Inferring Bird Migration from Bone Isotopes and Histology: a Fossil-Friendly Methodological Framework

Abstract Bird seasonal migration is a remarkable biogeographic phenomenon, yet its deep-time origin(s) and evolutionary history remain poorly understood, with the bird fossil record largely overlooked. This study explores the predictability of bird migratory behaviour from the oxygen isotope composition of their bone apatite phosphate (δ 18 O p ), a promising approach in this regard because: (i) sedentary and migratory birds tend to occupy distinct climatic niches year-round; (ii) their δ 18 O p values primarily reflect the climate-driven isotopic composition of their drinking water; and (iii) this isotopic signature can persist through fossilisation. Bone tissues were categorised based on their potential to yield spatio-temporally distinct climatic records: Early Bone Tissues (EBT), deposited before somatic maturity, and Late Bone Tissues (LBT), formed through bone remodelling over the lifespan. The predictability of migratory behaviour was theoretically assessed by modelling tissue-specific δ 18 O p values for thousands of birds across 77 migratory and sedentary species, using tracking and observational data along with a revised phosphate-water fractionation equation. These theoretical results were confronted with data from 11 extant bird species, obtained using a new experimental framework combining histological and isotopic analyses. A significantly positive correlation between δ 18 O p and the proportion of LBT in bone samples was observed both theoretically and experimentally in migratory birds – particularly in long-distance migrants – and is predicted to be virtually absent in sedentary species at temperate latitudes. Migratory birds that died outside their natal sites can also be identified when their δ 18 O p,EBT values fall outside the range observed in local juvenile and sedentary birds. We conclude that bird migratory behaviour can be inferred from the δ 18 O p values of their skeletal remains, provided the individuals migrated across sufficiently contrasting climatic zones. While this approach cannot detect all migratory behaviours – such as short-distance or longitudinal movements – it is unlikely to misclassify sedentary birds as migratory at temperate latitudes. We therefore argue that this approach can be extended to well-preserved fossil bones to infer past bird migratory behaviours, as long as the palaeoclimatic context is carefully considered. More broadly, this study establishes a framework for inferring migratory behaviour in any vertebrate from non-fully remodelled biomineralised remains.