Background information and definitions
Thinning is a forestry operation that involves removing some trees in order to allow remaining trees to grow faster, or straighter or otherwise to produce better quality timber. It may especially be applied in young forest, a few years after onset of regeneration or planting. Thinning increases light that reaches the forest floor, potentially adding to habitat diversity, and may enable remaining trees to produce higher quality forage for herbivores.
The evidence summarised for this intervention includes one case where trees were selectively thinned to increase overwinter browse availability for deer and one where combinations of thinning and felling of groups of trees were combined. See also Use thinning of forest instead of clearcutting.
Supporting evidence from individual studies
A replicated, site comparison study in 1990–1991 of aspen Populus tremuloides forest at four sites in Minnesota, USA (Christian et al. 1996) found that in thinned tree stands, there was a greater abundance of small mammals, but a similar species richness compared to in unthinned stands. The average yearly site abundance of small mammals was greater in thinned stands (12–29 individuals/grid) than in unthinned stands (9–19 individuals/grid). Species richness did not differ between stand treatments (thinned: 2.8–5.3 species/grid; unthinned: 3.0–5.7 species/grid). Aspen stands at four sites had been growing for 9–11 years at time of thinning. Two had been thinned one year prior to sampling, one seven years previously and one 11 years previously. Unthinned stands were also surveyed at each site. Stands were 6–74 ha in extent. Small mammals were surveyed using snap traps, over two nights and one day, in July–September 1990 and 1991. Stands had 2–7 grids, of 64 traps each.Study and other actions tested
A site comparison study in 1996 of forest in Quebec, Canada (St-Louis et al. 2000) found that, following tree thinning through a partial forest cut aimed at increasing browse availability, white-tailed deer Odocoileus virginianus made proportionally greater use of the cut area than of the forest as a whole. Deer use of the cut area (estimated at 15,170 deer-days/km2) was higher than in the forest as a whole (estimated 2,808 deer-days/km2). However, deer did not move home ranges and only animals whose ranges overlapped the cut area used it. A partial forest cut, across 43 ha, was made in January–February 1996. This thinned the forest by removing approximately 40% of deciduous tree stems (with conifers and understorey trees retained). Deer use of the cut area was determined by counting pellet groups, on 27 and 28 April 1996, in eighty-four 2 × 40-m plots. This was compared with estimated pellet density in the whole forest area (total 25 km2) that was based on pellet production from an estimate of the overall deer population. Habitat selections of 30 individual deer were monitored by radiotracking, in January–April 1996.Study and other actions tested
A replicated, controlled, before-and-after study in 1994–1996 of four coniferous forest sites and a replicated, site comparison study in 1995–1996 of eight coniferous forest sites, all in Oregon, USA (Suzuki & Hayes 2003) found that thinning trees increased abundances of some small mammal species. Out of 12 species, abundances of three, deer mouse Peromyscus maniculatus, creeping vole Microtus oregoni and white-footed vole Arborimus albipes, increased in thinned plots during the two years post-thinning relative to in unthinned plots. Pacific jumping mouse Zapus trinotatus increased in thinned plots relative to in unthinned plots between the first and second years post-thinning. Seven species had similar abundances in each treatment. Western red-backed vole Clethrionomys californicus was less common in thinned than in unthinned plots. Capture rates did not significantly differ between plots before thinning. See paper for data. Of nine species, five, Pacific shrew Sorex pacificus, Trowbridge’s shrew Sorex trowbridgii, vagrant shrew Sorex vagrans, creeping vole and Pacific jumping mouse, were more abundant in plots thinned 7–24 years previously than in unthinned plots. See paper for data. Four sites, each with three 35–45-year-old Douglas-fir stands (26–40 ha/stand) were studied. Two stands/site were thinned in 1994–1995 (to averages of 193–267 trees/ha) and one was unthinned (average 500 trees/ha). Also, at eight pairs of stands, 52–100 years old and <1 mile apart, one stand (10–28 ha) had been thinned 7–24 years before surveying and one (20–73 ha) was unthinned. Small mammals were surveyed within the controlled study using pitfall traps for six weeks/year in 1994 (before thinning) and in 1995 and 1996 (after thinning). In the site comparison study, pitfall traps were operated for 40 consecutive days in each 1995 and 1996.Study and other actions tested
