proximate factors affecting the larval life history of acanthocephalus lucii (acanthocephala)

742

J. Parasitol., 93(4), 2007, pp. 742–749� American Society of Parasitologists 2007

PROXIMATE FACTORS AFFECTING THE LARVAL LIFE HISTORY OFACANTHOCEPHALUS LUCII (ACANTHOCEPHALA)

Daniel P. Benesh and E. Tellervo ValtonenDepartment of Biological and Environmental Science, P.O. Box 35, FI-40014 University of Jyvaskyla, Finland. e-mail: [email protected]

ABSTRACT: The growth and eventual size of larval helminths in their intermediate hosts presumably has a variety of fitnessconsequences. Therefore, elucidating the proximate factors affecting parasite development within intermediate hosts should pro-vide insight into the evolution of parasite life histories. An experimental infection that resulted in heavy intensities of an acan-thocephalan (Acanthocephalus lucii) in its isopod intermediate host (Asellus aquaticus) permitted the examination of parasitedevelopmental responses to variable levels of resource availability and intraspecific competition. Isopods were infected by ex-posure to egg-containing fish feces, and larval infrapopulations were monitored throughout the course of A. lucii development.The relative rate of parasite growth slowed over time, and indications of resource constraints on developing parasites, e.g.,crowding effects, were only observed in late infections. Consequently, the factors likely representative of resource availability tolarval parasites (host size and molting rate) primarily affected parasite size in late infections. Moreover, at this stage of infection,competitive interactions, gauged by variation in worm size, seemed to be alleviated by greater resources, i.e., larger hosts thatmolted more frequently. The relatively rapid, unconstrained growth of young parasites may be worse for host viability than theslower, resource-limited growth of larger parasites.

Many parasite life history traits, e.g., growth rates, are pre-sumably determined by the level of host resource consumption.Therefore, the exploitation strategies employed by larval hel-minths in their intermediate hosts should be reflected in parasitegrowth rates and body sizes. The potential fitness benefitslinked to larger larval size may include better establishmentsuccess in the definitive host (Rosen and Dick, 1983; Steinauerand Nickol, 2003), higher adult fecundity (Fredensborg andPoulin, 2005), and less developmental time to sexual maturity(Poulin, 1998). Consequently, selection should promote fasterlarval growth and greater ultimate size. However, the direction-al selection on parasite size is probably stabilized by the generalneed to maintain host viability until transmission occurs (Laf-ferty and Kuris, 2002).

This evolutionary tradeoff has been modeled to predict growthstrategies of larval helminths in their intermediate hosts (Parker etal., 2003). Specifically, it was suggested that growth patterns maynot be entirely a function of resources, but instead they may reflectflexible, adaptive life history strategies. That is, individual para-sites in the presence of competing conspecifics decrease their size,and thus the parasite burden on the host, not as a response tolimited resources, but to maintain host viability. Parker et al.(2003) attempted to address the predictions of their model by ex-amining data taken from the literature on experimental infectionsof copepods with pseudophyllidean cestodes, but too little infor-mation was available to make comparisons robust. Recent exper-imental work, though, suggests that the larval growth of the ces-tode Schistocephalus solidus may vary in an adaptive manner (Mi-chaud et al., 2006). Like cestodes, there is a paucity of informationon the growth strategies of acanthocephalans in their intermediatehosts. Intermediate host size and the presence of competitors areknown to affect acanthocephalan development in some systems(Awachie, 1966; Uznanski and Nickol, 1980; Pilecka-Rapacz,1986; Dezfuli et al., 2001; Poulin et al., 2003; Steinauer and Nick-ol, 2003), but the pervasiveness and magnitude of these phenom-ena are poorly known. Therefore, further empirical work is nec-essary to assess theoretical expectations and reach general conclu-sions concerning life history strategies of parasites in intermediatehosts.

Received 6 October 2006; revised 14 January 2007; accepted 15 Jan-uary 2007.

