2018 ESA Annual Meeting (August 5 -- 10)

COS 47-7 - Urbanization and agriculture act as barriers to genetic connectivity in a fragmented serpentine barren grassland plant

Tuesday, August 7, 2018: 3:40 PM
353, New Orleans Ernest N. Morial Convention Center
Sarah L. Emel1,2, Shichen Wang3, Richard Metz3 and Rachel Spigler2, (1)Organismic and Evolutionary Biology Graduate Program, University of Massachusetts Amherst, Amherst, MA, (2)Biology Department, Temple University, Philadelphia, PA, (3)AgriLife Genomics and Bioinformatics Service, Texas A&M University, College Station, TX
Background/Question/Methods

The fragmentation of landscapes by human disturbance is one of the greatest threats to biodiversity globally. A critical component of conserving fragmented populations is understanding how landscape composition surrounding fragments impacts connectivity between them. The serpentine barren grasslands of the eastern US are a globally rare ecosystem characterized by harsh soils and unique plant communities that tolerate or adapt to them. Surrounded by a patchwork of encroaching forest, agriculture, and a gradient of urbanization, these highly fragmented grasslands represent an opportunity to explore the effects of human land use on population and landscape processes. We applied a landscape genetic approach to evaluate the relative resistance of these different land cover types to population connectivity and the extent to which they impact population genetic diversity in the facultative serpentine herb, Sabatia angularis. Using 4822 neutral single nucleotide polymorphisms (SNPs) developed by restriction site associated DNA sequencing (RAD-seq), we characterized spatial genetic structure and calculated pairwise-population genetic distance (FST) and population-level genetic diversity metrics. We then created 88 distinct land cover resistance layers and modeled their influence on FST using the distinct approaches of least cost paths and circuit theory. We also examined correlations between genetic diversity and surrounding landscape composition.

Results/Conclusions

Our results indicate that urban development and cropland act as major barriers to connectivity in S. angularis. Model selection criteria under both least cost path and circuit theory approaches identified models of isolation by environment as significantly better than the null model of isolation by distance. Additionally, resistance to urban development (both low and high intensity) was always equal to or greater than that of cropland, and these were always greater than resistance to forest and grassland (4:2:1:1 and 100:2:1:1 development:crop:forest:grassland resistance ratios based on circuit theory based and least cost paths, respectively). The relative impact of cropland becomes stronger when examining smaller spatial scales, within two distinct clusters of populations. We also found that the amount of development and cropland within 500 m of populations significantly influenced mean expected heterozygosity and inbreeding per population; heterozygosity decreased, while inbreeding increased, with surrounding land use. Our work therefore indicates that not only impervious surfaces like buildings and roads, but also low intensity development and cropland limit connectivity among serpentine barrens grasslands. Although grassland restoration is necessary to preserve these unique communities, our findings indicate that reforestation efforts may also be key to long-term population viability.