INCREASING AGE OF SPRING WATER: CLIMATE OR CONSUMPTION?

Groundwater represents the primary water resource in many karst terrains and its sustainability depends on a balance between recharge and withdrawal. We hypothesize that changes in karst groundwater resources may be reflected in shifts in the apparent age (i.e., residence time in the subsurface) of groundwater. To test this hypothesis, we dated water discharging from six springs in the Ichetucknee springs group in north-central Florida, USA over a 4-yr period ending in August 2013, a year after Tropical Storm Debby crossed the springshed. We compare our results with apparent ages previously measured between 1997 and 2003. Measured tracers included SF₆, ³H/³He, CFC-11, CFC-12, and CFC-113 but SF₆ and ³H/³He yielded discordant ages and we assume they are contaminated. The CFC values yielded concordant apparent ages of 20 to 45 yrs over the entire 16-yr sampling period. These ages divide the springs into two groups: Group 1 has younger apparent ages and higher dissolved oxygen and nitrate and lower Mg and sulfate concentrations than Group 2 springs. The different chemical compositions suggest long residence time lowered the redox state of the water and increased water-rock reactions. Tropical Storm Debby deposited 342 mm of rain between June 24 and 26, 2012 and increased flow by nearly 60% but apparent ages during the ~8 month long recession did not reflect discharge of young storm water. Instead apparent ages increased for both spring groups over the sample period, indicating most water recharged the aquifer around 1973 for the Group 2 springs and 1980 for Group 1 springs. These recharge dates correspond to the end of a 20-yr period characterized by ~2 m of rainfall in excess of average, followed by a 30-yr period with a ~4 m deficit from average. These changes in rainfall correspond to a shift from the cool to warm phase of the Atlantic Multi-decadal Oscillation around 1980, which tends to move winter El Niño rains from northern to southern Florida. Water withdrawal has remained nearly constant over the study period and represents only around 1 to 5% of the total spring discharge. We suggest that long-term climate variations impact discharge of the Ichetucknee springs more than withdrawals but withdrawal could aggravate natural climate variability.