Arboricultural Journal: The International Journal of

Urban Forestry

ISSN: 0307-1375 (Print) 2168-1074 (Online) Journal homepage: https://tandfonline.com/loi/tarb20

The recognition of functional units as an aid to

tree management, with particular reference to
veteran trees

David Lonsdale

To cite this article: David Lonsdale (2013) The recognition of functional units as an aid to tree
management, with particular reference to veteran trees, Arboricultural Journal: The International
Journal of Urban Forestry, 35:4, 188-201, DOI: 10.1080/03071375.2013.883214

To link to this article: https://doi.org/10.1080/03071375.2013.883214

Published online: 13 Feb 2014.

Submit your article to this journal

Article views: 285

Citing articles: 5 View citing articles

Full Terms & Conditions of access and use can be found at

https://tandfonline.com/action/journalInformation?journalCode=tarb20
Arboricultural Journal: The International Journal of Urban Forestry, 2013
Vol. 35, No. 4, 188–201, http://dx.doi.org/10.1080/03071375.2013.883214

The recognition of functional units as an aid to tree management,

with particular reference to veteran trees
David Lonsdale*

Independent Consultant and Lecturer in Arboriculture, Tree Biology and Tree Pathology, 33 Kings
Road, Alton, Hampshire GU34 1PX, UK

Guidance on the management of trees, especially pruning, is reviewed in relation to the

tendency for trees, especially in later life, to differentiate into semi-autonomous
“functional units”, represented by particular branches or stems. The recognition of such
units provides a basis for assessing the tolerance of particular trees, especially veterans,
to various degrees of pruning. The results are likely to be better than can be achieved by
applying whole-tree criteria for the total amount of leaf area that can “safely” be
removed.
Keywords: functional units; dysfunction; function; hydraulic segmentation; water-
conductivity; xylem

Introduction
While generally subscribing to the principle that trees should not be pruned too severely,
practitioners express a wide range of opinions over the definition of “severe”, with regard
both to the proportion of leaf area removed and the size and number of the wounds created.
The potentially harmful effects of severe pruning are both predictable in theory and often
observed in practice. Yet, critics of an arguably over-cautious approach can point out that
many trees have readily tolerated the loss of a large proportion of their overall leaf area,
either through pruning or major mechanical failure.
The relationship between the overall proportion of leaf area removed and the outcome
for tree vitality and vigour cannot easily be predicted, even taking account of differences
in the age or species of the trees concerned or of other factors such as growing conditions.
In the present article, however, I argue that the responses of trees can be better understood
by taking account of the physiological requirements of particular branches, which,
together with their associated sectors of the root system, can be recognised as semi-
autonomous units.
For the purpose of this article, I describe the above semi-autonomous units as
“functional units” (Lonsdale, 2013). I do so tentatively, however, with the intention of
encouraging discussion about other terms that might be preferable both scientifically and
from a tree manager’s perspective. Another possible term is “morphological unit” but this
does not convey the importance of protecting the viability of the three components shown in
Figure 1 (foliage, stem and roots). However, the term “functional unit” could be confusing
because it has previously been used to describe a wedge-like zone of conductive cells that is
connected only by limited pathways of water flow to the surrounding xylem (Canny, 1991).
Another term “viable unit” could also be considered but it has the disadvantage of being
applicable to almost anything that is viable.

*Email: [email protected]

q 2014 Taylor & Francis and Aboricultural Association

 Arboricultural Journal: The International Journal of Urban Forestry 189

COMPONENTS OF A
“FUNCTIONAL UNIT”

Foliage-bearing region: of an entire tree

(crown) or an “autonomous” branch.
Main physiological function:
• Production and storage of sugar
and starch for energy and growth

Stem (or “autonomous branch”)

Main physiological functions:
• Conduction of water from roots to
foliage and growing shoots
• Conduction of sugar from foliage
towards root system
• Seasonal storage of energy
reserves (sugar and starch)

Root system (including mycorrhizas)

Main physiological functions:
• Absorption of water and mineral
nutrients
• Seasonal storage of energy
reserves

Figure 1. In the present context, a “functional unit” consists of the three components shown here.
Thus, the unit could be an entire tree or it could be part of a tree as illustrated in the example shown
in Figure 2, where the tree consists of discrete ramets, each comprising foliage, roots and an
interconnecting stem (hence the term “autonomous branch” in the text above).

