skip banner Southern Forest Resource Assessment    Final Report: Technical

Search this site:


Home > Final Report > TERRA-3   

Previous PageNext Page

Effects of Urbanization on Forest Wildlife

Effects of urbanization on forest bird communities

A number of studies investigated changes to bird communities by comparing an urbanized site versus a less urbanized (or more forested) site. Many investigators found that urbanization decreased the species diversity of the avian community and increased avian density (or bird biomass), favoring dominance by a few species. Bird species vary in sensitivity to urbanization, leading to loss of sensitive species and a shift in the species composition of urban versus forest bird communities. Habitat specialists, including many forest insectivores, neotropical migrants, and forest interior species, have been documented to be less tolerant of urbanization. Beissinger and Osborne (1982), Smith and Schaefer (1992), Franklin and Wilkinson (1996), Kluza and others (2000), Croonquist and Brooks (1993), and Dowd (1992) all documented shifts in avian species composition with increasing urbanization.

Some investigators studied the response of bird communities across several sites or along a gradient of increasing urbanization. Gradient studies revealed a less clear pattern in bird species diversity and density peaks; in some cases the pattern shifted seasonally. However, shifts in the avian species composition were generally found as urbanization increased (Blair 1996, Clergeau and others 1998, Lancaster and Rees 1979, Rottenborn 1999).

Others investigated changes in the bird community at a single site through time as the area became urbanized or more forested. Butcher and others (1981), Askins and Philbrick (1987), Aldrich and Coffin (1980), Long and Long (1992), and Horn (1985) documented the loss of sensitive forest bird species after urbanization or their return after reforestation.

Table 3.3 lists selected forest bird species in the Southeastern United States and their tolerances to urban and suburban development.

Urban fragmentation and edge effects

Forest size and level of fragmentation and the effects on breeding birds—Increasing urbanization fragments forest habitat into smaller and more isolated tracts. Research on breeding forest birds has shown that some species have minimum area requirements. Many studies documented declines in the numbers of forest breeding migratory birds in small isolated forest patches (Danielson and others 1997). Fragmentation is considered to be a primary contributing factor to observed neotropical migrant declines.

Whitcomb and others (1981) found that many neotropical migrant species became increasingly rare as the size of the forest decreased. In addition, area sensitivities varied depending on the degree of isolation from larger forest tracts. They concluded that forest tracts needed to contain hundreds or perhaps thousands of acres to conserve populations of some forest bird species. Robbins and others (1989) suggested that when managing forests for wildlife, top priority should go toward providing for the needs of area-sensitive or rare bird species. When conservation of large contiguous forest tracts is not possible, they suggested that several moderately sized contiguous forests could be helpful in maintaining rare forest breeding birds.

Reduced reproductive success of forest nesting birds in small or fragmented forests may be due to increased nest predation or nest parasitism by brown-headed cowbirds. Nest parasitism is associated with brown-headed cowbirds, which lay their eggs in the nests of other species. These hosts then raise cowbirds at the expense of their own offspring. Nest predation can be caused by a combination of many avian, mammalian, and reptile species. Rates of nest predation have been found to be higher in small forest tracts than in large forest tracts, and small urban forest tracts experience higher rates of predation than comparably sized forest tracts in isolated rural areas (Wilcove 1985). Migratory songbird populations suffer the most serious effects from increased predation in small forest tracts. Keyser and others (1998), Donovan and others (1995), Robinson (1992), and Robinson and others (1995) all documented reduced reproductive success of neotropical migrants and other forest nesting bird species in fragmented forests due to higher rates of nest predation and/or nest parasitism.