To elucidate the evolutionary forces shaping parasite growthin their intermediate hosts, the proximate factors affecting par-asite development must be understood. Resource availability,possibly represented by host size, condition, and/or growth, isone such factor presumably affecting the rate of parasite ontog-eny. Another factor is the number of conspecifics present. Theresponse of developing parasites to variable resource pools andinfection intensities can thus provide insight into the effects ofcompetition on parasite life history strategies. An experimentalinfection with an acanthocephalan (Acanthocephalus lucii) inits isopod intermediate host produced high infection intensities(Benesh and Valtonen, 2007b). This provided an opportunity toexamine parasite growth under conditions of high yet variablelevels of intraspecific competition.

MATERIALS AND METHODSAnimal collection, maintenance, and experimental infection

The collection site, maintenance of isopods, and experimental infec-tion protocol were described previously (Benesh and Valtonen, 2007b).Briefly, adult isopods (�5 mm) of the species Asellus aquaticus wereindividually exposed to European perch (Perca fluviatilis) feces con-taining acanthocephalan (A. lucii) eggs. The exposure was terminatedafter 10 days, and the course of infection was monitored over a periodof 101 days.

Data collection

Isopod molting was followed throughout the experiment, and the dateof observed molts was recorded. Isopods were checked daily to deter-mine survival. The sex and length of isopods were recorded upon death.Dead isopods were dissected, and parasites from infected isopods werecounted and measured. Parasites were examined with a compound mi-croscope, and measurements were taken using an ocular micrometer.During the early stages of the experiment, i.e., before 50 days or so,the length and width of individual worms was measured to the nearest0.004 mm. Later in the experiment, because worms were much larger,the length and width of individual worms were measured to the nearest0.01 mm. When possible, worms were sexed. After 101 days postex-posure (PE), all surviving isopods were killed and dissected.

Data analyses

The volume of individual worms was used as a measure of worm size.If worms were greater than or equal to 1 mm in length, they were consid-ered cylindrical in shape and their volume was calculated using the formula(�lw2)/4, where l is worm length and w is worm width. Worms less than1 mm were approximately ovoid in shape, and their volume was computedusing the formula (�lw2)/6. For each infected isopod, 4 variables were

BENESH AND VALTONEN—LARVAL A. LUCII GROWTH PATTERNS 743

→

FIGURE 1. Ln total worm volume (TWV; A), ln average worm vol-ume (AWV; B), and ln maximum worm volume (MWV; C) in cubicmillimeters plotted against days survival (n � 230). For all 3 infrapop-ulation characteristics, the logarithmic function (Y � b � ln[X]) pro-vided a better fit to the data than linear or quadratic functions. Theresiduals from each of the logarithmic functions were taken as relativevalues of TWV, AWV, and MWV adjusted for time.

calculated to characterize the infrapopulation, i.e., total worm volume(TWV), average worm volume (AWV), maximum worm volume (MWV),and the coefficient of variation in worm volume (CV, standard deviation ofworm volume divided by the AWV). TWV, AWV, and MWV were de-pendent on the age of infection because worms grew throughout the ex-periment. On the other hand, CV is relative to the AWV of the infrapop-ulation, so it is not necessarily dependent on infection age. To make allinfrapopulations comparable, it was necessary to ‘‘correct’’ values of TWV,AWV, and MWV for the age of infection. Each of these variables was lntransformed and plotted against time, i.e., days of survival. Linear, loga-rithmic, and quadratic functions were fitted to the data (Figs. 1A–C). Alogarithmic function, of the form Y � a � m � ln(X), provided the bestfit to the data for all 3 variables (lowest residual sum of squares and con-sequent P value). Therefore, the residuals from each of the 3 logarithmicfunctions were taken as values of TWV, AWV, and MWV independent ofinfection age. In all cases, residuals were normally distributed with ho-mogenous variance.

