Towards a fully predictive model of flight paths in pigeons navigating in the familiar area: prediction across differing individuals

TL;DR

This study develops a predictive model for pigeon flight paths that can generalize across different individuals, enhancing understanding of avian navigation and potential cue encoding.

Contribution

It introduces a method to adapt individual-based flight path models to new birds using observations from others, advancing predictive accuracy in avian navigation modeling.

Findings

Model predicts flight paths better than naive approaches.

Prediction accuracy improves with data from multiple individuals.

Implications for understanding navigational cue use in pigeons.

Abstract

This paper will detail the basis of our previously developed predictive model for pigeon flight paths based on observations of the specific individual being predicted. We will then describe how this model can be adapted to predict the flight of a new, unobserved bird, based on observations of other individuals from the same release site. We will test the accuracy of these predictions relative to naive models with no previous flight information and those trained on the focal bird's own previous flights, and discuss the implications of these results for the nature of navigational cue use in the familiar area. Finally we will discuss how visual cues may be explicitly encoded in the model in future work.