The existence of functional units in trees, as defined here, is especially obvious in

multiple-stemmed trees like the one shown in Figure 2. A similar division into functional
units can be recognised in old pollards, which have often survived for centuries despite the
loss of particular “units”. Non-pollarded trees (“maidens”) of the same species appear
generally not to live so long, partly because the failure of a single stem is more likely to
lead to death. This does not necessarily mean that, like a very young tree, a mature maiden
tree consists only of a single recognisable functional unit.
The division of trees into functional (or morphological) units was illustrated in a model
depicting tree development from seedling to “late ancient” by Raimbault (1995, 2006).
The model defines morphological changes including the increasing tendency for branches
to become semi-autonomous, together with the hollowing of the older, central wood.
In old trees, the model shows the occurrence of “reiterative growth”, whereby young
branches sometimes develop on portions of the tree that have shed primary branches or
twigs during unfavourable conditions. In their early stages of growth, the new branches
resemble the juvenile growth of a young tree.
One of the purposes of Raimbault’s model was to help practitioners assess the
suitability of trees for particular kinds of management. In the present article, I attempt to
develop this idea further by considering functional units in relation to the management of
190 D. Lonsdale

trees, especially veteran trees. First, I will review existing guidance on pruning, in relation
to the following:
(a) Leaf area and energy reserves,
(b) The potential for shoot growth after pruning,
(c) Cross-sectional exposure of wood resulting from severance of twigs, branches or
stems.

Leaf area and energy reserves

Shigo (1986, 1991) emphasised the importance of energy reserves (in the form of sugar
and starch) for the health and survival of trees. These reserves are thought to be needed not
only for maintaining growth but also for resisting disease and controlling the extent of
decay. Accordingly, Shigo (1991) criticised severe forms of pruning (especially topping),
which result in the loss of a relatively large proportion of the leaf area, in which sugar and
starch are produced by photosynthesis.
Shigo (1985) explained also that the defensive reactions of sapwood (i.e.
“compartmentalisation”) to harsh pruning can impair a tree’s capacity to store and
translocate carbohydrate reserves. In addition, Clair-Maczulajtys, Le Disquet, and Bory
(1999) later showed that relatively harsh pruning can strongly affect the pattern of starch
accumulation in different parts of the stem and branches, thus causing a disproportionate
depletion of these reserves in certain zones and disrupting the normal processes of
resource-partitioning throughout the tree. If, however, only young branches are
periodically cut from long-established pollard heads, the defensive systems of the tree
can be maintained very strongly (Ferrini, 2006). Shigo (1985) argued that trees could even
die because of wound-induced energy depletion, although this argument was criticised
with regard to trees affected by vascular wilt diseases such as Dutch elm disease, in which
decline and death can be explained by loss of water conduction, irrespective of energy
relations. Nevertheless, it seems reasonable to infer that pruning-induced loss of
physiological function in a large volume of sapwood can lead to a serious loss of energy
storage capacity.

Potential for shoot growth after pruning

In order to survive, each branch of a tree (or the entire tree in the case of an unbranched
species) must consist of the three main components shown in Figure 2, i.e. foliage, roots
and an interconnecting section of conductive stem or branch. Since pruning involves the
removal of foliage, it will eventually result in some degree of dieback of the pruned part as
far as the nearest point that still forms part of a fully functional unit. Such dieback, which
typically extends as far as a morphological boundary such as a branch junction, is familiar
to everyone who works with trees and was described many years ago in fruit trees by
Swarbrick (1926). Owing to the possibility of dieback, amongst other considerations, most
pruning guidance does not advocate the retention of stubs. However, stubs do not always
die back completely, since they can remain partly functional by forming new shoots from
dormant or adventitious buds.
Clearly, plans for pruning should take account not only of the removal of leaf area
(i.e. photosynthetic capacity), but also the potential dieback of branches (recognisable
as “functional units”) that fail to re-foliate enough to sustain their survival. Thus, the
potential for re-foliation should be assessed, especially in old trees, where viable
dormant buds are sometimes confined mostly to the periphery of the crown.
 Arboricultural Journal: The International Journal of Urban Forestry 191

Figure 2. A multi-stemmed tree, in which each of the stems, together with an associated sector of
the root system, can be regarded as a semi-autonomous functional unit (photo: Chris Knapman).

