Long‐term assessment of nutrient flow pathway dynamics and in‐stream fate in a temperate karst agroecosystem watershed

Abstract

Nutrient dynamics in karst agroecosystems remain poorly understood, in part due to limited long-term nested datasets that can discriminate upland and in-stream processes. We present a 10-year dataset from a karst watershed in the Inner-Bluegrass Region of central Kentucky, consisting of nitrate (nitrate-N [NO 3 − ]), dissolved reactive phosphorus (DRP), total organic carbon (TOC), and total ammoniacal-N (TAN) measurements at nested spring and stream sites as well as flowrate at the watershed outlet. Hydrograph separation techniques were coupled with multiple linear regression and Empirical Mode Decomposition time-series analysis to determine significance of seasonal processes and to generate continuous estimates of nutrient pathway loadings. Further, we used model results of benthic algae growth and decomposition dynamics from a nearby watershed to assess if transient storage in algal biomass could explain differences in spring and downstream watershed nutrient loading. Results highlight statistically significant seasonality for all nutrients at stream sites, but only for NO 3 − at springs with longitudinal variability showing significant decreases occurring from spring to stream sites for NO 3 − and DRP, and significant increases for TOC and TAN. Pathway loading analysis highlighted the importance of slow flow pathways to source approximately 70% of DRP and 80% of NO 3 − . Results for in-stream dynamics suggest that benthic autotroph dynamics can explain summer deviations for TOC, TAN, and DRP but not NO 3 − . Regarding upland dynamics, our findings agree well with existing perceptions in karst for N pathways and upland source seasonality but deviate from perceptions that karst conduits are retentive of P, reflecting the limited buffering capacity of the soil profile and conduit sediments in the Inner-Bluegrass. Regarding in-stream fate, our findings highlighted the significance of seasonally driven nutrient processing in the bedrock-controlled streambed to influence nutrient fluxes at the watershed outlet. Contrary to existing perceptions, we found high N attenuation and an unexplained NO 3 − sink in the bedrock stream, leading us to postulate that floating macrophytes facilitate high rates of denitrification. KEYWORDS bedrock stream, hydrologic pathways, in-stream fate, karst agroecosystem watershed, nutrient loadings, time-series analysis

Impacts of preferential flow and agroecosystem management on subsurface particulate phosphorus loadings in tile‐drained landscapes

“…In the next step, a linear increase in slow flow was assumed from the beginning of the rising limb of the hydrograph, which represented the start of quickflow, to the determined inflection point on the falling limb from the previous step; this point signified the separation of quick and slow flow (Husic, 2018). Finally, event contribution by each pathway was calculated as the area between the two curves for the quick flow pathway and the area under the curve for the slow flow pathway (Ford et al., 2019). To quantify the impact of flowrate and quickflow on TP − DRP concentrations, we performed a multiple linear regression analysis.…”

“…Instead of selecting fixed functional forms of trends, the trends are adaptive over time (Ford et al., 2015; Wu et al., 2007). As a result, the method has recently been applied to nutrient concentration and flow datasets in karstic and tile‐drained watersheds (Ford et al., 2015, 2018, 2019). Although there is perceived utility for analyzing tile drain nutrient signals, application at the field scale has been limited, in part, by a lack of continuous long‐term datasets.…”

“…Hydrograph recession analysis is an empirically based flow partitioning approach used in karst springs that has applicability to tile‐drained landscapes (Ford, Husic, Fogle, & Taraba, 2019; Husic et al., 2019; Jarvie et al., 2014; Schilling & Helmers, 2008). In hydrograph recession, the receding limb of the hydrograph is conceptualized as the drainage of a series of reservoirs that have variable hydraulic conductivities and storage volumes (Husic et al., 2019).…”

“…A master recession curve can be generated for a site by compiling events from long‐term monitoring data to determine the number of statistically differentiable reservoirs in a system (Gregor & Malik, 2012). Hydrograph recession can also be applied on an event‐by‐event basis to quantify temporal variability in flow pathway dynamics (Ford et al., 2019; Jarvie et al., 2014). The applicability of hydrograph recession to tile‐drained landscapes is recognized given that reservoir‐style hydrologic models have been applied to tile‐drain hydrographs to reflect quick preferential flow through macropores and slow diffuse percolation through the soil matrix (Brauer, Teuling, Torfs, & Uijlenhoet, 2014; Ford et al., 2018).…”

Optimal Transport for Assessing Nitrate Source‐Pathway Connectivity

“…For example, since sinkholes and swallets pirate surface water fairly quickly, the time duration of quickflow connection is similar to that of surface runoff, which has been approximated to be no more than 2 days in the region (Mahoney et al, 2018). On the other hand, subsurface fracture and matrix networks can become saturated and their flow paths connected for much longer durations (Ford et al, 2019). This result indicates that functional (dis)connectivity of nitrate pathways is temporally variable since connectivity of nitrate source pathways is mediated hydrologically (Ali et al, 2018; Zingaro et al, 2019).…”