Table 3. Ranked linear mixed-effects models of anemone abundance explained by date, bleaching category (bleach_cat) and the interaction between these two variables.

mod	k	LL	AICc	ΔAICc	wAICc
Abundance∼date×bleach_cat+(1|id)	6	−199.7	417.79	0	1
Abundance∼date+bleach_cat+(1|id)	4	−212.6	438.24	20.45	<0.01
Abundance∼date+(1|id)	3	−217.5	445.6	27.81	<0.01
Abundance∼bleach_cat+(1|id)	2	−221.9	452.19	34.4	<0.01
Abundance∼1+(1|id)	1	−226.9	459.82	42.03	<0.01

All models have individual colony coded as a random effect (1|id). The models are ranked against the intercept-only model (NULL model). Also shown is the number of estimable model parameters (k), maximum log-likelihood (LL), Akaike’s information criterion corrected for small samples (AIC_c), the difference in AIC_c for each model from the top-ranked model (ΔAIC) and the model weight (wAIC_c).