文献类型：期刊文献

中文题名：长白落叶松林分进界模型的研究

英文题名：Tree Recruitment Model of Larix olgensis

作者：雷渊才[1] 张雄清[1,2]

第一作者：雷渊才

机构：[1]中国林业科学研究院资源信息研究所;[2]中国林业科学研究院林业研究所

年份：2013

卷号：26

期号：5

起止页码：554-561

中文期刊名：林业科学研究

外文期刊名：Forest Research

收录：CSTPCD;;Scopus;北大核心:【北大核心2011】；CSCD:【CSCD2013_2014】；

基金："十二五"科技支撑项目(2012BAD22B0201);"863"科技计划项目(2012AA12A306)

语种：中文

中文关键词：进界;Poisson模型;负二项模型;零膨胀模型;Hurdle模型;长白落叶松

外文关键词：recruitment; Poisson model; negative binomial model; zero-inflated models; Hurdle models; Larix olgensis

分类号：S791.22

摘要：利用吉林省汪清林业局金沟岭林场落叶松林分连续观测数据,以计数类模型为基础,分别利用Poisson回归模型、负二项模型、零膨胀模型和Hurdle模型拟合林木进界株数,并通过AIC值,Pearson残差图以及Vuong检验对这些模型进行了详细分析比较。结果表明:Poisson回归模型不适用于模拟林木枯损株数;负二项回归模型相对于Poisson回归模型比较适用,但是对于零枯损过多的数据,这两类模型拟合效果较差;零膨胀模型和Hurdle模型对这类数据有很好的解决办法,而且,零膨胀负二项模型拟合效果最好。
Tree recruitment model play an important role in simulating stand dynamic processes. Considering the fact that in permanent sample plots some of the plots have no occurrences of recruitment even over periods of several years, it means that data are bounded and characteristically exhibit varying degrees of dispersion and skewness in re- lation to the mean. Additionally, the data often contain an excess number of zero counts. If least squares method is still used to deal with the data with large proportion of zero counts, the estimated results will be biased. Based on the data from permanent plots of Larix olgensis in Wangqing Forest Farm, Poisson model, negative binomial model, zero-inflated models and Hurdle models were used to analyze tree recruitment. The best model was chosen according to the AIC value, Pearson residual plot and Vuong test. The results showed that Poisson model was not suitable for recruitment, and negative binomial was superior to the Poisson model. But both of them were not competent for the over-dispersion data. Zero-inflated model and hurdle model were fitted into the data. Additionally, zero-inflated negative binomial model （ZINB） outperformed than other models. The result provided a feasible method for analy- zing tree recruitment.