. 2020 Nov 5;15(11):e0240434. doi: 10.1371/journal.pone.0240434

Table 1. Top models (Δ_i < 2) predicting building occupancy by T. mauritianus based on AIC_C and Akaike weight W_i for comparisons between: (i) roosts and all controls (paired + random) and (ii) roosts and paired controls.

Model	AIC_C	D_i	W_i
*Roosts vs all controls*
Wall height + area + wall material^a + eaves depth^a + beams^a	49.2728	0.0000	0.2176
Wall height + area + wall material	49.6630	0.3901	0.1790
Wall height + area + wall material + eaves depth^a + beams^a + tree^a	50.4491	1.1762	0.1209
Wall height + area + eaves depth + beams	50.8716	1.5987	0.0978
Intercept only	174.8033	125.5305	0.0000
*Roosts vs paired*
Wall height + area + beams	32.5679	0.0000	0.5639
Wall height + area + beams + eaves depth^a	34.3921	1.8242	0.2265
Intercept only	122.6432	90.0752	0.0000

Models built with binomial distribution and logit link. Intercept models included for comparisons under conditions of no roost selection. D_i represents delta (AIC_C) and indicates the difference in the AIC_C value with the top model. The symbol ‘a’ denotes non-significant terms based on 0.05 significance threshold.

Table 1. Top models (Δi < 2) predicting building occupancy by T. mauritianus based on AICC and Akaike weight Wi for comparisons between: (i) roosts and all controls (paired + random) and (ii) roosts and paired controls.

Table 1. Top models (Δ_i < 2) predicting building occupancy by T. mauritianus based on AIC_C and Akaike weight W_i for comparisons between: (i) roosts and all controls (paired + random) and (ii) roosts and paired controls.