Life cycle abbreviation in the trematode Coitocaecum parvum: can parasites adjust to variable conditions?

Abstract

The complex life cycles of parasites are thought to have evolved from simple one-host cycles by incorporating new hosts. Nevertheless, complex developmental routes present parasites with a sequence of highly unlikely transmission events in order to complete their life cycles. Some trematodes like Coitocaecum parvum use facultative life cycle abbreviation to counter the odds of trophic transmission to the definitive host. Parasites adopting life cycle truncation possess the ability to reproduce within their intermediate host, using progenesis, without the need to reach the definitive host. Usually, both abbreviated and normal life cycles are observed in the same population of parasites. Here, we demonstrate experimentally that C. parvum can modulate its development in its amphipod intermediate host and adopt either the abbreviated or the normal life cycle depending on current transmission opportunities or the degree of intra-host competition among individual parasites. In the presence of cues from its predatory definitive host, the parasite is significantly less likely to adopt progenesis than in the absence of such cues. An intermediate response is obtained when the parasites are exposed to cues from non-host predators. The adoption of progenesis is less likely, however, when two parasites share the resource-limited intermediate host. These results show that parasites with complex developmental routes have transmission strategies and perception abilities that are more sophisticated than previously thought.

Full text

Page 1

Life cycle abbreviation in the trematode Coitocaecum parvum: can
parasites adjust to variable conditions?
C. LAGRUE & R. POULIN
Department of Zoology, University of Otago, Dunedin, New Zealand
Introduction
Life cycles involving several host species are observed in
numerous and phylogenetically distant groups of para-
sites (Poulin, 1998). Among those, the complex life cycles
of parasitic helminths are thought to have evolved from
simple one-host cycles by incorporation of new hosts
(Parker et al., 2003a). For example, a predator might
have been added to a parasite life cycle if it frequently
consumed prey used by a parasite as host, therefore
becoming a definitive host. In the case of trematodes, the
ancestral single host was a vertebrate and new interme-
diate hosts were added during the course of evolution
(Cribb et al., 2003). First, adopting a vertebrate as a host
may have allowed high reproduction rates in hosts of
large body size and long life-span (Parker et al., 2003a),
and a greater rate of cross-fertilization by concentrating
isolated individuals (Brown et al., 2001). Then, the
incorporation of intermediate hosts, used as transmission
vectors, might have increased the probability of reaching
a new vertebrate host individual (Choisy et al., 2003).
Trematodes are also known to undergo active asexual
multiplication within their mollusk first host, thus
increasing their likelihood of transmission (Cribb et al.,
2003). Nevertheless, complex transmission routes pre-
sent parasites with a succession of highly hazardous
transmission events in order to complete their life cycle,
reach maturity and reproduce. This is particularly true in
trophic transmission when the infected intermediate host
must be eaten by the definitive host for the parasite to
achieve its maximum growth and reproduce. It is well
known that parasites relying on trophic transmission
often alter the behaviour or appearance of their inter-
mediate host to increase the likelihood of predation by
the predator host (Moore, 2002). In several groups, some
Correspondence: Cle ´ment Lagrue, Department of Zoology, University of
Otago, PO Box 56, Dunedin 9054, New Zealand.
Tel.: +64 3 479 7964; fax: +64 3 479 7584;
e-mail: lagcl981@student.otago.ac.nz
ª 2006 THE AUTHORS 20 (2007) 1189–1195
JOURNAL COMPILATION ª 2006 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY
1189
Keywords:
abbreviated life cycle;
alternative strategies;
Coitocaecum parvum;
phenotypic plasticity;
progenesis;
trematode.
Abstract
The complex life cycles of parasites are thought to have evolved from simple
one-host cycles by incorporating new hosts. Nevertheless, complex develop-
mental routes present parasites with a sequence of highly unlikely transmis-
sion events in order to complete their life cycles. Some trematodes like
Coitocaecum parvum use facultative life cycle abbreviation to counter the odds of
trophic transmission to the definitive host. Parasites adopting life cycle
truncation possess the ability to reproduce within their intermediate host,
using progenesis, without the need to reach the definitive host. Usually, both
abbreviated and normal life cycles are observed in the same population of
parasites. Here, we demonstrate experimentally that C. parvum can modulate
its development in its amphipod intermediate host and adopt either the
abbreviated or the normal life cycle depending on current transmission
opportunities or the degree of intra-host competition among individual
parasites. In the presence of cues from its predatory definitive host, the
parasite is significantly less likely to adopt progenesis than in the absence of
such cues. An intermediate response is obtained when the parasites are
exposed to cues from non-host predators. The adoption of progenesis is less
likely, however, when two parasites share the resource-limited intermediate
host. These results show that parasites with complex developmental routes
have transmission strategies and perception abilities that are more sophisti-
cated than previously thought.
doi:10.1111/j.1420-9101.2006.01277.x

