Sexual harassment by males has been reported from several live-bearing fishes (Poeciliidae) and has been shown to inflict costs on females. For example, poeciliid females have reduced feeding opportunities when accompanied by a male because females dedicate attention to avoiding male copulation attempts. Poeciliid species differ considerably in male mating behavior, such as the presence or absence of courtship. Courting males display in front of the females, but males attempting to sneak-copulate approach females from behind, that is, in the blind portion of their visual field, and force copulations, which can be viewed as a male persistence trait. We predicted that poeciliid females need to be more vigilant in the presence of noncourting males, and costs of harassment by noncourting males might be stronger. In a comparative approach we examined the costs of male sexual harassment for females as reduced feeding time in 9 species of live-bearing fishes, including courting (Poecilia latipinna, Poecilia reticulata, Xiphophorus cortezi, Xiphophorus variatus) and noncourting species (Poecilia mexicana [surface-and cave-dwelling form], Poecilia orri, Gambusia affinis, Gambusia geiseri, Heterandria formosa). In all species examined except for the cave form of P. mexicana, focal females spent significantly less time feeding in the presence of a male than when together with another female. The time females spent feeding was found to significantly decline with increasing male mating activity (sum of all sexual behaviors), but there was no support for the idea that females would spend more time feeding in the presence of courting males compared with noncourting ones.

When multiple groups of organisms experience similar environmental gradients, their patterns of differentiation might exhibit both shared and unique features. Here, we investigated the relative importance of three factors in generating body shape variation in a livebearing fish, Gambusia caymanensis, inhabiting the Cayman Islands: (i) shared patterns of divergent selection between predator regimes (presence/absence of piscivorous fish) driving replicated morphological differentiation, (ii) historical island effects yielding different morphologies across the three islands and (iii) unique effects of predation on morphological differentiation within each island. Shared effects of predation proved much more important than historical or unique effects. Populations coexisting with piscivorous fish exhibited larger caudal regions and smaller heads than conspecifics found in the absence of predatory fish. These results match a priori predictions, and mirror recent findings in a number of fish species, suggesting predation might often drive predictable morphological trends in disparate fishes. However, interestingly, the sexes achieved this morphological pattern through different means: head depth, caudal peduncle length and depth in males; head length, caudal peduncle depth in females. In G. caymanensis, we quantitatively confirmed that predation intensity represents a primary driver of body shape differentiation.

Relatedness strongly influences social behaviors in a wide variety of species. For most species, the highest typical degree of relatedness is between full siblings with 50% shared genes. However, this is poorly understood in species with unusually high relatedness between individuals: clonal organisms. Although there has been some investigation into clonal invertebrates and yeast, nothing is known about kin selection in clonal vertebrates. We show that a clonal fish, the Amazon molly (Poecilia formosa), can distinguish between different clonal lineages, associating with genetically identical, sister clones, and use multiple sensory modalities. Also, they scale their aggressive behaviors according to the relatedness to other females: they are more aggressive to non-related clones. Our results demonstrate that even in species with very small genetic differences between individuals, kin recognition can be adaptive. Their discriminatory abilities and regulation of costly behaviors provides a powerful example of natural selection in species with limited genetic diversity.

The maintenance of sexual reproduction is still a largely unresolved question in evolutionary biology, and one of the most puzzling aspects of this is the coexistence of sexual and asexual females. This often leads to widespread niche overlap, posing an interesting challenge to competitive exclusion theory. In this study, we investigate how the aggressive behaviors between females in a sexual/unisexual mating complex of mollies (Poecilia latipinna and P. formosa) differ, and the effects these behaviors have on both the performer and receiver of aggressive acts. We exposed females to five treatments: (1) alone (control); (2) a large and small sexual female; (3) a large sexual female and a small unisexual female; (4) a small sexual female and a large unisexual female; and (5) a large and small unisexual female. We found that P. formosa females were on average more aggressive over time when compared to their sexual hosts, P. latipinna. The females of P. formosa, however, incurred a higher cost of being aggressive. Large, aggressor females of this species had a lower body fat content than their conspecific recipients. By contrast, P. latipinna females showed no differentiation between aggressors and recipients. Here, we combine life history methodology and behavioral observations to show that female competition can in fact be quantitatively measured and that these aggressive behaviors are costly to perform. The adaptive value of this interspecies aggression is not yet completely understood, but sexual competition, access to resources, and position within the shoal are all possible functions.