Unveiling the roles of temporal periodicity, the spatial environment and behavioural modes in terrestrial animal movement

Table 1 Overview of the fitted HMMs, each including a combination of the predictors modes, time and space

Model	Predictor set	#modes	μ, σ and κ covariates	γ_i→j covariates	ΔAIC eland	ΔAIC wildebeest	ΔAIC zebra
0	none	1	–	–	85,474	64,152	57,309
1a	modes	2	–	–	10,932	8974	7572
1b	time	1	DCC	DCC	74,877	49,193	50,430
1c	space	1	(tc, sl)²	tc, sl	84,674	63,346	55,937
2a	modes + time	2	DCC	DCC	822	918	2240
2b	modes + space	2	(tc, sl)²	tc, sl	10,157	8219	5327
2c	time + space	1	(tc, sl, DCC)²	tc, sl, DCC	74,117	48,110	48,601
3	modes + time + space	2	(tc, sl, DCC)²	tc, sl, DCC	0	0	0

ΔAIC was calculated for each species separately relative to the best model of that species. The AIC of the full models (Model 3) was 1,305,636; 1,304,360 and 1,286,972 for eland, wildebeest and zebra respectively. tc = tree cover, sl = terrain slope, DCC = daily cyclic covariate, μ, σ and κ are movement parameters and γ_1→2 = mode switching probability. Powers of 2 indicate all two-way interactions among the covariates between brackets

ISSN: 2051-3933