COS 47-2
Transience and scale-dependence in dispersal-phenotype correlations

Tuesday, August 11, 2015: 1:50 PM
324, Baltimore Convention Center
Christopher A. Searcy, Biology, University of Toronto Mississauga, Mississauga, ON, Canada
Benjamin Gilbert, Ecology and Evolutionary Biology, University of Toronto, Toronto, ON, Canada
Martin Krkosek, Ecology & Evolutionary Biology, University of Toronto, Toronto, ON, Canada
Locke Rowe, Ecology & Evolutionary Biology, University of Toronto, Toronto, ON, Canada
Shannon J. McCauley, Biology, University of Toronto Mississauga, Mississauga, ON, Canada
Background/Question/Methods: Ecological and evolutionary models often assume that dispersers have the mean phenotype observed across the entire metapopulation. However, there are strong theoretical expectation for phenotype-dispersal correlations to exist, and growing empirical evidence that they do. We took advantage of a unique experimental landscape to investigate three dispersal-phenotype correlations in the green frog (Rana clamitans). We captured green frogs at different time points as they colonized an array of 36 newly created and geographically dispersed experimental ponds. At the same time points, we also captured a sample of green frogs from the nearby source ponds for comparison. For each frog we measured body size, residual hindlimb length, and body condition.

Results/Conclusions: After only one month, a positive correlation was detected between dispersal and body size. After a second month, this correaltion had disappeared, but phenotypic differentiation between the entire population of dispersers and philopatric individuals remaining at the source ponds had increased. This indicates that phenotype-dispersal correlations are both transient and dependent upon the spatial scale being examined. This may explain why so few emperical examples of body size-dispersal correlations have been documented, despite strong theoretical expectation for their existence. We found no correlation between dispersal and residual hindlimb length, the trait previously most closely tied to anuran dispersal, and a negative correlation between dispersal and body condition, possibly indicating the energetic costs of dispersing. The fact that we detected a positive correlation between dispersal and body size, the trait most likely to represent fitness, indicates that dispersal-phenotype correlations need to be considered in models of metapopulation and metacommunity dynamics, as they may influence which patches are classified as sources and sinks and how comeptition-colonization tradeoffs can allow for coexistence in the metacommunity.