Environmental Issues > Wildlands Main Page > All Wildlands Documents

Top of Report

Chapter 1


Displacement of wildlife: Sensitive wildlife species are displaced by roads. They move or modify their home range as road density increases and avoid roads during daily movement activities.

Barriers to dispersal: Roads fragment populations of many small mammals, amphibians, and reptiles by creating barriers to dispersal. Direct mortality through roadkill also affects populations of both large and small animal species.

Loss of habitat: Wildlife species associated with forest interior habitat or old-growth are adversely affected by habitat degradation and by forest fragmentation due to logging or roads.

Reduced nesting success: The reproductive success of interior forest bird species decreases in areas fragmented and/or disturbed due to logging or roads. Some species are sensitive to disturbance; others suffer from increased rates of nest parasitism and nest predation.

Displacement of wildlife: Sensitive wildlife species are displaced by roads. They move or modify their home range as road density increases and avoid roads during daily movement activities.

Key Finding: Grizzly bear use of suitable habitat in Montana declined as road density and road traffic increased.

Source: Mace, R. D., J. S. Waller, T. L. Manley, L. J. Lyon and H. Zuuring. 1996. Relationships among grizzly bears, roads and habitat in the Swan Mountains, Montana. Journal of Applied Ecology 33: 1395-1404.

The authors investigated the relationship between grizzly bears (Ursos arctos horribilis), their habitat, and road density in the Swan Mountains, Montana. Radio-telemetry data from 1990 to 1994 were used to analyze home ranges. Female grizzly bears' ranges had lower road densities than non-range areas. The bears also avoided areas within a 0.5 km buffer surrounding roads that had more than 10 vehicles per day.

Eight of the grizzly bears marked in the study were killed by humans between 1988 and 1994. Improved road access resulted in illegal killings as well as management removal of bears conditioned to human foods. Combined with natural mortality, these death rates were considered too high to allow local population growth.

Key Finding: Grizzly bears used habitat near roads less than expected in the Northern Rockies, resulting in less habitat available in roaded areas.

Source: McLellan, B. N. and D. M. Shackleton. 1988. Grizzly bears and resource-extraction industries: Effects of roads on behaviour, habitat use and demography. Journal of Applied Ecology 25: 451-460.

The authors studied 27 grizzly bears (Ursus arctos) over seven years, using radio-telemetry in the Rocky Mountains at the British Columbia/Montana border. Road use or avoidance was analyzed. Bears used habitats within 100 m of roads less than expected in spring and habitats within 250 m of roads less than expected in summer through autumn. Avoidance of roads was independent of the volume of traffic on the roads.

An analysis of habitat characteristics supported the authors' conclusion that bears were displaced by the roads themselves rather than by unfavorable habitat types along roads. The authors note that many of the habitats close to roads had important bear foods, and avoidance of these areas resulted in a habitat loss of 8.7%.

Key Finding: Black bears crossed roads with higher traffic volumes less frequently than roads with lower traffic volumes.

Source: Brody, A. J. and M. R. Pelton. 1989. Effects of roads on black bear movements in western North Carolina. Wildlife Society Bulletin 17: 5-10.

The frequency of road crossing by bears was investigated using telemetry data from 17 black bears (Ursus americanus) in Pisgah National Forest, North Carolina. The bears rarely crossed the interstate highway within the study area. Crossing of abandoned, restricted, and unrestricted roads was also analyzed, with highest crossing rates found on abandoned roads. Roads of low traffic volume were crossed more frequently than roads with higher traffic volume.

Key Finding: Habitat occupied by wolves in Minnesota had a lower road density than unoccupied habitat.

Source: Mech, L. D., S. H. Fritts, G. L. Radde and W. J. Paul. 1988. Wolf distribution and road density in Minnesota. Wildlife Society Bulletin 16: 85-87.

Road density and wolf distribution were studied in northeastern Minnesota to evaluate threshold road densities for the occurrence of wolves. Wolf distribution data were obtained from authors who had a long-term knowledge of the area, from Department of Natural Resources personnel, and from surveys of 112 canid trappers.

Road density was found to be inversely correlated with current wolf populations. Road densities within the entire wolf range in Minnesota, as well as within the primary wolf range, peripheral areas, and disjunct areas, all fell below the threshold road density of 0.58 km/km2 previously determined by Thiel (1985) and Jensen et al. (1986).

Key Finding: Wolves showed a preference for areas with low road density rather than high road density when establishing packs in the northern Great Lakes region.

Source: Mladenoff, D. J., T. A. Sickley, R. G. Haight and A. P. Wydeven. 1995. A regional landscape analysis and prediction of favorable gray wolf habitat in the northern Great Lakes region. Conservation Biology 9: 279-294.

The authors analyzed recolonization by the eastern timber wolf (Canis lupus lycaon) into northern Wisconsin and upper Michigan from Minnesota. They used data from radiocollared wolves and GIS information to evaluate characteristics within new wolf pack areas. Road density and land cover complexity proved to be the most important variables in their models for predicting occurrence of wolf packs. Wolves strongly selected areas with low road density as opposed to high road density.

The authors believe that wolves moved through a wide area, including unfavorable habitat, but established successfully only in higher quality habitat, low human access being one of the characteristics of the latter. They report that upper Michigan and Minnesota, with their greater area of contiguous forest, are a source population of wolves for the state of Wisconsin, where forests are more fragmented.

Key Finding: Wolves in Alaska avoided roads that were open to regular public use.

Source: Thurber, J. M., R. O. Peterson, T. D. Drummer and S. A. Thomasma. 1994. Gray wolf response to refuge boundaries and roads in Alaska. Wildlife Society Bulletin 22: 61-68.

Sixty-four radiocollared gray wolves (Canis lupus) were studied from 1976 to 1980 in Kenai National Wildlife Refuge, Alaska. Their response to different road types and human presence was evaluated by computing levels of wolf activity at varying distances from roads and boundaries. Wolves avoided oilfield access roads, which were open to regular and frequent public use. Roads with limited human use (gated pipeline access roads and secondary gravel roads) were used as travel corridors. The authors concluded that the absence of wolves from settled areas and heavily used roads was due to behavioral avoidance.

Of the total radiocollared wolf mortality in the study area, 87% were killed by humans, with one illegal killing. Hunting by people was concentrated in areas of easy human access. A wolf pack using a den within 1 km of a highway did not obviously avoid the highway. Ten of the 12 pack members were killed by hunting in the last years of the study, between 1978 and 1979.

Key Finding: Roadless areas are important reservoirs for maintaining wolf populations in adjacent, high-road-density areas.

Source: Mech, L. D. 1989. Wolf population survival in an area of high road density. American Midland Naturalist 121: 387-389.

Seventy-one wolves were radiocollared from 1969 to 1986 in a study area in Superior National Forest, Minnesota, with an average road density of 0.73 km/km2. The area is adjacent to a roadless area - the Boundary Waters Canoe Area Wilderness. Mortality attributed to human causes (including roadkill and trapping) was high - 69% of radiocollared wolves were killed. Wolf survival in this study area was considered below the threshold necessary to maintain a wolf population.

No human-caused mortality was detected in the roadless area. Of 53 wolves tagged in a roaded area north of the study area, 22 had emigrated from the roadless area. The author concludes that wolf populations in an area of high road density could only be sustained if there are suitable roadless reservoirs nearby.