The distinction between removal of leaf area per se and the removal of potential leaf-
bearing twig structure can be appreciated by imagining a hypothetical form of “pruning”,
in which every leaf is individually picked from a tree without a single twig being removed.
This procedure would remove most of the short-term photosynthetic capacity of the tree,
with the only green tissue remaining in the unlignified shoots and the outer layers of the
primary bark. However, the intact twigs would provide a structure for re-foliation during
the next period of growth if the tree had the energy reserves required. Similarly, very light
pruning that preserves much of the existing twig structure can be regarded as less harmful
than the removal of most of the twig structure.
Although very light pruning seems less inherently harmful than a more significant
reduction in crown-size, a somewhat greater reduction in crown-size could be regarded as
preferable if this would lessen the energy demand during re-foliation. Such demand is
thought to be a major factor in the decline of trees affected repeatedly by defoliation by
pathogens or insect herbivores. In an experimental study, Wargo and Houston (1974)
found that manual leaf removal (for simulating the effects of insect defoliation) reduced
the resistance of sugar maple, Acer saccharum, to root colonisation by honey fungus,
Armillaria mellea. Furthermore, in evergreen trees, especially certain conifers, which
store a large proportion of their carbohydrate reserves in foliage, defoliation can lead to
severe decline and sometimes death.
192 D. Lonsdale

Figure 3. A tree with a well-formed lower crown, which would probably be self-sustaining in the
event of the upper crown being lost or removed by pruning (photo: Neville Fay).

The scope for reducing the size of the crown while not causing undue harm is perhaps
most apparent where there is already a “secondary”, well foliated, inner crown, which can
be preserved during pruning. The upper and lower portions of the crown of a tree like the
one in Figure 3 can be regarded as being morphologically distinct from each other to some
degree. They are not, however, semi-autonomous functional units of the kind shown in
Figures 1 and 2, each consisting of all three essential components, i.e. foliage, stem and
roots. In other words, the division between an outer and inner crown is axial rather than
lateral. Studies of functional divisions within trees have been largely concerned with the
junctions between axially sequential “segments”: e.g. leaves, twigs and branches, as
mentioned below.
Even if pruning entails the removal of a large proportion of the existing leaf area, this
is not necessarily very harmful if the tree has the potential to become densely re-foliated
soon after pruning. This potential might take the form of existing shoots that will develop
in response to pruning or of dormant or adventitious buds that can develop into new shoots.
Thus, assessment of the potential to retain a well-foliated crown soon after pruning is one
of the key considerations identified in British Standard BS 3998:2010 (Anonymous, 2010)
rather than merely taking account of the leaf area removed in pruning. This potential is
influenced largely by the species (or variety) of the tree concerned and by the age of the
bark at the cutting points. The latter factor is especially important in the pruning of veteran
trees (Lonsdale, 2013; Read, 2000).
 Arboricultural Journal: The International Journal of Urban Forestry 193

For the many veteran trees that have an apparently limited capacity to produce
vigorous shoots in response to pruning, the current recommendation is that pruning, if
required for biomechanical reasons, should be phased according to the process of
“retrenchment pruning” (Fay, 2008; Lonsdale, 2013). The specifications for each new
phase should take account of responses of the tree to the previous phases, both as a whole
and within each of its identifiable functional units.
In order to survive in the long term, the new branches that develop after pruning must
eventually establish vascular connections with the root system via their parent branches
and/or the main stem. This will normally happen within a single growing season but there
are instances where this appears not to have happened in severely pruned trees that have
lacked the photosynthetic capacity to fuel the growth of new sapwood all the way down
from the new branches to the roots (e.g. as in Figure 4). The new branches might survive
for several years, obtaining water via the pre-existing sapwood but this older wood
can eventually become non-conductive, especially because of fungal colonisation
(see Figure 5).

Figure 4. An ancient, lapsed hornbeam pollard, which initially survived severe pruning. The water
supply for the new shoots probably depended on the conductivity of pre-existing sapwood, which
was eventually colonised by a “sap-rot” fungus.
194 D. Lonsdale

Dead, pre-existing wood

Live, pre-existing wood
New, live wood
New wood, now dead

Figure 5. Different outcomes from total re-cutting of lapsed pollards: left, showing intact channels
of new xylem and all new branches alive; right, showing dieback of branches associated with xylem
dysfunction (adapted from Lonsdale, 1999). In both cases, total re-cutting would have been
inadvisable but the resulting extent of dysfunction will vary according to the condition of the tree
(and its constituent functional units) before cutting, together with the growing conditions, including
the weather during and after cutting.

