Make your own free website on
Wolf History, Conservation, Ecology and Behavior
Wolfology Item #  839
v.26, n4 (1998), 767-775

Click on the journal title (above) for information on acquiring the complete article
Extent and Connectivity of Habitat for Wolves in Eastern North America
Daniel J. Harrison & Theodore G. Chapin
Editor's Comment
In the years since Harrison and Chapin wrote this essay, wolf reintroduction in the Northeast has become increasingly problematic. Some wolf advocates hinge their hopes on the possibility of natural recolonization of this region by wolves from Canada. The results of this study seem to indicate that recolonization is also highly problematic, considering the barriers faced by Canadian wolves that might be predisposed to immigrate to the Northeast, not to mention the possibility of genetic swamping by coyotes if there is an insufficient number of recolonizing wolves -- JM
The eastern timber wolf (Canis lupus lycaon) once occupied the entire northeastern United States, but was extirpated from New England, New York, and extreme southeastern Canada by 1900 through direct human persecution, habitat alteration, and human-induced reductions in obligate prey species. By 1960, the only remaining population of wolves within the coterminous 48 United States occurred in Minnesota, and that population was subsequently protected in 1974 by the Endangered Species Act 16 U.S.C. 1531-1544).
The primary goal identified in the recovery plan for eastern timber "to maintain and reestablish viable as much of its former range as possible" (U.S. Fish and Wildl. Serv. 1992). Furthermore, the plan established that the eastern timber wolf will be considered officially recovered when the survival of the wolf in Minnesota is assured, and at least 1 viable population exists outside Minnesota and Isle Royale in the coterminous 48 United States. The criteria used to define a viable population were: >200 wolves in a population, or >100 wolves occurring in Wisconsin or Michigan within 160 km of the Minnesota population over a 5-year period. The population of wolves in Minnesota exceeded recovery goals by 1989, and wolves in Wisconsin and in Michigan's Upper Peninsula numbered 99 and 116 individuals, respectively, in 1996. The combined number of wolves in Wisconsin and Michigan has exceeded 100 wolves since 1994. Thus, the official recovery objective for the eastern timber wolf was achieved in 1998....
The 1992 recovery plan also identified 24,287 km2 in New York and 35,751 km2 in Maine as areas that warranted further consideration as potential habitat for wolves, but those areas were not quantified and mapped, based on habitat criteria identified in the plan, and the connectivity of potential habitat with occupied habitat in southeastern Canada was not considered. Human access to wolf populations is a primary consideration for evaluating habitat suitability. Researchers from the Great Lakes region have reported that resident wolves do not persist in areas where road densities exceed 0.58 km/km2 (Thiel 1985, Jensen et al. 1986). However, Mech et al (1988) reported that wolves can persist in areas with road densities as high as 0.73 km/km2, if these areas are adjacent to habitats with less human access. Similarly, Fuller et al. (1992) reported that 85% of packs and 80% of single wolves identified in Minnesota occurred in townships with <0.70 km roads/km2 and <4 humans/km2. Those thresholds for densities of roads and humans were subsequently adopted in the recovery plan as maximum levels of human presence that are compatible with viable habitat for wolves in the coterminous United States....
Quantification of potential habitat for the wolf east of the Lake Superior basin is particularly relevant because of a recent taxonomic reclassification of Canis lupus proposed by Nowak and Federoff (1996). This reclassification...placed the wolves that currently occupy Minnesota, Wisconsin, and Michigan within the same subspecies (C.l. nubilus) as the Rocky Mountain timber wolf. If formally recognized by U.S. Fish and Wildlife Service, this reclassification would change the status of the eastern timber wolf to extirpated from its former range in the United States and extant only in southeastern Canada. If this reclassification were accepted, recovery of eastern timber wolves would require natural recolonization or reintroduction from extant stocks in southeastern Canada, which themselves are being considered for the category of "vulnerable" or "threatened" by the Canadian government in some parts of their ranges.
