit and support the fragile structure as it was lifted from
provides information about sedimentation rates. the sea bed. At no time, before it was lifted out of the
mud, could the entire ship be seen or measured.
Sampling
Observations, particularly when meticulously 18.9
recorded, are an excellent starting point for envi-
ronmental investigation. Most research pro-
grammes, however, require data beyond the record-
ing capability of an observing diver, so physical
samples are needed for laboratory analysis.
Collection by diving is an expensive option to
choose in terms of time and money and has distinct
limitations, especially when many samples are
needed from a large area. Remotely operated grabs
deployed from ships sample soft mud and sand, but
as the sea bed becomes harder these are less effec-
tive. The diver is a ‘thinking sampling device’, and
will bring the essential skills of discriminating
observation, judgement, selection, and decision-
making to the work underwater. In other words,
the diver can choose a sampling method to suit the
job and adjust techniques to achieve the best
results. If the sea bed is mostly mud with occasion-
al rocks, the diver can, without positional bias,
sample the mud with a core tube and the rock with
a hammer, recording the relative proportions of
each (Figure 18.9). A corer dropped from the sur-
face would sample (or not), whatever it landed on.
In some situations a diver is the only possible
option. A specific material or species may have to
be searched for, recognised, and sampled with care.
Delicate sessile specimens can be detached from Figure 18.9 A diver collects core samples (for infaunal
their substrate with minimum damage, or samples analysis) of a known volume from a stony sea bed, which
are not easily obtained remotely. A metal tube is ham-
for chemical analysis collected by hand to avoid mered into the sea bed to a known depth. The tube is
contamination. In addition, a description of the then dug out and capped at each end for recovery. If nec-
surrounding sea bed can also be made to accompa- essary, a compass can be used to determine the tube’s
ny the sample to the laboratory, so that the non- magnetic orientation and it can be kept upright as it is
diving scientist is given a preserved specimen, a brought to the surface. Several replicates are usually col-
photograph of it in life, and a description of the lected to ensure a representative sample.

277
 J.J. Mallinson, A.C. Jensen, N.C. Flemming, and K.J. Collins

18.10 Figure 18.10 A diver, with

appropriate training, can
identify marine fauna and
flora on the sea bed. Detail of
a specimen’s undisturbed
behaviour and the habitat in
which it lives can be
observed and recorded. It can
be collected and brought to
the surface with a photograph
of that actual specimen in its
natural environment.

habitat and community of which it was a part within submarine cave systems, divers have collect-
(Figure 18.10). ed biological and geological samples yielding infor-
Sampling by divers need not be restricted to the mation new to science (Figure 18.11).
shallow sea bed. ‘Blue-water diving’ in oceanic
waters allows planktonic species (normally the Photography and Video
larger ones, like jellyfish) to be taken from the The viewpoint of the diver can be brought to a
water column without the damage caused by cap- wider audience by the use of photography and
ture in nets. These animals can then be maintained video. Early Hass and Cousteau films have inspired
in aquaria to reveal information about their life many a budding scientific diver and the realism of
cycles and physiology that would have been impos- more recent filming techniques is probably respon-
sible to obtain from stressed and damaged net- sible for today’s popularity of sport diving and
caught specimens. marine conservation. The ability to retain images
Divers are very skilled at penetrating caves and has a number of advantages. Photographs and films
wrecks where access for ROVs and sampling allow a better insight of the subject to those who
devices is difficult or impossible. By exploring deep do not dive. They can also provide proof of exis-

18.11
Figure 18.11 The vast and complex
underwater cave systems of the
Caribbean ‘Blue holes’ were creat-
ed by freshwater flows through the
limestone bedrock of islands during
periods in the past when sea levels
were much lower. Samples of sta-
lactites and stalagmites from the
caverns were collected and dated
by the uranium–thorium method.
These records provide evidence of
periods when the caves were in a
subaerial environment and when
they were under the sea, so that the
changes of sea level during the late
Ice Age can be determined accu-
rately. Only a diver can collect
water and rock samples from 20 m
underwater, 200 m into the Lucaya
cave system of Grand Bahama.
(Courtesy of R. Palmer, Technical
Diving Ltd, England.)

278
 18: Scientific Diving

tence to the cynical and a lasting record of a disap- 18.12

pearing or transient feature. Actual images of live
animals and plants now complement line drawings
and description in identification guides.
In the hands of a skilful diver, photography can
be a research technique in itself, as well as one to
support other recording and sampling methods.
Photogrammetry, used extensively in archaeology,
comprises a series of photographs, taken from a
fixed grid, which are pieced together to give a
detailed, accurate photograph of a large area.
Similarly, stereophotography, using two cameras
mounted above the site, provides three-dimensional
images from which measurements can be taken.
Time-lapse, a video as well as photographic tech-
nique, is used to determine changes over time in
diversity and growth. Divers can install equipment
above a marked station to record the rate of
increase in the size of a sponge or encrusting bry-
ozoan over weeks or months, or to record the
changes in diversity of a community from season to
season. Animal behaviour can be monitored with
minimal disturbance to the subject. While animals
may respond to the presence of a diver, a piece of
apparatus is quickly accepted as part of the sea bed
(Figure 18.12).
Divers do not generally work at depths greater
than 50 m for safety reasons. Their time underwa-
ter is also limited by a number of factors, such as
human physiology, equipment, location, logistics,
and, ultimately, financial cost. Scientific divers can
work directly with their subject in situ or bring
their work to the surface for the study to be contin- Figure 18.12 A diver installs a time-lapse video recorder
ued more conventionally. Advances in the equip- above a lobster pot to investigate the efficiency of this
ment available for diving continue to expand the fishing method. The camera runs intermittently for a few
scope of the diver and therefore the progress of seconds at a time over 24 hours. At night a lamp is auto-
matically switched on during recording. The lobster’s
underwater science. approach to the pot is captured and its behaviour inside
the trap monitored.

General References Dean, M., Ferrari, B., Oxley, I., Redknap, M., and Watson, K.
(eds) (1992), Archaeology Underwater: The NAS
Flemming, N.C. and Max, M.D. (eds) (1990), Scientific Diving: Guide to Principles and Practice, Nautical
A General Code of Practice, CMAS, UNESCO, Paris, Archaeology Society, London, 336 pp.
254 pp.
Woods, J.D. and Lythgoe, J.N. (eds) (1971), Underwater
Science – An Introduction to Experiments by Divers,
Oxford University Press, Oxford, 330 pp.
National Oceanic and Atmospheric Administration (1991),
Reference
NOAA Diving Manual: Diving for Science and 1. Flemming, N.C. and Max, M.D. (eds) (1990), Scientific
Technology, 3rd edn, US Department of Commerce, Diving: A General Code of Practice, CMAS,
Washington DC. UNESCO, Paris, 254 pp.

279
CHAPTER 19:

Marine
Instrumentation
G. Griffiths and S.A. Thorpe
Introduction delicate instruments or sensors may be damaged in
An ability to measure is central to any science. handling and as they are lowered over the side, per-
Data provide quantified measures of the composi- haps swinging before entering the water. With
tion and contents of the oceans and their support- experience, risks are reduced, but are inevitably
ing sediments and rocks, and are needed to test or high when weather changes suddenly, as the
develop hypotheses; in recent times, the data have research vessel is unexpectedly struck by a large
been needed to provide starting points (‘initial wave, or when data are required in just those con-
data’) to run models to forecast natural changes ditions in which hazards are greatest.
and the consequences of anthropogenic inputs to Once in the water, hazards to instruments are
the ocean or the atmosphere. still present. Their design must be appropriate to
Marine research requires investment in ships the conditions in which they operate, perhaps to
and, for the science to develop, technological devel- pressures 500 times that at the sea surface (e.g., on
opments, the construction of new or more accurate an abyssal plain at a depth of 5 km), in a salty and
instruments, and novel measurements to provide corrosive environment. Collapse of buoyancy
further insights into ocean processes. Ships have almost inevitably results in the loss of both record-
always been essential for marine studies (see ed data and instruments. Near the surface, instru-
Chapter 1), and will be for many decades into the ments are subjected to large and variable forces
future, but they are expensive so other methods of applied by breaking waves, and may be subjected
observing the ocean are evolving. The most striking to ‘marine fouling’ by organisms settling on and
of such methods to appear in the twentieth century attaching to sensors (see inset in Figure 19.12).
is satellite remote sensing. There are problems, too, of recovering samples
The information about the upper parts of the from depth without damage due to reduced pres-
ocean obtained in recent years by Earth-orbiting sure or contamination, and of preserving samples
satellites is immense (see Chapter 5). Measurements at depth, possibly for extended periods of time,
are made over vast areas of the ocean surface, often before recovery to the surface for study. Damage
in regions seldom visited by research vessels, and and loss of moorings by trawling is a severe risk in
are repeated at regular intervals so that variability areas of intensive fishing. The ‘dramatic’ events
over periods of months and years can be studied. which occur and result in damage are (sadly for the
Satellites have greatly improved the accuracy of reader’s entertainment) not those recorded by pho-
position fixing (Chapter 20), but satellite sensors tographs, since the focus of attention and activity is
are only able to obtain measurements at or near the then on the safety of the instruments, scientists, and
ocean surface. The movement and properties of the crew. However, some of the consequences, the
waters deeper than a metre or so, in particular damage to instruments, are recorded [see Figure
those responsible for carrying heat around the 19.30(b)].
globe, are often undetected or poorly resolved by Given the large expense of marine operations
satellite sensors, so we are forced to rely on other, and the importance of the data, the reliability,
more ‘conventional’, methods of sampling and accuracy, and precision – demanding careful cali-
measurement to probe the remote ocean depths, bration – are paramount. The resolution of the
usually (but not always, as we explain later) involv- data and the sampling intervals must be matched to
ing the use of ships. the sensors and fitted to the purpose for which
The deployment and recovery of instruments measurement is intended. There is little point in
from ships at sea carries a reliance on seamanship sampling with high frequency if the sensors them-
and navigation, and on the efficient operation of selves do not react to rapid change, or if high-fre-
winches and cranes. There are risks to those quency data are known to be of no use. Data stor-
involved, particular in rough weather when safe age and subsequent analysis are expensive and
working on deck is difficult (if not impossible), and internally recording instruments are limited by log-

280
 19: Marine Instrumentation

Figure 19.1 The research FLoating 19.1

Instrument Platform (FLIP) moored
in the deep sea while participating
in an experiment to study surface
waves. FLIP is a 120 m manned spar
buoy operated by the Marine
Physical Laboratory of the Scripps
Institution of Oceanography in La
Jolla, California. It is towed to an
experiment site in a horizontal posi-
tion. Ballast tanks are then flooded,
inducing the ‘flip’ into the vertical
position as shown. With 87 m of the
cylindrical hull below the surface,
FLIP is extremely stable. In a sea-
way, typical vertical motions are
less than 2% of the wave height. A
crew of five operate the platform,
hosting up to 10 scientists for ‘cruis-
es’ of as long as 35 days. Doppler
sonar (see text and Figure 19.28),
mounted beneath the water line, is used to profile the sea-surface and upper-ocean velocity fields. Conducting wires
to measure waves, current meters, conductivity, temperature, and depth instruments (CTDs) to measure the thermal,
salinity, and density of the ocean, and other instruments involved in the study of air–sea interaction (see Chapter 2) are
suspended from light-weight booms deployed when FLIP is vertical. (Courtesy of Dr R. Pinkel, Scripps Institution of
Oceanography, La Jolla, California, USA.)

ger capacity and available battery power. In many we have had to be very selective in making a
cases, ‘real-time’ data display and analysis are an choice. The selection is not comprehensive and
advantage, if not a requisite, so that more intensive there is no implication that the ‘best’ instruments
measurements or a different sampling strategy can are included, although some are.
be implemented when particular events or process- International protocols for data quality and col-
es occur; study in the ocean, unlike in the laborato- lection in the ocean may lag behind atmospheric
ry, does not have control over change, and prepara- ones (from which meteorological predictions are
tion is needed to respond to opportunities to make made) by as much as 30 years; this is the case even
measurement as they occur. Herein lies skill, excite- for basic physical properties. Detailed and continu-
ment, and reward to the observational scientist in ous measurements of biology and chemistry in the
the natural environment. water column are rarely possible. The ability to
This also illustrates a fundamental difference in forecast the oceans, to predict the important part
the science from that conducted entirely in the lab- which they play in climate change (Chapter 3) or
oratory; experiments in the sea often cannot, how they will respond to man-made wastes or pol-
except in statistical ways, be repeated, and a cus- lution (Chapter 22), is correspondingly less precise;
tomary axiom of science, that the conditions in this is magnified because the most energetic
experiments should be described in such detail that motions which must be resolved and measured are
they may be precisely repeated by others, cannot of a much smaller scale than those in the atmos-
apply in the natural environment where replication phere (Chapter 4), so far more measurements are
in identical conditions is rarely possible. needed to detect their presence and quantify their
Precise measurement in the ocean, for example effect.
of turbulent motions, may require ‘stable plat- Thus, an immense challenge is posed to marine
forms’ which do not move or vibrate. To solve such scientists – how are they to make measurements
difficult measurement problems is sometimes sufficient to verify and initialise predictive models
impossible by the available means, and new plat- of the ocean? In this chapter we provide a state-
forms or techniques may be the only answers (e.g., ment of some of the recent progress in developing
Figure 19.1). The challenge to instrument designers and using instruments in the ocean, and then dis-
is enormous! We describe in this chapter some of cuss the future prospects of improving sampling
the instruments that they have cleverly devised. techniques and attaining the goal of a Global
While the range of instrument types is illustrated Ocean Observing System (GOOS) to provide a firm
herein, the total numbers and variety are great, so basis for predictions in the twenty-first century.

281
 G. Griffiths and S.A. Thorpe

19.2 Figure 19.2 A deep-ocean hydrographic instrument

package consisting of a CDT, a 24 x 10 litre-bottle rosette
multisampler, fluorometer, and transmissiometer.
(Courtesy of S. Hall, SOC.)

tion, its latitude, longitude, heading, and speed.

High-precision satellite navigation – the Global
Positioning System (GPS) – is used routinely on
oceanographic research ships, giving position accu-
racy to about 40 m world-wide (see Chapter 20).
Acoustic instruments on ships provide a remote-
sensing capability. Echo-sounders provide accurate
depth measurements, both beneath and on both
sides of the ship’s track (Chapter 20). Acoustic
Doppler sounders give measurements of the cur-
rents in the water below the vessel (see later).
Most research cruises make use of automated
surface sampling – where sea water from an intake
on the ship’s hull is pumped to the laboratories for
analysis. Temperature, salinity, fluorescence (indi-
cating chlorophyll in phytoplankton), optical trans-
mission (indicating particulate matter), and the
nutrients, including nitrate, phosphate, and silicate,
may be measured in this way.
Volatile chemical substances must be measured
immediately, such as the chlorofluorocarbons
(CFCs) used as tracers of water pathways through
the ocean (Chapter 11). Some substances can
change their concentration within the samples as a
result of biological activity; such samples must be
‘In Situ’ Instruments deep frozen on board ship for later analysis.
We begin with instruments which take samples or Lowered instruments
measurements of the sea water or sediment sur- Knowledge of the global-scale circulation of the
rounding them. deep ocean has been obtained from measurements
Instruments connected directly to a vessel
19.3
Instruments on board research vessels
The earliest scientific measurements at sea were
made from vessels while underway. Medieval navi-
gators estimated ocean currents by observing the
drift of their vessels, comparing the observed posi-
tion with the estimated (or dead-reckoned) posi-
tion, and ascribing the difference to the currents.
By the 1820s the major surface currents of the
world’s oceans had been discovered and charted by
hydrographers (see Chapter 1).
Today, instruments fitted to the ship still provide
fundamental measurements of the ocean. Foremost
among these is the measurement of the ship’s posi-

Figure 19.3 Precision thermometers still have a role –

today’s instruments use platinum resistance sensors and
LCD displays rather than the mercury-in-glass type, the
mainstay of oceanic temperature measurements for over
a century. (Courtesy of M. Conquer, SOC.)

282
 19: Marine Instrumentation

Figure 19.4 Used for obtaining large diameter cores for 19.4
geochemical and geotechnical analysis (see, for example,
Chapter 8), a 2 m long Kastenlot box-section corer is
deployed from the midships winch of RRS Discovery.
(Courtesy of SOC.)

of temperature and salinity using equipment low-

ered from stationary ships. From the early nine-
teenth century to the late 1950s, oceanographers
were dependent on bottles closed at depth to collect
water samples for salinity determination on board
ship, and on mercury thermometers for accurate
temperature measurements. Exercising great care in
handling and calibration and with a practised eye,
the best of these deep-sea thermometers could be
read to an accuracy of some 0.003 K. The arrival
of the transistor and integrated circuit revolu-
tionised temperature measurements. It became
practical to use electrical-resistance thermometers,
to process the tiny electrical signals within a low-
ered instrument, and to send the results back to a
display on the ship.
The early Salinity, Temperature, and Depth
(STD) instruments, developed in about 1960, were
not as accurate as the mercury thermometers, but
they provided continuous, rather than discrete,
measurements of both temperature and salinity
from surface to sea floor; they contributed to the
discovery that the density of the ocean rarely
increases smoothly with depth, but is irregular with
microstructures on the scale of 1 mm to 10 m (see
Figure 18.6).
The STD instruments also measured salinity – ty corer. In its simplest form the design is usually a
previously, salinity could only be determined by large weighted head with a tubular or box-section
collecting a sample of water in a reversing water barrel, which is driven into the sediment. A catcher
bottle, and analysing the sample using chemical or mechanism closes as the corer is pulled out of the
electrical methods back on board the ship. sediment and this retains the core during retrieval.
Developed from STD instruments in the early Gravity cores are prone to compression as well as
1970s, the Conductivity Temperature Depth instru- disturbance, and the depth to which a core can be
ment (CTD) provides more accurate measurements, taken is restricted to a few metres due to friction.
and is still today the work-horse of physical Box corers (e.g., Figure 19.4) are of a more
oceanographers (Figure 19.2). Its temperature accu- sophisticated design, and of a larger diameter to
racy is now better than that of mercury thermome- reduce friction. The larger diameter also allows
ters, though modern mercury equivalents are still subsamples to be taken, and because these corers
used as a check (Figure 19.3). But the conductivity are lowered onto the sea bed, rather than dropped
measurements that give salinity still need to be cali- from a height (as are gravity corers), the sedi-
brated against water samples. Although this is a ment–water interface is preserved, which is a great
chore, water samples cannot yet be dispensed with advantage.
as they are still the only source of data on many Piston corers were designed to take long cores –
chemical compounds of interest. up to 50 m. The action of a piston within the barrel
reduces the internal friction, effectively sucking the
Bottom-sampling instruments sediment into the core barrel. Piston corers, like
Samples of sediments from the ocean floor are gravity corers, miss the sediment–water interface,
obtained by a broad class of devices known as which – being light unconsolidated material – is
grabs and corers. Some corers are general-purpose ‘blown away’ by the pressure wave ahead of the
tools, others are designed for specific subsampling core barrel.
and analysis needs. The most basic tool is the gravi- For biological and geochemical analysis this

283
 G. Griffiths and S.A. Thorpe

19.5 Figure 19.5 The multicorer takes eight short cores for
biological and geochemical purposes (see Chapter 11).
The stand rests on the sea floor and the corer is lowered
into the sediments. (Courtesy of SOC.)

interface needed to be sampled, a need that led to

the design of multicorers (Figure 19.5), with a grid
of short tubes mechanically pushed into the sedi-
ment, retaining the sediment–water interface and
causing minimal disturbance.
Towed sensors and samplers
The traditional oceanographic measurement sta-
tion, where the ship is stopped and wires are low-
ered to make measurements, is costly in ship time.
To save time, engineers and scientists have devised
instruments and vehicles that may be towed behind
a ship to give continuous coverage while underway.
Following tows of instruments at constant depth
came the need for vehicles which could sample the
water column more completely.
Canada led the way with the Batfish™ in the
early 1970s, a vehicle that could be towed at 5 m/s
while undulating between the surface and 400 m.
With a payload of a CTD and other instruments, it

19.6a 19.6b

Figure 19.6 Towed vehicles do away with the need for

the vessel to stop and lower instruments. The SeaSoar
undulator travels from the surface to 500 m depth in a
horizontal distance of 2 km at a ship speed of 9 knots
(about 4.5 m/s). It carries CTDs and other physical, chemi-
cal, and biological instruments to measure the structure of
the upper ocean (see, for example, Figure 4.10). (a)
Deployment through the stern ‘A’ frame of RRS Charles
Darwin is straightforward in calm seas, but can be difficult
in a gale (a spare vehicle is in the right foreground). (b)
The vehicle being deployed – SeaSoar’s ability to carry
several instruments makes it especially valuable for
research spanning more than one discipline. Increasingly,
the vehicle is used to make measurements for biological
oceanographers, including ocean colour, phytoplankton
pigments, and zooplankton counts using optical and
acoustic sensors.

284
 19: Marine Instrumentation

19.7 Figure 19.7 Introduced in the early 1970s, the UOR –

known commercially as the Aquashuttle – can be towed
from ships of opportunity at speeds of up to 10 m/s. The
vehicle can either be self-contained, following pre-set
undulations and storing data internally, or it can be con-
trolled from the ship through a conducting cable to give
real-time data. It can be launched by one person and can
record its data internally. It is ideally suited for deploy-
ment by non-experts on merchant ships, so extending the
geographical coverage of data collection beyond the
tracks of research vessels in a most cost-effective way.
(Courtesy of Plymouth Marine Laboratory, Plymouth, UK.)

19.8 19.9

Figure 19.8 The Rectangular Midwater Trawl (RMT) 1+8 Figure 19.9 The epibenthic sledge is designed to skim
acoustically controlled mid-water net system consists of a over the surface of the sea bed and collect the organisms
pair of rectangular trawl nets within a single frame. The living on the sea floor, immediately above it, and in the
outer net has an opening area of 8 m2 with a mesh of upper few centimetres of the sediment. An acoustic
4.5 mm to catch micronekton, the inner net an opening pinger indicates when the sledge has reached bottom,
area of 1 m2 with a finer mesh (320 μm) is for plankton. when the opening–closing mechanism has operated, and
Acoustic commands from the ship control a mechanical when the sledge leaves the bottom. Cameras and elec-
release gear to open and close the nets. Data on tempera- tronic flash may be mounted on the sledge to provide
ture, depth, speed and distance travelled, and net position images of the nature of the bottom and to give some indi-
are telemetered acoustically to the ship. (Courtesy of SOC.) cation of the efficiency of the net in capturing animals.
(Courtesy of SOC.)

gave new insights into the structure of the upper For many experiments this is too heavy a burden
ocean at 1 m vertical and 1 km horizontal resolu- (too large a team to support financially or to
tion. This idea was developed into the UK SeaSoar accommodate onboard ship); for these, simpler
vehicle, which can undulate from the surface to vehicles are available to carry out similar measure-
500 m in a distance of less than 2 km. Several expe- ments. The Undulating Oceanographic Recorder
ditions have used SeaSoar for tows of over (UOR; Figure 19.7) is one such device for covering
10,000 km, with the greatest danger being during the upper 200 m.
recovery or deployment (Figure 19.6). Research laboratories have developed many
Supporting SeaSoar at sea requires a team of other types of specialist towed instruments for sam-
electronic and mechanical engineers, as well as data- pling, many of which are unique, serving the needs
handling specialists, to enable the wealth of infor- of individual researchers or small teams – examples
mation to be processed aboard ship in real-time. are shown in Figures 19.8–19.11.

285
 G. Griffiths and S.A. Thorpe

19.10a 19.10b

Figure 19.10 A thermistor chain (a) laid out under cover

before deployment, and (b) being deployed. This chain of
thermistors to measure the thermal structure of the upper
ocean is 400 m long and consists of 100 sensing ‘pods’,
each providing a measure of temperature, and 28 mea-
sures of pressure so that the depth of the measurements
are well-established. A 2 tonne sinker weight keeps the
chain near-vertical under tow at a speed of 4 knots
(about 2 m/s). The chain uses digital data communication
and contains only six wires, two for power supply, two
for data, and two for control. Sampling the 100 pods
takes only 0.9 s. This particular chain has been used on
14 ocean measurement surveys, and more than 30 weeks
of continuous data have been collected, comprising
some 38,000 km of track. Much of the work has been
directed toward the detection and tracking of fronts and
eddies in the area between Iceland and the Faeroes, 19.11
where waters of the North Atlantic meet those of the
Norwegian Sea. (Courtesy of Dr J. Scott, Defence
Research Agency, Dorset, UK.)

Figure 19.11 The Lightfish is an instrument to measure

multispectral reflectance at high spatial resolution. There
are six sensors pointing up and down, the six white ports
in the upper black bar, within which are mounted the
irradiance sensors sampling at 410, 440, 490, 520, 550,
and 670 nm. It is towed just below the water surface
from a ship, collecting data from ocean transects which
can be compared with observations from airborne or
satellite sensors; this provides ‘ground-truth’ calibrations
of these sensors for ocean colour estimates, which can be
related to the near-surface distribution of phytoplankton
(see Chapter 14). Data are logged onto a ship-board com-
puter, together with other relevant data such as tempera-
ture, fluorescence, and transmittance. (Courtesy of Dr A.
Weeks, SOC.)

286
 19: Marine Instrumentation

Figure 19.12 (a) Deployment of an Aanderaa™ current 19.12a

meter mooring from the foredeck of RRS Discovery (cour-
tesy of SOC). (Inset) A recovery beacon that enables the
ARGOS satellite system to track and locate the mooring if
it surfaces prematurely (courtesy of M. Conquer, SOC).
(b) A release mechanism with biofouling after recovery
from a 12-month deployment. (Courtesy of M. Conquer,
SOC.)

Instruments fixed in position

Current meter moorings
Obtaining information on the behaviour of the
ocean over time-scales of more than a few days
requires scientists to resort to self-recording instru-
ments moored to the sea floor, or left drifting.
Ingenious non-electronic self-recording instruments
were designed before the arrival of the transistor,
but the rapid advances of recent years owe much to
the microprocessor and to high-energy density bat-
teries.
The highly successful Aanderaa™ current meter
was designed in the early 1960s. Using mechanical
encoding of the rotation of a Savonius rotor and
the direction of a large vane, it owed a great deal to
the weather vane and anemometer.
Nevertheless, the instrument became the stan-
dard for measuring deep-ocean currents, and in a
solid-state form is still popular today. Capable of
being deployed for periods of a year and more, its
simplicity, reliability, and relative cheapness make
it an almost ideal oceanographic tool.
In depths of up to some tens of metres, typical of
waters close to shore in the shelf seas, equipment
may be moored to the sea bed and to a surface 19.12b
buoy, enabling simple recovery.
Such a simple mooring is not practical in the
deep ocean, where the buoyancy is usually placed
beneath the surface, well away from the influence
of surface waves and shipping or fishing (see Figure
19.12).
The problems are then, first, to find the mooring
and, second, to retrieve it. A single instrument, the
acoustic release, gives the answer. When the contin-
ually listening unit hears a coded sound-pulse from
a ship it sends out a reply signal that indicates the
range from the ship to the mooring.
As the ship homes in, another signal activates
the release mechanism to separate the mooring
from its anchor, so the mooring returns to the sea equipment, and have a small probability of early
surface. failure, satellite position-indicating transmitters
When rising the release provides a beacon signal, may be fitted to the buoyancy. These alert the labo-
often augmented on the surface by radio transmit- ratory to moorings that have surfaced prematurely,
ters, flashing lights, or radar reflectors. giving position and drift information that aids their
As moorings contain increasingly more valuable recovery.

287
 G. Griffiths and S.A. Thorpe

19.13 19.14

Figure 19.13 A surface meteorology buoy, with Figure 19.14 A Wavecrest buoy, developed by the
anemometers (for mean wind speed and direction), sea Netherlands company Datawell, being deployed. The
and air temperature sensors, a 3 m path acoustic current buoy follows the sea surface and sensors measure the
meter, a buoy motion package (to give wave height and components of acceleration, which, integrated twice,
directional spectrum), and radio and satellite data give the wave-height variations. The buoy is moored to
telemetry (see Chapter 2 for a discussion of the way the the sea bed using a compliant tether – usually a length of
atmosphere affects the ocean). (Courtesy of A. Hall, thick rubber line – to avoid mooring forces affecting the
SOC.) record. (Courtesy of C. Griffiths, Dunstaffnage Marine
Laboratory, Oban, Scotland.)

Moorings measuring meteorology ment duration is short, or may be autonomous to

and surface waves give measurements over long periods.
Surface meteorological buoys (Figure 19.13) com- Geophysical experiments make extensive use of
bine accurate sensors, replicated for reliability, with sea-bottom seismographs that record the signal
data telemetry via satellite and terrestrial radio associated with natural earthquakes or from
links (see also Figure 3.24). Additional sensors that induced sound sources, such as explosives and air
measure the motion of the buoy give estimates of guns (Figure 19.15; also see Chapter 8).
wave period, height, and direction. An example of Scientists studying the geochemistry of the sedi-
an instrument designed purely for wave measure- ment–water boundary require water samples from
ment is shown in Figure 19.14. within the sediment (the pore waters). Bottom lan-
ders have been built that contain hydraulically dri-
Bottom-mounted instruments ven syringes to penetrate the sediments, take samples
Benthic landers are instrument packages that pro- at a range of depths, and retract in readiness for
vide observations near the sea floor. They may be recovery. Other sensors, such as those to measure
deployed from the ship’s warp, when the experi- pH and oxygen concentration within the sediments,

19.15

Figure 19.15 The Digital Ocean Bottom Seismometer

(DOBS) is a self-recording listening station with an in-
water hydrophone and geophones in contact with the sea
floor. Sound generated from ship-towed air guns or from
dropped explosive charges reaches the DOBS directly
through the water and also through the ocean floor,
through which the sound speed is much higher. From the
characteristics of the different propagation paths, the
nature of the sea floor can be inferred. (Courtesy of SOC.)

288
 19: Marine Instrumentation

19.16 Figure 19.16 This sampler contains hydraulically driven

syringes that penetrate the sea floor to take sediment and
pore-water samples at a range of depths. Other sensors,
such as pH and oxygen probes, may be fitted to the lan-
der. (Courtesy of SOC.)

