Initial response of small ground-dwelling mammals to forest alternative buffers along headwater streams in the Washington Coast Range, USA

Abstract

We assessed the short-term effects of alternative designs of forested buffer treatments along headwater streams on small ground-dwelling mammals in managed forests in western Washington, USA. Over three summers (one pretreatment and two posttreatment), we trapped 19 mammalian species along 23 streams in the northern Coast Range. We compared faunal communities in unlogged controls with riparian treatments: (1) clearcuts (no buffer) to the stream bank, (2) clearcuts with continuous buffer strips from a few to 20m each stream side, and (3) clearcuts with small patch (<0.6ha) buffers left to protect sensitive streamside terrain. We estimated abundance (Nˆ) of eight common and three less common species with simple models and capture indices, respectively, and analyzed effects of treatments and community measures on relative abundance compared to controls and over time. Habitat and forest stand changes after treatments included reduced live standing trees and canopy, more dead down wood, and increased ground plant species richness, evenness, and diversity, particularly in the patch and no buffers. The forest obligate Peromyscus keeni significantly declined in all treatments. No other species declined in strip buffers. Species with statistically greater Nˆ in treatments included the early-seral microtine Microtus oregoni, the semi-arboreal generalist Tamias townsendii and the old forest specialist Myodes gapperi in all treatments; the generalist Peromyscus maniculatus in strip and no buffers; and the early-seral riparian Zapus trinotatus in patch buffers. Postlogging trends in Nˆ were negative for two common shrews, Sorex monticolus and Sorex trowbridgii. Moist soil denizen Neurotrichus gibbsii capture rates were significantly lower in clearcut no buffers. Captures were low for two species of semi-aquatic shrews, Sorex bendirii and Sorex palustris, but in general, S. palustris responded poorly to treatments. Windfall was an important proximate influence on buffer function because persistence was low and created structural diversity in dynamic change. The alternative buffers were ineffective in retaining the habitat and relative abundance of individual small mammal species. Species composition similarity also significantly declined in the clearcuts. Change in species richness, evenness, diversity and total mammals captured in treatments however were not different. Control streams maintained faunal biodiversity but abundance of about half of the species changed following treatment. Steep topography, weather events, and small buffer size advanced habitat changes that influenced small mammal response to treatments. Site-specific data are necessary in logging unit planning along headwater streams, if the management goal is the maintenance of communities observed in undisturbed forest.