J. exp. Biol.

137, 513-527 (1988) 513

Printed in Great Britain © The Company of Biologists Limited 1988

INGESTIVE BEHAVIOUR AND PHYSIOLOGY OF THE

MEDICINAL LEECH

BY CHARLES M. LENT1, KARSTEN H. FLIEGNER*, EDWARD

FREEDMANt AND MICHAEL H. DICKINSON*
^Department of Biology, Utah State University, Logan UT 84322 and Division of
Biology and Medicine, Brown University, Providence RI 02912, USA

Accepted 16 February 1988

Summary
Ingestion lasts 25 min in Hirudo medicinalis and is characterized by pharyngeal
peristalsis which fills the crop. This peristalsis has an initial rate of 2-4 Hz which
decays smoothly to 1-2 Hz at termination of ingestion. During ingestion, the leech
body wall undergoes peristalsis which appears to aid in filling the crop diverticula.
Body peristalsis begins at a rate of 10 min"1 and decreases linearly to 2min~' at
termination. The body also undergoes dorsoventral flexions when blood flow is
occluded. Blood meal size increases slightly with leech size: 8-4 g for 1-g leeches
and 9-7 g for 2-g leeches. However, relative meal size decreases markedly with
increasing animal size; from 8-15 times body mass for 1-g to 4-80 times for 2-g
leeches. When intact leeches were exposed to micromolar concentrations of
serotonin, there was an increase in the rate of pharyngeal peristalsis and the size of
the blood meals.
Leeches excrete the plasma from their ingested blood meals. Excretion is
activated during ingestion, which increases feeding efficiency by increasing the
proportion of blood cells in the ingestate. Excretion continues for 4-6 days
following ingestion, removing all the remaining plasma from the ingestate. Leech
ingestion comprises stereotyped muscular movements, secretion of saliva and
excretion of plasma. A strikingly similar feeding physiology is seen in the blood-
sucking insect Rhodnius, and we suggest that efficient sanguivory may require the
convergent evolution of similar ingestive mechanisms.

Introduction
Ingestion is the consummatory phase of leech feeding behaviour. In Hirudo
medicinalis, it is initiated by both chemical (Elliot, 1986) and thermal stimuli
* Present address: School of Medicine, New York University, New York, NY 10016, USA.
t Present address: Department of Neurobiology and Behavior, Cornell University, Ithaca,
NY 14853, USA.
$ Present address: Department of Zoology, University of Washington, Seattle, WA 98195,
fcJSA.
Key words: behaviour, excretion, ingestion, leech, sanguivory, serotonin.
514 CM. LENT AND OTHERS

(Dickinson & Lent, 1984). A hungry leech attaches the anterior sucker to the host
and cuts the skin with three semicircular jaws. Blood flows into the buccal cavity
and is pumped into the crop by rhythmic pharyngeal peristalsis. Leech saliva
contains a powerful anticoagulant hirudin (Haycraft, 1884) and is secreted from
the cutting edges of the jaws (Marshall, 1984).
Hirudo are distended by meals of 7-9 times their pre-fed mass (Dickinson &
Lent, 1984), and the resulting distension of the body wall terminates ingestion and
inhibits biting (Lent & Dickinson, 1987). Clear fluid flows copiously from the skin
of leeches as they feed (Worth, 1951) and, after termination, excess water and salts
are excreted from the ingested blood via the paired, segmental nephridia (Zerbst-
Boroffka, 1973).
Leech ingestion is evoked by the activity of identified neurones (Lent &
Dickinson, 1984). These cells release serotonin, a neurotransmitter which affects
both behavioural and physiological components of ingestion. The practice of
leeching is now re-emerging in medicine (Lent, 1986), and here we describe the
stereotyped movements and glandular activities underlying ingestion by the
medicinal leech. An abstract of some of the research reported here has been
published (Lent, Fliegner & Freedman, 1986).

