Oikos 116: 1895Á1903, 2007 doi: 10.1111/j.2007.0030-1299.15883.x, Copyright # Oikos 2007, ISSN 0030-1299 Subject Editor: Jonathan Shurin, Accepted 13 June 2007
Competition and intraguild egg predation among freshwater snails: re-examining the mechanism of interspecific interactions
Andrew M. Turner, Rebekah R. Turner and Scott R. Ray
A. M. Turner ([email protected]), R. R. Turner and S. R. Ray, Dept of Biology, Clarion Univ., Clarion, PA 16214, USA.
Experimental and field studies suggest that freshwater snail species have negative effects on each other’s population growth rates. Because snails share similar diets, these interactions have been interpreted as the result of exploitative competition, but they could also result from intraguild predation. Here we present three experiments aimed at testing the hypothesis that interspecific interactions among three species of freshwater gastropod (Helisoma trivolvis, Physa acuta and Stagnicola elodes) are mediated by intraguild egg predation. Foraging trials, conducted in a laboratory, showed that some snails readily prey on eggs, but the extent of egg predation depended on both the identity of the snail predator and the identity of the egg mass. Of the three species considered, Stagnicola had the largest effect on egg mortality and Physa had no effect on egg mortality. Foraging trials also showed that the eggs of Physa were the most vulnerable to predators and that the eggs of Stagnicola were largely invulnerable. A study conducted in large outdoor mesocosms assessed the occurrence of egg mortality in an environment of more extensive spatial scale and complexity. The results largely mirrored those of the laboratory study, with Stagnicola being the most voracious predator and the eggs of Physa being most vulnerable to predation. The reproductive success of Physa and Stagnicola raised in sympatry and allopatry was assessed in a mesocosm study conducted over three months. Recruitment of both species was depressed in sympatry, but patterns of growth in the survivors suggest contrasting mechanisms of suppression: Physa suppressed Stagnicola via exploitative competition, but Stagnicola suppressed Physa via egg predation. These experiments support the hypothesis that freshwater snail assemblages are structured by strong interspecific interactions and that a rich interplay of egg predation and interspecific competition underlie interactions among the members of this guild.
Studies of competitive interactions among animal because it can produce community dynamics (e.g. species are often built around experiments that manip- alternate stable states) that are qualitatively and quanti- ulate the density of hypothesized competitors, and the tatively different than those produced by pure exploi- putative competitors are selected based on similarity in tative competition (Polis et al. 1989, Polis and Holt diet and habitat use (Connell 1983, Schoener 1983). 1992). The frequency and importance of intraguild Although such tests can reveal whether negative inter- predation in nature may be underestimated because it is specific interactions exist, they do not always reveal the difficult to identify, and it can produce patterns of mechanism of interaction, as a number of studies show species replacement that are sometimes mistakenly that animals originally thought to share similar diets interpreted as the result of competition. may sometimes prey on each other (Heyer et al. 1975, Freshwater snails provide an excellent example of a Wissinger et al. 1996). An interaction in which a species guild exhibiting predictable patterns of species turnover facultatively preys on another species that otherwise uses along an important environmental gradient, in this case similar resources, and thus is also a potential compe- the pond size/hydroperiod gradient (Pip 1986, Jokinen titor, is called intraguild predation (Polis et al. 1989). A 1987, Dillon 2000). Field studies (Barbosa 1973, Adam full understanding of community structure requires a and Lewis 1992) and experimental studies (Brown 1982, knowledge of the occurrence of intraguild predation Hershey 1990) of interspecific interactions conclude that
1895 freshwater snails have negative effects on each other’s (20 35 14 cm deep) were filled to a depth of 4 cm population growth rates. The negative interactions have (2.8 l) with aged well water. Adult snails were stocked been interpreted as the result of exploitative competition, into the containers, allowed to deposit egg masses, and but none of the above studies provide corroborating removed. Egg masses were then thinned to a constant evidence. In one important example, Brown (1982) density of five masses per container. Egg masses were found that Stagnicola elodes depressed the reproductive individually marked with inscriptions on the outside of output of Physa gyrina. Brown concluded that the the opaque container. A single adult snail was stocked reproductive suppression was due to resource competi- into each container and allowed to forage. After four days tion between Stagnicola and Physa, and this study is now the predator was removed and the remaining egg masses a textbook example of how exploitative competition were tallied. Finely ground fish food was added to each structures community composition (Townsend et al. container to provide an alternate food source for the 2003, Begon et al. 2006). Our field observations suggest, snails. Temperature was maintained at 228C, and light- however, that Stagnicola is a predator of Physa eggs ing was set at 14:10 h day: