Conservation Evidence strives to be as useful to conservationists as possible. Please take our survey to help the team improve our resource.

Providing evidence to improve practice

Action: Use prescribed burning on pine forests Bird Conservation

Key messages

Read our guidance on Key messages before continuing

  • Two studies of the 28 captured (all from the USA) found higher bird species richness in sites with prescribed burning, tree thinning and mid- or understorey control or just burning and tree thinning, compared to control sites. Five studies found no differences in species richness or community composition between sites with prescribed burning; prescribed burning, tree thinning and mid- or understorey control; or prescribed burning and tree thinning only, compared to control sites, or those with other management.
  • Eight studies found that some species or guilds (such as open habitat species) were more abundant or more likely to be found in burned areas of pine forest than control areas. One study found that the responses of Henslow’s sparrows to burning varied considerably with geography and habitat.
  • Three studies found that some species were more abundant in thinned and burned stands, compared to controls or other management. Three studies found that overall bird densities or abundances of red-cockaded woodpeckers were higher in open pine forests with prescribed burning, tree thinning and mid- or understorey control, compared with control areas or those thinned but not burned. One found differences were more marked in spring. A study found that a red-cockaded woodpecker population increased following the start of intensive management consisting of prescribed burning and other interventions.
  • Ten studies found that total bird densities or those of some species was the same or lower in sites with prescribed burning compared to control sites, or those with other management. Five studies investigated several interventions at once. Generally, closed-forest species and ground nesters appeared to be adversely affected by burning.
  • Three studies found higher productivities or survival of species in burned or burned and thinned areas, compared to control areas or those burned less recently. Seven studies found no differences in productivity, behaviour or survival (including of artificial nests) in burned areas or burned and thinned areas, compared to controls. One study found that northern bobwhite chicks had lower foraging success in burned areas, compared to other management regimes, whilst another found that different predators were dominant under different management.
  • The three studies that investigated it found that burning season did not appear to affect the effects of burning.

 

Supporting evidence from individual studies

1 

A replicated before-and-after study in four National Forests in Texas, USA (Conner et al. 1995), found that red-cockaded woodpecker Picoides borealis populations increased at all four sites in the early 1990s after management for woodpeckers was intensified in 1989. This followed declines in the 1980s. Management included: prescribed burning and mechanical mid-storey vegetation removal and pine thinning; provision of 716 artificial cavities and the translocation of 19 woodpeckers. Status (i.e. active or inactive) of woodpecker cavity-tree clusters (groups of trees occupied by a breeding group of woodpeckers) was monitored (1983-1993) and by 1993, 98% of active clusters and 58% of inactive clusters had been burned at least once.

 

2 

A replicated study in 1991 in Ocala National Forest, an area of sand pine scrub in Florida, USA (Greenberg et al. 1995), found similar bird densities and species richness in areas that were burned, compared to areas that were clearcut and ‘brake-seeded’. This study is discussed in detail in ‘Clearcut and re-seed forests’.

 

3 

A replicated controlled study in 33 pine-grassland stands in Ouachita National Forest, Arkansas, USA (Wilson et al. 1995), found that overall bird species richness was similar across a series of different managements aimed at red-cockaded woodpecker conservation. Bird densities were highest in the second growing season after both burning and midstorey and tree thinning, compared with those subject to only midstorey and tree thinning or untreated stands. Ground-nesting species were most abundant in untreated stands. Management appeared beneficial to several species of conservation concern e.g. Bachman's sparrow Peucaea aestivalis (formerly Aimophila aestivalis), as well as red-cockaded woodpeckers.

 

4 

A study in mixed pine Pinus spp. forests in South Carolina, USA (Franzreb 1997) found that a population of red-cockaded woodpeckers increased from four individuals and one breeding pair in 1985 to 99 and 19 pairs in 1996 following intensive management, including prescribed burning. The authors emphasise that hardwood midstory control using prescribed burning as well as cutting, herbicide applications and thinning of trees mimicked the natural fire regime and was essential to the success of the project. In addition, artificial cavities were installed; southern flying squirrels Glaucomys volans (a cavity competitor) were controlled and 54 woodpeckers were translocated into areas with artificial or natural cavities.

