Sexual segregation is widespread throughout the animal kingdom. Although a number of hypotheses have been proposed to account for observed patterns, the generality of the mechanisms remains debated. One possible reason for this is the focus on segregation patterns in large mammals such as ungulates, where the majority of studies are descriptions of a single population. Here, we present the results of a cross‐population comparison of patterns of sexual segregation in the Trinidadian guppy, Poecilia reticulata. We relate observed patterns to experimental quantification of predation risk and sexual harassment of females by males in eight populations. We find that the degree of segregation increases with predation risk, with deeper waters becoming increasingly female biased. Furthermore, we observed that levels of male harassment are lower in deeper water but only in those rivers that contain major guppy predators. We conclude that sexual segregation in guppies is consistent with the predation risk hypothesis: sexual segregation results from a combination of predation risk driving males (the more vulnerable sex) into less risky habitats and females gaining benefits of reduced sexual harassment by remaining in high‐predation environments.

Males and females are often defined by differences in their energetic investment in gametes. In most sexual species, females produce few large ova, whereas males produce many tiny sperm. This difference in initial parental investment is presumed to exert a fundamental influence on sex differences in mating and parental behavior, resulting in a taxonomic bias toward parental care in females and away from parental care in males. In this article, we reexamine the logic of this argument as well as the evolutionarily stable strategy (ESS) theory often used to substantiate it. We show that the classic ESS model, which contrasts parental care with offspring desertion, violates the necessary relationship between mean male and female fitness. When the constraint of equal male and female mean fitness is correctly incorporated into the ESS model, its results are congruent with those of evolutionary genetic theory for the evolution of genes with direct and indirect effects. Male parental care evolves whenever half the magnitude of the indirect effect of paternal care on offspring viability exceeds the direct effect of additional mating success gained by desertion. When the converse is true, desertion invades and spreads. In the absence of a genetic correlation between the sexes, the evolution of paternal care is independent of maternal care. Theories based on sex differences in gametic investment make no such specific predictions. We discuss whether inferences about the evolution of sex differences in parental care can hold if the ESS theory on which they are based contains internal contradictions.

Investigations of the micromorphology of rabbit tibial articular cartilage using scanning and transmission electron microscopy revealed that the collagenous elements in the tissue form fluid-containing tubular structures. The commonly described radial or deep zone longitudinal fibres were found to be tubular structures with internal diameters of 1-2 microm. The walls of the tubules were composed of tightly packed fibrils of collagen. The tangential zone, close to the tibial plateau, was composed mainly of a spongy arrangement of collagen fibrils, containing bunches of tangentially lying small (< 1 microm) diameter tubules. The application of conventional chemical fixation techniques resulted in the fine detail of this tissue being obscured. When the tissue was frozen, followed by cryo-scanning electron microscopy or freeze-drying, prior to observation in the scanning electron microscope the tubule structures were not obviously present. It was only by applying freeze-substitution techniques, followed by critical point drying or resin embedding, that the structure was revealed clearly. Segregation of water into ice crystals did occur during the freezing process, but the formation of those crystals played no part in creating the tubular morphology observed. A similar structure was still revealed following pre-treatment with glycerol, methanol or Triton X-100, provided that concentration of these additives was not too high. The walls of the tubules in the radial region were composed of straight, longitudinally arranged as well as helically arranged, 30 nm diameter fibrils. The lumen of the tubules appears to be lined by a circumferentially arranged array of approximately 10 nm diameter fibres, spaced at regular intervals of 50-70 nm.

No abstract