night. Eggs of three species (Turner unpubl.), leading us to question whether (S. elodes, P. acuta and H. trivolvis) were used as prey, interspecific interactions involving Stagnicola are driven and the same three species were employed as predators, by exploitative competition or egg predation. with all nine possible predator/prey combinations tested. This paper presents the results of three experiments The experiment consisted of n 9 replicate trials per testing the hypothesis that interspecific interactions predator/prey combination, for a total of 81 predation among freshwater gastropods are mediated in part by trials. Every set of nine trials was accompanied by a no- intraguild egg predation. We first examine whether predator control trial in order to assure that there was no three species of pulmonate snails, common in the ponds apparent egg mortality in the absence of predators. of eastern North America, are capable of preying on Effects of predator identity and egg identity on egg each other’s eggs in a laboratory setting. We then mortality were analyzed with 3 3 ANOVA. expand the spatial scale, and test whether egg predation occurs in mesocosms that better mimic natural ponds. Finally, we conduct a mesocosm experiment examining Egg mortality study the population level consequences of egg predation and interspecific competition. We assessed the importance of egg predation at broader spatial scales by monitoring egg mortality in large mesocosms (1.5 m Ø, 0.5 m deep, 900 l), meant to Methods mimic small ponds, placed outdoors at the Pymatuning Laboratory of Ecology in northwestern Pennsylvania, Study system USA. Three target species (P. acuta, H. trivolvis and S. elodes) were stocked by themselves and in pairwise Our study focuses on species interactions among three combinations with the two other species. Thus, each species of large-bodied pulmonate snails: Physa acuta, species was subject to three treatments, an allopatry Stagnicola elodes and Helisoma trivolvis. These three treatment and two sympatry treatments. In the third species are widespread and commonly co-occur in the week of June, we filled 36 mesocosms with well water, ponds and marshes of northeastern and midwestern inoculated them with zooplankton and phytoplankton, North America (Pip 1986, Jokinen 1987). Physa acuta and added 5 g dry red maple, Acer rubrum, to each tank. and H. trivolvis are especially widespread and abundant Growth of periphyton and decaying leaves provided food in the ponds of northwest Pennsylvania (Mower and for snails and some physical complexity, though the lack Turner 2004). Stagnicola elodes has a more narrow of macrophytes resulted in an environmental with less habitat distribution, being restricted to shallow ponds complexity than is typical of natural ponds. After four with some forest cover, but within such ponds they are weeks, reproductive snails were stocked at the rate of 15 also quite abundant (Brown 1982, Mower and Turner individuals per tank (allopatry treatments) or 30 snails 2004). All three species have traditionally been consid- per tank (sympatry treatments, 15 individuals of each ered detritivores and herbivores as their diet consists species). This density (8.5 and 17 snails m 2)is largely of detritus, periphyton, and senescent macro- considerably lower than snail densities typically observed phytes (Bovbjerg 1965, Clampitt 1970, Calow 1975, in ponds (Brown 1982, Dillon 2000). We observed egg reviewed by Dillon 2000). deposition and survivorship for one week following stocking of adults. Egg mortality was assessed by locating Laboratory foraging trials individual egg masses on the tank sides and monitoring their survivorship over a period of 72 h. The effect of The ability of snails to prey upon snail eggs was assessed sympatric neighbors on egg mortality was assessed with in laboratory trials. Polypropylene ‘‘shoebox’’ containers one-way ANOVA, with a separate test run for each target
1896 species, followed by Tukey’s post-hoc test. We mon- species, and mesocosms in high density treatments were itored multiple egg masses in each mesocosm, and mean stocked with eight snails per assigned species. Because the mortality was used in all analyses. The experimental density manipulation quadrupled the density of both consisted of n 6 replicate mesocosms for each treat- species, their relative density was maintained at 1:1. Eggs ment, and three treatments for each species (allopatry, were allowed to hatch and juveniles to grow without any sympatric with neighbor a, sympatric with neighbor b), further disturbance. The experiment was terminated yielding 18 mesocosms for each species. after 12 weeks as the largest individuals neared repro- ductive size. At this time, water was drained from the mesocosms and all snails were removed. We then sorted Recruitment study snails by species, enumerated individuals, and measured their collective wet biomass. Biomass and density were Longer-term experiments are necessary in order to test used to calculate the mean body size of snails in each whether interspecific interactions, mediated by either egg mesocosm. For each species, we analyzed treatment predation or exploitative competition, influence popula- effects on density and body size with a 2 2 ANOVA. tion size. Further, patterns of individual growth can Each species was stocked into a total of 24 mesocosms, so elucidate the mechanism of suppression, as predation there were six replicate mesocosms for each treatment will tend to allow higher individual growth rates, whereas