Materials and methods

Hirudo were obtained from European suppliers and maintained on a 12 h: 12 h
light: dark cycle at 15°C in glass aquaria filled with artificial pond water (Muller,
Nicholls & Stent, 1981). The leeches weighed between 0-9 and 2-7 g and were fed
at 37°C upon 'artificial blood' a 50vols% suspension of human erythrocytes in
Leibovitz L-15 culture medium (Microbiological Association, Bethesda, MD). In
most experiments, sodium heparin was added as an anticoagulant (0-2mgml~').
Blood was prepared volumetrically, warmed to 38°C and poured into 15 ml
polystyrene centrifuge tubes. The open end of the feeding tube was covered with a
single layer of Parafilm and the tapered end removed, a procedure which
facilitated flow and allowed for the addition of blood. In approximately half the
feeding trials, an additional volume of fresh blood at 38°C was added to the tube
after 10-15 min of feeding.

Behaviour
Hirudo were weighed individually and offered the film-covered end of the
feeding tube. Hungry leeches attached within a few seconds and their normal
exploration of the warm tube was restricted to the film surface by means of a sleeve
of Tygon tubing. Once a Hirudo had bitten the film and began to ingest blood, a
timer was started and the sleeve was removed. The feeding tube, with the leech
attached, was positioned in a 1-1 glass beaker containing pond water and secured to
a rigid bar (Fig. 1). Most leeches initiated feeding with only the anterior sucker
attached to the film, but in those instances in which a leech had both sucked
attached, we dislodged the posterior one to obtain a clear view of the head.
 Leech ingestion 515

Pharynx

Pond water

Fig. 1. A diagram showing Hirudo feeding on the artificial blood in a centrifuge tube
through Parafilm. Pharyngeal peristalses move the leech head visibly and body
movements can be readily seen. Hirudo and the Parafilm were digitized from a
photograph and the apparatus was drawn.

Behaviour during ingestion was observed visually. Pharyngeal contractions

were counted for 10 s at the beginning of an ingestive bout and then at 5-min
intervals until ingestion was terminated. Body wall peristalses and dorsoventral
flexions were recorded continuously on an event recorder (Esterline-Angus).
When leeches were satiated, they detached from the Parafilm and the timer was
stopped. Leeches were recovered from the beaker immediately and weighed to
calculate the size of ingested blood meals. The absolute meal size was calculated
by substracting the initial mass of the leech from its final mass. The relative mass of
the ingested blood was calculated as (final mass - initial mass) x(initial mass)" 1 .

Serotonin
To examine the effects of serotonin (5-hydroxytryptamine, 5-HT) upon inges-
tion, intact animals were bathed in dilute solutions of 5-HT in pond water
immediately prior to feeding (Glover & Kramer, 1982). The time of exposure
varied with the 5-HT concentration and leeches were bathed in either 3 or
1 0 x l O " 5 m o i r ' 5-HT for 20 or 2min, respectively. The 5-HT-treated leeches
^vere fed while suspended in pond water and behaviour was observed as above.
Weighing and timing procedures were also identical.
516 CM. LENT AND OTHERS

Excretion
To evaluate the effects of postingestive excretion upon body mass, fed leeches
were reweighed at daily intervals. Haematocrit (ratio of erythrocytes to total
blood volume) was used as an index of the effects of excretion upon the ratio of
blood cells to plasma in the ingestate. To obtain samples of ingested blood, the
posterior one-third of fed leeches was immersed in 3moll" 1 NaCl for approxi-
mately 15 s. This stimulus usually caused regurgitation, a response in which we
frequently assisted the leech by firmly squeezing the body. Crop contents were
regurgitated onto a stainless-steel pan, and immediately five replicate samples
were collected by capillarity into 1x75 mm microhaematocrit tubes. One end of
each tube was sealed with clay (Critoseal, Biological Research, St Louis, MO),
and tubes were spun for lOmin in a clinical centrifuge at 2500rev. min" 1 . The
packed cell to plasma ratio was measured with a standard haematocrit reader.
Haematocrit of the artificial blood samples was measured identically. Feeding
behaviour of the leech moved the Parafilm and appeared to prevent any settling of
blood cells. Nevertheless, to avoid settling, fresh blood at 38°C was added to the
feeding tube, and was stirred at least four times during excretion experiments.
This procedure maintained a relatively constant height of warm, homogeneous
blood in the feeding tube.
Animal size is reported as the median mass x ± range, in grams. Data-are
presented as the arithmetic mean x ± 1 S.E.M. (N, number of observations) unless
otherwise noted, and statistical tests were performed according to procedures
detailed by Sokal & Rohlf (1981).