These 3 sets of residuals were used as dependent variables in a mul-tivariate analysis of covariance (MANCOVA). Also, ln-transformed CVwas included as a fourth dependent variable. To assess whether thefactors affecting infrapopulation characteristics changed over the courseof the experiment, a categorical variable was generated representing thetime isopods died and were dissected. Specifically, infrapopulationswere designated as being from isopods that died less than 40 days PE,called early infections, or from those that died more than 40 days PE,dubbed late infections. Day 40 was chosen as the splitting point becauseit effectively halved the data, creating nearly equal sample sizes in eachgroup. Along with infection age, host size was used as an independentfactor. Isopods were assigned to 3 size classes, small (�7.5 mm), in-termediate (7.5–8.5 mm), and large (�8.5mm), designated to have equalsample sizes within each category. Infection intensity and molting rate(see Benesh and Valtonen [2007b] for description and calculation ofmolting rate) were used as covariates. Interactions between infectionage and host size, infection age and intensity, and infection age andmolting rate were also included in the model to evaluate whether theeffects of these factors varied with time.

Host sex was initially included in the MANCOVA model, but it wasremoved from the final analysis because the equality of covariancesassumption was not met (Box’s test, P � 0.001). When there are dis-similar sample sizes in factor levels, as was the case for host sex, i.e.,the data were male biased (males: n � 174, females: n � 56), violationof this assumption can distort P values (Zar, 1999). Samples in thelevels of the other factors, by contrast, were similar. Thus, the modelexcluding host sex was considered more robust. In any case, when in-cluded in the model, host sex did not have any main effects on thedependent variables, and all the other results were qualitatively un-changed, so removal of this factor did not seem to affect conclusions.

Parasite growth patterns may differ between isopods that died (n �196) during the experiment and those that did not (n � 34). Thus, theMANCOVA analysis was also conducted on a data set excluding hoststhat survived 101 days. Also, in some isopods, shriveled, melanized,and obviously dead worms were observed alongside live worms. Be-cause dead worms cannot interact with live, cooccurring conspecifics,they may bias some infrapopulation characteristics. Therefore, theMANCOVA was rerun after excluding obviously dead worms from thedata, recalculating estimates of worm volume, and adjusting intensityto only include live worms. The MANCOVA was thus rerun twice,once excluding surviving isopods and once excluding dead worms. Inboth cases, the dependent variables, independent factors, and covariateswere the same as described above.

An additional confounder is that A. lucii is sexually dimorphic, i.e.,females are larger than males as both adults and larvae. Thus, female-

744 THE JOURNAL OF PARASITOLOGY, VOL. 93, NO. 4, AUGUST 2007

TABLE I. Summary of the MANCOVA analysis of 230 infrapopulationsin which all measured worms were used. The 4 dependent variablesused to characterize each infrapopulation were total worm volume(TMV), average worm volume (AWV), maximum worm volume(MWV), and the coefficient of variation in worm volume (CV). Infec-tion age, whether infrapopulations were observed before or after 40days PE, and isopod size, i.e., small, intermediate, or large, were cate-gorical factors. Molting rate and infection intensity were covariates. Anidentical analysis in which obviously dead worms were excluded fromthe data set produced the same results. Significant terms are indicatedin bold.

