Northern Prairie Wildlife Research Center
In general, adults of all carnivore species we studied, except weasels and female minks, have large home ranges (usually several square kilometers or larger), travel extensively (usually several kilometers or more daily), and use a variety of habitats (Sargeant 1972; Sargeant and Warner 1972; Ewer 1973; Fritzell 1978a, 1978b; Andelt and Gipson 1979; Lampe and Sovada 1981; Greenwood et al. 1985; Arnold and Fritzell 1987a; Novak et al. 1987; Sargeant et al. 1987a). Thus, tracks can be a sensitive indicator of the presence and relative abundance of a carnivore species.
We used the percentage of searched quarter sections in which tracks were found to rate abundance of all carnivore species, except weasels. Ermines and long-tailed weasels are small and lightweight (Banfield 1974) and, therefore, reliable searches for their tracks could not be conducted. The only obtained data on weasels were observation rates, which were insufficient to rate abundance in individual study-areas. Data from track surveys were supplemented with locations of active rearing dens (coyote, red fox), systematic livetrapping (striped skunk), and sightings of individuals (mink). Data on removal of predators provided useful ancillary information for several species.
The Franklin's ground squirrel—the only studied rodent—is a secretive species whose presence in an area may not be readily apparent. However, Franklin's ground squirrels are easily trapped (Choromanski-Norris 1983; Greenwood 1986) and, because they are diurnal, are occasionally seen. We used observation rates to rate abundance of Franklin's ground squirrels. Capture rates were also considered a reliable indicator of abundance but were not available for all study area-years. Moreover, observation rates reflected abundance throughout each study area, whereas capture rates reflected abundance in only the most dense brush and herbaceous habitats. Although observation rates did not increase with capture rates (r = 0.33, 21 df, P = 0.129) in study areas with both data types, Franklin's ground squirrels were captured in only two study area-years where they were not seen (different study areas, one in each area). Franklin's ground squirrels were seen in both areas in another year. Systematic livetrapping provided useful supplementary data on distribution in study areas and removal of predators provided useful ancillary data on abundance in some study areas.
Based on all collected data, descriptions of Franklin's ground squirrel populations by Sowls (1948), and observations of Franklin's ground squirrels elsewhere in the prairie pothole region (A. B. Sargeant, R. J. Greenwood, and M. A. Sovada, personal observations), we assumed only the highest rate of our observations indicated the abundance of the species could be categorized as numerous.
Because of differences in nests and behavior of corvid species, we used different index methods to rate abundance of black-billed magpies and abundance of American crows and common ravens. The average percentages of quarter sections in which they were detected during the road-transect counts were the primary data used to rate abundance of black-billed magpies. Those data were augmented with observation rates and with incidental records of occupied nests. Based on literature (Brown 1957; Shaw 1967) and our findings, we assumed the highest index value we obtained indicated the abundance of the species could be categorized numerous. In study areas where black-billed magpies were not detected during counts along road-transects and in study areas where counts along road-transects had not been made, we categorized the abundance of these birds undetected if they were not seen, scarce if they were seen at less than or equal to 0.005 place/ h, and uncommon if they were seen more often.
The primary data for rating abundance of American crows were the density of occupied nests. The data on density were supplemented with information from the road-transect counts and observation rates.
We estimated numbers of American crow nests in study areas in which no or partial nest searches were conducted from detection rates of American crows in quarter sections during the road-transect counts. The expansions were based on the relation between nest densities and average percentage of quarter sections in which they were detected (r = 0.91, 11 df, P <0.001) in study areas in Canada where both types of data were available. Based on literature (Kalmbach 1937) and on our findings, we assumed that a density of >0.58 occupied nest/km2 was necessary to rate the species numerous. In areas without occupied nests, we rated American crows scarce if they were seen in less than or equal to 0.005 place/in and uncommon if they were seen more often.
Common ravens were not initially included in our surveys because we presumed they would be absent, but our surveys of American crows in all searched study areas inadvertently revealed common abundances of ravens. We used the same criteria as for American crows to rate abundance of common ravens.
We relied primarily on density of occupied nests to rate abundance of raptor species and, therefore, were unable to rate the abundance of raptors in the eight Central Flyway study areas where no systematic searches for nests and less overall field work was conducted. In searched study areas where no occupied nests of hawks were found, we rated each hawk species undetected if it was not seen, scarce if it was seen but was <5% of total sightings, and uncommon if it was >5% of sightings. In searched study areas where no nests of great horned owls were found, we rated the species undetected if it was not seen, scarce if it was seen in <0.005 place/h, and uncommon if it was seen more often. The difference in rating methods between hawks and the great horned owl was necessary because the observation rates of hawks were not species specific. No raptor species was rated common or numerous based on sightings alone.
We treated the 1983 data on nests of buteos in study areas in the prairie of Canada as complete counts, although no systematic searches for raptor nests were conducted during that year. In doing this, we assumed no buteo nests in study areas in the prairie of Canada were overlooked because of the conspicuousness of those nests and their tending adults, the limited wooded habitat available for nesting, and the extensive coverage by study personnel. We did not use the 1983 data on nests to rate abundance of buteos in study areas in the aspen parkland because some occupied nests were probably overlooked, especially in areas with considerable woodlands. After 1983, we assumed we found nearly all occupied nests of buteos in searched study areas because of the thoroughness of the searches. We used the same criteria to rate abundance of each buteo species. Based on the range of values from our and published information, we assumed a density of 0.08 occupied nest/km2 was minimal to rate the abundance of a buteo species common.
No systematic searches for nests of northern harriers were conducted. However, rather than forgo rating the abundance of northern harriers, we assumed the intensity of searches for nests in the stabilized regulations and small unit management studies was comparable among study areas and that at least half of northern harrier nests were found opportunistically during the extensive searches for duck nests and monitoring of predators. We then treated the data on nest density as an index and rated the abundance of northern harriers on the same basis as the buteos, with a minimal index value of 0.08 occupied nest/km2 to rate the species common. Although the proportion of located nests by which we rated abundance was smaller for northern harriers than for other buteos, we believe the ratings are comparable with those for the buteos because northern harriers often cluster their nests (Bildstein and Gollop 1988). We assumed that high populations of northern harriers include more breeding pairs than high populations of the highly territorial buteos. We treated the data on nest density in study areas in the prairie and in the aspen parkland as equally complete because northern harriers seldom nest in woods (Stewart 1975; Bildstein and Gollop 1988).
With the same rationale we used for treatment of data about buteos, we treated the number of nests of great horned owls in study areas in the prairie of Canada in 1983 as complete counts and did not use such data from the aspen parkland in 1983 to rate the abundance of great horned owls. Because searches for raptor nests were conducted in late May and early June and great horned owls begin nesting in late February, about 6 weeks before other surveyed raptor species (Stewart 1975), nests of great horned owls were vulnerable to destruction (e.g., from weather or predators) for longer periods than nests of other raptors. To rate the abundance of great horned owls, we concluded 0.04 occupied nest/km2 was minimal to rate the species common. However, nest densities of the great horned owl, as those of northern harriers, are indices.