Density-dependent costs of an inducible morphological defense in Crucian carp.
Predation is an important selective force in most habitats. Through direct, lethal effects, predators may have profound effects on abundance and size structure of prey populations and on species composition of prey communities (Sih et al. 1985, Kerfoot and Sih 1987). Likewise, predation has long been recognized as important in the evolution of anti-predator defenses, e.g., behavior (Lima and Dill 1990), crypsis (Endler 1986), unpalatability (Schultz 1988), and morphological defenses (Harvell 1990). While many anti-predator defense strategies are constitutive, i.e., fixed, a number of studies have recently demonstrated the importance of environmentally induced defenses, e.g., in plants (Baldwin et al. 1990), bryozoans (Harvell 1984), cladocerans (Dodson 1984), barnacles (Lively 1986a), gastropods (Appleton and Palmer 1988), and fish (Bronmark and Miner 1992). Constitutive defenses should be favored when predators are permanently present, whereas inducible defenses are favored in variable environments where predator attacks are intermittent (Harvell 1990, Clark and Harvell 1992). The evolution of inducible defenses also requires that the defense enhances prey survival and that prey have reliable cues for detecting predators (e.g., Dodson 1989, Adler and Harvell 1990). Further, the use of an inducible defense instead of a permanent one indicates that a fitness cost is associated with the defense, otherwise the evolutionarily stable strategy should be to use a permanent defense (Lively 1986b, Clark and Harvell 1992). Interestingly, these costs have proven difficult to demonstrate (Karban 1993, Tollrian 1995a), and recent work suggest that they may not always be detectable (Adler and Karban 1994), or even present (Karban 1993, Padilla and Adolph 1996). Nevertheless, allocation of resources to induced defense structures are likely to result in fewer resources being available for other purposes, e.g., growth, reproduction, or longevity (Levins 1968, Stearns 1989, 1992). Further, if possible, a defense should be resorbed when costs outweigh benefits (Stearns 1989), e.g., when the prey reaches a size when it is no longer susceptible to predation or when predators are no longer present (Riessen 1984, Havel and Dodson 1987, Bronmark and Pettersson 1994).
Most phenotypically induced morphological defenses are known from aquatic invertebrates that produce a defense structure in response to waterborne, chemical cues released by predators or injured conspecifics (reviewed in Havel 1987). The structure increases survival in the presence of predators (Harvell 1986, Lively 1986a, Tollrian 1995b), but also results in fitness costs in terms of reduced growth (e.g., Lively 1986c, Havel and Dodson 1987) or reproductive rate (e.g., Lively 1986c, Havel and Dodson 1987, but see Spitze 1992, Tollrian 1995a).
Recently, a predator-induced change in body morphology was discovered in a fish, the crucian carp, Carassius carassius (Bronmark and Miner 1992). Crucian carp is a widely distributed cyprinid that lives in ponds and lakes in Europe and central Asia (Maitland and Campbell 1992). In the absence of piscivores, crucian carp form dense, stunted populations (often [greater than] 30 000 individuals/ha), whereas populations coexisting with piscivores consist of few, (1-250 individuals/ha) large individuals (Bronmark et al. 1995). This suggests that crucian carp is very vulnerable to predation, which has been confirmed by experimental manipulations of predation pressure (Tonn et al. 1989, Bronmark and Miner 1992). When northern pike, Esox lucius, were introduced to small ponds in southern Sweden, crucian carp populations decreased markedly and surviving individuals became deeper bodied (Bronmark and Miner 1992). Laboratory experiments showed that body shape of crucian carp changes in response to waterborne chemical cues from piscivorous predators (Bronmark and Miner 1992, Bronmark and Pettersson 1994). The deeper body benefits crucian carp by decreasing predation efficiency of gape-limited piscivores, e.g., northern pike (Nilsson et al. 1995). However, theoretical estimates suggest that a deeper body increases total drag, thus increasing swimming costs (Bronmark and Miner 1992).
Few studies have investigated costs of inducible morphological defenses under field conditions (but see Lively 1986c, Harvell 1992). Whereas laboratory studies are important for detailed studies of, e.g., cost mechanisms, field experiments are essential for showing selective forces under natural conditions. Crucian carp is well suited for field studies because fish that have experienced induced body-shape changes still differ from non-induced control fish after 180 d in the absence of piscivores (Bronmark and Pettersson 1994). While the benefits of inducible defenses are manifested in the presence of predators, their costs should be most apparent during intense intraspecific competition when predators are absent. We hypothesized that shallow-bodied crucian carp should have a competitive advantage over deep-bodied fish as they avoid costs associated with the induced defense. Therefore, in this study, we investigated effects of intraspecific competition on shallow-bodied and deep-bodied crucian carp in a field situation with no predators present.
MATERIALS AND METHODS
The experiment was conducted from June to October 1993 in a 0.2-ha pond near Lund, southern Sweden where we built 20 enclosures (3 x 4 m, mean water depth 0.8 m). The enclosures were made of clear plastic walls sunk [approximately equal to] 20 cm into the sediment and supported by wooden frames. We added a 1-[m.sup.2] patch of sedge, Scirpus lacustris, to each enclosure to provide some structure; in addition a few pondweed, Potamogeton pectinatus, P. obtusifolius, and needle spike-rush, Eleocharis acicularis, were found in the cages. To equalize water levels in the enclosures, we placed two 15 x 15 cm plastic screen (1.5-mm mesh size) windows in the enclosure walls.