F df P

Infection age

TWV 0.082 1 0.775AWV 1.080 1 0.300MWV 2.962 1 0.087CV 9.973 1 0.002

Isopod size

TWV 1.308 2 0.272AWV 3.377 2 0.036MWV 1.127 2 0.326CV 1.866 2 0.157

Infection intensity

TWV 26.759 1 �0.001AWV 13.652 1 �0.001MWV 0.196 1 0.658CV 24.374 1 �0.001

Molting rate

TWV 10.832 1 0.001AWV 12.074 1 �0.001MWV 11.155 1 �0.001CV 0.083 1 0.773

Infection age � Molting rate

TWV 3.417 1 0.066AWV 2.788 1 0.096MWV 0.914 1 0.340CV 5.338 1 0.022

Infection age � Isopod size

TWV 3.206 2 0.042AWV 5.414 2 0.005MWV 2.342 2 0.099CV 4.766 2 0.009

Infection age � Intensity

TWV 23.008 1 �0.001AWV 12.912 1 �0.001MWV 15.945 1 �0.001CV 0.732 1 0.393

biased infrapopulations may have misleadingly high TWV, AWV,MWV, and CV. To check for this, the extent that sex ratio correlatedwith any of the 4 infrapopulation characteristics was determined. Analternative scenario in which worm sex ratio could bias results is if itchanges with time. For example, the sex ratio may shift toward a maleexcess during the experiment if isopods harboring more female worms,which may be more energetically demanding, die earlier. Therefore, thedegree that sex ratio changed over time was evaluated with regressionanalysis. Only infrapopulations in which more than half of the wormswere large enough to be sexed were used to investigate the potentialbiasing effects of sex ratio. In all tests, the log transformation of sexratio was used.

Alpha values less than 0.05 were considered statistically significant.SPSS 12.0.1 statistical software was used to conduct all analyses (SPSSInc., Chicago, Illinois).

RESULTS

Data from 3,202 worms in 230 infrapopulations were usedin the MANCOVA. This sample is less than the number ofinfected isopods (n � 248) for 2 reasons. First, in 9 infrapop-ulations it was possible to count the worms, but their conditionwas too poor for accurate measurements of worm size to bemade. Second, there were 9 isopods that harbored only 1 par-asite, so it was not possible to calculate the CV. The analysiswith the whole data set produced the same qualitative resultsas the analyses that either excluded surviving hosts or deadworms. Thus, only the results from analysis of the whole dataset are presented.

Dead worms were observed in 40 isopods. The average sizeof dead worms was 0.036 mm3, which is roughly the size ofearly acanthellae. The variance to mean ratio of intensity wasgreater for dead worms than for live worms, 7.14 vs. 4.29,suggesting dead worms were generally aggregated into fewerhosts than live worms. The sex ratio of dead worms was femalebiased, 1.96 �:1 � (112 females, 57 males), and a chi-squaretest indicated that this ratio was significantly different from1:1 ( � 17.90, P � 0.001).2�1

The sex ratio of all the sexed worms was slightly femalebiased at 1.1 �:1 � (785 females, 710 males). A chi-squaretest suggested this was a marginally significant departure fromthe expected 1:1 ratio ( � 3.76, P � 0.05). There was no2�1

relationship between worm sex ratio and any of the 4 infrapop-ulation characteristics (Pearson correlations, n � 97 in eachcase, all P � 0.094). Furthermore, infrapopulation sex ratio didnot change with time, i.e., days of survival (r � 0.03, P �0.77). Thus, sex ratio is considered to have little confoundingeffect on the analysis of larval infrapopulations.

The age of infection, late or early, had a significant effect onthe CV (Table I). Specifically, variation in worm size was greaterin late infections, i.e., after 40 days. There was no main effect ofinfection age on TWV, AWV, or MWV (Table I). Isopod size hada significant positive effect on AWV, but not on TWV, MWV, orCV (Table I). Infection intensity had different main effects on the4 dependent variables. Intensity was positively related to TWVand CV (Table I; Fig. 2A, D, respectively). In contrast, it had anegative effect on AWV (Table I; Fig. 2B). There was no maineffect of intensity on MWV (Table I; Fig. 2C).

There were significant interactions between intensity and in-fection age for 3 of the infrapopulation characteristics (TWV,AWV, and MWV), however, indicating the effects of intensitywere time dependent. TWV increased with intensity in earlyinfections, but not in late infections (Fig. 2A). Intensity did not

affect AWV in early infections, but in late infections a negativecorrelation between AWV and intensity was observed (Fig. 2B).In early infections, there was a slight positive relationship be-tween MWV and intensity, but in late infections this relation-ship was negative (Fig. 2C). For CV, there was not a significantinteraction between infection age and intensity, suggesting thatthe relationship between CV and intensity did not change withtime, i.e., it was consistently positive (Fig. 2D).