19.17

Figure 19.17 Work on research vessels proceeds through

the night – here deploying the Sediment Transport and
Boundary Layer Equipment (STABLE) benthic instrument
platform with current and high-frequency acoustic sedi-
ment transport sensors. (Courtesy of J Humphrey,
Proudman Oceanographic Laboratory, Bidston, UK.)

19.18
can be included on these landers (Figure 19.16).
Movement of sediment along the sea floor helps
to shape much of the coastline, so several instru-
ment platforms have been designed to study sedi-
ment concentration and transport close to the sea
bed. Rapid turbulent motions, critical in causing
fine sediment to become suspended, need to be mea-
sured along with the concentration or mass of the
suspended matter. Measurements are made several
times a second. Instruments based on the principle
of electromagnetic induction are suitable and can be
made sufficiently robust. A voltage sensed by elec-
trodes on the instrument is generated by the move-
ment of the conducting sea water through a magnet-
ic field produced by an internal solenoid. These cur-
rent meters can be combined with acoustic probes
to measure the sound scattered from the suspended
particles in bottom landers (Figure 19.17). Figure 19.18 Sea-level changes of centimetres can be
Increasing emphasis is being given to obtaining observed from the deep-ocean floor by the MYRTLE
data over many years from the deep ocean to moni- package. Designed for operation over 5 years, MYRTLE
tor natural and man-made change. Some locations releases data ‘podules’ to the surface, where, as they
lend themselves to regular visits by research ships, drift, they telemeter data to the laboratory via ARGOS
others are either too remote or visited too infre- satellites. The package shown here is being deployed on
quently. New bottom-mounted instruments have the continental slope off the Antarctic Peninsula, the
been designed to solve the problem of obtaining southern boundary of Drake Passage. With another MYR-
data over several years from remote locations. The TLE package on the northern slope off South America, the
slope of the sea level between the two sites provides a
MYRTLE (Multi Year Tide and Sea Level
measure of the transport of water through the Drake
Equipment) package releases data podules from the Passage by the Antarctic Circumpolar Current, and of its
sea bed that, on the surface, return their data via variability (see Chapter 4). (Courtesy of Proudman
satellite to the laboratory (Figure 19.18). Oceanographic Laboratory, Bidston, UK.)

289
 G. Griffiths and S.A. Thorpe

19.19a 19.19b

Figure 19.19 Here are two examples of free-fall instru-

ments. (a) EPSONDE4, a tethered ‘semi-free-fall’ device as
shown in the top right-hand sketch. Data from the slowly
falling probe is transmitted up the cable to recorders on
the deployment vessel. The cable is sufficiently light and
thin that it does not affect the smooth descent of the
instrument through the water. The central photograph
shows the instrument being deployed at sea. The lower
right-hand corner shows the shear probes used for mea-
suring subcentimetre velocity fluctuations in the turbulent
ocean, together with thermistors and a thin-film ther-
mometer to measure temperature (courtesy of Dr N.
Oakey, Bedford Institute of Oceanography, Dartmouth,
Nova Scotia, Canada.) (b) A High Resolution Profiler
(HRP) during recovery. This instrument, developed by a
team at Woods Hole Oceanographic Institution, Woods
Hole, Massachusetts, is internally recording and is free to fall through the water with full ocean-depth capability7.
Once the maximum depth has been reached, ballast weights are released by a solenoid mechanism on command from
an on-board instrument computer and the HRP returns to the surface. The spikes or whiskers at the top are to increase
the drag and so reduce the instrument fall speed. The instrument cradle to reduce damage in deployment and recovery
is visible to the left. The instrument carries sensors similar to those of EPSONDE (courtesy of Dr J. Toole, Woods Hole
Oceanographic Institution, Woods Hole, Massachusetts, USA).

Free-fall instruments and ocean microstructure

The discovery of ocean microstructure in the 1960s the water at about 1 m/s has been achieved by
drew attention to the turbulent mixing which must decoupling the instrument from the research vessel
accompany it and offered the challenge to design and allowing it to free-fall before recovery by line
instruments to measure velocity down to the scale [Figure 19.19(a)], or by the instrument’s releasing
of millimetres, at which energy is lost by viscous ballast which leaves it buoyant [Figure 19.19(b)].
dissipation. Osborn5 and Siddon’s solution was to [Alternatively, microstructure probes can be
use a pair of piezoceramic bimorph beams – allow- mounted on a very stable platform, such as a sub-
ing flow-induced pressure fluctuations to be regis- marine 6 or a mooring (Figure 19.20).] Accom-
tered as voltages – embedded together in a 6 mm panying measurements of pressure (from which the
diameter cylindrical probe which, when carried mean speed – the fall speed – is found), tempera-
steadily through the water pointing ahead of the ture, salinity, and tube orientation observations are
supporting body, responds rapidly to lift variations usually made from the free-fall instruments when
and therefore to the two components of velocity velocity microstructure is measured. Measurements
normal to the direction of motion of the probe. have been made right to the sea surface by con-
These small components can be determined once structing instruments which, having released bal-
the orientation of the probe and its mean speed last, rise upward with sensors extended on mount-
through the water are known. Smooth, vibration- ings above the tube containing the power supply
free, slow steady motion of the instrument through and the recording and control electronics.

290
 19: Marine Instrumentation

Figure 19.20 Operating free-fall devices to measure 19.20

microstructure is very expensive because of the cost of
maintaining ships at sea, typically $US15,000 per day.
Cost reduction and the need for very long-term measure-
ments prompted the development of a moored instrument
for the study of ocean mixing. The Tethered Autonomous
Microstructure Instrument (TAMI), shown here during
deployment, carries sensors (protected by the ring at the
right of the instrument), which detect currents as small as
0.1 mm/s and temperature changes of 10–5 K over a sam-
pling volume of diameter about 5 mm. On-board comput-
ers process the measurements and record data on disk dri-
ves for 6 months. The instrument is anchored in water up
to 5500 m deep. A 1.5 m diameter syntactic foam float
(about 670 kg buoyancy) holds the mooring taught while
the instrument (about 90 kg buoyant) floats some 7 m
above on a line attached to the shackle on the spar shown
(just below the fin on the instrument tube). The tube
swivels into the prevailing current. An acoustic command
to a release located about 20 m above the anchor on the
sea bed returns all but the anchor to the surface for recov-
ery. The instrument was developed by Dr R. Lueck at the
University of Victoria, Canada. (Courtesy of Dr R. Lueck,
University of Victoria, BC, Canada.)

Measurements using tracked devices 19.21

Surface drogue floats and drifters

Drifters or floats released at sea that carry requests
to the finder to report the time and location of
recovery have for centuries provided information
about the mean current drift at the sea surface. A
major question was always whether drifters were
driven by the wind or by the currents.
Radio and satellite navigation and telemetry has
led to a resurgence in using surface drifters. The
ARGOS system, developed by France and the US,
uses low-power satellite transmitters on buoys to
give positions with to an accuracy of <1 km, and
up to 32 bytes of data per transmission. Shown in
Figure 19.21, drifters are designed with a minimal
area above water, and are fitted with drogues at
depths of tens to hundreds of metres to tag and fol- Figure 19.21 An ARGOS satellite-tracked buoy and its
low the ocean currents and eddies (Chapter 4). drogue on deck. The drogue consists of concentric tubes
These buoys are used in large numbers. From of nets 18 m long, the outer tube of 0.5 m diameter.
1978–1981 some 300 were released in the Southern These drogues have been used down to 800 m. The buoy
Ocean. The buoys measure sea-surface temperature contains sensors for temperature and for detecting the
and barometric pressure, as well as currents. The presence of the drogue; weathering many storms over
World Ocean Circulation Experiment (WOCE) operating lives of years eventually leads to the loss of
drogues from many buoys. If this is not detected, false
called for some 3800 buoys to be deployed globally
conclusions may be drawn about the currents or ocean
from 1990–1996. eddies (see Chapter 4). (Courtesy of M. Conquer, SOC.)
Deep drifting buoys
New insights into the strength and variability of
currents in the deep ocean came from the invention less than that of sea water. Through careful balanc-
(in the early 1950s, by Dr John Swallow) of the ing, the float could be made neutrally buoyant at
neutrally buoyant float. The first experiments in any pre-set depth. Fitting an acoustic transmitter to
1955 proved the idea – that an aluminium tube, the float enabled it to be tracked from a ship using
negatively buoyant at the surface, would gain triangulation methods (Chapter 20). Examples of
buoyancy as it sank, because its compressibility is the tracks of such floats are given in Figure 4.6.

291
 G. Griffiths and S.A. Thorpe

19.22 Figure 19.22 A WOCE-style RAFOS float

being held before deployment on R/V
Knorr. This version of the float was
recently developed by the University of
Rhode Island and the Woods Hole
Oceanographic Institution, and is manu-
factured by Seascan Inc. RAFOS floats
are ballasted to be neutrally buoyant at a
pre-selected depth, where they drift for
as long as several years, showing the
water motions (see Chapters 3 and 4).
The floats receive acoustic signals from
moored sound sources, record the times
of arrival so that their position can be
determined, and (at the end of their mis-
sion) surface and transmit the data,
including temperature and pressure, back
to the laboratory via the ARGOS satellite
system. (Courtesy of Dr T. Kleindinst,
Woods Hole Oceanographic Institution,
Woods Hole, Massachusetts, USA.)

For a period of two decades the basic Swallow and pressure on the speed of sound (see later).
float was refined, which led to the development of These floats communicate over ranges of 1500 km
the SOFAR float in the early 1970s. The SOFAR and more, with autonomous listening stations, thus
channel is an acoustic wave guide in the ocean, eliminating the need for ships to attend the floats.
arising from the combined effects of temperature A derivative of the SOFAR float – termed RAFOS
(see Figure 19.22) – reverses the roles of the float
19.23 and listening station; the listening is done on the
float, and the sound sources are moored and recov-
erable. This reduces costs, since sound transmitters
are far more expensive than receivers.
The next logical step was to make floats inde-
pendent of a tracking network. The Autonomous
Lagrangian Circulation Explorer (ALACE) was
developed by the Scripps Institution of
Oceanography, along with the Webb Research
Corporation, in the late 1980s (see Figure 19.23).
These modern floats are used in large numbers; the
WOCE experiment has commitments for over 1000
to be used from 1990–1996.

Figure 19.23 An ALACE float, showing the satellite

telemetry and location antenna. Instead of obtaining
position fixes from moored sound sources, the ALACE
float periodically increases its buoyancy and rises to the
surface every 20 days. There, by using the ARGOS plat-
form location and data transmission satellite system, the
ALACE float can be located and data sent to shore before
the float returns to its pre-set depth. The major challenge
in designing the ALACE float was to devise a reliable,
high-efficiency buoyancy control system that could func-
tion at ambient pressures of up to 200 bar. The buoyancy
change is performed by changing the effective volume of
the float. An electrically driven hydraulic pump moves
oil from a reservoir internal to the float to an external
bladder. An example of the float tracks is shown in Figure
3.13. (Courtesy of Dr R. Davis, Scripps Institution of
Oceanography, La Jolla, California, USA.)

292
 19: Marine Instrumentation

19.24a 19.24b

A B

0 5 10

C D

Figure 19.24 The ‘electronic pebble’ is a tracer devel-

oped to monitor the mobility of shingle on beaches. (a)
The system comprises synthetic pebbles consisting of a
battery-operated circuit (A, B, and C) encapsulated in
resin (D), which emits a coded train of magnetic pulses
(scale bar, 0–10 cm). These are placed on the beach at
low tide and can then be detected and decoded by a
portable receiver (b) or a towed receiver (c) when in
close proximity. Their movement indicates the motion of
similar pebbles. (Courtesy of Dr M. Workman, SOC.)

19.24c

Movement of shingle remote sensing employed in the study of the ocean,

An example of a specialist development is the ‘elec- subsurface acoustics, of which the most familiar is
tronic pebble’, used to track the movement of shin- the conventional echo-sounder, and land-based
gle on beaches (Figure 19.24). radar. The use of optics is limited to ranges of only a
few metres generally, because of the high attenuation
Remote Sensing Instruments of light by sea water3 compared with the tens of
Excluding airborne and satellite remote sensing metres to thousands of kilometres distances (depend-
(Chapter 5), there are two principal methods of ing on frequency8) over which sound can propagate.

293
 G. Griffiths and S.A. Thorpe

19.25 19.26

Figure 19.25 GLORIA, the Geological Long Range Figure 19.26 Launch of the TOBI deep-towed vehicle.
Inclined Asdic, is housed in a launch/recovery gantry on Towed some 100–300 m above the sea floor at a speed
the stern of RRS Discovery. For more than 20 years of 1–2 m/s, TOBI does not need to be streamlined.
GLORIA has been used for long-range side-scan sonar Underneath the orange syntactic foam buoyancy are two
surveys, during which time it has surveyed about 6% of side-scan sonar arrays, one port and one starboard. It also
the ocean floor. The instrument can survey 20,000 km2 a has a magnetometer and a sub-bottom profiler (7 kHz).
day, and over a 7-year period covered the entire 7 mil- The vehicle can also carry CTDs, transmissiometers, and
lion km2 of the US Exclusive Economic Zone. GLORIA chemical sensors, which are useful for locating
images show large features on the ocean floor, such as hydrothermal vent sites. Examples of TOBI sonographs
volcanoes and sea-mounts, channels, sediment slides, are given in Chapter 10. (Courtesy of SOC.)
and geological hazards. Examples of GLORIA sono-
graphs are given in Chapters 9 and 10. (Courtesy of
SOC.)

Acoustic remote sensing Acoustic Doppler Current Profilers (ADCPs)

Sonar can be used in two principle ways, ‘active’ or The Doppler principle – the change in pitch of a
‘passive’. In the ‘active’ mode, sound is produced wave due to relative motion between the source
by an instrument, usually in short pulses, and the and receiver – was first used successfully in the
reflections are recorded by the same instrument or early 1960s by Kocsy and others at Miami to mea-
by another. The range of the reflecting ‘targets’ (for sure currents. However, it was not until the early
example, the depth of water below the ship carry- 1980s that the technique became widely adopted
ing the sonar) can be determined from the time for both ship-mounted and in situ instruments. Its
between emission of sound and its return, and the main advantage over conventional current meters is
known speed of sound in sea water, a time of that it is a remote sensing method. Sound from a
approximately 1500 ms. ‘Passive’ sonars receive transmitter is directed along a narrow beam away
sound over a broad band of frequencies and can be from the instrument and is scattered back toward
used, for example, to quantify the occurrence or the instrument, often from particles or zooplankton
detect the presence of physical processes that pro- being carried by the current. A receiver compares
duce sound, such as rainfall (producing sound at the pitch of the received and transmitted frequen-
about 12 kHz) or the cracking and break-up of cies, and from their difference, or Doppler shift,
winter ice. calculates the current at different distances from
the instrument – hence the ‘profile’. Modern
Survey of the sea bed ADCPs can measure currents accurate to 1 cm/s in
This is the subject of Chapter 20. Figures 19.25 up to 128 depth slices, each typically 4–8 m thick;
and 19.26 provide examples of modern deep-sea they can be used from a ship (Figure 19.27) to pro-
survey instruments. vide a detailed picture of the currents of the upper
Bottom-mounted single instruments using active ocean.
or passive sonar at frequencies of 1 MHz or more Moored and bottom-mounted ADCPs can pro-
(with a consequent sonar range of 30 m or less), vide high-resolution current measurements in dis-
are also used to detect and quantify sediment in tance and time. They contribute to studies of inter-
suspension, often alongside electromagnetic current nal waves in the open ocean and have been used
meters (Figure 19.17) to measure the vertical distri- extensively through Arctic ice floes, enabling mea-
bution of current and the stress of the turbulent surements that could not otherwise be obtained.
flow on the sea bed. The ADCP also has another use – it may be used to

294
 19: Marine Instrumentation

Figure 19.27 A four-transducer cluster 19.27

for the acoustic Doppler current profiler
on the hull of RRS Discovery used in the
survey of ocean currents and eddies (see
Chapter 4). Acoustic transducers on
ships’ hulls need careful positioning. Too
near the bow and bubbles from the bow
wave interfere with sound propagation
from the transducer; too near the stern
and the transducers pick up noise from
the propellers. (Courtesy of R. Bonner,
SOC.)

infer the distribution of zooplankton in the open nelled (the SOFAR channel, see above). Low fre-
ocean from the strength of the acoustic back-scatter quency (<100 Hz) sound propagates around the
signal. global ocean over distances exceeding 15,000 km,
unless impeded by land masses. Changes in the
Acoustic study of the upper ocean arrival time of coded pulses of sound, which can be
Upward-pointing sonar and hydrophone arrays measured very accurately, provide information
have been used to investigate the dynamics of the about the changing structure of the water in the
upper ocean, especially turbulence and breaking propagation path. Over ranges of 1000 km,
waves; examples are given in Figures 19.28–19.29. acoustic ‘tomography’ can provide information
about the structure, variation, and propagation of
Long-range acoustics meso-scale eddies; while over 10,000 km ranges,
The speed of sound in sea water depends primarily experiments repeated over a few years should pro-
on temperature (increasing by about 4 m/s per K) vide information about the variation of tempera-
and pressure; except in Arctic waters, the minimum ture averaged over the acoustic path, and hence
sound speed occurs in mid-water, typically at some about changes in ocean climate. This is an objective
800 m depth, where sound is trapped and chan- of the Acoustic Thermometry of Ocean Climate

19.28

Figure 19.28 An array of six long-range Doppler sonars

mounted on the hull of the research platform FLIP. With
FLIP in the vertical orientation (as shown in Figure 19.1),
the four two-panel sonars in the foreground point down-
ward and outward, defining the edges of a four-sided
pyramidal measurement array. The larger six-panel
sonars are directed horizontally, just below the sea sur-
face. Each beam transmits pulses of 20–30 ms duration at
a frequency of 67–80 kHz and a peak power of 2 kW.
The sound scatters from zooplankton drifting with the
water. From the Doppler shift of the echo, the radial
component of water velocity, and the characteristics of
internal waves and upper-ocean structure can be deter-
mined to ranges as great as 1.2 km. This system was
developed in 1979 and has been used extensively since.
(Courtesy of Dr R. Pinkel, Scripps Institution of
Oceanography, La Jolla, California, USA.)

295
 G. Griffiths and S.A. Thorpe

19.29 Figure 19.29 A hydrophone array, used to ‘listen to’ and

track the motion of breaking waves, which play an
important part in the transfer of momentum and gases
between the air and the sea (see Chapter 2), being recov-
ered using a Zodiac inflatable boat. Recovery of equip-
ment at sea is often a difficult and potentially dangerous
operation, requiring careful attention to stringent safety
precautions. For further details of the equipment and its
use, see Farmer and Ding1. (Courtesy of Dr D. Farmer,
Institute of Ocean Sciences, Sidney, British Columbia,
Canada.)

19.30a 19.30b

Figure 19.30 Acoustic transducers used in a successful pilot experiment to propagate coded pulses of sound at 57 Hz
from close to Heard Island in the Southern Indian Ocean in late January and early February 1991 to receiving stations
around the world’s ocean, some over 15,000 km away. Here, the acoustic sources are shown (a) ready to be lowered
off the support ship (fairings were fitted to smooth the flow around the cylindrical transducers) and (b) on recovery
after several days of severe weather. Fairings were torn away, exposing the transducers to the full force of the Southern
Ocean swells. The experiment was designed to test the feasibility of developing an ‘acoustic thermometer’, using
acoustics to measure the changes in mean ocean temperature (sound speed increases with temperature). An ocean
warming of 4 mK per year would result in an increase in sound speed of about 0.02 m/s per year which, at a range of
15,000 km reached by sound in about 10,000 s, would lead to a reduction in travel time of about 0.13 s per year,
which is easily measurable (see text and Figure 3.16). Plans have been made to establish a global network of transduc-
ers and listening stations to monitor climate change in the ocean. (Courtesy of Dr A. Forbes, Scripps Institution of
Oceanography, La Jolla, California, USA.)

(ATOC) program, of which the Heard Island Future instruments and methods
Experiment (Figure 19.30) was a forerunner2.
The development of sensors
Radar The greatest demands for automated methods of
Land-based radar provides another powerful tool ocean monitoring and survey are now from the dis-
to study the surface of the ocean. One example of ciplines of biology and chemistry. As knowledge of
its use is illustrated in Figure 19.31. the ocean advances, the limitations of nets (biolo-

296
 19: Marine Instrumentation

Figure 19.31 The Ocean Surface Current Radar 19.31

(OSCR), deployed at Crammag Head Lighthouse
on the Galway Peninsular, W. Scotland. In the
foreground are three elements of the 85 m long,
16-element receiver array. Each element stands
2.25 m high. Behind the receiver array are four
elements of the transmitter array, which are
5.5 m high. The lighthouse contains the radar
hardware and computer. The OSCR radar trans-
mits electromagnetic waves in the HF radio
band at 27.0 MHz, corresponding to a wave-
length of 11 m, which are scattered from the
waves on the sea surface. Scattering comes
mostly from waves with half the radar wave-
length, ‘Bragg scattering’, in a manner similar to
that of X-rays from a crystal lattice. The frequen-
cy (Doppler) shift of the returning radar waves is
measured by the receiver, and the speed of sea
waves can thence be measured. This is a combi-
nation of their movement through the water (which is well-known from wave theory) and the speed of the radial com-
ponent of the surface current, which can be estimated by subtraction. The use of a multi-element array produces a nar-
row beam, which can be digitally steered through 90˚ of azimuth. The OSCR system collects data from 1 km2 ‘cells’,
with a maximum of 700 data cells and a maximum range of 40 km. Two radars are used with orthogonal coverage of
the required survey area to resolve the two components of surface vector current. (Courtesy of Mr R.D. Palmer, SOC.)

gy) and laboratory analyses (chemistry) become The other approach is to devise novel sensors or
more obvious and, indeed, will limit scientific adapt existing sensors, for example:
progress. Improvements can only come about
through new chemical and biological sensors. • Ultraviolet and infra-red absorption – some ions
One approach is to create new instruments by exhibit an absorbance at well-defined wave-
miniaturising equipment found in the laboratory, lengths in the ultraviolet or infra-red regions.
reducing its power consumption, and removing any Nitrate concentration in the ocean is particular-
need for human intervention. These are testing ly important, and may be determined through
requirements, but may be achievable for: ultraviolet absorption. New ultraviolet lamps,
narrow ultraviolet filters, and in situ signal pro-
• Flow cytometry, a technique developed for med- cessing are now making this possible.
icine, in which the characteristics of single cells • Biosensors, already used in medicine, integrate a
are analysed at rates of thousands per second chemical substance or enzyme with a semicon-
using laser optical methods. Developments in ductor to give a transducer that converts direct-
solid-state lasers and fibre-optics now make it ly the concentration of a substance into an elec-
feasible to consider marine in situ flow cytome- trical signal (e.g., sensors for glucose are now
ters. The biomass of single-celled organisms in commonplace). As a technique, this has poten-
the ocean is immense and poorly known, so this tial to add to our knowledge of the microbio-
technique is potentially very powerful. logical and organic chemistry of the ocean.
• Broadband acoustics and miniature low-light
cameras, used in medicine for imaging, may Novel vehicles and platforms
provide a means of studying zooplankton The use of submersibles that carry scientists and
species distribution and behaviour, building on instruments to investigate the ocean depths is
the present use of acoustics for estimating abun- becoming routine (Figure 19.32). The advent of
dance. cheap disposable fibre-optic cable enables minia-
• Flow injection chemical analysis, a standard ture, remotely operated vehicles (ROVs) to be
laboratory technique, is being adapted for in designed for applications such as the acoustic and
situ use. Micromachining, with pumps and optical imaging of zooplankton. The fibre-optic
valves based on electrostrictive materials – link provides real-time control of the vehicle and
materials that change their physical dimensions video data from sensors (Figure 19.33).
in response to an applied electrical signal – The task of making routine measurements in the
could provide a route to miniature analysers ocean to meet the demand of present computer
that require minimal amounts of reagents and models is growing beyond the capability of the
maintenance. world’s fleet of research vessels, even with ROVs.

297
 G. Griffiths and S.A. Thorpe

19.32a 19.32b

19.32c

19.32d

Figure 19.32 Some of the submersibles used for investigation of the deep ocean. (a) The French submersible, Cyana,
capable of operating to depths of 4000 m. The hull is titanium, and the submersible carries a pilot, navigator, and one
scientist. It is usually equipped with underwater lights, an arm, claw, and video. (Courtesy of Professor P. Tyler, SOC.)
(b) The Russian Mir I submersible, capable of operating to 6000 m carrying two pilots and a scientist. The hull is titani-
um. The submersible is here carrying a (red) sediment trap, and still and video cameras; it has been used to investigate
hydrothermal systems in the North Atlantic (see Chapter 13). (Courtesy of Professor P. Tyler, SOC.) (c) The US Johnson
Sealink submersible in preparation for deployment. It is capable of operating to 900 m with a pilot and scientist in the
front compartment and with a technician and second scientist in an aft compartment. The sphere is acrylic. In addition
to lights and cameras, the submersible has a manipulation claw. (Courtesy of Professor P. Tyler, SOC.) (d) The US
Navy-owned Deep Submergence Vehicle (DSV) Alvin operated by the Woods Hole Oceanographic Institution. A typi-
cal 8 hr dive takes two scientists and pilot to 4500 m with 4 hours on the bottom for observation, photography, and
experiments. Three video and two 800 frame cameras are usually carried, together with two hydraulic arms and instru-
ments such as corers, temperature probes, water samplers, and a biological sample pump. (Courtesy of Woods Hole
Oceanographic Institution.)

298
 19: Marine Instrumentation

Figure 19.33 The Remotely Operated 19.33

Vehicle (ROV) Jason, named after the
Greek hero who searched for the
Golden Fleece. It is owned and operat-
ed by the Woods Hole Oceanographic
Institution and has been widely used in
studies of the deep ocean floor. An
early version was operated from DSV
Alvin [Figure 19.32(d)] during the expe-
dition, which discovered the wreck of
the ‘Titanic’ in 1986. Launched from
the support vessel to which it is con-
nected by a 10 km long fibre-optic
cable, Jason carries cameras and sonar
from which signals are transmitted back
through the cable to the pilot, who
steers the vehicle from the support ship,
and to the scientist directing operations
and the research team. Images and nav-
igation data are recorded for subse-
quent analysis. (Courtesy of the Woods
Hole Oceanographic Institution.)

Computer power is rising exponentially, continual- Leaman, K.D. (1990), Physical oceanographic measurement
ly increasing the gap between modelling and obser- techniques at sea, in The Sea: Ocean Engineering
Science, Vol. 9B, Mehaute, B. and Hanes, D.M. (eds),
vations. Autonomous vehicles and systems provide J. Wiley & Sons, New York, pp 1163–1192.
a low-cost way to scale up our observations. The Urick, R.J. (1975), Principles of Underwater Sound, 2nd edn,
ARCS vehicle from ISE Research Ltd, Canada, the McGraw-Hill, New York.
Odyssey vehicle from MIT/Woods Hole
Oceanographic Institution, and the GEC-Marconi
References
AUV have successfully demonstrated autonomy for 1. Farmer, D.M. and Ding, L. (1992), Coherent acoustical radi-
ation from breaking waves, J. Acoust. Soc. Amer., 92,
scientific data-gathering over limited depths and for 397–402.
short periods. The UK Autosub project aims to 2. Heard Island Principles (1991), The Heard Island experiment,
provide scientists with a deep-diving vehicle able to Oceananus, 34, 6–8.
collect routine physical, chemical, biological, and 3. Jerlov, N.G. (1968), Optical Oceanography, Elsevier,
Amsterdam.
geophysical data over transects of several thou- 4. Oakey, N.S. (1988), EPSONDE: an instrument to measure
sands of kilometres (Figure 3.15). turbulence in the deep ocean, IEEE J. Oceanogr.
Eng., 13, 124–128.
5. Osborn, T.R. (1974), Vertical profiling of velocity
microstructure, J. Phys. Oceanogr., 4, 109–115.
6. Osborn, T., Farmer, D.M., Vagel, S., Thorpe, S.A., and Cure,
M. (1992), Measurements of bubble plumes and tur-
bulence from a submarine, Atmosphere–Ocean, 30,
General References 419–440.
7. Schmitt, R.W., Toole, J.M., Koehler, R.L., Mellinger, E.C.,
Griffiths, G. (1992), Observing the ocean – recent advances in and Doherty, K.W. (1988), The development of a
instruments and techniques for physical oceanogra- fine- and micro-structure profiler, J. Atmos.
phy, Sci. Progr., 76, 167–190. Oceanogr. Tech., 5, 484–500.
Jerlov, N.G. (1968), Optical Oceanography, Elsevier, 8. Urick, R.J. (1975), Principles of Underwater Sound, 2nd edn,
Amsterdam. McGraw-Hill, New York.

