Factors Influencing Sexual Size Dimorphism in Temperate Waterstriders

Abstract

Sexual size ratios (mean female length divided by mean male length) of 12 species of temperate waterstriders in the subfamily Gerrinae (Hemiptera, Gerridae) are analyzed to test both potentially nonadaptive and adaptive hypotheses for the evolution of sexual size dimorphism. Females tend to be larger than males, and regression of female length on male length (both log-transformed) reveals significant hypoallometry for size dimorphism. Both adaptive and nonadaptive explanations for the evolution of this allometry are discussed. Of these, only a nonadaptive model is consistent with the available data. Two closely related genera, Limnoporus and Gerris, are included in the data set, but covariance analysis reveals no significant effect of phylogeny on patterns of variation in size dimorphism. It is suggested, however, that the basic female bias may reflect descent from a common, highly dimorphic ancestor, rather than adaptation within this taxon. Four adaptive hypotheses are considered. Comparisons of the development times of males and females reveal that the female-biased size dimorphism is not associated with protandry, and the developmental-constraints hypothesis is therefore rejected. The hypothesis that selection for migration by flight tends to reduce sexual size dimorphism is supported by intraspecific comparisons of long-winged and short-winged morphs. A positive correlation between mating duration and residuals from the body-size regression supports the hypothesis that loading constraints associated with prolonged pairing select for increased size dimorphism. The available data on mating systems and size-selective mating support the hypothesis that sexual selection for large males promotes reduced or even reversed size dimorphism in some species, whereas sexual selection for large females increases size dimorphism in other species. Thus, patterns of size dimorphism in this taxon may be influenced by at least three, potentially interacting, selective processes: selection for dispersal by flight, selection for reduced loading during prolonged pairing, and sexual selection. The relative influence of these factors in three groups of ecologically similar species is discussed. The general conclusions of this study are consistent with the results of recent comparative studies of size dimorphism in primates. In both taxa, sexual size dimorphism is most strongly influenced by potentially nonadaptive factors, and the residual variation reflects the action of several selective factors. The complexity of size-ratio evolution revealed by these studies suggests that patterns of sizeratio variation within and among taxa, or size ratios typical of single species, are unlikely to be satisfactorily explained by reference to single hypotheses or to selective hypotheses only.