A replicated, controlled study in 2000–2002 of three coniferous forest sites in British Columbia, Canada (Ransome et al. 2004) found thinning of lodgepole pine Pinus contorta stands resulted in higher numbers of northern flying squirrels Glaucomys sabrinus when resultant tree density was high, whilst thinning did not affect abundances of red squirrels Tamiasciurus hudsonicus. Average northern flying squirrel abundance was highest in thinned stands where remaining trees were at high density (4.6 squirrels/stand), intermediate in medium density stands (3.3/stand) and lowest in low density (1.3/stand) and unthinned (1.8/stand) stands. Red squirrel abundance did not differ between treatments (high density: 10.8/stand; medium density: 9.7/stand; low density: 13.5/stand; unthinned: 11.3/stand). In each of three sites, four forest stands, regenerating following felling and/or wildfire in 1960–1972, were studied. In 1988–1989, one stand each in each site was thinned to approximately 500 (low), 1,000 (medium), and 2,000 (high) stems/ha and one was unthinned (with 4,700–6,000 stems/ha in 1988). Squirrels were surveyed using Tomahawk live traps, at 4-week intervals, from May–October 2000 and 2001 and at 8-week intervals in 2002. One trapping grid (9 ha, 50 traps) was located in each stand.Study and other actions tested
A randomized, replicated, controlled, before-and-after study in 1994–2001 in a pine and oak forest area in Missouri, USA (Fantz & Renken 2005) found that thinning and partial harvesting of trees led to a higher abundance of Peromyscus mice spp. Two to five years after harvesting, the annual average number of mice caught in uneven-aged harvesting compartments, where single trees and small groups were felled (8.5–27.0 mice) and even-aged harvesting compartments, involving limited clearcutting and thinning (11.4–31.5 mice) were higher than in uncut compartments (5.9–10.0 mice). Catch data from two Peromyscus spp. were combined. Mice were live-trapped, in two blocks of 144 traps each, in nine compartments (312–514 ha), over six nights each year in April or May of 1994–1995 and 1998–2001. Compartments were grouped in three replicate blocks. Uneven-aged harvesting (three compartments) involved cutting single trees and small groups. Even-aged harvesting (three compartments) involved clearcutting and thinning 10–15 % of trees. Three compartments were uncut. Harvesting was carried out in 1996. Biomass removal was similar between harvesting treatments.Study and other actions tested
A replicated, site comparison study in 2000–2001 of coniferous forest across seven townships in Maine, USA (Homyack et al. 2005) found that thinned regrowing forest stands had more red-backed voles Clethrionomys gapperi and masked shrews Sorex cinereus, but not deer mouse Peromyscus maniculatus or short-tailed shrew Blarina brevicauda than did unthinned stands. More red-backed voles were caught in thinned (9.1/survey) than in unthinned (3.8/survey) stands. The same pattern held for masked shrew (6.8 vs 1.2). No significant abundance differences were detected for deer mouse (3.6 vs 4.4) or short-tailed shrew (6.0 vs 4.4). Twenty-four stands were felled in 1967–1983, herbicide-treated in 1977–1988 and thinned in 1984–1999. Thirteen stands were felled in 1974–1982 and herbicide-treated in 1982–1988 but not thinned. Small mammals were surveyed at 64 live-trapping stations/stand for six consecutive 24-h periods during June–August 2000 and again in 2001.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 thinning trees increased small mammal abundance relative to unthinned stands, but not to clearcut stands. Before management, average small mammal abundances were similar between stands planned for different treatments (thinning: 2.4 small mammals/100 trap nights; no thinning: 2.5; clearcut: 0.9). After management, more small mammals were caught in thinned stands (9.3/100 trap nights) than in unthinned stands (1.7) but a similar number was caught in clearcut stands (10.7). In each of four blocks of second-growth forest (59–69 years old at start of study), one stand was thinned, retaining 49–99 of the largest trees/ha, one was not thinned and one was clearcut. Tree removal was conducted in summer 1993. Stand extent was 13–28 ha. Small mammals were surveyed using an average of 67 Sherman live traps/stand, pre-management in 1991 and 1992, and post-management 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 2000–2002 of three coniferous forests in British Columbia, Canada (Sullivan et al. 2005) found that thinning of regenerating lodgepole pine Pinus contorta stands did not result in higher small mammal abundance or species richness 12–14 years later. Small mammal abundance varied between years but not between treatments (low remaining tree density: 13–26 individuals/stand; medium density: 11–23 individuals/stand; high density: 15–27 individuals/stand; unthinned: 10–26 individuals/stand). Similarly, species richness did not differ between treatments (low tree density: 2.3–4.3 species/stand; medium density: 3.7–3.9 species/stand; high density: 3.0–3.4 species/stand; unthinned: 2.5–3.7 species/stand). In each of three sites, four forest stands, regenerating following felling and/or wildfire in 1960–1972, were studied. In 1988–1989, one stand each in each site was thinned to approximately 500 (low), 1,000 (medium), and 2,000 (high) stems/ha and one was unthinned (with 4,700–6,000 stems/ha in 1988). Small mammals were live-trapped, over two nights and one day, at 4-week intervals, from May–October of 2000, 2001, and 2002. One trapping grid (1 ha, 49 trap stations) was located in each stand.Study and other actions tested
