USGS - science for a changing world

Northern Prairie Wildlife Research Center

  Home About NPWRC Our Science Staff Employment Contacts Common Questions About the Site

Effects of Fire on Bird Populations in Mixed-grass Prairie

Results and Discussion

Killdeers (Charadrius vociferus) were most common at Woodworth immediately after a burn (Fig. 1). Densities and adjusted densities declined sharply with age of burn. Killdeer prefer open habitats with bare ground or very short vegetation, little residual vegetation, and no woody vegetation (Sample 1989, Best et al. 1995). These habitats include cultivated fields, heavily grazed pastures, bare shorelines of lakes and ponds, and exposed gravel and sand (Stewart 1975, Kantrud 1981). Open habitat with sparse cover is created immediately following a burn, which accounts for Killdeer favoring such plots at Woodworth. Huber and Steuter (1984) reported a noticeable absence of killdeers on unburned areas.

Marbled Godwits (Limosa fedoa) were not common at Woodworth, but highest densities were found in plots during the first two years after a burn (Fig. 2). Godwits often forage in wetlands but they nest and forage in upland areas, sometimes a considerable distance from water (Stewart 1975). They often feed in upland areas immediately after a burn. Preferred habitats of breeding godwits include idle grasslands and pastures with short vegetative cover (Ryan et al. 1984). They nest in cover of low to intermediate density and height (Kantrud and Higgins 1992). At Woodworth, the occasional high densities of Marbled Godwits were in plots within two years of a prescribed burn, where the lack of litter and low vegetation profile facilitated feeding. Ryan et al. (1984) recommended periodic disturbance by fire, grazing, or mowing, to produce the shorter-grass habitats favored by this species.

Upland Sandpipers (Bartramia longicauda) at Woodworth had highest densities and adjusted densities immediately after a burn and for about a year following a burn (Fig. 3). This appropriately named shorebird spends little time near water, nesting and foraging on dry land. It favors open grassland, with short vegetation and little woody cover (Skinner 1982, Sample 1989), and uses bare-ground habitats such as cropland (Stewart 1975). Although the Upland Sandpiper favors somewhat heavier cover for nest sites (Kantrud and Higgins 1992, Bowen and Kruse 1993), it uses open habitats for foraging (Kantrud 1981). Such selection for feeding areas is consistent with the higher-than-average use of recently burned plots at Woodworth. Huber and Steuter (1984) also reported that Upland Sandpipers used burned fields more often than unburned ones.

The Willet (Catoptrophorus semipalmatus) was uncommon on the Woodworth study plots; both densities and adjusted densities appeared unrelated to the age of burn (Fig. 4). Willets are most common in marshy areas and nearby uplands (Ehrlich et al. 1988). In North Dakota, Stewart (1975) indicated that they use mostly semipermanent and seasonal wetlands. Ryan and Renken (1987), however, found that, relative to the available area, Willets prefer less-permanent wetland types and alkali wetlands. Willets nest in the upland, often a considerable distance from water (Stewart 1975, Ryan and Renken 1987), typically in short, grassy cover (Higgins et al. 1979). Although Willets make little use of uplands except for nesting, they tend to favor short, native vegetation (Ryan and Renken 1987, Kantrud and Higgins 1992). With the Willet's affinity for wetlands and its large territory size, no selection of habitats at Woodworth was evident.

Densities of Eastern Kingbirds (Tyrannus tyrannus) at Woodworth were slightly lower than average for the first three years after a burn, although there was much variation around the loess line (Fig. 5). Adjusted densities were virtually unaffected by the recency of burn. Eastern Kingbirds use woodlands with open canopies and habitats with scattered stands of small trees or shrubs (Stewart 1975). Sample (1989) indicated that the species uses a variety of open habitats, with the nearby presence of tall, woody vegetation. At Woodworth, Eastern Kingbirds require only a small thicket of chokecherry or other woody species, from which they range out to forage into the grassland. Despite repeated fires, Woodworth still retains sufficient thickets for nesting eastern kingbirds. Should the small trees and shrubs be eliminated, use of the area by Eastern Kingbirds will diminish (Arnold and Higgins 1986).