Results
Behaviour
Under the experimental conditions, Hirudo ingested blood for an average of
24-7 ± 1-2 min (N = 50). As ingestion began, there were extensive contractions of
the body for 1-3 min. The body then elongated and pharyngeal peristalsis could
easily be discerned outside the anterior end. The duration of ingestive behaviour
was influenced by the initial mass of the leech; small leeches ingested blood for
slightly longer than large ones (Fig. 2A). In contrast, duration of ingestion was
unaffected by the final mass of the leech and its blood meal (Fig. 2B). The
duration of leech ingestive behaviour ranged from 14-4 to 47-5 min and, to analyse
the muscular movements of ingestion statistically, durations were normalized to a
value of 1-0.

Pharyngeal peristalsis
Leech ingestion is effected by peristaltic contractions of the pharynx pumping
blood into a distensible crop. These rhythmic movements constitute the definitive
characteristic of ingestion and are visible externally on the head of the leech. Data
on pharyngeal peristalsis were collected from 27 leeches (1-2-5 g). As ingestioM
began, the initial pharyngeal frequency was 2-35Hz (Fig. 3A). The rate of
 Leech ingestion 517

1 2
Initial mass (g)

•2 50
a
B a

40 B B
a
a B

• a
30

1 B
BB° • •
° O Q ° _
20 D
B
a
BO a c * !B

1 0 ••

10 20
Post-ingestive mass (g)

Fig. 2. Duration of Hirudo ingestion. (A) Duration decreases with the initial mass of
the leech. The least squares regression differs significantly from zero (P<0-05).
(B) Duration of ingestion is independent of the final mass of the leech and its blood
meal.

pharyngeal movements then decreased smoothly and continuously until it reached

approximately 1-2Hz and ingestion was terminated. Peristalsis usually ceased a
few seconds before the leech detached from the Parafilm. The frequency was
linearly related to the natural logarithm of the period (r = 0-99) and appeared to
approach an asymptote of 1-17 Hz at the end of ingestion.
It will be seen below that leech size affects several components of ingestion, so
ffo resolve any effects of size on pharyngeal movements, the above data were
divided about the median of 1-75 g into two size groups: 1-1 -70g and 1-80-2-5 g.
518 CM. L E N T AND OTHERS

Small Hirudo contracted the pharynx at higher frequencies than large ones, and it
can be seen in Fig. 3B that small leeches had an initial rate of 2-7 Hz and large ones
2 1 Hz. The frequency of both size groups decayed smoothly with a similar
temporal relationship: a linear decrease in frequency over the logarithm of the
period. The asymptoticpharyngealfrequency of the small leeches was 1-60 Hz and
of large leeches 0-92 Hz. The pharyngeal rates of the two size groups are displaced

3-0 . B y=l-60-0-40(lnx), r = 0-91, F = 78-8

2-5

2-0

1-5

1-0 I—i—i—i i_
y = 0-92-0-38(ln x), r = 0-97, F = 271

0-2 0-4 0-6 0-8 10

Ingestive period

Fig. 3. Pharyngeal peristalsis of Hirudo decays logarithmically during ingestion.

(A) Pharyngeal frequency of 27 leeches (range 1-0-2-5 g, as arithmetic mean ±
I S . E . M . ) . The initial frequency of 2-35Hz decreases to 1-2Hz at termination of
ingestion. The plotted line represents the equation: frequency = 1-17—0-38xln period
0 = 0-99; F = 513). (B) Pharyngeal frequencies of small ( A , 1-0-1-70g) and large
( A , 1-80-2-5 g) leeches decay logarithmically. The slopes of the two groups are similar
but their asymptotic values differ. The 33 min range of feeding times enabled these data
to be divided into 10 modal increments. Horizontal error bars are omitted.
 Leech ingestion 519
vertically from one another by approximately 0-6 Hz throughout the ingestive
period.

