Effects of fire on home range size, site fidelity and habitat associations of grassland birds overwintering in southeast Texas

Abstract

Understanding habitat affinities of wintering grassland birds is pertinent to their conservation. The objectives of this study were to evaluate the 1) effect burn history has on vegetation structure in coastal prairie; 2) habitat associations of the most common avian species; and, 3) home range size and site fidelity of Le Conte's Sparrow (LCSP). Data were collected on sites with 1-, 2- and 3-year burn histories during the winter of 2002 at Brazoria National Wildlife Refuge, Texas. Line transects were conducted to estimate bird abundance and vegetation measurements were recorded for herbaceous density, shrub density, and community type. Twenty-six Le Conte's Sparrows (Ammodramus lecontii) (LCSP) were radio-marked, in 1- and 2-year burn units, and located for approximately 10 days. One hectare plots on each burn unit were flush-netted once each month. Burn history was highly correlated with herbaceous vegetation and shrub density. LCSP were most common in areas of medium herbaceous density, low to medium litter densities, and areas with tall vegetation. LCSP also occurred in areas with low shrub densities, but abundance was significantly higher in areas with dwarf wax myrtle (Morella cerifera). Savannah Sparrows (Passerculus sandwichensis) were most often found in areas with low herbaceous and shrub densities. Sedge Wrens (Cistothorus platensis) (SEWR) were most common in areas with dense herbaceous vegetation, but showed no relationship to shrub density. SEWR were associated with areas where saltbush and tallow were present. Swamp Sparrows (Melospiza georgiana) were most common in areas of high shrub densities, but demonstrated no relationship to herbaceous vegetation densities. LCSP were sedentary during winter with a mean home range of 2.41 ha (72% < 1 ha) with a 50% probability, and 10.31 ha (44% < 1 ha and 55% < 1.5 ha) with a 95% probability. Home range size did not significantly differ between burn year (p = 0.227). LCSP appeared to exhibit a behavioral response to capture (p < 0.001) with estimated capture probability of 0.462 and recapture probability of 0.056. Maintaining a mosaic of prairie, in three year burn rotations, and controlling woody invasives will provide sufficient overwintering habitat for SAVS, LCSP and SEWR.