Key Finding: Mountain lions avoided improved dirt roads and hard-surfaced roads and selected home range areas with lower densities of these road types.

Source: Van Dyke, F. G., R. H. Brocke and H. G. Shaw. 1986. Use of road track counts as indices of mountain lion presence. Journal of Wildlife Management 50: 102-109.

Mountain lions were radiocollared in Arizona and Utah on three study sites to assess their interactions with roads and the efficacy of track surveys for determining populations. Improved dirt roads and hard-surfaced roads were crossed less often than unimproved dirt roads, suggesting avoidance. These roads were also less likely to occur within lion home range areas. Of the home range areas evaluated, 58% had improved dirt roads, 23% had hard-surfaced roads, while 85% had unimproved dirt roads.

Key Finding:Female Roosevelt elk reduced their daily movements, core area size, and home range size, and therefore, their energy needs, when disturbance due to vehicular access on roads was limited by gate closures.

Source: Cole, E. K., M. D. Pope and R. G. Anthony. 1997. Effects of road management on movement and survival of Roosevelt elk. Journal of Wildlife Management 61: 1115-1126.

The effect of road closures on female Roosevelt elk (Cervus elaphus nelsoni) in the Southern Oregon Coast range was examined. Movements and survival were studied before and after limits were imposed on vehicle access. Roads were closed using gates; monitoring indicated that access was successfully limited to up to 4 vehicle trips per week (including legal entry for management purposes and illegal entry). Of the 29 radiocollared elk, 20 survived through the entirety of the study and were used for much of the analysis. Elk using the study area were compared with a control group that used less than 30% of the study area. During the period of limited vehicle access, core area size and home range size were significantly reduced for the former group than during open access. There was also a reduction in elk daily movement distances.

The authors conclude that limiting access by vehicles reduced human disturbance of elk. Their reduced movements as a result would suggest that the elk would use less energy, potentially increasing fat reserves, survival rates, and productivity. Poaching mortality was also reduced in the areas with limited vehicle access.

Key Finding:Mule deer and elk avoided roads and areas within 200 m of roads.

Source: Rost, G. R. and J. A. Bailey. 1979. Distribution of mule deer and elk in relation to roads. Journal of Wildlife Management 43: 634-641.

Mule deer (Odocoileus hemionus) and elk (Cervus canadensis) responses to roads during winter was assessed by counting fecal-pellet groups at varying distances from roads. The study was conducted in Colorado - east of the continental divide in Roosevelt National Forest, and west of the divide in White River National Forest. Transects perpendicular to roads, 400 m long, with 40 plots each, were established at each study site.

Only study sites having at least 50 fecal-pellet groups were included in the analysis. For deer, pellet group locations were analyzed at a total of 63 sites. Pellet group densities increased with distance from the road. Deer showed greater avoidance of roads east of the continental divide, with statistically significant results at 69% of the sites. West of the divide, this relationship was significant for 35% of the sites.

For elk, pellet groups were analyzed at 20 sites. East of the continental divide, elk avoided roads in 90% of the sites, with statistically significant results from 50% of the sites. West of the divide, elk pellet group density decreased at 60% of the sites, with 20% decreases being significant. The authors note that elk behavior relative to roads on the west slope could not be clearly established from these results. Deer and elk frequently avoided both heavily traveled roads and roads near human habitation.

Key Finding:Columbian black-tailed deer were displaced from their usual home ranges by increased vehicular traffic during the hunting season.

Source: Livezey, K. B. 1991. Home range, habitat use, disturbance, and mortality of Columbian black-tailed deer in Mendocino National Forest. California Fish and Game 77: 201-209.

The movements and home ranges of Columbian black-tailed deer (Odocoileus hemionus columbianus) were studied in an area on the west slope of California's Coast Range, including part of the Mendocino National Forest. Sixteen female deer were radiocollared and observed over almost 2½ years. Traffic increased considerably on secondary roads and smaller roads during the hunting season. Four of the radiocollared deer used areas 10-200 m from secondary roads in the summer and fall season. However, during the fall hunting period, these deer were displaced and moved 0.6-2.5 km away from their usual areas. The authors conclude that displacement by increased traffic reduced the amount of habitat available to the deer.

Key Finding: Eastern massasauga rattlesnakes avoided roads in all seasons.

Source: Weatherhead, P. J. and K. A. Prior. 1992. Preliminary observations of habitat use and movements of the eastern massasauga rattlesnake (Sistrurus c. catenatus). Journal of Herpetology 26: 447-452.

The authors used radiotelemetry to track 12 eastern massasauga rattlesnakes (Sistrurus c. catenatus) for 419 days. This species persists in small relict populations in the U.S. and Canada and is considered threatened. Tracking data showed that the snakes avoided open areas, including roads and trails, mixed forest, and open water at all times of the year. The snakes were most strongly associated with coniferous forests and wetlands. In contrast, capture rates were higher in open areas, and the authors believe this to be due to easier detection in these areas.

Barriers to dispersal: Roads fragment populations of many small mammals, amphibians, and reptiles by creating barriers to dispersal. Direct mortality through roadkill also affects populations of both large and small animal species.

Key Finding:Roads were a barrier to movement by the eastern chipmunk and the white-footed mouse.

Source: Oxley, D. J., M. B. Fenton and G. R. Carmody. 1974. The effects of roads on populations of small mammals. Journal of Applied Ecology 11: 51-59.

Trapping and observation were used to study small mammal population movements along roads in southeastern Ontario and in Quebec. Four types of roads were included - county gravel roads, county paved roads, two-lane highways, and four-lane highways. All sites except one were oak-maple mixed forest; the exception was primarily coniferous forest. A total of 589 individuals were trapped.

Out of a total of 651 recaptures, 98% were white-footed mice (Peromyscus maniculatus) and eastern chipmunks (Tamias striatus); therefore, analysis focused on them. Only eight of 254 trapped white-footed mice and six of 179 trapped eastern chipmunks crossed roads; none crossed highways involving more than 30 m of clearance.

Road clearance was determined to be the most important factor for crossing by forest mammals. Road surface (gravel vs. paved) was not a significant factor. Traffic was not necessarily an inhibiting factor - one of the divided highways had very low traffic volume (four vehicles per hour), but experienced very little crossing. Wider roads were crossed almost exclusively by medium-sized mammals, such as skunks and porcupines, rather than by small mammals.

Road mortality was highest in July, when traffic levels were highest and when the young of several species were emerging and dispersing. More than 380 mammals were found killed over 116 days, as well as 150 amphibians, 228 reptiles, and 217 birds.

The authors conclude that roads may affect the survival of populations by fragmenting gene pools and that this effect should be considered when planning roads.

Key Finding:Highways were a barrier to movement for seven of 10 rodent species studied.

Source: Wilkins, K. T. 1982. Highways as barriers to rodent dispersal. Southwestern Naturalist 27: 459-460.

The authors examined road-crossing behavior of rodents by capturing and marking 1,968 rodents of 10 species. The study was conducted adjacent to four highways in southeastern Texas. Individuals from only three of the 10 species crossed the roads - five Baiomys taylori (1.8% of marked individuals); one Reithrodontomys fulvescens (1% of those marked); and 86 Sigmodon hispidus (5.6% of those marked).

Key Finding: A narrow dirt road was a significant barrier to movement by prairie voles and cotton rats.

Key Finding:Prairie voles and cotton rats tended to move away from a narrow dirt road rather than toward it.