Body wall movements

The body of Hirudo undergoes visible movements as blood is ingested. The
most regular of these body movements is peristalsis, which also includes
antiperistaltic movements. Peristalses of the body wall were slower and more
variable than those of the pharynx and had an initial rate of lOmin" 1 (Fig. 4).
These movements decreased to about 2min~' at termination of ingestion and the
rate of change during the period of ingestion was linear (r = 0-97). Peristaltic and
antiperistaltic body movements appeared to be more robust during the earlier than
the later phases of feeding.
Sinusoidal, dorsoventral flexions of the leech body, which resembled those of
swimming (Kristan, Stent & Ort, 1974), occasionally interrupted the peristaltic
body movements. These flexions were less regular than peristalses, and are not
depicted graphically. The initial rate of flexions was 6min~' - approximately half
that of peristalses - and they decreased linearly until ingestion was terminated.
The rate decreased more slowly for flexions than for peristalses, and at termin-
ation of ingestion, both rates were 2 min~'. When dorsoventral flexions were seen,
their amplitudes appeared to be larger in the later phases of feeding. We found
flexions were correlated with the presence of clots inside the Parafilm covering the
feeding tube. After removing the occluding clots, flexions usually disappeared.
However, on some occasions flexions persisted, but with decreased amplitude. In
later experiments, we added the anticoagulant heparin to the artificial blood, and
this procedure usually prevented any flexions.

0-2 0-4 0-6 0-8 1-0

Ingestive period

Fig. 4. Body wall peristalsis of Hirudo decreases during ingestion. The frequency of
these movements was initially 10min~' and it decreased to 2min~1 at termination of
ingestion. The least squares equation for these points is linear.
520 CM. LENT AND OTHERS

Meal size
Hirudo ingest voluminous meals. The leeches in this study had an average mass
of 1-35 ± 0-038 g (N = 119) and they ingested 8-85 ± 0-22 g(N= 119) of blood. The
initial size of the leeches influenced meal size: large leeches ingested somewhat
more blood than small ones. Thus Fig. 5A shows that a 1-g leech ingested 8-4 g of
blood and a 2-g leech 9-7 g. However, most previous reports of leech feeding have
emphasized the relative size of their blood meals (Sawyer, 1986). The average
1-35-g leech ingested a volume of blood of 7-00 ± 0-22 times (N= 119) its mass.

A
g

y = 7- 13+l-28x
15
a
B g B
a D a a
increa

10 • Hi i
-a-
JU-, i a
a
a
•

a a a
1/5 B
CO a o 1 a a
% B a a a a
5 - a D
a

B a a

i i

0-0 0-5 10 1-5 2-0 2-5 3-0

B
y = ll-45-3-32x, r = 0-59
12 a
a
lve mass increase

10

8 a
Q
i
a B
6 a a
a**.
•

B
a a a °sL a
I 4 -

a
a B ;S*4> •
a
2 • a
a

0-0 0-5 1-0 1-5 2-0 2-5 3-0

Initial mass (g)

Fig. 5. Blood meal size of Hirudo. (A) Absolute size of blood meals as a function of
the initial mass of the leech. Size of meals increases slightly as a function of initial mass
and the linear equation differs significantly from zero (P<0-05). (B) Relative meal
size as a function of initial mass. Relative volume decreases linearly.
 Leech ingestion 521
Relative meal size decreased markedly with body size (Fig. 5B) and the relation-
ship was linear (r = 0-59). The 1-g leech ingested 8-15 times its mass in blood and
the 2-g leech 4-80 times. Thus, whereas absolute meal size increased slightly with
leech size, relative size of blood meals decreased markedly.

Serotonin
Salivation and pharyngeal contractions are physiological components of leech
ingestion activated by neuronal serotonin (Lent & Dickinson, 1984). Similarly,
behavioural components of leech ingestion were altered by exposure of leeches to
dilute serotonin. The initial rate of pharyngeal peristalsis increased from 2-3 Hz to
2-8Hz (P<0-05, Fig. 6A), but, between 0-15 and 0-35 of the ingestive period,
pharyngeal rate appeared to be depressed slightly below control values

0-2 0-4 0-6 10

Ingestive period

Fig. 6. Effects of serotonin (5-HT) on leech feeding behaviour. Control means (•)
and estimating equations are plotted together with means ± Is.E.M. of the 5-HT-
treated animals (•). (A) Pharyngeal peristalsis. The initial rate is elevated to 2-8 Hz,
significantly higher than that of controls, 2-3 Hz. The rate appears depressed
(0-05 < P < 0-10) between 0-15 and 0-35 of the feeding period. (B) Serotonin and body
wall peristalsis. The initial increase to 12min~' is slightly higher than the control value
of 10 and the probability that they are the same is P = 0-055.
522 CM. L E N T AND OTHERS