Experimental carp were collected by trap-netting and electroshocking in two ponds near Anderslov, southern Sweden. Crucian carp is common in ponds in this area, and the populations either consist of few deep-bodied fish coexisting with piscivores or, in the absence of predators, a large number of small, shallow-bodied fish (Bronmark and Miner 1992, Bronmark et al. 1995). The shallow-bodied crucian carp were collected horn a pond without piscivores while deep-bodied individuals were collected from a pond with piscivores. The fish were brought to the laboratory and sorted by size. We then selected fish with similar body masses (23.0 [+ or -] 0.3 g; mean [+ or -] 1 SE) and marked them with dye using a Pan Jet Injector (Hart and Pitcher 1969), either as focal fish (60 deep-bodied and 60 shallow-bodied individuals) or as competitors (120 shallow-bodied individuals). After weighing and measuring maximum body depth and total length we held the fish in 165-L tanks for 2 d to ensure fish survived the handling; no fish died.
The experiment started on 18 June 1993 when treatments were randomly assigned to the enclosures, giving five replicates to each of the following treatment combinations: deep-bodied fish in low density, shallow-bodied fish in low density, deep-bodied fish in high density, and shallow-bodied fish in high population density. We stocked each enclosure with six randomly chosen focal fish of one morph. In all high-density enclosures an additional 12 randomly chosen shallow-bodied crucian carp were added. Thus, the low-density treatments consisted of focal fish only whereas the high-density treatments consisted of focal fish together with non-focal competitors. The non-focal competitors were all of the same morph because we wanted the impact that competitors had on resource levels in high-density treatments to be as similar as possible. Unfortunately, four replicates were lost due to disturbance, and from then on the number of replicates were: n = 3 deep-bodied fish in low density, n = 5 shallow-bodied fish in low density, n = 4 deep-bodied fish in high density, and n = 4 shallow-bodied fish in high population density. The loss of four replicates caused initial mean biomass of focal fish to differ marginally between treatments ([F.sub.1,12] = 4.628, P = 0.053), with deep-bodied crucian carp tending to have higher biomass than shallow-bodied fish (25.2 [+ or -] 1.4 and 22.3 [+ or -] 0.4 g, respectively). After the competitors had been added, the total biomass was 11.8 g/[m.sup.2] in the low-density treatment and 35.3 g/[m.sup.2] in the high-density treatment. This is below the maximum density of crucian carp found in southern Sweden (upper limit: 55.5-62.5 g/[m.sup.2]; Miner and Bronmark, unpublished data). Deep-bodied fish were of shorter total lengths than shallow-bodied ones (105.6 [+ or -] 1.6 and 113.5 [+ or -] 0.6 mm, respectively, [F.sub.1,12] = 22.035, P [less than] 0.001). Relative depth (maximum body depth [divided by] total length) was 0.350 [+ or -] 0.002 in deep-bodied fish and 0.281 [+ or -] 0.002 in shallow-bodied fish. Focal fish were of similar age (deep-bodied crucian carp: mean age 5.9 yr, range 4-9 yr; shallow-bodied crucian carp: mean age 5.8 yr, range 4-8 yr) with no difference between morphs (t test, t = 0.479, 82 df, P = 0.634).
Crucian carp are diet generalists, feeding on plant material, invertebrates, and detritus (Prejs 1973, Penttinen and Holopainen 1992; Pettersson and Bronmark, unpublished data). Nevertheless, large cladoceran zooplankters constitute an important share of the diet for all size classes (Penttinen and Holopainen 1992). Thus, we used the total biomass of cladoceran zooplankters in each enclosure as a measure of food availability. Five-liter samples of water were collected from each enclosure using a tube sampler at the beginning of the experiment in June, and then in August and October. The samples were filtered through a 40-[[micro]meter] filter and fixed in 4% formaldehyde. In the laboratory, we identified, counted, and measured the cladoceran zooplankters. Lengths were converted to dry mass using length-mass relationships given in Bottrell et al. (1976).
After 2 mo, by mid August, we retrapped the crucian carp using electroshocking. The fish were weighed and measured and then put back in the cages. In October, 4 mo after the experiment was started, we removed all fish by electroshocking and seining. During the experiment, we had not replaced missing fish since we wanted all fish used in the subsequent analysis to have been in the experiment from the start. Fortunately, 186 of the original 192 crucian carp were retrapped. Missing fish (1-2) occurred in all four treatments. Despite frequent checking, no dead fish were found during the experiment. The retrapped fish were weighed and measured and then stored in the freezer at -20 [degrees] C for later analyses of liver and gonads. In the laboratory, gonads and livers were dissected from the focal fish and weighed. We calculated their gonadosomatic index (GSI) as 100 x the ratio of gonad mass to total body mass, and similarly, the liver index (LI) as 100 x the ratio of liver mass to total body mass. Glycogen content of liver tissue was determined by colorimetric methods from Lo et al. (1970). We estimated fish age from scales (Bagenal and Tesch 1978) and verified these measures against age-class data (Miner and Bronmark, unpublished data). Scale measurements were also used to compare previous growth histories of the focal fish. To do this, we applied linear models and regressed scale increments until the year before the present experiment against fish age (Weisberg and Frie 1987, Weisberg 1993). Individual regression coefficients were then compared using parametric tests.
To analyze treatment effects on cladoceran biomass, focal fish length, and focal fish biomass, we used mean values from each enclosure and date, and analyzed the data with multivariate ANOVA (MANOVA). The MANOVA with time as one factor allows detection of both overall differences and patterns through time (von Ende 1993), and in this particular case, it allowed a comparison of morphological differences that a univariate ANOVA on, e.g., the differences in data between June and October, would have obscured. As we used deep-bodied and shallow-bodied focal fish of similar body mass, they differed in total length (due to their morphology). But because the MANOVA handled data from each sampling date separately, these initial differences were explicitly taken into account. Unfortunately, the loss of four replicates made the design unbalanced, and we therefore had to restrict our analysis to the June and October samplings (due to lack of power, cf. von Ende 1993). We also used Type III sums of squares, to account for unbalanced designs (Scheiner 1993), throughout the analysis. Further, variables that were only measured once were analyzed with univariate ANOVAs. To meet assumptions of the models, cladoceran biomasses were transformed by natural logarithms, and relative depths were angular transformed.
