NUTRIENTS IN STREAMS DURING BASEFLOW IN SELECTED ENVIRONMENTAL SEVFINGS OF THE POTOMAC RiVER BASIN¹

Abstract

A regional assessment of water quality in small streams was conducted within four areas of distinct physiography and lithology in the upper Potomac River Basin. The Potomac River is a major tributary to the Chesapeake Bay, and this study provides new insight on the relationships between nutrient concentrations in small streams and watershed characteristics within this river basin. Nutrient concentrations were compared to land‐use data including categories for agriculture (cropland and pasture), urban areas, and forests. Among agricultural areas, streams draining areas of intense row cropping typically contained higher nitrate concentrations than did those draining pastures. Streams draining forested areas typically had the lowest nutrient concentrations. Streams in areas underlain by carbonate bedrock were more likely to contain elevated concentrations of inorganic nitrogen and potassium than did streams in areas underlain by fractured siliciclastic or crystalline rocks, and we suggest that this is a physical phenomenon related to high hydraulic conductivities in carbonate ground‐water systems. The median nitrate concentrations were highest in the Great Valley portion of the Valley and Ridge physiographic province, particularly in watersheds that have both carbonate bedrock and intensive row cropping. Values of nitrate in these streams ranged up to 8.99 mg/L as nitrogen. The soluble phosphorus concentrations during baseflow were generally low in all sub‐units, even in some settings with potential for high phosphorus inputs such as urban areas with municipal point sources or agricultural areas. The mobility of phosphorus in these environments may be hindered by adsorption and geochemical reactions.

EFFECTS OF LAND COVER AND GEOLOGY ON STREAM CHEMISTRY IN WATERSHEDS OF CHESAPEAKE BAY1

“…Factors can be related to actual processes that influence the behavior of stream chemistry variables. A number of studies have employed factor analysis, or the closely related principal component analysis (PCA), to investigate processes controlling the chemical composition of groundwater and streams (Dawdy and Feth, 1967;Reid et al, 1981;Christophersen and Hooper, 1992;Puckett and Bricker, 1992;Evans et al, 1996;Miller et al, 1997). Factor analysis can also simplify the interpretation of large multivariate data sets by identifying groups of variables that behave similarly.…”

“…This facilitates the interpretation of the factors (Johnson and Wichern, 1982). For the purpose of interpretation, factor loadings were classified into three categories: high loadings (> 0.75), moderate loadings (0.45-0.75), and low loadings (< 0.45) (Puckett and Bricker, 1992;Evans et al, 1996;Miller et al, 1997). For the sake of clarity, low loadings are not reported in Table 5.…”

“…For the same reason, the effects of land cover on P fluxes in the streams may not appear in the water samples. Some studies have reported similar levels of P in watersheds with different land uses, or lack of significant land use effects on P concentrations in streams (Jordan et al, 1997a, b;Miller et al, 1997). However, many studies in the literature have shown that when its major components are accounted for, P flux is strongly influenced by land use, especially intensive agriculture (Hill, 1981;Correll et al, 1999a, b).…”

“…The water quality of streams in the Chesapeake drainage basin has come under increasing study, especially during the past decade (Puckett and Bricker, 1992;Correll et al, 1994;Jordan et al, 1997a, b, c;Miller et al, 1997;Ator et al, 1998;Lindsey et al, 1998). The Chesapeake Bay basin has been included in a number of nation-wide water quality programs.…”

Identifying the hydrochemical characteristics of rivers and groundwater by multivariate statistical analysis in the Sanjiang Plain, China

“…Multivariate statistical analysis requires that the observations conform to normal distribution and the data transforms, such as log-transform (Cloutier et al 2008), rankordered (Miller et al 1997) and normalization (Ravichandrana et al 1996;Ruiz et al 1990), are carried out before doing multivariate statistical analysis. In this study, the normality of variables was checked by one-sample K-S test.…”