Recently, investigators stress the importance of overall forest cover or landscape levels of fragmentation surrounding a local area when evaluating the presence or nesting success of area-sensitive or forest-interior birds. As indicated by Villard (1998), preference for forest-interior habitat or avoidance of small fragments tends to focus attention on the local scale, whereas processes underlying these phenomena may take place over landscape or even continental scales. Therefore, forest-interior preference and area sensitivity should be considered in a landscape context. In one study, forest cover in approximately 40-square-mile study plots was found to be the most important factor affecting the distribution of forest birds (Trzcinski and others 1999). Comparatively, the independent measures of forest fragmentation produced effects that were inconsistent and far less important than overall forest cover. In addition, the reduction in nesting success of forest birds due to nest predation and parasitism was much greater in heavily fragmented landscapes with low forest cover than in heavily forested landscapes (Hartley and Hunter 1998, Robinson and others 1995). Similarly, no differences were detected in the breeding success of worm-eating warblers in small and large forest tracts when high amounts of forest canopy cover were present in the surrounding landscape (Gale and others 1997).

In addition, landscape-level factors may partially affect the distribution of mammalian nest predators and, potentially, songbird nest-predation rates. A combination of local features, such as proximity to some types of edge, as well as broader landscape-level features, such as land use patterns, was determined to influence the abundance of these mammals (Dijak and Thompson 2000). At a broader scale, raccoons were more abundant in agricultural landscapes with high densities of streams than in forested landscapes with low densities of streams. Opossums were more abundant in heterogeneous landscapes with widely spaced patches of forest and high densities of riparian habitat.

A review of Breeding Bird Survey trends for the southern Piedmont physiographic area might lead one to conclude that perhaps urbanization is not a serious threat to sensitive forest breeding birds. As indicated in Hunter and others (2001a), very few vulnerable species in the southern Piedmont have declined overall from 1966 to 1996. This apparent stability, however, may reflect an overall increase in forest acreage and maturation of the forests during this period. As further summarized in Hunter and others (2001a) wood thrushes and red-eyed vireos have shown consistent declines within patches of mature forest in Piedmont suburban areas, such as Atlanta, GA. In addition, a number of area-sensitive woodland bird species, such as northern parulas, black-throated green warblers, Swainson’s warblers, and worm-eating warblers, have population centers in relatively more forested areas, such as the southern Blue Ridge and the South Atlantic Coastal Plain, but are nearly absent as a breeding species over much of the southern Piedmont (Hunter and others 2001a). Perhaps more revealing than population trend data alone for woodland warblers and other sensitive mature forest species is the absolute abundances for those species as derived from the Breeding Bird Survey data (Hunter, W.C., May 2002. Unpublished analysis on Breeding Bird Survey data. 4 p. On file with: Kenneth L. Graham, U.S. Fish and Wildlife Service, Ecological Services, Suite 200, 1875 Century Blvd., Atlanta, GA 30345). Absolute abundances of these species in heavily fragmented physiographic areas, such as the southern Piedmont and the southern ridge and valley/southern Cumberland Plateau, are clearly much lower than those exhibited by more heavily forested, less fragmented physiographic areas, such as the southern Blue Ridge and northern Cumberland Plateau. In the face of very low absolute abundances of sensitive woodland bird species, positive or negative population trends within heavily urbanizing areas, such as the southern Piedmont, may reflect habitat conditions and population trends in nearby physiographic areas that actually support those species’ population centers and act as source populations. Ironically, some of the most forested physiographic areas in the Southeast have exhibited the steepest declines in forest birds in recent years. These areas have long been considered to be population sources for forest nesting birds (and still are, but to a more limited extent than previously thought) (Simons and others 2000). See chapter 4 for more information concerning population declines of forest birds in more forested physiographic regions and for trends in wood-warbler species in the Piedmont.

Connective corridors and offsetting the deleterious effects of fragmentation

The presence of connective corridors may help to reduce the isolation of wildlife populations in fragmented forests (MacClintock and others 1977, Machtans and others 1996, Wegner and Merriam 1979). Corridors may provide a connection that allows wildlife to move from one patch to another across an intervening, inhospitable landscape. This phenomenon has been especially well documented for disturbance-dependent grassland and scrub-shrub bird species, such as Bachman’s sparrow in largely forested areas (Dunning and others 1995). It is not obvious that animals possessing the mobility of birds need corridors to cross-fragmented landscapes, but it appears that the open space between forest islands is a barrier to movement of some songbirds (Whitcomb and others 1981). Gaps of 250 feet or more produced isolation characteristics for some songbirds in small forest fragments created by power lines and roads (Robbins and others 1989). Such gaps may not represent as serious a problem in largely forested landscapes, however (Gale and others 1997). Some investigators question the conservation value of corridors or question whether sufficient experimental evidence exists to draw conclusions on their benefits (Inglis and Underwood 1992, Simberloff and others 1992). Several potential negative effects and disadvantages of corridors should be considered prior to their use in overcoming fragmentation (Simberloff and others 1992). Disagreement over the value of corridors to overcome the effects of fragmentation for various species is likely to continue for some time. The use of corridors and the effect of fragmentation on movement patterns seem to be highly species-specific (Debinski and Holt 2000).