Cross-sectional exposure of wood

Shigo and Marx (1977) developed the concept of “Compartmentalisation of Decay in
Trees” (CODIT), which describes the confinement of wood decay within discrete columns
of a particular size and shape as influenced by natural boundaries, such as those that occur
between annual increments and between the radially and axially orientated xylem tissues.
The original concept of CODIT was eventually modified through the realisation that
compartmentalisation is a general form of defence against the spread of injury-induced
physiological dysfunction (especially loss of water-conduction) in sapwood rather than
being a specific defence against decay (Boddy & Rayner, 1983; Liese & Dujesiefken,
1988; Pearce, 1996). In addition, Boddy and Rayner (1983) emphasised the role of the
high moisture content of sapwood as a form of passive defence alongside active defences
that can develop in response to injury.
The original CODIT model was presented as a modern replacement for an allegedly
outdated heartrot concept, according to which decay began in exposed areas of heartwood
and then spread without any active defensive response on the part of the tree. Additionally,
Shigo and Shortle (1979) suggested that heartwood, rather like sapwood, can respond to
wounding or to fungal ingress by forming defensive barriers (albeit by a chemical reaction
only because there are practically no living cells in heartwood). The concept of CODIT
does not, however, invalidate the observation that heartrot is initiated in wounds that
expose older central wood. The rate of development of heartrot is of key importance in
determining the future mechanical integrity (and perhaps the health) of the tree.
Another matter to consider with regard to the original CODIT model (Shigo & Marx,
1977) is that it depicts the dysfunctional columns of wood as being mostly internal, except
 Arboricultural Journal: The International Journal of Urban Forestry 195

Heartwood Sapwood Ripewood

(durable in some species) (has active defences) (non durable)

Figure 6. Comparison between tree species with true heartwood (left) and with a ripewood core
(right). High moisture content acts as a passive defence whether the wood is living, non-living,
durable or non-durable. Wounding usually causes moisture loss and so impairs this defence.

where they include a dead stub. In old trees, however, dysfunction and decay often extend
to the surface of the main stem, forming columns that define the lateral boundaries of
adjacent living functional units.
In many tree species, the central wood of older branches and stems is durable
heartwood, in which decay is slowed by the presence of antimicrobial substances known in
the timber industry as extractives. In many other species, the central wood is either non-
durable heartwood or “ripewood”, the latter being sapwood that has gradually lost
physiological function over many years and has become non-living (Figure 6). Non-
durable heartwood and ripewood can remain free from decay over centuries if, owing to
high moisture content, there is insufficient aeration for fungal decay to occur. However,
such wood can undergo rapid decay when it becomes exposed to the atmosphere, thus
undergoing partial drying and aeration (Lonsdale, 1999). Exposure of a large proportion of
the cross-section in a tree that consists largely of ripewood or non-durable heartwood can
lead to very extensive decay, which in turn can shorten the life of the tree through
catastrophic mechanical failure or ingress of the decay into sapwood.

The recognition of functional units as an aid to understanding responses to pruning

To some extent, the response of a tree to pruning can be predicted by taking account of its
age, its species, the growing conditions and the intended intensity and pattern of pruning.
The latter has implications for the effects outlined above, i.e. (a) removal of leaf area, (b)
the potential for re-foliation after pruning and (c) the implications of exposing sapwood,
durable heartwood or non-durable central wood.
All the above factors can be usefully taken into account on a whole-tree basis but there
is also a need to recognise that individual branches and roots of a tree have a certain degree
of autonomy, owing to partial separation between their pathways of water-conduction. As
mentioned above, such separation is relatively obvious in a multi-stemmed tree or a multi-
branched pollard. It is perhaps even more obvious where contact between a low branch and
the soil has led to natural layering. In such cases, or where the entire tree, having fallen,
has undergone “phoenix regeneration” (Lonsdale, 2013), the newly rooted portion can
eventually become completely autonomous.
A branch or a stem that is partly autonomous can be regarded as a component of a
functional unit within the tree. Each such unit comprises foliage, a sector of the root
system and an interconnecting length of stem and/or branch (see Table 1). Such units are
not obviously present in a relatively young, single-stemmed tree that has not been
196 D. Lonsdale

Table 1. Components of a structural (functional) unit within the “architecture” of a tree.

