JOURNAL OF THE AMERICAN WATER RESOURCES ASSOCIATION AMERICAN WATER RESOURCES ASSOCIATION Sediment Fingerprinting to Delineate Sources of Sediment in the Agricultural and Forested Smith Creek Watershed, Virginia, USA A.C. Gellis and L. Gorman Sanisaca Research Impact Statement: Sediment fingerprinting helps identify and apportion sediment sources, includ- ing sediment derived from top soil and eroding streambanks, which are often overlooked sources of sediment to streams. ABSTRACT: The sediment fingerprinting approach was used to apportion fine-grained sediment to cropland, pasture, forests, and streambanks in the agricultural and forested Smith Creek, watershed, Virginia. Smith Creek is a showcase study area in the Chesapeake Bay watershed, where management actions to reduce nutri- ents and sediment are being monitored. Analyses of suspended sediment at the downstream and upstream sam- pling sites indicated streambanks were the major source of sediment (76% downstream and 70% upstream). Current management strategies proposed to reduce sediment loadings for Smith Creek do not target stream- banks as a source of sediment, whereas the results of this study indicate that management strategies to reduce sediment loads in Smith Creek may be effective if directed toward managing streambank erosion. The results of this study also highlight the utility of sediment fingerprinting as a management tool to identify sediment sources. (KEYWORDS: sediment fingerprinting; bank erosion; Chesapeake Bay; sediment TMDL.) INTRODUCTION and other contaminants (Owens et al. 2001; Gerbers- dorf et al. 2011). Sediment is a major contributor to ecological Worldwide, sediment is an important pollutant degradation in Chesapeake Bay (Gellis and Brakebill degrading aquatic habitat and impacting infrastruc- 2013). Smith Creek, along with two other streams in ture, such as reservoirs (Strayer and Dudgeon 2010; the Chesapeake Bay watershed, was selected by the Liu et al. 2017). In the United States (U.S.), sediment U.S. Department of Agriculture as a “showcase” is one of the leading causes of stream impairment study area, meaning that if successfully restored, it (USEPA 2017). Fine sediment can reduce light pene- would become a model for restoration efforts in the tration and suppress primary production in algae and Chesapeake Bay watershed (Epes 2010; Jenner 2010; macrophytes (Yamada and Nakamura 2002; Izagirre USDA-NRCS 2017). Biological monitoring conducted et al. 2009; Jones et al. 2012). Deposited sediment by the Virginia Department of Environmental Qual- can bury channel substrate and degrade habitat for ity indicated Smith Creek was violating the state’s macroinvertebrates (Jones et al. 2012) and fish (Sear general standard for aquatic life use where the et al. 2016). In addition, fine sediment provides a stream should support the propagation and growth of transport vector for bound nutrients, heavy metals, a balanced indigenous population of aquatic life Paper No. JAWRA-17-0153-P of the Journal of the American Water Resources Association (JAWRA). Received November 28, 2017; accepted July 19, 2018. © 2018 American Water Resources Association. This article is a U.S. Government work and is in the public domain in the USA. Discussions are open until six months from issue publication. Maryland-Delaware-DC Water Science Center (Gellis, Gorman Sanisaca), U.S. Geological Survey, Baltimore, Maryland, USA (Correspon- dence to Gellis: [email protected]). Citation: Gellis, A.C., and L. Gorman Sanisaca. 2018. “Sediment Fingerprinting to Delineate Sources of Sediment in the Agricultural and Forested Smith Creek Watershed, Virginia, USA.” Journal of the American Water Resources Association 1–25. https://doi.org/10.1111/1752- 1688.12680. JOURNAL OF THE AMERICAN WATER RESOURCES ASSOCIATION 1 JAWRA GELLIS AND GORMAN SANISACA (VADEQ 2009). The primary stressor on the aquatic beds) (Gellis and Walling 2011; Gellis et al. 2016). Dif- community was identified as sediment. In 2004, a ferentiating between these two broad categories (up- Total Maximum Daily Load (TMDL) was developed land and channel sources) is important because for Smith Creek to reduce sediment loadings (VADEQ sediment-reduction management strategies differ by 2009). The successful mitigation of sediment-related source and require very different approaches — reduc- impairments requires knowledge of the contribution ing agricultural sources may involve soil conservation of sediment from its various sources. and tilling practices, whereas reducing channel sources Attempting to relate sediment to its sources is a of sediment may involve stream restoration, bank stabi- difficult task. Approaches for estimating sediment lization, and (or) grade control to arrest downcutting. sources include sediment budget studies (Gellis and The objective of this study was to identify the relative Walling 2011; Gellis et al. 2016), geographic informa- contributions of sediment from cropland, pasture, for- tion system (GIS) and photogrammetric analysis (Fer- est, and streambanks in the Smith Creek watershed nandez et al. 2003; Curtis et al. 2005; Roering et al. using the sediment fingerprinting approach. 