299
 CHAPTER 20:

The Sea Floor –

Exploring a Hidden World
P. Riddy and D.G. Masson
Introduction about how much water was beneath their keel (safe-
Almost since the human race began to interact with ty) than exactly where they were (their position, see
the sea, a knowledge of the sea-bed structure and Figure 1.3). The days of navigation or position-fix-
composition has become interwoven with the lives ing by sextant [(Figures 20.1(a)–20.1(c)] and depth
of those who work in the marine environment. At measurement by weighted line are largely over,
first an experiential and intuitive knowledge proba- although these techniques illustrate well the basic
bly allowed boatman and sailors to find sources of tools of bathymetric surveying (indeed, they may
food and, avoiding hazardous rocks and reefs, to still be used where resources to purchase advanced
return safely to shore. Later, as foraging trips went technology are limited).
further afield and perhaps exploration began, some Today, we are also interested in the wide range
sort of record or simple chart would have been of material of which the sea floor is made, its
made of where shoals and deeps occurred. potential value (see Chapter 21), and how it is
Eventually, weighted lines were used to measure shaped and organised by the overlying water. So
depth; when such measurements were recorded the depth, the variations in sediment type, and the
along with the location at which they were made, outcropping rocks are together useful in the charac-
bathymetric surveying was born. Position-fixing and terisation of the sea floor. Other factors, such as
depth measurement have thus always been intimate- the relationship between the species which live on
ly related, although the criers of ‘by the mark’ when and in the sea bed (see Chapter 13) and the sedi-
‘heaving the lead’ were usually more concerned ments/rocks which are present (see Chapter 8),

20.1a 20.1b 20.1c

Figure 20.1 Sextants (a) are traditional navigation instruments

which have been in use (b) since 1731. The diagram (c) illustrates C
the use of sextants for position-fixing in good visibility up to a few
kilometres from the landmarks. Two sextants are used to measure
simultaneously the angles between three landmarks or objects of
already-known position – further calculation allows two intersect-
ing position circles to be drawn6. This technique can give accura- A
Error in position
cies of a few metres, but using sextants from the deck of a boat (an associated with
unstable platform) requires practice! (d) Lines A, B, and C are beacons A and B
known as position lines. Each position line has an associated error
which is dependent on the position-fixing system being used. This B
diagram shows the effect of errors associated with two of the posi-
tion lines. Instead of the position being defined at a single point, 20.1d
the vessel’s position lies somewhere within the parallelogram. No
matter what position-fixing system is used, an error will always be
associated with each position line. For high-accuracy positioning Parallelogram of
error from intersection
at least three position lines are normally used to try and minimise of position lines
the effect of errors. from A and B

300
 20: The Sea Floor – Exploring a Hidden World

Figure 20.2 GLORIA 6.5 kHz side-scan sonar image of a 20.2

submarine volcano in the eastern Pacific off South
America. Strong sonar targets are white, and acoustic
shadows are black. The volcano mouth is approximately
2 km in diameter.

contribute to the more complete picture. The sub-

surface structure of the sea floor influences the visi-
ble surface structure. The chemical composition of
the sediments can influence the range and diversity
of life present and the nature of the sea floor. These
factors, in turn, influence or are influenced by the
overlying currents. In this chapter, we focus on the
basics of the methodology and technology used to
investigate the geological and physical characteris- track over the area of interest, with the appropriate
tics of the sea floor in shallow and deep water envi- sensors being deployed to continually record all the
ronments, and illustrate some of the features found data (Figure 20.3).
(Figure 20.2). The structures of the sea bed and ocean basins
revealed by geophysical survey techniques are
Marine Geophysical Surveying described and illustrated in Chapters 8 and 9.
The goal of the marine geophysical survey is to Several bottom-survey tools use sound waves to
provide an integrated picture of the surface, the obtain information from the sea floor, while others
subsurface structure, and the composition of the measure the magnetic field, the gravity field, or the
sea bed. Most such surveys involve remote sensing electrical field. Any system which remotely senses
of the sea bed by tools deployed at or near the sea the sea floor requires a compromise between accu-
surface. Surveys follow an exploration or search racy and/or resolution, and range – this becomes

20.3
(a)

(b)

(c)

A Survey
area

B
Figure 20.3 A survey design strategy involves (a) the selected equipment being deployed from the vessel, and (b) the
vessel following a survey track designed to optimise coverage of the area. (c) A – profilers collect a narrow ‘line’ of data
from directly beneath the vessel; B – sonar and swath systems collect a broad band of data across the vessel’s track.

301
P. Riddy and D.G. Masson

Table 20.1 Electromagnetic spectrum and position-fixing systems. Only typical parameters for each
system are given. In particular, short-range systems can be given improved range capability by ele-
vating or enlarging the antennae and increasing the output power of the transmitter, although
power transmitted is limited by legislation. With very high-accuracy systems, the movement of the
vessel and the position of the measuring devices with respect to the ‘antenna’ become important
factors in the measurement of position. (Based on Ingham6.)
Frequency Wavelength Type of electromagnetic position- Typical
fixing systema with examples working
accuracy

Radio waves
10 kHz 30 km Very long range (world-wide):
Omega (10–15 kHz) 2–4 km
30 kHz 10 km Long range (200 km):
Loran C; Lambda (100 kHz) 0.2–2 km
Decca ( 112–196 kHz) 40–400 m
Pulse 8 (100 kHz) 16–30 m
300 kHz 1 km Medium range (80–150 km):
3 MHz 100 m Hyperfix (1.6–3.4 MHz) 8–15 m
30 MHz 10 m
300 MHz 1m TRANSIT satellite (150 and 400 MHz) 200 m
GPS satellite (1572.42 MHz ) 15 m
GPS satellite with SA on 100 m
GPS in differential mode 1–10 m
Syledis (406–488 MHz) 8–15 m
Microwaves
3000 MHz 100 mm Short range (0–80 km):
Radar (3000–10,000 MHz) >1 m
Trisponder (9000 MHz) 1–5 m
30 GHz 10 mm
Microfix (5.48 GHz) 1–5 m
300 GHz 1 mm

Radiated heat
3000 GHz 0.100 mm
30,000 GHz 0.010 mm

Infra-red
300,000 GHz 0.001 mm Infra-red distance measurers a few cm

Visible light
3,000,000 GHz 0.0001 mm Lasers a few cm

a
Short range systems measure range(s) and sometimes bearing; others are predominantly hyperbolic in operation.

clearer later when the systems are described. ples and systems, as other data are of little use if we
Descriptions of devices to obtain samples of the sea do not know where they come from!
bed (e.g., cores) are omitted as they are described in Position lines
Chapter 19. Position lines can be defined in a variety of ways,
the accuracy of each position-fixing system being
Position-Fixing – The Position calculated from the error associated with, or ‘thick-
of the Vessel ness’, of the position line [Figure 20.1(d)]. All
We begin with navigation or position-fixing princi- methods involve the measurement of angles or

302
 20: The Sea Floor – Exploring a Hidden World

Figure 20.4 Types of position-fixing systems 20.4

B
and their use (see also Table 20.1). Longer range
EM hyperbolic systems (A) can provide accura-
cies of 10–500 m at ranges of 50–2000 km. All-
weather satellite systems (B) provide 24-hour
positioning to around 100 m (differential users
can be accurate to within 1 m). Sextants used
with landmarks (C) in good visibility provide A
high accuracy close to shore, and are used in
mid-ocean for astronavigation. Range–range and
hyperbolic systems (D) provide 1–10 m accura-
cy at ranges up to 80 km. Optical and EM-bear-
ing systems (E) give accuracies of 0.01–10 m at
ranges of 2–20 km. (Based on Ingham6.)
E
C

ranges, or a combination of the two. Further infor- the various systems, the Navstar Global Positioning
mation on the wide range of methods and instru- System (GPS, see below), is fast becoming the most
ments used for position-fixing is given by Ingham6 used world-wide.
and Forssell4. Today’s requirements for high-accu-
racy, all-weather position-fixing systems have led Global position-fixing system (GPS)
to the development of sophisticated electronic sys- This is a satellite-based system developed for the
tems based around the transmission of electromag- US government to provide a high-accuracy, all-
netic (EM) waves at frequencies which range from weather position-fixing capability for 24 hours per
radio waves to visible light (Table 20.1). Such sys- day anywhere on the globe.
tems (Figure 20.4) offer a variety of accuracies and The system has a number of components, illustrat-
ranges, but until the arrival of satellite systems high ed in Figure 20.5. Users can be given different levels
accuracy was only possible over short ranges. of access to the system, the highest level of accuracy
EM waves are briefly discussed in Box 20.1. Of normally being reserved for the military. The receiv-
er-position calculations are based on the satellites’
orbital parameters as transmitted by the satellites.
20.5 Space segment A modification of the these signals, known as
Selective Availability (SA), has degraded the poten-
tial accuracy of the system from 15 m to >100 m,
although full accuracy is still available to autho-
rised users by decoding.
<
<

Monitor stations
Figure 20.5 GPS components and operation. Satellites
moving in known orbits, about 20,200 km above the
Earth’s surface, transmit information on their position
Upload continuously. This is received by users and allows them
station to calculate their position – the satellites become the
Master landmarks of the user. The satellites are monitored and
control updated through the control segment network, which can
station lead, for example, to correcting deviations in a satellite’s
Control segment orbit. Various errors can be coded into the satellite trans-
missions to give an error in the calculation of position for
User segment users not equipped with the appropriate decoding infor-
mation; this typically results in positional errors of about
100 m (see text).

303
P. Riddy and D.G. Masson

Box 20.1 What is an electromagnetic wave?

An electromagnetic (EM) wave is a combination of oscillating electric and magnetic fields which trans-
port EM energy. EM waves are able to propagate in a vacuum at a speed of approximately
299,776 km/s, and cover a wide frequency band which includes visible light (Table 20.1). EM waves are
reflected or absorbed by objects and refracted by changes in the refractive index of the medium
through which they travel. Waves undergo a velocity change when refracted as they travel through
different parts of the Earth’s atmosphere. This results in a potential source of error in measuring the
ranges from satellites, which becomes important in the calculation of accurate positions.

20.6

Satellite

Reference station

Figure 20.6 Differential GPS (DGPS) involves corrections based on the known position of a
reference station being calculated and transmitted to the vessel. The vessel uses the correc-
tion information to recalculate and correct its known position, which has been calculated
from directly received satellite signals. This provides positional accuracy of less than a metre.

Differential GPS (DGPS) The position of the sensor

This development of GPS uses a reference station Some survey sensors are towed at the end of a cable,
with knowledge of its own position to calculate the which can be several kilometres in length (Figure
errors in the satellite signals caused by SA and by 20.7). If the position of the sensor is not well-
atmospheric conditions (Box 20.1, Figure 20.6). known, it is impossible to produce an accurate sea-
The errors, or an error correction, are then trans- bed map of the property being measured. In practice,
mitted in real-time from the reference station to the acoustic devices on board or placed on the sea bed
vessel, where they are applied to give much greater (e.g., an array of accurately positioned transponders
accuracy than is provided by GPS. The accuracy of or navigation beacons) may be used to measure the
DGPS partly depends on how close the reference relative position of a sensor package.
station is to the survey. Where the reference station
is within 30 km of the survey vessel, ‘instantaneous’ The Sea Floor
accuracies of less than 1 m are consistently What are we looking for? It may be the size, orien-
obtained. When used in a special geodetic mode, tation, shape, and composition of natural features
centimetre accuracies can be obtained. Some com- (or natural obstacles), or the location of manufac-
panies have set up networks of reference stations to tured objects, such as sunken vessels and mines.
routinely provide corrections that allow a positional Surveys of shallow-water environments tend to be
accuracy of around 5 m over an area extending more detailed and to find wide local variations of
hundreds of kilometres from the base stations. features, sediment type, and their distribution. In

304
 20: The Sea Floor – Exploring a Hidden World

20.7 the deep-water environments, which cover a sub-

stantial area of the Earth’s surface, there is generally
more interest in the larger scale changes in sediments
and features (see Chapters 8 and 9). In either envi-
ronment, the minimum expected size of the feature
of interest usually determines the type of system
used, the survey line spacing, and, consequently, the
type of position-fixing system. Sedimentary features
vary in size from sand ripples of a few centimetres in
wavelength and height, to the massive slumps and
turbidite features described in Chapter 9. Small fea-
tures are often found superimposed on larger fea-
tures, which may be important in understanding the
general evolution of an area. Changes in sediment
type can be detected by changes in the related reflec-
tive properties of the sea floor, as can be seen, for
A example, in Figure 9.13. Deep underwater canyons
are easy to identify, but difficult to survey accurate-
ly. The ocean liner Titanic remained hidden in 5 km
of water for over 40 years before Bob Ballard1,2 used
sophisticated sonar systems and submersibles with
cameras to find and explore the wreck.

Principles of Surveying Equipment

The echo-sounder
The use of instruments which remotely sense the sea
floor is well-illustrated by the echo-sounder, which is
used for measuring water depth (Figure 20.8). A
short pulse of sound at a given frequency is transmit-
ted into the water column by an acoustic transducer,
either on the vessel’s hull or towed at a known depth
beneath the surface. The transducer can both trans-
mit and receive sound pulses. The pulse reflected
B back from the bed is collected; its travel time to
reach and return from the sea bed and its amplitude
C are measured. The amplitude of the returning pulse
relative to that transmitted depends on the depth of
water (since water attenuates sound energy) and the
reflection characteristic (or reflectivity) of the sea
Figure 20.7 Surveying at depth in the ocean involves a
bed. The size, shape, reflectivity, and orientation of
tow cable (A) and a sensing package (B), which is held
near to the sea bed (C). Scaled to a vessel length of the sea-bed features to the arriving sound pulse also
225 m, the water depth here is 3150 m, but sensors can affect the returning pulse amplitude.
be deployed at 5000 m or more. Some systems use the characteristics of reflected
pulses to obtain other information, such as the type
of sediment which makes up the sea bed. The rela-
20.8
tionship between a sound pulse and a sound wave

B Figure 20.8 In echo-sounding, the time a pulse of sound

takes to travel to the sea bed and back is measured; since
C A the speed of sound in water is known to be about
1500 m/s, it can be converted into depth. The amplitude
(strength) of the returning, reflected signal depends on the
depth and absorbence of the water and the type of sea
bed – e.g., rock will give a stronger reflection than sand.
(A) Depth of water, (B) path of outgoing pulse, (C) path of
D reflected pulse, and (D) sea bed.

305
 P. Riddy and D.G. Masson

Box 20.2 What is a sound wave?

Energy in the form of a sound wave can be transmitted through water, since it is a compressible
media. The speed (v, m/s) is proportional to the water density as modified by salinity, temperature,
and depth, as described by equation (20.1), where S is salinity (standard salinity units), T is tempera-
ture (°C), and D is depth (m)7. The equation indirectly describes the density variation through the water
column, density having the most significant effect on the velocity of sound.
–2
v = 1492.9 + 3(T – 10) – 6*10–3(T – 10)2 – 4*10 (T – 18)2 + 1.2(S – 35) – 10–2(T –18)(S – 35) + D/61 (20.1)
Many instruments make the inaccurate assumption of a constant velocity of sound in sea water,
although the error introduced is often not important in the qualitative interpretation of results. When
high accuracy is required, measurements of the temperature and salinity in the survey area can be
combined to estimate the variation of sound velocity with depth, and hence to correct depth and
range measurements.
Although sound waves travel outward from their source as spherical wave fronts, it is convenient
to represent the passage of sound waves through the water column as rays which travel in straight
lines. Sound waves are absorbed, refracted, and diffracted like other waves; for example, they obey
Snell’s Law when being transmitted through layers of sub-bottom rocks.

is illustrated in Figure 20.9, and the propagation of 10 m in, for example, the Bay of Fundy. In order to
sound waves is described in Box 20.2. compare water depths from one area with another
Water depth is measured as a profile beneath the and to provide tidal charts for mariners, depths are
vessel. Since the transducer sends out a sound beam generally referred to a common reference level (e.g.,
that is roughly cone shaped, the depth is measured Ordnance Datum in the UK). Accurate tidal correc-
from returns (frequently the first return) from with- tions, probably using a local tide gauge together
in the cone; this can lead to inaccurate results. with a numerical model of the local tides (see
Multibeam sounders (discussed later) have been Figure 1.6), are required to calibrate measurements
developed to provide a swath of depth measure- of depth made at different parts of a tidal cycle.
ments each side of the vessel, and to improve the
resolution of depth and sea-bed features. Accurate What size features can we see? The resolution of
measurement is important in areas where the bathy- features
metric data are used to produce charts for shipping The resolution of an instrument can be defined as is
or, for example, in planning a construction project. its ability to see two objects as separate (Figure
Changes in sea-floor level or in the shape of the 20.9). The closer the objects are, the harder they
topography may indicate convergence or divergence are to resolve, and the higher the resolution of the
of sediment flux or of varying current speeds. instrument required to see them as separate. To
All major bodies of water experience tidal improve the exact identification of reflected sound
changes in water level, caused mainly by the gravi- waves, marine instruments usually transmit a
tational attraction of the Sun and Moon. This can stream of short pulses of sound, each of a few
vary from a few centimetres in enclosed seas, like cycles grouped together. Each instrument’s trans-
the Mediterranean, or in the deep oceans, to over missions are characterised by a pulse rate and fre-

20.9 (a) Amplitude

Time

Pulse length Figure 20.9 (a) Pulses consist of several wave

(b) (c) cycles grouped together. (b) The wider (actually
longer) pulse reaches both features at the same
time, so they appear as one reflection. (c) Both
features are seen separately by the narrower
(actually shorter) pulse. The sea-bed resolution
of a system depends on the length of the sound
pulses transmitted from the sensor, and is gen-
Wide pulse
Narrow pulse erally taken as half the pulse length.

306
 20: The Sea Floor – Exploring a Hidden World

20.10 A (b) Towfish and targets Display

(a)
tow
fish Sonar
beam

Towfish

Scan 1
(i)
h Single Pair of
l object objects
pairs of equally spaced
objects on the sea bed

r (ii) Scan 2

Scan 1
Dir
tow ection
fish o
trav f Object not seen
el

(c)

tion
irec Sound
md pulse
Bea

Large footprint, Smaller footprint,

close range long range

Figure 20.10 The principles and resolution of side-scan techniques are shown. (a) The system records successive scans of
the sea bed and builds up a picture line by line. In this case, the conversion of the time into distance using the speed of
sound produces a slanted range of features from the transducer; this has to be converted into the true range using the
equation r2 = l2 – h2, where r is the true range, l is the slant range, and h is the height of the towfish. This conversion
assumes the sea bed is flat and the velocity of sound is constant. Any velocity correction would have to account for the
more difficult-to-define spatial velocity variation. The along track distance (A) is determined from vessel navigation. (b)
Here, the side-scan towfish position while emitting consecutive pulses is illustrated. Resolution along-track is complex,
being basically a function of the beam pattern, proximity of objects to the towfish, and the display device. (i) Two objects
may appear as one of slightly higher intensity-of-reflection than a single closer object. The further objects appear as one
more intense object on the display. (ii) The two closer objects are still not resolved with successive pulses, as they appear
as one continuous object on the display device. With a larger delay between pulses (long-range settings), the towfish can
move sufficiently far forward to miss small objects at a close range. (c) For resolution across track, the outgoing pulse
forms a larger footprint at short ranges than at long ranges.

Instrumentation
quency to which its receiving electronics is ‘tuned’,
thereby improving the quality of reception and Deep and shallow water
reducing the effect of other sound sources. The res- Owing to the difference in the scale of investiga-
olution is determined by the frequency, wavelength, tions in shallow and deep water (see above), sys-
beam shape, and pulse length of the transmissions tems tend to be designed for one or the other. A
which characterise the instrument. high-resolution, deep-water system generally uses
Instrument resolution is only part of the story – similar internal electronics to its shallow-water
detail is also lost if we take a picture which repre- counterpart, but is deployed in a different way,
sents 75 m of sea bed and compress it down to a and, to work at depth, must be designed to with-
30 cm print. The overall resolution of the system, stand substantial pressure. Lower frequency sound
of both transducers and recording devices, is used by longer range systems suffers less attenua-
important [e.g., see Figures 20.10(a)–20.10(c)]. tion in the water column, and is more suitable for
deep-water environments. Table 20.2 lists a num-
What size objects are visible? The effect of ber of types of system and gives examples of the
background operating ranges which can be expected.
Have you ever tried to find a particular pebble on a
pebbly beach? If the feature or object has the same Side-scan sonar
reflective properties as its neighbours and is at the The standard echo-sounder (see p. 305) only makes
limit of the system resolution, then it will be impos- measurements directly beneath the vessel, which is
sible to spot. Similarly, sea-bed features must have an inefficient use of ship time. Side-scan sonar is
a different reflective characteristic to their neigh- used to take the equivalent of a mosaic of aerial
bours to stand out. For example, wood saturated photographs of the sea floor (see Chapter 5), but
with water is difficult to distinguish from soft sedi- using sound [Figure 20.10(a)]. The general princi-
ments by using underwater acoustic measurements. ples of resolution are outlined above and in Figures

307
 P. Riddy and D.G. Masson

Table 20.2 Surveying instrumentation. The data given can only serve as a guide, since the parame-
ters depend on the configuration of each system, and the type of sea bed. The compromise between
system resolution and depth of penetration is clear. Chirp systems are a relatively recent develop-
ment in profiling technology, and employ a swept-frequency pulse and sophisticated processing of
the return signal to achieve high resolution, while giving good penetration of the sea bed.

System Frequency System Resolution Penetration

(kHz) (cm) (m)

Echo-sounder 200 5 0–0.5

Echo-sounder 35 20 0–2
Pinger 3.5 20 20–75
Boomer 1–6 40 50–100
Chirp systems 0.5–12.5 10 40
Sparker 0.1–1 200 100–200
Air gun 0.05–0.1 200–5000m 100–tens of km

20.11a Figure 20.11 Two annotated sections of

side-scan traces recorded in shallow water.
The grey scale is calibrated for the strength
of reflections, from black (maximum) to
white (minimum), and is typical of this type
S of system. Both traces were recorded dur-
ing the same survey with consistent system
settings. (a) This is from an area of soft sedi-
ment. The numbers 1–5 are fixed reference
W points marked on the trace at one minute
intervals. Note the distance scales marked
on the trace and the consequent scale dis-
tortion of the features – along track dis-
15 m
tances are proportionally larger than across
180 m
1 2 3 4 5 track distances. The bed has distinct zones
which are rippled (darker areas) and unrip-
pled, respectively, interspersed with other
20.11b dark areas, ambiguous because of the lack
of ripples. The sea bed in this area is pre-
dominantly sand with varying quantities of
R F shell fragments. The sand ripples are of
small amplitude, typically less than 0.2 m,
S
and the apparently narrow boundary
between rippled and unrippled areas
extends over several metres. (S, sea-bed
reflection; W, water-column noise.) (b) The
trace shows a rocky area which has under-
gone considerable folding and faulting, and
is directly adjacent to an area of rippled
15 m sand (R). Note the fault indicated by the
180 m strong reflection (F) running diagonally
across the centre of the trace. The layers of
1 2 3 4 5 rock are being viewed from ‘end on’, dip-
ping at a steep angle down into the page.

308
 20: The Sea Floor – Exploring a Hidden World

Figure 20.12 (a) The GLORIA side-scan 20.12a

vehicle during deployment from its dedi-
cated launch cradle. The vehicle is
7.75 m in length and weighs approxi-
mately 2 tonnes in air, but is neutrally
buoyant in sea water. It is towed at a
speed of between 8–10 knots. (b) GLO-
RIA 6.5 kHz side-scan image showing a
linear chain of submarine volcanoes in
the Pacific Ocean, west of San Francisco
(scale bar is 20 km) – strong sonar targets
are white, acoustic shadows are black
(note this is opposite to those in Figure
20.11).

20.12b

20 km

20.10(b) and 20.10(c). In the interpretation of side- GLORIA (Geological Long Range Inclined Asdic)
scan pictures, a number of other factors influenced GLORIA, or Geological Long Range Inclined Asdic
by the range of operation have to be kept in mind. (Figures 20.12 and 20.2), is a unique side-scan
The images built up, as illustrated in Figure sonar system, developed at the UK Institute of
20.10(a), are named sonagraphs and represent the Oceanographic Sciences, for rapid imaging of large
intensity of reflection plotted in the two space co- areas of sea floor. It operates at a frequency of
ordinates, range and distance, along the vessel’s 6.5 kHz and has a depth-dependent swath width
track. The intensity depends on the reflective with a practical maximum of 45 km. Towable at
nature of the sea floor as well as on the orientation speeds of up to 5 m/s, up to 20,000 km2 of sea
of the features [Figure 20.10(a)]. floor can be covered in a single day (an area the
Features may be compressed or expanded in the size of Wales or Massachusetts). Over a 25-year
along-track direction unless a correction is applied period it has evolved from a large and difficult-to-
to allow for the speed of the vessel (Figure 20.11). deploy single-sided sonar vehicle into a much more
Other distortions of the image can be caused by compact vehicle which produces both side-scan
electronic noise and interference, and the move- imagery and swath bathymetry (see later) on both
ment of the towfish in the water5. sides of the vehicle track. The neutrally buoyant
Sonagraphs provide much information about the vehicle has its own handling system and is config-
nature of the sea bed, (e.g., texture, composition, ured with buoyancy at the top and transducers
and the orientation of features). Their interpreta- below to give good roll stability. It is towed from
tion requires experience and some knowledge of an the nose at a depth of 40–80 m, depending on the
area in order to decide what features are likely to ship’s speed and its tow-cable length (usually about
be real. Natural features, for example, tend to fol- 400 m). The resolution of GLORIA, as with other
low particular patterns or shapes and have a reflec- side-scans, varies with range, being largely depen-
tivity which is often different to human artefacts. dant on the horizontal beam width (about 2˚ for

309
 P. Riddy and D.G. Masson

20.13 Figure 20.13 The more common types of deeper water

(a) multibeam systems (e.g., Simrad EM12, Seabeam, Krupp
m Atlas Hydrosweep) use a cross-shaped array of transduc-
Towfish depth ers to form a series of beams fanning out perpendicular to
m
Example the ship’s track. In these systems, the beam pattern is
beams controlled by transmitting from the along-ship arm of the
e

Sl
cross and receiving on the across-ship arm. Multibeam

an
t ran
systems are generally more accurate by a factor of 5–10

ge
(S
than the interferometer systems, but give lower swath

R)
widths. (a) The travel paths of a number of example
beams – continuous depth measurements are available
over the full width of the swath. Simplistically, depths are
calculated by converting the beam travel time into a slant
(b) (i) (ii) range (SR) and using the equation D = SRcose + T, where
D is depth, T is the towfish depth, and e, the beam angle
to the vertical, is known or measured. (b) Examples of
multibeam transducer array configurations: (i) line array
(discussed in the text); (ii) cylindrical array, which can be
directly hull-mounted, or placed on a retractable unit to
avoid turbulence associated with the hull. (c) In interfer-
View from below ometer measurements, a reflected wave-front strikes the
W
W

W
W

W
W

n
transducer assembly at an angle. By measuring the
change in phase as the wave moves across the transduc-
(c)
m er(s), its angle of incidence can be measured, allowing
m Transducer assembly the beam angle, e, to be calculated. The angle of inci-
Wave front
dence of the wave front is related to the distance of its
reflecting origin from the ship’s track.
Phase
difference
X point along the swath. The shallow-water versions
X of these systems can provide full coverage and an
Reflected overall accuracy of better than 10 cm in depth and
beam 5 cm in range over the swath. The configuration of
a system determines the depth range in which it is
used, with maximum swath widths (i.e., maximum
GLORIA). At mid-range, it can resolve features ranges) of shallow-water systems being typically
about the size of a football pitch in a water depth 5–8 times the depth of water beneath the transduc-
of 5 km, i.e., an area around 120 x 60 m. ers. Most multibeam systems do not provide side-
scan data, although these are available with some
Swath-sounding systems systems.
Echo-sounders provide depth along a profile; side- There are two main types of swath sounders,
scans provide both depth beneath the vessel from a multibeam and interferometer, both of which pro-
straight down–up return and a picture of an area vide line-by-line measurement of depth along the
based on the variation in the intensity of reflec- swath. Transducers can be hull- or towfish-mount-
tions. It would be more efficient if one system ed and are operated in a number of configurations,
could provide depth and intensity measurements at
several ranges (not only below the vessel). This is 20.14
achievable by swath sounding.
A development of the echo-sounder/side-scan
technology, swath sounders measure depth along a
swath which extends each side of the vessel’s track
[Figure 20.13(a)]; depths are calculated for each

Figure 20.14 Three-dimensional visualisation of a sec-

tion of the mid-Atlantic Ridge based on data from a
Hydrosweep multibeam swath system. The image shows
a series of deep basins (dark blues and purples) within
the central mid-ocean ridge rift valley, which crosses the
image from bottom left to top right. The higher rift valley
shoulders are seen toward the edges of the image (yel-
lows and greens).

310
 20: The Sea Floor – Exploring a Hidden World

20.15

B
A

B C
D A
B

Figure 20.15 TOBI 30 kHz side-scan sonar image of a deep-sea channel on Monterey Fan, off the western US – strong
sonar targets are white, acoustic shadows are black. The channel, here at a depth of about 4100 m, is between 1–2 km
in width and up to 50 m deep. Note the terraced walls (B) and ‘waterfalls’ (A) within the channel (D), showing strong
similarities with similar features associated with subaerial rivers (C is a sediment wave field).

of which two examples are illustrated in Figure few hundred metres of the sea floor, such vehicles
20.13(b). A number of overlapping beams are elec- can provide resolution far higher than that of sur-
tronically formed from the same set of transducers face-towed vehicles such as GLORIA. The penalty
to produce the multibeam scan necessary for swath is that this mode of operation can require the use
bathymetry (e.g., one system produces 52 beams of many kilometres of tow cable, resulting in very
from four transducers). The technology is changing slow tow speeds. TOBI (Figure 19.26), has a
rapidly3, and more recent developments have pro- swath width of 6 km, can cover 400–600 km2 per
duced compact, relatively portable systems which day, and can operate at depths of 6000 m, allow-
still function in depths of <1 m. ing its use over all but a few percent of the sea
Essentially a development of side-scan systems, floor. Figure 20.15 shows a TOBI sonagraph.
interferometer systems (e.g., GLORIA, SeaMARC Able to carry a wide range of instrumentation,
II) use a set of transducers to measure the interfer- TOBI currently carries a side-scan sonar operating
ence pattern of reflected sound [Figure 20.13(c)]. at 30 kHz with a beam width of about 0.8˚, giving
This sound ‘interferometer‘ allows the accurate a resolution on the scale of a few metres. Other
measurement of the travel time of sound waves instruments carried include a sub-bottom profiler
reflected from small, adjacent areas of sea floor. (7 kHz), and a transmissiometer to allow the mea-
Travel time is again converted into distance and, surement of temperature and depth.
knowing the beam characteristics and direction, the
depth can be calculated. The intensity of reflection Sub-bottom profiling
can also be used to provide more information Sub-bottom profilers are used to explore the even
about sediment type or sea-floor orientation. more invisible world beneath the sea floor. Still
A three-dimensional relief map, based on swath using reflected sound waves, the frequency of oper-
bathymetry, is shown in Figure 20.14. Swath ation is chosen such that the sound can penetrate
bathymetry provides an ‘ordnance survey’ or topo- the sea floor and be reflected from interfaces
graphic-type map, in contrast to the ‘aerial photo- between different types of rock or sediment [see
graph’ provided by side-scan sonar. Near geologi- Table 20.2 and Figure 20.16(a)]. As in the conven-
cal-quality interpretation can be achieved when the tional echo-sounder, the system records a profile of
results of both are combined using computational data along the vessel’s track; a three-dimensional
tools, such as a Geographical Information System picture of the subsurface structure requires the col-
(GIS). lection of many closely spaced profiles. The speed
at which sound waves travel varies in rocks of dif-
The Towed Ocean Bottom Instrument (TOBI) ferent type and density (see Table 20.3). This wide
TOBI is one of a family of deep-towed instrument range of velocity increases the potential for errors
platforms, which includes, among others, the orig- in the conversion of travel time into depth of a
inal Scripps deep-tow, the American SeaMARC 1, reflecting layer beneath the sea bed (depth = travel
and the French SAR system. By operating within a time times velocity). The variation of velocity with-

311
 P. Riddy and D.G. Masson

20.16a

Distance along profile

0

(i)

(ii)
1

2
(iv) (iii)
3

Time

20.16b 20.16c

20.16d

Figure 20.16 (a) Layers 1–4 represent rock layers of different types and/or density. In sub-bottom profiling a stream of
pulses travels through the water column and penetrates the sea floor. Reflections can occur at the interfaces between
rocks of different density, providing the conditions for reflection defined by Snell’s law are met. The return time of
reflected pulses is measured, and their travel time converted into depth to give a point-by-point profile of the position
of the reflecting boundary; reflections occur at (i) the sea surface, (ii) the sea bed, and (iii) the sub-bed. When the den-
sity of adjacent layers is very close, no reflection occurs, as illustrated by the absent reflector at level (iv), between lay-
ers 2 and 3. (b) Institute of Oceanographic Sciences’ echo-sounder and 3.5 kHz towfish. The tow cable is streamlined
by the attached ‘fairing’. Such towfish typically hold a set of four transducers and can weigh in excess of 100 kg. (c) A
pinger profile from Lac de Cazaux et Sanguinet, Aquitaine, southwest France, taken with a 3.5 kHz ORE system. The
profile shows a filled river channel cut through a horizontally layered sand stratigraphy and buried beneath fine, floc-
culated lake sediments. (Data acquired by Southampton Oceanography Centre.) (d) Shallow seismic image of
Bouldner Cliff, West Solent, UK, taken using a swept frequency (2–8 kHz) Chirp system (System Design, Geochirp©,
Geoacoustics Ltd). The sequence represents inclined reflectors of lithified Oligocene clays, marls, and limestones. The
780 m long section shows a penetration of 25 m and a final resolution of approximately 75 cm for individual beds
(<10 cm resolution is possible with this system). (Data acquired by Southampton Oceanography Centre.)