Density had a significant effect on the increase in total length of crucian carp over the experimental period (Fig. 1, Table 1). Crucian carp in the low-density treatment increased markedly in total length, whereas fish in the high-density treatment grew less. Crucian carp body morphology, however, had no effect on the increase in total length. Deep-bodied and shallow-bodied crucian carp increased on average 25.9 [+ or -] 2.4 and 24.6 [+ or -] 0.6 mm, respectively (means [+ or -] 1 SE), at low density, and 13.7 [+ or -] 4.0 and 16.9 [+ or -] 1.4 mm at high density.
Increases in crucian carp body mass over the experimental period were affected by both density and body morphology (Fig. 2, Table 1). High density and deep-bodied morphology both resulted in lower body-mass gains. In the low-density treatment, deep-bodied and shallow-bodied crucian carp practically doubled their body masses [ILLUSTRATION FOR FIGURE 2 OMITTED], gaining 25.8 [+ or -] 2.7 and 28.0 [+ or -] 0.9 g, respectively (means [+ or -] 1 SE). At high population densities, overall body mass gain was significantly reduced (65% lower in deep-bodied individuals, 35% lower in shallow-bodied individuals). Furthermore, the increase in body mass of deep-bodied crucian carp was only half of that for shallow-bodied fish (increases 9.1 [+ or -] 3.4 and 18.2 [+ or -] 2.2 g, respectively, [ILLUSTRATION FOR FIGURE 2 OMITTED]). Although this clearly suggests that the difference in body mass gain between deep-bodied and shallow-bodied fish mainly occurred at the high density, support from a significant time x morphology x density interaction was lacking, possibly because of low power. However, univariate analysis of final biomass showed that, by October, there was no difference between deep-bodied and shallow-bodied focal fish at low population density (ANOVA, contrast shallow-bodied vs. deep-bodied focal fish: [F.sub.1,12] = 0.181, P = 0.678), while deep-bodied fish had a significantly lower biomass than shallow-bodied fish at the high density (ANOVA, contrast shallow-bodied vs. deep-bodied focal fish: [F.sub.1,12] = 5.136, P = 0.043). Moreover, the deep-bodied fish gained body mass significantly more slowly than their shallow-bodied within-enclosure competitors (mean difference 7.5 g, paired t test, t = -4.478, df = 3, P = 0.021), [TABULAR DATA FOR TABLE 1 OMITTED] whereas there was no significant difference between focal shallow-bodied fish and their within-enclosure competitors (mean difference 0.4 g, paired t test, t = - 1.714, df = 3, P = 0.185). Increases in total length and body mass did not differ between competitors kept with deep-bodied focal fish and competitors kept with shallow-bodied focal fish (MANOVA, total length, time x morphology: [F.sub.1,6] = 0.190, P = 0.678; body mass, time x morphology: [F.sub.1,6] = 0.100, P = 0.763).
The shallow-bodied and deep-bodied crucian carp used in this experiment came from different populations but were affected in a similar way by the experiment, and differences in growth rates during the experiment were not due to interpopulation differences in responsiveness to changing resource levels (cf. Arendt, in press). Relative to average scale growth during the previous 2 yr, fish at low population density grew at a similar rate (deep-bodied crucian carp 102 [+ or -] 9% (mean [+ or -] 1 SE) of previous growth, shallow-bodied crucian carp 122 [+ or -] 9%) with no difference between morphs (Tukey's hsd, P = 0.340). At high population densities, growth was lower than in previous years (deep-bodied 57 [+ or -] 9% and shallow-bodied crucian carp 63 [+ or -] 5% of previous growth), with no difference between morphs (Tukey's hsd, P = 0.937). Further, scale analyses showed that growth rates prior to the experiment had been decreasing with age in both populations, but this effect had been much stronger in shallow-bodied fish, resulting in a significant origin x age interaction ([F.sub.1,80] = 8.804, P = 0.004). We therefore analyzed the major age classes (5 and 6 yr) separately, and found no difference in growth rate prior to the experiment between shallow-bodied and deep-bodied crucian carp of age 5 (t test on individual regression coefficients, t = 1.295, df = 23, P = 0.208), whereas for 6-yr-old fish, deep-bodied individuals grew significantly faster than shallow-bodied fish (t test on individual regression coefficients, t = 4.192, df = 39, P = [less than] 0.001). During the experiment, scale increments mirrored overall effects of competition. Deep-bodied fish grew slower than shallow-bodied fish (ANOVA, [F.sub.1,12] = 7.057, P = 0.021) and growth was reduced at high population density, ([F.sub.1,12] = 84.963, P [less than] 0.001).
Male and female gonad masses were significantly greater at low density than at high density (Fig. 3, ANOVA, [F.sub.1,12] = 11.403, P = 0.006, and [F.sub.1,11] = 14.597, P = 0.003, respectively) but were not affected by morphology (ANOVA, [F.sub.1,12] = 0.335, P = 0.574, and [F.sub.1,11] = 0.986, P = 0.342, respectively). However, the male gonadosomatic index (GSI), i.e., relative allocation to gonads, was unaffected by population density and body morphology (ANOVA, density: [F.sub.1,12] = 1.762, P = 0.209; morphology: [F.sub.1,12] = 1.104, P = 0.314). Female gonadosomatic index was significantly lower at high density than at low density (ANOVA, [F.sub.1,11] = 10.146, P = 0.009) but was not affected by fish morphology (ANOVA, [F.sub.1,11] = 0.093, P = 0.767). Female gonadal mass was closely correlated with body mass (Y = 3.2 + 0.15X, Pearson r = 0.936, df = 13, P [less than] 0.001), larger females having larger gonads. The same was true in males, although the correlation was weaker (Y = 0.07 + 0.025X, Pearson r = 0.546, df = 14, P = 0.029).