 

5 

A replicated study in mid-December 1995 to mid-February 1996 at a mixed pine forest at Fort Benning Military Reservation, Georgia, USA (King et al. 1998), found no significant differences between wintering bird abundances or community composition in nine plots burned during the growing season (April-August ) of 1994, compared to nine plots burned during the dormant season (January-March). A total of 48 species were recorded (during point count censuses): 41 in growing-season burn plots and 41 in dormant-season burn plots (34 common to both). On average, 27 species were recorded in a growing-season burn plot and 25 in a dormant-season burn plot. Abundance of individual species was similar between plot types.

 

6 

A replicated study in 2004-2005 in open pine forests at Eglin Air Force Base, Florida, USA (Hill et al. 1998), found that eight of 21 plots managed intensively for red-cockaded woodpeckers contained Bachman's sparrows. Management was mainly prescribed burning but included some midstorey clearing. Some areas with woody midstory vegetation and lacking dense ground cover were unsuitable for breeding sparrows. A burn rotation of 3-5 years (burning early in the growing season) appeared best to encourage the dense grassy understory and sparse midstory preferred by Bachman's sparrows for nesting, whilst not decreasing habitat suitability for red-cockaded woodpeckers.

 

7 

A replicated study in 1995-1996 in pine savanna in South Carolina, USA (Krementz & Christie 1999), found that there were fewer scrub-successional species in stands managed for red-cockaded woodpeckers (including prescribed burning) than in stands which were clearcut to remove non-native pines and replanted with longleaf pines Pinus palustris. This study is discussed in detail in ‘Clearcut and re-seed forests’.

 

8 

A replicated, controlled study in 1993-1995 in loblolly pine Pinus taeda-dominated forest in Piedmont National Wildlife Refuge, Georgia, USA (White et al. 1999), found that red-cockaded woodpeckers were found in 18 rotationally burned plots (each >100 ha) but not in six unburned plots. Average species richness was similar for burned and unburned plots (42 vs. 41 species) and all other species were found in both plot types. Of 29 species that showed significant differences in abundance between burned and unburned areas, 22 were more abundant in burned plots and seven were more abundant in unburned plots. During 1994-1995, 224 nests of 20 species were found in burned plots; only nine nests (six species) were found in unburned plots.

 

9 

At Piedmont National Wildlife Refuge, Georgia, USA, a four-year replicated controlled study found no evidence that winter burning and tree thinning (management primarily to improve red-cockaded woodpecker habitat) negatively affected wood thrushes Hylocichla mustelina in loblolly pine-dominated forest (Powell et al. 2000). Thrush density and adult or juvenile survival during the breeding season did not vary between compartments subject to thinning and prescribed burning and unmanaged ones. Overall, thrush numbers increased on treatment compartments (0.91-0.97 birds present before vs. 0.98-1.05 after management), and declined slightly in controls (0.98-1.05 vs. 0.87-0.92). Plots were surveyed eight times in 1993 and sixteen times a year in 1995-1996.

 

10 

A replicated study in 1996-1998 in a longleaf pine forest in South Carolina, USA (White & Seginak 2000), found that great crested flycatchers Myiarchus crinitus had larger clutches and slightly higher productivity in nest boxes in plots burned during the warm season (April-June: 4.9 eggs/clutch and 2.7 fledglings/clutch for 24 clutches) compared to those burned during the cool season (December-March: 4.5 eggs/clutch and 2.5 fledglings/clutch for 14 clutches). There were no differences in overall occupancy rates or hatching rates (21% occupancy and 61% hatching success for 210 boxes in plots burned in the warm season vs. 19% and 61% for 120 boxes in those burned in the cool season). This study is also discussed in ‘Provide artificial nest sites’

 