Source: Swihart, R. K. and N. A. Slade. 1984. Road crossing in Sigmodon hispidus and Microtus ochrogaster. Journal of Mammalogy 65: 357-360.

During a nine-year period, prairie voles (Microtus ochrogaster) and cotton rats (Sigmodon hispidus) were live-trapped in abandoned farmland on both sides of a lightly traveled narrow dirt road in Jefferson County, Kansas. Out of a total of 1,865 captured prairie voles, only 23 were recorded as crossing the road. Out of 823 captured cotton rats, 47 crossed the road. Preference for certain habitat rather than road avoidance was tested as a reason for this behavior, but found not to be a factor.

The effect of roads on vole and mouse behavior was examined in three ways. First, voles and mice near the road were found to move preferentially away from rather than toward the road: 92% of captured voles moved away from rather than across the road; 81.5% of cotton rats moved away from the road. Second, the behavior of individual roadcrossers was compared with non-roadcrossers. The majority of roadcrossers were found to cross the road only once each. The authors interpret this as indicating that the road-crossing individuals did not incorporate the road into their home range. Third, the likelihood of increased road crossing with increasing population density was tested. Road crossing was found to be density dependent for cotton rats but not for voles. That is, in years of lower population size, fewer cotton rats crossed the road, but no effect was found on voles, which always had a very low crossing rate.

The paper also addresses the question of whether the road functions as a barrier to gene flow. The authors observe that it is a difficult question to answer, but that the inhibitory effect of the road on voles may be strong enough to have ramifications on the population's genetic diversity.

Key Finding:A highway in southern Nevada acted as a barrier to crossing for all eight rodent species studied.

Source: Garland, T., Jr. and W. G. Bradley. 1984. Effects of a highway on Mojave Desert rodent populations. The American Midland Naturalist 111: 47-56.

A rodent trapping study was conducted from March 1979 through February 1980 near a divided four-lane highway in southern Nevada. The study area, a grid over 9 ha, was in a creosote bush community. A total of 612 rodents of eight species was captured over the course of the study. Of these, almost two thirds, 387 individuals, were recaptured.

The road was found to inhibit crossing by all rodent species. All species traveled distances large enough to permit crossing the highway. However, only one individual, an adult male antelope ground squirrel (Ammospermophilus leucurus), crossed the entire highway. There was no relationship between proximity to the highway and home range size or life span. There was no road mortality in the study area, and the authors believe that crossing inhibition prevented roadkill.

Key Finding: Roads impeded movement by amphibians and could result in population isolation.

Key Finding: Despite some speculation, road ruts and ditches have not been shown to provide successful amphibian breeding habitat rather than acting as ecological traps.

Key Finding:Amphibians play a key role in the forest ecosystem, affecting nutrient cycling and also serving as high-quality prey for many species.

Source: deMaynadier, P. G. and M. L. Hunter, Jr. 1995. The relationship between forest management and amphibian ecology: a review of the North American literature. Environmental Reviews 3: 230-261.

This article reviews the impact of a variety of forest management practices, including logging roads, on amphibians. Studies reported significant roadkill on busy roads. A more significant impact of roads, however, may be as barriers to dispersal. Isolated populations could suffer a loss of genetic diversity. Although the barrier effect on many small mammals has been shown, almost no studies have been performed on the effect of unpaved roads and amphibian movement. The authors report some of their own findings, however, from a drift fence study along a 5-m-wide dirt track and a 12-m-wide gravel logging road in Maine. Although the dirt track had no significant impact on amphibian movement, the wider, gravel track inhibited movement by salamander species.

Road puddles and roadside ditches have been discussed as potential new breeding habitat. Salamanders and frogs have been documented breeding in rut ponds on abandoned logging roads in Kentucky. However, deMaynadier and Hunter note that no study had been done as yet to prove the level of reproductive success in these sites compared to natural breeding pools and to confirm that these road breeding sites are not serving as ecological traps, with high mortality through higher drying rates or higher predation rates.

The authors also review the ecological importance of amphibians. They are an important part of forest food chains as high-quality prey for many predators including birds, small mammals, snakes, and other amphibians. Amphibians are also believed to play an important role in forest nutrient cycling. Salamander species, for example, are top predators within the detritus food web and regulate populations of soil microfauna. The authors therefore believe that any practice that modifies local salamander populations may affect decomposition and nutrient cycling rates.

Key Finding: Roads impeded dispersal of all six amphibian species studied.

Source: Gibbs, J. P. 1998. Amphibian movements in response to forest edges, roads, and streambeds in southern New England. Journal of Wildlife Management 62: 584-589.

Amphibian dispersal relative to roads, forest edges, and streambeds was examined on a 100-ha preserve near New Haven, Connecticut. Six species, for which more than 10 individuals were captured, were included in the analysis: spotted salamander (Ambystoma maculatum), marbled salamander (Ambystoma opacum), pickerel frog (Rana palustris), redback salamander (Plethodon cinereus), wood frog (Rana sylvatica), and red-spotted newt (Notophthalmus viridescens). Capture rates of each species varied. For all species, forest-road edges had lower permeability than forest-open land edges and forest-residential edges. Red-spotted newts exhibited the strongest avoidance of forest edges.

Key Finding: Road mortality of 7 amphibian species, 10 reptile species, 21 mammal species, and 62 bird species was documented during four years of study, exceeding 32,000 individuals on a 3.6 km stretch of highway.

Source: Ashley, E. P. and J. T. Robinson. 1996. Road mortality of amphibians, reptiles and other wildlife on the Long Point Causeway, Lake Erie, Ontario. Canadian Field-Naturalist 110: 403-412.

The authors censused the mortality of amphibians, reptiles, and other wildlife along 3.6 km of a two-lane paved causeway next to Big Creek National Wildlife Area on Lake Erie. Censusing was carried out from spring 1979 to autumn 1980 and again from 1992 to 1993, for a total of 716 days.

The total recorded mortality was greater than 32,000 individuals, from 100 different species - 7 amphibians, 10 reptiles, 21 mammals, and 62 birds. Road mortality rates averaged 11.65 amphibians/km/day, 0.34 reptiles/km/day, 0.51 birds/km/day, and 0.11 mammals/km/day. Amphibians accounted for 92.1% of the total road mortality, the majority being young leopard frogs (Rana pipiens). Amphibian and reptile mortality had seasonal patterns based on their dispersal behavior.

Key Finding: Frog and toad density near paved roads decreased with increasing traffic intensity.

Key Finding: Frog and toad mortality on roads increased with increasing traffic intensity.

Source: Fahrig, L., J. H. Pedlar, S. E. Pope, P. D. Taylor and J. F. Wegner. 1995. Effect of road traffic on amphibian density. Biological Conservation 73: 177-182.

Frog and toad populations were studied along two-lane paved roads in two regions of Ottawa, Canada. Traffic intensity was categorized as low, medium, or high. Dead and live frogs and toads were counted along each 1-km segment of the roads. Relative densities of the amphibians were estimated using breeding chorus intensity rankings. Choruses were identified to species and given an intensity rating of 1 (for 1 individual); 2 (distinguishable individuals); or 3 (indistinguishable individuals).

In total, 1,856 dead frogs and 591 live frogs were counted over a total road distance of 506 km. The proportion of dead frogs and toads increased with increasing traffic intensity. After correcting for effects of local habitat, date, time, and region, frog and toad density was found to decrease with increasing traffic intensity.

