Use selective harvesting instead of clearcutting
Overall effectiveness category Likely to be beneficial
Number of studies: 8
Background information and definitions
Clearcutting of large areas of forest can have substantial impacts on associated fauna. Selective logging is the removal of selected trees within a forest based on criteria such as diameter, height or species. Remaining trees are left in the stand, as opposed to clearcutting where all trees are felled. This intervention is similar to several others that involve harvesting some, but not all, trees. In this case, tree removal was largely based on forestry specifications, rather than designed spatially to retain undisturbed patches. This intervention covers a wide range of tree removal intensities. In some cases, management is for shelterwood, a specific forestry practice that involves gradually removing mature trees to allow growing space for younger trees that initially germinate in partial shade.
See also Fell trees in groups, leaving surrounding forest unharvested, Retain undisturbed patches during thinning operations, Use thinning of forest instead of clearcutting and Use patch retention harvesting instead of clearcutting.
Supporting evidence from individual studies
A replicated, site comparison study in 1980 of a forest in Nova Scotia, Canada (Swan et al. 1984) found that selectively harvested plots, cut as shelterwood, did not host more small mammals than did clearcut plots. In shelterwood plots, average capture rates (10–31 small mammals/100 trap nights) did not differ significantly from those in clearcuts (12–27 small mammals/100 trap nights). The forest had regrown following fire 80 years previously. Three plots (average 3.6 ha) were clearcut 3–5-years previously and two plots (average 1.9 ha) were shelterwood cut, entailing removing a proportion of harvestable timber. Shelterwood plots had an average tree stem basal area of 9.4 m2/ha (compared to 25.9 m2/ha in adjacent unharvested forest). Small mammals were surveyed using snap traps for four consecutive nights and days, one or twice in each plot in July–August 1980.Study and other actions tested
A site comparison study in 1974–1977 of three mixed forest blocks in Maine, USA (Monthey 1986) did not find more snowshoe hares Lepus americanus in partially harvested forest than in largely clearcut forest. In a partially harvested forest, a lower proportion of transect sections (7.9%) contained hare tracks compared to in a largely clearcut forest (17.6%). However, patches of unharvested trees were included within the clearcut forest sampled, and tracks were most numerous in or close to these. Hare tracks were surveyed, 1–2 days after snowfall, over the winters of 1974–1975, 1975–1976 and 1976–1977. Tracks were counted on 15-m sections along 50 km of permanent lines through clearcut and partially harvested forest. Partial harvesting occurred in 1974–1977 and the clearcut forest was harvested in 1960–1975.Study and other actions tested
A replicated, controlled, before-and-after study in 1994–1998 of a coniferous forest in British Colombia, Canada (Klenner & Sullivan 2003) found that when forest was harvested by single tree selection, one of four small mammal species was more abundant relative to clearcutting. Populations of all species did not differ between plots assigned for different treatments in the year before harvesting. After harvesting, there were more southern red-backed voles Clethrionomys gapperi in single tree selection plots (20.8–44.0/ha) than in clearcuts (0.1–10.8/ha). Long-tailed vole Microtus longicaudus was less abundant in single tree selection than clearcut plots (0.0–3.4 vs 2.6–16.2/ha) as was northwestern chipmunk Tamias amoenus (0.8–1.4 vs 1.9–6.0/ha). Deer Mouse Peromyscus maniculatus numbers were similar between treatments (single tree selection: 0.4–4.0/ha; clearcuts: 0.8–5.0/ha). Forest stands were c.30 ha. There were three replicates each of single tree selection (removing 33% of timber volume) and 10-ha clearcuts, harvested in winter 1994–1995. Small mammals were live-trapped in 1994–1998, over two consecutive nights, at 3-week intervals, from June or July to August or September.Study and other actions tested
A replicated, controlled study in 1997–1998 of a forest in Maine, USA (Fuller et al. 2004) found more small mammals in selectively harvested forest stands than fully harvested, regenerating stands. Annual average catches were higher in partially harvested than fully harvested stands for the three most abundant species; red-backed vole Clethrionomys gapperi (partially harvested: 12.4–22.1; fully harvested: 2.5–5.0 voles/grid), deer mouse Peromyscus maniculatus (partially harvested: 4.9–12.5; fully harvested: 0–2.5 mice/grid) and short-tailed shrew Blarina brevicauda (partially harvested: 4.3–5.0; fully harvested: 0–3.0 shrews/grid). These comparisons were not tested for statistical significance. Seven stands were selectively harvested between 1992 and 1995, with 52–59% of basal tree area removed and 13 m2/ha basal area remaining. Two forest stands were clearcut between 1974 and 1984 and treated with the herbicide, glyphosate, 3–8 years post-harvest. Small mammals were surveyed in live trap grids, between 22 June and 28 July 1997 and between 21 June and 31 July 1998.Study and other actions tested
