A geographical information system or GIS is a computer mapping system that allows researchers or managers to examine various layers of landscape simultaneously. By examining various landscape features, biologists can determine why a species occurs in a specific location.
Gray wolves lend themselves well to examining of their habitat selection using GIS. Wolf packs occupy fairly discrete areas that are maintained as territories, and represents the breeding potential of a wolf population. In the Great Lakes region wolves normally occupy territories that cover 20 to 120 square miles. By discerning the characteristics of suitable pack habitat (breeding habitat), we can determine the extent of area that wolves can occupy, and the size of a wolf population that an area can support.
GIS was used recently to determine the type of landscape features that packs occupy in Wisconsin and the adjacent states of Michigan, and Minnesota (Mladenoff et al 1995, 1999). Additionally, work was done to determine how many wolves could occur in Wisconsin and Michigan (Mladenoff et al 1997).
Various landscape features were initially examined in 14 wolf territories that were monitored by the Wisconsin DNR using radio-collared wolves during 1980-1992. These known territories were compared to 14 random areas the size of wolf territories scattered across northern Wisconsin. Wolf territories were also compared to the overall landscape of northern Wisconsin. Landscape features that were examined included human population density, prey (deer) density, road density, land cover, land ownership, and several spatial indices. An additional 23 new packs were examined in an update of the analysis (Mladenoff et al. 1999).
Table C1 illustrates some of the important features of wolf habitat in Wisconsin. In general the average wolf territory contained no urban land, very little farmland, and was 93% forest. Nearly 30% of an average territory was in wetlands, especially conifer swamps and bogs, compared to only 16% overall for northern Wisconsin.
Average Values for characteristics of wolf pack habitat versus overall Northern Wisconsin Study Area
|Wolf Pack Habitat
|Wolf Pack Habitat
90% Cut-off Level
|Agricultural and open land||93%||--||73%|
|Marsh or bog||4%||--||2%|
|Private industrial forest||10%||--||5%|
|Other private lands||21%||<50%||66%|
|Roads Density||0.4 mi/mi2||<1.0 mi/mi2||1.1 mi/mi2|
|Human Density||4.0 persons/mi2||<10.8 persons/mi2||11.3 persons/mi2|
|Deer Density||22.2 deer/mi2||--||21.3 deer/mi2|
Wolf territories consisted mainly of public and industrial forest land (80%), even though these areas cover only about 1/3 of northern Wisconsin. Wolf pack areas had about 1/3 the road density and human population density of northern Wisconsin in general.
Road density was the best predictor of suitable wolf habitat, as had been found by Thiel (1985) and others. Areas that contain less than 0.7 miles of road per square mile have a 50% chance or greater of being settled by wolf packs if adequate space and prey are available. Blocks of land with less or equal to 0.7 miles/mile2 was considered suitable wolf habitat for management purposed. Land with more than 1 mile of road/mile2 is least suitable and has less that 10% chance of being settled by wolf packs. An update of the GIS analysis confirms that road densities continue to be good predictors of suitable habitat in Wisconsin (Mladenoff et al. 1999).
Although road density is an important indicator of good wolf habitat, wolves do not have an aversion to roads. Wolves readily travel down roads for hunting and dispersing, especially dirt and gravel roads. The reason road density is important to wolf habitat, is because higher road densities equate to higher risks of vehicle collisions or illegal kills. In recent years vehicle collisions have become almost as high a mortality factor as illegal killing in Wisconsin. During an 8 month period in 1994-95, 5 wolves died in central Wisconsin due to vehicle collisions.
