A model-based hypothesis framework to define and estimate the diel niche via the 'Diel.Niche' R package

Area of Science:

• Ecology
• Evolutionary Biology
• Quantitative Biology

Background:

• Understanding how animals utilize the diel period (24-hour light-dark cycle) is crucial for defining ecological niches.
• Existing literature lacks clear, quantitative definitions for diel phenotypes (e.g., diurnal, nocturnal), leading to potential inference issues in research.

Purpose of the Study:

• To propose quantitative definitions for diel phenotypes to standardize research terminology.
• To develop a flexible Bayesian modeling framework for analyzing and comparing diel activity patterns.

Main Methods:

• Proposed four sets of mutually exclusive hypotheses for diel phenotypes based on activity probabilities during twilight, daytime, and night-time.
• Developed a Bayesian modeling framework using Bayes factors for hypothesis comparison and a multinomial model with linear inequality constraints for parameter estimation.
• Created the Diel.Niche R package and an R Shiny web application for model comparison, parameter estimation, and result visualization.

Main Results:

• The study provides extensive simulation results to guide researchers on discriminating between hypotheses across various sample sizes.
• Demonstrated the framework's ability to complement other analytical methods like circular kernel density estimators and animal movement modeling.
• Offered examples and online vignettes for practical application of the Diel.Niche package.

Conclusions:

• The proposed quantitative definitions and Bayesian framework aim to standardize the language and analysis of diel activity.
• Encourages bridging conceptual frameworks with data-driven models in diel research.
• Highlights the ecological and conservation importance of understanding animal diel time utilization.