By their nature, ambient water-quality monitoring programs cover large geographic areas and operate over long time periods. This broad spatial and temporal coverage makes them well suited to broad-scale assessment of transport of constituents in a large river basin, but can also restrict interpretations of these data. Resource limitations and logistics in operating a large, long-term monitoring network often force difficult design trade-offs, favoring a sampling design that is prescribed more by schedule than by targeting a wide variety of environmental conditions. Use of the data for interpreting constituent transport as a function of environmental conditions is, therefore, hampered. Heterogeneous characteristics of the large watersheds contributing to many of the ambient monitoring sites confound spatial comparisons of input to export; meaningful comparisons require a network of sites draining watersheds that are homogeneous with respect to both natural and human-influenced environmental setting.
Data-collection requirements depend on the water-quality evaluation to be made: the important water-quality indicator for evaluating risk to the ecological health of a receiving water body caused by nutrient overenrichment is nutrient loading rate from its tributaries, rather than concentrations in these tributaries. Estimation of loading rate requires continuous streamflow record at the tributary sampling site and requires a fully stratified sampling program that covers all possible combinations of season and runoff condition. Evaluation of ecological risk may also require estimates of loading rates during the period of the year when growth of aquatic plants responds most rapidly to nutrient influx: the period of long hydraulic-residence time and warm, clear water in the receiving water body. Although the data sets used in this report are sufficiently large for estimating annual loading rates, they are too sparse when stratified by season to allow accurate estimation of loading rate during a specified, critical season. Accurate estimation of transport during a critical season requires targeting sampling efforts to cover the full range of runoff and streamflow conditions during that season. Data from a fully stratified sampling program will also produce more accurate estimates of temporal trends and can be used to distinguish between loading patterns of nonpoint sources and loading patterns of point sources.
Estimates of 1992 annual flow-weighted mean concentration of total nitrogen ranged from 0.53 to 2.8 mg/L as nitrogen, representing a fivefold difference in instream transport among watersheds in the LTEN River Basin. The smallest estimate was for a minimally developed watershed, and the largest estimate was for a watershed with the largest areal percentages of urban and agricultural land use and largest amounts of wastewater discharge, suggesting that human activity increased exports of total nitrogen by as much as fivefold. The range in estimates of annual flow-weighted mean concentration of total phosphorus, from 0.02 to 0.73 mg/L as phosphorus, represents nearly a fortyfold difference in instream transport among the watersheds. The largest three estimates, 0.73, 0.53, and 0.52 mg/L, probably represent a natural source in those watersheds: the phosphatic limestones of the brown-phosphate districts. The outcrop pattern of these phosphatic limestones may be an important factor to consider as regional boundaries are established for attainable region-specific water-quality criteria for total phosphorus (U.S. Environmental Protection Agency, 1998).
Nutrient overenrichment caused impairment in 37 of the 109 impaired stream segments in the LTEN River Basin in 1996. Impairment is caused by influx of nutrients and growth of algae and aquatic macrophytes during critical periods of the year; therefore, estimates of seasonal flow-weighted mean concentration may be more useful than annual estimates in evaluating ecological risk to water bodies, and in establishing water-quality criteria. Seasonal estimates of flow-weighted mean concentration generally were less than half of the annual estimates, and ranged from 0.28 to 1.2 mg/L total nitrogen, as nitrogen, and from less than 0.01 to 0.18 mg/L total phosphorus, as phosphorus.
Nitrogen from wastewater discharge represents a small part (less than 27 percent) of the annual nitrogen export in 11 tributary basins, and variability in wastewater discharge among basins correlates poorly with annual export. Wastewater discharge may account for a larger part of nitrogen yield during low streamflow, however, and does correlate well (r = 0.71, p = 0.01) with total nitrogen concentration during low streamflow. Phosphorus from wastewater discharge represents as much as 1.2 times the annual phosphorus export, and correlates well with annual export except in watersheds with outcrops of the phosphatic limestones.
The estimates of input from other sources (atmospheric deposition, fertilizer application, and livestock waste) cannot be compared as a fraction of export in the same way as for wastewater discharges because these estimates are the land-phase inputs from these sources, rather than inputs directly to the stream channel. The fraction of export contributed by a source may be inferred indirectly based on correlations between inputs and export. The significant correlation (r = 0.65, p = 0.03) between the estimate of agricultural land-phase input within a watershed and exported nitrogen might mean that these sources contribute to annual instream loads in significant amounts.
Concentrations of total ammonia and total nitrogen decreased during the period 1985-94 at about half of the sites where temporal trends could be tested. The spatial distribution of decreasing trends corresponds with the spatial variation among basins in wastewater input, and the time period of observed trends corresponds to the period of improvements in municipal treatment; thus, decreases in wastewater effluent concentrations of nitrogen might be responsible for the decreasing trend in instream concentrations at these sites.
| Next | Title Page | Table of Contents | List of Illustrations | List of Tables | Conversion Factors and Vertical Datum | Glossary |