....Provided there is a sufficiently large source population and no barriers to movement, wolves are capable of colonizing distant habitats located hundreds of kilometers from a population source. Some dispersing wolves have crossed 4-lane highways, areas of high road density, and extensive areas unoccupied by wolves, while circumventing large urban areas.
Although wolves are physically capable of dispersing the relatively short distances from extant populations in Canada to potential habitat in Maine and New York, there are potential physical and habitat barriers that may preclude immigration of a sufficient number of wolves to promote population establishment south of the United States-Canada border. These potential barriers include the Saint Lawrence Seaway, which is situated 75 km from Maine and forms the boundary between New York and Ontario, as well as extensive areas of unforested agricultural land, areas of high human density, and areas with high road densities. Human activities may create barriers that slow or impede range expansion of wolves. For example, several individual wolves that dispersed south of the city of Minneapolis-St. Paul, Minnesota were documented as having been killed while crossing highways or shot when mistaken for coyotes (Canis latrans), a species that is legally harvested throughout the coterminous United States and southern Canada. There are considerably higher densities of humans, agricultural areas, and roads in areas of southeastern Canda, south of occupied wolf range, than in potential habitats identified for further study in the northeastern United States....
....Decisions among alternative approaches (e.g., reintroduction, natural recovery) to restore wolves to the northeastern United States will require understanding of habitat connectivity, barriers to natural recolonization, and the potential for population exchange between areas of suitable habitat.
We present an analysis that quantifies and maps the extent, distribution, and connectivity of potential habitat for wolves in the northeastern United States that is forested land cover, and below thresholds of 0.70 km roads/km2 and 4 humans/km2 (Harrison and Chapin 1997). To provide insight into the potential for natural reestablishment of wolves through emigration from extant populations, we mapped potential dispersal corridors between wolf populations in southeastern Canada and potential habitat in the northeastern United States. Further, we used published wolf densities from areas throughout the North American range of the timber wolf, in conjunction with our estimates of the extent of potentially suitable habitat, to evaluate whether areas of contiguous habitat in the northeastern United States are sufficient to satisfy the population viability criteria for the eastern timber wolf that were established in the recovery plan.
....We defined potential core wolf habitat in Canada and the United States as areas in a forested cover type with <4 humans/km2. We defined potential wolf dispersal habitat as areas in either forested or mixed forest-cropland cover types with <10 humans/km2. In the United States, we further restricted both potential core and dispersal wolf habitats to include areas with <0.7 km roads/km2....Our human and road density criteria for potential core habitat were based on documented thresholds of wolf occupancy in Minnesota. Based on accounts that dispersing wolves sometimes move through farmland and other areas with higher levels of human activity than are typically occupied by resident wolves, we expanded the definition of potential dispersal habitat to include 1 km2 cells classified as a mosaic of agricultural land and forest....
We used the range of wolf densities (1.0-4.0 wolves/100 km2) recorded in the literature from across North America (Ballard et al. 1987), eliminating high values from island populations, to estimate the minimum and maximum number of wolves that could likely occur in contiguous areas of potential core habitat within the northeastern United States. We excluded potential core habitat occurring in Canada, regardless of its proximity to habitat within the United States, when calculating the minimum population sizes of wolves that might be supported. We evaluated the viability of potential populations within contiguous areas by comparing the lower range of expected wolf density with the viability criteria stated in the recovery plan of "an isolated eastern timber wolf population" averagin at least 1 wolf per 129 km2 (i.e. 200 wolves) distributed within a minimum area of at least 25,906 continguous square kilometers (U.S. Fish and Wildl. Serv. 1992, 25-26).
Substantial contiguous areas of forested habitat with road and human densities below the thresholds identified by Fuller et al (1992) occur within Maine and New Hampshire (48,787 km2). These areas meet the criteria for defining potential core habitat, and exceed by 36% the habitat identified for future study (35,571 km2) in the recovery plan. Based on areas with road densities [less than or equal to] 0.38 km/km2 ... Mladenoff and Sickley (1998) estimated 52,804 km2 of potential wolf habitat in Maine and New Hampshire.