A replicated, controlled, before-and-after study in 1991–1997 of two second-growth forests in Arkansas and Oklahoma, USA (Perry & Thill 2005) found that selectively harvesting isolated trees did not increase small mammal abundance relative clearcutting. Before harvesting, average small mammal abundances did not differ significantly between stands planned for different treatments (single tree selection: 2.7 small mammals/100 trap nights; clearcut: 0.9). Similarly, after harvesting, small mammals numbers did not differ significantly between single tree selection stands (6.4/100 trap nights) and clearcut stands (10.7). In each of four blocks of second-growth forest (59–69 years old at start of study), one stand was managed by single tree selection and one was clearcut, harvested in summer 1993. Tree basal area after harvesting was 15–16 m2/ha in single tree selection plots (compared to 24–32 m2/ha in unharvested forest). Stand extent was 13–28 ha. Small mammals were surveyed using an average of 67 Sherman live traps/stand, pre-harvest in 1991 and 1992, and post-harvest in 1995, 1997 and 1999. Traps were operated for seven consecutive nights during winter (December–January).Study and other actions tested
A replicated, controlled study in 1994–1997 of Douglas-fir Pseudotsuga menziesii forest in British Colombia, Canada (Klenner & Sullivan 2009) found that selective harvesting of trees increased one of four small mammal species abundance in the third and fourth, but not first and second, year after harvesting relative to clearcutting. There were more southern red-backed voles Myodes gapperi in the third and four year in all selectively logged treatments (6–17/plot) than in clearcut stands (0–1/plot), but similar numbers between treatments in the first two years (selective cut: 33–42/plot; clearcut: 13–34/plot). There were no differences between treatments for deer mouse Peromyscus maniculatus (selective cut: 1–15/plot; clearcut: 6–21/plot) or northwestern chipmunk Tamias amoenus (selective cut: 0–6/plot; clearcut: 0–6/plot). There were more meadow voles Microtus pennsylvanicus in clearcut stands (selective cut: 0–2/plot; clearcut: 3–14/plot). Forest stands, 20–25 ha in extent, were partially harvested in winter 1993–1994. Two each had 20% of timber volume removed by individual-tree selection, 35% removed by individual-tree selection on 50% of the area and 50% volume removed by individual-tree selection. These were compared with two 1.6-ha clearcut areas. Small mammals were live-trapped, at 2–4-week intervals, in May–October of 1994, 1995, and 1996 and in April–May 1997.Study and other actions tested
A systematic review in 2008 of 56 studies of small mammal responses to partial harvesting, clearcutting or wildfire in North American forests (Zwolak 2009) found that partially harvested forests had more red-backed voles Myodes gapperi, but not deer mice Peromyscus maniculatus than did clearcut forests. Absolute abundances are not presented but vole numbers in partially harvested stands, 1–9 years after harvesting, were significantly higher than in clearcut stands. Deer mouse abundances did not differ significantly between partially harvested and clearcut stands. Meta-analyses were carried out on studies identified following a defined literature search procedure.Study and other actions tested
A replicated, site comparison study in 2006–2007 in a mixed temperate forest in Quebec, Canada (Le Blanc et al. 2010) found that harvesting trees selectively did not result in higher small mammal species richness or abundance compared to clearcutting. Small mammal species richness did not vary along a gradient of retained conifer basal area that resulted from different felling densities (result presented as statistical model coefficient). The combined abundances of red-backed voles Myodes gapperi, masked shrews Sorex cinereus and deer mice Peromyscus maniculatus (which comprised 92% of individuals caught) did not vary with conifer basal area (result presented as statistical model coefficient). Four tree blocks were harvested in 2004–2005. Three or four harvesting treatments (each 20 ha extent) were applied in each block. Selective harvesting resulted in retention of 17–23%, 57–69% or 60–73% of standing timber. Clearcut areas had <10% of timber remaining. Small mammals were live-trapped, between 3 July and 25 August in 2006 and 2007.Study and other actions tested