Caffeine as an Anthropogenic Source Indicator in Freshwater and Marine Systems

Bacteria serve as traditional indicators of water quality. In the past decade, methods have been developed to explore the use of more human specific indicators. One of these suggested indicators is caffeine. Caffeine is widely consumed in populated areas and has already been proven as a successful indicator in freshwater systems. This project aimed to modify the caffeine extraction method to obtain ng/L detection limits and prove its usefulness as a tracer in marine systems. Optimal extractions were performed with solid phase extraction using traditional C-18 sorbent columns. Extracts were then run on a GC/MS. Freshwater samples were examined in this project to determine a background level of caffeine in relatively isolated areas. Analysis of samples from creeks in southwest Georgia did not detect caffeine. An isolated wetland sample did contain caffeine and led to the identification of a natural local source of caffeine, the Yaupon Holly. Samples were also collected along the Ochlockonee river system. Some of its tributaries known to have treated wastewater effluent were also sampled. Samples from these tributaries contained measurable caffeine, but caffeine was not found in the main stem. The marine system sampled was Sarasota Bay, Florida. There was no interference of salts in marine samples to compromise the method. Caffeine concentrations were found to be highest along the mainland side of the bay, where septic systems were suspected sources of contamination. Caffeine had an inverse relationship with salinity and conductivity and positive correlations with nutrient concentrations. Fecal coliforms correlated well with caffeine, but not with the more resilient enterococcus bacterium. Samples of groundwater were analyzed and proved a definite source of caffeine to this system. A degradation experiment found that in a natural sample of marine water exposed to sunlight, caffeine has a half-life of approximately 30 days. However, for samples kept in the dark the half-life increases to 71 days, and samples refrigerated at a constant 4oC and in the dark have a half-life of 178 days.