RETRACTED: Impact of anthropogenic disturbance on the density and activity pattern of deer evaluated with respect to spatial scale-dependency

This article has been retracted: please see Elsevier Policy on Article Withdrawal (http://www.elsevier.com/locate/withdrawalpolicy).This article was retracted at the request of the authors Naoki Agetsuma, Ryosuke Koda, Riyou Tsujino and Yoshimi Agetsuma-Yanagihara.The authors are sorry to report that they found a bug in statistical software “R” used in their article published in Mammalian Biology. They would like to withdraw the paper.The authors found a bug of the “dredge” function in “MuMIn” package of statistical software “R” (R Development Core Team, 2012) they used. The “dredge” function generates all possible mathematical models such as generalized linear model (GLM) and generalized linear mixed model (GLMM) using all combination of variables, and supports model selection procedures by AIC. The “R” prepares several GLMM functions such as “glmmML”, “lmer” and “glmmadmb”, and manual of the “MuMIn” package shows the “dredge” function can be applicable for all these GLMM functions (Barton K. 2013). Then, they used “glmmadmb” and “dredge” for their analyses.However, when the authors made further analyses of the same data set using “glmmadmb” and “dredge” functions recently, they found a discrepancy of statistic results of “dredge” with other functions (“summary” and “model.sel”). The “dredge” seems not to take account of “mixed effect” of the models and it seems to generate results of GLM (i.e. “glm.nb” function). This bug of software has not been ever reported as far as the authors know and may occur only between “glmmadmb” and “dredge”. The “dredge” might work properly in “glmmML” and “lmer”.The authors found that their data in the paper were analyzed by GLM not GLMM accidentally. Authors re-analyzed the same data set using GLMM (“glmmadmb”) and “model.sel” function and they got different results from the published ones.