Fragmented forests have a greater proportion of edge habitats. Edges have generally been regarded by wildlife managers to have a positive effect on wildlife because the number of species increases near habitat edges (Yahner 1988). This positive effect likely remains true for birds in predominantly forested landscapes. In fragmented landscapes, however, maximizing species diversity is not always a desirable objective in light of the number of rare species that depend on large areas of habitat. Rates of nest predation and brood parasitism are greater at edges for some forest nesting birds (Gates and Gysel 1978), especially as overall forest cover becomes increasingly fragmented (Donovan and others 1997). Paton (1994) reviewed a number of studies that dealt with bird nesting success as a function of distance from an edge. Most studies found that nesting success decreased near edges as a result of increasing nest predation and parasitism rates. The strongest effects appeared to occur within about 125 feet of the edge. Indigo bunting nests along abrupt forest edges, such as agricultural edges, wildlife openings or campgrounds, had nearly twice the nest predation rate as those found along more gradual edges, such as those created by treefalls, streamsides, and gaps created by selective logging (Suarez and others 1997).

While the results of many investigations indicate that nesting success for forest birds is reduced by the proximity of edges, recent information indicates that such effects depend on the nature of the surrounding landscape. Hartley and Hunter (1998) reviewed various nest predation studies and concluded that nest predation rates decreased as the amount of overall forest cover increased. Edge effects were more apparent in largely deforested landscapes. Donovan and others (1997) found that nest predation rates were significantly higher near edges, but these increased rates were apparent only in highly and moderately fragmented landscapes and not in unfragmented landscapes. The ovenbird may be an exception, however. Even in an extensively forested landscape, slightly reduced rates of breeding success were documented for ovenbirds near forest edges (King and others 1996). Still, ovenbird reproductive success remains high overall, and other sensitive neotropical migrants fare better in highly forested landscapes (Gale and others 1997). Ovenbirds reproduce well in midsuccessional forests, and since such conditions are plentiful throughout eastern forests, the ovenbird is not considered a conservation priority species. See chapter 1 for more information about the effects of forest fragmentation on forest wildlife.

Not all investigators agree that higher nest predation rates occur in smaller forests or along forest edges (Friesen and others 1999, Haskell 1995, Matessi and Bogliani 1999, Yahner 1996, Yahner and Mahan 1996). Studies in large contiguous forest areas, such as the Great Smoky Mountains National Park, indicate that although these areas enjoy an overall higher nesting success rate for forest nesting birds (such as wood thrush), they may also support a more diverse and abundant predator community than more disturbed or less contiguous sites (Simons and others 2000). In addition, the magnitude and patterns of nest parasitism by brown-headed cowbirds is not consistent among studies (Coker and Capen 1995, Donovan and others 1997, Evans and Gates 1997, Gates and Gysel 1978, Hahn and Hatfield 1995, Robinson 1992, Robinson and others 1995).

Effects of urban environments on bird abundance and nesting success

In urban areas, forest-breeding birds may have lower abundances and lower nesting success. A 10-acre woodlot without any nearby houses had greater species richness and higher abundances of neotropical migrant species than did a 60-acre urbanized woodlot, indicating that the diversity and abundance of neotropical migrant birds decreased with increased urban development (Friesen and others 1995). Golden-cheeked warblers declined near urban development, apparently due to the increased presence of blue jays and greater nest predation (Engles and Sexton 1994). Declines of neotropical migrants were documented over a 50-year period in the North Carolina Highlands Plateau, likely due, in part, to the close proximity of residential development and urban fragmentation (Holt 2000). Nest predation rates were found to be greater for woodlands in the vicinity of human settlement (Matessi and Bogliani 1999). Mammalian nest predators were found to be more abundant in floodplain forests that adjoined residential and agricultural lands (Cubbedge and Nilon 1993).