Main physiological functions Possible requirements

Component of the component for management
Foliage-bearing region, i.e. † Production and storage of sugar † Control of sunlight
either the crown of an entire and starch for energy and growth and shade
tree or twigs borne on a partly
“autonomous” stem or branch
Stem (or the main axis of an † Conduction of water from roots † Protection against
“autonomous” branch) to foliage and growing shoots mechanical failure
† Conduction of sugar from foliage
towards root system
† Seasonal storage of energy reserves
(sugar and starch)
Root system (including † Absorption of water and mineral † Provision of suitable
mycorrhizal associations) nutrients soil conditions
† Seasonal storage of energy † Protection against
reserves uprooting

pollarded but all trees contain anatomical boundaries, which confer a certain degree of
autonomy on their component parts. This can be regarded as an important survival strategy
because it allows certain parts to become dysfunctional through ageing or damage while
adjacent parts are protected, by virtue of boundaries, from the spread of dysfunction.
In arboriculture, perhaps the most familiar concept of compartmented anatomy is the
above-mentioned CODIT model of Shigo and Marx (1977), in which decay (or
dysfunction) is often confined within pre-existing anatomical boundaries that have been
reinforced as barriers by virtue of active injury-induced responses. Within fully functional
sapwood, the pre-existing anatomical boundaries can be envisaged as providing some
degree of separation between the conductive pathways of adjacent functional units, such as
semi-autonomous branches.
It has long been recognised that individual branches of trees have a certain degree of
autonomy; i.e. while usually depending on the parent stem for mechanical support and for
a water supply, they “pay their own way” in the form of sugar and starch, produced by
photosynthesis. In this respect, they can be compared with hemi-parasitic plants such as
mistletoes, even though the comparison partly falls with regard to the translocation of
sugar.
The success of trees as large, long-lived organisms probably depends on the partially
compartmented anatomy of their stems. While providing a framework for walling-off
zones of dysfunction, this also provides enough lateral connections to allow fluid to move
in new pathways when a main pathway of fluid conduction has been damaged. This
concept of adaptability is borne out by research that has explored patterns of water flow
around obstructions such as saw cuts in stems (e.g. Preston, 1952).
In the context of pathways of water-conduction, palm trees are especially interesting,
even though their physiology is not strictly relevant to the subject of pruning. The stems of
palms develop only by primary growth; i.e. they do not lay down new wood by secondary
growth. Their water-conducting vessels lie within discrete vascular bundles, which
develop in association with particular fronds. The survival of the palm tree depends on the
continued water-conducting capacity of the first-formed vessels, long after the death of
their associated fronds. Zimmermann (1978) demonstrated the importance of connections
between the vascular bundles (Figure 7), as shown in earlier studies (Zimmermann &
 Arboricultural Journal: The International Journal of Urban Forestry 197

Figure 7. (from Zimmermann & Tomlinson, 1965): diagrammatic representation of vascular

bundles in the palm Rhapis excelsa, shown at the junction of a frond (leaf) and the stem. The major
bundle entering the bottom-right of the diagram serves the frond, together with the minor bundle on
the left and also bears branches (shaded black), which directly serve the stem or connect via bridges
to other vertical bundles (shown white), which belong to other fronds. After the eventual death of the
frond, the vertical bundles and the bridges remain conductive (the diagram is not to scale and shows a
horizontal exaggeration of about four times); copied with permission from Harvard College.