combination. All data were log-transformed prior to competition should lead to lower individual growth. analyses. Therefore, we set up a longer-term mesocosm study designed to test whether two species of snails, P. acuta and S. elodes, influence each other’s recruitment and to Results detect the mechanism responsible. We chose to focus on these two species because they are the most divergent in Laboratory foraging trials vulnerability to egg predation and ability to prey on eggs, Snails readily preyed on eggs, but the extent of egg and also because Brown (1982) reported on competition predation depended on both the identity of the snail between S. elodes and a congener of P. acuta. predator and the identity of the egg mass, as well as the As in the egg-mortality study, we used a reciprocal interaction of the two factors (Table 1). Stagnicola were target-neighbor approach to evaluate interspecific inter- the most voracious predator of snail eggs, consuming actions. Adult snails were stocked into tanks in allopatry an average of 41% of the egg masses presented to it and sympatry with the other species, allowed the (Fig. 1). Physa was the least dangerous of the putative opportunity to oviposit and prey on eggs, and then predators, as it failed to consume any eggs, and Helisoma removed before the eggs hatched. Hatchlings were was of intermediate ferocity, as it imposed egg mortality allowed to grow to maturity, at which time the of 26% (Fig. 1). Eggs of Physa were the most vulnerable experiment was terminated. Thus, any predator effects to predation. Averaged across predator species, Physa were exerted during the pre-hatching period that adults eggs experienced a mortality of 37%, and in the presence were present and competition among juveniles occurred of Stagnicola adults they experienced a mortality rate of during the post-hatching period. This reflects the life- 96%. The eggs of Stagnicola were the least vulnerable to history pattern of these species, which are annuals with predators, as only 0.7% of their eggs experienced a mid-summer peak in reproduction and adult mortal- mortality. No-predator controls confirmed that egg ity. The interspecific competition treatment was cross- survivorship in the absence of predators was 100%. factored with a intraspecific density treatment in order The non-additive effects of predator species and egg to determine whether interspecific interactions de- identity accounted for a large portion of the variation in pended on absolute density. The density treatment egg mortality (Table 1), showing that the effect of involved a four-fold manipulation of the stocking predator species was highly dependent on egg identity. density of both species. This design is very similar to Examination of cell means (Fig. 1) suggests that the most that employed by Brown (1982). striking non-additive effect involved the predation of The experiment was conducted in 130 l mesocosms, H. trivolvis on eggs of different species. When presented measuring 1.32 0.78 0.30 m deep, placed outdoors. with Physa or Stagnicola eggs, H. trivolvis was relatively The mesocosms were filled with well water on 7 June, ineffectual as a predator, but it readily cannibalized eggs inoculated with pond water, and covered with 50% of its own species. shadecloth. As in the egg mortality study, the mesocosms provided snails with abundant algal food, but less physical complexity than natural ponds. Adult snails Egg mortality study were stocked in mid-June, allowed to oviposit and prey on eggs for seven days, and then removed. Mesocosms in Treatment means for egg mortality in the mesocosm low density treatments received two snails per assigned study ranged from 0 to 55.4%, which is lower than the
1897 Table 1. Effect of predator identity and egg identity on egg 1.0 mortality in laboratory foraging trials. For each factor, h2 is the proportion of total variance attributable to the factor, and is calculated as the ratio of the effect variance to the total variance 0.8 P. acuta eggs 2 (h SSfactor/SStotal).
Factor h2 DF F p 0.6
Predator species 0.218 2 39.0 B0.001 Egg species 0.185 2 33.2 B0.001 0.4
Predator species egg species 0.396 4 35.4 B0.001 Egg mortality Error 0.201 72 0.2 egg mortality observed in the smaller-scale laboratory 0.0 feeding trials. Mortality of Physa eggs was elevated in P. acuta H. trivolvis S. elodes the presence of other species (Fig. 2 top; F2,15 26.3, pB0.001), and pairwise comparisons show that the 1.0 effect of Stagnicola on Physa egg mortality was significant (pB0.001), but the effect of Helisoma was 0.8 H. trivolvis eggs not significant (p0.05). Mortality of Helisoma eggs also depended on the presence of other species (Fig. 2 0.6 middle; F2,15 3.89, p 0.04), as it was elevated in the presence of Stagnicola (Fig. 2. middle; p B 0.05), but not by Physa (p0.05). Stagnicola egg survival was not 0.4 influenced by the presence of other species (Fig. 2 Egg mortality bottom; F2,15 1.00, p0.10). Taken together, these 0.2 results are consistent with the laboratory study in that Physa eggs were the most vulnerable to predation and 0.0 the Stagnicola eggs were the least vulnerable (Fig. 2). P. acuta H. trivolvis S. elodes Also consistent with the laboratory study, Stagnicola had the largest effect on egg mortality, and Physa had 1.0 no effect (Fig. 2). In contrast with the laboratory study, there was no evidence in the outdoor mesocosms that 0.8 S. elodes eggs Helisoma consumed Physa eggs, or that Helisoma cannibalized their own eggs (Fig. 2). 0.6
