Table 2. Different models for covariate effects on the transition probabilities between behavioural states.

The predictors η_ij relate to the transition probabilities as given in Eq (7). The spline effects of hour are denoted by $H_{ij}^t={\sum }_k{\delta }_{ij}^{\left(k\right)}{h}_k^t$, of τ_t by $T_{ij}^t={\sum }_k{\theta }_{ij}^{\left(k\right)}{s}_k^t$, and of d_t by $D_{ij}^t={\sum }_k{\zeta }_{ij}^{\left(k\right)}{d}_k^t$ for k = 1, 2, 3 and i, j = 1, 2, 3; i ≠ j. A list of all explored models can be found in S1 Table in the Supporting Information.

Predictors in the transition probabilities
Model	η12(t)	η13(t)	η21(t)	η23(t)	η31(t)	η32(t)
1	${\alpha }_{00}+T_{12}^t+H_{12}^t$	${\alpha }_{01}+T_{13}^t+H_{13}^t$	${\beta }_{00}+T_{21}^t+H_{21}^t$	${\beta }_{01}+T_{23}^t+H_{23}^t$	${\gamma }_{00}+D_{31}^t+H_{31}^t$	${\gamma }_{01}+D_{32}^t+H_{32}^t$
2	${\alpha }_{00}+H_{12}^t$	${\alpha }_{01}+H_{13}^t$	${\beta }_{00}+H_{21}^t$	${\beta }_{01}+H_{23}^t$	${\gamma }_{00}+H_{31}^t$	${\gamma }_{01}+H_{32}^t$
3	${\alpha }_{00}+T_{12}^t$	${\alpha }_{01}+T_{13}^t$	${\beta }_{00}+T_{21}^t$	${\beta }_{01}+T_{23}^t$	${\gamma }_{00}+D_{31}^t+H_{31}^t$	${\gamma }_{01}+D_{32}^t+H_{32}^t$
4	${\alpha }_{00}+T_{12}^t$	${\alpha }_{01}+T_{13}^t$	${\beta }_{00}+T_{21}^t$	${\beta }_{01}+T_{23}^t$	${\gamma }_{00}+D_{31}^t$	${\gamma }_{01}+D_{32}^t$