This Is AuburnElectronic Theses and Dissertations

Microhabitat Associations of Wintering Birds in a Southeastern Bottomland Forest within the Eastern Gulf Coastal Plain of Florida

Date

2009-07-16

Author

Brian, Rolek

Type of Degree

thesis

Department

Biological Sciences

Abstract

I examined microhabitat characteristics affecting the occupancy of wintering birds in a southeastern bottomland forest with repeated point counts at 186 sites. Models were derived a priori based on published literature and personal observations. I assessed 34 species and their microhabitat preferences within the Choctawhatchee River Basin. I then created microhabitat groups based on all habitat associations that allowed for a simplified interpretation of results. Microhabitat characteristics were important for 25 of 34 species of birds and were important predictors their occupancies 34 times. The occupancy of species was influenced similarly by physiognomic and floristic characteristics; the former influenced occupancy of 12 species and the latter occupancy of 19 species. The basal area of tupelo (Nyssa) was the most important floristic predictor, and for five of six species it negatively affected the presence of birds. Number of woody stems (< 10 cm dbh) was the most important physiognomic predictor of species occupancy. Woody stems affected occupancies of four species. Presence of standing water and oak-hickory (Quercus + Carya aquatica) community affected occupancies of three species. Oak-gum-cypress (Quercus + Nyssa + Taxodium distichum) community affected occupancies of two species. The probability of detecting birds was most frequently affected by date and observer differences. Also affecting the probability of detection were wind, temperature, and time after sunrise. Microhabitat is an important component of habitat selection by birds and should be incorporated into models of occupancy that are used for conservation purposes. My study provides empirically tested associations between occupancy and microhabitat characteristics, and can provide a starting point for future habitat modeling efforts.