High‐frequency dissolved organic carbon and nitrate measurements reveal differences in storm hysteresis and loading in relation to land cover and seasonality

Abstract

Storm events dominate riverine loads of dissolved organic carbon (DOC) and nitrate and are expected to increase in frequency and intensity in many regions due to climate change. We deployed three high‐frequency (15 min) in situ absorbance spectrophotometers to monitor DOC and nitrate concentration for 126 storms in three watersheds with agricultural, urban, and forested land use/land cover. We examined intrastorm hysteresis and the influences of seasonality, storm size, and dominant land use/land cover on storm DOC and nitrate loads. DOC hysteresis was generally anticlockwise at all sites, indicating distal and plentiful sources for all three streams despite varied DOC character and sources. Nitrate hysteresis was generally clockwise for urban and forested sites, but anticlockwise for the agricultural site, indicating an exhaustible, proximal source of nitrate in the urban and forested sites, and more distal and plentiful sources of nitrate in the agricultural site. The agricultural site had significantly higher storm nitrate yield per water yield and higher storm DOC yield per water yield than the urban or forested sites. Seasonal effects were important for storm nitrate yield in all three watersheds and farm management practices likely caused complex interactions with seasonality at the agricultural site. Hysteresis indices did not improve predictions of storm nitrate yields at any site. We discuss key lessons from using high‐frequency in situ optical sensors.

Drying–Rewetting Cycles Affect Nitrate Removal Rates in Woodchip Bioreactors

“…The spectrophotometer and MPS measurement sequence was adjusted such that the stock tank and each column outflow concentrations were measured at 2‐h intervals. Nitrate concentrations were calculated from the measured absorbance following methods previously described (Birgand et al, 2016; Etheridge et al, 2014; Vaughan et al, 2017). …”

Concentration–discharge relationships describe solute and sediment mobilization, reaction, and transport at event and longer timescales

“…Other studies have used index‐based approaches to assess solute (Butturini, Alvarez, Bernal, Vazquez, & Sabater, ; Lloyd, Freer, Johnes, & Collins, ; M. C. H. Vaughan, Bowden, et al, ) and particulate (Aich, Zimmermann, & Elsenbeer, ; Langlois, Johnson, & Mehuys, ; Lawler, Petts, Foster, & Harper, ; Sherriff et al, ) hysteresis during events. Various quantitative hysteresis indices (HIs) exist that variously compare the C‐Q regression relationships on the rising and falling hydrograph limbs (Langlois et al, ), the ratio of constituent concentrations on the rising and falling limbs (Lawler et al, ), and the size of hysteresis loops (Aich et al, ; Lloyd et al, ; M. C. H. Vaughan, Bowden, et al, ). The HI proposed by Lloyd et al () and modified by M. C. H. Vaughan, Bowden, et al () uses normalized discharge and concentrations to quantify differences on the rising and falling hydrograph limbs.…”

“…Various quantitative hysteresis indices (HIs) exist that variously compare the C‐Q regression relationships on the rising and falling hydrograph limbs (Langlois et al, ), the ratio of constituent concentrations on the rising and falling limbs (Lawler et al, ), and the size of hysteresis loops (Aich et al, ; Lloyd et al, ; M. C. H. Vaughan, Bowden, et al, ). The HI proposed by Lloyd et al () and modified by M. C. H. Vaughan, Bowden, et al () uses normalized discharge and concentrations to quantify differences on the rising and falling hydrograph limbs. Values of HI range from −1 to +1, with negative and positive values indicating counterclockwise and clockwise hysteresis, respectively.…”

“…Because hysteresis patterns depend on the order, magnitude, and mixing of hydrologic source contributions, additional metrics are useful for characterizing the sequence of hydrologic source contributions to streams during events. The flushing index (FI) presented by M. C. H. Vaughan, Bowden, et al (), and similar to that of Butturini et al (), quantifies the directional change in concentration at the onset of stormflow and at peakflow; coupled with HI, this index provides insight into the timing, flushing behaviour, and proximity of biogeochemical sources to the stream channel. Solutes characterized by positive HI values (i.e., clockwise hysteresis) and negative FI values (i.e., decreasing concentrations on the rising limb) might indicate proximal, high‐concentration sources early in the storm followed by increasing mixtures with additional low‐concentration sources during peakflow.…”

“…In this study, we use hydrologic and biogeochemical data to characterize patterns in event‐scale hysteresis and interannual C‐Q relationships of dissolved and particulate constituents in streams. Based on multiple metrics related to the C‐Q relationship, including the slope (β) and CV C /CV Q of the log(C)–log(Q) relationship described by Musolff et al () and the HI and FI proposed by M. C. H. Vaughan, Bowden, et al (), we describe the biogeochemical source and transport dynamics responsible for generating observed patterns in solute and particulate C‐Q relationships. More specifically, we use a large hydrochemical dataset from White Clay Creek, a characteristic mid‐Atlantic piedmont watershed, with data encompassing pre‐event baseflow, stormflow, and multiple extreme events, to answer the following three questions about C‐Q relationships at multiple timescales: (a) What are the multistorm event to interannual patterns of solute and particulate C‐Q relationships in White Clay Creek?…”