Simple behavioural rules can explain the complex migration patterns of monarch butterflies

Abstract Many animals undertake seasonal migrations, yet the mechanisms by which they navigate remain poorly understood. For example, North American monarch butterflies (Danaus plexippus) travel upwards of 4500 km to specific overwintering sites in central Mexico. Leading hypotheses pose that monarchs either possess an innate ability to navigate to these sites (true navigation) or are migrating southwest for most of their journey and fine-tune their direction as they near their destination (vector navigation), partly guided by landscape features. Here, we employed agent-based models (ABMs) to investigate how different navigation strategies and landscape features influence predicted population patterns. We then compared the predicted distributions from these simulations to observed data from iNaturalist. We found that simpler simulations, where agents migrated southwest and adjusted their direction southward mid-journey (multi-vector navigation), aligned more closely with observed patterns than those with agents that knew where to go (true navigation). Our results suggest that true navigation is not needed to explain monarch migration and that the use of landscape funnels and a general southward movement can allow them to reach their overwintering sites. More broadly, we show that ABMs provide a useful tool for exploring alternative navigation strategies of migratory animals.