Recent Demographic History Inferred From Allele Frequency Spectra: Evaluation by Simulations and Insights From Pool Sequencing of Wild Bees

Recent insect declines have raised global concern, but assessing insect population trends remains challenging due to limited long-term monitoring data. Understanding the timing of population changes is essential for discerning the relative impacts of drivers like land use modification and climate change. Importantly for conservation, species that have undergone recent declines should ideally also be distinguished from species that have been rare throughout evolutionary time. Here, we first evaluated the performance of Stairway Plot 2 using simulations of different effective population sizes (N_e), sample sizes and timing and magnitude of declines. We found reliable reconstructions over the past 100-1000 years for populations with smaller N_e (5000 and 50,000) using ≥ 40 diploid individuals. Next, we applied pooled whole-genome sequencing data to reconstruct demographic trajectories for two common and two rare and endangered species of wild bees. The two rare species showed historically lower N_e compared to the two common species, but their N_e trajectories varied in the timing and magnitude of population size changes. We further show that the use of reference genomes from non-focal but closely related species can inform N_e trajectories with only slight deviations from when using focal reference genomes, while more distantly related reference genomes compromised reliability. Our findings represent an important step towards reconstructing population size changes using genomic data, offering a long-term perspective on species rarity, current conservation status and potential drivers of ongoing insect decline.