Urban areas as habitat for birds

Urban woodlands are unsuitable habitat for many forest bird species, including many neotropical migrant birds, birds that require large habitat areas for breeding, birds that breed only in forest interior habitats, many scrub-shrub and grassland species, and those sensitive to urban disturbance. Urban and suburban preserves tend to be small and isolated from other forests. However, urban woodlands still provide habitat for some wildlife species and seasonally support migrating birds. Not all urban habitats are the same.

Woody vegetation volume is important in determining breeding bird diversity in urban settings (Goldstein and others 1986). Urban woodlots of 20 acres or more can support dense and diverse populations of breeding birds, provided that they have adequate shrub understory, mature and dead standing trees, and vegetative edge types of sufficient width and proper quality (Linehan and others 1967). Large urban parks with well-preserved natural forest habitat support bird populations more characteristic of native forests (Gavareski 1976). Urban parks, cemeteries, schoolyards, and other open spaces are prime sites for wildlife management (Bolen and Robinson 1995). For example, Washington, DC, has only house sparrows, pigeons (rock doves), and starlings in the downtown area, but nearby in the spring gardens surrounding the White House, 19 species are present.

In urban environments, the objective of wildlife management should be to maintain biological diversity by retaining sufficient habitat for the maximum number of wildlife species (Milligan and others 1995). Urban wildlife habitat designs must consider the size, composition, connectivity, dynamics of the habitat patches, and human perceptions of the habitat areas. At the same time, however, urban wildlife habitats must be at a scale compatible with the surrounding urban uses. Constraints are necessary to promote human health and safety, and to meet habitat requirements of the different wildlife species.

Urban habitats pose additional risks to resident avifauna. An estimated 98 million birds are killed each year in the United States from window collisions with high-rise buildings (Bolen and Robinson 1995). In addition, an estimated 2 to 4 million birds are killed each year in the Eastern United States due to collisions with communication towers (Weisensel 2000). The relative contributions of these mortality sources to the declines of any conservation priority bird species were not described in these references.

Effects of urbanization on birds of prey and scavengers

Birds of prey, such as hawks, eagles, and owls, can be vulnerable to the effects of urbanization because they are at the tops of food chains, and their home ranges are larger than those of most other birds (Adams 1994). Hawk species differ in their requirements for nesting habitat and tolerance for forest openings and human disturbance. Cooper’s hawks abandon nest sites when housing construction and residential disturbance encroach on established nest sites (Bosakowski and others 1993). There is evidence, however, of adaptability of various hawk species to urban settings. Broad-winged hawks are more tolerant of forest openings when selecting nest sites than red-shouldered, red-tailed, or Cooper’s hawks (Titus and Mosher 1981). Red-shouldered hawks in New York and New Jersey have higher nest productivity with increasing distance from human habitation (Speiser and Bosakowski 1995).

Bald eagles generally select well forested areas near water bodies and avoid areas of human development and areas of high boat and pedestrian traffic (Buehler and others 1991a, 1991b; Chandler and others 1995). On the lower Melton Hill Reservoir and the adjoining Clinch River in eastern Tennessee, residential and industrial development was found to be the primary factor limiting habitat suitability for eagle nesting (Buehler 1995).

When not searching for food, black and turkey vultures tend to prefer forested habitats free of buildings for roosting and nest sites (Coleman and Fraser 1989). Nests are frequently located away from human disturbance in rock crevices and in roadless, forested, and undeveloped areas. Nesting success for vultures was found to increase farther from buildings due to lower disturbance and less depredation by dogs.

