Movement behavior and stream network structure predict salamander abundance

Background/Question/Methods

Movement behavior is implicitly assumed to be important in determining the response of amphibian populations to habitat loss and fragmentation.  However, most studies that evaluate the distribution and abundance of amphibians across different spatial scales do not directly account for movement behavior.  This study uses information on movement behavior for two stream amphibians, the northern dusky salamander and the seal salamander, to predict spatial patterns of abundance.  Prior mark-recapture studies revealed upstream bias in movement for both species.  Upstream bias in movement should result in populations having higher abundances in upper reaches of headwater streams.  To test this prediction, temporary removal sampling was used to collect data on these species at 59 sites across the Shenandoah National Park, Virginia, USA.  Stream habitat, including the amount of cover objects and flow, was quantified at the time of sampling.  We also calculated the total length of streams and the number of tributaries within watersheds to quantify stream network structure, and we calculated catchment area and betweenness centrality to measure position within stream networks.  This information was used with the salamander survey data to estimate the abundance of both species with N mixture models.    

Results/Conclusions

Patterns of abundance for northern dusky and seal salamanders were largely similar.  At the scale of the survey site, both species were more abundant when more cover objects were present.  At the scale of stream networks, both species were more abundant in watersheds with more tributaries and more streams.  Within stream networks, both species were also more abundant in sites that had smaller catchment areas, which is consistent with predictions about abundance based on their movement behavior.  However, northern dusky salamanders were present at lower abundances in places with higher betweenness centrality, but seal salamander abundance did not respond to this.  This difference may result from stronger upstream bias in movement of northern dusky salamanders.  Even after accounting for local habitat structure and the total amount of habitat within the landscape, the position within stream networks still affects salamander abundance, and this is consistent with knowledge about the movement behavior of these species.  As ecologists seek to better understand amphibian population dynamics, including information about movement behavior and the structure of habitat networks may provide important insights into conservation planning.