Seasonal Abundance and Habitat Associations of American Kestrels on the Southern High Plains of Texas
American Kestrel (Falco sparverius) populations are generally declining across the species' North American distribution but the population in the Southern High Plains region currently appears to be stable. Historical evidence suggests the region formerly had a low abundance of kestrels, and that their current numbers are due to landscape changes associated with European settlement. We conducted monthly surveys for American Kestrels across 2 yr to estimate seasonal densities and identify land cover associations in the Southern High Plains of Texas. We found an overall estimated density of 0.99 birds/km2 (95% CI = 0.406, 1.582) across the 2-yr period, with seasonal estimated densities highest in autumn and winter (0.92–2.53/ km2), and lowest in spring (0.49–0.67/km2). Whereas other studies have found that temperature influenced detection of wintering kestrels, we found an interaction of drought conditions and snow most strongly influenced the number of kestrels in our study area. Kestrels largely used land cover types in proportion to availability but there was some evidence of seasonal shifts. Generally, they tended to avoid cotton fields and sometimes selected for areas with woodlots, abandoned or occupied houses, and barns, all of which likely provided nesting and roosting opportunities. Our study provides the first contemporary assessment of seasonal abundance and habitat associations of American Kestrels in the Southern High Plains, where their presence and abundance has been unintentionally facilitated by landscape changes following settlement. We provide a baseline for population monitoring and studies assessing response to additional landscape changes (e.g., development of wind energy facilities) and a changing climate.
Abstract
American Kestrel survey route (heavy dark line) located between the urban centers of Lubbock and Levelland in, respectively, Lubbock and Hockley Counties, Texas.
Estimated seasonal density (circles and triangles) and 95% confidence intervals (lower and upper limit of bars) for American Kestrels derived from monthly 141-km surveys in the Southern High Plains of Texas, December 2012–November 2013 (Year 1; denoted by circles) and December 2013–November 2014 (Year 2; denoted by triangles).
Ninety-five percent confidence intervals on resource selection index (Wi) values for American Kestrels by land cover type, season, and year based on monthly 141-km surveys in the Southern High Plains region of Texas for 2012–2013 (Year 1) and 2013–2014 (Year 2). Point estimate of index value indicated by point within interval. Horizontal dashed line at one indicates value of no apparent selection for or against land cover type. Land cover types are grazed grasslands (GGRAS), non-grazed grasslands (NGRAS), textile agriculture (TAGRI), grain agriculture (GAGRI), and other cover types that occurred in low frequency and amount (OTHER).
Solid triangles indicate standardized selection ratios (Bi) values for American Kestrels by land cover type, season, and year based on monthly 141-km surveys in the Southern High Plains region of Texas for 2012–2013 (Year 1) and 2013–2014 (Year 2). Standardized ratios sum to one for all land cover types and indicate relative importance of cover types when compared with each other. Open triangles indicate proportion of landscape (i.e., availability) in each of the land cover types. Land cover types are grazed grasslands (GGRAS), non-grazed grasslands (NGRAS), textile agriculture (TAGRI), grain agriculture (GAGRI), and other cover types that occurred in low frequency and amount (OTHER).
Contributor Notes
1 Present address: USDA Forest Service, Kiowa and Rita Blanca National Grasslands, Clayton, NM 88415 USA.
2 Email address: clint.boal@ttu.edu; cboal@usgs.gov
Associate Editor: Jean-François Therrien