Key Finding: Road mortality of 20 species of snakes was recorded along a 44-km stretch of highway passing through Organ Pipe Cactus National Monument, Arizona.

Source: Rosen, P. C. and C. H. Lowe. 1994. Highway mortality of snakes in the Sonoran Desert of southern Arizona. Biological Conservation 68: 143-148.

The authors studied snake mortality on Arizona State Routes 85 and 86 for four years, on a section of the highway passing primarily through Organ Pipe Cactus National Monument (ORPI). Because of their earlier mark-recapture work, the authors also had information on the abundance and movement ecology of snakes in ORPI.

Snakes were observed by driving slowly and recording all snakes, dead or alive, on the road surface ("quantitative road-cruising"). During their four years of research, the authors recorded the mortality rates for 20 species of snakes due to roadkill. Snake mortality was highest in spring. The authors computed an estimate of 13.5 snakes killed/km/year.

The Organ Pipe shovelnosed snake (Chionactis palarostris), which occurs only at and near ORPI, was found almost exclusively at road-modified habitat. Most individuals found were dead (70%). The authors believe that this species' population status may be strongly affected by highway mortality.

Key Finding: Mortality due to roadkill was documented for northern saw-whet owls and eastern screech-owls over a 10-year period in New Jersey.

Source: Loos, G. and P. Kerlinger. 1993. Road mortality of saw-whet and screech-owls on the Cape May peninsula. Journal of Raptor Research 27: 210-213.

Mortality of raptors due to roadkill was documented for 10 years along a 145-km route in New Jersey that included county roads, state roads, and highways. The location, age, and sex of birds were recorded. A total of 250 road-killed raptors representing six owl and six hawk species were found. Northern saw-whet owls (Aegolius acadicus) accounted for 45% of all roadkills, with 79% of these being less than one year old. Eastern screech-owls (Otus asio) accounted for 36% of all roadkills, with 88% of these being less than one year old.

Key Finding: Road mortality along a highway in Ohio was surveyed for one year and included 11 species of mammals, 12 species of birds, 11 species of amphibians, and at least 249 species of insects.

Source: Seibert, H. C. and J. H. Conover. 1991. Mortality of vertebrates and invertebrates on an Athens County, Ohio, highway. Ohio Journal of Science 91: 163-166.

The authors walked 1.6 km of a highway in Ohio, with 50 excursions over the course of more than a year to document the species killed along roads. The highway was bordered by a riverine elm-maple-sycamore woodland on one side and a red oak-white oak hillside forest on the other. Both sides of the road were surveyed, generally at weekly intervals.

They found 79 roadkilled mammals, representing 11 species, 21 birds representing 12 species (including non-roadside feeders), and 74 amphibians, representing 11 species. Amphibian road mortality was particularly high after a rainfall. Over 1,000 killed insects were collected from the road, belonging to at least 249 species.

Key Finding: Collision with a vehicle was the highest cause of death for female moose studied in Alaska.

Source: Bangs, E. E., T. N. Bailey and M. F. Portner. 1989. Survival rates of adult female moose on the Kenai Peninsula, Alaska. Journal of Wildlife Management 53: 557-563.

The authors radiocollared 51 adult female moose in two wintering areas on the Kenai National Wildlife Refuge. They tracked the survival of the tagged moose for six years and mapped their home ranges. Of the 13 deaths recorded, six were due to vehicle collisions. The remainder were due to hunting, old age, predation, and natural accidents. Roads were present in a high percentage of moose home ranges: of the tagged moose, 23 had highways within their home range, and 30 had gravel roads within their home range.

Key Finding: Mortality of white-tailed deer due to roadkill was documented for 18 months along an interstate highway.

Source: Puglisi, M. J., J. S. Lindzey and E. D. Bellis. 1974. Factors associated with highway mortality of white-tailed deer. Journal of Wildlife Management 38: 799-807.

The number of white-tailed deer (Odocoileus virginianus) killed along a 313-mile-long interstate highway in Pennsylvania was analyzed from monthly reports completed by state game protectors. Over a period of 18 months, 874 deer were killed due to collisions with vehicles. Mortality was highest during November and December. Vegetation, fence type, fence location, and topography were recorded along the highway in order to investigate mitigation possibilities.

Key Finding: Road mortality rates of white-tailed deer were documented after the construction of an interstate highway through their wintering area.

Source: Reilly, R. E. and H. E. Green. 1974. Deer mortality on a Michigan interstate highway. Journal of Wildlife Management 38: 16-19.

Information on white-tailed deer (Odocoileus virginianus) mortality along an interstate highway was collected from the Michigan Department of Natural Resources. The authors investigated a 5-mile stretch of a new interstate highway that intersected a deer wintering area (already with one road).

Roadkill data from 1960 through 1972 were available. After construction of the interstate, rates of deer mortality rose, and in the first year were 500% greater than the average of the previous four years. Mortality rates declined over the following years, but remained twice as high as pre-interstate mortality figures. The decline in mortality was attributed to deer moving away from wintering in areas affected by the highway as well as the deer families formally occupying the highway area having already been killed.

Key Finding: Mortality of mule deer due to roadkill was documented for two years along a highway and two state roads.

Source: Romin, L. A. and J. A. Bissonette. 1996. Temporal and spatial distribution of highway mortality of mule deer on newly constructed roads at Jordanelle Reservoir, Utah. Great Basin Naturalist 56: 1-11.

Mortality of mule deer (Odocoileus hemionus) due to roadkill was studied on segments of three highways in northeastern Utah. Deer roadkill data were collected over a total of 47.3 km for two years. One highway was a four-lane road, the other two were two-lane state roads. Vegetation characteristics, topography, traffic volume, and road variables were also noted.

The authors recorded 397 deer roadkills during their study period. Of the total, 51.6% were does, 18.9% bucks, 21.7% fawns. Roadkill rates were highest along the four-lane highway.

Loss of habitat: Wildlife species associated with forest interior habitat or old-growth are adversely affected by habitat degradation and by forest fragmentation due to logging or roads.

Key Finding: Marten capture rates declined as forest fragmentation increased, and the animals were rarely detected in sites with more than 25% non-forested area in a total 9 km2 area.

Source: Hargis, C. D., J. A. Bissonette and D. L. Turner. 1999. The influence of forest fragmentation and landscape pattern on American martens. Journal of Applied Ecology 36: 157-172.

The authors investigated the extent to which the abundance of American martens (Martes americana) changed with incremental increases in forest fragmentation. Their study was conducted for three years in the Uinta Mountains of northern Utah, during a moratorium on commercial marten trapping. Eighteen study sites were selected, each 9 km2 in area and large enough to potentially overlap several martens' ranges. Sites were composed of mature forest, dominated by Engelmann spruce, lodgepole pine, and subalpine fir. The area of open, non-forested habitat (natural openings and areas clearcut at least five years earlier) ranged from 2% to 42% of a site. The dominant vegetation in these openings was grasses and forbs.

Twenty-five traps were placed on each site over the summer. Martens' reproductive status, weight, body condition, and recapture rates were used as indicators of their health. Habitat fragmentation was quantified using five measures: the percentage of landscape in openings, edge density (m of edge per ha), isolation of each open patch, nearest-neighbor distances of openings, and mass fractal dimension. Clearcuts and natural openings were combined to determine the percentage of landscape in openings.