312
 20: The Sea Floor – Exploring a Hidden World

in the water column is significantly less. These sys-

tems provide a profile of reflections from beneath Table 20.3 Speed of sound in different rock
the sea floor [Figures 20.16(c) and 20.16(d)], from types. Compressional wave velocities are
which the geology of an area can be explored and given. The wide range reflects variations in
its evolution ascertained (see the examples in seismic velocity due to, for example, the
Chapter 9). degree of water saturation in sediments. In
As explained in Table 20.2, in surveys requiring the interpretation of seismic sections, selec-
detailed information from close to the sea floor tion of the appropriate velocity from the
(e.g., investigation of the thickness of gravel wide range possible for a particular rock type
deposits, the pre-site construction survey for a makes the conversion of pulse return times
coastal or offshore installation, and the relationship into depth complicated. In practice, a variety
between sea floor and subterranean topography), of techniques are available to measure a
high-resolution, shallow-penetration profiling sys- depth profile of the velocity variation.
tem are required. By contrast, low resolution and
low frequency systems are used where deep pene- Rock type Range of seismic
tration is needed, as in oil exploration and to char- velocities (km/s)
acterise the Earth’s structure.
Sediments 1.0–3.5
Satellite and airborne systems, Sandstones 2.0–6.0
and underwater photography Limestones 2.0–6.0
Salt 4.5–5.0
Images recorded from a variety of sensors deployed
Granite 5.5–6.0
in satellites (Chapter 5) or aircraft offer informa-
tion based primarily on reflected energy in the EM
spectrum. Such images provide a wide variety of
information related to the structure, composition,
and contents of the near-surface water. Since the ‘photographed’ with sound (see, for example,
penetration depth of EM waves into sea water is Chapters 10 and 13). There are various types of
very limited, little direct information about the sea equipment, ranging from hand-held devices (cam-
floor can be obtained from them. However, in shal- eras, probes) used by divers (Chapter 18) to
low-water environments, sea-bed reflections have remotely driven tethered systems used on sub-
been observed; indeed, in fairly shallow coastal mersibles. Optical systems provide images with
areas the circulation patterns seen in the images high resolution on a scale not possible with
often reflect sea-bed features (e.g., see Box 5.3). acoustic instruments, but with the disadvantage
Underwater photography or video offers an excit- that the operational range is restricted to a few
ing opportunity to provide a visual image of areas metres.

General References References

Belderson, R.H., et al. (1991), Sonagraphs of the Sea Floor, a 1. Ballard, R.D. (1985), How we found the Titanic, National
Picture Atlas, Elsevier, Amsterdam, 185 pp. Geogr., 168(6), 696–717.
Forssell, B. (1991), Radionavigation Systems, Prentice Hall, 2. Ballard, R.D. (1987), Epilogue to the Titanic, National
New York, 392 pp. Geogr., 172(4), 454–465.
Fish, J.P. and Carr, A.H. (1990), Sound Underwater Images, 3. de Moustier, C. (1988), State of the art swath bathymetry
Lower Cape Publishing, Orleans, MA, 189 pp. survey systems, Internat. Hydrogr. Rev., Monaco,
Ingham, A. (ed.) (1975), Sea Surveying, Vols 1 and 2, Wiley, 65(2), 25–54.
London, 306 and 233 pp. 4. Forssell, B. (1991), Radionavigation Systems, Prentice Hall,
Kayton, M. (ed.) (1983), Navigation, Land, Air, Sea and Space, New York, 392 pp.
IEEE Press, New York, 465 pp. 5. Fish, J.P. and Carr, A.H. (1990), Sound Underwater Images,
Urick, R.J. (1983), Principles of Underwater Sound, 3rd edn, Lower Cape Publishing, Orleans, MA, 189 pp.
McGraw-Hill, New York, 423 pp. 6. Ingham, A. (ed.) (1975), Sea Surveying, Vols 1 and 2, Wiley,
London, 306 and 233 pp.
7. Urick, R.J. (1983), Principles of Underwater Sound, 3rd edn,
McGraw-Hill, New York, 423 pp.

313
 CHAPTER 21:

Ocean Resources
C.P. Summerhayes

gas); of minerals, from diamonds to mundane sand

Introduction: The Ocean as a Resource and gravel; and of chemicals, fertilisers, and medi-
Humans have used the ocean as a resource since cines. In this chapter we discuss these resources.
their appearance on Earth. Early humans fed from With the continued growth of the human popula-
its shores, gathering seaweed and shellfish, and tion, it is inevitable that we will have to use more
hunting seals. As time went by they learned to trav- of them, and use them more efficiently, as time
el short distances over it and, eventually, to fish goes by. We also use the sea in other ways. It is a
from boats. Later, people learned to exploit the major tourist resource; we like to swim in it, sail on
ocean for trade, creating their ports from fine nat- it, and dive in it. We use it as a highway for trade,
ural harbours. They found other uses for it, too, and as a space for putting things in, like submarine
such as driving mills with tidal power. Maritime telephone cables, or various kinds of wastes (see
nations used the ocean to transport armies, and as Chapter 22). It can be a source of riches, through
a battleground. salvage and the discovery of sunken treasure. And,
With the dawn of the industrial age, man’s engi- in the future, we may expect to see some of our
neering ingenuity led to the steady seaward march grandchildren living in purpose-built cities in it.
of onshore industries. Mines extended out beneath What is now realised is that to exploit the
the sea, in the UK tapping coal off northeast resources of the ocean we have to learn to manage
England and tin off Cornwall. Large structures, like them better, so that they are kept truly sustainable.
bridges and piers, were built offshore. Dredgers Some things, like wave and tidal power, will always
began mining sand, gravel, and minerals from be sustainable, no matter what we do to our envi-
beneath the waves. Telephone companies laid ronment. Living renewable resources, such as fish,
cables across the ocean bed, linking far-flung conti- however, will not; nor will mineral resources.
nents by instantaneous communication. And in the Governments have made a start – first, by enacting
twentieth century, oil and gas production began legislation to define ownership of the resources;
offshore. second, by enacting legislation to regulate fishing
Nowadays we regard the oceans as resources of practices so as to conserve stocks 22. Under the
food; of renewable energy (from tides, waves, and United Nations’ Law of the Sea1, countries have a
the ocean’s thermal structure); of fossil fuel (oil and right to claim Exclusive Economic Zones extending

21.1

Figure 21.1 The locations and extents

of Exclusive Economic Zones extending
370 km (200 miles) offshore, within
which nations may claim right to all
renewable and non-renewable
resources. (© Institute of Oceano-
graphic Sciences Deacon Laboratory,
Wormley, England.)

314
 21: Ocean Resources

21.2 21.3

Figure 21.2 A school of mackerel, photographed off Figure 21.3 Bringing the catch in. (Photo, Tom Stewart;
Cornwall with a towed underwater camera from the research © Zefa Pictures.)
vessel Clione. (Courtesy of John Ramster, MAFF, UK.)

370 km (200 nautical miles) offshore3,9,16 (Figure 21.4

21.1). Further seaward, the Law of the Sea dictates
how the resources of the global commons are to be
managed and shared. Ownership is the first step
along the road to sustainable development, with all
that this implies for conservation and a managed
environment5.
Growing use of the ocean as a resource will lead
to the growth of ocean forecasting as a marine
industry – users will require more knowledge. The
offshore environment will have to be surveyed in
detail if it is to be used properly24,38, and the ocean-
ic environment will have to be monitored as the
basis for detecting and forecasting change (see
Chapter 3).
Food from the Sea
Fish have always been part of the human diet
(Figures 21.2–21.4). These days fish are caught not
only to eat, but also to turn (the less edible ones, Figure 21.4 Scottish fishing boats in Pittenweem
anyway – about a third of the catch) into fish-meal Harbour, Fife, Scotland, on the north coast of the Firth of
for pigs and poultry, or into fertiliser (Tables 21.1 Forth.

Table 21.1 Disposition of world fish produc-

tion20 – percentage of world total catch in
live weight.

1981 1991 Change

(%) (%) (%)

Fresh 21.7 22.6 +4.2

Freezing 23.8 25.0 +5.0
Curing 12.8 10.9 –14.8
Canning 14.5 12.9 –11.0
Feed 27.2 28.6 +5.1

315
C.P. Summerhayes

Table 21.2 World value of fishery exports20.

1981 1991 Change
(%) (%) (%)

Fresh/chilled fish 10.9 13.5 +23.8

Frozen fish 18.6 18.7 +0.5
Fish fillets (fresh/frozen) 9.8 10.9 +11.2
Cured fish 9.1 5.6 –38.5
Crustaceans (fresh/frozen) 20.6 22.3 +8.3
Molluscs (fresh/frozen) 6.6 8.3 +25.8
Canned fish 13.7 11.0 –19.7
Canned crustaceans/molluscs 4.1 5.5 +34.2
Fish-meal 6.3 3.9 –38.1
Fish oil 0.3 0.3 0

Table 21.3 World fish-catch statistics19 (million tonnes).

1985 1986 1987 1988 1989 1990 1991 Growth (%)

(1985–1991)

World total 86.3 92.8 94.4 99.0 100.2 97.4 97.0 +12.4
Marine waters 75.7 81.0 81.7 85.6 86.4 82.8 81.8 +8.1
Inland waters 10.6 11.8 12.7 13.4 13.8 14.6 15.2 +43.4

Inland waters, main subtotals

Freshwater fish 8.6 9.6 10.4 11.0 11.4 12.1 12.6 +46.5
Diadromous fish* 1.39 1.44 1.59 1.61 1.68 1.69 1.71 +23.0
Marine fish 0.08 0.08 0.08 0.10 0.09 0.11 0.17 +112.5
Crustaceans 0.26 0.29 0.35 0.37 0.41 0.42 0.44 +69.2
Molluscs 0.26 0.27 0.28 0.27 0.27 0.29 0.30 +15.4

Selected species, subtotals

Cod, hake, haddock 12.46 13.57 13.79 13.63 12.90 11.83 10.47 -16.0
Krill 0.19 0.46 0.37 0.37 0.40 0.37 0.23 +21.1
Squid 1.79 1.75 2.32 2.29 2.72 2.33 2.56 +43.0
Mussels 0.96 1.00 1.13 1.29 1.31 1.32 1.33 +38.5
Salmon 1.17 1.09 1.10 1.17 1.45 1.45 1.64 +40.2
Seaweeds 3.88 3.86 3.55 4.13 4.38 4.32 4.91 +26.5

* Migratory between fresh and salt water.

and 21.2). Some of the less edible ones are processed reached a limit beyond which further fishing will
by extracting surimi gel, a fish protein with a texture damage the ocean’s ecosystem, perhaps irrepara-
like that of shellfish, which forms the basis for pro- bly4,17,42. Some people suggest that further growth
duction of synthetic crab and lobster meat, so in the production of seafood should come from
adding considerable value to the original catch26. other organisms, like squid or krill, the catching of
Like the human population, the world fish-catch which could provide another 100 million
initially grew exponentially. Now that the catch tonnes17,42. Table 21.3 shows the growth in squid
approaches 100 million tonnes a year, the rate of and krill catches compared with the decline in tra-
growth has slowed and may be close to a plateau. ditional species (cod, hake, and haddock). The
Table 21.3 shows this plateau in the catch from problem is that nobody knows enough about the
marine waters, Table 21.4 shows the main species ocean and its ecosystems to be able to predict the
caught, and Table 21.5 shows the main fishing level of catch that can be maintained, no matter
nations. Many biologists feel that the catch has what creature is caught.

316
 21: Ocean Resources

Table 21.4 World fish-catch by principal species, 199119.

Top ten species Million tonnes Cumulative percent

Alaskan pollack 4.9 4.9

South American pilchard 4.2 9.0
Anchoveta 4.0 13.1
Chilean jack mackerel 3.9 17.0
Japanese pilchard 3.7 20.7
Skipjack tuna 1.6 22.3
Silver carp 1.4 23.7
Atlantic herring 1.4 25.1
European pilchard 1.4 26.5
Atlantic cod 1.3 27.8

Table 21.5 World fish-catch by principal producers,

199119.

Top ten fishing nations Million tonnes Cumulative percent

China 13.1 13.1

Japan 9.3 22.4
Former USSR 9.3 31.6
Peru 6.9 38.5
Chile 6.0 44.5
USA 5.5 50.0
India 4.0 54.0
Indonesia 3.2 57.2
Thailand 3.1 60.3
South Korea 2.5 62.8

Figure 21.5 Effects of climate 100

control on changes in pelagic Peru 21.5
Proportion of total pelagic fish catch (%)

fish catches for Ecuador, Peru, 90

and Chile, 1965–1989. Note 80
the decline off Peru and corre-
sponding increase off Chile 70
between 1970 and the warm El
60
Niño event of 1982–1983, fol-
lowing which the trends 50
reversed. The El Niño event of
1982–1983 appears to repre- 40
sent the peak of a much broad-
30
er warming period that dimin-
ished greatly the fish catch off 20
Peru. Catches ranged from over Chile
10 Ecuador
14 million tonnes (10% of total
world fish-catch) in the late
1960s to 7 million tonnes in
1955 1960 1965 1970 1975 1980 1985 1990
recent years. Species composi-
Year
tion shifted from 90%
anchoveta at the catch peak
through less than 10% during the mid-1980s to 50% in the mid-1990s. Although this analysis takes no account of
changes in the number of fishing vessels or method of fishing, it makes the point that climatic change can be very
important. (Based on Sharp and McLain37, Figure 3; redrawn at the Institute of Oceanographic Sciences Deacon
Laboratory, Wormley, England.)

317
 C.P. Summerhayes

21.6 Figure 21.6 Changes in biomass

(a) 10 (tonnes x 10–6) of (a) sardine and
Sardine
(b) northern anchovy off
Biomass (tonnes x 10–6)

5 California, showing the climate-

induced sardine–anchovy flip-
2 flop typical of environments
dominated by upwelling cur-
1 rents. Sardine periods are
believed to indicate warmer cli-
0.5 mates. (Based on Cushing 11 ,
0.3 Figure 30; redrawn at the
(b) 10 Institute of Oceanographic
Biomass (tonnes x 10–6)

Sciences Deacon Laboratory,

5 Wormley, England.)
Anchovy
2

0.5

1800 1820 1840 1860 1880 1900 1920 1940 1960

Year

With the expected population growth, there is studies show major changes in the herring popula-
no doubt that if fish are to remain a part of our tion in the North Sea (Figure 21.8), and in the cod
diet we will have to learn to manage the oceans population on the Grand Banks and off Greenland
sustainably. The only sensible way to increase the (Figure 21.9). Historical analyses show that these
catch without damaging the stocks is to farm fish populations grew during the world wars, when
rather than hunt them. To do this on a large scale fishing levels were reduced, and shrank afterward;
would require learning a great deal more than we clearly, overfishing causes the populations to
currently know about fish behaviour and the decline (see Table 21.3, cod, hake, and haddock).
response of fish to changes in the natural environ- However, the local patterns of their distribution
ment (see also Chapter 17). suggest that subtle alterations in the environment
One of the problems of the fishing industry is driven by changes in climate can be equally
that natural changes in the environment can cause
major changes in fish stocks, devastating the local
(a)
communities that depend on the fish catch. An 21.7
Number of scales (x103 cm2/yr)

example is Peru in the 1970s, where the anchovy 15

fishery collapsed due to climate-driven changes in 12

the environment that were related to a severe El
Niño event37 (Figure 21.5). Studies of the history of 9

fish catches off California, Peru, and Namibia 6

show that there have been substantial changes in 3
the balance between anchovies and sardines, with
only one species dominant at any one time; the pat- 0

terns look cyclical on time-scales of about 40 years,

and existed long before fishing began11,37 (Figures (b)
21.6 and 21.7). In the North Atlantic, historical
Number of scales (x103 cm2/yr)

25

20
Figure 21.7 Fluctuations in the (a) sardine and (b)
anchovy populations continue back through time off 15
California, based on studies of the number of fish scales
10
in laminated sediments from the anoxic Santa Barbara
Basin. These clearly indicating a control on population 5
by climate rather than by fishing. (Based on Sharp and
0
McLain 37 , Figure 4; redrawn at the Institute of
300
400

700

800
900
1000

1100
1200
1300
1400
1500
1600
1700
1800
1900
500

600

Oceanographic Sciences Deacon Laboratory, Wormley,

England.) Year

318
 21: Ocean Resources

Figure 21.8 Climatic control of 30 21.8

herring in the northern North Swedish
Atlantic, based on herring catch. Norwegian
During cool periods, coastal ice 25

Duration of coastal ice (weeks)

stays longer on the north coast of
Iceland and the herring move
20
south from Norway to Sweden. In
warm periods, when there is less
ice off the north coast of Iceland, 15
the herring move north from
Sweden to Norway. (Based on
Cushing11, Figure 29; redrawn at 10
the Institute of Oceanographic
Sciences Deacon Laboratory, 5
Wormley, England.)

0
1500 1600 1700 1800 1900 2000
Year

Figure 21.9 The rise and fall of the 21.9

West Greenland cod fishery: (A) the
catch of cod; (B) surface tempera- 1.0
ture anomalies for the Greenland B
West Greenland annual cod catch (x 103 tons)

0.8

Temperature anomaly (˚C)

coast south of Fredrikshab. Cod 500 0.6
appeared on the offshore banks B 0.4
between 1912–1923, initiating a
400 0.2
gradual build-up of stock, which
then crashed in the late 1960s. This 0

overall pattern is attributed to cli- 300 –0.2

matic factors, particularly the local –0.4
warming of 0.2–1.0˚C, which itself 200 –0.6
is a reflection of the regional warm-
ing of the North Atlantic at the A
100
time. (Based on Dickson and
Brander13, Figures 2 and 3; redrawn
at the Institute of Oceanographic
Sciences Deacon Laboratory, 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985

Wormley, England.) Year

Figure 21.10 A fin whale at the 21.10

Whaling Station at the head of
Hvalfjorder (whale fjord) on the
west coast of Iceland in 1978; the
station operated until the current
pause in commercial whaling, tak-
ing about 250 fin whales per year.
(Courtesy of Tony Martin, Sea
Mammal Research Unit, England.)

319
 C.P. Summerhayes

Table 21.6 Whale catches (thousands)19.

1985 1986 1987 1988 1989 1990 1991

Blue and fin whales 7.9 6.5 6.3 0.68 0.61 0.65 0.66
Sperm and pilot whales 83.2 153.8 135.6 132.7 168.4 105.8 50.0

important in determining stock levels from place to Specialist fisheries arose, like the cod fishery of the
place11. Grand Banks, or the herring fishery of the North
Clearly, we need to understand and to be able Sea. Whaling (Figure 21.10) provides a good exam-
numerically to forecast the natural variability in the ple of the evolution of a hunting-based fishery4. It
system in order to exploit the living world more can be done by primitive tribes in coastal waters,
effectively. By itself, however, such an understand- and in Europe it began to develop into an industry
ing will be inadequate; we also need to prevent by the twelfth century in northern Spain. Local
overfishing. The difficulties in policing fishing oper- overfishing drove whalers further afield, to
ations make effective control unlikely. This is Newfoundland and Spitzbergen, where whaling
another argument in favour of adopting the meth- had become big business by the mid-seventeenth
ods of the farmer, rather than those of the hunter, century. The business boomed in the nineteenth
to ensure the true potential of the ocean for feeding century, when technology made whaling easier and
the world’s growing population is achieved. turned it into a global industry. It reached its peak
When technology was primitive, fishing was in the 1930s and 1940s, but the combination of
indiscriminate; early man took what he could get. rapid technological advance, making the job easier,
Advances in technology brought greater control, and the limits of the whale population lead to
and an ability to target particular species systemati- decline by the 1960s. Large-scale commercial whal-
cally, no matter where they might be found. ing is now a thing of the past, although some

21.11a

21.11b
Figure 21.11 (a) Antarctic krill,
Euphausia superba (size approx. 2 cm),
plays a key role in carbon cycling in the
Southern Ocean. (b) The echo-sounding
trace, with horizontal scale lines spaced
25 m apart, shows a swarm of krill as a
large widespread acoustic signal (in blue
and red) centred at about 50 m (the red
horizontal line) over the edge of the
Antarctic continental shelf (shown at the
lower left). Whales feed on these dense
layers of krill and other zooplankton. (©
British Antarctic Survey, England.)

320
 21: Ocean Resources

21.12a 21.12b

Figure 21.12 (a) Development by the Japanese Marine Science and Technology Centre (JAMSTEC) of an artificial sea
floor (20 x 20 m), which can be raised to the surface (right) and lowered below the surface (left) to and from its work-
ing depth of 4 m; it forms a platform for growing abalones, and its tanks contain black rock-fish. (b) On the prototype
in Ryohri Bay, Japan, the abalone are fed with kelp and measured at the surface once a week. (© Mineo Okamoto,
JAMSTEC, Japan.)

nations, notably Norway and Japan, continue to seas more or less clean of fish if they continue
catch large numbers of whales 33 – Table 21.6 unchecked.
shows the decline in catch in recent years. The cre-
ation of a whaling reserve around Antarctica in Mariculture
1994 should help to renew and maintain whale One answer is mariculture, farming the sea (Figure
populations. 21.12). Oysters have been cultivated in Asia for
In the case of herring, the drift net was replaced over 2000 years. The Chinese developed freshwater
by the purse-seine net in the 1960s in Europe, aquaculture over 1000 years ago, and wrote the
which led to overfishing and bans on fishing in the first fish-farming textbook in 475BC; it described
North Sea at various times. Perhaps inevitably, the the farming of carp (fresh water), milkfish (brack-
response of the local industry to a decline in one ish), and mullet (marine)3. Mariculture has been
stock is to go for another. Declines in the herring common for centuries in Southeast Asia, where fish
catch by European fisheries were offset by a farmers seed ponds with the eggs of fish and crus-
tremendous rise in industrial fishing for sand-eels taceans, transferring the larvae and juvenile forms
and the like for fish-meal and fertiliser. The conse- to larger ponds and feeding them on algae and
quent decline in sand-eels is thought to be the cause plankton4. World mariculture production is cur-
of larger than usual numbers of deaths of sea-birds, rently around 10 million tonnes, excluding fresh-
which feed on them32. New fisheries are already water environments – fin fish account for about
developing to harvest new species of fish, like the 45%, shellfish for about 25%, and seaweeds for
Grenadier, from deep water, or different creatures, the rest. The Chinese are the largest producers.
like squid and krill 42 (Figure 21.11 and Table Japan derives about 15% of its total ocean produce
21.3). from mariculture, including such exotic items as
The lessons from the whale, cod, and herring sea-urchin roes and seaweed (which is grown on
fisheries show that, without protection, living nets35); note the growth of seaweed shown in Table
resources cannot survive the technological shift 21.3, which also shows the growth in fish taken in
from primitive to industrial fisheries. Bigger and inland waters (much of it from mariculture). The
better ships, with more sophisticated, expensive, UN Food and Agriculture Organisation (FAO) esti-
and durable equipment, are depleting a supply that mates that by 2000AD aquaculture (including
once seemed boundless. With the help of satellites, freshwater fisheries) will account for 20–25% by
sonars, computers, and refrigeration, combined weight of the world fisheries production, and about
with global-range ships, humans will sweep the 50% of the value3.

321
 C.P. Summerhayes

21.13a 21.13b

Figure 21.13 (a) Long-line mussel farm in Loch Etive, Scotland. The black objects are plastic floats from which hang
the ropes on which mussel spats settle (© Jim McLachlan, McLachlan Shellfish and Fish Farming Equipment). (b) At
harvest time, after around 3 years, there may be 250 kg of mussels per rope (© Nicki Holmyard, Association of Scottish
Shellfish Growers).

Around Europe, oysters and mussels have been ing could work on the Dogger Bank in the central
cultivated in coastal areas for centuries (Figure North Sea, where plaice hatchlings transplanted
21.13; note the growth in mussels shown in Table from the Dutch coast grew more rapidly and suc-
21.3). Since the late 1960s, salmon farms have cessfully42. Unfortunately, this operation has not
grown apace in the fjords of Scotland and Norway been taken up commercially.
(Figure 21.14); the salmon industry in Scotland
now equals the size of the beef and lamb industries Legislation
there3. The market is expanding and the outlook is Fishing can be controlled by legislation, although
good – note the growth in salmon shown in Table not without difficulty. It is a global business (Table
21.3. Interest is growing in farming other species, 21.5), so the ability of the large fishing countries,
especially halibut and turbot, and in developing like Russia, South Korea, and Japan, to fish just
salmon and other farms in the open sea rather than about anywhere means that the stocks of fish off
in lochs3. The advantage of the open sea is that the countries with less fishing capacity could be deplet-
farms could be larger, and are not at such great risk ed without any recompense. To counter this prob-
from local pollution. lem, most fishing countries agreed in 1977 to a
One variety of mariculture that does not use 370 km (200 mile) limit, to keep to themselves the
farms is ocean ranching, in which juvenile fish are right to fish their own waters and to prevent over-
released into the sea to be caught at a later stage. fishing 35. It was a declaration of extended geo-
Salmon ranching is a commercial success in Japan graphic limits around Iceland that kept British fish-
and Alaska28. It has been shown that plaice ranch- ing vessels out of Icelandic waters and led to the so-

21.14

Figure 21.14 Salmon farm in Loch Creran on

the west coast of Scotland. Young salmon,
bred in fresh water, are transferred to the
octagonal, 4.5 m (15 foot) diameter sea cages
at 18 months and fed on high-protein fish
food for about 2 years. Netbags (4–6 m deep)
are suspended from the floating cages; tides
help to flush effluents away. The farmer here
is putting wild wrasse (a ‘cleaner’ fish) into
the cages to keep the salmon free of lice.
Total production from this one farm is around
40 tonnes/year. (© Jim Buchanan,
Association of Scottish Shellfish Growers.)

322
 21: Ocean Resources

Figure 21.15 A desalination plant – removing salt from 21.15

sea water provides drinking water for people in arid
countries. (© Zefa Pictures.)

called cod wars between the two countries in the

mid 1970s3,35. The European deep-water fishery
was severely affected by the extension of the
Canadian fishery exclusive zone to 200 miles over
the Grand Banks. Factory ships from other nations
can circumvent the problem by ‘setting up shop’
seaward of the limit and paying local vessels for
their catches from within the zone35.
In some countries Governments have set quotas
on the size of fish or amount of catch in order to
control fishing and preserve stocks. Unfortunately, realise, unless they live in arid coastal areas, is that
controls work poorly as they are virtually impossi- one of the principal extracts of salt water is fresh
ble to police effectively. In protest against quotas, water. Distilling plants to produce fresh water from
French fishermen burned down the Town Hall of salt water have been common on ocean-going ships
Rennes, in Brittany, in 1993. Tempers will continue for well over a century. Desalination plants have
to run high as the hunting culture clashes with the become increasingly common on land, especially in
implementation of the principles of sustainable arid coastal regions like the Red Sea and the
development. Persian Gulf (Figure 21.15). At present, they make
We are at the beginning of what Borgese4 has commercially unattractive the once-popular idea of
called the Blue Revolution, in which sustainable obtaining fresh water from icebergs towed north
techniques, such as farming, will take over from from Antarctica9.
hunting as the method of harvesting food from the
sea. Farming in the coastal zone will grow in vol- Chemicals
ume and expand in area to become common every- About 3.5% of the weight of salt water is dissolved
where, with more and more species being farmed. solids; sodium chloride accounts for 71% of this
Farms will move offshore to take advantage of (see Chapter 11). It has been a prime ingredient for
space, on the one hand, and local supplies of nutri- cooking and a principal article of trade for well
ents (pumped from cold subsurface waters), on the over 5000 years. Roman soldiers were part paid in
other hand. salt (salarium argentium, from which the word
salary derives). Humans probably first came across
Chemicals and Medicines from the Sea natural salt in dried-up lakes and coastal ponds. It
The sea is a vast storehouse of dissolved minerals. was not a great leap of imagination to create artifi-
Unfortunately, most of the dissolved constituents cial coastal ponds, so bringing the process of evap-
are disseminated in such tiny amounts that extrac- oration under control, and the basic approach has
tion will never be profitable. Only a few are abun- not changed to the present day (Figure 21.16).
dant enough to be extracted on a commercial basis, However, raw sea-salt derived in this way can be
the most common of which is sodium chloride, or impure and rather bitter, containing iron, calcium,
common table salt. What many people do not and magnesium compounds as well as sodium chlo-
ride (to obtain pure salt, several stages of crystalli-
21.16 sation must be used). Evaporation accounts for
around one-third of the world supply of salt, most
of the world’s sea-salt being produced by India,
Mexico, France, Spain, and Italy16.
The only two elements commercially extracted
from the sea on a large scale are magnesium and
bromine16. After oxygen, hydrogen, chlorine, and
sodium, magnesium is the next most common ele-
ment in sea water. In recent years, some 18% of
the world magnesium production of 1.8 million
tonnes has come from sea water, mostly produced

Figure 21.16 Harvesting salt from salt-pans in Portugal.