11 

A replicated, paired, controlled study in 1995-1997 in two pine Pinus spp. habitats at Angelina National Forest, Texas, USA (Conner et al. 2002), found that bird species richness and abundances in spring were significantly higher in plots managed for red-cockaded woodpecker, compared to unmanaged plots (longleaf pine forests: 7 species and 22 individuals in managed forests vs. 6 and 13 in controls; loblolly pine-shortleaf pine P. echinata forests: 10 species and 32 individuals vs. 7 and 20). Differences were also present in loblolly-shortleaf, but not longleaf pine forests during winter (loblolly-shortleaf forests: 8 species and 42 individuals vs. 5 and 28; longleaf: 7 and 30 vs. 7 and 26). Management consisted of prescribed burning, mechanical mid-storey vegetation removal and thinning of pine trees.

 

12 

A replicated study in 1999-2000 across 40 shortleaf pine-hardwood stands in Ouachita National Forest, Arkansas, USA (Cram et al. 2002), found that northern bobwhite Colinus virginianus abundances were highest in stands thinned and burned three years previously (1.5 males heard/count), but were not significantly higher than in stands which were only thinned (1.1) and those thinned and burned two years previously (0.8). Control (unmanaged) stands and those burned a year previously had significantly lower abundances than those burned three years previously or thinned but not burned (control stands: 0.1 calls/count; 0.4 for stands one year after burning). The stands with the highest abundances also had greater understory shrub cover. The management was aimed at restoring pine-grassland habitats.

 

13 

A controlled study in 1998-2001 on Mt. Trumbull, Arizona, USA (Germaine & Germaine 2002), found that western bluebirds Sialia mexicana were more likely to successfully fledge young and fledged more chicks in restored ponderosa pine Pinus ponderosa forest, compare to unrestored forest (75% of 56 nests in restored forest fledging young, with an average of 3 chicks/nest vs. 39% and 2 chicks in control stands). Clutch size and number of nestlings per nest were similar between treatments but an average of 91% of nests were infested with parasitic blowfly Protocalliphora spp. larvae in treated forest compared with 46% in unmanaged forest; any effects on post-fledging survival are unknown. Restoration treatments comprised tree thinning, slash manipulation (e.g. chopping) and burning.

 

14 

A replicated, randomised and controlled study in May-July 2000 in 28 longleaf pine forest plots in Georgia, USA (Jones et al. 2002) found that survival of 770 artificial nests was similar between burned (42%) and unburned plots (41%). Bird predation was greater in burned (14%) than unburned (10%) plots. Small mammal predation was greater in unburned (31%) than in burned (15%) plots. There was no interaction between the supplementary feeding of predators, burning and nest predation.

 

15 

At Piedmont National Wildlife Refuge, Georgia, USA (Lang et al. 2002), a controlled before-and-after study found that wood thrush habitat use ands movements in loblolly pine stands were very similar in stands managed for red-cockaded woodpeckers and control stands. Management consisted of thinning forests and prescribed burning, mostly on small scales, in stands up to 50 ha. Juvenile and adult thrushes were monitored by radio-tracking in two breeding seasons before management (1993-1994) and two after (1995-1996) on an experimental compartment, and for four years on a control (1993-1996).

 

16 

A replicated study in 1999-2000 in longleaf pine forest at Conecuh National Forest Alabama, and Blackwater River State Forest, Florida, USA (Grand et al. 2004), found that Bachman's sparrow density was greater in forest patches during the first three years after burning (compared to four or more years). Density did not differ between stands burned during the growing season (April-September) or dormant season (October-March). More sparrows occurred in areas of denser, taller grass, regardless of burn season. As tree canopy cover increased, grass cover decreased. Fire suppression or burning at intervals of more than 4-5 years, resulted in a greatly reduced grass/herbaceous understorey.

