We've Moved!
Visit SDSU’s new digital collections website at https://digitalcollections.sdsu.edu
Description
Over a one-year period nitrate sampling was carried out in two streams within the Santa Margarita Watershed located in northern San Diego County, California. The two streams include Stone Creek, which is located in a largely undeveloped watershed, and Rainbow Creek, which is located in a watershed with diverse land uses. Nitrate results for Stone Creek showed that concentrations were at least one order of magnitude greater in the upper section of the creek than the lower section of the creek and generally peaked during late summer to early fall. The nitrate concentrations in Stone Creek are controlled by imported irrigation water used for the avocado and orange orchards located in the lower section of the creek. Rainbow Creek had high nitrate concentrations throughout the length of the creek with the highest nitrate concentrations being located in the upper section of the creek below a concentrated area of anthropogenic nitrogen sources. Nitrate concentrations generally peak during the spring in Rainbow Creek, and were generally lower in the summer and winter. Irrigation water from nurseries and orchards in the form of both point source discharges and groundwater discharges had the greatest influence in Rainbow Creek throughout the year, increasing the downstream nitrate concentration during the growing season. Though nitrate concentrations seemed to follow a pattern of increasing and decreasing concentration there was no correlation found between nitrate, discharge and the water quality parameters measured during this study. In addition to the sampling that was carried out for this study, two simple and one more complex watershed loading model were applied to the two watersheds. The three models used in this study include the Export Coefficient Method and the Simple Method, which are simple watershed loading models that predict total annual nutrient loads, and the Generalized Watershed Loading Function (GWLF) model, which is a more complex watershed loading model able to predict total monthly loads. The prediction of total nitrogen loading in the Stone Creek watershed was most accurate using the Simple Method, but all three models over-predicted the total nitrogen loading. Prediction of the percentage of the total nitrogen load contributed by each landuse was similar among the three models. Vacant land made up the bulk of the nitrogen loading followed by rural residential land use and orchards. Unlike the predictions in the Stone Creek watershed, the GWLF model was the most accurate in the Rainbow Creek watershed, primarily due to the model’s ability to predict loading from groundwater and especially point sources, which made up the bulk of the nitrogen load in the watershed. From this study it was determined that for small largely undeveloped watersheds more complex models did not increase the accuracy of nitrogen load predictions. However, more complex models are necessary even as a screening model to determine sources areas and nitrogen loads from larger and more complex watersheds. The Appendices are available for viewing on CD-ROM at the Media Center of Love Library.
