Investigation of dispersal ecology and evolution using laboratory populations of Drosophila melanogaster

Abstract

The spatial distribution of populations is shaped by phenomena such as range shifts and range expansions. The process underlying these phenomena (i.e. dispersal), however, is often demography- and context-dependent. The complex nature of dispersal has attracted much attention, particularly with respect to factors such as sex bias, density dependence and inter-individual variation. Investigating how dispersal is affected by these factors can help us obtain a better understanding of why and when organisms undertake the potentially risky and costly process of dispersal.

Using a combination of long-term and single-generation microcosm experiments on Drosophila melanogaster, I investigated both population-level and individual-level patterns of dispersal.

At the population level, I found that D. melanogaster showed negative density-dependent dispersal, even in the absence of established causes such as territoriality and sociality, likely via male mate harassment at high densities. Further experiments using dispersal-selected flies revealed that dispersal evolution countered these negative DDD patterns, implying that upon evolution, dispersal in these flies became context-independent. Finally, I demonstrated the occurrence of significant mate-finding dispersal in the aftermath of a sex-biased dispersal event, which can help ameliorate Allee effects in small populations, but slow down dispersal evolution in large populations.

At the individual level, I investigated the sex differences in phenotypic trait correlates of dispersal, showing that they not only depend on the ecological context, but may also change completely as a result of dispersal evolution. Next, I focused on the dual role of desiccation resistance as a cause vs. cost of dispersal in D. melanogaster, finding that, while desiccation-sensitive flies are the earliest dispersers, they cannot cover large distances owing to the energy costs of dispersal.

The population-level results highlight the role of density and mate availability in determining the effective patterns of dispersal. The individual-level results underline the importance of sex differences in phenotypic correlates of dispersal, including the role of stress in shaping dispersal patterns. Together, the results also showcase how evolution can dramatically change both population-level and individual-level patterns in unexpected ways. Through effects on the dynamics and distribution of newly found populations, these findings have potential consequences for higher-scale spatial processes such as range shifts and biological invasions.