Background/Question/Methods Increased impervious surface cover, leaky sewage systems and industrial and municipal discharge are consequences of urbanization that have substantial effects on both the abiotic and biotic properties of urban stream systems. Watershed-scale strategies to mitigate the chemical impact of urban runoff currently focus on reducing sources and/or increasing sinks, areas or processes that promote the uptake, immobilization and transformation of pollutants by physical, chemical or biological agents.
Studies have shown woody debris, the residue of living trees such as logs, branches, and whole fallen trees, to be a sink for pollutants in streams. For example, organic debris dams, containing wood, have been found to act as nitrogen cycling hotspots in urban streams. The role of wood as a pollutant sink can, in part, be attributed to the microbial biofilms found on its surface (epixylic biofilms). Microbial communities are thought to have high levels of taxonomic diversity and functional redundancy; suggesting a high level of functional resilience to environmental change. While land use and environmental properties, specifically impervious surface cover and pH have been shown to alter the composition of microbial communities, little is known about the impact of these environmental gradients on microbial community function in urban streams. This study seeks to characterize the taxonomic and functional composition of bacterial communities found in epixylic biofilms from streams spanning an urbanization gradient, in Baltimore Maryland, to determine whether microbial community function and composition respond in similar ways to environmental gradients.
Results/Conclusions
Epixylic biofilm samples were collected in the summer of 2015 from 20 logs in 5 streams in Baltimore, MD. DNA was extracted from biofilm samples and Bacterial 16S rRNA genes were PCR amplified and sequenced on an Illumina MiSeq. 16S sequences were assigned to operational taxonomic units (OTUs) using Mothur and OTUs were annotated using the Greengenes reference taxonomy. PICRUSt was used to predict metagenome functional content from 16S OTUs.
Unlike the NMDS ordination of predicted metagenomes, which showed no clear clustering pattern by stream site, an NMDS ordination of annotated taxa revealed a clustering of sampled communities by stream site. Fitting environmental variables to ordination scores revealed that community composition was correlated with environmental gradients. Stream pH, chloride concentration, impervious surface and tree canopy cover were identified as the environmental conditions most strongly associated with microbial community structure while canopy cover and nitrogen concentration were most strongly associated with community function.
