Data and R script from "Investigating a potential bias in resurrection experiments to measure adaptive evolution of flowering time"

Please refer to the README file for descriptions of files and organization.<b>Article citation:</b><b>Abstract:</b> Species persistence under climate change often depends on rapid evolutionary responses to increasingly extreme conditions. Resurrection experiments are a valuable approach for tracking the rate of adaptive evolution: ancestral propagules are stored, then later revived and grown alongside descendants, exposing shifts in trait values that indicate evolutionary change. However, even under the best storage conditions, ancestral propagules age and eventually a fraction die. If propagule longevity in storage is non-random and genetically correlated with a trait of interest, bias is introduced to the ancestral trait baseline, consequently distorting estimates of adaptive evolution. To understand the degree to which non-random ancestral propagule loss biases resurrection experiments, we simulated long-term seed aging of the annual Brassica rapa, testing whether seed “birth order”, a proxy for maternal resource allocation, influences seed survivorship. We then evaluated whether birth order correlates with flowering time, a trait previously demonstrated to evolve rapidly in response to climate change. Rapidly aged seeds had 26% reduced survivorship and flowered approximately one day later than unaged control seeds. This plastic storage effect disappeared in the F2 generation. Furthermore, simulations testing differential storage survivorship with birth order (i.e. between first and last produced seeds) reveal the bias on flowering time is low on average (&lt;1 day), even under extreme mortality asymmetries. Together, results from this study system suggest the resurrection approach remains a reliable experimental methodology in global change biology.