Incorporating animal spatial memory in step selection functions.

Area of Science:

• Ecology
• Animal Behaviour
• Movement Ecology

Background:

• Animal movement patterns are significantly influenced by memory, a factor often overlooked in ecological studies.
• Understanding the interface between spatial memory and animal navigation is key to explaining space use and home range behavior.
• Direct measurement of memory is challenging, complicating its integration into movement models.

Purpose of the Study:

• To incorporate spatial memory into step selection functions (SSF) for analyzing resource selection and space use in feral hogs (Sus scrofa).
• To evaluate the impact of land cover, time of day, and spatial memory on feral hog space use.
• To demonstrate how integrating spatial memory can enhance habitat selection estimates in animal movement models.

Main Methods:

• Utilized Biased Random Bridge kernel estimates linked to residence time as a proxy for spatial memory.
• Developed and tested four distinct dynamic spatial memory maps.
• Applied memory-based SSF to feral hog relocation data.

Main Results:

• Feral hogs exhibit a strong preference for previously visited areas, particularly those within recent memory.
• Habitat selection varied by time of day, with hogs favoring forests and water bodies and avoiding grasslands during the day.
• Animals demonstrated the ability to differentiate spatial memories based on circadian activity phases.

Conclusions:

• Incorporating spatial memory into movement models significantly improves habitat selection estimates.
• Memory-based SSF offers a practical method for investigating spatial learning in animal movement decisions.
• Feral hogs' reliance on recent memory is likely driven by the dynamic and uncertain distribution of local food resources.