A 'how-to' guide for estimating animal diel activity using hierarchical models.

Area of Science:

• Ecology
• Wildlife Biology
• Statistical Modeling

Background:

• Animal diel activity patterns are crucial for understanding behavioral adaptations, species coexistence, and ecosystem dynamics.
• Current methods like kernel density estimators (KDEs) often ignore spatial heterogeneity in activity, potentially leading to biased results.
• There is a need for more flexible statistical approaches to accurately model diel activity and test ecological hypotheses.

Purpose of the Study:

• To introduce and demonstrate hierarchical models as a superior alternative to KDEs for estimating animal diel activity patterns.
• To show how these models can incorporate spatial variability, temporal correlations, and sampling effort.
• To illustrate the quantification and testing of biotic and abiotic drivers influencing activity patterns.

Main Methods:

• Utilized trigonometric terms and cyclic cubic splines within hierarchical models.
• Applied frequentist and Bayesian approaches for both site-specific and population-averaged analyses.
• Demonstrated model fitting and interpretation using time-stamped data from camera traps and audio recorders.

Main Results:

• Hierarchical models successfully accounted for site-to-site variability and other sources of heterogeneity ignored by KDEs.
• These models effectively quantified the impact of seasonality, anthropogenic stressors, and species co-occurrence on activity patterns.
• Provided a framework for analyzing both conditional and marginal activity patterns.

Conclusions:

• Hierarchical models offer a viable, flexible, and effective alternative to KDEs for modeling animal activity patterns.
• This approach enhances the accuracy of uncertainty estimates and provides more reliable insights into the drivers of diel activity.
• The study provides practical guidelines and tutorials for implementing these advanced statistical methods in wildlife research.