2017 ESA Annual Meeting (August 6 -- 11)

PS 16-14 - Effects of ecosystem engineering by beavers on reproduction of pond-breeding amphibians

Tuesday, August 8, 2017
Exhibit Hall, Oregon Convention Center
John M. Romansic, School of Biological Sciences, Washington State University, Vancouver, WA, Deanna H. Olson, USDA Forest Service, Pacific Northwest Research Station, Corvallis, OR and Jimmy D. Taylor, Animal and Plant Health Inspection Service, US Department of Agriculture, Corvallis, OR
Background/Question/Methods

Ecosystem engineers directly influence other species by altering physical habitat characteristics, sometimes dramatically. Ecosystem engineers may also affect species indirectly by altering their interactions with predators, competitors, parasites, and mutualists. We investigated the effects of ecosystem engineering by North American beavers (Castor canadensis) on the relationship between amphibians and their trout predators in a Cascadian mountain meadow containing a combination of connected streams, beaver-dammed pools, and undammed pools as well as unconnected, isolated pools. We used visual encounter surveys to assess the distribution of trout across dammed pools, undammed pools, and dam cells (stream-pool sections bordered by dams) to evaluate whether dams exclude trout from some lentic habitat patches. In addition, we quantified the presence and absence of breeding by Pacific treefrogs (Hyliola [Pseudacris] regilla), Cascades frogs (Rana cascadae), and long-toed salamanders (Ambystoma macrodactylum), all of which breed in pools. We used these data to address two competing hypothesis: 1) habitat changes in beaver-dammed pools facilitate breeding by amphibians when trout are present, potentially decreasing negative effects of trout on amphibians, and 2) isolated, fishless pools are more important for amphibian breeding than pools connected to the trout-occupied aquatic network, regardless of the presence or absence of dams.

Results/Conclusions

We observed trout in 75% of dammed pools and 82% of dam cells, suggesting that beaver dams excluded trout from little or no habitat. Few stream-connected pools supported breeding of any amphibian species, regardless of the presence or absence of dams. Isolated, seasonal pools lacking fish were more likely than stream-connected pools to be used by amphibians for breeding, although this pattern was significant only for Pacific treefrogs. Despite spatial autocorrelation of isolated pools, our results are consistent with previous studies that highlight the importance of trout-free habitats for amphibian reproduction. We suggest that beaver are more likely to influence trout-sensitive amphibians by changing the availability of isolated pools rather than by restricting trout distribution or promoting successful reproduction in the presence of trout. We are now investigating whether beavers increase or decrease the amount of isolated pool habitat using a suite of beaver-dammed and reference meadows. We are also testing an assemblage of six amphibian species for positive and negative associations with beaver activity. First year results suggest that northwestern salamanders (A. gracile) preferentially breed in meadows containing inactive beaver dams, emphasizing the importance of a shifting mosaic of active and inactive beaver-influenced habitat patches for maintaining amphibian biodiversity.