J. Parasitol., 93(4), 2007, pp. 735–741 䉷 American Society of Parasitologists 2007 EFFECTS OF ACANTHOCEPHALUS LUCII (ACANTHOCEPHALA) ON INTERMEDIATE HOST SURVIVAL AND GROWTH: IMPLICATIONS FOR EXPLOITATION STRATEGIES Daniel P. Benesh and E. Tellervo Valtonen Department of Biological and Environmental Science, P.O. Box 35, FI-40014 University of Jyväskylä, Finland. e-mail: dabenesh@cc.jyu.fi ABSTRACT: Intermediate host exploitation by parasites is presumably constrained by the need to maintain host viability until transmission occurs. The relationship between parasitism and host survival, though, likely varies as the energetic requirements of parasites change during ontogeny. An experimental infection of an acanthocephalan (Acanthocephalus lucii) in its isopod intermediate host (Asellus aquaticus) was conducted to investigate host survival and growth throughout the course of parasite development. Individual isopods were infected by exposure to fish feces containing parasite eggs. Isopods exposed to A. lucii had reduced survival, but only early in the infection. Mean infection intensity was high relative to natural levels, but host mortality was not intensity dependent. Similarly, a group of naturally infected isopods harboring multiple cystacanths did not have lower survival than singly infected isopods. Isopods that were not exposed to the parasite exhibited sexual differences in survival and molting, but these patterns were reversed or absent in exposed isopods, possibly as a consequence of castration. Further, exposed isopods seemed to have accelerated molting relative to unexposed controls. Infection had no apparent effect on isopod growth. The effects of A. lucii on isopod survival and growth undermine common assumptions concerning parasiteinduced host mortality and the resource constraints experienced by developing parasites. For parasites with complex life cycles, intermediate hosts serve a variety of functions, 2 of the most important being that (1) they are an energy resource for parasites, allowing growth and in some cases asexual reproduction, and (2) they are vessels for transmitting parasites to the next host in the life cycle, commonly through predation (Parker, Chubb, Ball, and Roberts, 2003). These 2 roles of intermediate hosts presumably impose conflicting selective pressures on parasites. Parasite growth and maturation requires consumption of host resources, and heavy exploitation of intermediate hosts may benefit parasites by increasing growth rates, production of infective stages, and/or adult success. However, damage done is likely constrained by the need to keep the intermediate host alive long enough to mature and be successfully transmitted to the next host. Therefore, the optimal level of host exploitation (virulence by some definitions) is expected to reflect a tradeoff between the benefits of resource consumption and the costs of reduced host viability (Ebert and Herre, 1996; Poulin, 1998). There are examples of parasite-induced reductions in intermediate host viability within all major helminth groups, e.g., nematodes (Ashworth et al., 1996), trematodes (Sorensen and Minchella, 2001), cestodes (Rosen and Dick, 1983), and acanthocephalans (Hynes and Nicholas, 1958). On the other hand, there are also systems in which neither parasitism nor infection intensity affects host survival (e.g., Uznanski and Nickol, 1980; Wedekind, 1997; Hurd et al., 2001). Thus, the host exploitation strategies employed by larval helminths seem to vary widely. Recognizing which factors mediate the relationship between parasitism and intermediate host survival in different systems is a necessary step toward explaining this variability. For instance, host sex (Shostak et al., 1985), host condition (Krist et al., 2004), and infection intensity (Fredensborg et al., 2004) may all exacerbate or buffer parasite-induced mortality. Moreover, the probability of host death can vary with parasite development (Shostak et al., 1985; Schjetlein and Skorping, 1995; Sorensen and Minchella, 1998; Duclos et al., 2006). The aim of this study was to examine intermediate host exploitation, as well as some potential factors influencing it, by an acanthocephalan. The host-parasite system examined was Acanthocephalus lucii and its isopod intermediate host, Asellus aquaticus. Acanthocephalus lucii exhibits a typical acanthocephalan life cycle (Schmidt, 1985); the definitive host is a vertebrate, commonly European perch (Perca fluviatalis), and an arthropod, in this case an isopod (A. aquaticus), serves as intermediate host. Adult worms mate in the intestine of fish, and females release eggs into the environment with the feces. Isopods become infected by ingesting eggs. The parasite develops in the isopod to the infective cystacanth stage, and the life cycle is completed when an infected isopod is eaten by an appropriate definitive host. Brattey (1986) found that isopods exposed to A. lucii eggs experienced higher mortality than unexposed isopods. Conversely, Hasu et al. (2006) observed that infected isopods actually survive better than controls. Given