(0-05 < P < 040). During the last half of the ingestive period, 5-HT had no effect
on pharyngeal movements. The effect of 5-HT on body wall peristalsis had a
similar temporal pattern (Fig. 6B) and the initial rate was elevated to 12 min" 1 ,
only slightly higher than the control rate of 10 min"1 (P = 0-055).
After a 5-HT bath, leeches ingested blood meals of 10-07 ± 0-31 g (N = 61),
significantly higher than the 8-85 g of controls (Mest, unpaired comparisons,
P< 0-001). The relative blood meal size of 7-0 times body mass was also increased
to 7-6 ±0-29 (N = 67) by the exposure ( P < 0-015). The ingestive duration of
24-7 min was2min longer after 5-HT treatment, 26-8 ± 1-85 min (N = 67), but this
effect is not statistically significant.

Excretion
Leeches are obviously distended by ingestion, but over 1-2 weeks, the
distension appears to decrease and they regain some of their lost mobility. To
quantify this phenomenon, we measured the daily mass of 10 Hirudo (1-1-3 g) for
12 days immediately following ingestion. The post-fed mass of this group was 8-60
times their pre-fed mass. Total body mass decreased rapidly for 4 days and
stabilized at 3-60 times initial mass on day 6 (Fig. 7A). Even though some of their
distension remained, the mass reduction, of five times their pre-fed mass, partially
restored locomotor capacity. This postingestive mass loss represents approxi-
mately 60 % of the mass of the ingested blood, a proportion which approximates
the plasma volume of the ingestate. To determine whether postingestive mass loss
was accomplished by an excretion of plasma from ingested blood, we measured
the haematocrit of leech crop contents. Blood samples were taken from the crops
of 20 leeches which had fed 6-14 days earlier. The haematocrit of the artificial
blood upon which they had fed ranged between 25 and 40 %. The crop contents of
each leech had a haematocrit of 100 %. No plasma was detected in any of 100 crop
samples after sedimentation at 5000g.
We attempted to assess postingestive excretion directly by placing dyes in the
blood before ingestion, and then observing the appearance of colour in the pond
water which held the fed leeches. Fast green, neutral red and toluidine blue
coloured the artificial blood effectively, but none of these was excreted into the
pond water in visible amounts. However, the dyes were good indicators of
postingestive regurgitation, which occurred in about 10% of the leeches over a
period of 8 weeks.
To ascertain whether the changes in the blood cell to plasma ratio were
sufficient to account for the postingestive mass decreases, crop samples were
collected from different leeches at daily intervals after feeding. As Fig. 7B shows,
the haematocrit of the ingestate increased throughout the period of mass loss and
reached 100% by day four. Simple calculations on these haematocrit and mass
changes demonstrate readily that the excretion of blood plasma accounts for all
the postingestive mass loss by Hirudo.
To determine whether the clear fluid, which flows from the skin of the feeding
leech (Worth, 1951) resulted from excretion during ingestion, we sampled the crop
 Leech ingestion 523

9-0

8-0

7<0
j
u
•s 6-0
B

05
5-0 \

4-0

0 2 4 6 10 12

lOOr B o—o—o—o—o—o—o—o
Crop contents
80

60

40
Blood
20

4 6 8 10 12
Time after feeding (days)

Fig. 7. Ingestion and excretion by Hirudo. (A) Leech mass decreases following
ingestion. The average mass of 10 fed leeches (1-0-1-3 g) decreased from an initial
value of 8-7 (times pre-fed levels) and reached an asymptote of 3-60 times pre-fed level
by day 6. Most of the mass was lost in the first 4 days and the rate of change exceeded
that of an exponential function. (B) The proportion of blood cells (haematocrit) in the
leech ingestate increases over the first 4 days following feeding. From day 6 to day 20
the crops of 20 leeches contained 100% erythrocytes (no plasma). Error bars on the
haematocrit of the artificial blood represent ±1S.E.M.

