Location: Plant Science ResearchTitle: Capturing the Spatial Variability of Microbial Communties within Agricultural Soils
|CASTLE, SARAH - University Of Minnesota|
|GUTKNECHT, JESSICA - University Of Minnesota|
|KINKEL, LINDA - University Of Minnesota|
|ROSEN, CARL - University Of Minnesota|
|SADOWSKY, MICHAEL - University Of Minnesota|
|Samac, Deborah - Debby|
Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 10/21/2015
Publication Date: 10/21/2015
Citation: Castle, S., Gutknecht, J., Kinkel, L., Rosen, C., Sadowsky, M., Samac, D.A. 2015. Capturing the Spatial Variability of Microbial Communties within Agricultural Soils. Annual Argonne Soil Metagenomics Meeting. October 21-23, 2015. Lisle, IL.
Technical Abstract: Understanding patterns in spatial variability of soil microbial communities can provide important insights into the mechanisms that control ecosystem function. The level of replication required to adequately characterize the variability of soil communities across both small and large geographic and edaphic gradients, however, is not well understood. Here we asked: How extensive does sampling need to be in order to capture variability across different spatial scales? Furthermore, how much information do we lose by decreasing sampling effort? To address this, we utilized a recently established agricultural research network in three geographically and edaphically distinct sites across Minnesota: Grand Rapids, Waseca, and Lamberton. At each site, we identified five 24 x 24 meter plots and collected surface soil from six individual locations per plot. Illumina sequencing of 16S rRNA genes (V4 region) was used of to characterize sensitivity of community data to sample pooling (physical or computational). We rarefied individual samples and physical composites to a depth of 20,000 sequences per sample, and computational composites were created by first summing OTU occurrences across individual samples and then rarefying to this same depth. Community richness and diversity were greater in computationally-pooled soils than either individual or physically-pooled soils for two of the three sites (Grand Rapids and Waseca). By contrast, community richness and diversity did not differ between the two methods of compositing for Lamberton and a single physical composite provided a sufficient assessment of both richness and diversity. The effort and expense associated with extracting and sequencing replicate soil samples is often a motivation for physically pooling spatially explicit samples. However, our results suggest that the effectiveness of pooling may vary by soil type.