Five carnivore species representing three families were detected sufficiently to permit analysis. Carnivores were, in general, not distributed evenly over regional, landscape and local scales in the panhandle environment. Carnivores were also found to impact each others distributions both locally and at landscape level scales. Logistic regression modeling determined that carnivore distributions and canid distributions in particular, could be modeled according to their affiliations with remotely sensed Normalized Difference Vegetation Index data (NDVI) in the Oklahoma panhandle. This model provides a framework for predicting carnivore occurrence in habitats based upon remotely sensed vegetation spectral data.

Carnivores were sampled in the Oklahoma panhandle from October 1995 to February 1997 using aluminum, baited tracking plates and infra-red triggered cameras. Carnivore distributions on regional, landscape and local scales were examined using traditional univariate statistical techniques. A model of carnivore distribution in the panhandle was also developed using remotely sensed Advanced Very High Resolution Radiometer (AVHRR) data and logistic regression.