Page 2

parasites use more extreme adaptations to counter the
odds of trophic transmission: among trematodes, for
instance, several species are able to drop the definitive
host from their life cycle (Poulin & Cribb, 2002). Those
parasites produce eggs while still in their intermediate
host via progenesis; the worms achieve precocious
maturity and lay viable eggs by self-fertilization (Poulin,
2001). Although some species display extreme forms of
life cycle truncation with only one host remaining, most
life cycle truncations are facultative with both strategies
present in the population. Conditions that trigger the
adoption of one or the other strategy by parasites are
currently poorly understood.
Poulin (2003) demonstrated that a trematode parasite
can accelerate its development and reach precocious
maturity in its intermediate host in the absence of
definitive hosts. This result illustrates that some parasites
may adjust their life strategy with respect to their
immediate transmission opportunities to an appropriate
definitive host. However, in a wider ecological context,
other predators unsuitable as hosts may actively feed on
the same prey species and significantly influence the
benefit of one or the other strategy (Mouritsen & Poulin,
2003). Thomas et al. (2002a) suggested that selection
may have favoured parasites with the ability to perceive a
diversity of signals from environmental parameters
affecting their immediate and future fitness. Thus,
parasites might be able to strategically respond to a
diversity of signals emanating from their host environ-
ment such as the presence of predators (Combes et al.,
2002). Crustaceans are often used as intermediate hosts
by helminth parasites; in addition, crustaceans are
known to perceive the presence of fish predators by
detecting the chemical components of fish mucus in the
water (e.g. Lehtiniemi & Linde ´n, 2006) and change their
behaviour accordingly (Neill, 1990; Burks et al., 2000).
Consequently, parasites needing either to avoid preda-
tion by a fish or to finish their life cycle in a fish gut
should adopt different strategies based on their detection
of predator-induced stress in the intermediate host.
However, the actual perception ability of parasites and
how effectively they can adjust their life strategies is
currently unknown.
The present study addresses this important question
using the same model as Poulin (2003). The trematode
Coitocaecum parvum (Opecoelidae) is a common parasite of
freshwater fish in New Zealand (MacFarlane, 1939;
Holton, 1984a), principally the common bully (Gobio-
morphus cotidianus). Eggs are released in fish faeces and
hatch into free-swimming larvae (miracidia). Miracidia
penetrate snails in which they multiply and develop into
sporocysts. Sporocysts asexually produce cercariae, free-
living larvae that enter the amphipod Paracalliope fluvi-
atilis where they encyst as metacercariae in the body
cavity. At this stage, metacercariae can either stop
growing and await ingestion by a fish where they will