Molting rate was positively correlated with TWV, AWV, andMWV (Table I; Fig. 3A–C). Infection age and molting ratesignificantly interacted to determine CV (Table I). In late in-


FIGURE 2. The relationships between 4 infrapopulation characteristics and intensity. In the case of total worm volume (A), average wormvolume (B), and maximum worm volume (C), the data are relative to time, i.e., they are the residuals from the logarithmic regressions in Figure1A–C. The coefficient of variation in worm volume (D) is relative to average worm volume, so it is independent of time. Open circles representearly infections (infrapopulations observed before 40 days PE), whereas closed triangles depict late infections (infrapopulations observed after 40days PE). Dashed lines are fitted to the data from early infections, and solid lines correspond to late infections.

fections, the relationship between CV and intensity was slightlynegative, but in early infections it was weakly positive (Fig.3D). The interaction between infection age and molting ratewas not significant for TWV, AWV, or MWV (Table I). How-ever, for these 3 characteristics, the positive relationship withmolting rate seemed to be greater in late infections than in earlyinfections (Fig. 3A–C).

For TWV, AWV, and CV, there was a significant interactionbetween infection age and isopod size, suggesting the relation-ship between these infrapopulation characteristics and size de-pends on the time of the infection (Table I). In particular, TWVand AWV exhibit strong positive relationships with isopod sizelate in infections, but they do not seem to be affected by sizein early infections (Fig. 4A, B). MWV showed a similar trend


FIGURE 3. The relationships between 4 infrapopulation characteristics and molting rate, the number of molts per day survival. Total wormvolume (A), average worm volume (B), and maximum worm volume (C) data are relative to time, i.e., they are the residuals from the logarithmicregressions in Figure 1A–C. The coefficient of variation in worm volume (D) is relative to average worm volume, so it is independent of time.Open circles represent early infections (infrapopulations observed before 40 days PE), whereas closed triangles depict late infections (infrapop-ulations observed after 40 days PE). Dashed lines are fitted to the data from early infections, and solid lines correspond to late infections.

(Fig. 4C), but the interaction between infection age and isopodsize was not significant (Table I). In early infections, isopodsize had no effect on CV (Fig. 4D). For small and intermediatesize isopods, CV was much higher in late infections, but thiswas not the case for large isopods (Fig. 4D); CV in large iso-pods was similar before and after 40 days.

DISCUSSION

The relationship between time and the 3 measures of wormsize (TWV, AWV, and MWV) was not linear. Instead, it wasslightly curved, indicating that the relative rate of parasitegrowth decreased as the experiment progressed. Consequently,


FIGURE 4. The relationships between 4 infrapopulation characteristics and isopod size. In the case of total worm volume (A), average wormvolume (B), and maximum worm volume (C), the data are relative to time, i.e., they are the residuals from the logarithmic regressions in Figure1A–C. The coefficient of variation in worm volume (D) is relative to average worm volume, so it is independent of time. Open circles representmean values from early infections (infrapopulations observed before 40 days PE), whereas closed triangles depict the mean values from lateinfections (infrapopulations observed after 40 days PE). Bars represent standard error.

the impact of the factors affecting larval infrapopulations wastime dependent. For example, intensity had a positive main ef-fect on TWV, but a negative main effect on AWV. This con-forms to expectations; more worms should produce more wormbiomass, and the negative effect of intensity on AWV representsthe anticipated crowding effect observed in other acanthoceph-alans (Awachie, 1966; Dezfuli et al., 2001; Steinauer and Nick-

ol, 2003; Poulin et al., 2003). However, when the impact ofintensity is considered in the context of time, more complexpatterns emerge. Early on, higher intensities did yield higherTWV, but this relationship disappeared in late infections. Thissuggests that there is a threshold parasite biomass that can besupported by an individual isopod, regardless of the number ofparasites. For AWV, the expected crowding effect was ob-


served, but only in late infections. The average size of youngworms, however, did not seem to be affected by the presenceof conspecifics. The observation of crowding suggests that re-sources, or perhaps even space, are limited in late infectionswhen worms are larger. Temporal depletion of resources couldput worms under pressure to exploit resources aggressively ear-ly during development, since they may become more sparselater.

