A prominent interspecific pattern of sexual size dimorphism (SSD) is Rensch's rule, according to which male body size is more variable or evolutionarily divergent than female body size. Assuming equal growth rates of males and females, SSD would be entirely mediated, and Rensch's rule proximately caused, by sexual differences in development times, or sexual bimaturism (SBM), with the larger sex developing for a proportionately longer time. Only a subset of the seven arthropod groups investigated in this study exhibits Rensch's rule. Furthermore, we found only a weak positive relationship between SSD and SBM overall, suggesting that growth rate differences between the sexes are more important than development time differences in proximately mediating SSD in a wide but by no means comprehensive range of arthropod taxa. Except when protandry is of selective advantage (as in many butterflies, Hymenoptera, and spiders), male development time was equal to (in water striders and beetles) or even longer than (in drosophilid and sepsid flies) that of females. Because all taxa show female-biased SSD, this implies faster growth of females in general, a pattern markedly different from that of primates and birds (analyzed here for comparison). We discuss three potential explanations for this pattern based on life-history trade-offs and sexual selection. Online enhancement: appendix table.abstract: A prominent interspecific pattern of sexual size dimorphism (SSD) is Rensch's rule, according to which male body size is more variable or evolutionarily divergent than female body size. As-

This chapter uses data for 489 spider species from fifteen families to describe patterns of variation in sexual size dimorphism (SSD), and to evaluate hypotheses explaining these patterns. The direction and magnitude of SSD is found to depend strongly on the size measure chosen, and the use of carapace width is recommended because it is less affected by condition than body mass or length. Comparative analyses reveal that spiders do not exhibit allometry consistent with Rensch's rule. Instead, females appear to have diverged more than males over evolutionary time, and male and female body size show uncorrelated co-evolution, which is unusual for animals. Only two adaptive hypotheses — fecundity selection favouring large size in females and gravity selection favouring small size in males — have general explanatory power for patterns of SSD in spiders. However, processes may differ among species and comprehensive studies of selection within given species are needed.

Males of some cannibalistic species of spiders and insects appear to sacrifice themselves by allowing the female to eat them, and the adaptive significance of such drastic terminal reproductive investment has recently been demonstrated for a spider. Typically, the female has to kill the male, but it has been suggested that males of some species in the cannibalistic orb-weaving spider genus Argiope may die in copula without female 'collaboration'. Here, we provide the first experimental evidence to our knowledge of programmed sudden death after onset of copulation in males of the spider Argiope aurantia. Our observations reveal that males invariably die during the insertion of their second pedipalp, regardless of whether they mate with newly moulted, defenceless females or with older mature females that often attack them. We determined experimentally that the death of males is triggered immediately upon insertion of the second palp, when males become unresponsive, and heartbeat ceases within minutes of insertion. We discuss the possible adaptive significance of programmed death during copulation, and argue that male death has evolved in a context other than sexual cannibalism.

Mate search plays a central role in hypotheses for the adaptive significance of extreme female-biased sexual size dimorphism (SSD) in animals. Spiders (Araneae) are the only free-living terrestrial taxon where extreme SSD is common. The "gravity hypothesis" states that small body size in males is favoured during mate search in species where males have to climb to reach females, because body length is inversely proportional to achievable speed on vertical structures. However, locomotive performance of males may also depend on relative leg length. Here we examine selection on male body size and leg length during mate search in the highly dimorphic orb-weaving spider Argiope aurantia, using a multivariate approach to distinguish selection targeted at different components of size. Further, we investigate the scaling relationships between male size and energy reserves, and the differential loss of reserves. Adult males do not feed while roving, and a size-dependent differential energy storage capacity may thus affect male performance during mate search. Contrary to predictions, large body size was favoured in one of two populations, and this was due to selection for longer legs. Male size was not under selection in the second population, but we detected direct selection for longer third legs. Males lost energy reserves during mate search, but this was independent of male size and storage capacity scaled isometrically with size. Thus, mate search is unlikely to lead to selection for small male size, but the hypothesis that relatively longer legs in male spiders reflect a search-adapted morphology is supported.