mature and reproduce, or keep growing and reach
maturity while still inside the amphipod. Worms that
reach maturity in the crustacean intermediate host
reproduce by selfing and lay eggs in the amphipod’s
body cavity (Holton, 1984b; Poulin, 2001). Eggs pro-
duced by selfing hatch after host death into larvae that
are infective to the snail first host without the need to
pass through the fish host. Coitocaecum parvum can adopt
preferentially the shorter life cycle in the absence of
chemical cues emanating from its preferred fish definitive
host, the common bully (Poulin, 2003). However, Poulin
(2003) used naturally infected amphipods, and it is
possible that variation in the age of metacercariae may
have confounded the results. It is also important to
determine if this parasite is able to distinguish between its
fish definitive host and other nonhost predators and
adapt its life strategy accordingly as multiple nonhost
predators commonly co-occur with G. cotidianus.
Here, we tested the hypothesis, using experimental
infections, that the trematode C. parvum can use infor-
mation about current opportunities of transmission to its
fish definitive host to adjust its life-history strategy. We
expect that the parasite will shorten its life cycle not only
when the opportunities for transmission to the definitive
host are limited (low density of hosts), but also when
other predators not suitable as definitive hosts are
present, significantly decreasing the probability of trans-
mission to the appropriate host. In these two different
situations, the parasite may use progenesis to guarantee
the production of at least a few eggs before its amphipod
host dies or is eaten by a nonhost predator. However, the
proportion of worms adopting an abbreviated life cycle
and the rate of egg production might be different
between situations where there is no predator and
situations where nonhost predators are present. The
presence of any predator, even if not suitable as a host for
C. parvum, may induce stress in the amphipod that is
perceived by the parasite and that influences its life-
strategy.
Materials and methods
Animal collection
Naturally infected snails (Potamopyrgus antipodarum) and
amphipods (Paracalliope fluviatilis) were collected among
macrophytes in Lake Waihola, South Island, New
Zealand using dip nets. Snails and amphipods were kept
alive in aerated lake water before the experiments and
stocks of both animals were renewed every fortnight to
control for the delay before processing. Infected snails
were obtained by selectively choosing individuals that
displayed an altered shell shape, a sure sign of infection
by C. parvum (C. Lagrue, personal observation). Unin-
fected amphipods were obtained by inspecting each
amphipod under a microscope and discarding all amphi-
pods that showed any sign of infection, i.e. an opaque
mass in the body cavity corresponding to a metacercaria.
1190
C. LAGRUE AND R. POULIN
ª 2006 THE AUTHORS 20 (2007) 1189–1195
JOURNAL COMPILATION ª 2006 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

