Proportional sampling strategy often captures more genetic diversity when population sizes vary

Collecting and conserving genetic diversity from plant populations for ex situ collections is a major objective of seed banks and botanic gardens. However, current guidelines for collecting germplasm to preserve plant species ex situ might not adequately capture the genetic diversity present in real plant population systems. Here, we tested a previously unexplored challenge for sampling practices: when populations of a rare plant species vary in size. We hypothesized that sampling proportional to the population size would capture more genetic diversity than sampling an equal number of individuals from every population. Using simulations of a hypothetical rare species, we tested equal and proportional strategies and calculated how many alleles were captured by either strategy. The effects of migration rate, recent bottlenecks, and sampling intensity on genetic diversity capture were also examined. We found that when population sizes differ (e.g., one population 3 times the size of median population size), proportional sampling captures more genetic diversity under constant size populations. The relatively modest improvement (1–5% more allelic diversity for most cases) was observed across all parameters of migration and intensity tested–except when there were recent bottlenecks. We also created simulations tailored to three IUCN Red List threatened oaks (Quercus oglethorpensis, Q. engelmanii, and Q. acerifolia) and found similar results for these detailed case studies as for our generic simulations. We conclude that sampling proportional to population size may often be a useful sampling strategy to create genetically diverse ex situ plant populations, ultimately resulting in more efficient use of resources.