The liver index was rather variable [ILLUSTRATION FOR FIGURE 4 OMITTED] and did not differ between deep-bodied and shallow-bodied crucian carp (morphology: [F.sub.1,12] = 0.004, P = 0.951). The liver index was significantly greater at the higher density (Fig. 4, density: [F.sub.1,12] = 5.920, P = 0.032). Liver mass was well correlated with glycogen content (Y = 0.12X, Pearson r = 0.530, df = 19, P = 0.020), thus confirming liver indices as appropriate measures of crucian carp glycogen stores.
The most common cladoceran zooplankters in the enclosures were Daphnia longispina, Ceriodaphnia spp., Bosmina longirostris, and Chydorus sphaericus, whereas D. magna, D. pulex, D. cucullata, Polyphemus pediculus, B. coregoni, and Scapholebris mucronata occurred in lower numbers. Cyclopoid copepods were only found in low numbers. The total biomass of cladocerans present before the experiment was rather variable, but similar among treatments ([ILLUSTRATION FOR FIGURE 5 OMITTED], ANOVA, [F.sub.3,12] = 0.308, P = 0.819). During the experiment, total biomass was reduced in all enclosures (MANOVA, [F.sub.1,12] = 32.025, P [less than] 0.001), but it was significantly more reduced in enclosures with high densities of crucian carp than in low-density enclosures (MANOVA, [F.sub.1,12] = 5.824, P = 0.033). The presence of deep-bodied vs. shallow-bodied focal fish had no effect on cladoceran biomass (MANOVA, [F.sub.1,12] = 0.388, P = 0.545).
Throughout the experiment, shallow-bodied crucian carp remained more fusiform than the deep-bodied cruclan carp (MANOVA, [F.sub.1,12] = 243.540, P [less than] 0.001). Furthermore, there was an interaction between density and the maintenance of the fusiform body morphology, causing fish at the low-density treatment to increase in relative depth (MANOVA, [F.sub.1,12] = 4.874, P = 0.048).
Inducible defenses and their costs
Most theoretical models of inducible defenses assume that there are fitness costs associated with the defense that balance its advantages, otherwise the evolutionarily stable strategy should be to have a permanent defense (Lively 1986b, Clark and Harvell 1992). Different direct costs are possible, e.g., energy requirements for defense construction (Harvell 1986, Lively 1986a), increased maintenance costs caused by the defense (Dodson 1984, Bronmark and Miner 1992), or even resource needs for maintaining the ability to activate the defense, i.e., a cost of plasticity (Harvell 1990, Moran 1992). Also, increasing evidence suggests that more subtle, indirect costs can be important in the evolution of inducible defenses, e.g., lost opportunities faced by organisms with a particular defense investment when the environment or predation regime is changing (Tollrian 1995a; Harvell and Tollrian, unpublished manuscript). However, inducible defenses do not always result in measurable costs (Spitze 1992, Karban 1993, Tollrian 1995a, Padilla and Adolph 1996). One reason is that selection should generally force the evolution of inducible defenses towards cost minimization (Tollrian 1995a), thus making costs increasingly hard to detect. Further, recent theoretical work suggests that, under certain circumstances, inducible defenses can evolve even in the absence of costs (Adler and Karban 1994, Padilla and Adolph 1996). Nonetheless, the fact that these defenses are inducible makes them particularly suitable for experimental investigations, as relative merits of defended and undefended phenotypes can be directly compared.
The present study shows strong and density-dependent costs of the inducible morphological defense in crucian carp. At low density, both deep-bodied and shallow-bodied crucian carp increased markedly in length and body mass, whereas fish at the high density grew significantly less. The decrease in growth coupled at high density with reduced cladoceran densities suggests that crucian carp were limited by intraspecific competition for resources, as has been shown in natural populations (Tonn et al. 1994). Body mass gains at the high density revealed marked differences in performance between deep-bodied and shallow-bodied individuals. Deep-bodied focal fish gained body mass at a slower rate (50% between June and October) than shallow-bodied focal fish. A direct within-enclosure comparison of deep-bodied focal fish and their shallowbodied competitors showed a similar pattern, with deep-bodied fish gaining on average 7.5 g (45%) less body mass than shallow-bodied fish in the same enclosure over the same period.
Gonad mass was directly related to body mass, and thus, the deep-bodied fish suffered a substantial direct fitness cost in this situation. Furthermore, a reduced body-mass gain can have several other negative impacts on relative fitness. For example, fish body mass is well known to be positively related to both foraging success (Mittelbach 1984, Paszkowski et al. 1990), and competitive performance (Mittelbach 1981, Osenberg et al. 1988). Consequently, shallow-bodied crucian carp should have a competitive advantage in the absence of predators when resources are limiting.
Analysis of individual growth rates prior to the experiment demonstrated that the differences in growth rate found here between shallow-bodied and deep-bodied crucian carp were not caused by genetically determined interpopulation differences. No differences in growth rate were found for 5-yr-old crucian carp, whereas for 6-yr-old fish, deep-bodied individuals, i.e., from the population that coexisted with pike, had grown faster than had the shallow-bodied fish in the population without piscivores. If anything, our design was conservative insofar as deep-bodied individuals tended to have higher body mass than shallow-bodied individuals at the beginning of the experiment, which could have given them an initial competitive advantage (Mittelbach 1981, Paszkowski et al. 1990). Further, deep-bodied and shallow-bodied fish responded similarly at low density, with individuals of both morphs practically doubling their body masses, which clearly suggests no interpopulation difference.
