On-site, subannual climate records decouple Neolithic expansion from climate instability

How the life of prehistoric people was influenced by climate change and extreme events often remains elusive. Finding appropriate spatiotemporal scales is the major challenge for studies with socioenvironmental questions. This challenge can be overcome by sourcing climatic information from archaeological sites, alongside traces of human agency. For instance, extreme cold-events have been proposed as causing a standstill in early farming expansion across the Mediterranean 8600-8000 years ago. However, there is no immediate (local/short-term) evidence of such events and how they might have influenced people's lives, nor have previous records been sufficiently resolved to test this robustly. Here we statistically assess the influence of on-site and short-term extremes on human decision-making hypothesized by long-term climate records. Using the largest dataset yet of high-resolution temperature archives in mollusks, we test whether early farmers would have been exposed to extreme cold-events, compromising their subsistence by reducing harvest yields. This immediate agricultural risk is argued to have prevented the spread of new settlements and thus delayed the vigorous expansion that followed, one of prehistory's fastest known dispersals. Direct comparison of paleotemperature and subsistence data from the same archaeological contexts shows no winter cooling of the magnitude proposed for this period, and no strong relationship between climate and subsistence. Our study reveals how a major scale mismatch in climate-society research can be resolved at the site level, integrating climatic variability on the subannual timescale crucial to farming, and provides a direct, local empirical test of a long-assumed climate impact.