Abstract

Successful management practices of avian populations depend on understanding relationships between birds and their habitat, especially in rare habitats, such as riparian areas of the desert Southwest. Remote-sensing technology has become popular in habitat modeling, but most of these models focus on single species, leaving their applicability to understanding broader community structure and function largely untested. We investigated the usefulness of two Normalized Difference Vegetation Index (NDVI) habitat models to model avian abundance and species richness on the upper San Pedro River in southeastern Arizona. Although NDVI was positively correlated with our bird metrics, the amount of explained variation was low. We then investigated the addition of vegetation metrics and other remote-sensing metrics to improve our models. Although both vegetation metrics and remotely sensed metrics increased the power of our models, the overall explained variation was still low, suggesting that general avian community structure may be too complex for NDVI models.

First posted May 22, 2013