Model selection.

Shown are information criteria for fit of models including the fixed and random effects of (A) M. lucifugus, (B) P. subflavus, (C) M. sodalis, and (D) M. septentrionalis. Fixed effects are Day, smoothed Day, Year, and smoothed Year, and the random effects are Route and Route nested in Location. Best models were selected on the basis of Akaike's Information Criterion (AIC). DF are the degrees of freedom, Δi is the difference in AIC between the top-ranked and listed model, and wi is the Akaike weight, the weight of evidence for each model in the set given the data (where 1.00 represents the highest likelihood of the model relative to other models). The number of models examined varied for each species because some random effects were not applicable for some species, due to the particular survey routes used.本图展示了针对(A) M. lucifugus、(B) P. subflavus、(C) M. sodalis及(D) M. septentrionalis四个物种种群的、包含固定效应与随机效应的模型拟合信息准则。模型的固定效应包括日期（Day）、平滑化日期（smoothed Day）、年份（Year）以及平滑化年份（smoothed Year），随机效应则为样线（Route）与嵌套于样地（Location）内的样线。最优模型以赤池信息准则（Akaike's Information Criterion，AIC）为依据进行筛选。其中DF为自由度（degrees of freedom），Δi为最优排名模型与所列模型间的AIC差值，wi为赤池信息权重（Akaike weight），即给定数据集下该模型集合中各模型的证据权重（1.00代表该模型相较于其余模型拥有最高的似然性）。鉴于不同物种所使用的调查样线存在差异，部分随机效应无法适配部分物种，因此针对各物种开展检验的模型数量各不相同。