Area of potential wolf habitat in northern Wisconsin are illustrated in Table C2. A total of 5,739 square miles have greater than 50% probability of being settled by wolf packs and are listed as primary wolf habitat in Table C2. The majority of the primary (71%) occurs on public land or industrial forest land. Land that has a 10 to 50% probability of being settled by wolf packs is listed as secondary wolf habitat and covers 4,704 square miles; slightly over half the secondary habitat occurs on private land (Table C2). About 12,393 square miles of northern Wisconsin appears to be poorly suited as wolf habitat, and most unsuitable habitat occurs on private land. Some of the areas of less suitable habitat may be occupied by wolf packs if these areas occur close to areas of suitable habitat. Landscapes that are not likely to be settled by wolf packs, may still have potential for dispersing wolves, especially in forested habitats near existing packs.
Distribution of potential wolf habitat across Northern Wisconsin Habitat Potential is classified by the probability of being occupied by wolf packs
|AREA IN SQUARE MILES|
|Wolf Habitat Classes||Road Density||State||County||Federal||Tribal||Industrial Forest||Private|
|Primary Wolf Habitat
(>50% prob. class)
|Secondary Wolf Habitat
(>10% to <50% prob. class
|Unsuitable Wolf Habitat
(<10% prob. area)
|Total Land Area||1255||3179||2280||552||1171||14,399|
The initial analysis we conducted on potential wolf habitat examined about 23,000 square miles of northern Wisconsin, but did not examine land in central Wisconsin (Mladenoff et al. 1995). In fall 1994 a wolf pack was verified in central Wisconsin, therefore GIS analysis was conducted for the remainder of Wisconsin in spring 1996. A small area of favorable wolf habitat was identified in central Wisconsin (207 square miles) and included the three wolf territories located in the region in 1996. No other sizeable areas of primary or secondary potential habitat occur in the state, but a few small scattered parcels of secondary habitat exist in central and western Wisconsin. The chance of wolves settling into these small parcels is remote, but these areas may be used by dispersing wolves.
The potential wolf population for Wisconsin and Michigan were determined by Mladenoff et al. (1997) using two methods. A habitat based estimate used the average territory size (69 mi2) average sized pack (4.1 wolves), average space between territories (37%), and assumed 15% loners in the population within areas of primary wolf habitat. A wolf-prey based estimation developed by Fuller (1989) was also used to estimate the potential wolf population within primary wolf habitat, based on abundance of deer.
Table C3 illustrates estimated potential wolf population of 380 and 462 by the two methods. The habitat area based estimate is probably the more reliable projection of the potential population, because it has a more narrow confidence interval, and the prey based projection includes estimates of wolf densities that are higher than any mainland densities reported for wolves in the Great Lakes region. Therefore a reasonable estimation would be a potential wolf population of 300-500 wolves in northern Wisconsin, and 600-1000 wolves in Michigan.
Potential Wolf Population for Wisconsin and Michigan as Determined by Two Methods
|Habitat Area Model||380||324-461||751||641-911|
|Prey Based Model||462||262-662||969||581-1357|
The populations projections made by Mladenoff et al (1997) includes only potential habitat in northern Wisconsin. Based on the size of suitable habitat and wolf densities in other areas of Wisconsin, central Wisconsin could support an additional 20-40 wolves. More research is necessary to better assess habitat and wolf population potential in central Wisconsin.
Figure C1 shows the statewide potential habitat as calculated following the study. It shows 5,812 square miles of primary habitat and 5,015 square miles of secondary habitat in Wisconsin. Our potential wolf population was based on full occupancy of primary habitat, but if secondary habitat were also fully occupied, the potential wolf population could be 50% higher or more. Behavioral adaptions by wolves and greater acceptance by humans could allow for a considerably higher population. Conversely, if wolves are less accepted by people, and are unable to fully occupy even primary habitat, then the potential wolf population could be lower.
These results suggest that Wisconsin and Michigan could support far more than the goal of 100 for both states for delisting as neither endangered nor threatened. The current (1999) population for both states of about 370 wolves, already far exceeds the goal. The GIS results of delineating suitable habitat and potential populations will be very useful for future management planning for the Great Lakes States. The GIS data will provide an important bench mark for evaluating the success of wolf recovery in the Great Lakes region.