Table 2. Results of generalized linear mixed models of harpy eagle prey profile.

First model predicts probability that a given animal preyed by a harpy eagle is a sloth, while the second model predicts probability of prey being a nocturnal animal. Both models use a logit link due to the binomial natural of the data. Both models use tracked individuals and years sample as random effects over the intercept.

Model	Variable	Estimate	Standard error	p-Value	Random individual variance	Random yearly variance
Sloth	Intercept	0.588	0.470	0.2109	1.001	0.513
	Lunar disc (%)	−0.648	0.257	0.0116	–	–
	Minimum temperature (°C)	−0.508	0.263	0.0535	–	–
Night	Intercept	−0.933	0.422	0.0271	0.336	0.367
	Minimum lunar disc (3-day; %)	−0.392	0.269	0.1461	–	–