Page 3

This method allows the selection of only uninfected
individuals with an accuracy of about 95% (Lefebvre &
Poulin, 2005a). Twenty-five common bullies, the pre-
ferred definitive host of C. parvum, and 25 perch (Perca
fluviatilis), a nondefinitive host predator (Poulin, 2003),
were collected in Lake Waihola using a set net, kept alive
separately in 200 L tanks and fed with commercial fish
food. Twenty-five crayfish (Paranephrops zealandicus),
another nondefinitive host predator (Hollows et al.,
2002), were captured by hand in Ross Creek, within
Dunedin city, using a spot light after dark.
Experimental infections
Cercariae of the trematode parasite C. parvum were
obtained from snails under controlled conditions to
ensure that the cercariae used to experimentally infect
amphipods were freshly released and, therefore, more
likely to penetrate the amphipod. Randomly chosen
snails were transferred to Petri-dishes (about 50 snails per
dish) filled will 10 mL of filtrated lake water. Snails were
then incubated at 25 ?C for 20 min under constant light,
conditions known to induce cercarial release (Hay et al.,
2005). The Petri-dishes were then screened under a
microscope and the cercariae found were transferred to
500 lL Eppendorf tubes using a 20 lL micropipette. Two
cercariae were placed in each tube with 2.5 lL of filtrated
lake water and an amphipod was then added. Amphipods
were left in the tube along with the two cercariae for 5 h,
a time after which unsuccessful cercariae stop moving
and die. Amphipod survival, at this stage, was over 99%.
Amphipods were then separated randomly into groups of
about 50 individuals. Each group was placed in a plastic
container filled with 400 mL of aged and aerated lake
water; strands of macrophytes (Elodea canadensis) were
added for food. Aged water consisted of water collected at
least 1 week before the experiment to allow any chem-
ical cue to deteriorate (Poulin, 2003). Given that 3414
amphipods were experimentally infected for the purpose
of this study, controlled infections and treatments were
conducted over several weeks between July and October
2005.
Treatment
All types of scented water were prepared every second
day following the same protocol: 2.5 L of aged lake water
were placed in each of four identical plastic tanks, and
four individuals of either perch, common bully or
crayfish, randomly chosen from a pool of 25, were added
separately to three of the four tanks. The fourth tank was
left with only water (control water) to standardize the
treatment. Animals were left overnight and removed in
the morning. The different types of treated water were
immediately used and added to the tanks containing the
experimentally infected amphipods. Every 2 days and for
5 weeks, 60 mL of water was removed from each
amphipod container and replaced with either control
water or water conditioned by one of the three predators.
The same treatment was applied to a given container for
the whole experiment. In total, 730 amphipods served as
control, 1190 received bully water, 702 received perch
water and 792 were treated with crayfish-marked water.
Measures and statistical analyses
After 5 weeks, all surviving amphipods (309, 388, 80 and
277 respectively for each treatment) were killed in 70%
ethanol to facilitate dissections and measurements, rinsed
in distilled water and dissected immediately after.
Amphipods were measured (body length) and dissected
under the microscope. Any worm they contained was
measured (length and width) under a compound micro-
scope, and recorded as ‘normal’ (nonegg producing
worm) or ‘progenetic’ (egg producing worm); in the case
of progenetic parasites, eggs were also counted. The body
surface of each parasite was then determined and used as
a surrogate for body size. This was done using the
formula for an ellipsoid, (pLW)/4, where L and W are the
length and width of the parasite. As amphipods were
experimentally exposed to two cercariae, some individ-
uals contained two metacercariae of C. parvum, whereas
others had only one or were uninfected. Because
competition between individuals sharing the same host
is an important factor influencing life-history strategies
and growth of parasites (Thomas et al., 2002b; Parker
et al., 2003b), recovered worms were divided in two
different classes under ‘infection status’: single infections
(one worm per amphipod) and double infections (two
worms per amphipod).
Linear regressions were applied to test for the effect of
amphipod size on the size of the parasites. Effects of
treatment and infection status (single or double infec-
tion) on parasite strategy were tested using Fisher’s exact
tests; the proportions of parasites in each class were
compared in a pair-wise manner. Effects of treatment,
infection status and strategy (‘normal’ or ‘progenetic’) on
parasite body size were tested using a three-way ANOVA
with the size of the worm used as the dependent variable.
The body area of the parasite was log transformed before
analyses to normalize the data. The effect of the treat-
ment on egg production was tested using a nonparamet-
ric test (Kruskal–Wallis ANOVA ANOVA) with the number of eggs
used as the dependent variable and the type of treatment
as the independent variable; only parasites that had
produced at least one egg were included in this analysis.
Finally, a linear regression between the size of the
parasite and the number of eggs produced was used to
assess the effect of parasite size on egg production.
ANOVA
Results
Overall, 629 metacercariae of C. parvum were recovered
from the 1054 surviving amphipods. Of these 1054
Abbreviation of complex life cycles
1191
ª 2006 THE AUTHORS 20 (2007) 1189–1195
JOURNAL COMPILATION ª 2006 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