2013; Hackney and Clayton 2015; Gellis et al. 2016), models (Aksoy and Cavvas 2005; USEPA 2008), and the use of geochemical tracers or fingerprints (Wall- ing 2005; Gellis and Walling 2011; Mukundan et al. MATERIALS AND METHODS 2012; Collins et al. 2017). Each of these approaches has its advantages and disadvantages. Sediment bud- get approaches often rely on field measurements, Study Area which can provide useful data on erosion and deposi- tion rates, but are labor intensive, and can be spa- Smith Creek drains the Valley and Ridge Province tially limited. Photogrammetry/GIS analysis and in the Chesapeake Bay watershed, with land use in model analysis, although less intensive in terms of the area draining to the downstream station (in labor and time, may produce a wide range of results 2011) consisting of forest, 48%; pasture, 41%; devel- that need to be validated with data collected from the oped, 8%; and cropland, 3% (Figure 1) (Homer et al. watershed of interest. Ground-based and airborne 2015). The area draining to the downstream station lidar, as well as structure-from-motion photogramme- is underlain by dolostone and limestone (66%) and try with handheld cameras and unmanned aerial sys- sandstone and shale (34%) (Dicken et al. 2005). Ele- tems (drones), are being increasingly used to describe vations range from 270 m at the lower reaches to channel morphology and topographic change (Faux 890 m in the Massanutten Mountains on the eastern et al. 2009; Caroti et al. 2013; Roering et al. 2013; side of the watershed. In the area draining to the Caroti et al. 2015). The resultant scans, point clouds, upstream station, land use in 2011 was crop 2%; pas- and digital elevation maps can be overlain chronologi- ture 30%; forest 66%; and other 2% (Homer et al. cally to quantify the erosion and deposition of various 2015), and bedrock is 58% dolostone/limestone and channel sources. However, these techniques often 42% sandstone/shale (Dicken et al. 2005). Average require field validation and the representation of annual daily discharge recorded at the U.S. Geologi- morphological elements requires a high point density cal Survey (USGS) gage, 1961–2016, was 2.14 m3/s with large data processing demands. In addition, res- (USGS 2016). Precipitation in the watershed mea- olution and scale may restrict the applicability of sured at the Dale Enterprise rain station near Har- these techniques to quantify topsoil erosion and ulti- risonburg, Virginia, averaged 915 mm/yr with mately do not quantify the delivery of these sediment temperatures ranging from a July mean of 23°Ctoa sources out of the watershed. January mean of 0.44°C (University of North Caro- Sediment fingerprinting is an approach that has lina 2012). Most portions of Smith Creek are mean- been increasingly utilized to assist managers in identi- dering, pool-riffle systems on gravel to sand beds fying sources of sediment in a watershed (Collins, with occasional bedrock outcrops. Varying thick- Walling, et al. 2010; Mukundan et al. 2012; Miller nesses of sediment in channel storage (on the channel et al. 2015; Collins et al. 2017). The sediment finger- bed) were observed at select reaches, most noticeably printing approach entails the identification of specific in pools. Fine sediment was also present within the sediment sources through the establishment of a mini- interstitial coarse substrate. mal set of physical and (or) chemical properties that uniquely define each source in the watershed. In gen- eral, sediment fingerprinting results can provide infor- The Sediment Fingerprinting Approach mation on the relative contribution of upland (soil erosion from various land use and land cover types) The sediment fingerprinting approach provides a vs. channel contributions (streambanks and channel direct method for quantifying watershed sources of JAWRA 2 JOURNAL OF THE AMERICAN WATER RESOURCES ASSOCIATION SEDIMENT FINGERPRINTING TO DELINEATE SOURCES OF SEDIMENT IN THE AGRICULTURAL AND FORESTED SMITH CREEK WATERSHED,VIRGINIA, USA 78° Watershed boundary draining to F17 35' the North Fork Shenandoah River F16 F9 F18 38°45' B20 C1 C7 38°45' B2 P1 P20 C8 C12 P15 B16 P8 B7 B5 Smith Creek B6B F8 Watershed B4 C6 F10 B17 F11 B3 78° B8 F19 40' 38°40' C19 B9 C2 C3 38°40' Watershed boundary draining to the USGS station B18 C13 P2 B10 P9 EXPLANATION P3 C9 Land-use Sample points P10 C15 class source type B19 P15 P17 C14 Other Bank P7 B11 38°35' P11 Cropland Crop P12 P16 C16 C10 Urban Forest 38°35' F14 B15 F3 Forest Pasture 78° F13 Pasture 45' P18 F12 C18 P4 F20 F5 C11 Watershed boundary F4 F2 draining to Fridley’s Gap P14 F7 B12 38°30' B13 P5 F6 P19 P13 B14 P6 F1 C20 C17 38°30' B21 C4 78° 40' C5 02 4 8 KILOMETERS 78° 45' Base from NAD 1983 USGS Albers Equal Area Conic Land use data from Fry , 2011 02 4 8 MILES FIGURE 1. Source sample locations land use in 2011 of Smith Creek, Virginia Watershed (letter and number indicate sample ID; with letter denoting source type).