Robust predictions of species distribution: Spatial habitat models for a brood parasite
- 作者：Kelly J. Jewell ; Peter Arcese ; Sarah E. Gergel
- 关键词：Autologistic regression ; GIS ; Model validation ; Molothrus ater ; Predictive mapping ; Bird
- 刊名：Biological Conservation
- 出版年：2007
- 期刊代码：13_00063207
- 类别：env
- 出版时间：December 2007
- 卷：140
- 期：3-4
- 页码：259-272
- 文件大小：1217 K
摘要
Empirical habitat selection models aid managers by predicting distribution patterns, but recent critiques urge a fuller application of available methods. We used a suite of habitat modelling techniques to identify factors influencing the distribution of brown-headed cowbirds (Molothrus ater), a brood parasite capable of causing population declines in some host species. We used data on cowbird occurrence in the Southern Gulf Islands, Canada, to estimate the influence of landscape features on cowbird distribution beginning with a set of a priori models based on literature. Best models were also constructed using stepwise logistic and autologistic regression, and using only widely available data, and these models and their predictive maps were then evaluated. Cowbirds were predicted by proximity to potential feeding areas and landscape context. The best logistic model included landcover, cattle, and distance to urban area and agriculture. Model performance was improved by autologistic regression, and models restricted to widely available data were only slightly poorer. All models were robust to internal validation using bootstrapping and when compared to an independent empirical dataset of parasitism rates of a host on 12 sites monitored for up to 8 years. Our predictive maps for cowbirds should help land managers make spatial predictions about cowbird impact on particular hosts, as well as evaluate land-use decisions that could influence cowbird abundance and host fitness. We evaluated several approaches to habitat modelling of interest to conservationists and make suggestions for future studies when spatial predictions are likely to affect decisions in conservation and land use planning.