this contradictory evidence, additional investigation into the consequences of A. lucii infection on intermediate host survival seems warranted. The specific goals of this study were (1) to assess how A. lucii infection affects isopod survival throughout parasite development and (2) to evaluate whether A. lucii affects host growth, or molting behavior, or both. MATERIALS AND METHODS Animal collection and maintenance Isopods were collected at the end of August 2005 with a dip net from Niemijärvi, a small pond in central Finland (62⬚12⬘N, 25⬚45⬘E) in which the only fish species present is Carassius carassius, the crucian carp. Thus, all isopods were uninfected because the definitive host of the parasite is not present in the pond. In the laboratory, isopods were fed on a diet of leaves, primarily of alder (Alnus glutinosa). Leaves were conditioned in aerated lake water for at least 2 wk to allow microbial colonization prior to being offered to isopods. Animals were maintained at approximately 18 C under constant illumination. Experimental infection To collect A. lucii eggs, perch were isolated in a large tank during August 2005. Water flow through the tank was set at a very low level to allow feces accumulation. Perch feces were collected and stored in refrigerated lake water. Owing to mortality and the subsequent addition of new individuals, the number of fish in the tank fluctuated with time, but there were usually more than 50 fish. Some of these perch were dissected (n ⫽ 93, mean length ⫽ 12.99 cm, standard deviation [SD] ⫾ 2.4), and A. lucii prevalence was 75% with a mean infection intensity of 4.8 (SD ⫾ 4.8). Some of the collected feces were examined with a light microscope to confirm the presence of mature eggs. Received 6 October 2006; revised 14 January 2007; accepted 15 January 2007. 735 736 THE JOURNAL OF PARASITOLOGY, VOL. 93, NO. 4, AUGUST 2007 Individual isopods larger than 5 mm were isolated in plastic containers, 10 ⫻ 15 ⫻ 5 cm, containing 400 ml of lake water. Individuals were either exposed to fish feces (n ⫽ 339) or sham exposed with distilled water (n ⫽ 185). All the collected feces were combined in 5-L lake water to make a single suspension. For the exposed isopods, the fecal suspension was vigorously mixed before 10 ml was added to each container. Ten milliliters of distilled water was added to each container with unexposed, control isopods. Fish feces were left in the containers for 10 days before the exposure was terminated by removing all remaining feces. Control isopods were given conditioned leaves 5 days after being sham exposed. After 10 days and until the end of the experiment, all isopods were maintained on conditioned leaves. Every 5 days, the water in each container was replaced with fresh, aerated lake water. Data collection Molting was monitored throughout the experiment for all isopods. When shed exoskeletons were observed, they were removed from the containers and the date was recorded. Isopods were checked daily to determine survival. Dead isopods were sexed by examining their pleopods, and their length was measured to the nearest 0.5 mm. In A. aquaticus, male antenna length is a secondary sexual character (Bertin and Cezilly, 2003), and it was also recorded from dead individuals. Dead isopods were dissected to determine infection status, and parasites from infected isopods were counted. The experiment was terminated 101 days after the initial exposure. Naturally infected isopods The survival of naturally infected isopods was investigated to determine whether conclusions based on the experimental infection also seem applicable to natural infections. Naturally infected and uninfected isopods were collected in October 2005 from Lake Jyväsjärvi, Central Finland (62⬚14⬘N, 25⬚44⬘E). Isopods infected with A. lucii cystacanths have darkened respiratory opercula (Brattey, 1983), so they can be identified relatively easily. However, because only isopods infected with cystacanths can be recognized and collected, the effects of earlier acanthocephalan ontogeny on host survival cannot be determined. Instead, monitoring the survival of isopods harboring cystacanths allows the long-term effects of parasitism on host survival to be evaluated. Uninfected and infected isopods were isolated individually in plastic containers and maintained in a manner identical to the experimentally exposed isopods. Isopod size, sex, antenna length, and infection status, i.e., presence/absence of A. lucii and cystacanth intensity, were recorded upon isopod death. This group of isopods was observed for 75 days before surviving isopods were killed and dissected. Data analyses Survival: The factors affecting isopod survival in the experimental infection were assessed with Cox regression, a method commonly used for survival analyses (Andersen, 1991). An assumption of Cox regression models is that the ratio of the hazard function for any 2 individuals is dependent on their covariate values and the baseline hazard function, but not time (proportional hazards assumption). This assumption appeared to be violated, however, because