(© Centro de Coridade Nossa Senhora.)

323
 C.P. Summerhayes

in the US. Magnesium is an exceptionally light crude extracts from the mussel have been sold as
metal and has a variety of uses in structures of vari- food supplements under the name Seatone for 20
ous kinds. In the extraction process, sea water is years in 65 countries. The active component is a
mixed with dolomite rock (a calcium–magnesium glycoprotein which blocks the action of neu-
carbonate), which precipitates the magnesium from trophils, white blood cells that trigger the immune
sea water as a hydroxide. This is then converted system into action at sites of infection or tissue
into magnesium chloride, and the magnesium and injury. Heart stimulants may be obtained from cer-
chlorine are separated electrolytically. tain sea anemones. The antiviral compound idox-
Bromine is the ninth most common element in uridine and the antitumour compound arabinosyl-
sea water (see Chapter 11). It is used in antiknock cytidine have been developed from compounds
compounds in petrol, for instance, and is produced found in a marine sponge from the West Indies. A
as a by-product of both salt making and the pro- variety of red algae (seaweed), Digenia simplex, is
duction of magnesium from sea water. used to expel intestinal worms. A variety of brown
Uranium, the twenty-ninth most common ele- algae yields a choline derivative which can lower
ment in sea water, is 20,000 times less abundant blood pressure7,9.
than bromine in the sea. Even so, many countries Marine toxins, produced commonly as part of an
carry out research into extracting uranium from sea animal’s chemical defence system, show great
water as a means of ensuring stable supplies of ener- promise, not only as pharmacological compounds,
gy in the future. Japan, in particular, has a sizeable but also as models for the development of new syn-
interest, as befits a country without oil reserves. thetic chemicals7,9. For instance, lophotoxin is a
However, the current supplies of uranium on land lethal substance recently discovered in several
make sea-water extraction far from commercial. species of seawhips of the genus Lophogorgia. The
novel chemical structure of lophotoxin and its abili-
Medicines ty to induce neuromuscular blocking of a certain
Medicines have long been obtained from land type indicate that it may represent a new class of
plants, and nowadays several modern medicines are neuromuscular blocking agents with unique phar-
also produced from marine plants and animals9. macological properties. Several highly poisonous
The great diversity of marine species makes it high- fish, in particular the puffer fish, porcupine fish,
ly likely that many could have medicinal properties, and sunfish, contain tetrodotoxin, a very potent
so the study of these organisms for medicinal pur- neurotoxin which has some use as a local anaesthet-
poses is expanding. The discovery of other new ic and muscle relaxant in terminal cancer patients.
compounds is hindered, however, by the limited The blood of the horseshoe crab (Limulus
accessibility of the marine plant and animal world polyphemus) is perhaps the best-known source of a
beneath the sea. marine-derived compound with pharmacological
A wide variety of pharmaceuticals are marine- properties30 (Figure 21.17). In 1955 researchers dis-
derived. For instance, marine biologists in New covered the causative agent for clotting Limulus
Zealand have identified a compound in an extract blood. The medical benefit came in the form of a
from the native green-lipped mussel that appears to diagnostic reagent that detects bacterial infection of
relieve symptoms of arthritis. Capsules containing human blood by clotting in the presence of bacteri-

21.17

Figure 21.17 Horseshoe crab shell on a

Massachusetts beach. (© Diana
Summerhayes.)

324
 21: Ocean Resources

Figure 21.18 Sunlight streaming through a kelp forest. 21.18

Giant kelp grows up to 0.6 m/day and reaches lengths of
65 m. (© P. Glendell, Avon, England.)

al endotoxins. It can detect diseases, such as gonor-

rhoea, but is more widely used to test pharmaceuti-
cal products for the presence of contaminating
endotoxins, thereby improving the quality of drugs
and biological products for intravenous injection.
Biotechnology: Industrial Chemicals
Aside from medicines, marine organisms have given
rise to many other kinds of products. The kelp
forests of Asia and Europe, for instance, have long-
been harvested as a source of fertiliser, potash, and
food for livestock. Since 1929, kelp has also been
harvested, off California, for a unique substance,
algin, which controls the properties of mixtures
containing water 29 (Figure 21.18). The primary
industrial applications of algin are paper-coating
and sizing, textile printing, and welding rod coat-
ings, but it is also used in pharmaceutical and cos-
metic products, including dental impression com-
pounds, binding for tablets, and anti-acid formula-
tions. Algin is used in a wide range of foods, includ-
ing milk-shake mixes, and gels; it is used as an
emulsifier and stabiliser in salad dressings, to
enhance texture and moisture retention in bakery
products, and to stabilise beer foam!
Marine organisms provide a marketable insecti-
cide, produced in Japan since 1966 from an extract tending to pollute the air we breath and change the
called nereistoxin from marine worms (annelids); it temperature of the planet. Knowing that they will
is active against the rice-stem borer and other insect not last forever, more effort is now being put into
pests7. the development of alternative sources of energy
In yet another kind of application, it has been that are clean and renewable.
found that specific compounds of bacterial origin, The ocean has them in abundance, in its winds,
for example those related to melanin, can promote waves, tides, and currents, and in the differences in
settlement and development of oysters and heat between the surface layers and the deeps.
other marine invertebrates, including those of Paradoxically, although a library could be filled
commercial value. with writings on tapping ocean energy, by 1993
In contrast, there is considerable interest in using only one sizeable project had ever been carried
antifouling compounds made by marine organisms through to completion. It seems unlikely that ocean
as a natural means of preventing the build-up of energy will flourish while the price of oil stays low,
barnacles, algae, and other organisms on underwa- except in places difficult to supply with oil, natural
ter surfaces, such as pipelines, structures, and the gas, or electricity, such as capital-poor, labour-
hulls of boats7,12. Success would eliminate the use of intensive sites in less-developed countries.
highly toxic and environmentally unfriendly
antifouling compounds like tributyl tin (TBT). Much Wind
work on the possibility of bringing natural Wind has long-been a source of power at sea, but
antifoulants to the market is underway at the with the advent of the steamer its use has declined.
Marine Biotechnology Institute in Japan, where one Now, however, new ships are being built with alu-
promising compound has been found in a bryozoan. minium sails to supplement the power supplied by
engines. Ocean winds are a potential source of
Energy from the Sea energy for coastal communities and small island
As population grows inexorably, demand for energy nations 36. Sweden is planning to utilise marine
rises. Oil and gas, like coal, are finite, non-renew- windmills to generate electricity, and the US has
able resources. They are also dirty, their combustion installed a wind turbine on the coast of Hawaii40.

325
 C.P. Summerhayes

21.19a Figure 21.19 (a) The only surviving tide mill in the world
still producing wholemeal flour is the Eling Tide Mill at
Totton, near Southampton. The present building was
erected around 1785 on the site of a tide mill known to
exist since at least the mid-thirteenth century, when it
was purchased by the Bishop of Winchester. The mill ran
commercially until 1946, and was restored between
1975 and 1980 (photo, M. Conquer; © IOSDL). (b) The
mill operates as shown. A causeway blocks the River
Bartley to create a tide pool (the mill pond, MP), which is
filled twice a day by Southampton’s unusual double high
tide (the tidal range is about 4 m). At high tide the mill
pond is full (i), and the wheel stopped. A sluice gate
retains water in the mill pond until the ebbing tide drops
the level of water in the estuary to below the axis of the
mill wheel. The sluice gate is then opened and the head
of water in the mill pond drives the mill wheel until the
next rising tide stops the wheel turning and replenishes
the pond. (ii) 2 hours into the ebb tide – water leaving the
sluice gate starts to turn the wheel, which moves slowly
while part of it remains submerged. (iii) Low tide – the
wheel turns at full speed, driven by water leaving the
sluice gate as the level of the mill pond falls. The wheel
turns for about 7.5 hours, and at its maximum efficiency
for about 4 hours (based on an undergraduate project by
Green 21; redrawn at the Institute of Oceanographic
Sciences Deacon Laboratory, Wormley, England.)

21.19b (i) (ii) (iii)

▲
MP MP
MP

Tides with turbines that generate electricity when driven

Wherever tides are strong, humans have found the by both the flood and ebb tides3 (Figure 21.20).
means to harness their energy. For hundreds of The 1 km long Rance barrage, built in 1966, gener-
years tidal mills supplied mechanical power up and ates half a million kilowatts of power on each tide,
down the European seaboard by using either the and doubles as a road. It cost $100 million to
vertical rise and fall of the tides or the flood and build; its operating costs are lower than that of any
ebb of tidal currents4 (Figure 21.19). In the seven- power station in France; its fuel is free; and there
teenth century the Dutch took tidal mills to are no waste products4,6. A tidal range exceeding
America4. Tidal energy is traditionally put to use by 10 m seems to be a prerequisite; such ranges exist
lifting a mass that carries out work while falling in many other parts of the world, including the Bay
back into place, or by turning a paddle wheel. The of Fundy and the Severn Estuary (see Box 21.1),
more modern approach is to dam an estuary, like and barrages have been built in the former Soviet
that of the Rance in northern Brittany, and equip it Union, in Canada, and in China4,6,40.

326
 21: Ocean Resources

21.20

Figure 21.20 The Rance barrage near St Malo in Brittany, northern France, showing water pouring
through the sluice gates from upstream at low tide. The Rance barrage is equipped with 24 turbo-alter-
nator bulb sets; chromium–nickel steel is used for the turbine blades to prevent corrosion. There are six
sluice gates to empty and fill the estuary above the barrage. Operators may use the tide in one direc-
tion (as in a tide mill) or in both directions, and can use pumps to top up the estuary. The total capacity
(600 GWh) has been available since 1982. Contrary to expectation, the estuary above the dam has not
silted up; flooding of the estuary above the dam has greatly expanded pleasure boating; and the cause-
way over the dam provides a road that has contributed to the economic development of the region34.
(Photo, Ronald Toms; © Oxford Scientific Films.)

Box 21.1 The Severn Barrage Development Project

Tidal power barrages have certain environmental attractions compared with conventional power sta-
tions, notably the lack of contribution to acid rain, greenhouse gases, or radioactive waste. The com-
mercial feasibility of harnessing the extreme tidal conditions in the Severn Estuary to generate elec-
tricity has been under consideration for some time in the UK, where a detailed report was presented
to Parliament on 23rd October 1989. Before a barrage could be built, a full environment-impact assess-
ment needs to be carried out over several years. The design for the 16 km long barrage, which would
carry a public highway from near Weston-Super-Mare in England to near Cardiff in Wales, includes
216 turbo-generators of 9 m diameter, each rated at 40 MW, giving a total capacity of 8640 MW. The
barrage would be cut by 166 sluice gates, a major lock for shipping, and a lesser lock for small craft.
The annual output would amount to 17 TW hours of electricity, equivalent to burning 8 million tonnes
of coal and supplying 7% of the electrical demand of England and Wales. Early environmental-impact
studies suggest that the water above the barrage will be less turbid, hence clearer, than it is now,
allowing more light penetration and stimulating more photosynthetic activity. Mud flats exposed at
low water will be less disturbed by currents, and hence more biologically productive. Thus, the estu-
ary is likely to support more fish and birds in total, though there may be some changes in the
population supported.

327
 C.P. Summerhayes

21.21a 21.21b

3 2

Figure 21.21 (a) A schematic cross-sectional cutaway of

the prototype shore-line wave energy plant on the Scottish
island of Islay, demonstrating the operating principle: (1)
waves oscillate the water column; (2) upward water
motion in the chamber forces air through the turbine, dri-
ving the generator to produce electricity; (3) the turbine
converts reciprocal air flow into high speed uni-direction-
al rotation. The plant has a 75 kW capacity, an energy
cost of 7 p/kWh, and an energy output of 300 MWh/yr. It
indicates a future potential of 1000 kW, with an energy
cost of 3–4 p/kWh. This amount of production would
make wave energy plants economic for the islands (cour-
tesy of AEA Technology, Harwell, England). (b) The Islay
Wave Power Station showing the wave chamber (fore-
ground) and turbine housing (rear) (courtesy of Don
Lennard, England and Australia).

21.22a Figure 21.22 (a) Salter’s ducks are hollow floats made of
36 m
reinforced concrete, with a cone-shaped cross-section.
Each duck is about 33 m long and 20 m across. Named
after their inventor, Professor Stephen Salter of the
Water filled
bearing Duck University of Edinburgh, several would be connected in a
motion
in waves line by a hollow shaft attached to land or a fixed platform
at one end. Incoming waves tilt each float, or duck,
14 m diameter
spine Gyro unit
upward, absorbing the waves’ energy and leaving calm
water in its lee. Within the oscillating duck, gyroscopes
(inset) activated by the rocking motion, drive a high-pres-
Principal
Ballast wave sure hydraulic system; the circulating oil drives a turbine
pipes direction
that generates electricity. Power from each duck would
Power canister (steel)
flow to shore through a central cable. Compared with
Four gyros per Duck in two canisters other such plants, the ducks are extremely effective, pro-
Buoyancy tanks viding the highest output per metre of sea. A pilot plant
has been tested and research continues. (© ETSU,
Harwell, England.) (b) A model duck at about 1/150th
21.22b scale in a narrow tank, being tested for response to freak
waves. The duck (centre, in water) is being subjected to a
50-year design wave coming from the wavemaker (off-
stage to the right). The rig above water level, to which the
duck is attached by angled struts, constrains the axis or
‘spine’ of the single duck to move as if it were part of a
‘duck string’ of 50 or more such devices restrained by a
compliant mooring system, while allowing the
duck–spine surge and heave reaction forces to be studied
(© J. Taylor, EUWP, Edinburgh University, Scotland).

328
 21: Ocean Resources

Waves OTEC uses the difference in temperature between

Where tides are not powerful enough to warrant warm surface water and cold deep water to power
the construction of barrages to generate electricity, a turbine and generate electricity (Figure 21.23).
waves may be an alternative source of energy36. The potential for applying this principle is greatest
The power potential of an average wave per kilo- in the tropical regions, between 10˚N and 10˚S,
metre of beach is around 40 MW. It has been esti- where warm surface waters (25–30˚C) overlie cold
mated that a substantial part of Britain’s energy waters (4–7˚C) situated at depths of 500–1000 m.
needs could be met by putting wave energy to The first pilot OTEC plant was built by the
work3. French OTEC pioneer, Georges Claude, in Cuba in
Wave energy can be harnessed by fixed or float- 1930. Pilot plants have since been built or pro-
ing devices. A fixed device is operational on the Isle posed in the Ivory Coast, in Tahiti, in Nauru, in
of Islay, on the western coast of Scotland, where Japan, and elsewhere. The first closed-cycle Mini-
the wave energy offshore is equivalent to between OTEC plant to produce power was built in Hawaii
50–77 kW per metre of shore-line per year 3 , in 1979. Although this successful experimental
decreasing to 20 kW/m at the device. Here, plant generated an electrical output of 50 kW,
Britain’s first shore-line wave power station, deliv- about 80% of this energy was needed to pump up
ering 75 kW, was constructed by a team from the the cold water for the system23, leaving a net output
Queen’s University of Belfast in 1988 and commis- of 10–15 kW. Plants like this would be doing well
sioned in 1991 (Figure 21.21). It works on the to produce an annual return on capital of 1%,
principle of an oscillating water column, with which is unlikely to be commercially attractive to
waves pushing air through a turbine3. The con- investors.
struction of this pilot plant proved the concept and
opened the way to the construction of a commer-
cial demonstrator in the 0.5–1.5 MW range. Power Warm water in
21.23
stations of this type would be ideal for many small
island nations, and for local needs in larger coun- Evaporator
tries (such as China) where the electrical grid sys-
Gaseous ammonia
tems are not yet fully developed. Two large systems
have been installed on the coast of Norway.
Floating devices function on the principle that
wave motion can be converted into reciprocal Generator
motion 6 . Some of the systems currently being Turbine
refined, such as Salter’s nodding duck (Figure
21.22), seem to hold considerable promise as ener- Warm water out
gy sources for the future3,42. Gaseous ammonia
Liquid ammonia

Currents Demister

Although ships’ captains have used ocean currents

for decades to make speedy passage, the rivers in
the sea have only once been harnessed to generate Cold water out Gaseous ammonia
Condenser
electricity, off northwest Iceland. The potential is
there – the Gulf Stream4, for example, carries 30
million m3/s of water past Miami with a velocity of
2.5 m/s. The problem is that all this power, five
times larger than the flow of all the world’s rivers,
is too diffuse to harness easily by conventional
means. Scientists of the Woods Hole Oceano-
graphic Institution have, nevertheless, calculated Cold water in

that an array of turbines stretched across the cur-

rent in the Florida Straits would produce 1000 mil-
lion W/day, as much as two conventional nuclear Figure 21.23 Sketch of the workings of an OTEC plant.
power stations4. Most modern OTEC plants are closed-cycle systems, like
Thermal energy that shown here, in which warm surface water vaporises
an intermediate fuel, such as ammonia; the vapour pow-
Closer to reality is the conversion of the ocean’s ers a turbine and is then condensed back into liquid by
thermal energy into electricity3,6,27. The principle of the cool waters pumped up from the depths. These are
ocean thermal-energy conversion (or OTEC) is sim- called closed-cycle systems because the working fluid is
ple, and has been around for over 100 years. recirculated. (© Living Tapes Ltd.)

329
 C.P. Summerhayes

21.24 Figure 21.24 The Natural

Energy Laboratory of Hawaii
Authority’s plant at Keahole
Point, where there is an experi-
mental 210 kW open-cycle
OTEC plant. In an open-cycle
plant, sea water is evaporated in
near-vacuum conditions to cre-
ate low-pressure steam, which
drives a turbine; once the steam
condenses the resulting water is
discarded (compare with closed-
cycle OTEC, as described in
Figure 21.23). On the site, pipes
can be seen for accessing cold
ocean water (6˚C from 600 m)
and warm surface water
(25–28˚C); the large round tanks
are ponds for growing fish and
algae (kelp). (© Greg Vaughn,
Natural Energy, Laboratory of
Hawaii Authority, US.)

However, because of the considerable promise Energy from the Sea Bed
inherent in OTEC, especially for regions such as First, there was wood for the cooking fires of early
oceanic islands lacking in conventional energy humans; then came coal, to fuel the industrial revo-
sources, research is active 3. Between 1974 and lution; and now we live in the age of energy from
1994 the US Department of Energy invested oil and gas that occur in the ground. Although oil
around $250 million in OTEC research and devel- and gas are more or less ubiquitous in sedimentary
opment. Japan and Taiwan have also spent consid- rocks, in most places they are dispersed in such tiny
erable sums on it, along with some European coun- amounts that they cannot be extracted economical-
tries3,23,27. ly. Here and there, however, conditions were such
Luckily, there is another way of proving the that the oil and gas have been able to migrate out
commerciality of OTEC plants, which involves of the fine-grained mud rocks in which they were
considering the value of the cold deep water itself formed, and into coarser-grained strata. In some
and what can be done with it23. Because water from places, folds, domes, faults, and other structures
depths of 100–2000 m is much richer in nutrients have formed natural traps that prevented oil and
than the surface water (see Chapter 11), it can be gas from escaping from the coarser strata. Traps
used to encourage the growth of plankton to sup- full of oil or gas have become reservoirs, and tar-
port mariculture (e.g., the production of shellfish or gets for the drill (Figure 21.25).
fish-based protein foods), or to grow algae (such as Where the sandy strata reach the surface, or are
kelp) to provide biomass for feed, fertiliser, or cut by faults that offer avenues to the surface,
methanol. much of the oil and gas escapes by natural
Since the water is cold, it can be used for air seepage – such natural seeps on land were exploited
conditioning or be piped through the ground in by ancient peoples. American Indians used oil for
tropical areas to cool the soil so that fruits (e.g., fuel and medicine; the Chinese found natural gas
strawberries) that would not normally grow in the while drilling for salt, and used it as a fuel as far
tropics can be grown year-round3. Through con- back as 1000BC; ancient Egyptians coated mum-
densation or desalination the cold water can be a mies with pitch to help preserve them; and
source of fresh water for drinking or for irrigation; Nebuchadnezzer paved the streets of Babylon with
through electrolysis it can be a source of hydrogen asphalt around 600BC. Natural seeps at the coast
and oxygen. Combining OTEC and aquaculture, are still being exploited off Coal Oil Point in
and deriving a plethora of by-products, is the vision California16.
of the future for commercial success in converting All commercial accumulations of oil and natural
the ocean’s thermal energy into electricity in tropi- gas come from the decay of the buried remains of
cal regions3,23. Work along these lines has begun at marine and terrestrial plants deposited in sedimen-
the Natural Energy Laboratory of Hawaii, where tary basins 15. While some of these basins were
there is an open-cycle experimental OTEC plant depressions on land and accumulated their sedi-
(Figure 21.24). ments in lakes, most were along the drowned mar-

330
 21: Ocean Resources

21.25 Oil fields 21.26

Gas fields

Figure 21.25 North Sea production platform complex in Figure 21.26 Map of distribution of oil and gas fields in
BP’s Bruce gas field. (© British Petroleum.) the northeast Atlantic area, showing national borders.
Britain and Norway have the largest offshore sectors –
note the concentration of production along the boundary
gins of the continents and accumulated their sedi- between the British and Norwegian sectors, above the
ments beneath the sea. Over the aeons of geological deeply buried rift valley that forms the Viking Graben.
time, many marine sedimentary basins have Most fields in this area produce oil; most in the southern
become part of the landmass; it was in these that North Sea produce gas. The scale bar is 100 km; the light
the easily reachable oil, not far below the surface, blue sea is shallower than 200 m. Based on information
was discovered by the early drillers. The oil busi- from Institut Français du Pétrole and redrawn from Le
ness grew and prospered from this ‘easy’ oil, most Monde Diplomatique, Les Mers, Avenir de l’Europe,
of which has now been found. What remains is 1992, p. 47.
much more difficult to find, either because it is
buried deep, or because it is in places that are diffi-
cult to get to and work in, like the Arctic coasts or 1936 that operations began in the Gulf of Mexico,
beneath the ocean. culminating in the discovery of the Creole field in
Given that much oil forms from the decay of 1938. By 1948 the first offshore platform out of
marine planktonic remains in sedimentary basins, it sight of land had been completed off the Louisiana
is hardly surprising that oil reservoirs should exist coast, along with the first offshore pipeline 4 .
in the basins still beneath the sea. Seeps of oil at the Offshore oil had come of age.
coast (for instance, at Coal Oil Point, California, Things have moved apace since then.
and Kimmeridge Bay, England) are indications of Hydrocarbon exploration has been extended to the
the likelihood of oil offshore – indeed, there are waters of every major coastal nation. Oil is pro-
many seeps offshore39. Oil was first produced from duced in substantial amounts off the coast of
beneath the sea in 1894, off California, by drilling California, in the Gulf of Mexico, in the North Sea
from wooden wharves 4. Production was costly (Figure 21.26), in the Black Sea, off Angola, off
compared with that on land, and it was not until Brazil, off southeastern Australia, in the East Indies,

331
C.P. Summerhayes

Figure 21.27 The IXTOC-1 21.27

blow out of 1979–1980 in the
Bahia de Campeche, Gulf of
Mexico; blow outs are one of
the hazards of the offshore oil
industry. The well, drilled in
50 m of water, blew out on June
3rd, 1979, and was capped 290
days later after a loss of between
0.5–1.6 million tonnes of oil.
Drilling had reached 3625 m
when gas under high pressure
forced its way up the drill pipe
and onto the platform; the gas
caught fire, the fire and explo-
sion destroying the platform and
catching the sea alight (actually
burning gas and light hydrocar-
bons as they reached the sea
surface). The platform sank to
the bottom and damaged the
well casing at the sea bed,
beginning the largest marine oil
spill in the history of oil explo-
ration. Wind and current carried the oil northwest, where some of it fouled the beaches of northeastern Mexico and
southeast Texas; most of it evaporated or sank to the bottom of the Gulf. The blow out was stopped by drilling a relief
well into the side of the original well. (© Prof. O Linden, Stockholm University, Sweden.)

in the Persian Gulf, and to a lesser extent in many enough for 43 years at current rates of extraction.
other places besides9,16. Natural gas is also produced We do know that there are no large reserves of
in substantial amounts offshore, the North Sea oil in the deep ocean beyond the continental slopes,
being a major source (Figures 21.25 and 21.26). because sedimentary rocks need to be buried to cer-
As the shallow-water oil has been found, explor- tain depths and cooked by the flow of heat through
ers with advanced drilling technology have moved the Earth to produce oil, and deep-sea sediments
steadily into deeper and deeper water. The deepest are not buried deeply enough for the right condi-
production wells are now located down the conti- tions to have been met15. However, natural gas, or
nental slope in the Gulf of Mexico, where they have methane, is not subject to these same constraints;
progressed from 312 m in 1978 to 914 m in early much of it is created early on in the decomposition
1994. Plans are afoot for production from 1101 m of organic matter and it can readily be trapped in
off the Philippines in 1996. In deep water, produc- deep ocean sediments. The trapping process is
tion is usually not from the conventional drilling and somewhat unusual. The bottom waters of the deep
production platform, but from ‘subsea systems’, in ocean are so cold, and under so much pressure,
which the well-head is at the sea bed and connects that within the pore waters trapped in the underly-
directly to a pipeline9,16. Safety is a primary consider- ing sediments, mixtures of water and methane form
ation to avoid blow outs, like that at the IXTOC-1 gas hydrates, or clathrates, cages of water mole-
well in the Gulf of Mexico (Figure 21.27). cules that surround gas molecules25. With depth in
According to the UK’s Institute of Petroleum, the sediment, the flow of heat from the interior of
annual world-wide production of oil and natural the Earth is sufficient to melt these ice cages, free-
gas liquids reached 3.2 billion tonnes in 1992, rep- ing the gas, which may then be trapped beneath the
resenting 65 million barrels of oil per day (1 tonne hydrate layer. It has been suggested that there are
= about 7.3 barrels). Of this, about two-fifths came 100,000 trillion cubic feet (3 x 1015 m3) of clean
from offshore. Offshore production is gradually natural gas trapped in the deep ocean in this way,
increasing, as onshore oil production peaks. Almost which is as much as has been discovered by drilling
every year, it seems, we are given forecasts that to date in conventional exploration and production
there will be enough oil to last another 40 years; wells. Clathrates are concentrated on continental
then, because of technical advances, more is found, slopes and rises deeper than 300 m, so cover
or it becomes possible to extract more than was around 10% of the total ocean area25. If some way
thought possible from existing reservoirs. Total can be found of extracting this deep-ocean gas, pre-
world reserves stood at 137 billion tonnes in 1992, sent resource estimates will have to be revised.

332
 21: Ocean Resources

21.28 21.29

Figure 21.28 Gravel being pumped aboard an offshore Figure 21.29 Piles of sand and gravel from offshore on
dredger. (© United Marine Dredging Ltd, England.) the quayside awaiting distribution. (© United Marine
Dredging Ltd, England.)

SOUTHAMPTON 21.30

PORTSMOUTH Brighton
50˚50' Shoreham
Lymington
Bournemouth
Ryde
50˚40'
Swanage

Portland
50˚30'
it
ile lim
12 m

50˚20'

2˚20' 2˚00' 1˚40' 1˚20' 1˚00' 0˚40' 0˚20'

Figure 21.30 Dredging areas in the English Channel. (Courtesy of the Institute of Oceanographic Sciences Deacon
Laboratory, Wormley, England.)

Minerals from the Sea Bed Use of the beach for pleasure has moved the
search for aggregates offshore, to tap deposits that
Aggregates formed during the rise and fall of sea-level that
Beaches are beautiful, but like many other things in accompanied the ice ages. At least half of the sur-
life they may interest people for reasons other than face of the world’s continental shelves is covered
the aesthetic. They are made of sand or gravel, the with sands and gravels from fossil beaches, dunes,
raw materials of the building trade, and so they and river channels (Figure 21.28). This is a great
have been used for generations in some parts of the boon to those countries that are highly industri-
world to feed the construction industry. This par- alised and densely populated, and where land is
ticular use of beaches is on the decline, because scarce. Dredging for these bulk materials is a grow-
with the galloping growth of tourism, even in the ing industry off their coasts (Figure 21.29). About
remotest spots the amenity value of most beaches is 25% of the sand and gravel for construction in
considered to be worth more than their value as southeastern England comes from offshore (Figure
building materials. 21.30), but about 50% of the world’s offshore

333
C.P. Summerhayes

aggregate production is from Japan10,16. Where the 21.31

deposits are close to shore, care must be taken to
ensure that the natural offshore supplies of sand for
beaches are not destroyed by dredging; this is why
around the UK, for instance, dredging is regulated,
being confined to specified concession areas2, and
why little offshore dredging is allowed around the
US, even though there are substantial resources of
sand and gravel off the US coasts10. Most dredging
is from depths of less than 45 m, which may be
extended to 50–60 m over the next few years.

