Table 2. Co-occurrence detection models used to evaluate the effect of detection and/or presence of one ground-dwelling tinamou species on the detection of the other (the brown tinamou (Crypturellus obsoletus) and tataupa tinamou (C. tataupa)) in a seasonal Atlantic forest remnant in Brazil.

Detection Model†	Detection Covariates ‡	ΔAIC	K	wi	LL
pA = rA ≠ pB = rBA = rBa	SlopeA ≠ SlopeB,Temp	0	9	0.50	560.4
pA = ra ≠ pB ≠ rBA = rBa	SlopeA ≠ SlopeB,Temp	2.00	10	0.18	560.4
pA = rA ≠ pB = rBA = rBa	No covariate	2.43	6	0.15	568.83
pA ≠ rA ≠ pB ≠ rBA ≠ rBa	SlopeA ≠ SlopeB,Temp	3.94	12	0.07	558.34
pA = ra ≠ pB ≠ rBA = rBa	No covariate	4.43	7	0.05	568.83
pA ≠ rA ≠ pB ≠ rBA ≠ rBa	No covariate	5.12	9	0.04	565.52
pA = rA = pB = rBA = rBa	No covariate	18.58	5	0	586.98

Models indicate the same (=) or different (≠) β parameters for the conditional p or r probabilities. Models with ΔAIC < 2 are marked in bold. For detailed description of detection parameters see Methods section. K = no. of parameters. w_i = Akaike weight. LL = twice the negative log-likelihood. Slope = terrain slope. Temp = temperature.

^†All detection models included the best model for occupancy from the co-occurrence models (Ψ^A ≠ Ψ^BA = Ψ^Ba; Elevation^A ≠ Elevation ^BA = Elevation^Ba.

^‡ Covariates indicate that the effect of terrain slope is different for each species (Slope^A ≠ Slope^B), while temperature is the same.