Effect of source integration on the geochemical fluxes from springs

Geochemical fluxes from watersheds are typically defined using mass-balance methods that essentially lump all weathering processes operative in a watershed into a single flux of solute mass measured in streamflow at the watershed outlet. However, it is important that we understand how weathering processes in different hydrological zones of a watershed (i.e., surface, unsaturated, and saturated zones) contribute to the total geochemical flux from the watershed. This capability will improve understanding of how geochemical fluxes from these different zones may change in response to climate change. Here, the geochemical flux from weathering processes occurring solely in the saturated zone is investigated. This task, however, remains exceedingly difficult due to the sparsity of subsurface sampling points, especially in large, remote, and/or undeveloped watersheds. In such cases, springflow is often assumed to be a proxy for groundwater (defined as water residing in fully saturated geologic formations). However, springflow generation may integrate different sources of water including, but not limited to, groundwater. The authors’ hypothesis is that long-term estimates of geochemical fluxes from groundwater using springflow proxies will be too large due to the integrative nature of springflow generation. Two conceptual models of springflow generation are tested using endmember mixing analyses (EMMA) on observations of spring chemistries and stable isotopic compositions in a large alpine watershed in the San Juan Mountains of southwestern Colorado. In the “total springflow” conceptual model, springflow is assumed to be 100% groundwater. In the “fractional springflow” conceptual model, springflow is assumed to be an integration of different sources of water (e.g., groundwater, unsaturated flow, preferential flow in the soil, etc.) and groundwater is only a fractional component. The results indicate that groundwater contributions in springflow range from 2% to 100% overall and no springs are consistently composed of 100% groundwater; providing support for the fractional springflow conceptual model. Groundwater contributions are not strongly correlated with elevation, spring contributing area, spring discharge, or seasonality. This variability has a profound effect on long-term geochemical fluxes. The geochemical fluxes for total springflow overestimate long-term solute release by 22–48% as compared to fractional springflow. These findings illustrate that springflow generation, like streamflow generation, integrates many different sources of water reflecting solute concentrations obtained along many different geochemical weathering pathways. These data suggest that springs are not always ideal proxies for groundwater. Springs may be integrating very distinct portions of the groundwater flow field and these groundwater contributions may become mixed at the spring emergence with much younger sources of water that have never resided in the groundwater system.