some factor effects seemed to vary with time. Thus, a time-dependent covariate was incorporated into the model. Exposure (unexposed vs. exposed to A. lucii) and isopod sex were included in the model as categorical covariates. For each individual, the lengths of the first antennae were averaged and used as the dependent variable in a linear regression with isopod size as the independent variable. The residuals of this regression, which are a measure of antenna length corrected for body size, were used as a covariate in the model. A separate Cox regression was conducted with only exposed isopods to examine the factors affecting the survival of infected isopods. Infection intensity and antenna length, corrected for isopod size, were included as covariates. Isopod sex was incorporated into the model as a categorical covariate. As before, a time-dependent covariate was added to the model. A third Cox regression was used to evaluate the factors affecting survival of naturally infected isopods. Three categories of isopods were defined: (1) uninfected (n ⫽ 64); (2) isopods with a single cystacanth (n ⫽ 41); and (3) isopods with more than 1 cystacanth (n ⫽ 17; mean intensity ⫽ 2.4). Infection status was incorporated into the model as a categorical covariate. Additionally, isopod sex and relative antenna length were included in the model. The terms included in each of the 3 Cox regression models were evaluated using both forward and backward selection procedures. Isopod growth and molting: Isopod growth was evaluated with an analysis of covariance (ANCOVA) using isopod size as the dependent variable, exposure as an independent factor, and time, i.e., days of survival, as a covariate. An identical analysis was conducted to test whether naturally infected and uninfected isopods differed in growth. Naturally infected isopods harboring either a single or multiple cystacanths were pooled for this analysis. Molting was only recorded for the isopods from the experimental infection. It was considered to have 2 components, frequency and timing. That is, an isopod molted a particular number of times during the experiment, and those molts transpired during a given number of days after exposure. Both the quantity and timing of molts was dependent on the length of time individual isopods were observed, i.e., days of survival. Thus, molting frequency was made relative by dividing the number of times an isopod molted by the number of days it survived. This value will be referred to as the molting rate. Timing of molts was summarized by calculating the average days postexposure (PE) on which an isopod molted. Specifically, for individual isopods, the days PE on which each observed molt occurred were summed and divided by the total number of molts. This value was made relative by dividing it by the number of days an isopod survived, and will henceforth be referred to as molt timing. Two variables were thus used to summarize the molting behavior of individual isopods, number of molts per day survival, i.e., molting rate, and the average day PE on which molting occurred divided by days of survival, i.e., molt timing. To examine whether exposed and unexposed isopods differed in molting behavior, molting rate and molt timing were used as dependent variables in multivariate analysis of variance (MANOVA). Molting rate was ln transformed to homogenize variance. Exposure and isopod sex were used as independent factors in the analysis. When all the isopods that had molted at least once were included in the analysis, the assumption of equal covariance was violated (Box’s test, P ⬍ 0.001). This was considered problematic because more male isopods were sampled than female isopods, and, when covariance matrices are heterogeneous, biased sample sizes can distort P values (Zar, 1999). However, when only isopods that molted at least twice were used in the analysis, the null hypothesis of equal covariance was not rejected (P ⫽ 0.26). Thus, the MANOVA only included isopods that molted at least twice. All analyses were conducted with SPSS 12.0.1 (SPSS Inc., Chicago, Illinois) statistical software, and alpha values less than 0.05 were considered statistically significant. Untransformed values are shown in figures. RESULTS Survival of all isopods After 101 days, 427 of 524 isopods (82%) had died. Antenna length, corrected for body size, had a weak, positive effect on survival, i.e., the hazard function decreased 6% for each unit increase in residual antenna length (Table I). There was a significant interaction between exposure and the time-dependent covariate, suggesting the baseline hazard function differed between unexposed and exposed isopods (Table I). Specifically, exposed isopods seem to have lower survival during the initial 40 days or so of the experiment (Fig. 1). After approximately 40 days, the survival functions of unexposed and exposed isopods appear quite similar (Fig. 1). Male isopods survived slightly longer than female isopods, on average (47 vs. 44 days). However, the effects of isopod sex on survival were both time and exposure dependent (Table I). For unexposed isopods, females exhibited higher mortality than males during the middle of the experiment (between days 20 and 50 approximately; Fig. 2A). In contrast, exposed male isopods had slightly greater mor- BENESH AND VALTONEN—HOST EXPLOITATION BY A. LUCII 737 TABLE I. Terms included and excluded from the final model produced by Cox regression survival analysis of all isopods (n ⫽ 524). Forward and backward selection procedures gave the same best model. Odds ratio indicates the ratio of the estimated survival probabilities of contrasted classes, e.g., the survival probability of male isopods was estimated to be 1.4 times higher than female isopods. Terms Wald score df P Odds ratio Terms in the final model Antenna length corrected for body size Isopod sex Exposure ⫻ time-dependent covariate Isopod sex ⫻ exposure ⫻ time-dependent covariate 5.29 4.13 50.33 7.95 1 1 1 1 0.021 0.042 ⬍0.001 0.005 0.94 1.37 0.00 0.19 0.45 1 1 1 1.00 0.89 0.83 Terms excluded from the final model Exposure Isopod sex ⫻ time-dependent covariate Isopod sex ⫻ exposure tality than females during the early part of the experiment (Fig. 2B). males tended to have lower survival early in the experiment (Fig. 2B). Survival of exposed isopods Survival of naturally infected isopods Around 16 days PE, parasites were large enough for infection intensity to be reliably quantified. Thus, the analysis of exposed isopod survival only included individuals that lived at least 16 days (n ⫽ 252). Infection prevalence was 98% with only 4 of 252 isopods uninfected, and the mean infection intensity was 13.5 (SD ⫾ 8.8) parasites. Contrary to expectations, infection intensity had a significant positive effect on survival (Table II). Furthermore, there was a significant interaction between infection intensity and the time-dependent covariate, suggesting the effects of intensity on survival varied with time (Table II). This interaction is perhaps best exemplified by the fact that the relationship between intensity and survival was better described by a curve than a line (F1,250 ⫽ 22.07, P ⬍ 0.01). Intensity had a small, positive effect on survival initially, but, as the experiment progressed, intensity had negligible impact on survival (Fig. 3). There was also a significant interaction between isopod sex and the time-dependent covariate (Table II), since exposed Only 39 of 122 isopods (32%) died during the 75-day observation period. The survival of these isopods was not affected by infection status, isopod sex, relative antenna length, or any of their interactions (all P ⬎ 0.18). FIGURE 1. Cumulative survival of 185 unexposed and 339 exposed isopods over the course of 101 days. Isopod growth and molting For the experimental isopods, there was a significant positive relationship between time (days survival) and isopod size (F1,520 ⫽ 537.7, P ⬍ 0.001). The intercept (exposure in ANCOVA, F1,520 ⫽ 0.47, P ⫽ 0.49) and slope (exposure ⫻ time interaction, F1,520 ⫽ 0.001, P ⫽ 0.97) of this correlation were similar for unexposed and exposed isopods, suggesting their pattern of growth did not differ. Similarly, the relationship between time and isopod size was positive for uninfected and naturally infected isopods (F1,118 ⫽ 135.3, P ⬍ 0.001), but intercept and slope parameters were unaffected by infection status (infection status, F1,118 ⫽ 1.49, P ⫽ 0.23; infection status ⫻ time, F1,118 ⫽ 0.59, P ⫽ 0.44). Of the 524 isopods observed, 394 (75%) molted at least once during the experiment. The number of times an isopod molted ranged from 1 to 7, with an average of 1.8 molts per isopod. For those isopods molting multiple times, the average interval between molts was 17.9 (SD ⫾ 7.6) and 23.6 (SD ⫾ 11.3) days for exposed (n ⫽ 122) and unexposed (n ⫽ 85) isopods, respectively. The intermolt period did not change over the course of the experiment for either exposed or unexposed isopods, e.g., the interval between the first and second molt did not significantly differ from the interval between the second and third molt, etc. (paired t-tests, all P ⬎ 0.1). Exposure had no effect on molting rate (F1,203 ⫽ 0.17, P ⫽ 0.68), but isopod sex did (F1,203 ⫽ 6.35, P ⫽ 0.01). On average, females tended to have a higher molting rate than males. However, there was a significant interaction between exposure and isopod sex (F1,203 ⫽ 7.79, P ⫽ 0.01). Unexposed females had a higher molting rate than unexposed males, whereas exposed male and female isopods had very similar molting rates (Fig. 4A). Furthermore, unexposed females molted more frequently than exposed isopods, while unexposed males molted less frequently than exposed isopods (Fig. 4A). 