Species ranges likely different & smaller than today
Ranges shrank c. 25% from Pleistocene to Holocene
Range of crop progenitors extended to Western Anatolia, Cyprus, Aegean
Roe, J. & Arranz-Otaegui, A., 2026. Biogeography of crop progenitors and wild plant resources in the terminal Pleistocene and Early Holocene of West Asia, 14.7–8.3 ka. Open Quaternary 12. https://doi.org/10.5334/oq.163
Is the model wrong?
Heavy spatial bias in occurrence data
No hyperparameter tuning
Is the archaeology wrong?
Small samples
Assemblages aggregated by site-phase
Poor chronological control
…or is the concept wrong?
Fundamental vs. realised niche
Plasticity unaccounted for
Human impact unaccounted for
Time-averaged in different ways
Summary
Better model
Spatially stratified sampling
Per-species hyperparameter tuning
Better training data
More biodiversity surveys in the Middle East
Century-scale palaeoclimatologies
Climatic uncertainty/interannual variability
Better verification
Context-level archaeobotanical database
Refined chronology
But can ecological models predict the occurrence of species in the archaeological record?
Can ecological models predict the occurrence of species in the archaeological record? Can I?
