Active acoustic telemetry tracking and tri-axial accelerometers reveal fine-scale movement strategies of a non-obligate ram ventilator

Table 2 Linear mixed effects models (LMEs) to estimate predictors of time spent within a patch for California horn sharks (Heterodontus francisci)

Model Parameters
Response	Predictors	Random	Likelihood	AIC	Total Deviance from Model	Deviance Explained (%)	Chi-Squared	Degrees of freedom	p value
Time in Patch	1 (null model)	(1\|Track)	ML		141.05
Time in Patch	Distance travelled to get to patch	(1\|Track)	ML	145.7	135.7	3.79%	0.584	1	0.444
Time in Patch	Number of patches per individual per night						5.028	1	0.025
Time in Patch	Activity within patch (% of time spent resting)	(1\|Track)	ML	146.1	136.1	3.53%	0.165	1	0.685
Time in Patch	Number of patches per individual per night						5.324	1	0.021
Time in Patch	Area of patch	(1\|Track)	ML	136.7	126.7	10.18%	13.349	1	0.0002
Time in Patch	Number of patches per individual per night						7.446	1	0.006

LMEs were done in the lmer framework in R. Time in patch was calculated as the total time (min) spent in a patch and derived from the start and end times of patch use from a First Passage Time (FPT) analysis. Distance travelled to get to patch (m) and area of patch (m²) were measured in ArcGIS. Activity within patch (% of time spent resting) was calculated using the k-means analysis. Number of patches per individual per night was added to each model to account for temporal autocorrelation associated with time spent, and individual per night was included as a random factor. Maximum Likelihood (ML) was used to obtain AIC and deviance values. Percent deviance explained was calculated by subtracting the total deviance from the model from the null deviance and dividing by the null deviance. The AIC values are from the model summary and for each model independently, as no model comparison was made. Chi-squared values, degrees of freedom, and p values were obtained from the Anova function in the car package in R

ISSN: 2051-3933