CAUSES OF VARIATION IN DISPERSAL DISTANCE IN THE STREAM SALAMANDER GYRINOPHILUS PORPHYRITICUS

Dispersal is expected to evolve as an adaptive mechanism to optimize individual fitness across the landscape. While there is evidence that active dispersers base emigration decisions (i.e., stay vs. leave) on perceived costs associated with environmental variation and inbreeding, it is less well understood how and whether these same factors influence dispersal distances– a more comprehensive measure of dispersal. More generally, the challenge of quantifying dispersal in the field has resulted in a paucity of data on the fate and fitness of dispersing individuals, leaving us with little understanding of the factors influencing individual variation in dispersal distance. In my dissertation, I use a combination of morphological, performance, demographic, and genetic data to understand the selective forces shaping variation in dispersal distances in the stream salamander Gyrinophilus porphyriticus. I found that phenotypic attributes that facilitate long-distance dispersal restrict other locomotor performances. Specifically, salamanders that dispersed farther in the field had longer forelimbs, but swam at slower velocities under experimental conditions. These results suggest that salamanders disperse by walking, and that longer limbs may lower the cost of transport by increasing stride length. Longer limbs also impose more drag, potentially explaining the reduced swimming performance of long-distance dispersers. These results are novel in demonstrating a trade-off associated with variation in dispersal distance, and, more broadly, suggest that this and other trade-offs associated with continuous variation in dispersal distance may constrain dispersal evolution. I show that large-scale, long-term environmental variation – reflected in survival probabilities of G. porphyriticus – better predicts dispersal distances than current, local variation in habitat quality. These results provide the first empirical support for early theory that treated dispersal as an innate, ‘fixed’ quality of individuals that evolves in response to a history of spatiotemporal environmental variability at large spatial scales. Importantly, these results challenge the current paradigm that most dispersal is conditional and based on gathering information about local habitat quality. Based on these findings, I develop a conceptual model of dispersal evolution where informed strategies explain short-distance dispersal, and fixed strategies explain long-distance dispersal. I provide rare empirical support for the basic prediction that inbreeding risk decreases with dispersal distance. Further, I show that the degree to which dispersal functions to reduce inbreeding risk in G. prophyriticus is mediated by other environmental conditions influencing dispersal distance. Specifically, dispersal effectively reduced inbreeding risk in downstream reaches where dispersal distances were greater. In contrast, dispersal did not reduce inbreeding risk in upstream reaches, where dispersal distances were shorter. These results suggest that selective pressures influencing dispersal distances in G. porphyriticus can vary at fine spatial scales (i.e., reach-scale), with resulting consequences on inbreeding risk. Population genetic data indicated that inbreeding avoidance is likely not the primary driver of dispersal distance, but downstream and upstream reaches differ in many abiotic and biotic factors (i.e., discharge, streamwater chemistry, substrate size, prey and predator communities) that may explain differences in dispersal distances. Collectively, my dissertation research provides empirical insight on the causes of individual variation in dispersal distance and constraints on the evolution of dispersal. My work demonstrates that data on dispersal distances are crucial for disentangling the relative importance of the many selective pressures influencing dispersal in natural populations. Previous studies have shown that long-distance dispersal is predicted to contribute disproportionately to range shifts in response to climate change and persistence in fragmented habitats. Therefore, understanding the processes promoting and constraining long-distance dispersal in natural populations may help to address several pressing applied issues