TRIBUTARY CREEK WATER QUALITY INFLUENCED BY LAND USE CHARACTERISTICS AT MARINE COPRS BASE CAMP LEJEUNE, NC

Abstract

Marine Corps Base Camp Lejeune is located in the flat coastal plains of North Carolina. The base is home to ecologically important and sensitive habitat. It is also the largest amphibious Marine Corps base with high impact training areas and a growing population. Creek water quality was monitored in ten tributary creeks throughout the base, representing paired plots of forested and developed drainage areas. It was my aim to assess land use contributions to creek nutrient concentrations. By building watersheds using hydrology tools in ArcMap and a 1-meter horizontal resolution DEM, I was able to delineate watershed area and calculate land use proportions within each of the ten watersheds. In total, I defined twenty-two land use characteristics which were used as predictors of seven nutrient concentration response variables (TSS, NOx, NH4, PO4, ON, TN, Chla). Linear model results isolated significant predictors of water quality. For each response variable I fit the best GLM, indicated by StepAIC, using each predictor with a p-value less than 0.05. Nitrate/Nitrite (NOx) was best predicted by the percentage of paved roads within each watershed for annual averaged concentrations. When separated into storm and base flow events, NOx, again, is predicted by the percentage of paved roads. Ammonium (NH4) concentrations are correlated the percentage low intensity development which was defined by the National Land Cover Dataset (NLCD 2001). These positive correlations make ecological sense and are consistent with similar land use studies. However, this study site is unique because of its low elevation, low flow accumulation, and lack of agricultural lands which are of high proportion in surrounding basins. It is important for base environmental managers to apply these results and models to answer the following type of question: How will creek water quality change if we convert one land cover type to another? Using the best model predicting NH4, I developed an interactive tool for managers to map out a proposed construction site and obtain changes in average NH4 concentrations in tributary creeks.