EFFECTIVENESS OF AGRICULTURAL BMP'S IN REDUCING STORMFLOW SUSPENDED SEDIMENT IN TELLICO CREEK, MACON COUNTY, NORTH CAROLINA

Sediment tops the list of water quality impacts in North Carolina, and nationwide. Agriculture is the dominant cause of such nonpoint source impairment of streams, affecting more than 50% of the surface waters in the southeast US. North Carolina's Division of Water Quality identified sedimentation as the biggest threat to mountain streams. Tellico Creek (Macon Co) is a tributary of the Little Tennessee River, in the southern Appalachian Mountains in western North Carolina. The region is largely forested, with limited agriculture.Total suspended solids (TSS) have been monitored at up to seven sites in the 1591-ha Tellico Creek study area since 1990. The initial study documented the successful use of construction BMP's to improve and stabilize roads, and the subsequent reduction of instream stormflow sediment concentrations. Additionally, a small family farm was identified as a significant (90%) source of stormflow sediment. The current agricultural BMP assessment focuses on that farm, and on three sampling sites: Upper Tellico and Sugar Cove Creek (the main stream and a major tributary, both just above the farm), and Lower Tellico (just below the farm, and also below the confluence of Tellico and Sugar Cove Creeks). A total of 856 baseflow samples (280 to 292 per site), and 524 stormflow samples (152 to 189 per site) was collected. From the baseline period (Nov 1990 – Mar 1994) through the period of post-BMP monitoring (May 1995-Mar 1999), mean baseflow sediment concentrations at Upper Tellico, Sugar Cove, and Lower Tellico decreased from 16.5 to 7, 6.5 to 5.4, and 8.5 to 6 mg/L, respectively. From the baseline to post-BMP period, mean stormflow sediment concentrations in the control watershed (Sugar Cove) increased over five-fold, from 108 to 531 mg/L. At Upper Tellico, mean stormflow sediment increased over eight-fold, from 282 to 2400 mg/L; and below the BMP work, at Lower Tellico, it increased only 30%, from 1340 to 1740 mg/L.A sediment transport assessment indicated that 90% of stormflow sediment originated on the farm before BMP's; this dropped to 62% originating on the farm during the BMP application period, and dropped further to 32% for the post-BMP period, based on mean sediment concentrations. Of five storms analyzed individually, a pre-BMP storm in 1991 indicated that 99% of stormflow sediment originated on the farm; and four storms between 1995 and 1999 indicated that between 25 and 62% of stormflow sediment originated on the farm. Two major floods were also compared: the flood of April 1994, in which the farm reach aggraded significantly, and a 1998 flood in which the farm served as the source of an estimated 61% of the downstream stormflow sediment. While much of the sediment leaving the farm is now attributed to upstream sources, it is also suspected that some sediment still originating on the farm (post-BMP's) may in fact be due to unsampled (nonpoint source) road runoff, rather than farm runoff, that is delivered to the stream just above the Lower Tellico site.Analysis of the frequency distributions of stormflow sediment concentrations reinforced the conclusion that while the agriculture BMP's were effective, the Upper Tellico site and the paired control watershed both deteriorated during the study. From the baseline to post-BMP period, the median stormflow sediment concentration, C50 (50% of samples being greater than the given concentration), in the control watershed increased from 60 to 124 mg/L; and at Upper Tellico it increased from 103 to 125 mg/L. Water quality below the BMP work, at Lower Tellico, improved as the median concentration decreased from 159 to 94 mg/L. While the C80 at Sugar Cove increased from 24 to 38 mg/L, at Upper Tellico it was similar (40 and 36 mg/L), and at Lower Tellico the C80 decreased from 44 to 29 mg/L. Similarly, on the high end of concentrations, the C20 at Sugar Cove increased from 182 to 365 mg/L; at Upper Tellico it increased from 325 to 2900 mg/L; and at Lower Tellico the C20 decreased from 1960 to 1050 mg/L.Assessment of the effectiveness of the BMP's was compounded by development activity in (and apparent degradation of) both our paired control and Upper Tellico basins. Natural events such as catastrophic floods disrupted stream profiles, cross-sections, and sediment sources. Stormflow sediment concentrations were highly variable. Nevertheless, although sediment concentrations below the farm increased slightly over the years, this was primarily due to higher concentrations above the farm. Temporal trends indicate that the percentage of stormflow sediment that is attributable to agricultural practices on the farm has dropped significantly.