Gonad mass of focal fish was mainly affected by fish density, with both shallow-bodied and deep-bodied males and females having significantly lower gonadal mass at high density, but with no difference between deep-bodied and shallow-bodied fish. Male relative allocation to gonads, gonadosomatic index (GSI), was unaffected by population density and body morphology, whereas females increased their relative allocation to gonads at low density. However, crucian carp is a multiple spawner with several batches per year (Astanin and Podgornyy 1968), which complicates estimation of fecundity from gonad mass or gonadosomatic indices, as several batches are developing at the same time (Macer 1974). Further, it can take several months before changes in body mass are reallocated to gonads (Reznick and Braun 1987, Tanasichuk and Mackay 1989), and although the following breeding season's eggs and milt were formed by the time we terminated our experiment (S. E Hamrin, personal communications, L. B. Pettersson and C. Bronmark, personal observations), the gonads were not yet fully developed by then. It would have been preferable to census gonadal allocation the following spring, allowing more time for reallocation of resources, but unfortunately this was logistically impossible. However, as both male and female gonad mass was significantly correlated with body mass, it seems reasonable that the lower body mass of deep-bodied crucian carp at the high density would eventually have adverse effects on fitness here as in other species (cf. Wootton 1985, Harvell 1986, Lively 1986c).
Internal resource-allocation patterns in crucian carp are also affected by the presence of dynamic, seasonally large, liver glycogen reserves. Glycogen that is stored in the liver acts as a buffer between resource intake and somatic or gonadal growth (Wootton 1985, Chellappa et al. 1989). In crucian carp, it plays an additional role, as these fish can survive winterkills (i.e., severe oxygen depletion during winter) by anaerobically metabolizing their glycogen reserves. The reserve buildup takes place during the autumn, and the relative size of the liver can then increase from a few percent up to 15% of the body mass (Hyvarinen et al. 1985). As liver mass is correlated with liver glycogen content (e.g., Hyvarinen et al. 1985; Results), crucian carp with livers that are large relative to their total body mass have larger glycogen stores and can therefore survive longer anoxic periods (Hyvarinen et al. 1985). Tonn et al. (1989, 1994) have shown that intense intraspecific competition at high population densities can lead to reduced liver indices. In the present study we found just the opposite, i.e., lower liver indices at low density than at high density. The reason for this is not obvious but may reflect regional differences in growth and glycogen storage. For example, crucian carp continue to grow throughout October in southern Sweden (L. B. Pettersson and C. Bronmark, unpublished data), whereas the growing season ends in mid August in eastern Finland (Holopainen et al. 1991). Further, the probability of winterkill is lower in southern Sweden than in eastern Finland and this should affect glycogen storage, e.g., by lowering the optimal size of the glycogen reserves. However, both these explanations suggest similar effects at low and high density, whereas we found that the two densities were affected differently. An alternative explanation is provided by Clark and Ekman (1995), who have shown that the predictability of food resources can influence storage strategies, causing individuals that experience an uncertain food supply to keep larger stores whereas a more certain food supply allow smaller stores. Some support for this have been found in crucian carp, e.g., larger reserves associated with more intense competition (Paszkowski et al. 1990; Results), indicating that crucian carp glycogen storage may indeed be more complex than previously assumed.
Mechanisms and implications
A number of explicit mechanisms by which inducible morphological defenses may result in direct fitness costs have been suggested, e.g., investment in permanent structures (Harvell 1984), molt losses (Barry 1994), and hydrodynamic disadvantages (Dodson 1984, Bronmark and Miner 1992). In the crucian carp, estimates of theoretical drag suggest that the increase in body depth induced by the presence of piscivores may result in a cost by increasing energy expenditure during swimming (Bronmark and Miner 1992). Generally, a more deep-bodied morphology should result in a larger cost of swimming (Webb 1975), as fish having a larger surface area experience greater drag and therefore spend more energy on movement. The theoretical total drag (Webb 1975) for the body of a 140-mm crucian carp was calculated to be 32% higher for deep-bodied fish (mean depth/length ratio 0.378) than for shallow-bodied fish (mean depth/length ratio 0.309) (Bronmark and Miner 1992). Although energy expenditure during routine swimming normally accounts for a minor portion of the energy budget in fish (Priede 1985), such small but consistent costs could still be sufficient to direct the course of evolution (Adler and Karban 1994). Therefore, all else being equal, more deep-bodied individuals should have a hydrodynamic competitive disadvantage.
In the present study, differences in performance between the deep- and shallow-bodied fish were not due to costs of defense construction, as the morphological differences were present already at the beginning of the experiment. Further, the deep-bodied individuals suffered no substantial reduction in body mass gain until facing intense intraspecific competition. This corresponds well with effects predicted by hydrodynamic theory in this situation, as energy costs of defenses can be expected to be greater when resources are limited, because animals then need to allocate resources in the order of importance (Black and Dodson 1990, Walls et al. 1991). In fact, if crucian carp at the high population density had to forage more actively to find prey, this could have even increased the relative disadvantage of deep-bodied fish. Furthermore, whereas cladoceran densities were similar in both high-density treatments, suggesting that the overall energy intake of crucian carp was similar in these enclosures, the total gain in mass per enclosure (6 focal fish and 12 competitors) was 17% lower in enclosures with deep-bodied fish (273 g) than in enclosures with shallow-bodied fish (328 g), which may indicate that both morphs fed at the same rate, but that deep-bodied individuals had higher energy expenses. However, the energy cost may be explained by other mechanisms, e.g., increases in basal metabolism in deep-bodied individuals and, thus, direct measures of crucian carp routine energetics are needed to confirm that decrease in body mass gain is due to a hydrodynamic disadvantage.