Mineral ores
Most sand and gravel is made of mixtures of quartz
and feldspar, the main rock-forming minerals of
the continents. But the precise mineralogy of a
beach depends on the mineral make-up of the local
rocks, the local climate (which dictates how the Figure 21.31 Phosphatic limestone with a P2O5 content
of around 15–18% dredged from the continental shelf off
rocks weather), how the eroded minerals were
Morocco (maximum dimension 12 cm). This example is
transported to the sea, and the processes they were a conglomerate of pebbles of phosphatised limestone and
subjected to by sea waves and currents. phosphorite set in a phosphorite matrix, in which are
On beaches and offshore, the continual action of embedded dark green to black, sand-sized grains of the
waves and tides helps to concentrate hard and iron–potassium aluminosilicate mineral glauconite. The
heavy minerals in deposits known as placers10,16. rock is believed to have formed in Miocene times. Being
Placers generally form toward the base of offshore solid, rather than pelletal, and having a moderate P2O5
sand deposits, at the interface with the underlying content, it is a rather low-grade resource of fertiliser
rock, and in depressions and channels. A wide vari- phosphate, so is uncompetitive with the extensive high-
ety of types of dredgers can be used to mine these grade pelletal phosphorite deposits mined onshore in
Morocco. Miocene glauconitic conglomeratic phosphatic
unconsolidated mineral ores, which can be of many
limestones like this are also common off South Africa and
different types depending on the local geology. California. All three sites are today washed by nutrient-
Beaches around the world have been mined for rich upwelling currents; similar, but richer, currents may
many minerals, including diamonds (Namibia), gold have prevailed at these sites during the Miocene in
(Alaska and Nova Scotia), and chromite (Oregon). response to a global reorganisation of oceanic nutrients.
Offshore, diamond placers are mined off Namibia, (Photo, M. Conquer; © IOSDL.)
cassiterite (for tin) off Malaysia, Indonesia, and
Thailand, and (in the past) off Cornwall,
England8,10,16. Other minerals, such as chromite (for hotels to house tourists and boost local income.
chromium), rutile (for titanium), ilmenite (for iron Paradoxically, these short-term gains increase the
and titanium), magnetite (for iron), zircon (for zir- risk of damage from severe storms, by wrecking the
conium), monazite (for rare earths), and scheelite islands’ natural coastal protection (see also
(for tungsten), have been or are currently being Chapter 17).
dredged in various places around the world, like Sri Some minerals have been mined from beneath
Lanka and Australia10,16. Most dredging is from the sea bed, through offshore extensions of mines
depths of less than 50 m, but the Japanese have that began on land, including coal off the east coast
dredged cassiterite from up to 4000 m deep. of the UK, scheelite (for tungsten) off Tasmania,
Beaches have even been mined for their most basic iron ore off Newfoundland, and tin off Cornwall.
mineral, quartz; pure quartz is the basic constituent Other minerals are deposited on or just beneath
of glass sand, mined from the beach in western the sea bed, and could be mined from the sea bed if
North Island, New Zealand, for instance8. economic conditions were right. Among these is
On tropical islands fringed with coral reefs, the calcium carbonate–fluorapatite, a phosphate miner-
white sands consist not of quartz but of coral frag- al that forms the main component of the phosphat-
ments, made of calcium carbonate, the basic con- ic rock, phosphorite (Figure 21.31). Sand-grain size
stituent of cement. Calcium carbonate in the form pellets and pebble- to boulder-size nodules of phos-
of shell remains is also common in places on the phorite abound in certain areas, usually off the
continental shelves, and has been dredged offshore west coasts of continents in mid-latitudes where
(e.g., off Iceland) for the production of cement10. surface waters are exceptionally well supplied with
Extraction of carbonate by the demolition of coral phosphate10 (e.g., Peru, California, Morocco, South
reefs on some tropical islands helps build holiday Africa, and Namibia). Usually they occur where the

334
 21: Ocean Resources

21.32a 21.32b

Figure 21.32 Manganese nodules carpet the sea bed at a depth of around 4100 m in the Peru Basin, as seen from a
photo-sledge during RV Sonne cruise S079; the diameter of the outer ring of the compass is 25 cm. (a) From 8˚35.0’S,
90˚42.5’W, showing the sea bed densely covered with nodules; to one side of the compass a benthic organism has
pushed the nodules apart in its search for food from the underlying sediment. (b) From 8˚43.0’S, 90˚43.0’W, showing
less dense coverage by larger nodules, with the trails of benthic organisms obvious between the nodules to the side and
a star fish near the centre. Most nodules look like small black potatoes or tennis balls; these ones look more like round
cauliflower heads. (Courtesy of Dr von Stackelberg, Bundesanstalt für Geowissenschaften und Rohstoffe, Germany.)

water is 100–1000 m deep, so they are potentially seamounts may be encrusted with manganese oxides
accessible with the right mining gear. Similar that are particularly rich in cobalt10 (up to about
deposits on land are mined for most of the world’s 2%, Figure 21.33). Constraints on utilising these
phosphate fertiliser. Where the economics are right, various deposits include the high costs of mining in
offshore deposits could provide the basis for a local deep water, the variability of ore grades, the distance
phosphate fertiliser industry. But land reserves are to market, and the competing costs of mining on
abundant and cheap to mine, so it seems unlikely land. Deep-sea metal-mining will stay unattractive as
that offshore phosphorites will become commer-
cially viable before 2000AD. A likely first con-
tender is actually an east-coast deposit, on the 21.33
Chatham Rise off New Zealand. There are rich
deposits also off the east coast of the southeastern
US, from Florida to the Carolinas.
Elsewhere on the sea bed, iron–manganese oxide
minerals are deposited in the form of nodules or
crusts10. In certain places on the deep-ocean floor
(4000–5000 m deep), where the rate of sedimenta-
tion is extremely slow, the sea bed is carpeted with
small potentially mineable manganese nodules the
size and shape of potatoes or tennis balls (Figure
21.32). Nobody is very interested in their manganese
content; the attractive feature is their high content of
combined copper, nickel, and cobalt (averaging
2.4% in places 18). In shallower waters (around
1000–2000 m), the flanks of volcanic islands and

Figure 21.33 Manganese encrustation on pillow basalt

from Nod Hill, an abyssal hill rising from the western part
of the Madeira Abyssal Plain close to the eastern flank of
the Mid-Atlantic Ridge. (1) Basalt core; (2) oxidised glassy
surface of pillow basalt; (3) 1.6 cm thick manganese
encrustation. Crust has been removed to show the
smooth oxidised surfaces in the glassy section.
Encrustations41 can reach 25 cm. (Photo, M. Conquer; ©
IOSDL.)

335
 C.P. Summerhayes

21.34 Figure 21.34 Multicoloured layers representing different

(a) (b) minerals deposited from metal-rich Red Sea brines. (a) A
grey–green section from 125–145 cm in CHAIN core
119K in DISCOVERY DEEP. It comprises detrital sedi-
ments (foraminiferal tests made of high magnesium cal-
cite) in the middle (pale, greenish grey, and homoge-
neous), and a finely laminated mixture of these sediments
with iron sulphides at the top and bottom. (b) A red,
brown, and yellow section from 125–165 cm in CHAIN
core 161K from the saddle between the CHAIN DEEP and
the ATLANTIS II DEEP. It comprises a mixture of detrital
sediments (foraminiferal tests of high magnesium calcite)
and an orange–yellow mix of goetthite and amorphous
‘limonite’ (both are iron oxides). Precipitation of the goet-
thite–amorphous material requires oxidation of dissolved
ferrous iron14. (© D A Ross, Woods Hole Oceanographic
Institution, US.)

long as metal prices are depressed, which is likely to phide deposits rich in copper and zinc are common
persist until early in the twenty-first century10,16,18. along the crests of the world-encircling mid-ocean
The constraints have not stopped international con- ridge system, where the several plates of the Earth’s
sortia of companies from Japan, the UK, France, rigid outer shell are pulling apart from one another
Germany, and the US, and the Indian Government, (see Chapter 10). The Red Sea is one end-member of
from exploring for manganese nodules and/or devel- this ridge system, where not much sea floor has yet
oping the means to harvest them. The prime sea-bed been created, where the sediment supply is high, and
sites are in the eastern Pacific between Hawaii and where deeply buried evaporite deposits make the
California at about 12˚N10,16. hydrothermal fluids exceptionally salty, forming
We have known about manganese and phospho- brines. Hydrothermal systems and associated miner-
rite nodules since their discovery during the late al deposits also occur where sea-floor spreading
nineteenth century. Many were recovered during takes place in the marginal basins of the Pacific10.
the first global oceanographic expedition, by HMS A typical mid-ocean ridge hydrothermal field
Challenger in 1872–1876 (see Chapter 1). Much consists of several hydrothermal vents, each exiting
more recently two new kinds of mineral deposit through a tube of sulphide minerals (see Chapter
have been discovered on the deep-sea floor: metal- 10). Individual tubes may be anything up to 200 m
rich muds and hydrothermal mounds. high, or perhaps more. Drilling shows that massive
In the early 1960s hot brines were discovered in sulphide deposits permeate the rock below the sea
deeps in the central graben of the Red Sea 14 . bed in these locations. Because the fields are small
Investigations in 1965 showed that beneath these and hard to find, and because most of them consist
were muds rich in metal (iron, manganese, zinc, of hard rock and are far from land, mining them
copper, cadmium, lead, and silver; Figure 21.34). before 2005 seems out of the question. However,
The brines are hot hydrothermal fluids that have studying them gives us valuable clues as to how to
leached salt and metals from the underlying rocks prospect more accurately for similar deposits now
(which include evaporites). The metals precipitate raised up on land.
as minerals of different kinds14 (oxides, sulphides,
sulphates, or carbonates). Hydrothermal activity is
concentrated along the central graben because it is
here that Africa and Arabia are pulling apart from
General References
Attway, D.H. and Zaborsky, O.R. (eds) (1993), Marine
one another. German researchers, funded by Saudi Biotechnology, Vol. 1, Pharmaceutical and Bioactive
Arabia and the Sudan, have established that the Natural Products, Plenum, New York, 500 pp.
muds are rich enough in metals to be of economic Barnaby, F. and Barnaby, W. (1990), Oceans of Wealth, Living
potential, but the general depression in metal prices Tapes Ltd, London, 94 pp.
Borgese, E.M. (1975), The Drama of the Oceans, Abrams, New
makes them non-viable at present. Of all the deep- York, 258 pp.
sea metal deposits, these are the ones most likely to Couper, A. (ed.) (1983), The Times Atlas of the Oceans, Times
be mined first10,16. Books, London, 272 pp.
The second discovery, in 1975, was that hydro- Cronan, D.S. (1992), Marine Minerals in Exclusive Economic
thermal fluids discharging in the axial valley of the Zones, Chapman and Hall, London, 209 pp.
Earney, F.C.F. (1990), Marine Mineral Resources, Routledge,
East Pacific Rise were also associated with metal- New York, 387 pp.
rich sulphide deposits10,16,31. We now know that Elder, D. and Pernetta, J. (1991), Oceans, Mitchell Beazley
hydrothermal activity and associated massive sul- Pub., London, 200 pp.

336
 21: Ocean Resources

Faulkner, D.J. (1992), Biomedical uses for natural marine chem- 21. Green, M. (1990), Physics of a Tide Mill, Oceanography
icals, Oceanus, 35(1), 29–35. Department, University of Southampton Student
Penny, M. (1990), Exploiting the Sea, Wayland Pub., Hove, 32 Report (unpublished), 12 pp.
pp. 22. Hinds, L. (1992), World marine fisheries: management and
Porter, G. and Brown, J.W. (1991), Global Environmental development problems, Marine Policy, 16(5),
Politics, Westview Press, Oxford, 208 pp. 394–403.
Rogers, D. (1991), Food from the Sea, Wayland Pub., Hove, 32 23. Johnson, F.A. (1990), Energy from the oceans: a small land
pp. based ocean thermal energy plant, in Ocean
Simpson, S. (1990), The Times Guide to the Environment, Resources, Vol. 1, Assessment and Utilisation, Ardus,
Times Books, London, 224 pp. D.A. and Champ, M.A. (eds), Kluwer Acad. Pub.,
Whipple, A.B.C. (1983), Restless Oceans, Time-Life Books, Dordrecht, pp 201–211.
Amsterdam, 176 pp. 24. Kunzendorf, H. (1986), Marine Mineral Exploration,
Elsevier Oceanography Series No 41, Elsevier,
References Amsterdam, 300 pp.
1. Anderson, D.H. (1993), Efforts to ensure universal participa- 25. Kvenvolden, K.A., Ginsburg, G.D., and Soloviev, V.A.
tion in the United Nations Convention on the Law of (1993), Worldwide distribution of subaquatic gas
the Sea, Int. Comp. Law Quarterly, 42, 654–664. hydrates, Geo-Marine Lett., 13, 32–40.
2. Ardus, D.A. and Harrison, D.J. (1990), The assessment of 26. Lee, C.M. (1984), Surimi gel and the US seafood industry,
aggregate resources from the UK continental shelf, in Oceanus, 27(1), 35–39.
Ocean Resources, Vol. 1, Assessment and Utilisation, 27. Lennard, D.E. (1990), OTEC developments out of Europe,
Ardus, D.A. and Champ, M.A. (eds), Kluwer Acad. in Ocean Resources, Vol. 1, Assessment and
Pub., Dordrecht, pp 113–128. Utilisation, Ardus, D.A. and Champ, M.A. (eds),
3. Barnaby, F. and Barnaby, W. (1990), Oceans of Wealth, Kluwer Acad. Pub., Dordrecht, pp 191–200.
Living Tapes Ltd, 94 pp. 28. McNeil, W.J. (1984), Salmon ranching: a growing industry
4. Borgese, E.M. (1975), The Drama of the Oceans, Abrams, in the North Pacific, Oceanus, 27(1), 27–31.
New York, 258 pp. 29. McPeak, R.H. and Glantz, D.A. (1984), Harvesting
5. Borgese, E.M. (1992), Ocean mining and the future of world California’s kelp forests, Oceanus, 27(1), 19–26.
order, in Use and Misuse of the Sea Floor, Hsü, K.J. 30. Novitsky, R.H. (1984), Discovery to commercialisation: the
and Thiede, J. (eds), Wiley and Sons, Chichester, pp blood of the Horseshoe Crab, Oceanus, 27(1), 13–18.
117–126. 31. Oceanus (1984), Deep sea hot springs and cold seeps,
6. Charlier, R.H. and Justus, J.R. (1993), Ocean Energies, Oceanus, Vol. 27, No.3, Woods Hole Oceanographic
Elsevier Oceanographic Series, No. 56, Elsevier, Institution, Massachusetts, 100 pp.
Amsterdam, 534 pp. 32. Penny, M. (1990), Exploiting the Sea, Wayland Pub., Hove,
7. Colwell, R.R. (1984), The industrial potential of marine 32 pp.
biotechnology, Oceanus, 27(1), 3–12. 33. Porter, G. and Brown, J.W. (1991), Global Environmental
8. Cook, P.J., Fannin, N.G.T., and Hull, J.H. (1992), The Politics, Westview Press, Oxford, 208 pp.
physical exploitation of shallow seas, in Use and 34. Rodier, M. (1992), The Rance tidal power station: a quarter
Misuse of the Sea Floor, Hsü, K.J. and Thiede, J. of a century, in Tidal Power: Trends and
(eds), J. Wiley and Sons, Chichester, pp 157–180. Developments, Clare, R., Price, R., Madge, B., Binnie,
9. Couper, A. (ed.) (1983), The Times Atlas of the Oceans, C.J.A., and Wilson E.A. (eds), Proceedings of the
Times Books, London, 272 pp. Fourth Conference on Tidal Power, London, Thomas
10. Cronan, D.S. (1992), Marine Minerals in Exclusive Economic Telford Pub., London, pp 301–309.
Zones, Chapman and Hall, London, 209 pp. 35. Rogers, D. (1991), Food from the Sea, Wayland Pub., Hove,
11. Cushing, D.H. (1982), Climate and Fisheries, Acad. Press, 32 pp.
New York, 373 pp. 36. Senior, A.G. (1990), Renewable energy from the sea, in
12. Dayton, L. (1994), Sea’s green clean for boats, New Ocean Resources, Vol. 1, Assessment and Utilisation,
Scientist, May, 21. Ardus, D.A. and Champ, M.A. (eds), Kluwer Acad.
13. Dickson, R.R. and Brander, K.M. (1993), Effects of a Pub., Dordrecht, pp 171–181.
changing windfield on cod stocks of the North 37. Sharp, G.D. and McLain, D.R. (1993), Fisheries, El Niño-
Atlantic, Fish Oceanogr., 2(3/4), 124–153. Southern Oscillation and upper-ocean temperature
14. Degens, E.T. and Ross, D.A. (1969), Hot Brines and Recent records: an eastern Pacific example, Oceanogr., 6(1),
Heavy Metal Deposits in the Red Sea, Springer- 13–22.
Verlag, New York, 600 pp. 38. Smith, A.J. (1990), The potential resources of the sea areas
15. Demaison, G. and Murris, R.J. (1984), Petroleum around the remaining dependencies of the United
Geochemistry and Basin Evaluation, Am. Assoc. Kingdom, in Ocean Resources, Vol. 1, Assessment
Petrol. Geol. Memoir, No 35, AAPG, Tulsa, 426 pp. and Utilisation, Ardus, D.A. and Champ, M.A. (eds),
16. Earney, F.C.F. (1990), Marine Mineral Resources, Kluwer Acad. Pub., Dordrecht, pp 21–48.
Routledge, New York, 387 pp. 39. Spiess, R.B. (1983), Natural submarine petroleum seeps,
17. Elder, D. and Pernetta, J. (1991), Oceans, Mitchell Beazley Oceanus, 26(3), 24–29.
Pub., London, 200 pp. 40. Vadus, J.R., Bregman, R., and Takahashi, P.K. (1992), The
18. Exon, N.F., et al. (1992), What is the resource potential of potential of ocean energy conversion systems and
the deep ocean?, in Use and Misuse of the Sea Floor, their impact on the environment, in Use and Misuse
Hsü, K.J. and Thiede, J. (eds), Wiley and Sons, of the Sea Floor, Hsü, K.J. and Thiede, J. (eds), Wiley
Chichester, pp 7–28. and Sons, Chichester, pp 373–402.
19. FAO (1993a) FAO Yearbook, Vol. 72 (1991), FAO Fisheries 41. von Stackelberg, U., Kunzendorf, H, Marchig, V., and
Series No. 40, Catches and Landings, United Nations Gwozdz, R. (1984), Growth history of a large ferro-
Food and Agriculture Organisation, Rome, 654 pp. manganese crust from the equatorial North Pacific
20. FAO (1993b) FAO Yearbook, Vol. 73, (1991), FAO nodule belt, Geologisches Jahrbuch, A75, 213–215.
Fisheries Series No. 41, Commodities, United Nations 42. Whipple, A.B.C. (1983), Restless Oceans, Time-Life Books,
Food and Agriculture Organisation, Rome, 395 pp. Amsterdam, 176 pp.

337
 CHAPTER 22:

Waste Disposal
in the Deep Ocean
M.V. Angel
Introduction land may be kept within bounds by better agricul-
Satellite pictures of the Earth clearly show that far tural practices. The ‘Green Revolution’ of the
more of its surface is covered with sea than with 1980s saw considerable increases in the amounts of
earth. The oceans cover just over seven-tenths of its rice grown in developing countries through the use
surface, but what the satellite images fail to show is of improved strains of rice. So, being optimistic, if
that half is covered by ocean deeper than 3000 m. one assumes that the agronomists and biological
At present, apart from the transit of shipping and a engineers can repeat their past successes, then the
small amount of oceanic fishing, this half of the next major foreseeable problem is what to do with
world’s surface remains largely unused by humans, the vast increases in waste that this growing popu-
albeit modified to some extent by anthropogenic lation will generate. Every year every person in the
contaminants (as, for example, clinker, see Chapter developed world generates about 2 tonnes of waste.
13). Contrast this with the startling data on how lit- In developing countries, the average per capita pro-
tle land is now left for natural systems. Allowing for duction is only about a quarter of a tonne, but
the land area which is covered with ice, or is desert, these nations have ambitions to develop their
or is used either agriculturally or for managed industries, material wealth, and life-styles, and in
forestry, less than one hectare of land per capita of so doing their waste production will increase. How
the population (less than 30% of the whole land are we to manage our waste streams to prevent
surface) remains for natural ecosystems (Figure them causing ecological havoc?
22.1). This is at the world’s present population size
of just over 5 billion. The most optimistic – that is The Nature of Waste
to say, the lowest – forecast of world population is What are the sorts of waste with which mankind is
for it to double to 10 billion by 2100; more pes- littering and contaminating the environment? The
simistic forecasts are for this doubling to occur by detritus of twentieth-century life-styles and the
2050! Even using the more optimistic forecast, there debris of industries include both natural and artifi-
will be growing pressure to utilise every usable cial products. The natural substances, such as
square metre of land, including most National Parks sewage sludge, can be dealt with by environmental
and Nature Reserves, within a few decades. systems so long as these systems are not over-
It is possible that the demand for agricultural loaded. There are bulky, more-or-less inert wastes,
such as mining tailings, fly ash from power sta-
tions, and dredged spoils from harbours and estu-
22.1 Deserts aries. There are the gases we discharge: carbon
dioxide, sulphur dioxide, and oxides of nitrogen.
Croplands
There is a great variety of chemical wastes, some of
23.0
which are highly dangerous not only to ourselves,
10.0 but also to other life, and which threaten the eco-
Ice
logical processes that keep the Earth habitable.
11.0
Among the most dangerous of these are some of
the man-made organic compounds, a few of which
have been purposely designed to kill. Being selected
20.1
for their high toxicity and persistence in the envi-
Grazing
ronment, they are tailor-made to create environ-
29.4 mental problems. Being synthetic, they are resistant
to degradation by micro-organisms which, never
6.5 Nature

Managed Forest Figure 22.1 Percentage distribution of usage of land

space globally (based on McAllister3).

338
 22: Waste Disposal in the Deep Ocean

having encountered them before, have not evolved bers of sites suitable for land-fill are rapidly dwin-
the biochemical systems to cope with them. Many dling. Not only is there a serious risk of ground
of these substances, despite their hazardous proper- waters becoming contaminated, but also sites con-
ties, are considered to be ‘essential’ not only for tinue to generate gases, like methane, long after dis-
industry to compete more effectively, but also to posal has ceased. Even if such difficulties can be
control weeds in lawns and keep rats from infesting overcome, is it really sensible to continue to squan-
the sewers. der one of the most basic of resources – land
Another major source of waste is the social space – in this way? The more land area that
demand in developed countries that all goods be becomes restricted in its use, the greater will
highly packaged, partly to maintain their quality, become the pressure to exploit the dwindling
but more importantly to make them increasingly wilderness. When will such losses of wilderness
irresistible to the purchaser. As a result, massive begin to compromise the ability of natural systems
quantities of paper, plastic, aluminium foil, glass, to keep the planet habitable?
and cans are discarded daily. We have become too There is no single universal solution to these
busy or too lazy to clean our own vegetables and problems. Even if human life-styles can be changed
yet over-concerned about the slightest contamina- so that less waste is generated, there needs to be
tion of food. Economic considerations are used as more recycling and reductions in the manufacture
an excuse to neglect the recycling of garbage and of the more damaging synthetic chemicals like
the adequate processing of sewage. In Britain alone, PCBs and chlorofluorocarbons (CFCs).
about 100 million tonnes of domestic waste are As mentioned, some of the more toxic materials
generated annually, to which must be added indus- can be destroyed by incineration. But some residues
trial wastes – mining tailings, dredge spoils, and fly will always remain a problem – the toxic residues
ash – plus the 200 million tonnes or so of carbon from sewage treatment, the fly ash from incinera-
dioxide that are released into the atmosphere. tors, the dredge spoils that are too contaminated to
Estimates of the waste stream in the US for 1988 use for land reclamation, and our legacies of conta-
amounted to 180 million tonnes of municipal solid minated land from the industrial revolution and
waste, 300 million tonnes of sewage sludge (wet), stockpiles of nuclear wastes. Despite some innova-
400 million tonnes of dredged material, and 400 tive ideas, such as the creation of artificial reefs
million tonnes of industrial waste (wet and solid), using blocks of fly ash (see Chapter 20), bulky
but did not include waste gases. How are humans wastes of mixed origin present some of the most
to ensure that the environment is kept healthy? awkward problems. Where can these be put with
the minimum environmental damage? Is the deep
Options for Waste Management ocean a suitable place?
The most sensible approach is to reduce the
amounts of waste produced. However, while this is Marine Disposal – The Background
feasible for some materials, it is not for others. For At present, the trend in international law is to
example, the amount of sewage produced is direct- move toward the banning of all waste disposal into
ly proportional to the size of the population, and the ocean. This attitude originated from the real
the more sewage that is treated, the more sludge need to clean up coastal seas which, since the
has to be disposed of. More wastes can and should industrial revolution, have been increasingly sub-
be recycled, but for some recycling can have high jected to environmental abuse. In 1984, prior to the
environmental costs and is not necessarily always Interministerial North Sea Conference, the UK
the best option. It may prove better to incinerate dumped just over 9 million tonnes of sewage sludge
organic materials to generate energy, rather than into her coastal seas (Figure 22.2). Added to this
recycle them. For some of the persistent organic were numerous outfalls discharging smaller
compounds, like polychlorobenzenes (PCBs), amounts, which were often highly contaminated
destruction by properly controlled incineration is with industrial and oily wastes. Many bathing
probably the only safe way of disposal. Other beaches had become contaminated with unaccept-
materials, such as some radioactive wastes, have to ably high levels of sewage-derived micro-organ-
be placed in properly engineered repositories where isms, and there was a real fear that many people
they can be stored until they become safe, which, in using the sea for recreation were being unnecessari-
all practical terms, for isotopes with long half-lives ly exposed to pathogens. The sewage was a small,
is for ever. but still significant, source of various heavy metals
At present, about 70% of Britain’s domestic entering UK coastal waters each year (e.g., account-
refuse is disposed of into land-fill sites. In the short ing for 5% of 98 tonnes of cadmium, 3.5% of 38
term, this is a cheap and easy option, but one that tonnes of mercury, and 7% of 2370 tonnes of
in the longer term is creating problems. The num- lead). Most of the discharges of sludge were into

339
 M.V. Angel

22.2 Figure 22.2 Localities and the quantities (in units of mil-
lions of tonnes) at which sewage sludge was dumped in
UK coastal waters in 1984 (BD, Barrow Deep in the
Thames Estuary; GH, Garroch Head in the Firth of Clyde).

Bight near Heligoland, in the North Sea.

Eutrophication is more readily induced in warm
tropical seas, because less oxygen dissolves in
0.01 warmer water and also the higher temperatures
0.14 increase the metabolic activity of bacteria. If anoxic
0.01 GH conditions persist, sulphur bacteria begin to flour-
( ish. These bacteria gain their energy by oxidising
1.7
0.43 North organic material through reducing sulphate ions to
0.3 sulphide ions. Dissolved sulphide ions form hydro-
Sea
gen sulphide, which not only makes the water stink
of bad eggs, but also is lethal to most animals.
When water rich in hydrogen sulphide is stirred up
1.3 0.08 to the surface, there is no escape for the fish and
other animals, so massive mortalities ensue.
Mortalities can also be the result of another
effect of eutrophication. The highly productive con-
BD 0.04 ditions favour opportunistic species; some of the
( 4.2 phytoplankton species which flourish under these
0.3
enriched conditions form dense blooms and secrete
0.01 powerful toxins. Such massive blooms of toxic
0.23 algae (many of which are dinoflagellates) are
0.04 referred to as ‘Red Tides’ (see Chapter 6). In 1988,
0.1
Chrysochromulina polylepis (not a dinoflagellate)
bloomed in the Skaggerak and was then carried by
tidally dynamic regimes where it was very rapidly the currents northward along the coast of Norway.
dispersed. So, in the Barrow Deep in the Thames This bloom killed almost every pelagic and benthic
Estuary, the biggest dump of all, little trace of any organism in its path, and threatened to destroy the
dumped material remained after just a single tidal stocks of the many fish farms along the Norwegian
cycle. However, there were reports of heavy metals coast (in Norway, production of farmed salmon
accumulating in local shellfish, and marked increas- now exceeds its land-based agricultural production,
es in the frequency of skin lesions in local fish. so this was a major threat to its economy).
Biochemical and physiological characteristics of At one site in the UK, off Garroch Head in the
fish caught in the region showed them to be under Clyde, sewage sludge was being dumped in an accu-
some physiological stress, but these signs become mulative regime involving about 1.7 million tonnes
more apparent in the centre of the North Sea, each year (Figure 22.2). The impact of this heavy
where residual tidal currents probably lead to the enrichment (Figure 22.3) resulted in the abundances
accumulation of contaminants from both the of bottom-living animals increasing more than ten-
sewage dumps and riverine discharges of contami- fold at the centre of the dump site. Biomass showed
nants from all the coastal countries. a five-fold increase. The numbers of species per unit
Coastal seas that are continually enriched with area increased around the fringes of the dump site,
nutrients or additional organic material suffer from but then decreased very sharply toward the centre.
eutrophication. This results in a few species with At the centre the populations were dominated by
very high growth and reproductive rates becoming vast numbers of a very few, small, opportunistic
dominant, with a consequent reduction of available species, mostly polychaete and nematode worms.
oxygen in the bottom sediments because of bacteri- The impact of the dump appears to have been
al oxidation of the excess organic matter. In restricted to about 50–75 km2 of sea bed. Not all
regions where the tidal mixing is low and the the effects were negative; the dump site attracted
waters are density-stratified, animals inhabiting the large numbers of fish and so became the centre of a
sea bed die as the oxygen level diminishes; in thriving local fishery. The fish showed no evidence
extreme cases, even those inhabiting the overlying of having accumulated heavy metals or human
water are killed. Recently, such anoxic conditions pathogens. Although conditions in sediments in the
have occurred almost every summer in the German centre of the dump site were highly reducing, the

340
 22: Waste Disposal in the Deep Ocean

22.3
5

✦ ✦ ✦
3
✦ ✦ ✦
✦
2 ✦ ✦
✦S

1 ✦
✦ B

0 A

k
5 4 3 2 1 Centre
1 2 3 4 5

Distance (km)
Figure 22.3 Characteristics of the benthic communities across the Garroch Head sewage sludge
dump site: A, abundances of animals per cm2; B, biomass in 100 g wet weight per m2; S, num-
bers of species per 0.1 m2. The arrow indicates the centre of the dump site (redrawn from
Pearson4).

overlying water remained well oxygenated. The cri- that the deep ocean may prove to be environmen-
teria used to judge the acceptability of such a dump tally acceptable, with due care:
site include the absence of deleterious effects on liv-
ing resources and on other legitimate uses of the sea • First, the vast volume of the oceans, around
(shipping, recreation, and the exploitation of non- 1370 x 106 km3, will serve to dilute concentra-
living resources), and an absence of transmission of tions of any toxins to insignificance. On the
organic and inorganic contaminants and pathogens abyssal plains, living biomass is too sparse (see
back to the human population. Despite the rather Chapter 13) for bio-accumulation significantly to
benign effects of the Garroch Head dump, coastal enhance transfers back to humans.
dumping of sewage has generally been considered to • Second, no living resources are presently exploit-
be unacceptable by European coastal countries, and ed from the abyssal depths, nor is there any likeli-
so the UK agreed to phase out marine disposal by hood of any such resource being discovered.
1997.
• Third, at temperate latitudes the annual deposi-
tion of sedimenting material scavenges metals
The Deep Ocean Option from the water column (Chapter 7).
Internationally, such unacceptable use of shallow • Fourth, heavy metals may not present as serious
seas has been banned under the London Dumping an environmental problem as they do in shallow
Convention, which is managed by the International seas. We now know that hydrothermal vents emit
Maritime Organisation (IMO); these embargoes substantial quantities of heavy metals into the sea
have automatically been extended to deep-ocean water overlying the mid-oceanic ridges (Chapter
disposal as well. Has this been wise? Other chap- 10), and yet these plumes of dissolved metal
ters demonstrate the fundamental importance of remained undetected until very recently. They are
the integrity of oceanic processes to the mainte- very rapidly diluted and, more importantly, the
nance of the global environment. Does this mean geochemical processes within the water column
that no use can be made of the oceans in seeking result in the rapid redeposition of the metals back
solutions to some of the problems of waste man- onto the sea floor. If the high concentrations of
agement; for example, by discharging large-bulk dissolved and particulate heavy metals that are
wastes directly onto the sea bed at abyssal depths normal around vents were found in inshore
of around 4000 m, where the ensuing environmen- waters they would, quite rightly, cause great con-
tal problems will be less? cern. But around the vents, the rich communities
There are a number of factors which suggest of animals have adapted to cope with them.