Willow Flycatchers (Empidonax traillii) at Woodworth were seen only in thickets of chokecherry, hawthorn, or similar vegetation. The species occurred regularly only in the control plot, where they used a large thicket in almost every year. Adjusted densities for this species showed no pattern following a burn (Fig. 6). The Willow Flycatcher is a species of swamps and thickets, especially of willow (Salix spp.; Ehrlich et al. 1988, Sedgwick and Knopf 1992). Kahl et al. (1985) described the species' breeding habitat as having intermediate to tall ground vegetation; a low, open canopy; dense ground vegetation; at least a few woody stems ≥ 2.5 cm dbh; an intermediate to high number of smaller woody stems; with a litter layer of intermediate depth and intermediate to dense coverage. In North Dakota, Willow Flycatchers use natural prairie thickets, consisting of species such as chokecherry, hawthorn, and wild plum (Stewart 1975). Similar vegetation occurs at Woodworth; these habitats are reduced by repeated fires. Clearly the species would disappear once the habitat it requires was eliminated (Arnold and Higgins 1986).

Sedge Wrens (Cistothorus platensis) were common at Woodworth only in some years. Densities were mostly zero, but none of the positive values occurred within a year after a burn (Fig. 7). Adjusted values likewise demonstrated little response with time post-burn. In the study plots at Woodworth, the species was rarely found in emergent wetland vegetation during years with average water conditions, but was frequent in upland grasses during unusually wet years, such as 1995. The breeding habitat of Sedge Wrens consists of tall and dense vegetation (Skinner 1982, Schramm et al. 1986) found in emergent wetland vegetation (Stewart 1975), dry marshes or wet meadows (Bent 1968), moist grasslands, old fields, dense cultivated grainfields (Dobkin 1992), or retired cropland (Stewart 1975). Tall and dense vegetation and dense, prostrate residual vegetation appear to be important (Sample 1989). Such habitats are eliminated by a burn, suggesting why the species did not occupy recently burned plots at Woodworth. In contrast, for the tallgrass prairie of Illinois, Schramm et al. (1986) suggested Sedge Wrens prefer a clear understory and ground area, and indicated that the wrens were most attracted to recently burned areas, although they used, and perhaps required, unburned areas nearby to gather material for nest construction. In that study, the recently burned areas had taller and denser vegetation than did the unburned areas.

Yellow Warblers (Dendroica petechia) were most common at Woodworth in the control plot. Adjusted densities showed no pattern following a burn (Fig. 8). The Yellow Warbler favors thickets of small deciduous trees or tall shrubs (Stewart 1975, Dobkin 1992). Knopf and Sedgwick (1992) defined more precisely the habitat features associated with nest sites, primarily characteristics of the vegetation patch. At Woodworth, Yellow Warblers were most common in the control plot, especially in the large woody thicket. This species likely would disappear once the taller woody vegetation was eliminated (Arnold and Higgins 1986).

The Common Yellowthroat (Geothlypis trichas) had highest densities at Woodworth in the control plot. Adjusted densities indicated no trend, except for a modest depression immediately after a burn (Fig. 9). The yellowthroat usually nests in tall, dense herbaceous vegetation. Breeding habitat can be either in emergent wetland vegetation or in uplands with lush vegetation, possibly including shrubs or small trees (Stewart 1975, Kahl et al. 1985). Sample (1989) noted that the species prefers the presence of some standing residual cover and dense, prostrate residual vegetation. At Woodworth, wetland habitats were not usually affected markedly by fires, so yellowthroats used those habitats even immediately after a fire. Upland thickets sometimes were affected by the burns, and the species would avoid those areas until the vegetation had a chance to regrow. Huber and Steuter (1984) found yellowthroats more common, but not significantly so, in unburned than in burned grassland. Repeated fires that eliminate the brushy vegetation will reduce the suitability of the upland habitat for this species (Arnold and Higgins 1986).

Densities of Bobolinks at Woodworth were low immediately post-burn, peaked one to three years after a burn, and began to decline about five years post-burn (Fig. 10). Densities in the control plot remained high, however. Adjusted densities showed the same pattern. Breeding habitats of the bobolink include mixed-grass and tallgrass prairies, wet-meadow zones of wetlands, domestic haylands, retired croplands, and occasionally active croplands (Stewart 1975). Bobolinks prefer grasslands with a high coverage of fairly lush vegetation of intermediate height, with some residual vegetation (Wiens 1969, Sample 1989). The species favors areas with deep litter and a preponderance of grasses over legumes (Wiens 1969, Bollinger and Gavin 1992). Its preference for lush vegetation and deep litter is consistent with its reduced use of recently burned plots at Woodworth. Huber and Steuter (1984) also found that Bobolinks avoided burned grasslands. In contrast, Johnson and Temple (1986) found Bobolink nests most frequently in tallgrass prairie a single growing season after a burn, and less frequently in fields burned less recently.