Recruitment study 0.4 Egg mortality The recruitment study revealed strong interspecific interactions, and these interactions were independent 0.2 of the four-fold variation in overall density. Recruit- ment of Physa from eggs was suppressed by Stagnicola 0.0 (Table 2). Averaged across density treatments, the P. acuta H. trivolvis S. elodes number of Physa recruiting in sympatry was just 65% Predator species of recruitment in allopatry (Fig. 3). Physa recruitment, unsurprisingly, was higher in the high density tanks, Fig. 1. Egg predation by three species of snails, measured in but the effect of Stagnicola on Physa recruitment did laboratory foraging trials. Trials evaluated mortality of Physa, Helisoma and Stagnicola eggs (top, middle, and bottom) in not depend on density. Most importantly, Physa the presence of the same three species as putative predators (x- were substantially larger in the presence of Stagnicola axis). Bars represent the mean proportion (91 SE) of eggs (Table 2), as the mean mass of Physa was 2.0-fold suffering mortality in each treatment. higher in the sympatry treatment relative to the allopatry treatment (Fig. 3). Increasing Physa density had a significant negative effect on their body size, and recruitment in the presence of Physa was just 53% of there were no nonadditive effects of density and recruitment in the absence of Physa (Fig. 4). The effect Stagnicola on Physa body size (Table 2) of overall density on Stagnicola recruitment was Both recruitment and growth of S. elodes was significant, though weaker than the effect of Physa, strongly suppressed by P. acuta (Table 2). Stagnicola but the effect of P. acuta on S. elodes recruitment did
1898 1.0 Table 2. Effect of interspecific interactions and absolute density on recruitment and body size of P. acuta and S. elodes. Effect size (h2) is the proportion of total variance attributable to a 0.8 P. acuta eggs factor, and is calculated as the ratio of the effect variance to the 2 total variance (h SSfactor/SStotal).
0.6 Factor h2 DF F p
Physa recruitment 0.4 Stagnicola 0.263 1 10.9 0.004
Egg mortality density 0.247 1 10.2 0.004 Stagnicola density 0.006 1 0.3 0.602 0.2 error 0.484 20 Physa body size 0.0 Stagnicola 0.485 1 27.4 B0.001 P. acuta P. acuta P. acuta density 0.135 1 7.6 0.012 alone + H. trivolvis + S. elodes Stagnicola density 0.026 1 1.4 0.243 error 0.354 20 1.0 Stagnicola recruitment Physa 0.360 1 21.9 B0.001 density 0.327 1 19.9 B0.001 0.8 H. trivolvis eggs Physa density 0.002 1 0.1 0.739 error 0.311 19 0.6 Stagnicola body size Physa 0.251 1 7.1 0.015 density 0.032 1 0.9 0.351 0.4 Physa density 0.048 1 1.4 0.256
Egg mortality error 0.669 19 0.2 but in contrast with the effect of S. elodes on P. acuta, 0.0 the presence of P. acuta resulted in smaller bodied S. H. trivolvis H. trivolvis H. trivolvis elodes (Fig. 4). Stagnicola density and the nonadditive + P. acuta alone + S. elodes effects of density and P. acuta did not influence body 1.0 size (Table 2).
0.8 S. elodes eggs Discussion 0.6 Because freshwater snail taxa overlap broadly in their use of resources, exploitative competition has been 0.4 assumed, quite reasonably, to be the mechanism
Egg mortality underlying negative interspecific interactions among 0.2 snails (Brown 1982, Hershey 1990). The results presented here show strong negative interspecific inter- 0.0 actions among pond snails and reveal that the mechan- S. elodes S. elodes S. elodes ism of suppression involves both competition and + P. acuta + H. trivolvis alone intraguild egg predation. Laboratory foraging trials and outdoor egg mortality experiments show that Assemblage composition Stagnicola elodes is a voracious predator of snail eggs, and that Physa acuta eggs are the most vulnerable to Fig. 2. Egg mortality suffered by three snail species in allopatry and in sympatry with other species, measured in predation. The recruitment study revealed reciprocal an outdoor mesocosm study. Horizontal lines join assem- negative interactions between species, but body size data blages not statistically different from each other (Tukey’s test, suggest contrasting mechanisms of suppression. Stagni- p 0.05). Bars represent the mean proportion (91 SE) of cola were smaller in sympatry than in allopatry. This eggs suffering mortality in each treatment. pattern would be expected if recruitment were limited by interspecific competition and is consistent with the observation of substantial overlap in the diets of Physa not depend on density (Table 2). Stagnicola body size and Stagnicola (Hunter 1980, Brown 1982). Physa, depended on whether P. acuta was present (Table 2), however, were larger in sympatry than in allopatry,
1899 1000 High density High density Stagnicola elodes Physa acuta Low density Low density 400
500 200
100 Snail recruitment (no. / mesocosm) Snail recruitment (no. / mesocosm) 250 Allopatry Sympatry Allopatry Sympatry
40 Physa acuta High density Stagnicola elodes High density Low density Low density 4 20
2 10 Snail body size (mg) Snail body size (mg)
1 5 Allopatry Sympatry Allopatry Sympatry
Fig. 3. Recruitment (top) and mean body size (bottom) of Fig. 4. Recruitment (top) and mean body size (bottom) of Physa acuta in allopatry and sympatry with Stagnicola elodes, Stagnicola elodes in allopatry and sympatry with Physa acuta, as measured in a long-term mesocosm experiment. Recruit- as measured in a long-term mesocosm experiment. Recruit- ment and body size are shown on logarithmically scaled axes. ment and body size are shown on logarithmically scaled axes. Data represent means91 SE. Data represent means91 SE.
detritivores and herbivores, but intraguild predation consistent with the ideas that Physa recruitment is requires a shift to carnivory. The facultative shifts of limited by egg predation and that reduced intraspecific herbivores and detritivores to a carnivorous diet may be competition induced by the thinning of Physa by more common in aquatic animals than is generally Stagnicola more than offset the increased interspecific appreciated (Atema and Burd 1975, Wissinger et al. competition from Stagnicola. Selective grazing by Stagnicola could conceivably benefit Physa if their diets 1996, Danielpol 1999, Vera Candioti 2005). were complementary (e.g. an indirect mutualism), but In light of these results, we turn to a reexamination of the negative effect of Physa on Stagnicola growth and Brown (1982), a study widely cited as providing the known similarity in their diets (Hunter 1980, empirical support for the important role played by Brown 1982) argue against this interpretation. Physa interspecific competition in shaping communities and Stagnicola commonly co-occur in the vernal ponds (Townsend et al. 2003, Begon et al. 2006). Brown of North America, and the structure of gastropod (1982) investigated the interactions between S. elodes assemblages in these ponds is likely shaped by a and P. gyrina, a congener to P. acuta. Field experiments complex interplay of competition and egg predation. were conducted by stocking Stagnicola and Physa alone Intraguild predation usually involves predators feed- and in sympatry into small (1 l) plastic containers placed ing on smaller species or size classes that are otherwise into in three ponds. In each pond the fecundity of Physa ecologically similar, and thus does not often involve was depressed in the presence of Stagnicola. Brown radical diet shifts (Polis et al. 1989, Polis and Holt (1982) also showed that Physa and Stagnicola overlap 1992). This study provides a more surprising example broadly in their use of food and habitat, and thus of intraguild predation as it requires a significant concluded that exploitative competition between the two departure from the typical pattern of snail resource species led to the depression of Physa fecundity within use. Freshwater snails have long been viewed as obligate the experimental enclosures. Physa growth, however, was
1900 unaffected by the presence of Stagnicola (Brown 1982), likely a very small contribution to their overall foraging an observation that is not consistent with the hypothesis activity. of exploitative competition. Both observations are con- Theory suggests that intraguild predation will make sistent, however, with the hypothesis that intraguild egg coexistence on a local scale less likely and that predation was responsible for the suppression of Physa coexistence is possible only if the intraguild prey is a reproduction. Another study also found that S. elodes more efficient competitor for shared resources than the reduced reproduction of coexisting gastropods, but had intraguild predator (Polis and Holt 1992). Our results no effect on growth (Hershey 1990). These studies have provide some support for the second prediction, but not been accepted as evidence that resource competition the first. Although our experiments do not allow for a plays the central role in determining the composition of strong test of competitive ability, in part because the gastropod assemblages, underscoring the degree to which density of Physa in the sympatry treatment was higher intraguild predation may be an overlooked mechanism than Stagnicola density, Stagnicola suffered a large in community ecology. depression of growth in sympatry, whereas Physa did Taken together, the results of our experiments show not show any depression of growth in sympatry, which that the nature of the relationship between Stagnicola suggests that Physa is the better competitor. Field and Physa shifts over the ontogeny of both species. surveys show that although Stagnicola has a quite Adult Stagnicola prey on Physa eggs, thereby limiting limited distribution relative to Physa, it is usually found Physa recruitment, but juvenile Stagnicola face a coexisting with Physa (Brown 1982, Mower and Turner recruitment bottleneck created by competition with 2004). Thus, in this case, it appears that the intraguild Physa. Such size-structured, mixed predation-competi- predator, an inferior competitor, does not exclude its tion interactions have been documented in several other intraguild prey from a community. Competitive sup- aquatic systems. Late-instar Chaoborus prey on Daph- pression by Physa may, however, be partially respon- nia, but early instar