Marten captures decreased with increasing loss of forest habitat as measured by the percentage of each site in open areas. Their capture rates were lowest in landscapes that had large, closely spaced open areas. Landscapes with an average distance between open areas of less than 100 m had no marten captures. Landscapes with high edge densities also had few marten captures. Prey availability did not correlate with marten captures.

Key Finding: Mountain lions avoided logging areas and established home ranges in areas with lower road densities than the average in the area.

Source: Van Dyke, F. G., R. H. Brocke, H. G. Shaw, B. B. Ackerman, T. P. Hemker and F. G. Lindzey. 1986. Reactions of mountain lions to logging and human activity. Journal of Wildlife Management 50: 95-102.

Radiocollared mountain lions (Felis concolor) were studied in northern Arizona (Kaibab Plateau) and south-central Utah (Escalante Study Area) to evaluate the lions' reactions to logging, road densities, and human activity. Mountain lions using the same area for six or more months were considered residents; all others were considered transients. Lion locations were classified as: lions in an active timber sale, within 1 km of an active sale, in an inactive sale, within 1 km of an inactive sale, or not in or near a sale. The area of timber sales within a mountain lion's home range was also examined. Daily lion activity patterns were recorded.

Lions used timber sale areas less frequently than expected given the proportion of the landscape in which there was sale activity. Resident lions rarely occurred in areas within 1 km of sites logged within the previous six years. Only one of seven lion home ranges in the Kaibab Plateau included timber sales, although six included portions of the overall timber zone. Lions in the Escalante study site had larger home ranges than did lions in the Kaibab Plateau. Their home ranges included all or part of at least one timber sale, perhaps because timber sales were more dispersed through the landscape. Home ranges for established residents and for newly established young lions were in areas with lower road densities than the study area average.

Key Finding: Northern flying squirrels, the primary prey of northern spotted owls, occurred at lower densities in logged, shelterwood stands than in unmanaged, old-growth forest.

Source: Waters, J. R. and C. J. Zabel. 1995. Northern flying squirrel densities in fir forests of northeastern California. Journal of Wildlife Management 59: 858-866.

The density of northern flying squirrels (Glaucomys sabrinus) was compared in three different stand types: old-growth forest, shelterwood stands that had been logged and burned, and young, naturally regenerated stands 75 to 95 years old. All study sites were located in Lassen National Forest in the Cascade Range of California. White and red fir were the dominant tree species. Flying squirrels were trapped at locations 40 m apart. Vegetation characteristics, including logs, snags, stumps, and canopy cover were measured. Sporocarps of hypogeous fungi, the preferred food of northern flying squirrels, were also sampled.

The mean density of flying squirrels was greatest in old stands and lowest in shelterwood stands. Mean densities did not differ between old and young stands. Body mass (a measure of habitat quality) of adult males and females did not differ between stand types. Fungal sporocarps were found most frequently in old stands and least frequently in shelterwood stands. Sporocarp density was found to be correlated with mean flying squirrel density.

Key Finding: California red-backed voles were more abundant in old-growth forest and naturally regenerated stands than in young, managed stands. Their higher abundance correlated well with the deeper organic soil layers measured in unmanaged stands.

Source: Rosenberg, D. K., K. A. Swindle and R. G. Anthony. 1994. Habitat associations of California red-backed voles in young and old-growth forests in western Oregon. Northwest Science 68: 266-272.

The abundance of California red-backed voles (Clethrionomys californicus) was investigated in the Willamette National Forest in the Oregon Cascades. Ten stands were included in the study: five stands were old-growth, dominated by Douglas-fir and western hemlock, four stands were young, managed stands 30 to 60 years old (clearcut and planted), and one stand was a young forest that had naturally regenerated after a wildfire. Voles were trapped using Sherman live-traps spaced at 20-m intervals on trapping grids of 3.2 ha.

During the four years of the study, 209 individual voles were captured. Significantly more voles (80.4%) were captured in old-growth than in young stands (19.6%). Of the young stands, the majority captured were from the unmanaged, naturally regenerated stand (80%). No voles were captured in one of the managed young stands.

Vegetation and soil characteristics were also recorded. The density of large trees, large snags, and organic soil layer depth were most highly correlated with vole abundance. The authors concluded that of these, forest floor depth may be most predictive of vole abundance since the animals nest in underground burrows or under debris, forage in the organic layer of soil, and use coarse woody debris for cover, travel, and as a source of fungi to feed on. Average organic soil depths were similar in the naturally regenerated young stand (7.8 cm) and in the old-growth (7.9 cm) stand.

The authors note that other scientists' research on California red-backed voles has shown that these voles are an important prey species for many birds and animals and are key dispersers of mycorrhizal fungi, and, therefore, may be an important part of forest ecological processes.

Key Finding: California red-backed voles were adversely affected by habitat fragmentation: they were absent in clearcuts, had low densities at the edges of forest remnants, increased in density toward the forest interior, and had higher abundances in large forest fragments compared to small fragments.

Key Finding: Truffles, the primary food source of red-backed voles, were absent in clearcuts and near the edges of forest remnants, but occurred in forest interiors.

Source: Mills, L. S. 1995. Edge effects and isolation: red-backed voles on forest remnants. Conservation Biology 9: 395-403.

The author investigated the influence of forest edges on the distribution of California red-backed voles (Clethrionomys californicus) in the Klamath Mountains of Oregon. Thirteen mature-to-old-growth forest remnants were studied, ranging from 0.6 to 2.5 ha in size. Each was surrounded by 1- to 30-year-old clearcuts. Five control sites, mature-to-old-growth forests more than 250 ha in size, were also included.

Voles were trapped for two years between June and September. Trap locations were assigned to one of four edge classes based on the trap's distance into the forest interior from the edge: 0-15 m, 16-30 m, 31-45 m, and 46-90 m. Log number and volume was sampled in each edge class. The distribution of truffles (a vole food source) was also sampled in forest remnants and clearcuts.

Voles were exceptionally rare in clearcuts, and out of 1,404 trap nights, only three voles were captured in clearcuts. Voles were captured on all five control sites and on 10 of 13 forest remnants, with marginally fewer voles per trap in remnants than in controls. Edges had a significant negative effect on vole numbers, with fewer voles captured closer to the edge compared to the forest interior. Vole density increased as forest remnant area increased, confirming the negative impact of edges.

Truffles (sporocarps of mycorrhizal fungi) were essentially absent on clearcuts and in the first edge class, but occurred further into the forest interior. The authors reported that truffles account for three quarters of the diet of this species of vole, and their presence or absence may be an important mechanism explaining the decrease of voles in clearcuts and edge habitat. Log presence or absence did not explain the negative effect of edges on voles, which they speculated may have been because degree of log decay was not included in the sampling.

Key Finding: Red-backed salamanders, sensitive to forest moisture and temperature levels, were more abundant in old-growth forest and 60-year-old second-growth than in clearcuts or selectively logged forest.

Key Finding: Salamanders are a critical part of the forest food chain: they are important food sources for birds and mammals, and as predators themselves, they cycle large amounts of energy through the forest ecosystem.

Source: Pough, F. H., E. M. Smith, D. H. Rhodes and A. Collazo. 1987. The abundance of salamanders in forest stands with different histories of disturbance. Forest Ecology and Management 20: 1-9.

Four pairs of study plots were established in New York State - four in old-growth forest and four in adjacent disturbed stands, the latter of which included a seven-year-old clearcut forest, a 25-year-old clearcut planted with conifers, a forest cut selectively for firewood, and a 60-year-old second-growth forest. Two 50 x 2 m transects were established in each plot. Understory vegetation cover, leaf litter depth, and soil properties were sampled. Salamanders were counted at night, with each pair of stands being surveyed a total of five times during the study.

The authors found two species of salamanders along their transects - red-backed salamanders (Plethodon cinereus) and the terrestrial eft stage of red-spotted newts (Notophthalmus viridescens). The largest source of variation in salamander numbers, other than night-to-night differences, was stand stage. Red-backed salamanders were most abundant in second-growth forest and its adjacent old-growth control plot, and least abundant in the recent clearcut. This species is believed to be particularly susceptible to changes in microhabitat as it spends all its life on land rather than partly in water and must have suitable forest habitat through all its life stages. Depth of leaf litter was the best predictor of the frequency of above-ground activity by these salamanders.

The authors report red-spotted newts as having been previously shown to be more tolerant of heat and dry conditions than red-backed salamanders. This species spends its larval and later adult life in the water and is only on land during the eft stage. Efts were most abundant in the old-growth plot next to the conifer plantation and rarest in the plantation, but more abundant in the firewood stand than its adjacent old-growth. The authors suggest that this may be because of the abundant piles of down wood present in the firewood forest stand.

The authors also review research on salamanders in various forests of the United States and report that these amphibians' high biomass makes them an important part of the forest food chain. In places such as the Hubbard Brook Experimental Forest in New Hampshire, the biomass of salamanders is twice that of birds and equal to that of small mammals. They are important food sources for birds and mammals and themselves exploit much of the small prey, thereby contributing greatly to the forest ecosystem energy flow.

Key Finding: The abundance of amphibians was significantly lower in clearcuts, plantations, and forest edges than in mature forest interior sites.

Key Finding: Lungless salamanders, such as the red-backed salamander, are particularly vulnerable to population declines due to clearcut logging.

Source: deMaynadier, P. G. and M. L. Hunter, Jr. 1998. Effects of silvicultural edges on the distribution and abundance of amphibians in Maine. Conservation Biology 12: 340-352.

Five sites in Maine were chosen to study the effects of logging and edges on amphibian populations. Three clearcut stands (ranging from 2 to 11 years old) and two conifer plantations (5 and 25 years old) were paired with adjacent mature forest stands as controls. Transects 140 m long were established perpendicular to the forest edge and 70 m into each stand type. Drift fences and pitfall traps were used to capture the amphibians. Habitat variables were also recorded, including ground cover, vegetation characteristics, litter depth, and ambient light levels.

A total of 2,394 amphibians of 14 species were captured. This included six salamander species and eight anuran (frog and toad) species. All statistical analyses were based on a catch-per-unit effort (number of animals per 100 trap nights) to standardize sampling efforts.

The overall amphibian capture rate was significantly lower in clearcuts and plantations than in the mature forest control sites. The abundance of all six salamander species and seven out of eight anurans increased significantly on plots closer to the forest interior than to the edge. Four species were identified as being particularly sensitive to forest management: red-backed salamanders (Plethodon cinereus), spotted salamanders (Ambystoma maculatum), blue-spotted salamanders (Ambystoma laterale), and wood frogs (Rana sylvatica). Red-backed salamanders were the most sensitive to clearcutting and forest edge effects. All four management-sensitive species occurred in higher numbers in forest interior habitat than at edges.

Both the distance that edge effects extended into the forest interior and edge contrast were analyzed. The four management-sensitive amphibian species were found to be negatively affected at distances up to 25 to 35 m from silvicultural edges. For salamanders as a group, high-contrast forest edges had a stronger negative impact on abundance than low-contrast edges.

The microhabitat variables that potentially limited populations were also identified. There was a strong positive association between species abundance and canopy cover levels, percent cover by snags, stumps, and root channels, and litter coverage and depth.

Key Finding: Clearcuts had a significantly lower abundance and fewer species of salamanders compared to mature, 50- to 70-year-old forest stands in the southern Appalachians.

Key Finding: Plethodon salamanders are unlikely to survive logging because individuals are closely tied to small home ranges and unlikely to relocate to intact forest from logged areas.

Source: Petranka, J. W., M. E. Eldridge and K. E. Haley. 1993. Effects of timber harvesting on southern Appalachian salamanders. Conservation Biology 7: 363-370.

The authors compared the species richness and abundance of salamanders in recent clearcuts and mature forest stands in Pisgah National Forest in the southern Appalachian Mountains of North Carolina.

Plots 50 m by 50 m in size were established on 47 sites, including six recent clearcuts (2-10 years old) and 34 mature, mixed deciduous forest stands (more than 50 years old). Plots in clearcuts were at least 20 m from adjacent uncut forest. A total of 12 species of salamanders were collected (828 individuals) during day searches, primarily in the genus Plethodon and Desmognathus. Statistical analysis was restricted to the five most abundant species.

Logging adversely affected almost all species. Clearcuts contained on average about half as many species of amphibians as mature forest sites. Catches of salamanders from plots in mature forest stands were about five times higher than catches in clearcuts.

The authors' discussion section reviews literature indicating that southern Appalachian salamander species, sensitive to moisture and temperature stress, are adversely affected by clearcutting, which degrades forest floor habitat by eliminating shading, reducing leaf litter, increasing soil surface temperature, and reducing soil surface moisture. In addition, due to having small home ranges and being closely tied to their home ranges, salamanders do not disperse away from logged areas. The authors believe that these factors make high mortality after clearcutting likely.

Key Finding: The relatively abundant land salamander Plethodon jordani, an important part of the food chain, disappeared from forest sites in the southern Blue Ridge Mountains after they were clearcut.

Source: Ash, A. N. 1988. Disappearance of salamanders from clearcut plots. The Journal of the Elisha Mitchell Scientific Society 104: 116-122.

The effects of clearcutting on the salamander species Plethodon jordani were studied for four summers after logging in the Nantahala National Forest in the southern Blue Ridge Mountains. This species was selected for study because, as an abundant predator in southern Appalachian forests, it plays an important role in ecosystem energetics.

Four 225 m2 plots were established: two in clearcuts and two in the adjacent oak-hickory forest. Plots were at least 40 m from the edge of the patch in which they were located. Salamanders were toe-clipped to permit individual identification. Searches were done at night, from June through August.

Throughout the study, between 19 and 49 salamanders were caught on forested plots. During the first summer after logging, salamander abundance in clearcuts was 40% of the numbers on forested sites. During the following summers, no salamanders were found in clearcuts. Due to the limited numbers of plots, statistical analysis was not performed.

Key Finding: In the first two years after clearcutting, salamander numbers, including Plethodon jordani, declined to almost zero on all three forest sites studied.

Source: Ash, A. N. 1997. Disappearance and return of plethodontid salamanders to clearcut plots in the southern Blue Ridge Mountains. Conservation Biology 11: 983-989.

The impact of clearcutting on terrestrial salamanders was studied at three sites in the southern Blue Ridge mountains, North Carolina. Clearcuts ranged in size from 10 to 23 ha. Salamander abundance was compared on plots (225 m2 in size) in clearcuts and in adjacent forest. Salamander sampling was conducted before and after clearcutting, with searches done at night.