Page 4

individuals, 486 had been successfully infected: 343 had
one metacercaria and 143 had two. However, because it
was impossible to know the proportion of dead amphi-
pods that was parasitized, we could not determine the
actual percentage of individuals infected at the beginning
of the experiment.
Amphipod length did not differ among treatments (F3,
482 ¼ 1.39, P ¼ 0.24) and, thus, all infected amphipods
were used to test for the effect of amphipod size on the
size of the parasites. Across all treatments, the body size
of normal worms was not related to amphipod length
either in single infections (r ¼ 0.035, n ¼ 222, P ¼ 0.61)
or in double infections (r ¼ 0.093, n ¼ 245, P ¼ 0.15);
there were also no significant correlations between these
variables within any of the treatments. On the contrary,
the body size of progenetic parasites found in single
infections was significantly related to host length (r ¼
0.19, n ¼ 121, P < 0.05). This weak trend was also
observed in progenetic worms found in double infections
but was not significant probably because of the smaller
sample size (r ¼ 0.17, n ¼ 41, P ¼ 0.29).
In single infections (one parasite per host), the
proportion of progenetic worms in the bully treatment
was significantly lower than in any other treatment
(Table 1A and Fig. 1). The control treatment showed the
significantly highest percentage of progenetic parasites
while there was no difference between the perch and
crayfish treatments (Table 1A and Fig. 1). In double
infections, those differences in occurrence of progenesis
were less pronounced (Table 1B) and, apart from the
bully treatment, the percentage of progenetic parasites
was clearly lower than in single infections (Table 2 and
Fig. 1). However, the difference was not significant in the
perch treatment, probably due to the small sample size.
The growth achieved by the parasites was significantly
different among the four treatments (Table 3): worms
exposed to control water were significantly larger
(0.125 ± 0.007 mm2)than
(0.04 ± 0.003 mm2), crayfish (0.081 ± 0.005 mm2) and
perch (0.095 ± 0.011 mm2) treatments (Fisher’s LSD,
d.f. ¼ 613, all P < 0.001), while trematodes from the
bully treatment were the smallest (Fisher’s LSD, d.f. ¼
613, all P < 0.001). The life-history strategy also proved
to have an effect on parasite mean body size: progenetic
parasites were significantly larger than normal ones
(Table 3). In contrast, the infection status (single or
double) had no significant effect on trematode size.
However, there was a significant interaction between the
in the bully
Table 1 Results of Fisher’s exact tests for pair-wise comparisons of
the proportion of progenetic parasites between treatments: (A) for
parasites in single infections and (B) for parasites in double
infections.
Treatments comparedv2
P-value
A
Control vs. bully
Control vs. crayfish
Control vs. perch
Bully vs. crayfish
Bully vs. perch
Crayfish vs. perch
124.67
19.37
17.35
51.66
21.87
0.66
< 0.0001
< 0.0001
0.0001
< 0.0001
0.0001
0.276
B
Control vs. bully
Control vs. crayfish
Control vs. perch
Bully vs. crayfish
Bully vs. perch
Crayfish vs. perch
4.03
0.32
0.00
8.65
2.10
0.13
0.045
0.367
0.627
0.003
0.160
0.501
0
10
20
30
40
50
60
70
80
90
Control Crayfish Perch
Single infection
Bully
Double infection
0
10
20
30
40
50
60
70
80
90
Percent of progenetic parasites
Percent of progenetic parasites
72
88
29
154
27
54
9
53
Control Crayfish PerchBully
Fig. 1 Proportion of egg producing (i.e. progenetic) parasites in each
treatment as a function of their infection status (i.e. single or double
infection). Numbers above bars are the number of infected amphi-
pods in each category.
Table 2 Results of Fisher’s exact tests for comparisons of the pro-
portion of progenetic parasites between single and double infections
within each treatment.
Single vs. doublev2
P-value
Control
Bully
Crayfish
Perch
50.58
0.03
15.28
2.40
< 0.0001
0.537
0.0001
0.110
Table 3 Results of the three-way ANOVA
tion status and life-history strategy as independent variables and the
size of the parasites as the dependent variable.
ANOVA using treatment, infec-
Main effectsd.f.
FP-value
Treatment
Infection status
Life-history strategy
Treatment · infection status
Treatment · strategy
Infection status · strategy
Treatment · infection status · strategy
3
1
1
3
3
1
3
5.543
2.813
568.723
1.147
2.859
10.405
0.310
0.0009
0.094
< 0.0001
0.329
0.036
0.0013
0.818
1192
C. LAGRUE AND R. POULIN
ª 2006 THE AUTHORS 20 (2007) 1189–1195
JOURNAL COMPILATION ª 2006 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

