Northern Prairie Wildlife Research Center
We surveyed breeding birds using standardized procedures for fixed-distance circular point counts; 5-minute counts within a 100 m radius, wherein all birds were identified by sight and/or sound and counted (Ralph et al. 1993, 1995). We recorded whether birds occurred within 0-50 m, 51-100 m, over 100 m from the point, and flying overhead. In addition, we indicated whether birds were identified in 0-3 or 3-5 minutes from initiation of the count. Only trained observers proficient at identification of birds by sight and sound performed counts. We started counts within one-half hour of local sunrise and ended no later than 0930. Counts were not conducted in rain, high winds, or other inclement weather that could affect our detection of birds. We calculated relative abundance for each species by dividing the number of individuals of a species within 100-m radius points by the total number of all birds of all species observed within the points.
We distributed point locations throughout all habitat types at TAPR (Fig. 2). To facilitate future location of points for NPS personnel, we placed points at least 500 m apart on roads and Trails (Fig. 3). Most of the roads on TAPR were small 4-wheel trails (2-tracks) and the one gravel road was used only by tour buses during a few designated times. Fifty points were distributed in proportion to habitat type present at TAPR: 23 in burned and grazed prairie, one in unburned and ungrazed prairie (Schoolhouse Prairie), five in brome field, nine along Fox Creek gallery forest, eight along Palmer Creek gallery forest, one in the cedar grove, and three in mixed riparian/prairie areas (areas near a creek, spring, or pond). We recorded each point location using Global Positioning Systems (GPS) (Appendix A). Point counts were conducted 23-26 May, 11-13 June, and 23-27 June 1998, and 8-9 June 1999. We compared number of species at Fox and Palmer Creeks with t-tests using the maximum number of species within each point over all point-count dates.
|Figure 2. Distribution of bird points within habitat types at Tallgrass Prairie National Preserve, Kansas, 1998-1999.|
|Figure 3. Point-count and transect locations at Tallgrass Prairie National Preserve, 1998-1999.|
We also surveyed breeding, migratory, and resident birds using fixed-distance (100 × 500 m) strip transects. Transects were located throughout all habitats in the preserve and placed approximately 2414 m apart (Fig. 3). Of the 11 transects, six were located within burned and grazed native prairie, one in unburned and ungrazed prairie (including the Ranch Headquarters cedar grove), one in brome field, one in gallery forest along Fox Creek, one in gallery forest along Palmer Creek, and one at Red House Spring. We recorded the start and end locations of each transect using GPS (Appendix A). We walked slowly within each strip, trying to cover as much ground as possible by walking in a zig-zag pattern. We recorded all birds seen or heard within each strip. Counts were conducted bimonthly: 9-11 September 1998, 13 November 1998, 24 January 1999, 11 March 1999, 23-25 May 1999, and 15-16 August 1999. We calculated densities (birds per hectare) for each species by dividing the number of individual birds for each species by the area of the strip (100 × 500 m = 5 hectares). We compared densities and species diversity at Fox and Palmer Creeks using paired t-tests with sampling dates (n = 6) as pairs.
We quantified basic habitat characteristics of the 23 burned and grazed, one unburned and ungrazed, five brome field, and three riparian/prairie points. We did not quantify habitat along Fox or Palmer Creeks or the Cedar Grove because the primary focus of this study was on grassland birds. We measured slope using a clinometer and aspect using a compass at each point. We then centered two 90-m transects on each bird point after randomly selecting their compass bearings. We then sampled vegetation every 15 m along each transect, resulting in 12 vegetation samples per point.
We used an ocular tube made by taping threads of monofilament line across a small, open ended steel can to determine percent canopy and ground cover (James and Shugart 1970). We determined canopy cover by looking directly overhead through the tube and recorded presence or absence of canopy cover where the threads crossed. We determined ground cover hits by looking through the tube at the ground, and recorded whether bare ground, rock, litter, grass, forb, or woody plant occurred where the threads crossed.
We measured litter depth and heights of the nearest grass, forb, and woody plant at each sampling point using a 1.2-m tall dowel marked at 1- and 10-cm intervals. We estimated the height of plants greater than 3.6 m. We used the same dowel to measure distribution of vegetation by recording contacts or "hits" in each 10-cm interval (Rotenberry and Weins 1980). Average hits in the first 10-cm interval were used as indices of spatial variation. We also created an index of spatial variation using a checkerboard with nine 3-cm² squares (Bibby et al. 1993). We placed the board vertically into the vegetation then walked away from the board until 50% of it was blocked by vegetation. We also calculated the coefficient of variation for grass, forb, and maximum vegetation height, and both of the indices of spatial variation to determine variation among sampling points within each bird point.
We estimated grass and forb density by the point-quarter sampling procedure (Weins 1973). At each point, we measured the distance to the nearest grass and forb. Both grass and forb density indices were calculated for each bird point by the equation:
where d = distance in cm to the nearest forb or grass, and N = number of samples per bird point (twelve).