Arboreal route navigation in a Neotropical mammal: energetic implications associated with tree monitoring and landscape attributes

Table 2 Results of two GLMMs testing the influence of different landscape attributes (slope, presence of canopy gaps, elevation and FTs density) on the probability occurrence of a route segment used at least twice and four times within a certain quadrat. Group ID (N = 5) was included as random effect in the model. We compared the full model to a corresponding null model (with only random and control variables) using likelihood-ratio tests. All p-values < 0.05 are shown in bold for clarity

Response variable	Probability of locating a route used at least twice					Probability of locating a route used at least four times
Full null model comparison	χ2 = 23.2, d.f. = 5, p < 0.001					χ2 = 18.8, d.f. = 5, p = 0.002
Predictor variable	Est.	s.e.	CI_lower	CI_upper	p-value	Est.	s.e.	CI_lower	CI_upper	p-value
(Intercept)	−0.156	0.574	−1.567	1.209	a	− 3.126	0.298	−6.359	−2.531	a
Slope	0.139	0.104	−0.128	0.481	0.182	0.390	0.139	0.006	0.752	0.058
Presence of canopy gaps	−0.803	0.115	−1.088	−0.482	0.001	−0.354	0.186	−0.452	0.023	0.128
Elevation	0.446	0.328	−0.323	1.298	0.198	0.944	0.550	−0.363	2.237	0.173
FTs density	0.896	0.213	0.372	1.390	0.006	0.967	0.246	0.401	1.533	0.010
Elevation * FTs density	−0.280	0.210	−0.783	0.340	0.273	−0.439	0.140	−0.809	−0.374	0.041
Overlapping area ^b	0.274	0.136	−0.014	0.689	0.054	−1.000	0.203	−3.398	−0.437	0.003
Location within the HR ^b	−0.538	0.442	−1.609	0.518	0.255	−0.724	0.499	−1.414	0.838	0.190
Autocorrelation term ^b	2.487	0.072	2.347	2.632	< 0.000	2.943	0.100	1.636	3.252	< 0.000

ISSN: 2051-3933