Generalized Pareto for Pattern-Oriented Random Walk Modelling of Organisms' Movements

Area of Science:

• Ecology and Behavioral Ecology
• Movement Ecology
• Quantitative Biology

Background:

• Understanding organism movement and dispersal is key to linking population dynamics with environmental spatial heterogeneity.
• Random walk (RW) models are standard for describing movement patterns, but the debate continues regarding whether Lévy or alternative RWs better represent forager movements.

Purpose of the Study:

• To introduce and apply the Generalized Pareto Distribution (GPD) as a unified framework for modeling diverse random walk patterns.
• To characterize and quantify spatial movement strategies of different species and individuals using GPD parameters.
• To explore the utility of GPD-derived movement metrics for ecological applications, including ecosystem indication and behavioral studies.

Main Methods:

• Utilized the Generalized Pareto Distribution (GPD) to model movement data, encompassing Normal, exponential, and power-law distributions as special cases.
• Applied the GPD framework to GPS-tracked seabird foraging movements and Vessel Monitoring System (VMS) data from fishing vessels in the Peruvian pelagic ecosystem.
• Analyzed the scale and shape parameters of the fitted GPD to synoptically characterize movement scale and diffusive properties.

Main Results:

• The GPD successfully models diverse movement patterns, allowing the most likely random walk model to emerge directly from the data.
• Fitted GPD parameters quantified the variability in spatial strategies among different seabird species and individual foragers.
• Demonstrated the applicability of GPD parameters as metrics for describing movement patterns and individual behavioral differences.

Conclusions:

• The Generalized Pareto Distribution offers a flexible and powerful tool for analyzing and synthesizing movement data across various taxa and systems.
• GPD parameters provide valuable metrics for describing movement patterns, utilizing top predators as ecosystem indicators, and studying individual behavioral variability.
• This approach advances our understanding of how movement strategies influence ecological dynamics in heterogeneous environments.