期刊
POPULATION ECOLOGY
卷 54, 期 3, 页码 467-474出版社
WILEY
DOI: 10.1007/s10144-012-0319-4
关键词
Compound Poisson model; Negative binomial model; Overdispersion; Trend estimation; Zero-inflated model
类别
资金
- Japan International Research Center for Agricultural Sciences (JIRCAS)
- Joint Institute for the Study of the Atmosphere and Ocean (JISAO) under NOAA [NA17RJ1232, 1823]
- Grants-in-Aid for Scientific Research [21580241] Funding Source: KAKEN
Overdispersed count data are very common in ecology. The negative binomial model has been used widely to represent such data. Ecological data often vary considerably, and traditional approaches are likely to be inefficient or incorrect due to underestimation of uncertainty and poor predictive power. We propose a new statistical model to account for excessive overdisperson. It is the combination of two negative binomial models, where the first determines the number of clusters and the second the number of individuals in each cluster. Simulations show that this model often performs better than the negative binomial model. This model also fitted catch and effort data for southern bluefin tuna better than other models according to AIC. A model that explicitly and properly accounts for overdispersion should contribute to robust management and conservation for wildlife and plants.
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