Biology Ph.D. Dissertations


An Integrative Approach to Conservation of the Crested Caracara (Caracara Cheriway)in Florida: Linking Demographic and Habitat Modeling for Prioritization

Date of Award


Document Type


Degree Name

Doctor of Philosophy (Ph.D.)


Biological Sciences

First Advisor

Karen Root


The Crested Caracara (Caracara cheriway) occurs in Florida as an isolated population in the south-central region. Caracaras currently occur almost exclusively on ranchlands, most of which are privately owned. The status of Florida's Caracara population has been a subject of concern as conversion of grasslands and pasture to residential and commercial development has accelerated, particularly within the past 10 years. Our objectives for this work were to provide a better understanding of the specific ecological needs of the Caracara with regard to habitat and to use known demographic information to develop a population model suitable for assessment of extinction risk. To achieve this, we used population viability analyses and a GIS. Using all available data, we constructed a population viability model to determine the most important demographic parameters and assess viability. The most important demographic variable impacting viability was adult survival, and our demographic model indicated that this population is stable under current conditions. However, current conditions are unlikely to persist and multiple habitat loss scenarios were explored. Results indicate abrupt habitat loss increases risk compared to gradual loss of the same magnitude. Using GIS, we evaluated habitat composition of known caracara home ranges and used the Mahalanobis distance statistic to build a predictive habitat model. Geographic layers used were recoded to reflect the landscape context surrounding each cell within an average home range. Models were tested using an independent data set. Results indicated that heterogeneity is critical and we identified seven landscape characteristics, including five vegetation types and both lentic and lotic water, necessary in maintaining suitable habitat in Florida. Our final habitat model performed well in both predictive accuracy and in reducing the portion of the study area judged suitable. Finally, we used occurrence data to identify distribution patterns of habitat use. A core area of habitat use was readily identified, and habitat use decreased consistently across buffers extending out from the core. This was integrated with our demographic and habitat models to prioritize habitat for conservation, and construct models simulating habitat loss in particular regions in Florida and their associated risk of population decline.