A replicated, site comparison study in 1999–2003 of two pine forest sites in British Columbia, Canada (Sullivan et al. 2006) found that thinning lodgepole pine Pinus contorta stands did not lead to greater use by mule deer Odocoileus hemionus or moose Alces alces. The average number of mule deer faecal pellet groups did not differ between thinned and unthinned stands in summer (thinned stands: 219/ha; unthinned stands: 73/ha) or winter (thinned: 378/ha; unthinned: 190/ha). Similarly, there was no significant difference between stands in the quantity of moose faecal pellet groups in summer (thinned: 7/ha; unthinnged: 7/ha) or winter (thinned: 16/ha; unthinned: 30 pellet groups/ha). Across the two sites, three forest stands in total were thinned in 1993 (to 1,000 stems/ha) and three were left unthinned. Stands had been clearcut in 1978–1982 and lodgepole pine had regenerated naturally. Faecal pellet groups were counted over a two-week period, five times in May and four times in October, in 55–145 plots/stands (plots were circles of 1.26 m radius), in 1999–2003.Study and other actions tested
A randomized, replicated, controlled study in 2000–2004 of five second-growth lodgepole pine Pinus contorta forests in British Colombia, Canada (Sullivan et al. 2007) found that in thinned stands, the abundances of snowshoe hare Lepus americanus, mule deer Odocoileus hemionus and moose Alces alces were not greater than in unthinned stands. Faecal pellet counts for snowshoe hares were not significantly different between low-density thinned plots (70,000 pellets/ha), medium-density thinned plots (60,000 pellets/ha), high-density thinned plots (38,000 pellets/ha) or unthinned plots (13,000 pellets/ha). Similarly, despite large count variations, no significant differences between treatments were detected for mule deer (low: 259 pellet clumps/ha; medium: 79; high: 33; unthinned: 13) or moose (low: 365 pellet clumps/ha; medium: 133; high: 188; unthinned: 93). In each of three areas, four stands (17–27 years old) were studied. One stand each was thinned to low (approximately 500 stems/ha), medium (1,000 stems/ha) and high (2,000 stems/ha) tree density in 1988–1989. One was unthinned (4,700–6,000 stems/ha at time of thinning). Treatments were assigned randomly within study areas. Mammal faecal pellets and clumps were surveyed in one hundred 5-m2 plots in each stand. Plots were cleared of pellets in early October 2000. Pellets were counted in spring 2004.Study and other actions tested
A replicated, controlled study in 2003–2008 of four lodgepole pine Pinus contorta forests in British Colombia, Canada (Sullivan et al. 2010) found that thinning did not increase forest stand use by snowshoe hares Lepus americanus, mule deer Odocoileus hemionus or moose Alces alces, relative to unthinned stands, 15–20 years after thinning. Hare faecal pellet density did not differ significantly between low (26,000 pellets/ha), medium (25,000 pellets/ha) or high (49,000 pellets/ha) density thinning or unthinned forest (106,000 pellets/ha). Similarly, there were no significant differences between treatments for mule deer (low: 495 pellet-groups/ha; medium: 500; high: 447; unthinned: 195) or moose (low: 190 pellet-groups/ha; medium: 88; high: 131; unthinned: 71). Naturally regenerated young lodgepole pine stands were studied at four sites. Stands were thinned, in 1988–1993, to target densities of 500 (low), 1,000 (medium) and 2,000 (high) stems/ha. Unthinned stands had >3,000 stems/ha. Mammal faecal pellets and pellet-groups were surveyed in 5-m2 plots (55–145 plots/stand). Plots were cleared of pellets in autumn 2003. New pellets and pellet-groups were counted in spring 2008.Study and other actions tested
A replicated, controlled study in 2007–2008 of a Douglas-fir Pseudotsuga menziessi forest in Oregon, USA (Manning et al. 2012) found that, 11–13 years after thinning, northern flying squirrels Glaucomys sabrinus were not more numerous in thinned than in unthinned stands. Flying squirrel density was lower in thinned (0.4 squirrels/ha) than unthinned (2.0/ha) stands. Among thinned stands, there were more flying squirrels in those that were lightly thinned with gaps (0.5/ha) than in heavily thinned stands (0.2/ha). The numbers in lightly thinned stands without gaps (0.4/ha) did not differ significantly from that in lightly thinned stands with gaps. Treatments were applied to 16 stands (15–53 ha), in four blocks (2.5–21 km apart), of 55–65-year-old forest, in 1994–1997. In each block, treatments were heavy thinning (to 125–137 trees/ha), light thinning (250–275 trees/ha), light thinning with gaps (as light thinning but also with 20% of the stand harvested leaving 0.2-ha gaps) and unthinned. Flying squirrels were surveyed using 100 traps/stand for four nights and three days, between late September and late November, in 2007 and 2008.Study and other actions tested