Published 2006-10-09
Keywords
- community analysis,
- birds,
- multivariate statistic,
- predictive models
Copyright (c) 2006 Italian Journal of Forest and Mountain Environments
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Abstract
In this study the composition and structure of an avian community inhabiting a farmland habitat was investigatedusing remotely sensed data and field surveys. The data were analysed both at the community and the species levels. In the first part of the study, standard multivariate techniques (such as nMDS, cluster analysis, ANOSIM test) were employed to associate peculiar avian communities to different habitat types. In the second part, species-specific models of habitat preference were produced, using both GLM and GAM techniques. The results proved to be useful to build habitat suitability maps and to implement quantitative models to predict changes in the composition of community structure and species presence, following environmental disturbance or land use changes.