According to much current theory, organisms should be able to balance the conflicting demands of the need to feed efficiently and the need to avoid preadtors while feeding. In an experimental conflict situation, it was possible to evaluate the relative fitnesses associated with the available choices and to compare the observed behaviors with predictions derived from fitness considerations. A backswimmer, Notonecta hoffmanni, was capable of balancing these two conflicting factors adaptively.

Communities of larval amphibians in fishless habitats often differ strikingly from those in habitats with fish. We surveyed larvae of 15 amphibian species to determine if presence or absence of specific defenses against fish was correlated with breeding habitat. Each species was tested for two key defenses: unpalatability and chemically mediated predator avoidance. In eight of nine cases, larvae of species that often encounter fish had at least one of these defenses. In contrast, larvae of seven species that breed in fishless pools consistently lacked defenses against fish. Lack of appropriate defenses appeared to be a primary reason why temporary pool species cannot successfully coexist with predatory fishes in permanent habitats. Palatability and responses to chemical cues from fish often differed among closely related taxa and were correlated strongly with frequency of encounter with fish. Thus, natural selection rather than phylogeny best explains interspecific variation in antipredator defenses. Our data show that members of at least two orders and four families of amphibians use chemical cues to reduce predation risk from predatory fish.

No abstract

We examined the effects of the patch size of catnip, Nepeta cataria, on pollinator visitation rates and pollinator limitation. The most important floral visitors were honey bees (Apis mellifera), solitary bees (Halictidae), and bumble bees (Bombus spp.). Our first goal was to see how spatial variation in patch size affected the rate at which individual flowers received pollinator visits (visitation rate). Visitation rate was higher in larger patches for honey bees and bumble bees, but lower for solitary bees. Patch size explained 74-83% of the variation in visitation rate. Intraspecific isolation also had an effect: isolated patches received relatively few visits. Visitation rate depended both on visitor abundance and on the proportion of flowers entered during one visit. All three visitor types wee more abundant in larger patches, i.e., flowers showed mutual attraction of pollinators. Relative to a visit to a small patch, during a visit to a large patch, honey bees visited more flowers but a lower proportion of flowers; solitary bees visited fewer flowers and thus a lower proportion of flowers; and bumble bees visited not only more flowers but a higher proportion of flowers as well. Thus within patches, flowers competed for visits from honey bees and solitary bees but showed facilitation regarding bumble bee visits. Our second goal was to relate patch size and visitation rates to immediate pollinator limitation. Comparisons of the percent of capsules setting seed (fruit set) of open-pollinated, hand-pollinated, and bagged flowers were used to evaluate pollinator limitation. The ranking of fruit set was: hand-pollinated > open-pollinated > bagged. Patch size did not affect fruit set in hand-pollinated or bagged flowers; however, for open-pollinated flowers, fruit set was lower in smaller patches. patch size explained 63% of the variation in pollinator limitation. The effect of patch size comes through its effects on visitation rates. A multiple regression model using visitation rates of the three types of visitors as independent variables explained 67% of the variation in pollinator limitation.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.Abstract. Field and laboratory experiments demonstrated that the degree to which an aquatic insect, Notonecta hoffmanni, shows avoidance behavior towards an important predator is proportional to the magnitude of the risk of predation. Laboratory experiments showed that I, II, and III instar juvenile N. hoffmanni can suffer significant mortality due to predation from adult N. hoffmanni whereas IV and V instars suffered no mortality from adults. In particular, the relative risk of predation from adults was I > II > III > IV = Vno risk. Based on these data I predicted that I, II, and III instars should avoid adults whereas IV and V instars should not, and that the relative degree of avoidance should be I > II > III > IV = V = no avoidance. Two measures of predator avoidance were documented: alterations in the use of space, and reductions in movement.In both the laboratory and the field the effects of adults on the behavior of juvenile notonectids were determined by comparing the juvenile's behavior when adults were experimentally removed to that in controls where adults were present. Alterations in spatial patterns were as predicted, i.e., in both the laboratory and the field only I, II, and III instars avoided adults by altering their use of space, and the degree to which spatial patterns were altered was highly correlated with the relative risk of predation. In the laboratory, reductions in movement were also as predicted, i.e., only I, II, and III instars reduced their movement in the presence of adults and the reduction in movement was highly correlated with the risk of predation. However, in the field, only I and III instars avoided adults by moving less; i.e., II instars did not conform to expectations. Accordingly, in the field the correlation between the reduction in movement and the risk of predation was positive but not significant.In the laboratory, the effects of adults on juvenile feeding rates were also documented. Because I and II instars avoided adults by altering their spatial and movement patterns, they suffered a cost in reduced feeding rates. Unexpectedly, III instars altered their feeding behaviors but suffered no reduction in feeding rates. Partial correlation analysis showed that a decrease in the juvenile's use of the central region of water bodies (where prey were four times more dense) resulted in a significant decrease in their feeding rates, but that a reduction in their movement had no such effect.I then discuss the effects on the predictions of optimal foraging theory of some common mechanisms by which notonectids avoid their predators. Because the degree of avoidance is related to the risk of predation I also discuss the possibility that feeding behavior...