Correlations between groundwater bacteria types and geochemistry of springs in Nashville, Tennessee

Very little is known about the connection between bacteria, groundwater geochemistry and water quality in karst terrains. The objective of this project was to determine if there was a correlation between bacteria types and water chemistry in two limestone bedrock springs draining an urban basin in Nashville, Tennessee. These springs, Tumbling Rock and Trough Springs are located in a karst terrain that is highly susceptible to contamination due to sinkholes and modified surface hydrology. The two springs were sampled every 1 to 4 weeks from June 2007 through March, 2009. The flow from the springs converged, and the combined discharge ranged from 150,000 gallons/day during a severe drought to 1 million g/d. An unexpected waste leak from a poultry-waste storage unit caused ammonia levels to rise to 40 mg/L for several weeks in the spring water. Additional water-quality parameters measured included temperature, specific conductance, dissolved oxygen, pH, sulfate, nitrogen, and bacteria Biological Activity Reaction Tests (BART). A continuous water-quality monitoring device was installed in Tumbling Rock spring to measure changes associated with different weather patterns. The pH values ranged from 6.5 to 8, temperature ranged from 16.5°C to 19°C, and specific conductance values were approximately 20% higher in the winter. There was a rapid rise in ammonia-oxidizing bacteria affiliated with the high ammonia levels, but both subsided after the leak was fixed. Additional analysis of bacteria types and geochemistry found a strong positive correlation between sulfur-related bacteria and sulfate concentrations (R-sq = 0.82). There was a moderate negative correlation between iron-oxidizing bacteria and dissolved iron concentration (R-sq = 0.25), which means when dissolved iron was low, iron-oxidizing bacteria were high. There was a moderate positive correlation between heterotrophic aerobic bacteria (HAB) and rain events, however it required a 10 day waiting period between rain and increased HAB (R-sq = 0.41). This 10 day delay probably reflects the time it takes for rain to enter the aquifer and stimulate bacteria growth with food. These data show there is a connection between geochemical conditions and bacteria populations in karst terrain