| dc.description.abstract | I surveyed Harris' hawk (Parabuteo unicinctus) grouping relative to rainfall and landscape patterns at eight southern Texas study sites during the fall non-breeding season. Objectives were to: I) determine relationships among landscape use, grouping and abundance across sites, 2) characterize variation in grouping associated with rainfall patterns in the brushland and grassland ecoregions of the study area, and 3) explore which landscape measures best explained within site variation in grouping, considering 3 spatial scales (mesohabitat, territory, home range). Although abundance was higher (17.1 ± 4.1 hawks/100km) than reported for more arid regions, observations of groups (>2 hawks) varied across sites (0.113 ± 0.037 groups/observation). Groups were not observed at 3 of 4 sites in the wetter, eastern grassland, where abundance was slightly lower than brushland. One grassland site with abundant hawks and groups was a drier, habitat fragment in a landscape devoid o f woody vegetation. Abundance correlated with grouping but not with the landscape variables measured. However, hawks were typically found in areas with less fragmented woody vegetation. Vegetation strata differed between random points and hawk locations at 5 sites, four of which had groups, but differences could not be generalized across sites. Sightings of 3 hawks were more likely to be at brushland than grassland locations (z = 2.34, p < 0.05). Larger groups (4-5 hawks) and groups with young-of-the-year were sighted infrequently with no differences between ecoregions. Two indices of grouping were significantly negatively correlated with rainfall recorded near each site in the same year, but not with indices o f rainfall during the previous year. Only subtle differences in landscape vegetation were found between areas around grouped and non-grouped Harris' hawk locations. I compared vegetation patterns at three ecologically appropriate scales. The greatest differences were observed at the finest scale (simulating mesohabitat), leading to hypotheses related to prey availability for this group-hunting predator and distribution of scarce resources (e.g., water). Results were interpreted relative to the Ecological Constraints Model predicting groups in harsh environments. To refine testable predictions from this model, I recommend using spatially explicit landscape measures to document distribution of resources relative to groups. | en |
