Northern Prairie Wildlife Research Center
Habitat and History. The described geographic range of the Franklin's ground squirrel includes the eastern and northern portions of the prairie pothole region and excludes Montana, most of Alberta except the east-central portion, and southwest Saskatchewan (Banfield 1974; Hall 1981). Franklin's ground squirrels prefer bush and dense grass and herbaceous cover; they are seldom in habitats with sparse or short vegetation (Sowls 1948; Jones et al. 1983; Choromanski-Norris et al. 1989). Little is known about the abundance of this species before settlement or about population trends after settlement. However, because of its specific habitat requirements, Franklin's ground squirrels probably were scarce or absent throughout the prairie before settlement, except in locations with brush, and probably were numerous in the aspen parkland, except in western Alberta.
Agricultural development in the prairie pothole region has had both positive and negative effects on the distribution and abundance of Franklin's ground squirrels. Expansion of the aspen parkland created new habitat, whereas conversion of wooded and brush areas to cropland and intensive grazing by livestock reduced habitat. The establishment of farmsteads and tree plantings in the prairie created small parcels of suitable habitat and probably aided range expansion by this species. Banfield (1941) often saw Franklin's ground squirrels in the aspen parkland of central Saskatchewan in 1939. Soper (1946:141) found them "fairly common to abundant" in southwestern Manitoba during 1927-44 but "scarce or absent on the treeless prairies west of the Turtle Mountains."
Population Structure. Adult Franklin's ground squirrels tend to be asocial although numerous individuals with overlapping home ranges may inhabit the same area (Choromanski-Norris 1983). Population densities in preferred habitat may fluctuate greatly (Sowls 1948); causes of fluctuations are largely undetermined. Densities of Franklin's ground squirrels are usually <2.5 adults/ha, although up to 12 adults and juveniles/ha have been reported (Sowls 1955; Murie 1973). In North Dakota, biweekly home ranges of Franklin's ground squirrels in dense grass and herbaceous cover averaged 7.4 ha for adult males and 2.8 ha for adult females (Choromanski-Norris et al. 1989).
Distribution and Abundance. Franklin's ground squirrels were present in all but three study areas in Canada and in all but six study areas in the United States (Fig. 13). However, they were not caught or seen each year in each of those areas (Appendix Tables 10 and 11). Nearly all study areas in which they were undetected were along the western and southern edge of the region (Fig. 13). Populations were low in nearly all study areas where they were detected. The percentage of study areas where Franklin's ground squirrels were common or more abundant did not differ between Canada and the United states (Appendix Table 6). In Canada, Franklin's ground squirrels were numerous in one study area (6%), common in two (13%), and uncommon or less abundant in all others. They were uncommon or less abundant in all study areas in the United States. The capture and sightings of Franklin's ground squirrels in two study areas (Ceylon, Shamrock) represent small range extensions by this species (Beck 1958; Hall 1981).
The highest populations of Franklin's ground squirrels were in aspen parkland in the north-central portion of the prairie pothole region (Fig. 13). All three study areas where the species was common or numerous were in that portion of the region. The percentage of study areas where Franklin's ground squirrels were common or more abundant was greater in the aspen parkland than in the prairie (Appendix Table 7).
Populations of Franklin's ground squirrels were low in most study areas and individuals were patchily distributed. Differences in local abundance were large. For example, populations were highest and distribution was among the most uniform in one study area (Hanley; captures in 70-90% of sample units each year [Appendix Table 11]). However, in a nearby study area (Craik-65 km south of Hanley), Franklin's ground squirrels were not caught and were seen at only three places during the 2 study years. Although the latter area was more intensively cultivated than the former (Table 1), it contained many brush sites that seemed suitable for Franklin's ground squirrels. In the study area where the population was highest (Earl Grey), all captures and most observations of Franklin's ground squirrels occurred in the western 30% of the area. The population was much higher in those 7.7 km2 than in any other comparably sized part of any study area; the population in the remainder of that area was low. In most other study areas, the captures of Franklin's ground squirrels indicated the species was restricted to small scattered habitat units. They were caught in ≥40% of the sample units in only four study areas (Appendix Table 11).
Further evidence of generally low populations of Franklin's ground squirrels was provided by the data on removal of predators from three study areas (Eldridge, Fredonia, Lake Park). Upland where trapping was conducted had dense grass and herbaceous cover with scattered patches of brush that is favorable habitat for Franklin's ground squirrels. Trapping of Franklin's ground squirrels was conducted continuously throughout May and June each year in each of the sites. Twenty-five Franklin's ground squirrels were removed from the three sites in 1987 during 1,654 trap-days and 11 in 1988 during 3,243 trap-days.
Fig. 13. Observation rates of Franklin's ground squirrels by study area in the prairie pothole region (places seen per 1,000 h) during May-early July. Results from each study area for >=1 year, 1983-88 (Appendix Table 10); results from areas studied in >1 year were averaged.