contents of 10 leeches immediately upon termination of ingestion. Every crop

sample (50 replicates) had more cells than the blood in the feeding tube. The
haematocrit of the blood in the feeding tubes was 25-7 ± 0-5 % (N= 10) and that
of the crop contents 28-2 ± 2 - 1 % (TV =10). This increase in haematocrit is
statistically significant (P<0-05; Mest for paired comparisons) and represents a
10-2% increase in the relative volume of blood cells in the ingestate.
It was necessary to ensure that leeches did not increase the blood cell volume of
their crop contents by selectively ingesting cells over plasma. We tested this

t ossibility in five experiments by measuring the haematocrit of the blood within a

Pfeding tube immediately before and after a leech fed upon it. The haematocrit in
each tube never changed after feeding by more than 0-2 % of its pre-feeding value
524 CM. LENT AND OTHERS

(25 paired replicates). The haematocrit of every leech crop sample was at least
1-0 % higher than that of its feeding tube (50 replicates). There is, therefore, no
indication that cells settle to the bottom of the feeding tube and/or that cells are
selectively harvested by the leech. Thus, leeches excrete plasma during ingestion
and this process increases feeding efficiency. The ingestate has more cells, and
therefore more protein, than the whole blood upon which Hirudo feed.
Anecdotal evidence supports these observations of ingestive excretion. The
fluid which flows from Hirudo feeding on a human is usually clear. In one such
experiment, the fluid exuding from the leech feeding on an individual was yellow.
A blood sample, taken from that individual, revealed a yellow pigment in his
plasma. This observation suggested that the exuding fluid was coloured because
the leech was excreting plasma during ingestion, and it provided the impetus for
these experiments.

Discussion
The context in which Hirudo feeds affects the duration of ingestive behaviour.
When submerged and feeding upon whole blood, Hirudo ingest for about 25 min.
If leeches feed upon tissue culture fluid lacking erythrocytes they ingest one-third
as much in about 8 min (Lent & Dickinson, 1987). Further, if leeches feed in air
rather than submerged, they ingest about half as much blood in half the time
(unpublished observations). The rhythmic movements of the pharynx pump blood
into the crop, and an average leech contracts its pharynx approximately 2500 times
in ingesting a satiating meal. The rate of pharyngeal peristalsis decays logarithmi-
cally; but in leeches which do not distend during ingestion (i.e. those with incised
body walls; Lent & Dickinson, 1987), the rate does not decrease as rapidly. This
observation suggests that body wall distension has a role in slowing pharyngeal
peristalsis. A logarithmically decreasing time course could be produced mechan-
ically. Increasing pressure on the crop fluid as ingestion proceeds would require
each contraction to develop more contactile force, lowering pharyngeal frequency.
The decay could also be produced by a central mechanism. Identified neurones
release serotonin peripherally which excites the pharynx (Lent & Dickinson,
1984), and distension receptors in the body inhibit these neurones (Lent &
Dickinson, 1987). An increasing inhibition arising from distensive synaptic input
as ingestion proceeds could reduce the firing rates of serotonergic neurones and
result in decreased levels of peripheral serotonin which, in turn, would slow
peristalsis. It should be noted that the addition of warm blood to the feeding tube
during ingestion does not increase pharyngeal frequency, and thus the decay does
not result from a cooling of the blood in the feeding tube.
The peristaltic and antiperistaltic movements of the leech body wall appear to
be a mechanism for distributing the ingested blood uniformly throughout the crop.
The Hirudo crop is a capacious storage organ which fills the body and has 11 paj
of lateral diverticula, the most posterior pair of which is exceptionally elonga
(Sawyer, 1986). To maximize meal volume, the incoming blood must fill these 22
I
 Leech ingestion 525

