Provide or retain un-harvested buffer strips
Overall effectiveness category Trade-off between benefit and harms
Number of studies: 8
Background information and definitions
Provision or retention of forest strips in areas subject to timber extraction may be undertaken with the purported objectives of helping to mitigate the effects of loss of forest cover for woodland flora and fauna, as well as reducing potential problems such as soil erosion. Nature conservation aims include retaining valuable old forest features such as older trees with cavities and dead wood affording nest site and foraging opportunities for woodland birds. A similar intervention is ‘Ensure connectivity between habitat patches’, discussed in ‘Habitat protection’.
Supporting evidence from individual studies
A replicated study in balsam fir Abies balsamea stands in Quebec, Canada (Darveau et al. 1995) in 1989-92 found that 60 m-wide riparian forest buffer strips retained forest-dwelling breeding bird abundances and a species composition more similar to uncut areas than that of narrower strips. For one year before and three years following clearcutting, birds were surveyed in five buffer strips: 20 m-, 40 m-, 60 m-, and more than 300 m-wide (i.e. undisturbed control) strips, and a 20 m-wide thinned (33% of trees removed) strip. After initial increases in bird densities (30-70%) in all strips in the year after cutting, the 20 m- and 40 m-wide strips exhibited greatest decreases. Three years after cutting, forest species were less abundant (four songbirds becoming virtually absent) than habitat generalists in the 20 m strips (the thinned 20 m strip had densities around 20% less than the unthinned 20 m-wide strip).Study and other actions tested
A replicated, controlled study from June-July in 1994 in five mainstem buffer strips (60-80 m wide), five tributary buffer strips (20-40 m wide) and five continuously forested control sites within commercial forests in Maine, USA (Vander Haegen & Degraaf 1996) found that avian nest predation rates were significantly higher in the buffer strips than in control sites (31% predation in tributary buffer strips, 23% in mainstem buffers vs. 15% in control sites). Red squirrels Tamiasciurus hudsonicus and blue jays Cyanocitta cristata were responsible for > 50% of depredations. The authors suggest leaving wide (> 150 m) buffer strips along riparian zones to reduce edge-related nest predation. Artificial nests were placed at five points (100 m apart) along transects that ran parallel to the stream.Study and other actions tested
A replicated, controlled study from May-June in 1994 in 12 riparian buffer-strip (18-70 m wide) sites and four unlogged riparian sites of old-growth forest in Oregon, USA (Hagar 1999) found that 27 species were recorded at unlogged and 42 species at buffer-strip sites: eight species were more abundant in unlogged and 12 species more abundant in buffer-strip sites. Four species that were more abundant in unlogged stands increased with increasing width of riparian buffers. However, four other species were rarely observed in even the widest buffers sampled (40-70 m). Overall, bird species richness and abundance were not related to buffer-strip width, but the author recommends buffer widths >40 m and maintaining a high density of trees within the buffer.Study and other actions tested
A replicated, controlled paired sites study from June-July in 1994-5 in 16 pairs of forested buffer strips (20-50 m) and undisturbed riparian coastal forest in Newfoundland, Canada (4), found that bird abundance was higher in the buffer strips (average of 10.5 individuals/transect for buffer strips vs. 7.9 for control sites), total species richness was similar (7.2 species/transect in buffers vs. 6.2 in controls) but that three of six specialist forest species were absent. Abundance of forest generalist, interior forest, and riparian species were similar between buffers and controls and did not increase in wider buffers. Buffer strips were adjacent to 3-5 year-old clear-cuts (> 10 ha) and were typically > 300 m long.Study and other actions tested
A replicated controlled before-and-after study in a managed Douglas-fir Pseudotsuga menziesii forest in Washington, USA, (Pearson & Manuwal 2001), found that 31 m wide riparian buffer strips contained bird communities that were more similar to control (unharvested) forests than 14 m strips. Forest species (black-throated grey warbler Dendroica nigrescens, golden-crowned kinglet Regulus satrapa and brown creeper Certhia americana) decreased in buffer treatments (especially the narrow buffer) relative to controls. Species of shrubby habitats (dark-eyed junco Junco hyemalis, cedar waxwing Bombycilla cedrorum and song sparrow Melospiza melodia) increased in one or both buffer treatments. Birds were surveyed in 18 sites (six of each treatment) in both pre-harvest (spring 1993) and post-harvest (1995 and 1996) years.Study and other actions tested
A replicated controlled trial along a stream in Minnesota, USA (Hanowski et al. 2005), found that bird species responded differently to timber harvest in riparian buffers, and that any amount of harvest affected breeding bird communities. Along the stream, 30 m wide forest buffers were established within plots with four treatments (3 plots/treatment): 1) no harvest in buffer; 2) reduction of tree basal area to 7-10 m²/ha; 3) reduction to 2 m²/ha (i.e. clear-cut); and 4) no harvest in buffer or adjacent upland forest. Bird surveys were conducted 1 year prior to and for 4 years after, harvest. In the first year after harvest, bird community composition changed in all buffer treatments relative to control plots, and diverged over time. More individuals and species (primarily associated with edge or early-successional habitats) colonized harvested buffers; abundances and species richness of interior forest species declined.
A replicated, randomised, controlled study from May-June in 2001 and May-July in 2002 in 24 buffer-strip blocks and 18 continuous, old-growth forest blocks in coastal, temperate forests in Alaska, USA (Kissling & Garton 2008) found that species richness was similarly distributed across treatments and controls (both averaged 15 species / 100 detections). Species richness and diversity were greatest in the narrowest buffers (< 100 m) but species composition in the largest buffers (? 400 m) was most similar to that in control blocks. Only 3 of 10 common species differed in abundance across buffer treatments and controls (2 were positively and 1 was negatively related to buffer width). The authors conclude that forested beach buffers ? 250 m wide can support densities of forest-associated birds similar to that of natural stands but rare or uncommon species will benefit most from buffers ? 400 m in width.Study and other actions tested
A replicated study in 2003-4 in old-growth forest on Prince of Wales Island, Alaska, USA (Sperry et al. 2008), there were no significant differences in average clutch size or number of young fledged across six species between nests in narrow (<250 m) buffers at four sites, compared to wide (>350 m) buffers at three sites. The buffers surrounded areas of 8-18 ha of forest and 76 nests of six species (Pacific-slope flycatcher Empidonax difficilis, chestnut-backed chickadee Poecile rufescens, winter wren Troglodytes troglodytes, Swainson’s thrush Catharus ustulatus, hermit thrush C. guttatus and varied thrush Ixoreus naevius) were monitored. Of the 25 (18%) of nests that did not fledge young, 23 failed due to predation. Daily survival rates were slightly higher (0.2 to 2.5%) in wide buffers.Study and other actions tested