Top candidate models that predict mean maximum sap velocity and mean annual basal area increment across seven tropical lowland rainforest tree species in Daintree, northeast Australia.

AICc refers to Akaike Information criterion corrected for small sample size, ∆AICc to the difference between each model’s AICc and the minimum AICc found, w to Akaike weights, and w1/wi to evidence ratios where w1 is the Akaike weight of the best fitting model. Akaike weights may be interpreted as relative model probabilities. Models with a higher evidence ratio are less likely to be the best model. The summed weights (Σw) for predictors is the relative likelihood that the predictor should form part of the model [48]. The top candidate models include ≥95% of Akaike weights.Top candidate models that predict mean maximum sap velocity and mean annual basal area increment across seven tropical lowland rainforest tree species in Daintree, northeast Australia.