Predicting the continuum between corridors and barriers to animal movements using Step Selection Functions and...

Area of Science:

• Landscape ecology
• Movement ecology
• Wildlife conservation

Background:

• Habitat loss and fragmentation necessitate identifying animal movement corridors and barriers.
• Existing corridor prediction models often assume unrealistic movement patterns (optimal or random).
• A dynamic corridor-barrier continuum framework is proposed to better represent animal movement.

Purpose of the Study:

• To develop a novel methodological approach for predicting realistic corridor-barrier continua using individual-based movement data.
• To introduce and apply the randomized shortest path (RSP) algorithm in movement ecology.
• To model animal movement as a trade-off between exploration and exploitation.

Main Methods:

• Utilized step selection functions (SSF) to create friction maps representing corridor-barrier continua at a tactical movement level.
• Applied the randomized shortest path (RSP) algorithm on friction maps for strategic movement predictions between functional areas.
• Modulated the RSP parameter Ѳ to bridge optimal and random walk movement models.

Main Results:

• The RSP algorithm, with an intermediate Ѳ value, accurately predicted wild reindeer migration corridors in Norway.
• This intermediate value signifies a movement trade-off between optimization and exploration, reflecting realistic animal behavior.
• The RSP approach outperformed traditional models assuming pure optimality or random walk in predicting empirical movement data.

Conclusions:

• The proposed approach provides a more realistic prediction of corridor-barrier continua by modeling animal cognitive navigation.
• The RSP algorithm generalizes common corridor identification methods and accounts for non-random, suboptimal movement strategies.
• This framework has broad applicability for predicting realistic movement pathways for diverse species facing habitat changes.