The consequences of this pressure may be manifested in therelationship between maximum worm volume (MWV), inten-sity, and time. In early infections, MWV was positively relatedto intensity. The presence of multiple competing conspecificsmay promote rapid initial exploitation of host resources whilethey are still abundant (Parker et al., 2003). This scramble com-petition may result in some parasites growing larger than mightbe expected. On the other hand, the positive association be-tween intensity and MWV could be entirely probabilistic; ifmore worms are measured, the chance of recording an abnor-mally large one increases. The negative correlation betweenMWV and intensity in late infections is probably analogous tothe situation with AWV in that it represents a crowding effectcaused by limited resources.

The impact of isopod size on parasite growth was also timedependent. Isopod size seemed to have little effect on A. luciisize in early infections, but TWV and AWV increased with hostsize in late infections. In early infections, the impact of hostsize on worm size may be negligible because resources arerelatively abundant, but as worms grow and resources dwindle,the energetic constraints imposed on parasites increase. Largerhosts may ease these constraints, thereby allowing more para-site growth. In several other systems, intermediate host size hasalso been recognized as an important determinant of acantho-cephalan size, at least for late developmental stages (Awachie,1966; Dezfuli et al., 2001; Steinauer and Nickol, 2003). MWVincreased with isopod size in late, but not early, infections.However, this trend was not significant, suggesting that hostsize, and by association resource availability, does not affectMWV to the same degree as TWV and AWV. The reason forthis is unclear, but it could represent the competitive superiorityof some individual parasites in that, regardless of the availableresource base, they can achieve a relatively large size.

Isopods that molted more frequently tended to harbor largerworms, suggesting the growth or physiological processes as-sociated with molting enhance parasite development. The pos-itive impact of host molting on parasite growth has not beenpreviously documented in other systems, though host growthseems to increase parasite growth in some systems (Barber,2005). Hosts that molt more presumably have higher food in-take, assuming molting is energetically costly, and may, there-fore, provide more resources to developing parasites. Thoughthe interactions were not significant, the impact of molting onparasite size seemed to be greater in late infections. This couldresult from the observation of more molts as time progressed,thereby making the correlation easier to detect in late infections.Another possibility is that, like the situation with isopod size,molting rate primarily affects parasite size in late infectionsbecause resources have become less abundant at this stage.

The variability in worm size increased with infection inten-sity. Some of this may be attributable to infrapopulation size;if more worms are observed, the likelihood of measuring ab-

normally large or small individuals increases. Some variationprobably also stemmed from age differences between parasites,which may range up to 10 days given the length of exposure.Much parasite size variation, however, likely derived from in-traspecific competition, i.e., individuals were not equal in theirability to acquire nutrients and grow. Competitive interactionsbetween developing parasites can take 2 forms, i.e., exploita-tion/indirect competition, or interference/direct competition. Inthe former case, competitive interactions are predicted to varywith resource abundance, whereas in the latter case competitionis expected to be largely independent of resources. Variation inworm size tended to increase as resources decreased, i.e., CVwas greater in late infections. Furthermore, CV was negativelyrelated to the 2 factors (molting rate and isopod size) likelyrepresenting resource availability for developing parasites.These factors only affected CV in late infections when resourc-es were presumably constrained, suggesting that competitionbetween larval parasites may be diffused in larger or fastergrowing hosts. Variation in worm size, thus, seems to vary withresource availability, which suggests that exploitation compe-tition, as opposed to interference, is primarily operating amonggrowing A. lucii.

