Calibrating an individual-based movement model to predict functional connectivity for little owls

Area of Science:

• Ecology
• Conservation Biology
• Computational Biology

Background:

• Dispersal is vital for population persistence and conservation efforts.
• Estimating animal movement and functional connectivity is methodologically challenging.
• Individual-based models (IBMs) offer realistic simulations but are difficult to parameterize.

Purpose of the Study:

• To develop and validate a novel analysis workflow for calibrating IBMs of animal dispersal.
• To improve the estimation of functional connectivity using a hybrid approach.
• To predict the dispersal potential of little owls and identify critical habitat corridors.

Main Methods:

• Utilized radio-tracking data from 126 juvenile little owls.
• Employed generalized additive models (GAMs) and biased random bridges (BRB) for habitat utilization estimation.
• Integrated statistical parameter estimation with Approximate Bayesian Computation (ABC) for IBM calibration.
• Applied random forest (RF) regression for dimension reduction of summary statistics.

Main Results:

• The hybrid approach efficiently parameterized the IBM, outperforming least squares regression.
• Identified significant inter-individual and inter-sexual variation in dispersal behavior.
• Females exhibited longer flights and greater directional persistence, consistent with sex-biased dispersal.
• Predicted limited but possible recolonization of northern Switzerland by little owls.

Conclusions:

• The developed analysis chain provides a robust framework for assessing species dispersal connectivity.
• The workflow effectively incorporates landscape influences and individual behavioral responses.
• Highlighted narrow dispersal corridors requiring enhancement for successful recolonization.