Tomlinson, 1965). These connections enable the water conducting pathways to persist
within the older part of the stem where the fronds have died. Even if injury occurs, the
connections enable water conduction to continue. Extensive colonisation by decay fungi
such as Ganoderma species is, however, often sufficient to cause death by destroying the
vascular pathways (Flood, Bridge, & Holderness, 2000).
Trees other than palms also depend for their long-term survival on the potential for
water to flow laterally between the main axial pathways of flow. Lateral flow is especially
important when injury has interrupted a main pathway, especially an injury that has
penetrated a main branch or stem. Because pruning involves the severance of the distal
ends of the main pathways, the subsequent survival of the tree depends partly on lateral
flow, which enables new shoots to develop and eventually establish new vascular
connections with the root system. Lateral flow probably also helps to meet a potentially
increased water-demand of the foliage that remains after pruning.
As well as allowing lateral flow, the water-conducting system includes many natural
points of constriction (Figure 8), which help to prevent the entry of gas bubbles (or
vacuum bubbles under strong negative pressure) into the main conductive pathways. In a
review of various studies, Tyree and Ewers (1991) concluded that these constrictions are
partly the result of differences in the water conductivity in the nodal and internodal regions
of twigs or similarly between branch junctions and the inter-branch regions of secondarily
thickened stems. Abscission zones can act as additional hydraulic bottlenecks (Rust,
Solger, & Roloff, 2004). Low conductivity in such zones can be attributed to factors such
as the length and diameter of the conductive cells together with the relative proportion
of such cells in the woody cross-section. In particular, in accordance with the
Hagen –Poiseuille equation, the lumen conductivity of a vessel increases with the fourth
198 D. Lonsdale

1a

2a 3b

3a

1b
2b 3c 4

Figure 8. Notional categories of vascular constriction at various morphological junctions in a

veteran tree: 1a, leaf/shoot (abscission zone); 1b, non-woody/woody root; 2a, twig/branch; 2b, root/
higher-order root; 3a, branch/higher-order branch; 3b, co-dominant junction; 3c, stem-root; 4,
pollard branch/bolling or coppice stem/stool.

power of the lumen diameter (Sperry, Hacke, & Wheeler, 2005; Zwieniecki, Elcher, &
Olbrook, 2001).
The anatomical nature of the constrictions depends on the type(s) of conductive cells
and on the water-conducting systems in the tree species concerned. The main categories
are as follows:

. Ring porous dicotyledonous trees: large-diameter springwood vessels, with much

smaller latewood vessels; also with tracheids in certain species.
. Diffuse porous dicotyledonous trees: small to medium-sized vessels, with a
decreasing diameter gradation from springwood to latewood; also with tracheids in
certain species.
. Palm trees (monocotyledons): primary xylem only (no secondary thickening of the
stem), with vessels in vascular bundles.
. Conifers: conductive xylem cells are all tracheids (in most species).
 Arboricultural Journal: The International Journal of Urban Forestry 199

Vascular constrictions can be regarded as inherently dividing a tree of any age into
units or “segments”, according to the hydraulic segmentation hypothesis of Zimmermann
(1978, 1983).
In their review, Tyree and Ewers (1991) stated as follows “ . . . plant segmentation . . .
can be defined as any structural feature of a plant that confines cavitations to small, distal,
expendable organs in favour of larger organs representing years of growth and
carbohydrate investment”. Thus, segmentation helps to prevent the spread of dysfunction
from the parts of the tree (e.g. topmost twigs) that are most vulnerable to damage from
events such as drought. Also, under all weather conditions, the constrictions are thought to
regulate the flow of water within the tree. The constrictions are partly a function of
hydraulic resistance in different parts of the system as defined by hydraulic attributes such
as leaf specific conductivity (LSC) and the Huber value (HV). [The LSC is the ratio of
stem conductivity to leaf area, thus representing the capacity of a stem or segment of a
stem to supply water to its attached leaves. The HV is the ratio between the sapwood cross-
section and the leaf area distal to the part of the stem concerned (Tyree & Ewers, 1991)].
By regulating water flow, hydraulic segmentation is believed to play a part in the
autonomy of individual branches, at least in conifers and other gymnosperms. In their
review of hydraulic segmentation, Tyree and Ewers (1991) wrote as follows:
The effect of this pattern of leaf specific conductivity in a gymnosperm is to confine most of
the resistance to water flow to the last metre, or less, of pathway from the base of the tree to the
leaf-bearing branches. Consequently, most small branches can compete for water on a more-
or-less equal basis, i.e. each small branch behaves like an independent plant rooted in a
common, highly conductive bole.
Also in citing the work of Zimmermann (1983), Tyree and Ewers (1991) pointed out that
the long-term survival of a tree depends partly on its capacity to shed or to lose function in
small, short-lived parts such as leaves and twigs. Indeed, the shedding of redundant twigs or
small branches is a normal feature of crown development, occurring either through an active
process of cladoptosis, involving abscission, or through dieback where the twig or branch
lacked sufficient light or other vital resources. In many instances, trees can also tolerate the
loss of relatively small branches that are shed because of canker development or other
disease, if this does not occur extensively throughout the crown. The possible occurrence of
hazards from branch shedding (for example in the case of “Massaria disease of plane”,
caused by Splanchnonema platani) is of course another matter.
The shedding of twigs or small branches can be much better tolerated than the loss of
function in larger, high-investment parts (especially stems) but even a major loss of
function can sometimes eventually be overcome by the formation of new conductive
pathways when further radial increments are laid down. Tyree and Ewers (1991) also
pointed out, however, that the more distal and expendable parts are, in any case, the most
vulnerable to moisture stress and would tend to die back first anyway without any need for
a mechanism of hydraulic segmentation. They suggested that hydraulic segmentation is
undoubtedly important in certain situations, for example where a rapidly growing leading
stem might otherwise be more vulnerable than twigs that are attached to the main stem
closer to the ground.

