Background: Dormant propagule pools can serve as natural archives for reconstructing demographic and genetic change over time. Though soil-stored seed banks can be well-stratified reservoirs of genetic information, they are rarely leveraged to as a tool for examining temporal trends in genetic variation. Question: Here, we examine the utility of soil-stored seed banks for reconstructing genetic change over time. Methods: Seeds of the brackish marsh sedge Schoenoplectus americanus were sieved from 2-cm increments of soil cores taken from Kirkpatrick Marsh (Edgewater, MD) that were dated using 210-Pb and 137-Cs. Seeds were then germinated and resulting genets genotyped at 11 microsatellite loci. In addition to reconstructing seed density profiles and estimating germination rates across soil depths, we calculated measures of genetic variation and effective population size for each cohort for comparison. We also compared patterns of temporal genetic to variation among extant S. americanus genotypes in Kirkpatrick Marsh, from across Chesapeake Bay.
Results/Conclusions
Results: Our study showed that seed banks can serve as resources for seeds that can be germinated, grown out, and genotyped in numbers sufficient for characterizing genetic variation over time. We harvested, germinated and genotyped five cohorts spanning the 20th century. An average of 15 individuals were genotyped per downcore cohort, with a minimum of 5 individuals and a maximum of 40 individuals. However, cohorts from 6-8 cm (1990 ± 1.3) and 10-12 cm (1976 ± 1.2) exhibited the highest germination rates. There is a shift in alleles present in oldest to youngest soil cores, without any loss of genetic diversity. We also found evidence of significant differentiation between downcore and extant individuals in Kirkpatrick Marsh, and evidence of low im/emigration between Kirkpatrick and other Chesapeake marshes. Conclusions: Overall, soil layers are dateable, well-stratified, and contain seeds capable of germination. In addition, we detected clear temporal genetic shifts that are more likely due to local demography (e.g., in situ adaptive variation or drift) as opposed to attrition over time. Together, our results indicate that seed banks can serve as informative tools for studying how ecological and evolutionary processes unfold over time.