Dead parasites, almost exclusively young acanthellae, oc-curred in some infrapopulations. Dead acanthocephalans intheir intermediate hosts, particularly early developmental stag-es, have been reported previously (e.g., Hynes and Nicholas,1958; Crompton, 1967; Robinson and Strickland, 1969; Nickoland Dappen, 1982; Gleason, 1989). Because the larval mortalityof A. lucii, as well as other species, seems to be greater duringearly development, the time spent in these ‘‘risky’’ stagesshould be minimized. Thus, the high initial relative growth ratemay be advantageous, even though it might negatively affecthost survival (Benesh and Valtonen, 2007b). There are 2 non-exclusive possible explanations for the size-specific mortality.First, the host immune response is most effective at killingyoung acanthellae, or, second, the parasites are heavily con-strained by resources at this stage. The overrepresentation offemales among the dead worms suggests that if females requiremore energy to develop, limiting resources are the cause ofworm death. However, the aggregation of dead worms into rel-atively few hosts also could imply that some isopods have anespecially efficient immune response against A. lucii.

Another unexpected result was that the A. lucii sex ratio wasslightly female biased. Given that the departure from 1:1 wasslight and the sex ratio of acanthocephalans from natural andlaboratory infections of intermediate hosts is generally 1:1(Crompton, 1985), this result must be interpreted with caution.Despite the apparent female bias in the component population,sex ratio did not seem to affect the characteristics of infrapop-ulations. The rather high parasite burdens may have dampenedany biasing effects of sex ratio on worm sizes.

All the proximate factors (host size, molting rate, and inten-sity) affecting A. lucii growth in some way revolved aroundresource availability, though the relevance of resources seemedto change over time. Interestingly, the temporal changes in re-source availability and parasite growth rates may affect hostsurvival. Infected isopods had reduced survival, relative to un-infected isopods, during the early stages of A. lucii develop-ment, but not at later stages of infection (Benesh and Valtonen,2007b). Though the mechanism underlying this mortality is un-


known, it is possible that relatively rapid and apparently re-source-independent growth by young parasites is worse for hostviability than the slower, resource-limited growth of larger par-asites. This goes against some hypothetical expectations in thatheavier parasite burdens (total parasite mass) are assumed toresult in greater host mortality (Parker et al., 2003). The nullrelationship between total worm volume and intensity in lateinfections suggests that developing A. lucii approached and/orreached an upper size limit defined by each individual host, yetthis did not result in lower host survival. The growth patternsof several pseudophyllidean cestode species, on the other hand,suggest that parasites are not exploiting host resources at max-imum levels, possibly reflecting adaptive life history strategies(Parker et al., 2003; Michaud et al., 2006). Such submaximalgrowth may also occur in single A. lucii infections, particularlyfor male parasites (Benesh and Valtonen, 2007a), leaving openthe possibility that larval growth patterns are not exclusivelydetermined by host resources (Parker et al., 2003). In any case,identifying the proximate factors impacting the growth of dif-ferent parasites, such as resource availability, host defenses, andhost mortality, and evaluating them within a comparativeframework may reveal the selective forces shaping the evolu-tion of parasite life history strategies.

ACKNOWLEDGMENTS

T. Hasu graciously shared her experiences with isopods in the labo-ratory and gave insightful criticisms on a draft of this manuscript. EeroVestola helped in the maintenance of isopods. D.P.B. was provided fi-nancial support by Fulbright and CIMO grants.

LITERATURE CITED

AWACHIE, J. B. E. 1966. The development and life-history of Echino-rhynchus truttae Schrank 1788 (Acanthocephala). Journal of Hel-minthology 40: 11–32.

BARBER, I. 2005. Parasites grow faster in faster growing fish hosts.International Journal for Parasitology 35: 137–143.

BENESH, D. P., AND E. T. VALTONEN. 2007a. Sexual differences in larvallife history traits of acanthocephalan cystacanths. InternationalJournal for Parasitology 37: 191–198.

———, AND ———. 2007b. Effects of Acanthocephalus lucii (Acan-thocephala) on intermediate host survival and growth: Implicationsfor exploitation strategies. Journal of Parasitology 93: 735–741.