Although some raptors are sensitive to urban disturbance, there may be differences among individuals, species, and regions of the country. Raptors that are tolerant of urban environments include Mississippi kites, sharp-shinned hawks, Cooper’s hawks, red-shouldered hawks, and red-tailed hawks (Adams 1994). Urban woodlands, even those composed primarily of exotic vegetation, lawns, and urban development, are acceptable to some red-shouldered hawks (Bloom and others 1993). One pair of red-shouldered hawks successfully fledged young within 65 feet of people engaged in jogging, picnics, and baseball games. American kestrels also have adapted to urban environments where suitable nesting cavities are available (Adams 1994).

The screech owl thrives in some suburban environments, especially those with large wooded lots (Gehlbach 1986). Burrowing owls, barn owls, and, occasionally, great horned owls have also been found in metropolitan environments (Adams 1994). Burrowing owls benefit from light levels of urban development and reach their highest densities in areas 55 to 65 percent developed. Other population-limiting factors are encountered beyond that development level, however.

Effects of urbanization on mammals

In general, urban environments support fewer species of mammals than surrounding rural areas (Adams 1994). The species that occur in urbanized environments tend to be habitat generalists rather than specialists. Urbanized areas can support high populations of exotic species, such as the house mouse and Norway rat. In less urbanized areas where large green spaces remain, more species are likely to be encountered. Downtown Boston cemeteries support 20 species of resident mammals (Bolen and Robinson 1995).

Small and medium-sized mammals, especially granivores, are the most abundant mammals found in urban and suburban environments (Adams 1994). In one study, mammals found in urban greenspaces were primarily habitat generalists that utilize a mosaic of habitat types (VanDruff and Rowse 1986). Deer mice, meadow voles, tree squirrels, ground squirrels, chipmunks, and woodchucks are common residents of urban areas (Adams 1994). Some small mammals, however, are habitat specialists that do not easily adjust to changes brought about by urbanization. Fragmentation of habitat in the Great Dismal Swamp of Virginia and North Carolina by residential subdivisions and industrial parks may be contributing to the decline of five indigenous subspecies of mammals (Rose 1991). The Allegheny woodrat is restricted to only a few habitats and is listed as threatened in Pennsylvania because of statewide declines (Balcom and Yahner 1996). Increases in residential and agricultural development were observed near sites of extirpation. The few sites still occupied by the woodrat generally had less fragmented surroundings (agricultural lands) than sites of extirpation.

Large herbivores do not easily find suitable habitat in highly urbanized settings (Adams 1994). Their large body sizes and correspondingly large home ranges exclude them from many urban environments. Nevertheless, many cities in North America have very high densities of white-tailed deer. Problems with damage to urban vegetation in sensitive areas, such as flower gardens and parks coupled with high instances of deer-vehicle accidents, have prompted some cities to initiate population control activities (Bolen and Robinson 1995).

Small insectivorous mammals, such as shrews, moles, and bats, are commonly encountered in most residential areas. Suburban residential areas often make excellent habitat for medium-sized omnivores, such as raccoons (Hoffmann and Gottschang 1997), opossums, armadillos, and skunks (Adams 1994).

Red foxes are more tolerant of urban areas than gray foxes. They occasionally den in large wooded areas within some larger cities. Urban foxes are common in many British cities, even in the districts most densely populated by humans (MacDonald and Newdick 1982). In a Boston cemetery, resident red foxes hunt a burgeoning gray squirrel population (Bolen and Robinson 1995). Gray foxes are more wary of urbanized areas, but can be found in rural residential areas (Harrison 1997). The threshold for avoidance of residential areas by gray foxes is between 130 and 325 residences per square mile. Coyotes are becoming more common in urban and suburban settings (Adams 1994). Coyotes occur in suburban Seattle and Los Angeles, in residential areas north of New York City, and in Lincoln, NE. In Lincoln, one coyote spent more than 70 percent of his time in a 35-acre residential subdivision (Bolen and Robinson 1995).

Large predators, such as wolves, cougars, and bears, are not part of urban mammal communities (Adams 1994). They have been eradicated from most rural areas as well. Black bear distribution in coastal North Carolina is negatively correlated with human density and positively correlated with percent of total forested land (Jones and others 1998).