341
 M.V. Angel

22.4 0 1 2 3 0 50 100 150 0 0.5 1 Figure 22.4 Concentration profiles

(a)
▼
▼ ▼ (nmol/kg) of dissolved substance at two
▼ ▼ ▼
▼
▼
▼ ▼
stations (t) at 34˚N 66˚W in the North
▼ ▼ ▼
▼
Atlantic and (l) at 32˚N 145˚W in the
1 1 ▼ 1 ▼
North Pacific. (a) Marked reductions in
▼ ▼ ▼
concentration in the near-surface waters
2 ▼ 2 ▼ 2 ▼ as a result of biological utilisation; these
▼ ▼ ▼ are categorised as bio-limited sub-
▼
stances. (b) Some reduction in concen-
3 3 ▼ 3 ▼
tration in the near-surface waters, but the
▼ ▼ influence of biological processes on the
4 PO4 ▼ 4 SiO2 4 Cd profiles is relatively small; these are cate-
gorised as bio-intermediate substances.
(c) No reduction in concentration in the
0 5 10 0 4 8 0 2 4 surface waters, and biological processes
(b)
▼ ▼
▼ play little or no role in determining the
▼
▼ ▼
▼▼ ▼ ▼ shape of the profiles; these are cate-
▼
1 ▼ 1 ▼ 1 ▼ gorised as bio-unlimited substances
▼ ▼ ▼
(redrawn from Kester et al.1).
Depth (km)

2 ▼ 2 ▼ 2 ▼
▼ ▼

▼ ▼ ▼
3 3 3

4 Ni 4 Zn 4 Cu

0 0.1 0.2 0 1 2 3
(c)
▼ ▼▼ ▼ ▼ North Atlantic
▼
▼ North Pacific
▼
▼
1 ▼ 1 ▼
▼ ▼

2 ▼ 2
▼

3 ▼ 3 ▼

4 Pb 4 Mn

In ocean waters, chemical constituents can be detected. Biological studies revealed little change in
classified by their concentration profiles into being the communities around the drums, nor was it evi-
bio-limited, bio-intermediate, and bio-unlimited dent whether these minimal changes had been
(Figure 22.4). If wastes contaminated with metals, induced by the physical presence of the drums or
such as copper, cobalt, nickel, lead, and even mer- by the impact of any leakage of radioisotopes.
cury, were to be discharged directly onto the deep- While not advocating the resumption of the dispos-
ocean floor, even with the remarkable sensitivity of al of radioactive wastes in the ocean, a fuller inves-
modern chemical analytical techniques it is doubt- tigation of this old low-level dump site could pro-
ful if increases in dissolved concentrations of these vide much clearer evidence of the environmental
metals would be detectable, and so there would be acceptability of disposing other types of waste
no biological impact. This hypothesis can easily materials into the deep ocean at a few licensed
tested by a serendipitous ‘experiment’ that has sites.
already been conducted. Prior to 1983, when all Similarly, any pathogens in the waste would
ocean dumping of radioactive waste was banned by need to survive for two or three centuries if they
international agreement, drums containing low- were to return back into the surface waters by the
level radioactive waste were routinely dumped over slow stirring of thermohaline circulation. High
abyssal plains at one or two licensed sites, one hydrostatic pressures certainly prevent the metabo-
being in the Bay of Biscay. A few of the drums have lism of micro-organisms in sewage, but studies are
been recovered and found to be leaking, but no needed to establish how long the pathogens remain
contamination of the surrounding sediments was viable. Even if some pathogens do remain viable,

342
 22: Waste Disposal in the Deep Ocean

22.5 Figure 22.5 Temperatures of bottom water in the

Atlantic, showing how from its source in the Weddell Sea
(WS) it flows northward up the western side, its north-
ward flow in the east being blocked by the Walvis Ridge
(WR) off southwest Africa. Water flows along the Mid-
2.1 Atlantic Ridge through the Romanche Trench (RT) on the
equator and the Vema Fracture Zone (VF) at 10˚N.
DG Eventually, the water flows into the Western European
Basin through Discovery Gap (DG), opposite the Straits
of Gibraltar, at 1–2 Sverdrups (i.e., 1–2 x 106 m3/s), by
1.4 1.8 which time the geothermal flux from the Earth and mix-
VF ing has warmed it to 2.1˚C. The hatching indicates the
approximate position of the mid-ocean ridges.
1.3
the other major oceans. The greater age of these
RT
0.5 bottom waters is reflected in their relatively low
1.9
dissolved oxygen content. So, the deep waters of
the North Pacific and Indian Oceans are more vul-
nerable to oxygen stress. However, microbial
2.0
0.1 degradation rates are slowed by the cool in situ
WR 0.7 temperatures and high hydrostatic pressures, so if
the waste piles up, much of the organic matter will
remain buried and will not be oxidised.
If we assume the worst-case scenario, that the
–0.9
full chemical oxygen demand is realised, will the
WS supply of oxygen in deep water be sufficient to
cope with the demand? First, we need to know how
much reduction in oxygen availability can be toler-
ated by the abyssal ecosystems. If, for example, the
they would have to be concentrated by many biological systems at abyssal depths in the north-
orders of magnitude to be infectious and to be eastern Atlantic can tolerate a reduction of 10%
transferred by some as yet unknown process back (i.e., the in situ oxygen concentrations being
to humans. reduced from 5.5 to 5.0 ml/l), then the supply of
bottom water will allow the disposal of 150 million
Oxygen Demand tonnes of the type of sludge dumped in Barrow
Would oxygen concentrations be reduced signifi-
cantly if the deep ocean were to be used for the dis-
Latitude (˚) 22.6
posal of organically enriched waste, such as sewage 60 S 40 20 0 20 40 60 N
sludge? Most deep waters in the oceans are rich in
dissolved oxygen; the bottom waters are formed at 7 Atlantic
the surface of polar seas, when sea ice is forming. ▼ Indian
Sea water freezes at –1.9˚C and, since the colder the Pacific
water the more gas will dissolve in it, where the 6 ▼▼
▼▼
bottom water is being formed in the Weddell Sea ▼
O2 ml/l

and to the west of Greenland, the dissolved oxygen ▼

▼ ▼
5
concentrations exceed 8 ml/l. The flow of bottom ▼
▼
▼
▼ ▼
water northward in the Atlantic can be traced by ▼
▼
▼

temperature (Figure 22.5), where the bottom 4

▼
▼
▼
topography plays a key role in determining the pat-
terns of flow. The concentrations of dissolved oxy- ▼
▼

gen reflect the age of the deep water (Figure 22.6). 3

These concentrations are enhanced in the north- ▼

western Atlantic by the bottom water formation off

Figure 22.6 Latitudinal distribution of dissolved oxygen
Greenland, but in the northeastern Atlantic concen- (ml/l) in bottom waters of the western regions of the three
trations over the Porcupine Abyssal Plain to the main oceans (modified from Mantyla and Reid2). Note
west of Ireland remain at about 5.5 ml/l. In the the rise in oxygen concentration in the bottom waters of
North Atlantic, the bottom water mixes with the the northwestern Atlantic resulting from bottom water
overlying waters to form North Atlantic deep formation off Greenland. There is no such poleward rise
water; this water mass supplies bottom waters to in the oxygen concentrations in the northeastern Atlantic.

343
 M.V. Angel

22.7 Figure 22.7 Eurythenes gryllus, a large amphipod

species, which can grow to lengths of over 12 cm and is
one of the most abundant and voracious species attracted
to baits on the sea bed. These sorts of scavengers will
probably become more abundant around an organically
enriched dump site, and may prove to be an important
vehicle for the dispersal of some of the contaminants. (©
Heather Angel.)

Deep in the Thames Estuary. During the widespread, but this needs to be verified. However,
Quaternary (i.e., the last 2 million years), the con- there are some obvious exceptions, such as the spe-
centrations of oxygen dissolved in ocean bottom- cialised inhabitants of hydrothermal vents and
water fluctuated beyond this range as a result of seeps (Chapters 10 and 13), and in hadal environ-
the climatic oscillations between glacial and inter- ments (>6000 m) at the bottom of ocean trenches.
glacial conditions. The proximity of these and any other special envi-
ronments would have to be accounted for if any
Biological Impact selection is made of sites for such waste disposal.
How much damage might be caused to the deep-
living communities and would it be a serious loss of The Future
biodiversity? If dumped in one pile, the material The question that urgently needs to be addressed is
would totally destroy the community beneath it, whether continuing with the extensive use of land-
just as everything natural is destroyed at a land-fill fill sites is likely to be environmentally more damag-
site. Such an impact would be trivial compared ing and ultimately more expensive than introducing
with the effects of major geological events (Chapter other waste-management options, including the use
9), and would probably be preferable to dispersing of a limited number of sites on the ocean floor.
the waste widely. The waste would probably be What should not be disposed of in the ocean?
invaded by large numbers of opportunistic species, • Anything that can be considered to be a resource.
as happened in the shallow waters off Garroch
• Any persistent synthetic organic substance which
Head (Figure 22.3), and in turn these dense concen-
is totally unknown in the natural marine or ter-
trations of ‘detritivores’ would attract large num-
restrial environment (for example, organohalo-
bers of predators and scavengers, in much the same
way that scavengers are attracted to baited cameras
and traps (Figures 22.7 and 22.8; see also Chapter
13). These mobile animals may play a greater role 22.8
in dispersing the contaminants than would the
water currents, which at abyssal depths are general-
ly weak, except in regions where meso-scale eddies
generate benthic storms (Chapter 4).
Oceanographers have still not sampled even a
hundred millionth of the total area of the ocean
bed, so we still know very little about the broad-
scale distribution patterns of deep-living species.
However, preliminary data for the larger animals
imply that most benthic species have widespread
distributions, both within and between the ocean
basins. There seems to be no reason to doubt the
assumption that the smaller species are equally

Figure 22.8 The remains of a mackerel used as bait in a

trap set on the sea bed for 2 days at a depth of 4500 m
off the Cape Verde Islands. Three species of amphipods,
including Eurythenes gryllus, were in the trap and had
taken part in the rapid consumption of the bait. (Courtesy
of the Southampton Oceanography Centre, UK.)

344
 22: Waste Disposal in the Deep Ocean

gens and PCBs) and is truly essential to manufac- These are complex issues which need further
ture should be chemically destroyed. research and large-scale experiments to assess. But
• Radioactive isotopes? Maybe there is a case for will these be conducted before the urgency to find
reconsidering how the isotopes produced for solutions to waste problems becomes so over-
weapons and peaceful purposes are to be con- whelming that decisions have to be made (because
tained and disposed of. Some of the present reposi- of socio-economic pressures on the land) in the
tories are vulnerable to natural catastrophes and absence of proper scientific evaluations based on
terrorist acts. sound ecological and biogeochemical principles?
• How about carbon dioxide? Can humans slow
down the rate of increase in the atmosphere
(which promises to change our climate) by dis-
charging carbon dioxide into the deep ocean,
where it will dissolve? The Japanese are actively General References
developing technologies to do so, but there are Kullenberg, G. (ed.) (1986), The Role of the Oceans as a Waste
potential dangers which require research. Adding Disposal Option, NATO ASI Series, D. Reidel
more carbon dioxide to deep waters will lower Publishing Co, Dordrecht, 725 pp.
their pH. If these more acidic waters erode and Spencer, D.W. (1991), Report of a Workshop to Determine the
destabilise calcareous sediments at the base of Scientific Research Required to Assess the Potential of
the Abyssal Ocean as an Option for Future Waste
the continental slopes, the result might be mas- Management, Woods Hole Oceanographic
sive failures of the continental margins (Chapter Institution, 111 pp.
9). There may be substantial biological impact –
an effective way of anaesthetising aquatic ani-
mals is to squirt carbonated water into their con- References
tainer. Would the disposal of so much carbon
1. Kester, D.R., Burt, W.V., Capuzzo, J.M., Park, P.K.,
dioxide into the deep ocean put benthic commu- Ketchum, B.H., and Duedall, I.W. (1985), Wastes in
nities to sleep permanently, and what would be the Ocean, Vol. 5, Deep-Sea Waste Disposal, Wiley
the scale of this impact – over 10, 100, or even a Interscience, 346 pp.
1000 km? And what scale of impact might we 2. Mantyla, A.W. and Reid, J.W. (1983), Abyssal characteristics
of the world ocean waters, Deep Sea Res., 30A,
consider to be acceptable in the context of even 805–833.
greater problems associated with global warm- 3. McAllister, D.E. (1993), How much land is there on Earth?
ing? The scale of such impacts is of serious con- For people? For nature?, Global Biodivers., 3, 6–7.
cern, because terrestrial ecologists repeatedly find 4. Pearson, T.H. (1986), Disposal of sewage in dispersive and
it impossible to extrapolate the results of experi- non-dispersive areas: contrasting case histories in
British coastal waters, in The Role of the Oceans as a
ments conducted in plots of a few metres to fore- Waste Disposal Option, Kullenberg, G. (ed.), NATO
cast what will happen at scales of tens and hun- ASI Series, D. Reidel Publishing Co, Dordrecht, pp
dreds of kilometres. 577–595.

345
Some Commonly Used Words and Terms
Depth Zones Epifauna. Animals that live on the sea bed, either
attached or moving freely over it.
Abyssal. A subdivision of the benthic zone
encompassing the ocean floor between a depth of Epiflora. Plants that live in contact with the sea bed.
2–6 km. Eutrophication. The process whereby water becomes
Abyssopelagic. Open-ocean (oceanic) environment anoxic through decomposing organic matter.
below 4 km depth. Foraminifera. Planktonic and benthic protozoans that
Aphotic. The dark region of the ocean that lies below have a skeleton or shell composed of calcium carbonate
sunlit surface waters. (CaCO3).
Benthic. That part of the ocean adjoining the sea bed. Holoplankton. Plants and animals that are plankton for
their entire life.
Epipelagic. The upper region of the ocean extending to a
depth of about 200 m. Infauna. Animals that live within or burrow through the
substrate (sand or mud).
Euphotic. The surface layer of the ocean that receives
enough light to support photosynthesis. Macroplankton. Large plankton (such as jellyfish and
Sargassum weed).
Eutrophic. That with an abundance of nutrients.
Meiofauna. Small species of animals that live in the
Littoral. The benthic zone between the highest and low-
spaces among particles in a marine sediment.
est normal water marks; the intertidal zone.
Meroplankton. Planktonic larval forms of organisms
Neritic. The water that overlies the continental shelf,
that are members of the benthos or nekton as adults.
generally of water depth less than 200 m.
Microplankton. Plankton of length 0.06–1 mm.
Oceanic. The waters beyond the shelf break, generally
of water depth greater than 200 m. Nanoplankton. Plankton of length <50 μm.
Pelagic. All water in the oceans, including the neritic Nektobenthos. Those members of the benthos that are
zone and oceanic zones. active swimmers and spend much time off the bottom.
Sublittoral. That portion of the benthic environment Nekton. Pelagic animals, such as adult squids, fish, and
extending from low tide to a depth of 200 m, often taken mammals, that are active swimmers to the extent they
as the surface of the continental shelf. can determine their position in the ocean by swimming.
Subneritic. The benthic environment extending from the Phytoplankton. Plant plankton, the primary producers
shoreline across the continental shelf to the shelf break. of the oceans.
Picoplankton. The smallest plankton, with a body
Plankton, Bacteria, and Marine Animals length of <2 μm.
Aerobic bacteria. Bacteria that undergo respiration in Plankton. Organisms that float or have weak swimming
the presence of free oxygen (O2). abilities.
Anaerobic bacteria. Bacteria that undergo respiration in Sulfur-oxidizing bacterium. Any bacteria which use
the absence of free oxygen (O2). energy released by oxidation to synthesise organic matter
Autotroph. Plants and bacteria that synthesise food chemosynthetically.
from inorganic nutrients. Ultraplankton. Plankton with body length <5 μm.
Benthos. Organisms that live on or within the sea bed. Zooplankton. Animal plankton.
Coccolithophores. Microscopic, single-celled plant
plankton having exo-skeletons composed of tiny, calcare-
ous plates or discs called coccoliths. Further definitions may be found in the Glossaries of:
Demersal organisms. Organisms which rest on the sea • Pinet, P.R. (1992), Oceanography. An Introduction
bed, but swim and feed in the water column. to the Planet Oceanus, West Pub. Co., New York,
Diatoms. Microscopic, unicellular phytoplankton pos- 570 pp.
sessing silica valves. • Thurman, H.V. (1994), Introductory Oceanography,
Dinoflagellates. Microscopic unicellular phytoplankton 7th edn, Macmillan, New York, 550 pp.
that propel themselves using tiny, whip-like flagella.

346
Acronyms
ADCP Acoustic Doppler Current Profiler JAMSTEC Japan Marine Science and Technology
AGCM Atmosphere General Circulation Model Centre
ALACE Autonomous Lagrangian Circulation Explorer MAFF Ministry of Agriculture, Food and Fisheries,
ARCS Autonomous Remote Controlled Submersible UK
ASP Amnesiac Shell Poisoning MODE Mid-Ocean Dynamics Experiment
ATOC Acoustic Thermometry of Ocean Climate MORT Mean Ocean Residence Time
ATSR Along-Track Scanning Radiometer MYRTLE Multi-Year Tide and Sea Level
AVHRR Advanced Very High Resolution Radiometer Equipment
BP Time in years Before Present day NAEC Normal Atmosphere Equilibrium
CASI Compact Airborne Spectrographic Imaging Concentration
CCD Calcite Compensation Depth NEADS North East Atlantic Dynamics Study
CFCs Chlorofluorocarbons NSP Neurotoxic Shellfish Poisoning
CLIMAP Climate: Long-Range Interpretation, ODP Ocean Drilling Programme
Mapping, and Prediction OSCR Ocean Surface Current Radar
CTD Conductivity, Temperature, and Depth OTEC Ocean Thermal Energy Conversion
measuring Instrument PCBs Polychlorinated Biphenyls
CZCS Coastal Zone Colour Scanner POC Particulate Organic Carbon
DGPS Differential Global Positioning System POM Particulate Organic Matter
DIC Dissolved Inorganic Carbon PSP Paralytic Shellfish Poisoning
DMS Dimethyl Sulphide RAFOS SOFAR backwards – to indicate that the
DOBS Digital Ocean Bottom Seismometer RAFOS system works in the opposite mode
DOC Dissolved Organic Carbon to conventional acoustically tracked floats
DSP Diarrhoetic Shellfish Poisoning ROV Remotely Operated Vehicles
DSV Deep Submergence Vehicle SA Selective Ability
EEZ Exclusive Economic Zone SAR Synthetic Aperture Radar
EKE Eddy Kinetic Energy SCUBA Self-Contained Underwater Breathing
EM Electromagnetic Apparatus
ENSO El Niño Southern Oscillation SCV Sub-mesoscale Coherent Vortex
EPSONDE This is not an acronym, but the name of an Sigma t (st) A convenient measure, commonly used in
instrument derived from running together place of density, equal to 1000[(l/lm) – 1)]
the words ‘epsilon’ (the Greek letter where l is the density
normally used to denote the rate of dissipa- and lm = 999.975 kg/m3, the maximum
tion of turbulent kinetic energy – which is density of pure water [see Gill, A.E. (1982),
what the probe is used to measure) with Atmosphere–Ocean Dynamics, Academic
‘sonde’, the German word for probe. Press, London, 662 pp]
ERBE Earth Radiation Budget Experiment SMBA Scottish Marine Biological Association (now
ERS Earth Resources Satellite the Scottish Association of Marine Science,
FAD Fish Attracting Device SAMS)
FAO United Nations Food and Agriculture SOC Southampton Oceanography Centre, UK
Organisation SOFAR Sound Fixing and Ranging
FCA Floc Camera Assembly SST Sea Surface Temperature
FLIP Floating Instrument Platform STD Salinity, Temperature, and Depth measuring
FRAM Fine Resolution Antarctic Model instrument
GCOS Global Climate Observing System TAG Trans-Atlantic Geotravers
GEOSECS Geochemical Ocean Sections Programme TBT Tributyl Tin
GLORIA Geological Long Range Inclined Asdic TEP Transparent Exopolymer Particles
GOOS Global Ocean Observing System TM Thematic Mapper
GPS Global Positioning System TOBI Towed Ocean Bottom Instrument
ICES International Council for the Exploration of TOC Total Organic Carbon
the Sea TOGA Tropical Ocean–Global Atmosphere
IFOV Instantaneous Field of View Experiment
IMO International Marine Organisation UML Upper Mixed Layer
IOSDL Institute of Oceanographic Sciences, Deacon UOR Undulating Oceanographic Recorder
Laboratory, UK WHOI Woods Hole Oceanographic Institution, USA
IPCC Intergovernmental Panel on Climate Change WOCE World Ocean Circulation Experiment
ITCZ Intertropical Convergence Zone

347
Index Atlantic, 9, 15, 118, 120–121
Atlantis, 89
Atlantis II Deep, 336
Bragg scattering, 297
Brandt, Karl, 24
Brazil–Falklands Confluence, 77, 82
A bold number indicates major ATLAS buoys, 56 Brisingella multicostata 200, 201
information on an entry (including atmosphere, 27–35 brittle stars, 197, 212, 225
relevant illustrations); an italic num- atmospheric, circulation, 33–35, 54–56 broadband acoustics, 297
ber indicates an illustration or cap- ATOC (Acoustic Thermometry of Ocean bromine, 324
tion (and any relevant text). Climate), 50, 51, 297 bryozoans, 259, 267, 269
Aurelia aurita, 256, 257 Buccinum undatum, 268
Aanderaa current meter, 287 authigenic sediments, 129, 131, 172–173 buffering, 184
abalone, 267, 321 Autosub, 299
Acanthephyra curtirostris, 222 AVHRR sensor, 65, 76 cadmium, 177
Acanthephyra purpurea, 216, 222 axial valley, 153, 541 caesium, 175
acid rain, 94 calcite, 90, 103, 129
acoustic sensors, 294–297 back-scatter, 78, 80–81, 86, 139, 142, 149 calcite coccoliths, 65
acuity, 219–220 bacteria, 100, 107, 110, 111, 131, 206, 250 calcite compensation depth, 103, 129–130,
ADCP (Acoustic Doppler Current Profilers), Baffin Island, 107 173, 179, 185
294, 295 Bahamas, 62, 196, 210 calcium, 171
Adriatic Sea, 94, 106, 264 Baltic Sea, 246, 267 calcium carbonate, 131–133, 176, 179–180
aeolian transport, 129, 130, 169–170, 180 Barents Sea, 147 California, 56, 150, 261, 318
Aequorea, 215 barnacles, 255, 325 Calocidaris micans, 209
aerosols, 168, 169 baroclinic instability, 67 Calyptogena magnifica, 158
Afar rift junction, 154 barotropic instability, 67 camouflage, 213–217, 224–225
aggregates, 97, 100, 101, 104, 106–108 barrages, 326–327 Campaniclava clionis, 234
Type A, 100, 101 Barrow Deep, 340, 343 Canary debris flow, 137, 139, 140–141,
Type B, 100, 107–108, 110, 111 basalts, 124–125, 126, 152, 153–155, 156, 170 142–143
Agulhas Current, 53, 68 Batfish, 284 Canary Islands, 141–142, 145
Agulhas Retroflection, 63, 82, 83 bathymetry, 22, 80 Cape Basin, 167
Agulhas slide, 137 Bathymodiolus thermophiolus, 158 Cape St Vincent, 66
air–sea exchange, 189, 190, 191 Bathynomus giganteus, 207 Cape Verde Islands, 344
ALACE (Autonomous Lagrangian Circulation Bathysnap, 102–104, 195, 204 carbohydrate, 157, 167
Explorer), 292 bathythermographs, 25, 48 carbon, cycle, 54, 110, 131–132, 171, 178,
Albatross, 22 Bay of Biscay, 76, 342 180, 185–186
Albert I, Prince of Monaco, 20–22 Beaufort eddies, see sub-meso-scale coherent carbon dioxide,
Aleutian Chain, 150 vortices (SCVs) atmospheric, 132, 170–171, 178–180,
algin, 325 Benthalbella, 219–220 182–183, 189–190, 192
alkalinity, 187 benthic fauna, 96, 103–104, 105, 239–241 dissolved, 129, 157, 162, 167, 178–179,
altimetry, 78, 82, 84, 155 benthic landers, 288–289 184–186, 191–193
Alvin, 158, 161, 298–299 Benthogone rosea, 105, 199 levels, 41, 54, 65, 182, 339, 345
Alvinocaris, 222 Bering Sea, 150 carbonates, 129–130, 167, 173, 178–179,
Amazon, River, 169 Bermuda, 61, 109 186–189, 193
ammonia, 107 Bermuda Plateau, 196 Caribbean ‘Blue Holes’, 278
amphipods, 204 Beroe, 215 Carson, Rachael, 96
Amphitretus, 220 Bigelow, H.B., 23 catagenesis, 130–131
Anachalyspicrinus nefertiti, 200, 201 biodeposition, 251–252 CDT (Conductivity, Temperature, Depth)
anchoveta, 317–318 biodiversity, 158–159, 267 instrument, 282, 283, 284
anemones, 206, 259 biogeochemical cycles, 96, 108, 110–111 Chaenophryne, 225
angler fish, see Chaenophryne biological, Chaetoceros, 101
angular momentum, 35 modelling, 100, 109–110 Challenger, HMS, 14, 18, 18–20, 19, 195
anhydrite, 160 pump, 176, 179, 181, 191 Challenger Report, 19–20, 165, 195
anoxia, 94, 107, 131, 171 utilisation, 342 Charles Darwin, RV, 91
Antarctic Bottom Water, 167 bioluminescence, 202, 212, 222–224 Chatham Rise, 335
Antarctic Circumpolar Current, 32, 35, 36, 44, biomass, 89, 92, 192, 237–238, 240–241, 267, chemo-autotrophy, 96
50, 53, 63, 77, 83, 228, 231, 289 340 chemosynthesis, 157, 158
Antarctic Convergence, 233 biomass profiles, 233 Chesapeake Bay, 94
Antarctic Ocean, 133 biophysical models, 90 chimneys, 156–157, 157, 159–160
Antarctica, 138, 181 biosecretion, 252, 253 Chirp systems, 308, 312
anthropogenic effects, 48 biosensors, 297 chlorofluorocarbons (freons), 165, 167, 282,
anticyclones, 27, 33, 35, 55 biotechnology, 325 339
antifouling, 325 bioturbation, 250–252, 251 chlorophyll, 64, 71, 79, 89–90, 92–93, 98, 192
appendicularian, 100 bivalves, 202 chlorophyte, 108
aquaculture, 266–267 Black Sea, 13, 48, 157 Chorocaris Chacei, 206
Aquashuttle, see UOR black smokers, 156–157, 159, 160, 161–162 chromite, 334
aragonite, 129, 187–188, 191 Blake Plateau, 172 chronometers, 15
ARCS vehicle, 299 Blake Spur, 127 Chrysochromulina polylepis, 340
Arctic Ocean, 133, 169 Blake, USCG, 20, 22 ciliate, 90–91
Arenicola marina, 251 blooms, 65, 79, 89–94, 107, 110, 237, 241, Circumpolar current, see Antarctic
Argentine Basin, 167 254–255, 340 Circumpolar Current
ARGOS satellite system, 287, 289, 291–292 Blue Revolution, 323 Clam Acres, 158
Argyropelecus, 215, 219, 224–225 Bohai Sea, 270 clathrates, 332
Aries, 61 Bosphorus, 13, 14 Claude, Georges, 329
Artemia salina, 249 bottom-mounted instruments, 288–299 climate change, 4–6, 41, 48, 51, 54, 56–57, 296
artificial reefs, 259–272, 339 bottom water, 43, 44, 343 climate system, 41–42, 42
artificial sea floor, 321 Bouldner Cliff, 312 Clio recurva, 234
Asopos, 276 Bouma sequence, 144 Clione, 315
asthenosphere, 113, 116, 120–122, 124 boundary layer, 37–39 closed cycle systems, 329
Atlantic Convergence, 233 Bourne, William, 10 Clytia, 256
Atlantic North Equatorial Current, 61 Boyle, Robert, 11 cnidarians, 200, 201
Atlantic Ocean, see also North Atlantic; South brachyuran crabs, 158 coal-ash concrete, 268, 270, 271