Chaoborus are food-limited and are sible for the limited distribution of Stagnicola. Fish suppressed by Daphnia (Neill and Peacock 1980). predators may also play an important role in limiting Similarly, adult largemouth bass prey on juvenile the distribution of Stagnicola and mediating coexistence bluegill, but juvenile bass compete with bluegill (Olson at the landscape scale. Physa is adept at detecting fish et al. 1995, see also Persson and Greenberg 1990, predators and moving into safe microhabitats but Bystro¨m et al. 1998). In each of these cases, the smaller Stagnicola is not, leaving Stagnicola more vulnerable species functions as prey to the larger species during to fish predation than Physa (Mower and Turner 2004, certain life history stages (a positive effect), and as a Turner et al. 2006). Predicting the net effect of competitor during other life-history stages (a negative intraguild prey on the intraguild predator remains a effect). A mixed predation-competition interaction complex task, especially when they share a common involving egg-predation has to our knowledge not enemy, and is a topic that deserves more attention. been previously described, but should not differ Egg predation (oophagy) is taxonomically wide- qualitatively from interactions in which predation falls spread (Polis 1981) and may be an overlooked on other life-history stages. Predicting the net effect of a mechanism of interspecific interaction. Given that change in intraguild prey population size on intraguild eggs offer a convenient package of energy to a forager predators is a complex task, as the outcome of the and are often poorly defended, it is no surprise that egg interaction will depend on the symmetry and intensity predation is prevalent. The role of egg predation in of the competitive and predatory interactions, as well as structuring animal communities has received the most the details of life-history, phenology, and growth attention in studies of avian ecology (Martin 1988), but (Persson and Greenberg 1990, Olson et al. 1995). in most cases egg predation has been approached from a Considering these factors, it seems likely that the population perspective rather than a community per- negative effects of Physa on Stagnicola recruitment, spective. Future studies could profitably address the mediated by resource competition, will override any community level consequences of egg predation. positive effects of Physa on Stagnicola as prey because Studies to date on freshwater gastropods have shown Physa egg masses are small and not likely to be an that ampullarids (apple snails and their relatives) can be important resource to Stagnicola. A typical adult Physa, egg predators, but oophagy was previously unknown held in the laboratory and fed ad libitum, will deposit an outside of this group. Marisa cornuarietis, an ampul- egg mass weighing 2.2 mg (dry mass; Turner unpubl.) larid gastropod native to central and south America but every three days, and the egg mass is small relative to widely introduced to Florida, preys on the eggs and Stagnicola adult body size (about 30 mg dry mass, soft neonates of other snails (Hunt 1958, Chi et al. 1971) tissue only). Thus, if an assemblage were evenly divided and this ability may explain its success in invading new between Physa and Stagnicola and egg predation rates habitats and displacing native gastropods. Other am- were 100%, Physa eggs would constitute a daily ration pullarids, including the apple snail Pomacea, have also equal to 2.4% of the body mass of Stagnicola, which is been documented to feed on the eggs of pulmonates
1901 (Paulinyi and Paulini 1972). Although there is only not reported for other freshwater snails. The preference limited study of egg predation among freshwater for carrion suggests that Stagnicola has an unusual gastropods, there is substantial evidence that interac- ability to consume and process animal material. tions among marine gastropods are mediated by egg Intraguild egg predation by snails is likely to be predation. Race (1982) reports that the exotic mudsnail context dependent, with the availability of physical Ilyanassa obsoleta has excluded the ecologically similar structure and alternate food sources important mediat- native species Cerithidea californica from much of its ing factors (Holbrook and Petranka 2004). We have preferred habitat, and that the mechanism at least conducted studies testing whether the availability of partially responsible is predation on the eggs and alternate food sources affects the rate of Stagnicola juveniles of C. californica (see D’Asaro 1970 and predation on Physa eggs (Turner et al. unpubl.). In Rawlings 