Guidance on pruning, based on the recognition of functional units

Current guidance on pruning, as set out in BS 3998:2010, takes account of our knowledge
of the overall effects of the removal of leaf area and the exposure of wood at pruning cuts.
BS 3998 also includes a certain amount of guidance on the selective pruning of particular
200 D. Lonsdale

branches or of sectors of the crown. In any case, it is obvious that good practitioners have
long taken account of the potentially different responses of different parts of the same tree.
However, better recognition of functional units could help to reconcile some of the
apparently opposing views on “how much is too much?” Certain aspects of guidance that
could be developed on this basis are as follows:
. Take account of the estimated age of the bark at the intended point of cutting,
because this will influence the potential for new shoot growth, as well as the amount
of exposure of heartwood or ripewood. The age could be very different between one
branch (or co-dominant stem) and another, especially if the tree is old and has a
mixture of primary and secondary branches borne on the same parent stem.
. Try to assess the vigour or vitality of each branch or stem that is to be considered for
cutting in order to gain some idea of its individual tolerance to pruning. (This should
be in addition to an assessment of the tree as a whole; overall signs of good health
often indicate relatively good tolerance to pruning.)
. Assess the viability of the branch and twig structure that will remain after pruning
rather than using an arbitrary concept of the amount of pruning the tree overall can
tolerate. For example, a tree with an existing well-developed inner or lower crown
might tolerate much heavier pruning than one that would be left with a very small
potential leaf-bearing structure after pruning.
. When planning the selective pruning of particular branches or of a portion of the
crown, take account of the amount of sunlight that will reach each of the remaining
functional units within the tree. Too much could lead to desiccation and/or sun-
scorch; too little could lead to suppression of the new shoots.
. Take account of the size and shape of different functional units when trying to
improve the growing conditions of the tree, for example in relation to the removal of
excess shading by other trees.

Conclusions
It is relatively easy to recognise the differentiation of trees, especially older ones, into
“functional units”. By realising that particular units might respond differently to
environmental conditions or to pruning, it is possible to reconcile some of the opposing
arguments about the degree of pruning that trees can reasonably tolerate. It should also be
possible to achieve better standards of tree care than can be provided by making decisions
based only on “whole tree” criteria.

Notes on contributor
David Lonsdale is a consultant, author and educator, specialising in the biology, pathology and
mechanical integrity of trees. After graduating from the universities of Southampton and
Manchester, he worked for Forest Research for 26 years before going freelance in 2002. Since then,
he has participated in the revision of British Standards 5837 and 3998. He has published many
scientific articles and is the author or editor of several books, including “Principles of Tree Hazard
Assessment and Management” (1999) and “Ancient and Other Veteran Trees: Further Guidance on
Management” (2013). He also works on a voluntary basis for the conservation of invertebrates and
ancient trees.