348
Coal Oil Point, California, 330–331 DOBS (Digital Ocean Bottom Seismometer), 268, 272, 319, 320–321
Coal Waste Artificial Reef Program (CWARP), 288 fishing limits, 322–323
268, 270, 271 Dohrn, Anton, 17, 20 flagellates, 108
cobalt, 335 Dolichopteryx, 221 Floating Instrument Platform (FLIP), 281, 295
coccolithophores, 90–91, 108, 129, 147, 172 doliolids, 96 Florida, 232, 271
cod, 197, 316–318, 319, 320–321 domestic waste, 339 Florida Straits, 329
Coelopleurus floridanus, 209 Doppler, 294 flow cytometry, 297
Columbus, Christopher, 9 Dornchus thomsoni, 199 flow injection chemical analysis, 297
Conchoecia valdiviae, 238 Drake Passage, 289 fluorescence, 99, 282
continental, dredging, 333–334 fly ash, 338–339
crust, 113, 124 drifters, 62, 291–292 foraminifera, 24, 129, 132–133, 172, 241
drift, 41 drogues, 291 fossil-fuel emissions, 189
shelf, 104, 128, 132–133, 195–197, 320 DSV (Deep Submergence Vehicle), 298 Fredrikshab, 319
slopes, 61, 97, 104, 118, 136, 137, 141–143, dump sites, 205, 339–341 free-fall instruments, 290, 291
145, 332 dyke systems, 125–126, 152, 153 freons, 165, 167
convection, fronts, 76, 233
heat, 156 earthquakes, 114, 120, 136, 140, 143, 145,
sea water, 46 150, 160 gabbros, 124, 126, 126
convective circulation, 170 East African Rift, 154 Gaia hypothesis, 178, 180, 194
convergence zones, 54–55, 233 East Indies, 232 Galapagos Islands, 56, 192
Cook, Captain James, 13 East Pacific Rise, 156, 157, 158, 161, 336 Galapagos Rift, 156–157, 158
copepods, 225, 230, 232, 253, 255 Eastern Pacific, 184 galatheid crabs, 158, 204
Copernicus, Nicolas, 9 echinoderms, 200, 204–205 Gammarus duebeni, 249
copper, 160, 165–166, 167, 168, 336 echinoids, 105 Garroch Head, 241, 267, 340, 341, 344
corals, 13, 56, 232, 261, 267–268, 334 Echinus affinis, 105, 200 gas, 131, 136, 143, 150, 330, 332
core samples, 128, 131, 144, 277 echiuran worms, 202, 203 genetic diversity, 242–243
corers, 103, 104, 283, 284 echo-sounders, 99, 162, 163, 305–307 geoid, 78, 82, 83
coriolis force, 23, 38 echo-sounding, 24 geomagnetic field, 24, 117, 123, 126
Coryphaena 216 eddies, hydrothermal, 162–163 geometric correction, 78
Côte d’Ivoire, 120 eel, 219 geophysical calibration, 75
crabs, 197, 199, 248, 259 EEZ (Exclusive Economic Zones), 314, Geosat, 84, 85
Crammag Head Lighthouse, 297 314–315 GEOSECS (Geochemical Ocean Sections), 177,
Cranchia scabra, 216 EKE (eddy kinetic energy), 62, 66, 67, 78 181
Crateromorpha, 200 Ekman spiral, 23 geostationary orbits, 73–75, 74
Crenilabrus melops, 264, 267, 322 El Niño–Southern Oscillation (ENSO), 48, geothermal energy, 116, 157
crinoids, 200, 201, 204, 208, 209 54–56, 78, 317, 318 German Bight, 94, 340
crust, electromagnetic spectrum, 69, 302 Geryon trispinosus, 197
continental, 113, 116, 120–121, 123, 124 electromagnetic waves, 304, 313 giant clams, 158
oceanic, 113, 116–118, 120–121, 124, 126, electronic pebble, 293 Gibraltar, Straits of, 12, 14, 48, 231
133 Eling Tide Mill, 326 glacial cycles, 41, 46–47, 133
crustaceans, 202, 204–205, 209, 221, 242, 255, Emiliania Huxleyi, 65, 89, 91, 94 glauconite, 172
264, 321 Endoxocrinus, 208 Global Climate Observing System (GCOS), 48
cryptic fish, 259 energy, 325–330 Global Ocean Observing System (GOOS), 48,
cryptomonads, 90 energy exchanges, 27–29, 31 57, 281
ctenophores, 225, 256 English Channel, 76, 333, Global Thermohaline Conveyor Belt, 43, 46–47
Curie temperature, 117 Enypniastes eximia, 207 global warming, 51, 182
current measurement, 12, 60–63, 287, 291–292 epibenthic sledge, 285 global water cycle, 28, 168, 169
currents, 15, 23, 35–37, 50, 60–62 EPSONDE, 290 globigerina ooze, 20
Cyana, 298 Equatorial currents, 35, 35–36 GLORIA (Geological Long Range Inclined
Cyathura carinata, 245 ERS-1, European Space Agency satellite ERS-1 Asdic), 137, 140, 144, 148–149, 150, 294,
Cystisoma, 220 74, 81, 84, 85–87 300, 309–311
estuaries, 244–257 Gnostoma bathyphilum, 216
debris flows, 137–143, 144 eukaryotes, 230 Goban Spur Basin, 122
decapods, 206, 221 Euler’s Theorem, 113, 115 gold, 334
Deep Sea Drilling Project, 131 Euphausia superba, 23, 235, 316, 320 GPS (Global Positioning System), 282, 303,
deep water circulation, 176 euphausiids, 219, 220–221, 235–236, 237, 239 304
DeMets, C., 115–116 euphotic zone, 89, 107, 110 graben, 123, 336
denitrification, 178 Euplectella sp., 200 Grand Banks, 318, 323
density, 14, 18, 52, 64 Eurythenes gryllus, 344 Grand Banks turbidity current, 136, 144–145
deposit feeders, 250 eutrophication, 94, 241, 340 gravel, 333, 334
depth soundings, 11, 13, 17 evaporation, 9, 12, 28, 29, 30, 32, 37–40 gravity measurements, 24, 115, 123
desalination, 323, 330 Ewing, M., 24, 145 Great Salinity Anomaly, 45
diagenesis, 130–131 Exmouth Plateau, 124, 125 greenhouse gases, 51, 57, 132, 178, 182
diamonds, 334 Greenland, 41, 92, 318, 319, 343
diatoms, 89–94, 97, 100–101, 108, 253 faecal pellets, 96, 100–101, 111, 199, 203, 251 grenadier, 205, 321
Dictyosysta elegans, 108 Falkland Current, 77 Gulf Stream, 35, 36, 60, 61, 62–63, 329
diel variation, 101, 102, 103, 111 Fasham, M.J.R., 100 Gulf Stream rings, 63–64, 67
differential GPS, 304 faults, 113, 114–115 gyres, 32, 35, 42, 50, 62, 83
dikes, see dyke systems FCA (Floc Camera Assembly), 99
dimethyl sulphide, 65, 94, 180 Ferrel cells, 34 habitat variability, 232–233
dinoflagellates, 90, 93–94, 225, 254, 340 ferromanganese nodules, 172 haddock, 316
Discovery, RRS, 23, 25, 106, 294–295 Fine Resolution Antarctic Model (FRAM), 51, Hadley cells, 33–35
Discovery Deep, 336 53 hake, 316
Discovery Gap, 343 fine structure, 276 halibut, 197, 322
Discovery II, RRS, 23, 60 fish-attracting devices (FAD), 263 haline catastrophe, 39
dissolved inorganic carbon (DIC), 185, 187, fish eyes, 220–222 Halley, Edmond, 11, 12
191 fish farms, 267, 322, 322–323, 340 Halosphaera viridis, 230
diurnal migration, 101, 102, 103 fish scales, 318 hatchet fish, see Argyropelecus
divers, 274–279 fisheries, 22–23, 54, 67, 78, 259, 263–264, Hatteras, Cape, 63

349
Oceanography

Hatteras Abyssal Plain, 196, 207 Kerala, 268 78, 79, 82, 92, 295
Hawaii, 183 Kiel School, 24 meso-zooplankton, 97, 254–255, 257
Hawaiian Deep, 141 Kimmeridge Bay, 331 Mesodinium rubrum, 90, 91–92, 254, 255,
Hawaiian-Emperor chain, 120 Kophobelemnon stelliferum, 198 256–257
Hawaiian Islands slides, 137, 140–141 krill, see Euphausia superba metalliferous sediments, 160–161
Heard Island, 50, 296, 297 Krupp Atlas Hydrosweep, 310 Meteor, 23–24
heat flux, 31–32, 36–37, 42, 43, 52, 163 Kuroshio, 35, 36, 60, 63, 96, 97 meteorology, 27, 39
Helland-Hansen, Bjorn, 23, 60 meteorology buoy, 288
herring, 197, 246, 317, 318, 319, 320–321 Lac de Cazaux et Sanguinet, 312 Meteosat, 74, 86
Hilnia slump, 140 laminar flow, 138, 143, 276 Meterythrops picta, 235
Hippolyte coerulescens, 214 land utilisation, 338 methane, 131, 132, 161–162, 182, 332, 339
Histoire Physique de la Mer, 13 Landsat TM, Thematic Mapper 74, 154, 256 Mexico, Gulf of, 98, 232, 259, 332
holoplankton, 255 Langmuir circulation, 39 Michael Sars, 23
holothurians, see sea cucumbers LAPS (Large Aggregate Profiling System), 99 micro-algae, 89–93
Homarus americanus, 265 latent heat, 28 microbiota, 107
Homarus gammarus, 266 Law of the Sea, 314–315 micronekton, 285
hot brines, 336 levees, 148, 150 microplankton, 97
hot spots, 19, 114, 120, 124, 126 light characteristics, 212, 213 microwaves, 70, 78, 80, 86
HRP (High Resolution Profiler), 290 light scattering, 79, 89 microzooplankton, 111
hurricanes, 86, 87, 151 lightfish, 286 Mid-Atlantic Ridge, 19, 154, 155, 159, 162,
Hvalfjorder, 319 Lightning, HMS, 18 196, 205–207, 310
hydrates, 332 Limulus polyphemus, 324, 325 Mid-Ocean Dynamics Experiment (MODE), 61
hydrocarbons, 131, 132, 147 lithium, 175 mid-ocean ridges, 152–156, 152
hydrogen, 162 lithosphere, 113, 116–117, 121–122, 124, 171 migration, diel, 101, 102, 103
hydrogen sulphide, 157, 158 Loano, 268, 269, 270 Milankovitch cycles, 41, 232
hydroids, 259, 267 lobsters, 198, 199, 259, 264–266, 279 mineral precipitation, 156, 157, 159, 160–161
hydrological cycle, 28–29, 30 Loch Etive, 322 minerals, 333–336
hydrophones, 296 London Dumping Convention, 271, 341 mining, 334
hydrosphere, 169 low light cameras, 297 Mir I, 298
Hydrosweep, 310 Lucaya caves, 278 Mississippi Fan, 149, 151
hydrothermal, luciferin, 222 models,
deposits, 125, 341 luminous bacteria, 223, 225–226 lithosphere, 116–117
fluids, 156, 157, 170 luminous defences, 225 ocean circulation, 51–54
mounds, 159–161, 336 Lydekker’s Line, 232 Moho, 117, 124, 126, 127
plumes, 156, 161, 162 lysocline, 129, 130, 179, 185 molluscs, 202, 205, 248, 255
vent fauna, 157-159, 162 monsoons, 9
vents, 156-163, 205–207, 222, 336 mackerel, 315, 344 Monterey Fan, 311
Hymenodora, 222 macrobenthos, 239, 240, 241 Mount Rainier, 139
hyperbolic systems, 303 macrophytes, see seaweeds mucus, 100, 107, 252, 253
hypertonic regulation, 248–249 Madeira Abyssal Plain, 103, 137, 141, multibeam swath sounders, 310–311
hypotonic urine, 249 142–145, 146–147, 205, 335 Mundus Subterraneus, 10
Maelstrom, 10 Munida tenuimana, 198
Iberia, 123 Mafia Island, 275 Munidopsis crassa, 204
ice cores, 41, 180, 183 magma, 117, 119, 122–124, 152, 153, Munidopsis susquamosa, 158
ice coverage, 39, 40, 45, 57, 132–133 155–156 Murray, John, 20, 22
Iceland, 124, 126, 154, 162, 163 magnesium, 165–166, 170, 323–324 Musée Océanographique, 20, 21
Idothea metallica, 214 magnetic polarity, 24, 117, 123, 126 mussels, 266–267, 322
Indian Ocean, 9, 61, 84, 85, 118, 121, 296, magnetic variation, 17, 117 MYRTLE (Multi Year Tide and Sea Level
343 Malacosteus, 223–224 Equipment), 289
Indonesian Throughflow, 32, 43 manganese, 20, 131, 161, 170, 177, 335, 336 mysids, 235–236, 237, 249
Indus Fan, 148 mantle, 14, 113, 116, 120, 121, 123–124, 152, Mytilus edulis, 246
infra-red radiation, 70–71, 76, 297 153, 170
Institut Océanographique, Monaco, 267 mantle convection, 113 Nansen, Fridtjof, 23, 60
Institute of Oceanographic Sciences, 312 mariculture, 321–322, 330 nasse triédrique, 21
instrumentation, 280–299 marine, Natural Energy Laboratory of Hawaii, 330
Interministerial North Sea Conference, 339 biology, 16–18 natural gas, 136, 143, 150, 330, 332
internal waves, 23, 100, 276 carbonate system, 182–194 navigation, 13, 15, 282, 300, 302, 303, 304
ion-pair formation, 188 Marine Biotechnology Institute, 325 Navstar, see GPS
ion-transport, 250 marine snow, 90, 96–111 nekton, 101, 110
Iridogiorgia sp., 200 Type A, 100, 101 Nematonurus armatus, 205, 321
iron, 131, 160, 161, 166, 336 Type B, 100, 107–108 Neomysis integer, 245
iron enrichment, 192 Marsigli, Count Luigi, 11, 13–14 nephelometer, 162
Islay Wave Power Station, 328, 329 Mary Rose, 260, 277 Nereis diversicolor, 245
isohalines, 246 mass extinctions, 228 Netherlands Institute for Sea Research (NIOZ),
isopods, 207, 214, 249 Matthiessen’s ratio, 220 94
isostasy, 45, 122, 123 Maury, Matthew Fontaine, 15, 60 Newfoundland, 136, 320
ITCZ (Intertropical Convergence Zone), 30, mean oceanic residence time (MORT), 174, Nezumia, 217
32, 35 177 Nice, 145, 147, 150
IXTOC-1, 332 meddies, see sub-meso-scale coherent vortices Ninetyeast Ridge, 120
(SCVs) nitrate, 165, 175
Japan, 236, 267, 270, 271, 334 medicines, 324–325 nitrogen cycle, 176, 178
Japanese fishing reefs, 262, 271 Mediterranean Sea, 12–13, 66, 232, 265, 268 Noctiluca, 90
Japetella, 225 medusae, 215, 225 Nod Hill, 335
Java, 241, 243 Megacalanus princeps, 230 nodules,
jelly fish, 215, 223 Meganyctiphanes, 219 ferromanganese, 172
Johnson Sealink, 298 meiobenthos, 241 manganese, 20, 335, 336
meridional fluxes, 29–32 phosphorite, 334
Keahole Point, Hawaii, 330 meroplankton, 255 normal atmospheric equilibrium concentration,
kelp, 261, 262, 267, 325, 330 meso-scale eddies, 59, 62, 64, 68, 73, 76–77, 190–191

350
 Index

North Atlantic Current, 46 Pennatula aculeata, 198 red clay, see pelagic clays
North Atlantic Deep Water, 36, 39, 43–44, 47, Perotrochus midas, 210 Red Sea, 120, 124, 336
167, 176, 200, 343 Peru, 317, 318 red tide, 79, 89, 93, 255–255, 256, 340
North Atlantic Ocean, 36, 60, 84, 91, 92, 127, Peru Basin, 335 Redfield ratios, 176
177 , 196, 238, 342 pesticides, 165 relict species, 232
North East Atlantic Dynamics Study (NEADS), Phaeocystis sp., 90, 94 remote sensors, 294–296
62 pharmaceuticals, 324 Rennell, James, 15, 18, 60
North Pacific Ocean, 36, 44, 84, 172, 177 , Pheronema carpenteri, 198, 240 Reykjanes Ridge, 163
342, 343 Phialidium hemisphericum, 257 rhabdoms, 222
North Sea, 80, 90, 93, 94, 241, 243, 318, 319, Philippines, 259, 262–263 Rhizosolenia delicatula, 92
331, 340 phosphate, 165, 175, 177 rifted margins, 121–124
Norway, 124 phosphorite, 172, 334, 335 Riftia pachyptila, 158
Norwegian Sea, 23, 84, 121, 143, 176 Photoblepharon, 226 rifts, 118, 152–153, 154, 156–157
Novodinea antillensis, 208 photographic techniques, 99, 102, 103, 313 Rimicaris exoculata, 160, 206
Nuuanu slide, 137, 140, 141 photosynthesis, 89–91, 92, 171, 191 Rio Grande Rift, 118
Phronima, 220 RMT (Rectangular Midwater Trawl), 285
ocean circulation, 14, 18–19, 23, 41, 51–54, phyla distribution, 229–230 Rockall Plateau, 123, 124
231 phytodetritus, 102–103, 104, 105, 199, 200, Romanche Trench, 343
ocean conveyor belt, 47 202 Rose Garden Vent, 158
ocean eddies, 51, 53, 59–68, 82 phytoplankton, 20, 24, 79, 96–97, 98–99, 104, Ross, Sir James Clark, 14, 17
ocean microstructure, 290 110, 111, 168, 180, 230 Rossby radius, 67
ocean models, 42, 47, 51, 52–53, 65, 68, 88 phytoplankton blooms, 54, 65, 79, 89–94, 102, ROV (Remotely Operated Vehicles), 274, 299
oceanic, 110, 198, 238, 254–257 Royal Society, 10–11, 18
ridges, 113, 114, 116, 155 Pillsbury, 20 rubidium, 175
trenches, 113, 114 pinger, 308, 312
oceanic crust, 152, 170 Pittenweem Harbour, 315 Sagitta setosa, 256, 257
octocorals, 198 pixels, 27, 74, 78, 80, 88 Sahara Desert, 146
octopus, 225, 272 placers, 334 Saharan debris flow, 137, 139, 141–142
ODP (Ocean Drilling Program), 124, 161 plaice, 197, 322 salinity, 12–14, 24, 165, 174, 246–247
Odyssey vehicle, 299 plankton, 97, 255 surface, 30, 44–45
oil, 121, 131, 132–133, 136, 330–332 see also phytoplankton, meso-zooplankton, salmon, 340
oil production platforms, 259, 331–332 microplankton, zooplankton salps, 96
Oncaea, 111 plate tectonics, 113–116, 119, 152, 153 salt, 323
Oneirophanta mutabilis, 203 Plesiopenaeus, 222 salt marshes, 244
oozes, Plesiopenaeus armatus, 204, 205 Salter’s ducks, 328, 329
calcareous, 20, 129 Pleurobrachia pileus, 256, 257 sampling bottles, 99
siliceous, 129–131 pollution, 205, 242, 338–341 sand eels, 321
ophiolites, 126 Polygon experiment, 61 sandwaves, 80
Opisthoproctus, 219 Polymode experiment, 62, 65 Santa Barbara Basin, 318
orbits, Poole Bay, 263, 266–267 SAR, Synthetic Aperture Radar, 74, 78, 80–81
geostationary, 73, 74, 75 Porcupine, HMS, 18, 195 sardines, 260, 317–318
near polar, 72, 73 Porcupine Abyssal Plain, 108, 195–196, 200, Sardinia, 259, 260
sun-synchronous, 73 202–204, 343 Sargasso Sea, 109
Ordnance datum, 306 Porcupine Bank, 199, 200 satellites, 50, 70–88, 280, 303
OSCR (Ocean Surface Current Radar), 297 Porcupine Seabight, 104 , 195, 197, 200, scallops, 197
Oslo–Paris Convention, 272 239–240 scatterometer, 78, 79, 86–87
osmoregulation, 246–249 pore waters, 131, 173, 332 scavengers, 204, 205, 344
ostracods, 225, 238 porosity, 107, 131 scavenging, 173, 175, 176
OTEC (Ocean Thermal Energy Conversion), Porphyrocrinus thallassae, 200, 201 Scripps Institute of Oceanography, 17, 22, 24,
329, 330 porphyropsin, 218, 219 281, 292
oxygen, Porpita, 214 Scyllarides latus, 265
depletion, 90–91, 93–94, 107, 131, 256, 340 Portugal, 81 SDT (Salinity, Temperature, Depth)
dissolved, 343, 344 Posidonia, 268, 269 instruments, 283
isotopes, 41, 56, 118, 132, 133, 146–147 position-fixing systems, 302–304 sea bed surveying systems, 80, 308
production, 171 pouting 260, 263, 263 sea cucumbers, 105, 199, 202–203, 207, 212,
ozone, 182 precipitation, 9, 29, 30, 32, 37–40 239, 241
primary productivity, 92, 107, 233, 245, sea floor spreading, 113, 117, 118–120, 123,
Pacific–Antarctic Ridge, 115 253–254 124, 152
Pacific–Indian throughflow, see Indonesian primum mobile, 9 sea grass, 245, 268, 269
Throughflow Princesse Alice I, 21 sea level, 45, 48, 56, 128, 132, 145, 146, 150,
Pacific Ocean, see also Eastern Pacific; North Procerodes (Gunda) ulvae, 247 182
Pacific Ocean; 9, 54–56, 120, 133 prokaryotes, 228, 230 sea pens, 197, 198, 224–225
Paelopatides gigantea, 199 protozoa, 90, 107, 202 sea stars, 200, 201, 208
Palaemonetes varians, 245 Psychropotes longicauda, 203 sea surface,
Panama, Isthmus of, 231 pteropod ooze, 20 colour, 71, 78
Panama Seaway, 133 pteropods, 96 roughness, 71, 78
Pangaea, 118, 228, 231 Pygospio elegans, 252 slope 71, 78
Panulirus argus, 264 temperature maps, 76, 78
Paragnathia formica, 245, 249 quadrats, 274, 275 temperatures (SST), 14, 37, 50, 52, 54, 56,
paramuriceidae, 200, 201 59, 63, 71,76, 286, 319
Paroriza pallens, 203 radar, 74, 78, 80–81 topography, 82, 83
Particle Conspiracy, 173–174 radiation balance, 28–29, 30, 31, 42 sea urchins, 209, 267, 321
particles, 96–97, 98–99, 100, 103, 110 radioactive waste, 165, 339, 342, 345 sea water, composition, 165–181
particulate flux, 108–110, 129, 171 radiolarian, 172 Seabeam, 310
particulates, river-borne, 169, 170, 174 radiometric dating, 118 SeaMARC I and II, 311
PCBs (polychlorobenzenes), 339, 345 RAFOS float, 292 seamounts, 120, 145
pelagic clays, 20, 129–130, 131, 140, 161 Rainbow hydrothermal field, 162 Seasat, 80
pelagic taxa, 232, 234, 239 Rance barrage, 326, 327 SeaSoar, 64, 92, 285
Pendleton artificial reef, 262 rat-tail fish 205, 217, 321 seasonal variation, 237, 241, 254

351
Oceanography

seaweeds, 96, 316, 321, 324 Storegga slide/debris flow, 137, 143 turbidites, 130, 137, 143, 144–147
sediment, Strass, V., 92 turbidity, 94, 99, 130
cores, 46, 56 stringer, 107 turbidity currents, 136–138, 141, 143–147
fans, 147 Stylocheiron suhmii, 220 turbot, 322
slides, 136–143 Stylocidaris lineata, 209
transport, 136–150 Styracaster, 204 ultramafics, 124
traps, 103, 106, 108, 109, 110 sub-bottom profiling, 311, 312, 313 ultraviolet absorption, 297
sedimentary cycle, 169–170, 171 sub-meso-scale coherent vortices (SCVs), 65, Umbellula, 224
sedimentation, 128–130, 151 66, 67 UML (Upper Mixed Layer), 97–100, 110–111
sediments, 96, 102, 103–104, 118, 124, 128 subduction, 113, 120, 152, 170 Umnak turbidite system, 150
Segonzacia, 206 submarine, UOR (Undulating Oceanographic Recorder),
seismic profiling, 136 cables, 16, 22 285
seismic reflection, 127, 128 canyons, 147, 148 upwelling, 64, 89, 176, 184, 233, 237, 245
seismic velocity, 24, 116, 124, 125–126 fans, 147–151 coastal, 54, 76–77, 87, 172, 184, 233, 237,
seismographs, 288 volcanoes, 300, 309 318, 334
Selective Availability, 303 submersibles, 298 uranium, 110, 324
Self-Contained Underwater Breathing Suez Canal, 233, 265 urine, 249–250
Apparatus (SCUBA), 262, 273–274 sulphates, 156
sensors, active, 70–71 sulphides, 131, 156, 157, 160, 336 Vargula, 225
Sergestes, 217 sulphur dioxide, 94 Vema Fracture Zone, 343
Severn barrage, 327 Super Ridge, 271 vent fields 157, 159
sewage sludge, 241, 338-339, 340–341, 343 surface temperature anomalies, 319 vent shrimps, 206, 222
sextants, 300, 303 surveying, 301–313 vertical migration, 217, 233
sexual light signals, 225 suspension feeders, 198, 250 Viking Graben, 331
shellfish poisoning, 93 Sverdrup, H.U., 91 visual acuity, 219–221
shingle mobility, 293 Swallow, Dr John, 60, 61, 291 visual pigments, 218–220
shrimp, 204, 206, 214, 216–217, 218, 219, swath sounding, 306, 310–311 volcanism, 120, 123, 132, 140, 152–153,
222, 225 Systellaspis debilis, 218, 219, 222 154–155
side-scan sonar, 307, 307–310 vortex shedding, 162
silica, 156, 160 TAG hydrothermal mound, 159, 160 vorticity, 37–38, 68
silicate, 90, 92, 123, 130 Taiwan, 270
silicoflagellate, 172 TAMI (Tethered Autonomous Microstructure Wadden Sea, 94
silicon, 165, 175, 177 Instrument), 291–292 Waitz, J.S. von, 12, 14
silvering, 215–216 tapetum, 222, 224 Wallace’s Line, 232
Simrad, E.M., 12, 310 TBT (tributyl tin), 325 Walvis Ridge, 19, 343
siphonophores, 212 tectonic plates, 113–116, 119, 152, 153 wastes, 268, 338–345
Skeletonema costatum, 92 temperatures, Watasenia scintillans, 226
slumps, 140 bottom, 19, 46 water,
SMBA (Scottish Marine Biological Association) sea surface (SST), 14, 37, 50, 52, 54, 56, 59, balance, 28, 29–31, 32
multiple corer, 103, 104 63, 71, 76, 286, 319 cycle, 28, 29, 52
smeddies, see sub-meso-scale coherent vortices TEP (transparent exopolymer particles), 100, masses, 44, 59, 233
(SCVs) 101, 107 sampler, 11
snails, 206 terrestrial radiation, 27 vapour, 28, 30–31, 54, 132
snatcher, 106, 108 Test, River, 244, 247 wave,
Snell’s window, 212, 306, 312 Tethys Ocean, 118, 120 energy, 328, 329
sodium chloride, 323 thermal energy, 329–330 heights, 25, 78, 84, 85
SOFAR (Sound Fixing And Ranging), 61, 62, thermistor chain, 286 recorders, 25, 84
66, 292 thermocline, 54, 94, 101, 233, 237, 275–276 Wavecrest buoy, 288
Sohm Abyssal Plain, 138, 145 thermohaline circulation, 27, 39, 42, 43, 46–47 weather prediction, 51
solar radiation, 27–28, 56, 182, 183 thermometers, 14, 282–283 weathering, 169–170
sole, 197 Thomson, C. Wyville, 18, 96 Weber’s Line, 232
Somali Current, 60 thorium, 107, 110 Weddell Sea Bottom Water, 36, 343
sonographs, 309, 311 Thouarella sp., 200 West Wind Drift, see Antarctic Circumpolar
sonar, 24, 294–296 tidal, Current
Sonne, RV, 335 correction, 306 Western Mediterranean Deep water, 48
sound, energy, 326–327 whaling, 319, 320–321
low frequency, 50, 295 mills, 326 whelks, 268
speed in rock, 313 tide gauge, 16, 56 white smokers, 157, 159, 160
speed in water, 50, 296, 305–306, 313 tides, 10–11, 16, 82 whiting, 197
sounding machine, 11–13 tintiniid lorica, 108 Wight, Isle of 81
South Atlantic Ocean, 82, 83, 84, 167 Titanic, 299, 305 Wilson cycle, 118, 119, 120
South Pacific Convergence Zone, 30 TOBI (Towed Ocean Bottom Instrument), 136, wind charts, 15
South Pole, 183 138–140, 142, 294, 311 wind-driven circulation, 26, 35, 52
Southampton Water, 91–92, 254, 255, tomography, 295 wind energy, 325
256–257 Tomopteris, 223 wind velocity, 15, 86–87
Southern Ocean, 36, 53, 63, 84, 85, 87, 115, Tongue of the Ocean, 196, 207–210 Woods Hole Oceanographic Institution, 17, 22,
155 TOPEX/POSEIDON, 63, 84, 85 61, 290, 292, 298–299, 329
species-richness, 234, 238–240 towed sensors, 284–286 World Ocean Circulation Experiment (WOCE),
spectral distribution, 79 toxins, 93, 324, 340 25, 31, 42, 48, 49, 84
Spitzbergen, 241, 243, 320 trace metals, 165, 166 worms, 202, 223, 225, 248, 251, 252
sponges, 198, 200, 217, 259, 267 trade winds, 34, 54, 55 wrasse, corkwing, 264, 267–268, 322
spring blooms, 91–93, 254–255 Trans-Atlantic Geotraverse, 159 Wüst, Georg, 23–24
squid, 25, 212, 216, 223–224, 226, 316, 321 transparency, 214–215
squirrel fish, 265 Trisopterus luscus, 263 yellow substance, 79, 213
STABLE (Sediment Transport and Boundary tritium, 167 Younger Dryas, 46–47
Layer Equipment), 289 Tropical Ocean Global Atmosphere
starfish, 197, 204, 208 experiment, 55, 56 zinc, 160, 161, 336
Stazione Zoologica, 17, 20 tsunamis, 136, 141, 143 zooplankton, 24, 100–101, 110, 253, 255–257
Steinahöll, 162, 163 tube worms, 158
Stommel, H., 61 Tugurium caribaeum, 209

352