Breeding ecology nest site selection and human influence of white-tailed hawks on the Texas barrier islands

I conducted fieldwork on Matagorda, Mustang and North Padre Islands along the Texas coast in 2006 and 2007. I located breeding White-tailed Hawk (Buteo albicaudatus; WTHA) pairs using road and point surveys. I monitored productivity, and nest success for 64 actively nesting pairs. I compared the proportion of nesting pairs per occupied territory between islands. I used a nonparametric Mann-Whitney U test to determine differences between nestling and fledgling production on Matagorda and Mustang Islands. North Padre Island was omitted due to a small sample size. The mean earliest clutch initiation for all islands was March 3 (± 3.2 days). Observed nest success on North Padre Island (22.2%) was markedly lower than Matagorda (56.8%) and Mustang (58.3%) Islands. Mayfield nest success estimate for all sites was 41.6%. There was no difference in nestling production (Matagorda ¯x = 1.05 ± 0.91; Mustang ¯x = 1.00 ± 1.83) or in fledgling production (Matagorda ¯x = 0.89 „b 0.87; Mustang ¡Âx = 1.08 ± 1.16) between the two islands. Productivity on Mustang Island is not different than Matagorda Island despite having a high level of human disturbance. This may be a result of WTHAs adapting to human disturbance and breeding pairs having larger territories and no density-dependence influence, such as what may be occurring on Matagorda Island where WTHAs may be at population saturation. I measured nest site selection at 38 nest sites and 38 paired random sites. I created a resource selection probability function for WTHA nest site selection using a logistic regression model of the characteristics measured at a subset of 19 nest sites and paired random sites on Matagorda Island. Models were evaluated using Akaike Information Criterion. The best model on Matagorda Island used the parameters of shrub category, nearest-neighbor distance, and distance to road to correctly differentiate 83% of nest sites from random sites on Matagorda Island, 70% on Mustang Island, and 50% on North Padre Islands. I then created model sets for Mustang and North Padre Islands, of which the best model for both islands used the parameters of shrub height, shrub circumference, and their interaction. The resource selection probability function from the best model on Matagorda Island should be used with caution. Overall, it appears that densely branched and thorned shrub species are important to WTHA nest site selection. I conducted behavioral observations on breeding WTHA pairs on Matagorda, Mustang and North Padre Islands, Texas in 2007. These islands were classified into high human impact (Mustang and North Padre Islands) and low human impact (Matagorda Island). Observations were conducted only during 2-3.5 hours after sunrise, after which visibility decreased due to shimmer caused by radiated heat. I used a generalized liner model with a logit link function that tested for differences between islands. Data collected from two breeding stages were analyzed with a repeated measures analysis. Pairs on Matagorda Island spent more time flying than pairs Mustang and North Padre Islands. This may be a result of an increase in interspecific and intraspecific territorial defense on Matagorda Island. There appeared to be no direct human induced influences to WTHA behavior.