Phillip G. deMaynadier

Malcolm L. Hunter, Jr.

Department of Wildlife Ecology

Nutting Hall, University of Maine

Orono. Maine 04469

Logging is a dominant land use practice in landscapes, such as those in northern New England that still retain a large proportion of their original forest cover. Unlike timber removal, which may produce only a temporary change in habitat quality for wildlife, the construction of forest roads to access harvest sites constitutes a relatively permanent change in habitat structure. Because the construction of forest roads involves a major investment, the incentive for long-term maintenance to provide future access is high. Across much of the eastern United States a relatively small proportion of the land base is publicly owned, and very few areas have been designated as roadless wilderness. Private landholders in New England, for example, comprise 93% of the region's forest ownership and maintain an extensive network of primary and secondary logging roads. Even on the small percentage of ownerships where conservation of wildlife is an explicit management objective (e.g., national forests, national wildlife refuges, state and national parks, private nature preserves. etc.), secondary forest road development is often the most common form of natural area disturbance.

Besides constituting a direct and often permanent loss of wildlife habitat, roads can also impact surrounding animal populations as a source of mortality, particularly where roads cross important dispersal and migration routes. In Germany, Kuhn (1987) demonstrated that 24 to 40 cars on a road per hour are sufficient to kill 50% of migrating common toads. Fortunately, on most logging roads traffic is too light, especially at night, to warrant concern about direct mortality. A more important concern is that a linear strip of gravel or dirt under an open canopy can serve as a physical or psychological barrier to animal movements (Stamps et al. 1987), especially for less mobile species. Formerly continuous populations that become reduced in size and isolated by barriers are more prone to higher extinction rates. Further, the barrier effect of less trafficked, unpaved roads has been convincingly demonstrated for small mammals and invertebrates. In one study in Kansas, a lightly trafficked road (10-20 vehicles per day) strongly inhibited the movement of prairie moles and cotton rats.

Most studies of the potential barrier impact of roads has focused on small mammal populations, but the consequences of barriers on amphibians and reptiles that require annual migrations among discrete sites for breeding, foraging, and overwintering may be more significant. This study examines the effects of logging on amphibian movements in a heavily forested region of Maine.

Over half of all anuran captures occurred in roadside habitats, and no barrier effects were detected for the species we studied, suggesting that forest roads are not avoided and, in fact, roadside habitat may even be selected by some species and life stages.

Salamanders

In contrast to the observations made for anurans, forest road habitats appeared less suitable for salamanders. Salamanders were generally less abundant in roadside habitats, particularly at the wider, industrial road site where 2.3 times as many animals were found in forest control locations. The dichotomy between salamander and frogs and toads also existed for road-crossing rates. Wyman (1991) reported average mortality rates of 50.3% to 100% for salamanders (red-spotted newts and redback salamanders, respectively) attempting to cross a paved rural road in New York, and suggested that individual salamanders generally fare worse than frogs because they are slower moving and often "freeze" in response to approaching vehicles. Our data suggest that salamanders may also be less inclined to attempt crossing wide road barriers. The major road appeared to serve as a barrier to movements for two of the species examined (red-spotted newts and redback salamanders). The fact that no significant difference was detected between the relative number of salamanders captured in forest interior locations and the number of salamanders approaching the edge of the major road suggests that the lowered number of observed crossings is due either to a reluctance of salamanders to cross a large logging road, or to decreased success for those that made an attempt.

Salamander abundance and crossing rates were depressed at the wider logging road. Width may affect road-crossing rates if salamanders perceive a wide road as a psychological barrier - a phenomenon described for some tropical forest interior bird species, which, although capable, are not willing to cross relatively narrow water gaps, even of a few hundred meters (Willis 1974, Karr 1982). Road width could also be important if increased exposure to edge-adapted amphibian predators (e.g., skunks, raccoons, hawks, snakes) or exposure to a harsh substrate or ambient micro-climate leads to mortality or to fewer crossing attempts.

Buech and Gerdes (U.S. Forest Service, Forestry Sciences Lab, Grand Rapids, Minn., unpublished data) found hundreds of dead blue-spotted salamanders along a forest road in Minnesota's Superior National Forest; the cause of mortality appeared to be desiccation from exposure to calcium chloride, not vehicles. In addition, predation on road-crossing amphibians can be intense when predators congregate during the short migratory season (Langton 1989).

Many animal species make at least four different types of movements including:

(1) home-range (i.e., regular movements within a fixed area),

(2) migratory (e.g., seasonal movements between breeding and non breeding habitats),

(3) dispersal (i.e., movements by juveniles away from the place of birth), and

(4) geographic range shifts (i.e., long-term movements by organisms in response to climatic or environmental changes.

Amphibians are particularly suitable to study the relationship between the first three types of movements and the effects of forest fragmentation because many species exhibit clear examples of each movement type. Observations of amphibian movements at the major road suggest that conclusions about the potential role of logging roads in fragmenting landscapes depends not only on species differences but also on the animal's age and the type of movement it is making.

Conclusions

The total area of land converted to road surface and shoulder clearance for permanent logging roads can represent a significant loss of former habitat in densely roaded regions. In this study, six acres of forest habitat were lost for every linear mile of road.  Stewards of natural areas and managed forests who are concerned about the potential impacts of secondary roads on sensitive species should construct fewer and narrower roads with little or no edge clearance. Large-crowned trees, particularly mature hardwoods, left in place at the road's edge will help to provide shade and litter to the surface of the road and will minimize forest opening disturbance. Avoidance of desiccating or toxic road-surface binding chemicals will serve to lessen the contrast between the microclimate of the road’s surface and that of the surrounding forest floor.