Application of Bayesian methods to habitat selection modeling of the northern spotted owl in California: new statistical methods for wildlife research

We compared a set of competing logistic regression habitat selection models for Northern Spotted Owls (Strix occidentalis caurina) in California. The habitat selection models were estimated, compared, evaluated, and tested using multiple sample datasets collected on federal forestlands in northern California. We used Bayesian methods in interpreting Akaike weights calculated for the estimated models. This approach combines Akaike weights with prior probabilities to provide posterior probabilities for the set of competing models for each dataset. This process can be iterated with multiple sample datasets to calculate a succession of posterior probabilities that provide revised assessments of the relative credibility of the models. The posterior probabilities also provide weights for model averaging. They can be used to measure the importance of the covariates in the models, and they provide the weights for model averaging of the predictive values and estimates of the coefficients of the covariates, along with error. This approach offers a robust solution to modeling habitat associations, providing a more realistic assessment of error and uncertainty in the results. We illustrate these methods with sample datasets for the Northern Spotted Owl in California.