References
Anonymous. (2010). Tree work – recommendations: British Standard 3998:2010. London: British
Standards Institution.
 Arboricultural Journal: The International Journal of Urban Forestry 201

Boddy, L., & Rayner, A. D. M. (1983). Origins of decay in living deciduous trees: The role of
moisture content and reappraisal of the expanded concept of tree decay. New Phytologist, 94,
623– 641.
Canny, M. J. (1991). The xylem wedge as a functional unit: Speculations on the consequences of
flow in leaky tubes. New Phytologist, 118, 367– 374.
Clair-Maczulajtys, D., Le Disquet, I., & Bory, G. (1999). Pruning stress: Changes in the tree
physiology and their effects on the tree health. Acta Horticulturae (ISHS), 496, 317– 324.
Fay, N. (2008). Retrenchment pruning guidance (update). Bristol: Treework Environmental
Practice. Retrieved from http://www.treeworks.co.uk/downloads
Ferrini, F. (2006). Pollarding and its effects on tree physiology: A look to mature and senescent tree
management in Italy. Proc. Premier colloque européen sur les trognes, Vendôme, October 26 –
28, 2006.
Flood, J., Bridge, P., & Holderness, P. D. (Eds.). (2000). Ganoderma diseases of perennial crops.
Wallingford: CABI.
Liese, W., & Dujesiefken, D. (1988). Wundreaktionen an Bäumen. In D. Dujesiefken (Ed.),
Wundbehandlung an Bäumen (pp. 22 – 32). Braunschweig: Thalacker.
Lonsdale, D. (1999). Principles of tree hazard assessment and management. Research for Amenity
Trees No. 7. London: The Stationery Office.
Lonsdale, D. (Ed.). (2013). Ancient and other veteran trees: Further guidance on management.
London: The Tree Council.
Pearce, R. B. (1996). Antimicrobial defences in the wood of living trees. Tansley Review No 87.
New Phytologist, 132, 203– 233.
Preston, R. D. (1952). Movement of water in higher plants. In A. Frey-Wyssling (Ed.), Deformation
and flow in biological systems (pp. 257– 321). Amsterdam: North Holland.
Raimbault, P. F. (1995). Physiological diagnosis. The tree in its various states: Diagnosis and
architectural training. Proceedings of second European congress of arboriculture, Versailles,
September 1995.
Raimbault, P. F. (2006). A basis for morpho-physiological tree assessment. Proceedings of seminar,
arboricultural association/treework environmental practice, 23rd and 24th, March 2006, Ashton
Court, Bristol.
Read, H. (2000). Veteran trees: A guide to good management. Peterborough: English Nature.
Retrieved from www.naturalengland.gov.uk
Rust, S., Solger, A., & Roloff, A. (2004). Bottlenecks to water transport in Quercus robur L.: The
abscission zone and its physiological consequences. Basic and Applied Ecology, 5, 293– 299.
Shigo, A. L. (1985). Wounded forests, starving trees. Journal of Forestry, 83, 668– 673.
Shigo, A. L. (1986). A new tree biology. Durham, NH: Shigo and Trees.
Shigo, A. L. (1991). Modern arboriculture. Durham, NH: Shigo and Trees.
Shigo, A. L., & Marx, H. G. (1977). Compartmentalization of decay in trees. Agriculture
Information Bulletin No 405. Washington, DC: USDA Forest Service.
Shigo, A. L., & Shortle, W. C. (1979). Compartmentalisation of discolored wood in heartwood of red
oak. Phytopathology, 69, 710– 711.
Sperry, J. S., Hacke, U. G., & Wheeler, J. W. W. (2005). Comparative analysis of end wall resistance
in xylem conduits. Plant, Cell & Environment, 28, 456– 564.
Swarbrick, T. (1926). The healing of wounds in woody stems. Journal of Pomology and
Horticultural Science, 5, 526– 534.
Tyree, M. T., & Ewers, F. W. (1991). The hydraulic architecture of trees and other woody plants.
New Phytologist, 119, 345– 360.
Wargo, P., & Houston, D. R. (1974). Infection of defoliated sugar maple trees by Armillaria mellea.
Phytopathology, 64, 817– 822.
Zimmermann, M. H. (1978). Vessel ends and the disruption of water flow in plants. Phytopathology,
68, 253– 255.
Zimmermann, M. H. (1983). Xylem structure and the ascent of sap. Berlin: Springer.
Zimmermann, M. H., & Tomlinson, P. B. (1965). Anatomy of the palm Rhapis excelsa. I: Mature
vegetative axis. Journal of the Arnold Arboretum, 46, 160– 179.
Zwieniecki, M., Elcher, P. M., & Olbrook, N. H. (2001). Hydraulic properties of individual xylem
vessels in Fraxinus americana. Journal of Experimental Botany, 52, 257– 264.