Processes Affecting Nitrogen Speciation in a Karst Aquifer

Like many karst aquifers, the Barton Springs segment of the Edwards aquifer, in central Texas, is in an area undergoing rapid growth in population, and there is concern as to how increased amounts of wastewater might affect groundwater quality. We measured concentrations and estimated loads of nitrogen (N) species in recharge to and discharge from the Barton Springs segment of the Edwards aquifer, central Texas, to evaluate processes affecting the transport and fate of N species in groundwater. Water samples were collected during 17 months (November 2008-March 2010) from five streams that contribute about 85% of recharge to the aquifer segment and from Barton Springs, the principal point of discharge from the segment. The sampling period spanned a range of climatic conditions from exceptional drought to above-normal rainfall. Samples were analyzed for N species (organic N + ammonia, ammonia, nitrate + nitrite, nitrite); loads of organic N and nitrate were estimated with LOADEST, a regression-based model that uses a time series of streamflow and measured constituent concentrations to estimate constituent loads. Concentrations of organic nitrogen and dissolved oxygen were higher and concentrations of nitrate were lower in surface water than in spring discharge, consistent with conversion of organic nitrogen to nitrate and associated consumption of dissolved oxygen in the aquifer. During the period of the study, the estimated load of organic N in recharge from streams (average daily load [adl] of 39 kg/d) was about 10 times that in Barton Springs discharge (adl of 9.4 kg/d), whereas the estimated load of nitrate in recharge from streams (adl of 123 kg/d) was slightly less than that in Barton Springs discharge (adl of 148 kg/d). The total average N load in recharge from streams and discharge from Barton Springs was not significantly different (adl of 162 and 157 kg/d, respectively), indicating that surface-water recharge can account for all of the N in Barton Springs discharge. Differences in the timing of recharge and discharge of loads were marked, however: During the drought period, loads of total N in recharge from streams (adl 2.4 kg/d) were much less than those in Barton Springs discharge (adl 70 kg/d), and during the wet period those in recharge from streams exceeded those in Barton Springs discharge (average daily loads of 395 and 282 kg/d, respectively). This indicates that total N entering the aquifer during periods of high recharge likely goes into groundwater storage, then discharges gradually as water-level altitudes in the aquifer decrease.