Search

Phenomenological and behavioural studies have greatly advanced the study of natural selection. Field studies of selection well appraise the natural situation, but is this also true for laboratory studies, which are typically more mechanistic? We compared precopulatory sexual selection (mating differential based on pairing success) in field and laboratory of several closely related, ecologically similar black scavenger dung flies (Diptera: Sepsidae). Selection on fore femur (sexual trait) and wing size (nonsexual trait) and shape varied considerably among seven species and continental populations in agreement with variation in their mating system and sexual size dimorphism. Selection on trait size was mostly positive or nil, but never significantly negative, implying mating advantages of large males in most species. Strongest selection was found in species/populations with male-biased size dimorphism, associating evolutionary shifts from female- to male-biased dimorphism with intensified sexual selection for large male size by adding male –male competition to a mating system previously driven primarily by female choice. Although sexual selection on shape was closely aligned with allometric shape variation, selection on fore femur shape was more consistent than selection on wing shape, which was absent in most species. Sexual selection intensities, but not necessarily the underlying behavioural mechanisms, were overall similar in field and laboratory, suggesting that laboratory assessments well represent the natural situation. If this conclusion can be generalized, it would lend credence to the strategy of using controlled laboratory mating studies to better understand natural selection, behaviour and ecology, at least for smaller animals that can be held in captivity.

Reproductive traits often evolve rapidly, and some suggest that behavioural traits, in particular, can diverge faster than morphology, resulting in sexual isolation between populations/species. An earlier study of a Neotropical dung fly, Archisepsis diversiformis (Diptera: Sepsidae), reported anecdotally that two central American populations that were approximately 500 km apart (Costa Rica and Panama) differed in male courtship behaviour despite being morphologically similar. Here, I present results of an in-depth study designed (1) to test whether these two populations show qualitative and/or quantitative differences in mating behaviour and morphology, (2) to test whether individuals from either population show some degree of reproductive isolation and (3) to characterize population variation in a particularly fast evolving mitochondrial gene fragment, cytochrome oxidase c subunit I (COI), to estimate genetic differences between the two populations. Despite similarities in overall courtship, I identified behaviours that were clearly population specific, and report that both populations showed strong premating isolation in one-on-one crosses. However, after extended exposure in mass-container group crosses, individuals did produce adult F1 offspring, suggesting that isolation is incomplete. Surprisingly, morphometric analyses indicated that these two populations differed significantly in sexually monomorphic adult wing shape but differed only moderately in sexually dimorphic male forefemur shape, and not at all in male genital clasper shape. Finally, both populations were genetically similar, forming a single, monophyletic cluster with low uncorrected pairwise distances (COI threshold