The recovery plan considered potential wolf habitat in northwestern and eastern Maine as discreet areas; however, our analysis suggests that potential habitat is contiguous throughout northern, western, and eastern Maine, and extends well into northern New Hampshire. Contiguous core habitat in Maine and New Hampshire could likely support 488-1,951 wolves; the lower extent of this range is 2.4 times higher than the minimum population viability criterion defined in the recovery plan. Similarly, Mladenoff and Sickley (1998) estimated that Maine and New Hampshire could support 784-1,575 wolves. Viability of potential habitat in Maine and New Hampshire would be further enhanced by connectivity with large areas (51,282 km2) of currently unoccupied, but potentially suitable, habitat in Quebec south of the Saint Lawrence River. Although our habitat assessment did not include New Brunswick, low human populations and extensive forest in the northern part of that province comprise additional suitable habitat that is contiguous with the potential habitat we identified in Maine and Quebec.
The Adirondack Mountains region of northern New York also represents a large, contiguous area (14,618 km2) of land meeting our criteria as potential core habitat for wolves. Our estimate of suitable habitat in the Adirondack region in 11.6% lower than estimates by Mladenoff and Sickley (1998)....Potentially suitable wolf habitat in northern New York is apparently isolated from other suitable habitat. Significant geographic barriers (e.g., Saint Lawrence River, Lake Champlain), as well as expansive areas that do not meet our criteria of either core or dispersal habitat (e.g., extensive agriculture in Saint Lawrence River valley), likely isolate potentially suitable areas in New York from extant populations in Canada and potential wolf populations in Maine and New Hampshire. New York would likely support 140-584 wolves, but our estimate of potential habitat equates to only 56% of the area requirements for maintaining long-term viability of an isolated wolf population, according to the criteria established in the recovery plan. New York state lacks a significant moose (Alces alces) population; thus potential densities of wolves there may be lower than in other regions of eastern North America where sympatric populations of moose, white-tailed deer (Odocoileus virginianus) and beaver (Castor canadensis) occur....
There is limited and widely scattered potential core (2,470 km2) and dispersal (1,430 km2) habitat for wolves in Vermont. Furthermore, Lake Champlain in Vermont and Lake George in New York have a north-south orientation that may hinder east-west movements of wolves between large areas of potential habitat in New York and habitat in other northeastern states. Therefore, it is unlikely that Vermont would either support substantial numbers of resident wolves or serve as an effective dispersal corridor linking a potential future wolf population in Maine and New Hampshire with a potential population in New York....Potential core and dispersal habitat for wolves in Massachusetts, Connecticut, Rhode Island, and southern New York is either absent or too isolated to contribute significantly to restoration or recolonization of the eastern timber wolf.
Our analyses...resulted in lower estimates of suitable habitat...than Mladenoff and Sickley (1998). Their model was based on previous research in Wisconsin, which suggested that density of roads was the primary predictor of whether colonizing wolf packs inhabited an area. Thus, density of roads, most of which were paved, was the sole variable used by Mladenoff and Sickley to map the extent of potential habitat suitability in the northeastern United States. Road density is considered an index of human activity and of the potential for humans to kill wolves (Fuller et al 1992), and probably correlates with human population density in most areas of North America. Within potential wolf habitat in northern Maine, however, extensive areas that are uninhabited by humans on a permanent basis occur within industrially owned forestlands. Densities of unimproved roads, which are traveled extensively by recreationists, sometimes exceed 1km/km2 on these managed forests. Therefore, we used the higher road-density thresholds (<0.7 km/km2) identified for long-established wolf populations in Minnesota. The classification of roads we used included many unimproved forest roads passable by a 2-wheel-drive vehicle during periods without snow cover. Also, our analyses treated human density separately from roads, using the thresholds (<4 humans/km2) identified for Minnesota (Fuller et al. 1992). Our analyses excluded areas that were not dominated by forest landcover from estimates of potential wolf habitat because most wolves inhabiting the Great Lakes Region of the United States and Canada occur in heavily forested regions. The inclusion of forest landcover as a criterion for defining potential habitat likely resulted in the exclusion of some areas that were considered suitable, based solely on road densities, by Mladenoff and Sickley.