CROMPTON, D. W. T. 1967. Studies on the haemocytic reaction of Gam-marus spp., and its relationship to Polymorphus minutus (Acantho-cephala). Parasitology 57: 389–401.

———. 1985. Reproduction. In Biology of the Acanthocephala, B. B.

Nickol and D. W. T. Crompton (eds.). Cambridge University Press,Cambridge, U.K., p. 213–272.

DEZFULI, B. S., L. GIARI, AND R. POULIN. 2001. Costs of intraspecificand interspecific host sharing in acanthocephalan cystacanths. Par-asitology 122: 483–489.

FREDENSBORG, B. L., AND R. POULIN. 2005. Larval helminthes in inter-mediate hosts: Does competition early in life determine the fitnessof adult parasites? International Journal for Parasitology 35: 1061–1070.

GLEASON, L. N. 1989. Movement of Pomphorhynchus bulbocolli larvaefrom the hemocoel to the peripheral circulation of Gammarus pseu-dolimnaeus. Journal of Parasitology 75: 982–985.

HYNES, H. B. N., AND W. L. NICHOLAS. 1958. The resistance of Gam-marus spp. to infection by Polymorphus minutus (Goeze, 1782)(Acanthocephala). Annals of Tropical Medicine and Parasitology52: 376–383.

LAFFERTY, K. D., AND A. M. KURIS. 2002. Trophic strategies, animaldiversity, and body size. Trends in Ecology and Evolution 17: 507–513.

MICHAUD, M., M. MILINSKI, G. A. PARKER, AND J. C. CHUBB. 2006.Competitive growth strategies in intermediate hosts: Experimentaltests of a parasite life-history model using the cestode, Schistoce-phalus solidus. Evolutionary Ecology 20: 39–57.

NICKOL, B. B., AND G. E. DAPPEN. 1982. Armadillidium vulgare (Iso-poda) as an intermediate host of Plagiorhynchus cylindraceus(Acanthocephala) and isopod response to infection. Journal of Par-asitology 68: 570–575.

PARKER, G. A., J. C. CHUBB, G. N. ROBERTS, M. MICHAUD, AND M.MILINSKI. 2003. Optimal growth strategies of larval helminths intheir intermediate hosts. Journal of Evolutionary Biology 16: 47–54.

PILECKA-RAPACZ, M. 1986. On the development of acanthocephalans ofthe genus Acanthocephalus Koelreuther, 1771, with special atten-tion to their influence on intermediate host, Asellus aquaticus L.Acta Parastologica Polonica 30: 233–350.

POULIN, R. 1998. Evolutionary ecology of parasites. Chapman and Hall,London, U.K, 212 p.

———, K. NICHOL, AND A. D. M. LATHAM. 2003. Host sharing andhost manipulation by larval helminthes in shore crabs: Cooperationor conflict? International Journal for Parasitology 33: 425–433.

ROBINSON, E. S., AND B. C. STRICKLAND. 1969. Cellular responses ofPeriplaneta americana to acanthocephalan larvae. ExperimentalParasitology 50: 694–697.

ROSEN, R., AND T. A. DICK. 1983. Development and infectivity of theprocercoid of Triaenophorus crassus Forel and mortality of the firstintermediate host. Canadian Journal of Zoology 61: 2120–2128.

STEINAUER, M. L., AND B. B. NICKOL. 2003. Effect of cystacanth bodysize on adult success. Journal of Parasitology 89: 251–254.

UZNANSKI, R. L., AND B. B. NICKOL. 1980. A sequential ranking systemfor developmental stages of an acanthocephalan, Leptorhynchoidesthecatus, in its intermediate host, Hyalella azteca. Journal of Par-asitology 66: 506–512.

ZAR, J. H. 1999. Biostatistical analysis. Prentice-Hall, Upper SaddleRiver, New Jersey, 929 p.

proximate factors affecting the larval life history of acanthocephalus lucii (acanthocephala)

Documents