Effects of urbanization on reptiles and amphibians

Some amphibians and reptiles have characteristics that make them vulnerable to the effects of urbanization (Adams 1994). They are less mobile than birds or mammals, and dispersal rates are slower. With habitat fragmentation, many amphibians and reptiles exist in localized distributions rather than one continuous population. Urbanization tends to exclude specialized reptiles and amphibians, while species with broad ecological tolerances and more general habitat needs tend to be more successful. Many reptiles and amphibians are eliminated when wetlands and aquatic habitats are lost due to drainage, channelization, or filling. Removal of ground cover and underbrush eliminates habitat for many salamanders and snakes (Adams 1994).

Amphibians are especially susceptible to local extirpations and constraints on recolonization due to the short distances traveled, site fidelity, and physiological constraints (Blaustein and others 1994). The effects of forest habitat loss during urbanization may be especially severe for forest-dwelling salamanders. Schlauch (1976) found that woodland salamanders, such as the blue-spotted, spotted, marbled, and eastern tiger salamander, were reduced in distribution in urbanized areas of Long Island. Loss of ponds, lowered water tables, urban pollution, reduced amounts of woodlands, and collections for pets were contributing factors. In addition, the northern two-lined salamander disappeared from most areas on Long Island due to destruction of suitable springs. This species needs cool and flowing spring water to breed. In western North Carolina, the abundance and diversity of salamanders were drastically reduced following clearcutting of the forests (Ash 1997, Petranka and others 1993). There is substantial debate about the recovery and long-term stability of salamander communities in managed forests (Ash 1999, Petranka 1999), but deforestation associated with urban development would be permanent, with little likelihood of recovery for many salamander species.

Recolonization of suitable areas can also be problematic for some reptiles, especially those that are habitat specialists. The Florida scrub lizard is a rare endemic, and its largest remaining population is in Florida sand pine scrub on the Ocala National Forest (Tiebout and Anderson 1997). The lizard has limited vagility and can only occupy young seral stages of a regenerating forest (less than 7 to 9 years of age). Scrub lizards probably do not disperse through forests older than about 12 years of age. Fire suppression and the lack of forest successional dynamics have contributed to the rarity of this lizard.

The threatened gopher tortoise also is sensitive to urbanization. Egg and hatchling mortality can be quite high in urban areas (see The effects of exotic animals on forest wildlife: exotic insect pests and forest wildlife and The effects of exotic animals on forest wildlife: effects of exotic wildlife on native forest wildlife). This problem is compounded by low reproductive rates (Adams 1994). The gopher tortoise has been extirpated from urban areas in Mobile County, AL (Nelson and others 1992). Populations are more stable, however, in areas with less severe habitat disturbance. Habitat modifications and land use changes associated with urbanization and agricultural development have eliminated the timber rattlesnake from much of its historic range in east Texas (Rudolf and Burgdorf 1997).

Although urbanization excludes some sensitive forest reptiles and amphibians, urban environments may provide habitat for some species. The heavily urbanized western end of Long Island still supported 28 of the 37 species documented to historically exist on Long Island (Schlauch 1976). The less developed, eastern end supported 35 of the 37 species. Herpetofauna found to be urban tolerant by Schlauch (1976) included the red back salamander, Fowler’s toad, the brown snake, the garter snake, and the eastern box turtle. Due to pet collection, box turtles disappeared quickly from areas near any ground-level nature trails, however.

Other general effects of urbanization on forest wildlife

Many habitats, such as the longleaf pine ecosystem or pine-oak woodlands of the Southern Appalachians, are dependent upon fire for maintenance. Fire suppression has affected the quality of wildlife habitats in some southern forests. In many forest areas, management now includes prescribed burning. However, the increasing presence of roads and residential areas has interfered with the use of prescribed fire. For more information on the effects of fire suppression and prescribed burning, see chapter 4, and chapter 25.

For more information about the effects of air pollution on forest health, see chapter 18. For more information about the effects of increasing demand for timber products on southern forests, see chapter 13.

Previous PageNext Page

Glossary | Sci.Names | Process | Comments | Draft Report


content: Ken Graham
webmaster: John M. Pye

created: 4-OCT-2002
modified: 08-Dec-2013