The shortest, straight-line distance from potential core habitat in Maine to the nearest occupied wolf range in Quebec is approximately 70 km; the distance to the long-established wolf population in Laurentides Provincial Park is approximately 140 km. Further, the distance from potential core habitat in New York to occupied wolf range in southern Ontario is approximately 230 km. Thus, if suitable dispersal corridors exist, potential habitat for wolves in the northeastern United States is well within dispersal capability of extant wolf populations.
Two potential corridors may link wolf populations occurring in Quebec, north of the Saint Lawrence River, with potential habitat in Maine and New Hampshire. One corridor occurs upstream from Quebec City in an area where occasional, intermittent ice cover may provide wolves opportunities to cross the river. Another potential corridor occurs near the mouth of the Saint Lawrence River, downstream from Quebec City, where home ranges of radiocollared wolves have occurred just north of the river, and where other mammals (e.g., white-tailed deer) have been observed to successfully cross the river. A verified wolf and a second large wolf-like canid were recently killed in Maine. These animals could have emigrated from extant populations of wolves in southeastern Canada.
Agencies and organizations interested in aiding the recovery of wolves in the northeastern United States could benefit from collaboration and exchange of information with scientists conducting ongoing research on wolves in southern Quebec. Information on population density and movement patterns of wolves in southern Quebec may provide useful information for estimating potential numbers of wolves that might disperse to Maine and New Hampshire. Further, the permeability of the Saint Lawrence River as a filter to dispersal needs to be better documented to evaluate the potential for natural recolonization of wolves to Maine and New Hampshire. Some dispersing coyotes in Maine successfully crossed a large river, and 1 dispersing juvenile swam to a coastal island, suggesting that the Saint Lawrence River may serve as a filter rather than a barrier to wolf dispersal. However, maintenance of an active shipping channel and unconsolidated ice in the Saint Lawrence River during most of the winter, coupled with the presence of dense human development and 4-lane highways parallel to the river, could preclude successful dispersal of a significant number of wolves from Quebec....
Given the relative isolation of potential wolf habitat in New York, natural recolonization of potential habitat is unlikely. Further, the success of potential reintroduction efforts for wolves in the Adirondack region of New York would be uncertain because the estimated suitable habitat is less than the area considered in the recovery plan as required to sustain an isolated population of wolves.
If numbers of dispersing wolves moving from extant populations to potential habitat are insufficient to provide opportunities for dispersers to pair with conspecifics of the opposite sex, then substantial hybridization between dispersing wolves and resident coyotes may occur. Roy et al. (1994) presented compelling genetic evidence suggesting that substantial hybridization occurs between coyotes and wolves along the southern edge of wolf range in southeastern Canada. Genetic swamping by coyotes could occur when occasional wolves disperse into the northeastern United States; therefore, strategies promoting slow, natural recolonization of wolves to the northeastern United States should consider potential genetic consequences of hybridization with coyotes.
Although large contiguous areas in Maine and New Hampshire meet the criteria for habitat suitability established in the recovery plan, well stratified, scientific surveys of public attitudes towards wolves in the northeastern United States have not been conducted. Our habitat criteria are based on factors that influence the extent of human contact with wolves and, presumably, the potential for human-induced mortality of wolves. Thus, our analyses assume that human attitudes towards wolves in the northeastern United States are similar to attitudes of humans towards wolves in the Lake Superior basin. Wolves are not intolerant of humans; however, some humans are intolerant of wolves. For example, in some regions of Europe and Asia where human attitudes and cultures differ significantly from the United States, wolves persist in areas with high human density. Thus, prior to establishing specific recovery strategies for wolf restoration, significant public education and involvement would be required.
Figure 1: Distribution of occupied and potential habitat for eastern timber wolves in northeastern North America
Table 1: Potential habitat (km2) for wolves estimated for 7 northeastern states based on thresholds for road density and human density...and extent of forest land cover.
References: 29