*K= number of parameters in model, AICc = Akaike Information Criterion corrected for small sample size, ΔAICc= difference between the AICc value of each model and the lowest AICc model, ꞷi= AICc weight, LL= negative log likelihood.