Urban infrastructure drives aquatic bacterial communities

Type of Presentation: 
Thu, 2017/11/16
What is the name of the event (if any) where you presented?: 
BYU Plant and Wildlife Sciences Department Graduate Research Conclave
Please enter a brief description of your presentation.: 

Urban infrastructure affects a critical but often overlooked aspect of stream ecosystems and water quality: bacterial communities. We were interested in learning how different types of urbanization altered aquatic bacteria communities along an elevational gradient (2368 m to 1353 m.a.s.l.) in three Utah watersheds—Red Butte Creek (RB), Provo River (PR), and Logan River (LR). Each of the three watersheds has one or more reservoirs for water resource management, ranging in surface area from 13.5 km2 to 0.38 km2 . The watersheds are relatively pristine and undeveloped at high elevations, with varying degrees of urbanization below the reservoirs: dense urban (RB), rapidly transitioning to urban from agriculture (PR), slowly transitioning to urban from agriculture (LR). We collected grab samples and filtered water from five sites in each watershed in November, February, and May to capture hydrologic and temporal variation on the effects of urbanization on bacterial communities. We extracted DNA from filters (MoBio Power Soil DNA kits), amplified, and sequenced the 16s rDNA V4 region using Illumina HiSeq and interpreted community composition using MOTHUR and R. Water quality parameters were analyzed by the USU Aquatic Biogeochemistry lab, the Carling lab (BYU Geology), or downloaded from multiparameter sondes installed at each sampling location. Water quality factors that were associated with bacterial community composition included specific conductivity, nitrogen (multiple species of N grouped together), dissolved organic carbon, and pH (P<0.001, Canonical Correlation Analysis). The bacterial communities notably decreased in richness and diversity below reservoirs with the exception of Logan River, where multiple smaller dams are spread over a long distance as opposed to a single large impoundment. In a test of community composition (how similar each community is to others), the interaction of location and watershed was significant (P=0.001, Anosim) while the interaction of location and season was not (P=0.589). A PCoA analysis of the community structure grouped below reservoir sites more tightly than watershed or season, except for LR sites. High elevation sites for RB and PR clustered with all LR sites, and subsequent downstream sites converged with the high elevation sites. This suggests that the presence and nature of these reservoirs is a significant determinant of bacterial diversity, and potentially function, more than downstream land use or changes in season or climate.