738 THE JOURNAL OF PARASITOLOGY, VOL. 93, NO. 4, AUGUST 2007 FIGURE 3. The relationship between intensity and isopod survival. A curvilinear (quadratic) function (Y ⫽ ⫺2.59 ⫹ 0.57x ⫹ ⫺0.004x2; R2 ⫽ 0.21) provided a better fit to the data than a line (Y ⫽ 7.4 ⫹ 0.12x; R2 ⫽ 0.14). ⫽ 3.56, P ⫽ 0.06), unexposed females tended to molt later than all other isopods (Fig. 4B). Molt timing did not differ between exposed males and females (Fig. 4B). DISCUSSION FIGURE 2. Cumulative survival of male and female isopods that were either (A) unexposed or (B) exposed to Acanthocephalus lucii. Exposed isopods molted sooner than unexposed isopods (F1,203 ⫽ 9.66, P ⬍ 0.01), but isopod sex did not affect molt timing (F1,203 ⫽ 2.85, P ⫽ 0.09). Although the interaction between isopod sex and exposure was not quite significant (F1,203 Parasites are generally assumed to decrease the survival of their intermediate hosts. In accordance with this premise, increasing infection intensity should negatively affect host survival. Isopods exposed to A. lucii experienced a higher mortality rate than control isopods but only during the early part of the infection. Furthermore, survival did not have the expected negative relationship with intensity; if anything, it was positive. Thus, some of the common assumptions concerning host-parasite interactions may not apply to this system. Brattey (1986) found isopods exposed to A. lucii to have consistently reduced survival during 60 days of observation. The exposed isopods observed here also had lower survival, but it was not consistently lower; late in the infection exposed and unexposed isopods exhibited similar survival. In contrast, Hasu et al. (2006) found infected isopods to survive better than TABLE II. Terms included and excluded from the final Cox regression survival analysis of exposed isopods which lived at least 16 days (n ⫽ 252). Forward and backward selection procedures gave the same best model. Terms Terms in the final model Infection intensity Infection intensity ⫻ time-dependent covariate Isopod sex ⫻ time-dependent covariate Terms excluded from the final model Antenna length corrected for body size Isopod sex Isopod sex ⫻ infection intensity Isopod sex ⫻ infection intensity ⫻ time-dependent covariate Wald score df P Odds ratio 44.99 29.58 4.92 1 1 1 ⬍0.001 ⬍0.001 0.027 0.87 0.97 1.24 0.37 0.13 1 1 1 1 0.33 0.27 0.54 0.72 BENESH AND VALTONEN—HOST EXPLOITATION BY A. LUCII FIGURE 4. (A) The number of molts per day survival (molting rate) of exposed and unexposed isopods of each sex. Molt timing (B), defined as the average day of molting divided by the number of days survival, of exposed and unexposed isopods of each sex. Only isopods that molted at least twice were included (n ⫽ 207). Bars represent standard error. controls. However, their study focused exclusively on gravid, female isopods, which may differ in their response to infection. Nickol (1985) hypothesized that early mortality in laboratory acanthocephalan infections results from massive penetration of acanthors through the intermediate host gut. This could explain some of the initial mortality of exposed isopods, but additional mechanisms seem necessary to explain the depressed survivorship of exposed isopods during the first few weeks of infection. 739 The relative growth rate of A. lucii is nonlinear; parasite volume increases rapidly initially, but the growth rate slows over time (Benesh and Valtonen, 2007). Therefore, later infections may be less energetically burdensome for the host, resulting in lower mortality. The similar survival rates of cystacanth-infected and uninfected isopods collected from nature support this notion. Periods of rapid parasite growth have been observed to be detrimental to intermediate host viability in other systems as well (Shostak et al., 1985; Duclos et al., 2006). Alternatively, given that intermediate hosts respond immunologically to acanthocephalan infections (Robinson and Strickland, 1969; Nickol and Dappen, 1982), host defenses and their associated costs could also play a role in reducing host survival during early infections (Hasu et al., 2006). Surprisingly, antenna length relative to body size had a weak positive effect on survival. The mechanism underlying the relationship between mortality and shorter antennae is not clear, but it might reflect the number of antagonistic conflicts experienced by an isopod, or it may be a consequence of immunological costs associated with wounding (Plaistow et al., 2003). Contrary to expectations, intensity positively affected survival, particularly early in the experimental infection. A common explanation for parasite-induced increases in host survival is that parasites divert host resources away from reproduction (Hurd et al., 2001; Sorensen and Minchella, 2001). Like several other Acanthocephalus species (Oetinger, 1987; Dezfuli et al., 1994; Kakizaki et al., 2003), A. lucii is a castrating parasite, at least in female hosts (Brattey, 1983). Whether castration is responsible for the apparent positive effect of intensity in early infections is unknown, and better knowledge of the mechanism and time course involved in resource reallocation would be needed to assess this phenomenon. It should be noted that considerable variation was observed in the relationship between intensity and survival; the quadratic function in Figure 3 explained only 21% of the variation in this association. Consequently, any statements about the effects of intensity on survival in this system are tenuous. Nonetheless, the survival