 

17 

A site comparison study in the winters of 2001-2002 at three pine savanna sites in southeast Louisiana, USA (Bechtoldt & Stouffer 2005), found that wintering Henslow's sparrow Ammodramus henslowii abundance was significantly higher in areas burned the preceding growing season (average of 3 birds/ha) than in areas burned two or three years previously (1 bird/ha). Eight areas were burned once in May-August 1999-2001. There was a trend for decreasing sparrow abundance with time since burn: 1.0/ha in areas burned one year previously, around 0.7/ha in areas burned two years previously and 0.2/ha in those burned three years previously. Of vegetation characteristics measured, average seed abundance (which decreased with time since burn) was the best predictor of sparrow abundance.

 

18 

A replicated, controlled study in 1994-1997 in open-longleaf pine and pocosin woodlands at Fort Bragg, North Carolina, USA (Allen et al. 2006), found that species associated with open longleaf habitats (e.g. red-cockaded woodpecker and Bachman's sparrow) were most common in burned areas of forest. Fire-suppression-associated species (e.g. wood thrush and ovenbird Seiurus aurocapilla) were confined to denser vegetation around pocosins (woodland with a dense understorey around stream-heads) in burned areas, but were abundant in fire-suppressed areas with a dense understorey. Overall bird abundance and diversity was greater closer to the pocosins.

 

19 

A replicated study in 2000 in pine and mixed forests in eight ‘Sky Island’ mountain ranges in Arizona, USA (Kirkpatrick et al. 2006), found that the distributions of 11 of 65 species of birds were affected by burning: 73% of species were positively associated with burned areas and showed stronger associations with severe rather than less severe burns.  Strong positive associations with severe fires were apparent for western wood-pewee Contopus sordidulus and house wren Troglodytes aedon, but negative associations were found for warbling vireo Vireo gilvus and red-breasted nuthatch Sitta canadensis.

 

20 

A replicated experiment in 1999-2001 in 13 longleaf pine forest plots within Conecuh National Forest, Alabama, USA (Tucker et al. 2006), found that a significantly lower proportion of Bachman's sparrow territories in compartments burned four years previously (20% of 20) successfully produced young than those in compartments burned less than three years previously (52% of 50). In addition, a higher proportion of male sparrows remained unpaired in compartments burned four years previously compared to those in more recently burned plots (50% of 20 males in plots burned four years before remain unpaired vs. 28% of 50 in more recently burned areas). There was no significant difference between growing or dormant season burn plots. Daily survival rates were similar in compartments burned 4-years previously (93%) and those burned more recently (94%), and those burned during the growing or dormant season (89% and 95% respectively).

 

21 

A controlled cross-over study over four breeding seasons (2003-2006) in old-growth longleaf pine forest at Wade Tract, Georgia, USA (Cox & Jones 2007), found no significant effect of breeding season prescribed burns on the territory size or site fidelity of male Bachman's sparrows. Approximately half (40 ha) of the study area was burned each year, with the unburned half burned the following year. On average, sparrow territories that were burned (1.9 ha) were similar in size from those that were not (2.1 ha). The proportion of ringed males observed after a prescribed burn was similar in burned (73%) and unburned (77%) areas.

 

22 

A replicated, randomised, controlled study in 1998-2005 in 12 ponderosa pine stands (15-20 ha) in the North Cascade Range, Washington, USA (Gaines et al. 2007), found that there was a trend towards higher bird density in restored stands, compared to controls (13 birds/ha in eight restored stands vs. 10 in four controls). Management consisted of thinning and understorey burning. Thinning took place in 1998-1999, with burns in spring 2000 and 2004. Breeding birds were censused in 2001 and 2005. White-headed woodpecker Picoides albolarvatus and western bluebird Sialia mexicana, Cassin’s finch Carpodacus casinii and yellow-rumped warbler Dendroica coronata had higher densities in treated stands. Mountain chickadee Poecile gambeli, western tanager Piranga ludoviciana and red-breasted nuthatch Sitta canadensis were more common in control stands.

 

23 

A controlled study within a loblolly pine plantation in Louisiana, USA, in 2003-2005 (Burke et al. 2008) found that northern bobwhite chicks were 50% less likely to successfully capture arthropods in burned areas of forest than in mown areas, areas burned and treated with imazapyr herbicide, or control areas.

