期刊
ECOGRAPHY
卷 27, 期 2, 页码 165-172出版社
BLACKWELL MUNKSGAARD
DOI: 10.1111/j.0906-7590.2004.03673.x
关键词
-
We examine the consequences of restricting the range of environmental conditions over which niche-based models are developed to project potential future distributions of three selected European tree species to assess first, the importance of removing absences beyond species known distributions (naughty noughts) and second the importance of capturing the full environmental range of species. We found that restricting the environmental range of data strongly influenced the estimation of response curves, especially towards upper and lower ends of environmental ranges. This induces changes in the probability values towards upper and lower environmental boundaries, leading to more conservative scenarios in terms of changes in distribution projections. Using restricted data analogous to not capturing the fun species' environmental range, reduces strongly the combinations of environmental conditions under which the models are calibrated, and reduces the applicability of the models for predictive purposes. This may generate unpredictable effects on the tails of the species response curves, yielding spurious projections into the future provided that probability of occurrence is not set to zero outside the environmental limits of the species. Indeed, as the restricted data does not capture the whole of the response curve, projections of future species distributions based of ecological niche modelling may be only valid if niche models are able to approach the complete response curve of environmental predictors.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据