of both experimentally and naturally infected isopods did not decrease with infection intensity, as expected. Other acanthocephalan species similarly fail to conform to the anticipated negative relationship between intensity and host viability (Lackie, 1972; Uznanski and Nickol, 1980). The parasite burdens in this study were far higher than those found in nature; the majority of infected isopods collected from nature harbor a single parasite (Brattey, 1986). Consequently, the results concerning host survival and the effects of intensity may not reflect the situation in nature. On the other hand, some of the trends observed in the experimentally infected isopods were also observed in the naturally infected isopods, e.g., lack of intensity-dependent mortality and similar growth patterns between uninfected and infected isopods. Thus, the observed effects of A. lucii infection on isopod viability seem likely to operate at lower intensities as well. Additionally, some acanthocephalan populations exhibit very high infection levels in their intermediate hosts (Seidenberg, 1973; Gleason, 1987; Brown and Pascoe, 1989; Sparkes et al., 2004), so the consequences of heavy infections may be relevant to some natural situations. Sexual differences in isopod mortality depended on exposure to A. lucii. For unexposed isopods, females suffered higher 740 THE JOURNAL OF PARASITOLOGY, VOL. 93, NO. 4, AUGUST 2007 mortality than males, but this trend was reversed for exposed isopods; males had slightly higher mortality than females. Unexposed females molted more frequently and at a later point in the experiment than unexposed males, which could reflect attainment of sexual maturity because female A. aquaticus, like all freshwater isopods, must undergo a nuptial molt to become sexually mature (Jormalainen, 1998). Indeed, late in the experiment, some control females were observed bearing eggs in their marsupium. Thus, the lower survival of unexposed females may stem from resource allocation toward reproduction. Exposed females, on the other hand, were likely castrated as a consequence of A. lucii infection (Brattey, 1983), so energy potentially dedicated to host reproduction was perhaps available for host maintenance instead. This might explain the slightly greater survival of exposed females in the first few weeks of the experiment. The difference in survival of exposed male and female isopods was not a result of dissimilar infection intensities, since the sexes appear rather similar in susceptibility to A. lucii (Brattey, 1986; Hasu et al., 2007). Host castration may also account for the similar molting pattern exhibited by exposed males and females; inability to achieve sexual maturity could disrupt normal, sex-specific molting behavior. Exposed isopods tended to molt slightly more frequently than unexposed, male isopods. Additionally, they tended to molt earlier during the experiment than both male and female unexposed isopods. These results suggest that infection with A. lucii increases host molting. Previous studies have found molting to be inhibited (Thomas et al., 1996; Kokkotis and McLaughlin, 2006) or unaffected by parasites (Calado et al., 2005). The accelerated molting of exposed isopods did not seem to translate into a larger size at the end of the experiment, suggesting that molting may not reflect isopod growth (Marcus, 1990). Interestingly, the size of A. lucii was positively related to molting rate, indicating that increased host molting is favorable for parasite growth and may, therefore, represent adaptive manipulation (Benesh and Valtonen, 2007). Alternatively and/or additionally, earlier molting may serve the purpose of repairing damage done to the intestine by invading acanthors. Early in the experiment, many exposed isopods shed exoskeletons that included intestines still packed with food, suggesting that molting was premature and it was either induced, or necessary, or both. This was never observed in control isopods. Beyond simply mending damage done to the intestine, accelerated molting could reduce parasite burden by removing young acanthors attached to the intestine. However, since neither molting rate nor timing seemed to affect intensity (data not shown), this hypothesis was not supported. The relationship between isopod size and time did not differ between uninfected and infected isopods, so the initial size and subsequent growth of these groups were very similar. In contrast, juvenile isopods, smaller than those used here, seem to exhibit accelerated growth after exposure to A. lucii (Hasu et al., 2007). Thus, the effects of A. lucii on host growth may be dependent on the size of the host at the time of infection. Some acanthocephalans reduce the growth of their intermediate hosts (Hynes and Nicholas, 1958; Awachie, 1966), but others do not (Uznanski and Nickol, 1980). Thus, in a situation analogous to survival, the effects of acanthocephalans on host growth appear to vary between species. This has important consequences for parasite development and exploitation strategies. For example, if host growth is not