Legacy retention versus thinning: Influences on small mammals

Management strategies for promoting late-seral attributes in second-growth forest need evaluation for their efficacy in maintaining biodiversity, incl uding complete forest-floor, small-mammal communities. Two common strategie s in the Pacific Northwest are (1) management with thinnings to promote lar ge trees with developed understories and (2) retention of legacies, defined as live trees, logs, and snags from the preceding forest, at harvest, foll owed by protection but not thinnings of the new stand. We compared small-ma mmal communities resulting from >65 yr of application of these strategies i n the Puget Trough, Washington. We also compared these communities with the small-mammal communities found in old-growth, naturally young, and extensi vely managed forests elsewhere in western Washington. Forests managed with thinnings had 1.5 times the individual mammals and 1.7 times the mammal bio mass of forests managed with legacies of coarse woody debris and snags-diff erences similar to those between old-growth and naturally young forest (1.2 times more individuals in old-growth) and old-growth and extensively manag ed forest (1.6 times more individuals in old-growth). Management strategy h ad a profound impact on community structure, with the Columbian mouse (Pero myscus oreas), the small mammal most associated with old growth, much reduc ed in Puget Trough forests (absent from most stands) and the creeping vole (Microtus oregoni) (a species commonly associated with early seral stages, but found in all seral stages in Washington) third-ranked in thinned stands but seventh ranked in legacy stands. The montane shrew (Sorex monticolus) was second-ranked, after Trowbridge's shrew (S. trowbridgii), in marked con trast to codominance by the southern red-backed vole (Clethrionomys gapperi ), S. monticolus, and P. oreas in old growth. Thus, neither strategy produc ed communities typical of late-seral forests.