Abstract: A deep understanding of the interactions between multispecies invasions and impact factors facilitates more pertinent control of invasions. Plot survey was carried out in Laohuling Forest, Nanning, and the vegetations were recorded in six districts; and relative cover of all alien plants in a plot, the alien ratio, was used as invasion index. R-lme4 mixed effect model with random district intercept and SPSS GLM-GEN were used to calculate impacts of environmental on invasions, and R-effects package to visualize the complex interaction effects. Multivariate mixed effect model showed that roadside had positive main effect on invasion(P=0.000), canopy cover and dominant plant height both had negative main effect on invasion(P=0.000); native richness' main effect on invasion was not significant, but its interaction with roadside had significant effect on invasion(P=0.007); on roadside, native richness strongly raised resistance to invasion; while within forest, native richness weakly increased resistance to invasion; canopy cover and dominant plant height's interaction effect on invasion was significant(P=0.004), but their filter effects were non-additive. GLM-GEN results indicated that forest age and tending times did not influence invasion; and we did not find significant difference between introduced and native species plantations in impacts on invasions; furthermore, plot locations relative to forest lanes influenced alien invasions, and plots below forest lanes were more vulnerable to invasions. In future works of monitoring and controlling alien plants in forest, emphasis should be focused on stands those below forest lanes.

Key words: alien plant, roadside, interaction effect, canopy cover, invasion