Aquatic Bacterial Communities Associated With Land Use and Environmental Factors in Agricultural Landscapes Using a Metabarcoding Approach.

Wen Chen,David R Lapen,Katarina D M Pintar,Edward Topp,Janis L Thomas,Julie T Chapados,Graham Wilkes,Izhar U H Khan,C André Lévesque

doi:10.3389/fmicb.2018.02301

Abstract

This study applied a 16S rRNA gene metabarcoding approach to characterize bacterial community compositional and functional attributes for surface water samples collected within, primarily, agriculturally dominated watersheds in Ontario and Québec, Canada. Compositional heterogeneity was best explained by stream order, season, and watercourse discharge. Generally, community diversity was higher at agriculturally dominated lower order streams, compared to larger stream order systems such as small to large rivers. However, during times of lower relative water flow and cumulative 2-day rainfall, modestly higher relative diversity was found in the larger watercourses. Bacterial community assemblages were more sensitive to environmental/land use changes in the smaller watercourses, relative to small-to-large river systems, where the proximity of the sampled water column to bacteria reservoirs in the sediments and adjacent terrestrial environment was greater. Stream discharge was the environmental variable most significantly correlated (all positive) with bacterial functional groups, such as C/N cycling and plant pathogens. Comparison of the community structural similarity via network analyses helped to discriminate sources of bacteria in freshwater derived from, for example, wastewater treatment plant effluent and intensity and type of agricultural land uses (e.g., intensive swine production vs. dairy dominated cash/livestock cropping systems). When using metabarcoding approaches, bacterial community composition and coexisting pattern rather than individual taxonomic lineages, were better indicators of environmental/land use conditions (e.g., upstream land use) and bacterial sources in watershed settings. Overall, monitoring changes and differences in aquatic microbial communities at regional and local watershed scales has promise for enhancing environmental footprinting and for better understanding nutrient cycling and ecological function of aquatic systems impacted by a multitude of stressors and land uses.

Highlights

Bacterial communities in freshwater are phylogenetically and metabolically diverse, being comprised of a complex mixture of common lineages within Actinobacteria, Proteobacteria, Bacteroidetes, Verrucomicrobia, Planctomycetes, and Firmicutes, candidate divisions (e.g., OD1, OP11, TM6, WS1, WS6), and genetic lineages unique to specific hydrological conditions (Zwart et al, 2002; Warnecke et al, 2004; Briée et al, 2007; Hu et al, 2014)
Empirical evidence in our studies suggests that compared with the cutoff determined by random matrix theory (RMT), the use of a larger threshold resulted in smaller networks with higher modularity (Supplementary Figures 12A,C and Supplementary Table S2), while a smaller threshold resulted in larger networks with strongly correlated modules being disguised by too many retained interconnections (Supplementary Figures 12B,D and Supplementary Table S2)
The current study investigated bacterial communities in freshwater from a suite of mixed-use but predominately agricultural watersheds

Summary

Introduction

Bacterial communities in freshwater are phylogenetically and metabolically diverse, being comprised of a complex mixture of common lineages within Actinobacteria, Proteobacteria, Bacteroidetes, Verrucomicrobia, Planctomycetes, and Firmicutes, candidate divisions (e.g., OD1, OP11, TM6, WS1, WS6), and genetic lineages unique to specific hydrological conditions (Zwart et al, 2002; Warnecke et al, 2004; Briée et al, 2007; Hu et al, 2014). The abundance of Polynucleobacter and Candidatus Planktophila limnetica were found to be associated with elevated nitrogen levels in urban streams, while Albidiferax, a Proteobacteria genus was considered an indicator of tetrachlorethene contamination of groundwater (Hosen et al, 2017) These pollutants may act as nutrient sources or toxins, depending on bacterial functional lineages pre-existing at or near a water body (Saxton et al, 2011; Jurelevicius et al, 2013). Due to the large number of usually dynamic and interacting factors that can affect community structure in natural systems, it is difficult to identify precisely causal effects (Lear and Lewis, 2009; Wang et al, 2011; Garrido et al, 2014; Van Rossum et al, 2015)

Methods

Results

Discussion

Conclusion