obstructed, the resources available to developing parasites are not temporally constrained; there may always be an influx of resources permitting additional parasite growth. The size of several acanthocephalan species, including A. lucii (Benesh and Valtonen, 2007), is positively correlated with host size (Awachie, 1966; Dezfuli et al., 2001; Steinauer and Nickol, 2003), so it would seem advantageous not to hamper host growth. The impact of A. lucii on its intermediate host’s viability, growth, and molting undermines common assumptions concerning host-parasite interactions. First, the effects of A. lucii on host mortality were not consistent over time, so the presumed tradeoff between host exploitation and host viability may primarily operate at particular times of parasite ontogeny. Second, host mortality was not intensity dependent. Third, the theoretical, resource-based, maximum size attainable by parasites in their intermediate hosts (Parker, Chubb, Roberts et al., 2003) may not be temporally fixed. Host growth was not impeded by A. lucii, and host molting was accelerated, so the asymptotic, maximum parasite size, if it is defined by host size, may increase over time. The observations on this system, as well as others like it (Uznanski and Nickol, 1980), demonstrate that assumptions concerning the interaction of acanthocephalans with their intermediate hosts cannot be made lightly. ACKNOWLEDGMENTS We would like to thank T. Hasu for sharing her experiences with infecting and maintaining isopods in the laboratory and for providing useful comments on earlier drafts of this manuscript. Eero Vestola helped in the maintenance of isopods. D.P.B. was supported by Fulbright and CIMO grants. LITERATURE CITED ANDERSEN, P. K. 1991. Survival analysis 1982–1991: The second decade of the proportional hazards regression model. Statistics in Medicine 10: 1931–1942. ASHWORTH, S. T., C. R. KENNEDY, AND G. BLANC. 1996. Density-dependent effects of Anguillicola crassus (Nematoda) within and on its copepod intermediate hosts. Parasitology 113: 303–309. AWACHIE, J. B. E. 1966. The development and life-history of Echinorhynchus truttae Schrank 1788 (Acanthocephala). Journal of Helminthology 40: 11–32. BENESH, D. P., AND E. T. VALTONEN. 2007. Proximate factors affecting the larval life history of Acanthocephalus lucii (Acanthocephala). Journal of Parasitology 93: 742–749. BERTIN, A., AND F. CEZILLY. 2003. Sexual selection, antennae length and the mating advantage of large males in Asellus aquaticus. Journal of Evolutionary Biology 16: 698–712. BRATTEY, J. 1983. The effects of larval Acanthocephalus lucii on the pigmentation, reproduction and susceptibility to predation of the isopod Asellus aquaticus. Journal of Parasitology 69: 1172–1173. ———. 1986. Life history and population biology of larval Acanthocephalus lucii (Acanthocephala: Echinorhynchidae) in the isopod Asellus aquaticus. Journal of Parasitology 72: 633–645. BROWN, A. F., AND D. PASCOE. 1989. Parasitism and host sensitivity to cadmium: An acanthocephalan infection of the freshwater amphipod Gammarus pulex. Journal of Applied Ecology 26: 473–487. CALADO, R., C. BARTILOTTI, AND L. NARCISCO. 2005. Short report on the effect of a parasitic isopod on the reproductive performance of a shrimp. Journal of Experimental Marine Biology and Ecology 321: 13–18. DEZFULI, B. S., L. GIARI, AND R. POULIN. 2001. Costs of intraspecific and interspecific host sharing in acanthocephalan cystacanths. Parasitology 122: 483–489. ———, E. ROSSETTI, E. A. FANO, AND R. ROSSI. 1994. Occurrence of larval Acanthocephalus anguillae (Acanthocephala) in Asellus BENESH AND VALTONEN—HOST EXPLOITATION BY A. LUCII aquaticus (Crustacea, Isopoda) from the River Brenta. Italian Journal of Zoology 61: 77–81. DUCLOS, L. M., B. J. DANNER, AND B. B. NICKOL. 2006. Virulence of Corynosoma constrictum (Acanthocephala: Polymorphidae) in Hyalella azteca (Amphipoda) throughout parasite ontogeny. Journal of Parasitology 92: 749–755. EBERT, D., AND E. A. HERRE. 1996. The evolution of parasitic diseases. Parasitology Today 12: 96–101. FREDENSBORG, B. L., K. M. MOURITSEN, AND R. POULIN. 2004. Intensitydependent mortality of Paracalliope novizealandiae (Amphipoda: Crustacea) infected by a trematode: Experimental infections and field observations. Journal of Experimental Marine Biology and Ecology 311: 253–265. GLEASON, L. N. 1987. Population dynamics of Pomphorhynchus bulbocolli in Gammarus pseudolimnaeus. Journal of Parasitology 73: 1099–1101. HASU, T., J. C. HOLMES, AND E. T. VALTONEN. 2007. Isopod (Asellus aquaticus) size and acanthocephalan (Acanthocephalus lucii) infections. Journal of Parasitology 93: 450–457. ———, E. T. VALTONEN, AND J. JOKELA. 2006. Costs of parasite resistance for female survival and parental care in a freshwater isopod. Oikos 114: 322–328. HURD, H., E. WARR, AND A. POLWART. 2001. A parasite that increases host lifespan. Proceedings of Royal Society of London, Series B 268: 1749–1753. HYNES, H. B. N., AND W. L. NICHOLAS. 