Multiple linear regression models (a - p) of environmental response variables regressed against the habitat transformation variables (predictors), given the spatio-temporal covariables.

Only significant relationships are presented here (α = 0.05, with the exception of agriculture, where α = 0.10). To maximise parsimony, covariable subsets were pre-selected for each model using step-wise regression of each response variable on the full list of possible covariables (see Table 1). For each predictor variable, results are listed in decreasing order of relationship strength based on P values. Only partial relationships between the response and predictor variables are reported here, not the full model results. Significant relationships at αNatural - indigenous vegetation; Invaded - alien invasive vegetation; Agriculture - agricultural land; Urban - urban area. The areal cover of these variables is represented within 100 and 500 m radii of each wetland edge, measured on an ordinal scale. Time - Number of days since the first sampling event; β – standardized regression coefficient; SE – standard error of regression coefficient; Partial r2 – coefficient of partial determination for each respective predictor variable; Res. df – residual degrees of freedom; r2 (Covariables) = Full model r2 - Partial r2 (predictor).