caecal pockets. Body wall peristalses force the ingested blood back and forth along
the central chamber of the crop, and this longitudinal flow probably facilitates the
lateral movement of blood through the small orifices of the diverticular chambers.
Dorsoventral flexions of the body were seen when clots formed in the feeding
tube and were absent, or highly attenuated, when anticoagulant was added to the
artificial blood. Flexions were more robust towards the end of feeding bouts, when
clots would have enlarged and become more occlusive. Dorsoventral flexions
might be a leech behaviour which dislodges clots and maintains an adequate flow
of blood from skin incisions whenever the salivary anticoagulant hirudin proves
ineffective.
Excretion is a major physiological component of leech ingestion. Fluid excretion
during ingestion increases the efficiency of leech feeding. By excreting the plasma
from incoming blood, the crop can be made to contain more cells than an equal
volume of whole blood. For a 1-g leech the observed increase represents
approximately 4xlO y erythrocytes more than the number in an equal volume of
whole human blood. The haemoglobin in mammalian erythrocytes is a primary
source of Hirudo dietary protein (Windsor, 1970) and linking excretion with
ingestion provides Hirudo with more digestible protein and enables the leech not
to feed for a full year.
After ingestion is terminated, the continued excretion of plasma restores some
of the mobility which the leech lost as a result of ingesting a voluminous meal.
Although blood cells are concentrated by excretion for 4-5 days, their volume is
not insubstantial and this distension is alleviated only slowly by digestion. Hirudo
lyse their ingested erythrocytes, digest the protein of the haemoglobin within their
short intestine and egest the porphyrin, haem (Windsor, 1970).
Small animals generally have higher rates of mass-specific metabolism and
rhythmic movements than large ones (Schmidt-Nielsen, 1984). This size principle
pertains to leech metabolism (Kulkarni & Nagabushanum, 1978) and is reflected in
the differing pharyngeal frequencies reported here. Small leeches are likely to
require relatively more energy than large ones during the long period of satiation
because of higher mass-specific metabolism. Thus, metabolic differences resulting
from leech size should affect meal volume such that relative meal size should
decrease with increasing leech mass. This is precisely what we see in the sizes of
the meals ingested by adult Hirudo (Fig. 5B). However, we have also fed five 0-5-g
leeches and these juveniles ingested only between 1 and 1-5 g of blood: 2-3 times
their initial mass. Such small meal sizes in leeches with high metabolic rates
combine a limited source of energy with a high rate of consumption. Thus, juvenile
Hirudo probably feed more often than once a year. Juveniles of the South
American leech Haementeria feed more frequently than mature specimens
(Sawyer, 1986).
Leech ingestion comprises stereotyped muscular movements, salivary secretion
nd nephridial excretion. During ingestion, pharyngeal contractions and salivation
activated by serotonin which is released peripherally by impulse activity of
identified neurones (Lent & Dickinson, 1984). Serotonin neurones are synaptically
526 CM. LENT AND OTHERS

excited into impulse activity by ingestive stimuli and are synaptically inhibited by
the distensive stimulus, and cease firing (Lent & Dickinson, 1987). Excretion is
activated concomitantly with ingestion but continues for several days. Hence, the
mechanism responsible for the long-term activation of excretion differs, at least
quantitatively, from the mechanism for the short-term activation of pharyngeal
muscles and salivary glands.
Exposing intact Hirudo to serotonin affects components of feeding behaviour:
swim initiation, biting frequency and relative blood meal size (Lent & Dickinson,
1984). We find that exogenous 5-HT increases the absolute, as well as the relative,
size of Hirudo blood meals. A brief exposure to serotonin also transiently
increases the rate of pharyngeal peristalsis. Since the exposure has no significant
effect on duration of ingestion, serotonin probably increases leech meal size by its
effect on pharyngeal peristalsis. Serotonin also reduces the resting tension of leech
body muscles (Mason & Kristan, 1982), a peripheral effect which undoubtedly
assists the pharynx in filling a more compliant body. Exogenous serotonin
probably exerts behavioural effects by activating peripheral receptors via the leech
skin. At least three chemicals, 5-HT (Glover & Kramer, 1982), its neurotoxic
analogue 5,7-dihydroxytryptamine (Glover, 1987) and neutral red dye (Lent,
1981), reach internal structures of the intact leech from the bathing medium.
Sanguivorous feeding by the nymphs of the bug Rhodnius resembles that of
Hirudo in several aspects. Rhodnius ingest blood meals of 6-7 times their mass
and the resulting distension terminates ingestion (Maddrell, 1963). The saliva of
Rhodnius has anticoagulant activity (Ribeiro & Garcia, 1980). Excretion is
activated as Rhodnius feeds, which must increase the proportion of blood cells to
plasma in the ingestate and, after feeding, excretion continues as postingestive
diuresis (Maddrell, 1964). These are striking functional similarities in feeding by a
blood-sucking annelid and a blood-sucking insect and yet these functions rely upon
anatomically disparate organ systems. Efficient sanguivory upon mammalian
blood may necessitate the convergent evolution of a similar set of physiological
mechanisms.