Densities of Western Meadowlarks at Woodworth were low immediately post-burn, highest two to four years after a burn, and lower after about five years following a burn (Fig. 11). The species was frequently common in the control plot, however. The loess line fitted to adjusted densities indicated a depression immediately post-burn, and a slight peak about three years after a fire. Western Meadowlarks are most common in native grasslands and pastures, but also occur in hayfields, roadsides, retired cropland, and other open areas (Stewart 1975, Lanyon 1994). A preference by meadowlarks for habitats with grass and litter cover was identified by Wiens and Rotenberry (1981) and Sample (1989). Ground cover and litter seem especially important, as nests are often constructed with a dome of interwoven grasses (Lanyon 1994). Fire removes most litter, which suggests why meadowlark densities at Woodworth were somewhat depressed for about a year after a burn. Bock and Bock (1987) noted the adaptability of the species and observed that it was equally common on burned and unburned sites. Pylypec (1991) observed that Western Meadowlarks were less common on a burned area than on an unburned area for two years following a fire, but densities in the two areas were comparable in the third year.

Red-winged Blackbird densities at Woodworth varied only slightly with age of burn, although adjusted densities were high about two to four years post-burn (Fig. 12). Their overall numbers varied dramatically among years, as is demonstrated by the variation in densities for the control plot. Red-winged Blackbirds use a wide variety of habitats for breeding, including marshes, riparian areas, ditches, hayfields, weedy fields, and active and retired croplands (Stewart 1975, Sample 1989). Sample (1989) reported that Red-winged Blackbirds prefer lush habitats with fairly tall, dense vegetation and with standing and prostrate residual vegetation. At Woodworth, Red-winged Blackbirds are most commonly associated with wetlands, and their numbers fluctuated in close relation to water conditions. For that reason, redwing densities did not exhibit any marked response to the fire regime but were widely variable around the loess line. Huber and Steuter (1984) indicated that Red-winged Blackbirds used the burned and unburned treatments similarly during June, but by July use was reduced on the burned treatment. Eddleman (1974) stated that Red-winged Blackbirds increased when long-term protection from burning causes increased shrub and forb cover. Arnold and Higgins (1986) also noted higher densities of Red-winged Blackbirds in grassland areas with more shrubby vegetation. Although Red-winged Blackbirds used brushy areas at Woodworth, they appeared to rely on the wetland habitat.

Densities of Brown-headed Cowbirds (Molothrus ater) at Woodworth were depressed for the first year or so after a burn, but were constant after that time (Fig. 13). Cowbirds are habitat generalists, preferring habitats with low or scattered trees among grassland vegetation (Lowther 1993). The slight reduction in cowbird densities following prescribed burns at Woodworth may reflect the reduction of potential host nests in those plots. Huber and Steuter (1984) found no response by cowbirds to spring burns.

Savannah Sparrows (Passerculus sandwichensis) at Woodworth had highest densities one to five years post-burn, although they were occasionally common in the control plot (Fig. 14). The loess line fitted to adjusted densities indicated that the species was less common for about a year or so immediately after a burn, with possibly a longer-term depression as well. Favored breeding habitat of the Savannah Sparrow has been described in various ways: dense ground vegetation, especially grasses, and moist microhabitats (Wiens 1969); dry or wet habitats with little woody cover, moderate-to-high cover of fairly short vegetation, and medium litter depth (Sample 1989); and sparse ground cover and moderate above-ground cover (Skinner 1982). Wheelwright and Rising (1993) included grassy meadows, cultivated fields (especially alfalfa), lightly grazed pastures, roadsides, and sedge bogs as Savannah Sparrow habitats. In North Dakota, the species selects tallgrass prairie, lightly grazed mixed-grass prairie, wet-meadow zones bordering wetlands, hayfields, weedy fields, and retired croplands (Stewart 1975). Within plots at Woodworth, Savannah Sparrows were not regularly seen in areas with extensive shrub patches; Arnold and Higgins (1986) reported a similar response. At Woodworth, Savannah Sparrows avoided plots for a year or so after a burn, but no other responses to burns were evident. Halvorsen and Anderson (1983) also reported that Savannah Sparrow densities were reduced on burned fields, a finding they attributed to the reduction of residual cover. Huber and Steuter (1984) likewise noted that the species was absent from a field burned earlier in the spring. Pylypec (1991) indicated that Savannah Sparrows in a fescue prairie were reduced for three years following a burn.

