PS 74-160
Fragmentation affects germinable soil seed banks from South Dakota Badlands

Friday, August 15, 2014
Exhibit Hall, Sacramento Convention Center
Kelsey Ducheneaux, Department of Natural Resource Management, South Dakota State University, Brookings, SD
Lora Perkins, Natural Resource Management, South Dakota State University, Brookings, SD
Background/Question/Methods

Fragmentation of wildlands is one of the most important components of global change with profound but critically under examined impacts. A need exists to understand how fragmentation impacts vegetation dynamics in order to mitigate and adapt management as land-use change and climate change progress. This research will develop a mechanistic understanding of how seed bank dynamics are altered by fragmentation

South Dakota Badlands contain characteristic geological features of vegetated relict pediments (colloquially, ‘Sod Tables’), formed by erosion of the surrounding land to create a distinct island/matrix landscape. Each Sod Table is a remnant fragment, or island, of the prairie surrounding the Badlands. A matrix of sparsely vegetated, nutrient-poor clay soils surrounds Sod Tables.  Thus, Sod Tables are a unique system to examine the effects of fragmentation on soil seed banks.

To examine the effects of fragmentation on soil seed banks, we collected samples from the top of Sod Tables and from the surrounding matrix.  We germinated the samples for 5 months with alternating cycles of wet and dry.  All germinates were identified to species.  Species richness and diversity were calculated and similarity of the germinable seed bank was compared between sod tables and the surrounding matrix. 

Results/Conclusions

Overall, 29 species were germinated from the seed banks including species not previously recorded in South Dakota (Malcolmia africana).  Assemblages of species that emerged from the sod table samples and from the matrix samples were dissimilar.  Between each sod table-matrix pair Jacquards Index of Similarity values range 0-0.25 and Bray-Curtis Dissimilarity Index values range 0.5-1.  When pooled all sod tables and matrix samples again were dissimilar (Jacquards= 0.18 , Bray-Curtis= 0.81). 

These results suggest that the vegetation communities on sod tables are distinct from communities in the surrounding matrix and that fragmentation may have substantial effects on soil seed banks.