Response of breeding bird communities to forest harvest around seasonal ponds in northern forests, USA

We examined response of breeding bird communities to varying levels of timber harvest in and around 16-forested seasonal ponds in northern Minnesota, USA. This experimental study employed a before-after-control-impact design with three different harvest treatments. Treatments were assigned randomly (n = 4 ponds/treatment) and were applied within 17 m wide buffers outward from the ponds' edge: clear-cut harvest (reduction of basal area to < 2m(2)/ha), partial cut harvest (reduction of basal area to 7-10 m(2)/ha), and no harvest (no cut). Forest stands around treatment buffers (n = 12) were clear-cut harvested (ranging from 6.5 to 12.5 ha). Ponds with no harvest in the adjacent forests (controls) or buffers surrounding the ponds (n = 4) were maintained throughout the 5-year study. Prior to harvest, we found no significant difference (P > 0.05) in bird community composition around seasonal ponds versus nearby forest habitat, suggesting that seasonal ponds do not affect bird community structure in a mature forest setting at this scale. Overall bird numbers and species richness increased (P < 0.05) in all pond buffers compared to controls over the 4 years after harvest. Increases in bird numbers on treated versus control pond buffers were found across all migration and nesting guilds, and among the forest edge guild. Bird community species composition also changed within the treated versus control pond buffers after harvest. Differences in bird communities among treatments were small the first year after harvest, but continued to diverge from controls over the 4 years after harvest. Bird communities of the clear-cut treatment were most dissimilar to controls. Both the partial and no cut buffer bird communities were more similar to the control than the clear-cut treatment. Treated pond buffers had more birds associated with early successional habitat. In contrast, many interior forest-associated bird species did not occur in any of the buffers after harvest. We found no difference in breeding bird community composition between pond buffers and other residual patches left on harvested sites, but there was a significant difference between harvest treatments when we combined pond buffer and residual patch birds on each site. Early successional habitat-associated bird species were more abundant in residual patches on sites that had a clear-cut pond buffer and forest-associated species were more abundant in residual patches on sites with no cut pond buffers. Habitat for mature forest-associated bird species can be maintained on harvest sites by leaving no cut or partial cut buffers around seasonal ponds or in similar sized residual patches in other areas of the harvest. (c) 2006 Elsevier B.V. All rights reserved.