LAUSR

Abstract Restoration practices are much needed on clearcut openings in commercial forest landscapes where some mammal species have declined in abundance from a loss of preferred food, cover, and other components of stand structure. Retention of excess woody debris in piles and windrows provides habitat for forest-floor small mammals and some of their predators such as small mustelids. However, it is unknown if these retention habitats are used over longer periods (>10 years) as new forests grow and develop on harvested sites, or do they become unoccupied? We tested the hypotheses (H) that (H1) abundance, species richness, and diversity of the forest-floor small mammal community, and (H2) reproduction and survival of the southern red-backed vole (Myodes gapperi), long-tailed vole (Microtus longicaudus), and deer mouse (Peromyscus maniculatus) would decline on sites with woody debris structures, compared with sites of dispersed woody debris or uncut forest, up to 12 years post-construction in south-central British Columbia, Canada. Woody debris structures provided habitat on new clearcuts for M. gapperi at comparable or higher abundance than in uncut forest and 5.0 to 7.6 times higher than on dispersed sites in the first 5-year period. Although numbers declined after the initial three years, populations of M. gapperi in debris structures recovered to earlier abundance levels at 11–12 years post-construction. Mean abundance of M. longicaudus was consistently higher (2.8 to 3.5 times) in piles and windrows than in sites with dispersed woody debris over the first 5-year period. Populations of M. longicaudus were high in all three treatment sites at 11 years post-construction reaching mean annual peak numbers of 24, 42, and 36 voles per ha in the dispersed, piles, and windrow sites, respectively. Mean abundance of P. maniculatus was similar among treatment sites and consistent over time. Mean abundance of total small mammals was consistently higher (1.8 to 2.4 times) in piles and windrows than dispersed or forest sites in the first 5-year period and this pattern was continued at 11–12 years post-construction. At 11 years post-construction, all treatment sites had the highest peak numbers per ha in the study: dispersed (40.3), piles (64.1), windrows (56.1), and forest (29.0). Our results did not support H1 as abundance, species richness, and diversity were increased or maintained in the debris structures over the 12-year period. Reproduction and survival followed the pattern of abundance for the major species, and hence H2 was not supported. Our study is the first to measure long-term (up to 12 years) responses of forest-floor small mammals to constructed piles and windrows of woody debris as a means of habitat retention on clearcuts. These mammalian species, particularly voles, may then serve as prey for marten and other mustelids. This relationship provides further support for piles and windrows to act as baseline trophic structures in ecological restoration of cutover forest land.