A total of 1,355 salamanders were captured during the study, mostly of the species Plethodon jordani. Salamander populations decreased dramatically at all three clearcut sites in the first two years after clearcutting. Salamander numbers on clearcuts were 30-50% of populations on adjacent forest sites. By the second year, salamanders were essentially absent from clearcuts. Data on salamander abundance was collected for up to eight years after cutting on one site and 15 years after cutting on a second site.

Key Finding: Adult and juvenile wood frog and spotted salamander capture rates declined along a gradient from closed-canopy forest to recently clearcut habitat.

Key Finding: Juvenile wood frogs, dispersing from breeding pools at the forest edge, preferred to migrate toward closed-canopy forest habitat and away from open habitat.

Source: deMaynadier, P. G. and M. L. Hunter, Jr. 1999. Forest canopy closure and juvenile emigration by pool-breeding amphibians in Maine. Journal of Wildlife Management 63: 441-450.

The authors examined habitat selection by natural populations of wood frogs (Rana sylvatica) and spotted salamanders (Ambystoma maculatum) in three upland, mixed-forest sites in Maine. Sampling was conducted using drift fences along transects that extended 70 m into relatively mature forest and 70 m into an adjacent clearcut. The abundance of both captured adults and migrating juveniles significantly declined across the transect from closed-canopy forest to recently clearcut areas.

The authors also released an experimental population of wood frogs just before metamorphosis into artificial dispersal pools along the forest edge of a 75-m-wide power line right-of-way. Dispersal of the frogs was evaluated using pitfall traps in the adjacent closed canopy, at the edge itself, and in an adjacent open right-of-way. Juvenile frogs showed a strong immigration preference for closed-canopy habitat rather than edge or power line habitat. Highest capture rates were in habitat with dense understory and canopy.

Reduced nesting success: The reproductive success of interior forest bird species decreases in areas fragmented and/or disturbed due to logging or roads. Some species are sensitive to disturbance; others suffer from increased rates of nest parasitism and nest predation.

Key Finding: The density of bald eagle nests in southeast Alaska decreased with proximity to clearcuts.

Source: Gende, S. M., M. F. Willson, B. H. Marston, M. Jacobson and W. P. Smith. 1998. Bald eagle nesting density and success in relation to distance from clearcut logging in southeast Alaska. Biological Conservation 83: 121-126.

The authors studied bald eagle (Haliaeetus leucocephalus) nesting success in the shoreline forests of Chichagof and Catherine Islands in the Tongass National Forest, southeast Alaska. Old-growth Sitka spruce and western hemlock were the dominant tree species. Study areas included forest units clearcut logged between 1970 and 1979 and forest units with little or no clearcutting. The distance from each nest to the nearest clearcut was recorded in 100-m intervals, up to a distance of 500 m. The distance between active nests was also recorded. This information was used as an indirect estimate of the density of active nests along the shoreline. Nesting success was recorded as the number of young per active nest.

Results showed that the distance between active nests increased as their proximity to clearcuts increased, i.e. nest density decreased near clearcuts.

Key Finding: Productivity of nesting bald eagles decreased with proximity to clearcuts.

Source: Anthony, R. G. and F. B. Isaacs. 1989. Characteristics of bald eagle nest sites in Oregon. Journal of Wildlife Management 53: 148-159.

The authors surveyed 201 bald eagle nest sites in the Oregon Cascades between 1979 and 1982. Nests were located in three forest types: Douglas-fir, mixed-conifer, and ponderosa pine. Forest stand characteristics were sampled in a 100- m radius around the nest trees - this approximated the buffer zone recommended for nest trees by the U.S. Fish and Wildlife Service. Transects were laid out in the four cardinal directions around the nest tree, with sample points established every 25 m. Human activity within 1.6 km of the nest tree was also recorded.

The productivity of nests was averaged over 1978 though 1982 and accounted for distinctions between unoccupied nests, occupied nests, and nests where young had successfully fledged. In mixed-conifer and ponderosa pine forests, mean productivity of bald eagle nests was negatively correlated with their proximity to clearcuts and main logging roads, decadence of nest trees, and the level of nonrecreational activities. In Douglas-fir forests, nest productivity was negatively correlated to proximity of clearcuts, but the other variables showed no clear relationship. Recently used nests, compared to old nests in the same territory, were located in areas farther from all types of roads and from recreational facilities.

Key Finding: Three of four forest interior bird species declined in abundance after logging, whether clearcutting or lower intensity logging.

Key Finding: The brown-headed cowbird, a species that parasitizes other birds' nests, increased in abundance after logging.

Source: Baker, M. D. and M. J. Lacki. 1997. Short-term changes in bird communities in response to silvicultural prescriptions. Forest Ecology and Management 96: 27-36.

The impact of logging on bird communities and on individual bird species was investigated in the Daniel Boone National Forest, Kentucky. Sixteen forest stands with four silvicultural treatments were included: no logging, clearcut logging, two-age high-leave logging (with 7m2 basal area left uncut) and two-age low-leave logging (3.5 m2 basal area left). Trees left uncut in the two-age stands were 15.2-40.6 cm dbh. Birds' relative abundance was obtained by using fixed-radius point count surveys. Data were collected before and after logging.

Data for 29 bird species were analyzed. Only bird species with 10 or more detections after logging were included in this analysis. Four species were classified as forest interior species. Of these, two (the ovenbird and the red-eyed vireo) were less abundant in all three treated stands than in unlogged stands. One species (Acadian flycatcher) was less abundant in the low-leave and clearcut stands than in stands with no logging, and showed no significant pattern for high-leave stands. The fourth species, the hooded warbler, was most abundant in low-leave and high-leave stands. The authors attributed this to the hooded warbler's preference for well-developed understory vegetation, which was higher in these stands.

The brown-headed cowbird, a nest parasite, increased in abundance in all logging treatments. The species was not recorded in any of the unlogged stands.

Key Finding: As forest fragmentation increased, nests of all nine bird species studied suffered higher rates of parasitism and predation.

Source: Robinson, S. K., F. R. Thompson III, T. M. Donovan, D. R. Whitehead and J. Faaborg. 1995. Regional forest fragmentation and the nesting success of migratory birds. Science 267: 1987-1990.

The impact of forest fragmentation was studied for nine bird species (eight being neotropical migrants) in the Midwest. More than 5,000 nests on nine study areas were monitored for five years. Mean percent forest cover within a 10-km radius of the center of each site was estimated from forest cover maps. As percent forest cover decreased, nest parasitism by brown-headed cowbirds increased for all species, with statistically significant increases for five of the nine species. Nest predation rates increased for all species as percent forest cover decreased, with three of nine species having significant correlations.

Key Finding: The reproductive success of ovenbirds, a forest interior species, was significantly lower in forest fragments than in continuous forest, partly due to cowbird parasitism of their nests.

Key Finding: The density of breeding male ovenbirds was lower in forest fragments than in continuous forest, with birds avoiding habitat within 100 m of the forest edge.

Source: Porneluzi, P. A. and J. Faaborg. 1999. Season-long fecundity, survival, and viability of ovenbirds in fragmented and unfragmented landscapes. Conservation Biology 13: 1151-1161.

