Masters Theses

Date of Award

8-1995

Degree Type

Thesis

Degree Name

Master of Science

Major

Wildlife and Fisheries Science

Major Professor

David A. Buehler

Committee Members

Ralph W. Dimmick, Mary Sue Younger

Abstract

Apparent declines of breeding neotropical migrant land bird populations in the United States over the last 25 years have been associated with increases in populations of brown-headed cowbirds (Molothrus ater). Because cowbirds parasitize nests of many of these species, such parasitism may be a major contributing factor to these declines. I used avian-census data and landscape-level habitat variables to develop models predicting the presence/absence of cowbirds and nest parasitism in forest openings in a predominately forested (89%) landscape at Land Between the Lakes in Kentucky and western Tennessee. Cowbird populations were surveyed in forest-opening and interior-forest settings between 15 May and 1 July 1993 and 1994, using 10-minute, 50-m fixed-radius point counts. Cowbirds were present within 50 m in 62 (27%) of 227 forest-opening and interior forest points. Cowbirds were present at forest-opening edges more often than at interior forest points (P̲ ≤ 0.05). Cowbird presence did not differ by distance from edge for interior-forest points (200, 300, 400, and 500-m points) (P̲ > 0.05). Overall cowbird abundance per point (forest-opening and interior-forest points) was 0.53 cowbirds/point. Cowbirds were more abundant at forest-opening edges than at interior forest points (P̲ ≤ 0.05). Cowbird abundance did not differ by distance from edge for interior-forest points (P̲ > 0.05). Potential cowbird host nests were located and monitored between 15 May and 1 July 1994. A total of 195 active nests of 20 species known to be parasitized by brown-headed cowbirds were located and monitored. Of these nests, 23.6% (46 nests) were parasitized by cowbirds. The percentage of nests parasitized did not differ across distance from nest to nearest edge categories (P̲ > 0.05). Forest species were more heavily parasitized (30.5% of 82 nests) than were edge species (18.6% of 113 nests parasitized) (P̲ = 0.03). A logistic regression model was developed to predict cowbird presence/absence in forest openings based on 17 landscape and avian census variables associated with each opening. Forest opening size (ha), average height of vegetation in the opening (m), and the number of potential cowbird foraging locations within a 2-km radius of the opening were the best predictors of cowbird presence/absence in forest openings. Larger forest openings, with lower average vegetation heights that were near a high concentration of foraging locations were the most likely to have cowbirds present. Using these model variables, predicted and observed responses were concordant 90.0% of the time, and forest openings were correctly classified for cowbird presence/absence 82.9% of the time. A logistic regression model was developed to predict the presence/absence of nest parasitism in forest openings based on 20 landscape and avian census variables associated with each opening. Forest opening size (ha), average height of vegetation in the opening, and the number of potential cowbird foraging locations within 2 km of the opening were the best predictors of the presence of nest parasitism presence/absence in forest openings. Smaller forest openings, with higher average vegetation heights that were near a high concentration of foraging locations were most likely to have nest parasitism present. Using these model variables, predicted and observed responses were concordant 85.2% of the time and forest openings were correctly classified for nest parasitism presence/absence 81.8% of the time. A total of 148 point counts and recorded playback sessions of cowbird vocalizations were conducted in forest-opening and interior-forest settings between 15 May and 1 July 1993. I assessed the effects of point-count duration cowbird .detection and estimates of cowbird abundance. I also compared the efficiency of point-count versus recorded-playback techniques for detecting cowbirds and estimating cowbird abundance. Ten-minute point counts were suggested to be the best option for managers initiating efforts to monitor cowbird populations. Point count and playback techniques yielded similar results (P̲ > 0.05) for detecting cowbird presence and estimating cowbird abundance.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS