Temporal changes in the relative abundance of amphibians relative to riparian buffer width in western Washington, USA

In certain regions of Canada and the United States, amphibians have been shown to be sensitive to forest management and the short-term adverse effects of timber harvesting on terrestrial amphibians have been well documented in the literature. However, the effects of silvicultural treatments on amphibians may not be realized until many years after treatment. Manipulative forest management experiments with pre- and post-treatment data, random assignment of replicates to treatments, and analyses over a large spatial scale are rare in the literature. Furthermore, even studies with pre- and post-treatment data can be affected by ecological time lags; the longevity of some species may mean that it could take several years to observe a treatment effect. Riparian areas in forested landscapes of the Pacific Northwest have been recognized for high species richness of amphibians yet little is known about the persistence of amphibians in mature forest fragments such as unharvested riparian buffers surrounded by clearcuts. Our study investigated the efficacy, 2 and 10 years post-logging, of two different riparian management zone prescriptions established around third- and fourth-order streams in western Washington to mitigate the effects of upland timber harvest on amphibians. Our findings suggest that there is no global response by terrestrial amphibians to logging or to the retention of riparian management zones. Rather, species showed individual responses that varied over time and between treatments and transects. Specifically Coastal Tailed Frogs (Ascaphus truei) declined in clearcut upland habitats 2- and 10-years following logging while Western Red-backed Salamander (Plethodon vehiculum) and Ensatina (Ensatina eschscholtzii) appear to benefit from the recommended riparian management zone width. The influence of site effects was evident for some species, whereas for others there is some suggestion of treatment effects, but the analyses were confounded by patterns of natural population change at both local and regional scales, which were evident 2 years following logging. Our results exemplify the need for longer-term studies to ensure that potentially confounding factors, such as natural population fluctuation, are considered when interpreting the results of time series data. (C) 2012 Elsevier B.V. All rights reserved.