 

24 

A controlled before-and-after study in 2000-2006 at three ponderosa pine forest sites in northern Arizona, USA (Hurteau et al. 2008), found no differences in species richness or evenness between blocks (16-30 ha) which were thinned, those which were burned, those both thinned and burned and control blocks. Western bluebird, and pygmy nuthatch Sitta pygmea were significantly more abundant after both burning and thinning (with bluebirds also being more abundant in burned-only sites), compared to controls.  Mountain chickadees Poecile gambeli were less abundant under burning and burning-and-thinning, than in controls, whilst yellow-rumped warblers Dendroica coronata were less abundant after thinning-and-burning, compared to control sites and dark-eyed juncos Junco hyemalis did show a response to treatments. Thinning was undertaken in autumn 2002 and burns in autumn 2003. Birds were surveyed (point sampling) in May-July 2000-2002 and 2003-2006.

 

25 

A replicated controlled trial in 2001-2003 in pine-dominated forests in Klamath National Forest, California, USA (Sperry et al. 2008), found that the average number of dark-eyed juncos, the number of territories, nest survival and nestling weights and sizes were all similar between five 40 ha plots which were burned in autumn 2001 and unburned plots (8-17 birds, 3-6 territories and 95-97% survival for burned plots vs. 8-18 birds, 3-5 territories and 95-96%). Juncos were surveyed in 2002 and 2003. All ten plots had small trees removed in 1998-2000.

 

26 

A replicated controlled trial in November-January 2005-2008 in longleaf pine plantations in Pebble Hill and Arcadia, Georgia, USA (Cox & Jones 2009), found that there were more Bachman's sparrows in burned plots (10-60 ha), compared to unburned plots (0.6 birds/count vs. 0.3). Approximately 50% of habitat was burned each May with individual blocks burned every second year. Sparrow numbers were positively correlated with percentage bare ground cover, and negatively so with numbers of low woody shrubs, grass cover and grass height.

 

27 

A paired, controlled study in the winters of 2004-2006 in ponderosa pine woodlands in Coconino and Kaibab National Forests, Arizona, USA (Pope et al. 2006), found that hairy woodpecker Picoides villosus densities were higher in burned plots than controls (11 birds/100 ha in burned plots vs. 2 in controls). Pygmy nuthatch (45 in burn units; 40 in controls) and white-breasted nuthatch Sitta carolinensis (10 and 12) densities were similar between treatments. Activity of bark beetles (potential bird food) was greater in burn units (10% of trees having signs of beetles vs. 5% of controls). Burns took place in autumn 2003 (Coconino) and autumn 2003 and spring 2004 (Kaibab).

 

28 

A replicated study in the winters of 2005-2007 in 19 longleaf pine and flatwoods pitcher plant savannas in Louisiana, USA (Palasz et al. 2010), found that the response of over-wintering Henslow's sparrows to burning varied regionally: abundance in burn plots increased over the first three years after burning in western plots (from approximately 1 bird/ha to 2 birds/ha), but decreased in the east (from 2 to 1  birds/ha). Different habitats may explain the regional responses, as seven of ten eastern plots were in wet flatwoods pitcher plant bogs, five of six western plots were in (drier) longleaf savanna. Throughout Louisiana 19, 2.25 ha plots were established, with 18 burned every 2-3 years. Sparrows were surveyed by flush netting in November-April. Habitat generally became unsuitable for sparrows by about 5-years post-burn due to woody plant encroachment.

 

Referenced papers

Please cite as:

Williams, D.R., Child, M.F., Dicks, L.V., Ockendon, N., Pople, R.G., Showler, D.A., Walsh, J.C., zu Ermgassen, E.K.H.J. & Sutherland, W.J. (2018) Bird Conservation. Pages 95-244 in: W.J. Sutherland, L.V. Dicks, N. Ockendon, S.O. Petrovan & R.K. Smith (eds) What Works in Conservation 2018. Open Book Publishers, Cambridge, UK.