COS 47-3
The importance of terrestrial dispersal for connectivity among headwater salamander populations

Tuesday, August 11, 2015: 2:10 PM
324, Baltimore Convention Center
William L. Miller, Department of Biological Sciences, Towson University, Towson, MD
Joel W. Snodgrass, Department of Biological Sciences, Towson University, Towson, MD
Gail E. Gasparich, Department of Biological Sciences, Towson University, Towson, MD
Background/Question/Methods

Headwater species are organisms that are primarily constrained to the upstream terminus of river networks either permanently or during critical phases in their life cycles.  Many have multi-phasic life cycles with limited capacities for both in-stream and overland dispersal.  Previous research has emphasized the utilization of multiple dispersal pathways as an important component of the population dynamics of headwater species.   Overland dispersal, in particular, is suggested to contribute to the stability and persistence of headwater populations.  However, the relative importance of overland dispersal to the connectivity of vertebrate headwater populations, such as salamanders, remains largely unexplored.  We used six microsatellite markers to examine regional genetic population structure and assess dispersal patterns for a headwater salamander, Desmognathus fuscus.  We collected tissue samples from 312 salamanders from a total of 16 streams.  Streams were sampled using a nested hierarchical design that specifically addressed the contribution of overland movements to headwater population connectivity.  

Results/Conclusions

Genetic divergence was significant among all populations (pairwise Fst = 0.027 – 0.405) and at all hierarchical spatial scales.  However, lower degrees of genetic population structuring were observed among proximal headwater streams that shared no downstream hydrologic connections (Fst = 0.027) than among headwater streams of the same drainage (Fst = 0.091).  Additionally, Mantel tests indicated a significant relationship between genetic divergence and both overland and in-stream distances.  However, the relationship between genetic and overland distances was significantly greater (β = 0.00134; R2 = 0.193) than that observed between genetic and stream distances (β = 0.00018, R2 = 0.157).    Estimates of effective migration rates suggest average rates of dispersal are low both within and among watersheds (Nem < 1 individual / generation).  Our results suggest that overland dispersal, while rare, does significantly contribute to the connectivity of headwater salamander populations.  These findings highlight the importance and need for integrative management approaches that incorporate the conservation of both upland and stream habitat in headwater areas.  Future research should address how terrestrial fragmentation may influence dispersal resistance among headwater populations.