97th ESA Annual Meeting (August 5 -- 10, 2012)

PS 1-32 - Effects of habitat composition on movement and optimal dispersal strategies in juvenile spotted salamanders (Ambystoma maculatum)

Monday, August 6, 2012
Exhibit Hall, Oregon Convention Center
Shannon E. Pittman, Fisheries, Wildlife, and Conservation Biology, University of Minnesota, St. Paul, MN, Grant M. Connette, Fisheries & Wildlife, University of Missouri, Columbia, MO and Raymond D. Semlitsch, Division of Biological Sciences, University of Missouri, Columbia, MO
Background/Question/Methods

Natal dispersal is a critical movement phase for many species, particularly those undergoing ontogenetic habitat shifts. Behavioral strategies of natal dispersers in response to human-altered habitat have far-reaching implications for functional connectivity and local population dynamics. Juvenile pond-breeding spotted salamanders (Ambystoma maculatum) are forest-dependent and exhibit considerable variation in natal dispersal behaviors such as speed of movement, path linearity, and settlement propensity. Habitat alteration and landscape change may alter the benefits associated with movement strategies. We quantified the fine-scale movement behavior of juvenile spotted salamanders during initial natal dispersal in different habitat types using powder tracking. We experimentally-released 200 recently-metamorphosed salamanders in 3 different habitat types (mature forest, grassland, and 5-year old clearcut) and followed their movements for 1-3 nights. We used this empirical data to parameterize spatially-explicit, individual-based movement models to investigate the effects of landscape composition on dispersal success. Specifically, we investigated how distance of natal wetland from forest and the size of terrestrial buffer zones affected optimal movement and settlement strategies of dispersing salamanders. 

Results/Conclusions

Habitat type affected path linearity (P=0.05), with salamanders moving straighter in low-quality (grassland) habitat. Movement of salamanders in forest habitat was adequately described by a correlated random walk model (CRWDiff=0.595, P=0.1617, n=84). Juveniles exhibited orientation to forest habitat when released distances of up to 10m away from forest (P=0.05, n=92), but did not significantly orient at distances of 20m or farther from forest (P=0.39, n=74). Results from our movement model indicated that the size of buffer zone affected the optimal responsiveness of juveniles to microhabitat. Continuous forest and large buffer zones favored straighter movement paths and less attention to microhabitat, while smaller buffer zones favored more tortuous paths and high settlement propensity.  All edge simulations favored high settlement propensity and more tortuous paths. These results indicate that juvenile movement is affected by habitat type and that modified landscapes may favor alternative dispersal strategies.