The authors compared the breeding success of ovenbirds (Seiurus aurocapillus) in a fragmented landscape versus an unfragmented landscape. There were seven study sites: three located in large forest patches (greater than 2,000 ha) and four within continuous forest (1.8 million ha). All sites were located in oak-hickory forest in Missouri. Analysis was based on the number of territorial males at each study site. The majority were captured and color-banded. Territories were mapped throughout the breeding season. The locations of all nests were mapped, nests were monitored every three to four days, and nests were inspected for the presence of cowbird eggs or nestlings. A pair of ovenbirds was considered reproductively successful only if it was observed caring for a fledgling out of the nest.

Reproductive success over the entire breeding season was lower in the fragmented forests than in the continuous forest sites. In the fragmented landscape, 72% of nests experienced cowbird parasitism, in contrast to 4% of nests parasitized in the unfragmented landscape. Annual productivity in the fragmented landscape was 0.70 juvenile female ovenbirds per female, compared to 1.47 juvenile females per female in the unfragmented landscape.

The average density of males (number per 10 ha) was also lower in the fragmented landscape (1.61) than on sites in the unfragmented landscape (2.2). Avoidance of forest edges was observed, with territorial males occupying significantly less habitat within 100 m of an edge than habitat greater than 200 m from an edge.

Key Finding: All three species of tanagers studied were sensitive to forest fragmentation, with a declining probability of breeding tanagers occurring at a given site as fragmentation increased.

Source: Rosenberg, K. V., J. D. Lowe and A. A. Dhondt. 1999. Effects of forest fragmentation on breeding tanagers: a continental perspective. Conservation Biology 13: 568-583.

The impact of forest fragmentation on tanagers (Piranga spp.) was evaluated by collecting data from more than 1,000 study sites throughout the United States and Canada. Data were analyzed for three species of tanagers - scarlet, western, and summer tanagers (P. olivacea, P. ludoviciana, and P. rubra). Volunteer participants established census points, which they visited during the season of bird territory establishment as well as during nesting. The presence of nest predators and brown-headed cowbirds was also recorded.

Data were used to construct models to predict the probability of a breeding tanager occurring at a study site. The probability of finding breeding tanagers decreased with increasing fragmentation for all three tanager species. Sensitivity to fragmentation varied geographically and was highest in the Midwest and Atlantic Coast regions.

Key Finding: Nesting success of forest birds decreased within 50 m of forest edges.

Key Finding: In five of six studies, nesting success of forest birds decreased as forest patch size decreased.

Source: Paton, P. W. C. 1994. The effect of edge on avian nest success: how strong is the evidence? Conservation Biology 8: 17-26.

To investigate the decline in neotropical migrant bird populations, the author reviewed research on nesting success and its relationship to habitat fragmentation and artificial edges. The author reanalyzed data from a number of studies in order to be able to compare results from research conducted under varying experimental designs. Study sites included forest as well as shrub-grassland and prairie habitat in North America and Europe, with one study from Central America.

Data from 14 studies using artificial bird nests in forests were reanalyzed. The majority (71%) demonstrated that nest success was lower near forest edges, with nest predation rates greatest at distances within 50 m of an edge. Results on effects further than 50 m from an edge were less conclusive. In addition, information on the influence of the type of edge (abrupt or feathered, for example) was inconclusive.

Of the seven studies that used natural nests, four (57%) demonstrated significantly higher nest predation rates near forest edges. Two of the other studies were located in grasslands and found no effect of edges. A third study, based on exposure days, could not be reanalyzed. Of the five studies on parasitism of natural nests, three (including forest and grassland study sites) demonstrated that cowbird parasitism increased near edges. A fourth indicated a similar trend, but results were not significant. Of the six studies investigating forest patch size and its relation to nest predation rates, five demonstrated that nest success decreased as patch size decreased.

Key Finding: Nest predation rates in southern Appalachian forest fragments increased as fragment size decreased.

Source: Keyser, A. J., G. E. Hill and E. C. Soehren. 1998. Effects of forest fragment size, nest density, and proximity to edge on the risk of predation to ground-nesting passerine birds. Conservation Biology 12: 986-994.

A study of nest predation rates relative to forest fragment size was conducted at Fort McLellan in the southern Appalachians, Alabama. Forest fragments ranged in size from 4 to 849.4 ha and were located in a variety of surrounding habitat types. Artificial ground nests contained two fresh quail eggs and two clay eggs. The former were included to detect large predators and to provide visual and olfactory cues, and the latter were included to obtain a record of claw and tooth marks from small-mouthed predators. Predator species were not identified for this study, and were only categorized as small (including small rodents such as chipmunks) or large (including blue jays, American crows, gray foxes, raccoons, Virginia opossums, and white-tailed deer). Snake predation could not be accounted for by this study design.

The authors laid out 22 linear transects in 10 forest fragments, each with 30 nests 20 m apart. Larger fragments had several transects. To test for clustering effects, nests were also grouped in grid formations rather than in linear transects. Nests were checked seven to eight days after placement.

Total nest predation rates per transect increased as forest fragment size decreased. Large predator activity increased as fragment size decreased, but no difference in small predator activity was detected. Large predator activity also increased among clustered nests, but total predation rates for clustered versus unclustered nests showed no significant difference. There was no difference in predation rate based on distance from edge. The authors concluded that edge types and the adjoining habitat were too variable among the fragments to discern any effects specific to edges.

Key Finding: Artificial nests had higher rates of predation on the edges of forest fragments than in the interior of fragments.

Source: Marini, M. A., S. K. Robinson and E. J. Heske. 1995. Edge effects on nest predation in the Shawnee National Forest, southern Illinois. Biological Conservation 74: 203-213.

The authors studied the effect of edges (forest-farm edges) on nest predation rates in the Shawnee National Forest, Illinois. Five transects (1,000 m long) were established 10-15 m in from the forest edge and five transects were established in the forest interior, at least 300 m from the edge. Artificial nests with quail eggs were placed at thirty 25-m intervals along each transect. Nests were placed on the ground, in shrubs, and in saplings. Predation levels were checked after five, 10, and 15 days. Nests were also placed in groups of five to compare predation on low versus high nest densities. Bird abundances were estimated from singing male counts along each transect through the breeding season. Mid-sized mammal abundance was determined from track censuses. Small mammals were trapped using Sherman live-traps.

Sapling nests had higher levels of predation on forest edge sites than in the forest interior. Ground and shrub nests had higher predation levels as well, but these differences were not statistically significant.

The authors were not able to determine the mechanism by which edges may increase predation. Predation rates were independent of singing bird abundances. Nor did nest density seem to be a factor. The authors note that their methodology did not incorporate possible predation by squirrels.

Get Updates and Alerts

See the latest issue >

Give the Gift That Will Make a Difference: Leader of the Pack

NRDC Gets Top Ratings from the Charity Watchdogs

Charity Navigator awards NRDC its 4-star top rating.
Worth magazine named NRDC one of America's 100 best charities.
NRDC meets the highest standards of the Wise Giving Alliance of the Better Business Bureau.

Donate now >

Related Stories

Q&A: Documentary Filmmaker Ken Burns on National Parks
Ken Burn spoke to OnEarth about his motivation for his new documentary series on America's national parks.
In the Canadian Boreal Forest, a Conservation Ethic at Work
After fighting successfully for years to keep destructive logging, hydropower and mining projects out of their traditional territory, the people of Poplar River are now working to secure permanent protection for their boreal forest homeland.
Shop Smart, Save Forests
Share | |
Find NRDC on