1990). Many marine prosobranchs (the laboratory foraging trials, the addition of lower quality Neogastropoda) are carnivores and use their well foods (decaying leaves, periphyton) had little effect on developed radula to prey on smaller snails, providing egg predation, but the addition of higher quality foods another (if unsurprising, given their carnivorous food (dried fish food) slowed the rate of egg predation by habits) example of intraguild predation (Paine 1963). Stagnicola by half. Our outdoor mesocosm experiments In sum, work done to date suggests that oophagy is provided the food resources present in natural ponds spread quite unevenly across the gastropod clan, and (periphyton, decaying leaves), so the egg predation rates only further study will reveal how widespread oophagy measured in mesocosms are roughly indicative of egg might be among freshwater snails. predation rates experienced in ponds. The eggs of Physa were the most vulnerable to In sum, our results show that some species of predators, whereas Stagnicola eggs were largely invul- freshwater gastropods are adept egg predators and that nerable. These differences in egg vulnerability are likely the pronounced interspecific interactions among gastro- related to the traits of the eggs themselves. The eggs of pod taxa are in some cases mediated by egg predation, pulmonate snails are deposited in groups enclosed by a and in other cases mediated by interspecific competi- gelatinous sheath, with up to several dozen eggs in each tion. These experiments support the hypothesis that mass. The gelatinous sheath is composed of muccopro- freshwater snail assemblages are structured by strong teins and muccopolysaccharides covered with a mem- interspecific interactions and that a complex interplay brane. The function of the external sheath is not well of egg predation and interspecific competition mediate established, but investigators have suggested that the interactions among the members of this guild. A more gelatinous mass may screen UV light, limit dessication, complete understanding of the structure of freshwater or deter predators (Pechenik 1986). The vulnerability gastropod assemblages will be achieved by integrating, of an egg mass to a forager employing a scraping radula for the interacting species, a knowledge of competitive is likely related to the hardness of the gelatinous mass ability with a knowledge of their propensity to consume and toughness of the outer membrane. Qualitative the eggs of other snail species. observations suggest that the egg masses of Stagnicola and Helisoma are harder and tougher than are Physa egg Acknowledgements Á We thank Michael Chislock, Amanda masses (Dillon 2000), consistent with their relative Harold, Hannah Turner, Marty Heben and Sharon Mon- vulnerability to predation. It is possible that a tougher tgomery for assistance with field work. Jennifer Jones first membrane could limit oxygen diffusion to the eggs, but noted egg predation by gastropods, and we thank her for the nature of the tradeoffs associated with variation in bringing this observation to our attention. Sharon Montgom- egg mass morphology are not well understood. ery, Michael Chislock and Jason Hoverman provided helpful The specific traits underlying variability in the comments on the manuscript. This study was supported by ability of snails to forage on eggs are not completely the National Science Foundation (DEB-0444939) and is a known either. Of the three species considered, Stagni- contribution of the Pymatuning Laboratory of Ecology. cola had the largest effect on egg mortality. Gastropods forage by using a tongue-like structure (the odonto- phore) bearing a ribbon of teeth (the radula) to scrape References material from a substrate. Barnese et al. (1990) conducted a study of radular morphology for five Adam, M. and Lewis, J. 1992. The lack of coexistence species of Stagnicola, Physa and Helisoma, and found between Lymnaea peregra and lymnaea auricularia (Gas- that Stagnicola possessed the largest radular teeth and tropoda:Pulmonata). Á J. Mollusc. Stud. 58: 227Á228. Atema, J. and Burd, G. D. 1975. A field study of chemotactic Physa had the smallest radular teeth. There is some responses of the marine mud snail Nassarius obsoletus. Á J. suggestion in the literature that the feeding ability of Chem. Ecol. 1: 243Á251. Stagnicola differs from other snails as Stagnicola will Barbosa, R. 1973. Possible competitive displacement and forage on carrion whenever it is available (Bovbjerg evidence of hybridization between Brazilian species of 1968, Brown 1982), but similar foraging behavior is planorbid snails. Á Malacologia 14: 401Á408.
