Results of General Lineal Mixed Models describing the effect of patch- and transect-level environmental variables on the abundance of six epiphytes and vines present in 18 forest fragments at Chiloé Island (S Chile).

Results of a model describing the effect of all other independent variables on tree DBH, the best predictor of epiphyte abundance, are also provided. Figures represent parameter estimates (estimate ± standard error) and significance levels for the variables retained in the best (reduced) model. Variables excluded from the reduced models in all cases (i.e. for all species) are not shown in the table. ∫ Epiphyte abundance  =  number of ramets per tree (N = 2467 at transect level).§Epiphyte abundance  =  number of ramets per 10-m transect. † Epiphyte presence  =  proportion of trees occupied per 10-m of transect (30 m for A. ovata; N = 680 and N = 197 at transect level, respectively). Patch level: N = 18. NS P>0.10, * P<0.05, ** P<0.01, *** P<0.001.

本研究同时提供了描述所有其余自变量对树木胸径（DBH, Diameter at Breast Height）影响的模型结果，而树木胸径是附生植物丰度的最佳预测因子。图表展示了最优（简化）模型中保留变量的参数估计值（估计值±标准误）及其显著性水平。所有情形（即针对所有物种）下均被排除在简化模型之外的变量未在本表中呈现。
∫ 附生植物丰度=每株树木的分株数（样带水平样本量N=2467）。
§ 附生植物丰度=每10米样带的分株数。
† 附生植物存在率=每10米样带内被占据的树木占比（A. ovata为30米；样带水平样本量分别为N=680与N=197）。
斑块水平：样本量N=18。
显著性标注规则：NS代表P>0.10，*代表P<0.05，**代表P<0.01，***代表P<0.001。