1958. The resistance of Gammarus spp. to infection by Polymorphus minutus (Goeze, 1782) (Acanthocephala). Annals of Tropical Medicine and Parasitology 52: 376–383. JORMALAINEN, V. 1998. Precopulatory mate guarding in crustaceans: Male competitive strategy and intersexual conflict. Quarterly Review of Biology 73: 275–304. KAKIZAKI, T., T. SAITO, A. OHTAKA, AND K. NAGASAWA. 2003. Effects of Acanthocephalus sp. (Acanthocephala: Echinorhynchidae) on the body size and reproduction of isopods (Asellus hilgendorfi). Limnology 4: 43–46. KOKKOTIS, T., AND J. D. MCLAUGHLIN. 2006. Pathogenicity of the hymenolepidid cestode Microsomacanthus hopkinsi in its intermediate host Hyalella azteca: Implications for transmission, host fitness, and host populations. Canadian Journal of Zoology 84: 32–41. KRIST, A. C., J. JOKELA, J. WIEHN, AND C. M. LIVELY. 2004. Effects of host condition on susceptibility to infection, parasite developmental rate, and parasite transmission in a snail-trematode interaction. Journal of Evolutionary Biology 17: 33–40. LACKIE, J. M. 1972. The course of infection and growth of Moniliformis dubius (Acanthocephala) in the intermediate host Periplaneta Americana. Parasitology 64: 95–106. MARCUS, J. H. 1990. Moulting as an indicator of growth in Asellus aquaticus (L., 1758) and A. meridianus Racovitza, 1919 (Isopoda). Crustaceana 58: 136–143 NICKOL, B. B. 1985. Epizootiology. In Biology of the Acanthocephala, B. B. Nickol and D. W. T Crompton (eds.). Cambridge University Press, Cambridge, U.K., p. 307–346. ———, AND G. E. DAPPEN. 1982. Armadillidium vulgare (Isopoda) as an intermediate host of Plagiorhynchus cylindraceus (Acanthocephala) and isopod response to infection. Journal of Parasitology 68: 570–575. OETINGER, D. F. 1987. Effects of Acanthocephalus dirus (Acanthoceph- 741 ala) on morphometrics and reproduction of Caecidotea intermedius (Crustacea: Isopoda). Transactions of the American Microscopical Society 106: 240–248. PARKER, G. A., J. C. CHUBB, M. A. BALL, AND G. N. ROBERTS. 2003. Evolution of complex life cycles in helminth parasites. Nature 425: 480–484. ———, ———, G. N. ROBERTS, M. MICHAUD, AND M. MILINSKI. 2003. Optimal growth strategies of larval helminths in their intermediate hosts. Journal of Evolutionary Biology 16: 47–54. PLAISTOW, S. J., Y. OUTREMAN, Y. MORET, AND T. RIGAUD. 2003. Variation in the risk of being wounded: An overlooked factor in studies of invertebrate immune function? Ecology Letters 6: 489–494. POULIN, R. 1998. Evolutionary ecology of parasites. Chapman and Hall, London, U.K., 212 p. ROBINSON, E. S., AND B. C. STRICKLAND. 1969. Cellular responses of Periplaneta americana to acanthocephalan larvae. Experimental Parasitology 50: 694–697. ROSEN, R., AND T. A. DICK. 1983. Development and infectivity of the procercoid of Triaenophorus crassus Forel and mortality of the first intermediate host. Canadian Journal of Zoology 61: 2120–2128. SCHJETLEIN, J., AND A. SKORPING. 1995. The temperature threshold for development of Elaphostrongylus rangiferi in the intermediate host: An adaptation to winter survival? Parasitology 111: 103–110. SCHMIDT, G. D. 1985. Development and life cycles. In Biology of the Acanthocephala, B. B. Nickol and D. W. T. Crompton (eds.). Cambridge University Press, Cambridge, U.K., p. 273–306. SEIDENBERG, A. J. 1973. Ecology of the acanthocephalan, Acanthocephalus dirus (Van Cleave, 1931) in its intermediate host, Asellus intermedius Forbes (Crustacea: Isopoda). Journal of Parasitology 59: 957–962. SHOSTAK, A. W., R. B. ROSEN, AND T. A. DICK. 1985. The use of growth curves to assess the crowding effect on procercoids of the tapeworm Triaenophorus crassus in the copepod host Cyclops bicuspidatus thomasi. Canadian Journal of Zoology 63: 2343–2351. SORENSEN, R. E., AND D. J. MINCHELLA. 1998. Parasite influences on host life history: Echinostoma revolutum parasitism of Lymnaea elodes snails. Oecologia 115: 188–195. ———, AND ———. 2001. Snail-trematode life history interactions: Past trends and future directions. Parasitology 123: S3–S18. SPARKES, T. C., V. M. WRIGHT, D. T. RENWICK, K. A. WEIL, J. A. TALKINGTON, AND M. MILHALYOV. 2004. Intraspecific host sharing in the manipulative parasite Acanthocephalus dirus: Does conflict occur over host modification? Parasitology 129: 335–340. STEINAUER, M. L., AND B. B. NICKOL. 2003. Effect of cystacanth body size on adult success. Journal of Parasitology 89: 251–254. THOMAS, F., O. VERNEAU, F. SANTALLA, F. CEZILLY, AND F. RENAUD. 1996. The influence of intensity of infection by a trematode parasite on the reproductive biology of Gammarus insensibilis (Amphipoda). International Journal for Parasitology 26: 1205–1209. UZNANSKI, R. L., AND B. B. NICKOL. 1980. Parasite population regulation: Lethal and sublethal effects of Leptorhynchoides thecatus (Acanthocephala: Rhadinorhynchidae) on Hyalella azteca (Amphipoda). Journal of Parasitology 66: 121–126. WEDEKIND, C. 1997. The infectivity, growth and virulence of the cestode Schistocephalus solidus in its first intermediate host, the copepod Macrocyclops albidus. Parasitology 115: 317–324. ZAR, J. H. 1999. Biostatistical analysis. Prentice-Hall, Upper Saddle River, New Jersey, 929 p.