Correlating Sub-basin Sediment Fingerprints with Land Use in the Southern Piedmont

This study seeks to further our ability to directly determine sediment provenance by utilizing the sediment fingerprinting technique and Rapid Geomorphic Assessments (RGAs) to determine both sediment contributions from potential sources and the stability of stream channels. Two sub basins of the North Fork Broad River (NFBR) were sampled for suspended sediment. Potential sources fall into three categories 1: surface (pastures and forests) 2: stream banks 3: upland subsurface (dirt roads, construction sites). Three tracers are being used in the study: total Carbon (TC), 15 N, and Fatty Acid Methyl Esters (FAME). The Multivariate Mixing Model was used to determine relative contributions from source components. Results from the fingerprinting study were compared to RGA data in an attempt to establish a relationship between the two techniques. Currently we have sample data for 7 events in 2009 and 2010. Utilizing TC and 15 N, the model output suggests a contribution of about 85% from stream banks and another 10% from pastures. The upland subsurface category is showing only a minimal contribution of about 5%. RGA data collected in 2008 show both tributaries to be unstable with mean stability indexes ranging from 17.2 to 17.6. INTRODUCTION In the United States, 12% of assessed streams are considered threatened or impaired (USEPA 2006). In the Southeast, many Piedmont streams are considered impaired due to high sediment levels. These large concentrations of suspended sediment have an adverse impact on stream biota from primary producers to upper food chain predatory species (Dunne 1978). In an effort to reduce sediment loading, Total Maximum Daily Loads (TMDLS) and Best Management Practices (BMPs) have been developed for sediment and runoff control. In an effort implement TMDLs and BMPs in the Southern Piedmont, research is needed to better identify the source of these suspended sediments. The Southern Piedmont had elevated rates of erosion during the intensive cotton farming era of 1830-1930 and as a result, channels and floodplains were inundated with the estimated 9.7 km 3 of soil that were eroded in the region (Trimble 1974). Figure 1. North Fork Broad River Basin In modern times, erosion rates in the Piedmont have waned to levels approaching if not equal to their preEuropean settlement rates because agriculture has waned and soil conservation measures have been put in place (Trimble 1974). It is apparent however that the effects of this period are still being felt as fluvial processes continue the task of transporting the legacy sediments deposited a century ago. The North Fork Broad River (NFBR) is in Northeast Georgia. In 1998 it was placed on the 303(d) list for impacted biota and habitat with sediment being the pollutant of concern. In 2004 the USEPA conducted a macroinvertebrate study on the watershed. Based on the results of the study, the watershed was removed from the 303(d) list, however they reported that “habitat concerns are present but not to an extent impacting biota.” In 2004 a grant was appropriated to implement BMPs and monitor sediment loads in the NFBR. A subsequent grant in 2007 funded our current research which involves the use of rapid geomorphic assessments and sediment fingerprinting to examine channel stability and determine the source contributions of suspended sediment. We are examining the spatial variability of sediment contribution among several sub basins of the NFBR in relation to the entire basin. Also, we are looking at the relationship between Rapid Geomorphic Assessment (RGA) values and bank contributions to suspended sediment with the hypothesis that higher bank contributions will be present in those basins whose channels seemed more unstable. A previous study was conducted here by Mukundan et al. (2010) with the intent of determining the sediment contributions from different land use types for the main stem of the NFBR. Their results were that around 65% of sediment was of bank origin, 25% from upland subsurface inputs, and 10% from pastures. This study uses the concept of the channel evolution model in streams created by Andrew Simon and his associates (Simon and Hupp, 1986; Simon 1988). Simon’s model was developed in western Tennessee in an attempt to understand process response mechanisms which followed channelization in that region. It was noted that upstream of disturbance (channelization) degradation was occurring with bed levels lowered up to 6 meters. Simon created a 6 stage model which documented the progression from predisturbance stable to post disturbance stable (or pre and post equilibrium) and is described in detail by Andrew Simon (1988). Rapid geomorphic assessments (RGAs) are used to determine the stage of channel evolution and overall stability. The RGAs carried out in this study followed the channel stability ranking scheme (Klimetz and Simon, 2007). There are 9 criteria used in performing an RGA . These are: primary bed material, bed/bank protection, degree of channel incision (percentage) , degree of downstream constriction (percentage), dominant bank erosion type (fluvial vs. mass wasting), percentage of each bank failing, established riparian woody buffer (percentage), occurrence of bank accretion (percentage) , and finally the stage of channel evolution from Simon’s model. A score above 20 indicates a very unstable reach; a score below 10 indicates a stable reach. There have been numerous sediment fingerprinting studies in the past and it has proven itself an effective tool in determining sediment source type and spatial origin (Walling, 2005). The technique involves the characterization of source types based on chemical, physical and/or biological properties establishing individual source “fingerprints”. The tracers used must be measurable in both source soils and sediment and must be conservative in that they don’t undergo any chemical alterations between generation and delivery. Properties used include sediment color (Grimshaw and Lewin, 1980), plant pollen (Brown, 1985), mineral magnetic properties (Walden et al., 1997), rare earth elements (Kimoto et al., 2006), fallout radionuclides (Collins and Walling, 2002; Nagle and Ritchie, 2004; Walling, 2005; Mukundan et al.,2009), and stable isotopes of C and N (Papanicolaou et al.,2003; Fox and Papanicolaou, 2007), and Fatty Acid Methyl Esters (Banowetz et al., 2006)