Furthermore, crucian carp were placed in relatively homogeneous habitats, i.e., open sediment and a patch of vegetation, and there were no obvious differences in foraging or habitat use by the two morphs (L. B. Pettersson, personal observations). However, depending on habitat availability, the dimorphism of crucian carp can potentially lead to a trophic polymorphism, i.e., that deep-bodied and shallow-bodied individuals have different foraging specializations (Ehlinger and Wilson 1988, Ehlinger 1990, Robinson et al. 1996). In pumpkinseed sunfish, Lepomis gibbosus, differences sufficient for a specialization are rarely [greater than] 3% for any single body measure and require multivariate techniques to detect (Robinson et al. 1996). The general pattern found in several fish species is that more shallow-bodied, fusiform individuals specialize on pelagic zooplankters whereas more deep-bodied individuals specialize on benthic food, in particular inside littoral vegetation (Ehlinger and Wilson 1988, Robinson et al. 1996). Habitat use in crucian carp is strongly dependent on the perceived risk of predation (Tonn et al. 1989, Pettersson and Bronmark 1993) and the size of individuals (Tonn et al. 1989; L. B. Pettersson and C. Bronmark, personal observations). Little is known about crucian carp foraging and movement patterns in relation to habitat use in the field, but the littoral zone seems to be more important than the pelagic: for deep-bodied crucian carp coexisting with piscivores (Tonn et al. 1989) and less important when piscivores are absent (Penttinen and Holopainen 1992). These differences may affect relative performance of foraging strategies, e.g., steady cruising when feeding on zooplankters vs. slow movement and precise maneuvering when feeding on benthos, and it is therefore imperative to take foraging specializations into account when modelling energy expenditure of the two morphs during routine movement, i.e., crucian carp defense cost models.
Relations to population dynamics
Direct and indirect costs of morphological defenses have been demonstrated in a number of organisms (reviewed in Harvell 1990; Harvell and Tollrian, unpublished manuscript), but few studies have evaluated their importance in the field. To our knowledge, only two authors (Lively 1986c, Harvell 1992) have measured significant disadvantages associated with a defense structure under natural conditions. Lively (1986c) showed that defended barnacles suffered decreases in growth and reproduction, whereas Harvell (1992) showed that spined bryozoan colonies grew more slowly and also senesced sooner than unspined colonies.
Here, we show that the deep body of crucian carp that coexist with piscivores is a structure that is costly in terms of energy, causing substantial reductions in growth rate compared to more shallow-bodied conspecifics. However, in presence of piscivores, the major selective force acting on crucian carp is size-limited predation, resulting in sparse populations dominated by large individuals (Bronmark et al. 1995). In contrast, following an environmental disturbance that eliminates piscivores, such as oxygen depletion during severe winter conditions, crucian carp may face a situation where their morphological defense, a deep body, is no longer advantageous. The disappearance of piscivores coupled with the high reproductive potential of crucian carp, may lead to a hundredfold increase in density over a few years (Tonn et al. 1994). At the resulting high population densities, intraspecific competition is intense, resulting in reduced growth rates, lower liver glycogen reserves, and lower mean relative depth (Tonn et al. 1994). Under such circumstances, where deep-bodied individuals from the period with piscivores are forced to compete with more shallow-bodied conspecifics, the costs of the deep body should be most apparent.
Given the high cost of the deep body form at the high population densities present in the absence of piscivores, a deep-bodied individual should be expected to resorb (Havel and Dodson 1987) or reallocate the defense structure to decrease costs. In a laboratory experiment we found that the defense structure was, to a certain degree, reversible when northern pike were removed from experimental aquaria (Bronmark and Pettersson 1994). Further, as animals grow and reach a size refuge where they are no longer susceptible to predation, energy should be reallocated towards growing longer (Bronmark and Miner 1992). In ponds with piscivores, Holopainen and Pitkanen (1985) found that the ratio of body depth to length decreased for crucian carp [greater than] 150 mm, which may suggest that crucian carp that have reached a size refuge indeed grow proportionally more in length, thus decreasing hydrodynamic costs.
In conclusion, while stochastic environmental disturbances like winterkills coupled with piscivore recolonization may result in a variable predation pressure that is essential for the evolution of inducible defenses (Schultz 1988, Harvell 1990), the present experiment provides another factor promoting this evolutionary process (Harvell 1990), namely a disadvantage that deep-bodied crucian carp suffer when competing for limited resources with undefended, shallow-bodied conspecifics in the absence of predators. This effect, a reduced growth in total body mass, is density dependent, and is found under realistic field conditions, illustrating its importance in nature.
We thank Tina D'Hertefeldt, Anders Nilsson, and Erik Svensson for field work, Henrik G. Smith for statistical advice, Jeff Miner for providing data on crucian carp, and Siv Billberg for glycogen analyses. Valuable comments on an earlier draft were given by Larry Greenberg, Curt Lively, Gary Mittelbach, Erik Svensson, Ralph Tollrian, and an anonymous referee. Financial support was received from the Swedish Board for Agriculture and Forestry Research (C.B.), the Royal Swedish Academy of Sciences (L.B.P.) and Lund University (L.B.P).
Adler, F. R., and C. D. Harvell. 1990. Inducible defences, phenotypic variability and biotic environments. Trends in Ecology and Evolution 5:407-410.
Adler, F. A., and R. Karban. 1994. Defended fortresses or moving targets? Another model of inducible defenses inspired by military metaphors. American Naturalist 144: 813-832.
Appleton, R. D., and A. R. Palmer. 1988. Water-borne stimuli released by predatory crabs and damaged prey induce more predator-resistant shells in a marine gastropod. Proceedings of the National Academy of Sciences, USA 85:4387-4391.
Arendt, J. In press. Adaptive intrinsic growth rates: an integration across taxa. Quarterly Review of Biology.
Astanin, L. P., and M. I. Podgornyy. 1968. Features of the fertility of Carassius carassius L. and Carassius auratus gibelio Bloch. Problems in Ichthyology 8:209-214.
Bagenal, T B., and F. W. Tesch. 1978. Age and growth. Pages 101-136 in T B. Bagenal, editor. Methods for assessment of fish production in fresh waters. Blackwell Scientific, Oxford, England.
Baldwin, I. T, E. L. Sims, and S. E. Kean. 1990. The reproductive consequences associated with inducible alkaloidal responses in wild tobacco. Ecology 71:252-262.
