Temporal and Spatial Patterns in Water Quality Constituents in Watersheds along an Urban-Rural Gradient in the Gwynns Falls Watershed (presentation)
BELT, KENNETH T., Groffman, P., Hopkins, J., and A. Lorefice.
USDA Forest Service, Site Manager, Baltimore Ecosystem Study, Rm. 134 TRC Bldg., University of Maryland at Baltimore County, 5200 Westland Blvd., Baltimore, Md. 21227, Phone: (410) 455-8011. Email: firstname.lastname@example.org
Institute of Ecosystem Studies, Box AB (Route 44A), Millbrook, NY, 12545-0129
Institute of Ecosystem Studies, Rm. 134 TRC Bldg., Baltimore Ecosystem Study, University of Maryland at Baltimore County, 5200 Westland Blvd., Baltimore, Md. 21227.
We used the watershed approach to characterize and monitor streams with a variety of land usess along an urban-rural gradient in the 17,150 ha. Gwynns Falls watershed located in Baltimore County and City. Nitrate, sulfate, chloride and phosphorus data collected from October 1998 to May 1999 were presented for fourteen stream stations.
The BES long term water quality monitoring station network includes nine continuously recording stream gages (Figure 1) maintained by the USGS, complimented by a network of ten rain gages, most of which has been in operation for the past year. There are three mainstem stations on the Gwynns Falls (at Carroll Park, Villa Nova, and Gwynnbrook), two subwatershed sites (at Dead Run and Baismans Run), and four small watersheds that represent high density urban (Rognel Heights), suburban (Glyndon), agricultural (McDonogh), and forested (Pond Branch) land uses. Additionally, six ungaged sites in the Gwynns Falls are sampled weekly through a one year collaborative arrangement with a doctoral student from Johns Hopkins University/US Army Corps of Engineers.
Patterns among the different watersheds have emerged from the early results (figure 1). Nitrate concentrations tended to decrease in the downstream reaches (the older, dense urban core) of the Gwynns Falls. This was counter to what was expected along the urban-rural gradient, where leaking sewage infrastructure could have increased nitrate levels. Unexpectedly low nitrate levels were also seen for Dead Run (an older highly developed urban subwatershed) and Horsehead Branch (a developing suburban subwatershed). These nitrate levels were lowest of all the sites except for Pond Branch, where low levels were expected since it is completely forested. Conversely, unexpectedly high nitrate concentrations were seen at Baismans Run (largely forested with some low density residential land use) and the Glyndon small watershed (medium density residential).
All four constituents showed some kind of temporal trend, although the magnitudes and directions of these trends varied with different watersheds. These included 1) trends over the period of sampling, 2) correlation with flow rates, 3) differences between the dry periods (Fall 1998 and Spring 1999) and the higher flow period (Winter 1998-99) and 4) short term increases. Chlorides, for example, exhibited large and sustained spikes during cold weather only at the Glyndon, Gwynnbrook, Carroll Park, and Dead Run stations, probably due to road salt operations (figure 3).