DO CARBONATE SPRINGS EXHIBIT THRESHOLD DISCHARGE RESPONSES TO ANTHROPOGENIC STRESSORS?

Congruent weathering of carbonate bedrock leads to highly permeable and deep critical zones that feature rapid fluxes of solids, solutes, and gases through the subsurface. Discharge in carbonate systems may therefore respond more quickly to anthropogenic perturbations (e.g., climate or land use change) relative to silicate-dominated lithologies. However, few studies compare carbonate aquifers and instead focus on evaluation of single sites. To look for global trends for these systems, we extracted long-term records of annual spring discharge from the World Karst Spring hydrograph (WoKaS) database for Europe (37 sites total with datasets of ≥50 years) and the United States Geological Survey (USGS) database for North America (12 sites total with datasets of ≥34 years) that span discharge magnitudes, are located in varying climates, and feature rural and urban land uses. Rather than observing thresholds (i.e., inflection points) in the discharge records, we saw decreasing, constant, or increasing flow trends. In Europe, discharge remained constant over time in most of the springs we assessed (64.9%), with fewer sites featuring decreasing (32.4%) or increasing (2.7%) flow. Most of the North American springs also had constant (58.4%) flow over the observation period, but more frequently exhibited discharge trends that increased (33.3%), rather than decreased (8.3%), with time. Using the Meinzer (1923) spring discharge variability index, we found that spring flow in both Europe and North America was most frequently classified as “variable” (71.4%), followed by “somewhat variable” (28.6%). We also observed that the variability index did not predict long-term trends in the discharge responses of the selected springs. An ongoing challenge in our efforts to assess threshold responses in carbonate springs has been limited long-term discharge datasets, particularly for North America. Differences in site-specific climate, land use, and resource management practices (e.g., groundwater extraction) further complicate our attempts to interpret discharge trends over time given the relatively small datasets available from WoKaS and USGS. Our future research goals are to obtain more spring discharge archives from additional databases and to consider how site-specific attributes influence long-term flow behavior.