1902 Barnese, L. E. et al. 1990. Comparative grazing efficiency of Mower, C. and Turner, A. M. 2004. Behavior, morphology pulmonate and prosobranch snails. Á J. N. Am. Benthol. and the coexistence of two pulmonate snails with Soc. 9: 35Á44. molluscivorous fish: a comparative approach. Á Am. Begon, M. et al. 2006. Ecology: from individuals to Malacol. Bull. 19: 39Á46. ecosystems. Á Blackwell. Neill, W. E. and Peacock, A. 1980. Breaking the invertebrate Bovbjerg, R. 1965. Feeding and dispersal in the snail bottleneck: interactions of invertebrate predators and Stagnicola reflexa (Basommatophora: Lymnaeidae). nutrients in oligotrophic lakes. Á In: Kerfoot, W. C. Á Malacologia 2: 199Á207. (ed.), Ecology and evolution of zooplankton commu- Bovbjerg, R. 1968. Responses to food in lymnaeid snails. nities. Univ. Press of New England, pp. 715Á724. Á Physiol. Zool. 41: 412Á423. Olson, M. H. et al. 1995. Competition between predator and Brown, K.M. 1982. Resource overlap and competition in prey: the importance of resource based mechanisms and pond snails: an experimental analysis. Á Ecology 63: implications for size-structured interactions. Á Ecology 412Á422. 76: 1758Á1771. Bystro¨m, P. et al. 1998. Competing predators and prey: Paine, R. T. 1963. Trophic relationships of 8 sympatric juvenile bottlenecks in whole lake experiments. Á Ecology predatory gastropods. Á Ecology 44: 63Á72. 79: 2153Á2167. Paulinyi, H. M. and Paulini, E. 1972. Laboratory observa- Calow, P. 1975. The feeding strategies of two freshwater tions on the biological control of Biomphalaria glabrata gastropods, Ancylus fluviatilis and Planorbis contortus L. in by a species of Pomacea (Ampullariidae). Á Bull. WHO terms of ingestion rates and absorption efficiencies. 46: 243Á247. Á Oecologia 20: 33Á49. Pechenik, J. A. 1986. The encapsulation of eggs and embryos Chi, L. et al. 1971. Predation of Marisa cornuarietis on by molluscs: an overview. Á Am. Malacol. Bull. 4: 165Á Oncomelania formosana eggs under laboratory conditions. 172. Á Á Veliger 14: 184 186. Persson, L. and Greenberg, L.A. 1990. Juvenile competitive Clampitt, P. T. 1970. Comparative ecology of the snails Physa bottlenecks: the perch (Perca fluviatilis)-roach (Rutilus gyrina and Physa integra. Á Malacologia 10: 113Á151. rutilus) interaction. Á Ecology 71: 44Á56. Connell, J. H. 1983. On the prevalence and relative Pip, E. 1986. The ecology of freshwater gastropods in the importance of interspecific competition: evidence from central Canada region. Á Nautilus 100: 56Á66. field experiments. Á Am. Nat. 122: 661Á696. Polis, G. A. 1981. The evolution and dynamics of intraspe- Danielopol, D. L. et al. 1999. High Amphipoda species cific predation. Á Annu. Rev. Ecol. Syst. 12: 225Á251. richness in the Nert groundwater system (southern Polis, G. A. and Holt, R. D. 1992. Intraguild predation: the France). Á Crustaceana 72: 863Á882. dynamics of complex trophic interactions. Á Trends Ecol. D’Asaro, C. N. 1970. Egg capsules of prosobranch mollusks Á from south Florida and the Bahamas and notes on Evol. 7: 151 154. spawning in the laboratory. Á Bull. Mar. Sci. 20: 414Á Polis, G. A. et al. 1989. The ecology and evolution of 440. intraguild predation: potential competitors that eat each Dillon, R. 2000. The ecology of freshwater molluscs. other. Á Annu. Rev. Ecol. Syst. 20: 297Á330. Á Cambridge Univ. Press. Race, M. S. 1982. Competitive displacement and predation Hershey, A. E. 1990. Snail populations in arctic lakes: between introduced and native mud snails. Á Oecologia competition mediated by predation? Á Oecologia 82: 54: 337Á347. 26Á32. Rawlings, T. A. 1990. Associations between egg capsule Heyer, W. R. et al. 1975. Tadpoles, predation, and pond morphology and predation among populations of the habitats in the tropics. Á Biotropica 7: 100Á111. marine gastropod, Nucella emarginata. Á Biol. Bull. 179: Holbrook, C. T. and Petranka, J. W. 2004. Ecological 312Á325. interactions between Rana sylvatica and Ambystoma Schoener, T. W. 1983. Field experiments on interspecific malculatum: evidence of interspecific competition and competition. Á Am. Nat. 122: 240Á285. facultative intraguild predation. Á Copeia 4: 932Á939. Townsend, C. R. et al. 2003. Essentials of ecology. Hunt, B. P. 1958. Introduction of Marisa into Florida. Á Blackwell. Á Nautilus 54: 53Á55. Turner, A. M. et al. 2006. Learning, memory, and predator Hunter, R. D. 1980. The effects of grazing on the quantity avoidance by freshwater snails: effects of experience on and quality of freshwater aufwuchs. Á Hydrobiologia 69: predator recognition and defensive strategy. Á Anim. 251Á260. Behav. 72: 1443Á1450. Jokinen, E. 1987. Structure of freshwater snail communities: Vera Candioti, M. F. 2005. Morphology and feeding in speciesÁarea relationships and incidence categories. tadpoles of Ceratophrys cranwelli (Anura:Leptodactyli- Á Am. Malacol. Bull. 5: 9Á19. dae). Á Acta Zool. 86: 1Á11. Martin, T. E. 1988. Processes organizing open-nesting bird Wissinger, S. A. et al. 1996. Intraguild predation and assemblages: competition or nest predation? Á Evol. Ecol. cannibalism among larvae of detritivorous caddisflies in 2: 37Á50. subalpine wetlands. Á Ecology 77: 2421Á2430.
1903