We thank David R. Smith and David Zundel for technical assistance with some
of these experiments. This research was supported by PHS grants NS-14482 and
NS-24077 to CML.

References
DICKINSON, M. H. & LENT, C. M. (1984). Feeding behavior of the medicinal leech, Hirudo
medicinalis L. J. comp. Physiol. A 154, 449-455.
ELLIOT, E. (1986). Chemosensory stimuli in leech feeding: An important role for NaCl and
arginine. J. comp. Physiol. A 159, 391-401.
GLOVER, J. G. (1987). Serotonin storage and uptake by identified neurons in the leech
Haementeria ghilianii. J. comp. Neurol. 256, 117-127.
GLOVER, J. G. & KRAMER, A. P. (1982). Serotonin analogue selectively ablates identified
neurons in the leech embryo. Science 216, 1012-1014.
HAYCRAFT, J. B. (1884). On the action of a secretion obtained from the medicinal leech on th?
coagulation of blood. Proc. R. Soc. Ser. B 36, 478-487.
 Leech ingestion 527
KRISTAN, W. B., JR, STENT, G. S. & ORT, C. A. (1974). Neuronal control of swimming in the
medicinal leech. I. Dynamics of the swimming rhythm. J. comp. Physiol. 94, 97-119.
KULKARNI, G. K. & NAGABUSHANUM, R. (1978). Respiratory metabolism of Indian leech
Poecilobdella viridis: influence of body size. Marathwada Univ. J. Sci. 17, 97-100.
LENT, C. M. (1981). Morphology of neurons containing monoamines within leech segmental
ganglia. J. exp. Zool. 216, 311-316.
LENT, C. M. (1986). New medical and scientific uses of the leech. Nature, Lond. 323, 494.
LENT, C. M. & DICKINSON, M. H. (1984). Serotonin integrates the feeding behavior the
medicinal leech. J. comp. Physiol. A 154, 457-471.
LENT, C. M. & DICKINSON, M. H. (1987). On the termination of ingestive behaviour by the
medicinal leech. J. exp. Biol. 31, 1-15.
LENT, C. M., FLIEGNER, K. & FREEDMAN, E. (1986). Ingestive behavior of the medicinal leech.
Soc. Neurosci. Abstr. 12, 407.
MADDRELL, S. H. P. (1963). Control of ingestion in Rhodnius prolixus Stal. Nature, Lond. 198,
210.
MADDRELL, S. H. P. (1964). Excretion in the blood-sucking bug, Rhodnius prolixus Stal. II. The
normal course of diuresis and the effect of temperature. J. exp. Biol. 41, 163-176.
MARSHALL, C. G. (1984). The salivary glands of the Hirudinea: Electrophysiology, structure and
secretion. Ph.D. dissertation, Brown University, Providence, RI.
MASON, A. & KRISTAN, W. B. JR (1982). Neuronal excitation, inhibition and modulation of
leech longitudinal muscle. /. comp. Physiol. 146, 527-536.
MULLER, K. J., NICHOLLS, J. G. & STENT, G. S. (eds) (1981). Neurobiology of the Leech. New
York: Cold Spring Harbor Laboratory Press.
RIBEIRO, J. M. C. & GARCIA, E. S. (1980). The salivary and crop apyrase activity of Rhodnius
prolixus. J. Insect Physiol. 26, 303-307.
SAWYER, R. T. (1986). Leech Biology and Behaviour. Oxford: Clarendon Press.
SCHMIDT-NIELSEN, K. (1984). Scaling: Why is Animal Size so Important? London: Cambridge
University Press.
SOKAL, R. R. & ROHLF, F. J. (1981). Biometry, 2nd edn. San Francisco: Freeman.
WINDSOR, D. A. (1970). Faeces of the medicinal leech, Hirudo medicinalis, are haem. Nature,
Lond. 227, 1153-1154.
WORTH, C. B. (1951). Description and discussion of the biting of an Indian land leech (Annelida:
Hirudinea). J. Bombay nat. Hist. Soc. 510, 423-426.
ZERBST-BOROFFKA, I. (1973). Osmo- und Volumenregulation bei Hirudo medicinalis nach
Nahrungsaufnahme. J. comp. Physiol. 127, 343-347.