Barry, M. J. 1994. The costs of crest induction for Daphnia carinata. Oecologia 97:278-288.
Black, A. R., and S. I. Dodson. 1990. Demographic costs of Chaoborus-induced phenotypic plasticity in Daphnia pulex. Oecologia 83:117-122.
Bottrell, H. H., A. Duncan, Z. M. Gliwicz, E. Grygierek, A. Herzig, A. Hillbricht-Ilkowska, H. Kurasawa, P. Larsson, and T. Weglenska. 1976. A review of some problems in zooplankton production studies. Norwegian Journal of Zoology 24:419-456.
Bronmark, C., and J. G. Miner. 1992. Predator-induced phenotypical change in body morphology in crucian carp. Science 258:1348-1350.
Bronmark, C., C. A. Paszkowski, W. M. Tonn, and A. Hargeby. 1995. Predation as a determinant of size structure in populations of crucian carp (Carassius carassius) and tench (Tinca tinca). Ecology of Freshwater Fish 4:85-92.
Bronmark, C., and L. B. Pettersson. 1994. Chemical cues from piscivores induce a change in morphology in crucian carp. Oikos 70:396-402.
Chellappa, S., F. A. Huntingford, R. H. C. Strang, and R. Y. Thomson. 1989. Annual variation in energy reserves in male three-spined stickleback, Gasterosteus aculeatus L. (Pisces, Gasterosteidae). Journal of Fish Biology 35:275-286.
Clark, C. W., and J. Ekman. 1995. Dominant and subordinate fattening strategies: a dynamic game. Oikos 72:205-212.
Clark, C. W., and C. D. Harvell. 1992. Inducible defences and the allocation of resources: a minimal model. American Naturalist 139:521-539.
Dodson, S. I. 1984. Predation of Heterocope septentrionalis on two species of Daphnia: morphological defenses and their cost. Ecology 65:1249-1257.
Dodson, S. 1989. Predator-induced reaction norms. BioScience 39:447-452.
Ehlinger, T. J. 1990. Habitat choice and phenotype-limited feeding efficiency in bluegill: individual differenees and trophic polymorphism. Ecology 71:886-896.
Ehlinger, T. J., and D. S. Wilson. 1988. Complex foraging polymorphism in bluegill sunfish. Proceedings of the National Academy of Sciences, USA 85:1878-1882.
Endler, J. A. 1986. Defense against predators. Pages 109-134 in M. E. Feder and G. V. Lauder, editors. Predator-prey relationships. University of Chicago Press, Chicago, Illinois, USA.
Hart, P. J. B., and T. J. Pitcher. 1969. Field trials of fish marking using a jet inoculator. Journal of Fish Biology 1: 383-385.
Harvell, C. D. 1984. Predator-induced defense in a marine bryozoan. Science 224:1357-1359.
-----. 1986 The ecology and evolution of inducible defenses in a marine bryozoan: cues, costs, and consequences. American Naturalist 128:810-823.
-----. 1990. The ecology and evolution of inducible defenses. Quarterly Review of Biology 65:323-340.
-----. 1992. Inducible defenses and allocation shifts in a marine bryozoan. Ecology 75:1567-1576.
Havel, J. E. 1987. Predator-induced defenses: a review. Pages 263-278 in W. C. Kerfoot and A. Sih, editors. Predation: direct and indirect impacts on aquatic communities. University Press of New England, Hanover, New Hampshire, USA.
Havel, J. E., and S. I. Dodson. 1987. Reproductive costs of Chaoborus-induced polymorphism in Daphnia pulex. Hydrobiologia 150:273-281.
Holopainen, I. J., and A. K. Pitkanen. 1985. Population size and structure of crucian carp (Carassius carassius (L.)) in two small, natural ponds in Eastern Finland Annales Zoologici Fennici 22:397-406.
Holopainen, I. J., W. M. Tonn, and C. A. Paszkowski. 1991. Ecological responses of crucian carp populations to predation by perch in a manipulated pond. Proceedings of the International Association of Theoretical and Applied Limnology 24:2412-2417.
Hyvarinen, H., I. J. Holopainen, and J. Piironen. 1985. Anaerobic wintering of crucian carp (Carassius carassius L.) - I. Annual dynamics of glycogen reserves in nature. Comparative Biochemistry and Physiology 82A:797-803.
Kerfoot, W. C., and A. Sih, editors. 1987. Predation: direct and indirect impacts on aquatic communities. University Press of New England, Hanover, New Hampshire, USA.
Levins, R. 1968. Evolution in changing environments. Princeton University Press, Princeton, New Jersey, USA.
Lima, S. L., and L. M. Dill. 1990. Behavioral decisions made under the risk of predation: a review and prospectus. Canadian Journal of Zoology 68:619-640.
Lively, C. M. 1986a. Predator-induced shell dimorphism in the acorn barnacle, Chthamalus anisopoma. Evolution 40: 232-242.
----- 1986b. Canalization versus developmental conversion in a spatially variable environment American Naturalist 128:561-572.
-----. 1986c. Competition, comparative life histories, and maintenance of shell dimorphism in a barnacle. Ecology 67:858-864.
Lo, S., J. C. Russell, and A. W. Taylor. 1970 Determination of glycogen in small tissue samples. Journal of Applied Physiology 28:234-236.
Macer, C. T. 1974. The reproductive biology of the horse mackerel Trachurus trachurus (L.) in the North Sea and English Channel. Journal of Fish Biology 6:415-438.
Maitland, P.S., and R. N. Campbell. 1992. Freshwater fishes of the British Isles. Harper Collins, London, England.
Mittelbach, G. G. 1981. Foraging efficiency and body size: a study of optimal diet and habitat use by bluegills. Ecology 62:1370-1386.
-----. 1984. Predation and resource partitioning in two sunfishes (Centrarchidae). Ecology 65:499-513.
Moran, N. A. 1992. The evolutionary maintenance of alternative phenotypes. American Naturalist 139:971-989.
Nilsson, P. A., C. Bronmark, and L. B. Pettersson. 1995. Benefits of a predator induced morphology in crucian carp. Oecologia 104:291-296.
Osenberg, C. W., E. E. Werner, G. G. Mittelbach, and D. J. Hall. 1988. Growth patterns in bluegill (Lepomis macrochirus) and Pumpkinseed (L. gibbosus) sunfish: environmental variation and the importance of ontogenetic niche shifts. Canadian Journal of Fisheries and Aquatic Sciences 45:17-26.
Padilla, D. K., and S.C. Adolph. 1996. Plastic inducible morphologies are not always adaptive: the importance of time delays in a stochastic environment. Evolutionary Ecology 10:105-117
Paszkowski, C. A., W. M. Tonn, J. Piironen, and I. J. Holopainen. 1990. Behavioral and population-level aspects of intraspecific competition in crucian carp. Annales Zoologici Fennici 27:77-85.
Penttinen, O.-P., and I. J. Holopainen. 1992. Seasonal feeding activity and ontogenetic dietary shifts in crucian carp, Carassius carassius. Environmental Biology of Fishes 33: 215-221.
Pettersson, L. B., and C. Bronmark. 1993. Trading off safety against food: state dependent habitat choice and foraging in crucian carp. Oecologia 95:353-357.
Prejs, A. 1973. Experimentally increased fish stock in the pond type lake Warniak. IV. Feeding of introduced and autochthonous non-predatory fish. Ekologia Polska 30: 465-505.
Priede, I. G. 1985. Metabolic scope in fishes. Pages 33-64 in P. Tytler and P. Calow, editors. Fish energetics: new perspectives. Croom Helm, London, England.
Reznick, D. N., and B. Braum. 1987. Fat cycling in the mosquitofish (Gambusia affinis): fat storage as a reproductive adaptation. Oecologia 73:401-413.
Riessen, H. P. 1984. The other side of cyclomorphosis: why Daphnia lose their helmets. Limnology and Oceanography 29:1123-1127.
Robinson, B. W., D. S. Wilson, and G. O. Shea. 1996. Tradeoffs of ecological specialization: an intraspecific comparison of pumpkinseed sunfish phenotypes. Ecology 77: 170178.
Scheiner, S. M. 1993. MANOVA: multiple response variables and multispecies interactions. Pages 94-112 in S. M. Scheiner and J. Gurevitch, editors. Design and analysis of ecological experiments. Chapman and Hall, New York, New York, USA.
Schultz, J. C. 1988. Plant responses induced by herbivores. Trends in Ecology and Evolution 3:45-49.
Sih, A., P. H. Crowley, M. A. McPeek, J. W. Petranka, and K. Strohmeier. 1985. Predation, competition and prey communities: a review of field experiments. Annual Review of Ecology and Systematics 16:269-311.
Spitze, K. 1992. Predator-mediated plasticity of prey lifehistory and morphology: Chaoborus americanus predation on Daphnia pulex. American Naturalist 139:229-247.
Stearns, S.C. 1989. The evolutionary significance of phenotypic plasticity. BioScience 39:436-445.
----- 1992. The evolution of life histories. Oxford University Press, New York, New York, USA.
Tanasichuk, R. W., and W. C. Mackay. 1989. Quantitative and qualitative characteristics of somatic and gonadal growth of yellow perch (Perca flavescens) from Lac Ste. Anne, Alberta. Canadian Journal of Fisheries and Aquatic Sciences 46:989-994.
Tollrian, R. 1995a. Predato-induced morphological defenses: costs, life history shifts, and maternal effects in Daphnia pulex. Ecology 76:1691-1705.
-----. 1995b. Chaoborus crystallinus predation on Daphnia pulex: can induced morphological changes balance effects of body size on vulnerability? Oecologia 101 151-155.
Tonn, W. M., I. J. Holopainen, and C. A. Paszkowski. 1994. Density-dependent effects and the regulation of crucian carp populations in single-species ponds. Ecology 75-824834.
Tonn, W. M., C. A. Paszkowski, and I. J. Holopainen. 1989. Responses of crucian carp populations to differential predation pressure in a manipulated pond. Canadian Journal of Zoology 67:2841-2849.
von Ende, C. N. 1993. Repeated-measures analysis: growth and other time-dependent measures. Pages 113-137 in S. M. Scheiner and J. Gurevitch, editors. Design and analysis of ecological experiments. Chapman and Hall, New York, New York, USA.
Walls, M., H. Caswell, and M. Ketola. 1991. Demographic costs of Chaoborus-induced defences in Daphnia pulex: a sensitivity analysis. Oecologia 87:43-50.
Webb, P. W. 1975. Hydrodynamics and energetics of fish propulsion. Bullentin of the Fisheries Research Board of Canada 190:1-159.
Weisberg, S. 1993. Using hard-part increment data to estimate age and environmental effects. Canadian Journal of Fisheries and Aquatic Sciences 50:1229-1237.
Weisberg, S., and R. V. Frie. 1987. Linear models for the growth of fish. Pages 127-143 in R. Summerfelt and G. Hall, editors. Age and growth of fish. Iowa State University Press, Ames, Iowa, USA.
Wootton, R. J. 1985. Energetics of reproduction. Pages 231-254 in P. Tytler and E Calow, editors. Fish energetics: new perspectives. Croom Helm, London, England.
|Printer friendly Cite/link Email Feedback|
|Author:||Pettersson, Lars B.; Bronmark, Christer|
|Date:||Sep 1, 1997|
|Previous Article:||Nonadditive effects of terrestrial and aquatic predators on juvenile estuarine fish.|
|Next Article:||Stoichiometry of fishes and their prey: implications for nutrient recycling.|