Prey species that are unprofitable to attack often share conspicuous colours and patterns with other coexisting defended species. This phenomenon, termed müllerian mimicry, has long been explained as a consequence of selection on defended prey to adopt a common way of advertising their unprofitability. However, studies using two unpalatable prey types have not always supported this theory. Here we show, using a system of humans hunting for computer-generated prey, that predators do not always generate strong selection for mimicry when there are two unprofitable prey types. By contrast, we demonstrate that when predators are faced with a range of different prey species, selection on unprofitable prey to resemble one another can be intense. Here the primary selective force is not one in which predators evaluate the profitabilities of distinct prey types independently, but one in which predators learn better to avoid unprofitable phenotypes that share traits distinguishing them from profitable prey. This need to simplify decision making readily facilitates the spread of imperfect mimetic forms from rarity, and suggests that müllerian mimicry is more likely to arise in multispecies communities.

Female-limited colour polymorphism occurs in many damselfly species, where one morph resembles the male (andromorph) and the other is dissimilar (gynomorph). Explanations for this phenomenon vary, but most assume that andromorphism has arisen in odonates, as a response to excessive male harassment. Here, we quantify the extent of continental and seasonal variation in female morph frequencies in a widely-distributed damselfly and ask whether the spatiotemporal patterns in andromorph frequency can be understood on the basis of sexual harassment theory. We sampled the damselfly, Nehalennia irene (Hagen) among regions across Canada, and at several sites, over the reproductive season, within Central Canada. Andromorph frequencies ranged from 0 to > 90% across Canada. In particular, sites in Western Canada had consistently high andromorph frequencies, whereas andromorph frequencies among Central sites were lower and variable and, among Eastern sites, were lower still (except one site) and relatively invariant. For populations in Central Canada, both andromorph frequencies and population densities varied significantly over time, reaching a peak mid-season. As expected, morph frequency covaried significantly with estimates of male harassment in some cases, but estimates of male harassment did not consistently account for variation in morph frequencies within all regions. Additional factors such as genetic drift may influence morph frequency at the edge of a species' range. Future work also should test, and attempt to explain causation, for seasonal variation in morph frequency.

2006. Sex biases in dispersal and philopatry: insights from a meta-analysis based on capture Ámark Á recapture studies of damselflies. Á Oikos 113: 539 Á547.Sex-biased dispersal is well known for birds and mammals, typically by females and males, respectively. Little is known about general patterns of sex-biased dispersal in other animal taxa. We reviewed return rates for a model group of invertebrates (damselflies) and explored putative costs and benefits of dispersal by males and females. We used published capture Ámark Árecapture data and examined whether a sex bias existed in likelihood of recapture at least once, at both emergence and/or breeding sites. We assessed whether this metric of likelihood of recapture was indicative of dispersal or philopatry, and whether any emerging pattern(s) were consistent across damselfly families. Using a meta-analysis, we found a higher likelihood of recapture at least once for males than for females at both natal sites and breeding sites, which seemed attributable to higher female-biased dispersal, although female-biased mortality cannot be discounted particularly for some species. Sex biases in dispersal among damselflies may be understood based on sex differences in maturation rate and foraging behaviour, both of which should affect the costs and benefits of dispersing. This hypothesis may be useful for explaining patterns of dispersal in other animal taxa.

No abstract

No abstract