Densities of Grasshopper Sparrows (Ammodramus savannarum) at Woodworth were depressed for about a year post-burn; they increased after that time and appeared to decline gradually after about five years post-burn (Fig. 15). Adjusted densities were more marked, with values depressed for two years post-burn, and no effect of burning thereafter. Over its broad breeding range, the Grasshopper Sparrow uses a variety of habitats, but prefers grassland, old-field, and retired cropland (Smith 1963, Stewart 1975, Kahl et al. 1985, Johnson and Schwartz 1993). Dense ground vegetation and shallow-to-moderate litter accumulation were identified as habitat features by Huber and Steuter (1984) and Kahl et al. (1985). Preference for patchy vegetation was noted by Skinner (1982) and Sample (1989). Whitmore (1981) observed that Grasshopper Sparrows favor bunchgrass habitats, with gaps in vegetation that facilitate movement. Smith (1963) indicated that Grasshopper Sparrows abandon fields once they become filled with shrubs. The requirement by the species for some litter accumulation is consistent with the reduced use of Woodworth plots for a year or so post-burn. Although Grasshopper Sparrows used the control plot regularly, they tended to concentrate in areas that lacked heavy shrub cover, similar to the finding of Arnold and Higgins (1986). In Montana, Bock and Bock (1987) observed fewer Grasshopper Sparrows in burned than unburned shrubsteppe habitats. Huber and Steuter (1984) reported Grasshopper Sparrows recolonizing a spring-burned field by mid-July of the same year. In tallgrass prairie in western Minnesota, Johnson and Temple (1986) found Grasshopper Sparrow nests more commonly in fields with four or more growing seasons since a burn. Eddleman (1974) indicated that burning will partially provide the interspersion of cover heights necessary for the species, but a severe lack of litter will result in few nesting sites and a lack of nesting material.

Baird's sparrows (Ammodramus bairdii) occurred infrequently at Woodworth. Densities tended to be highest in plots two to five years after a burn, although birds occasionally used the control plot (Fig. 16). The adjusted densities indicated a slight depression for the first couple of years post-burn, followed by a modest increase. The low observed densities caution against drawing any firm conclusions, however. The Baird's Sparrow is a mixed-grass prairie species, endemic to the northern Great Plains. In North Dakota it favors idle or lightly grazed areas, although it also uses more heavily grazed sites and, particularly in more arid locations, lowland areas (Stewart 1975). Baird's Sparrows favor abundant residual vegetation (Salt and Salt 1976) and a dense understory (De Smet and Miller 1989), which is consistent with its occurrence at Woodworth in plots several years after a burn. The species avoids areas with extensive bare ground (Davis and Duncan 1996). It has been reported to favor areas with some but not extensive coverage of shrubs (Arnold and Higgins 1986; Winter 1994; Dale et al., in press). Anstey et al. (1995) detected a positive association between Baird's Sparrow densities and litter depth. Winter (1994) suggested the species favored intermediate depths of litter. At Woodworth, the species was absent in recently burned plots, which typically have little remaining litter. Pylypec (1991) observed that Baird's Sparrows avoided a recently burned area, returned to the site the second year after burn, and in the third year had densities comparable to an unburned area.

Densities of Clay-colored Sparrows (Spizella pallida) at Woodworth increased almost monotonically following a burn (Fig. 17). The occasional high densities in recently burned plots was due to the occupancy by Clay-Colored Sparrows of brush thickets that were not markedly affected by the fires. Adjusted densities showed a similar pattern. The enormous variability among those points reflects the high annual variation in abundance of the species. Although Clay-colored Sparrows breed in a variety of habitats, ranging from retired cropland and abandoned fields to forest edges (Stewart 1975, Knapton 1994), in North Dakota the prime habitats are low thickets of wolfberry and other shrubs (Stewart 1975). Those habitats are common at Woodworth, accounting for the abundance of the species. Thickets were especially prominent in the control plot, which had the highest densities of Clay-colored Sparrows. Densities tended to increase with the number of years since the most recent burn. Clay-colored Sparrows were common even in recently burned plots, however, particularly if the burn was incomplete and did little damage to thickets. Huber and Steuter (1984) noted that the species favored the dense grassland that resulted from light grazing and the absence of fire. Halvorsen and Anderson (1983) observed declines exceeding 90% for Clay-colored Sparrows immediately following a burn, a change they attributed to the reduction of residual cover. In Saskatchewan, Pylypec (1991) noted that a burned area supported Clay-colored Sparrows at only one-third the density of an unburned area, for the first three years after the burn. Long-term idling of grassland habitat will permit the encroachment of shrubby and other woody vegetation, which favors this species (Eddleman 1974, Arnold and Higgins 1986), as was observed in the control plot.

Previous Section -- Analytic Methods
Return to Contents
Next Section -- Conclusions

Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo logo U.S. Department of the Interior | U.S. Geological Survey
Page Contact Information: Webmaster
Page Last Modified: Saturday, 02-Feb-2013 04:46:27 EST
Sioux Falls, SD [sdww55]