Page 5

life-history strategy and the infection status (see Table 3
and Fig. 2): normal parasites were significantly larger
when found in double infections than in single infections
while there was no significant difference among pro-
genetic individuals. There was also a significant interac-
tion between the life-history strategy and the type of
treatment (see Table 3 and Fig. 3): the mean size of
progenetic parasites was less affected by the treatment
than that of normal parasites. Finally, there were no
other significant interactions (Table 3).
The mean number of eggs produced per progenetic
parasite (27.4 ± 2.6, 19.7 ± 2.4,
28.6 ± 6.1 for Control, bully, crayfish and perch treat-
ments respectively) was not influenced by the type of
treatment (Kruskal–Wallis
P ¼ 0.517) but increased significantly with the size of
the parasite (r ¼ 0.581, n ¼ 162, P < 0.0001; Fig. 4).
23.3 ± 2.4and
ANOVAANOVA, H3, 162 ¼ 2.273,
Discussion
The likelihood of completing all transmission events is
considered as the main force acting on the evolution of
abbreviated life cycles (Poulin & Cribb, 2002). Poulin
(2001) suggested that the high density of amphipods and
the presence in New Zealand freshwaters of other
crustacean and fish predators, not suitable as definitive
hosts, induce an extremely low probability of transmis-
sion to the definitive host for C. parvum. Therefore,
hazardous transmission events may have promoted the
evolution of progenesis as an alternative strategy, giving
the trematode a wider range of transmission options.
Coitocaecum parvum should then be able to adjust its
developmental strategy to its immediate transmission
opportunities. The present study shows that this trema-
tode notonly adoptspreferentially
the absence of chemical cues from common bullies
(G. cotidianus), its definitive host (as found by Poulin,
2003), but a high proportion of individual parasites grow
larger and produce eggs within amphipods exposed to
chemical cues emanating from other nonhost predators,
the perch (Perca fluviatilis) and the crayfish (Paranephrops
zealandicus). In other words, the developmental strategy
progenesisin
0
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
0.20
Normal parasites Progenetic parasites
Mean size (mm2)
Single infection
Double infection
222
245
121
41
Fig. 2 Effects of life-history strategy and infection status on the
mean size (± SE) of parasites (see Table 3 for results of ANOVA
hoc tests showed a significant influence of the infection status on the
size of normal parasites (Fisher’s LSD, d.f. ¼ 613, P < 0.0001) while
no significant difference was found in progenetic parasites between
single and double infections (Fisher’s LSD, d.f. ¼ 613, P ¼ 0.191).
Numbers above bars are sample sizes.
ANOVA). Post
0
0.05
0.10
0.15
0.20
Control
Bully
Crayfish
Perch
Normal parasitesProgenetic parasites
Mean size (mm2)
59
242
133
33
67
18
63
14
Fig. 3 Effects of treatment and life-history strategy on the mean size
(± SE) of parasites (see Table 3 for results of ANOVA
showed significant effects of the different types of treatment on the
size of normal parasites: worms in the bully treatment were
significantly smaller than in any other treatment (Fisher’s LSD,
d.f. ¼ 613, all P < 0.0001) while parasites exposed to control water
were significantly larger than those in crayfish treatment but not in
perch treatment (Fisher’s LSD, d.f. ¼ 613, P < 0.0001 and P ¼ 0.67
respectively). No significant effect was detected within progenetic
parasites (Fisher’s LSD, d.f. ¼ 613, all P > 0.13).
ANOVA). Post hoc tests
0.08 0.100.120.140.160.18 0.200.22 0.240.26
Parasite size (mm2)
0
10
20
30
40
50
Mean number of eggs
2
5
27
30
27
20
23
13
12
Fig. 4 Relationship between the mean (± SE) number of eggs
produced and the size of the parasite. Only progenetic worms were
used for this analysis. Numbers above bars are sample sizes.
Abbreviation of complex life cycles
1193
ª 2006 THE AUTHORS 20 (2007) 1189–1195
JOURNAL COMPILATION ª 2006 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY