文献类型：期刊文献

英文题名：On the use of observational data in studying biodiversity-productivity relationships in forests

作者：Chu, Xiuli[1] Yang, Hua[2] Jiang, Yong[3] Man, Rongzhou[4]

第一作者：楚秀丽

通信作者：Man, RZ[1]

机构：[1]Chinese Acad Forestry, Res Inst Subtrop Forestry, Hangzhou, Zhejiang, Peoples R China;[2]Beijing Forestry Univ, Coll Forestry, Beijing, Peoples R China;[3]Guangxi Normal Univ, Minist Educ, Key Lab Ecol Rare & Endangered Species & Environm, Guilin, Peoples R China;[4]Ontario Minist Nat Resources & Forestry, Ontario Forest Res Inst, Sault Ste Marie, ON, Canada

年份：2019

卷号：95

期号：1

起止页码：24-28

外文期刊名：FORESTRY CHRONICLE

收录：;EI(收录号：20192206988462);Scopus(收录号：2-s2.0-85066114757);WOS:【SCI-EXPANDED(收录号:WOS:000468798000005)】；

基金：The authors acknowledge the editorial contributions from Lisa Buse and Mya Rice of the Ontario Ministry of Natural Resources and Forestry (OMNRF), the financial support from OMNRF, Canada to R.M. and Special Funds for the Basic Scientific Research Service of the Central Public Welfare Research Institute, China to X.C. (CAFYBB2018GC003).

语种：英文

外文关键词：species richness; confounding variables; regression analysis; causal effects; multicollinearity

摘要：The biodiversity-productivity relationship is one of the focus areas in ecological research that is studied primarily through mixed species experiments. Recent efforts in forests, however, increasingly involve the use of observational data, due to the difficulty in establishing long-term, multispecies plantations. Caution is warranted in the observational data-based causal relationships between biodiversity and productivity due to the potential confounding effects by environmental variations. In this article, we use a recent forest example to demonstrate how erroneous results could be generated in studying biodiversity-forest productivity relationships when species diversity is highly correlated with environmental variables (multicollinearity). In forestry, erroneous biodiversity-productivity relationships can mislead future research, industry decisions, and policy development. Forest researchers and managers should be aware of the issues associated with collinear data and validate research results with literature reports and professional knowledge. Options to deal with observational data are discussed.