needed to counteract runoff decreases resulting from a warming of 4 Fahrenheit degrees; a 15-percent increase would be needed for a warming of 7 Fahrenheit degrees. A warming of from 4 to 7 Fahrenheit degrees, without corresponding precipitation increases, would cause a 9to 25-percent decrease in total annual runoff. When reservoir data are factored into the model, a significantly greater risk of a drought emergency exists in which reservoir levels are too low to meet demand-for scenarios in which precipitation does not increase sufficiently to offset the CO2-induced warming.

A topographically based hydrologic model was developed and linked with the wet/dry climate model to analyze the sensitivity of daily streamflow. Results of this model, without reservoir data, indicate that the overall effect of warming is to decrease daily streamflow. Most of this decrease would occur in the warmer seasons. In the northern part of the basin where snow accumulation currently is significant, the warming actually would result in an increase in the February average and maximum daily flow, regardless of precipitation changes, because of the increased precipitation as rain. In general, the model showed that watershed runoff was more sensitive to changes in precipitation intensity than to changes in precipitation duration or temperature.

Estuary Modeling

A sea-level rise, which is likely to accompany global warming, would alter estuarine salinity. Scientific consensus indicates a global warming of about 7 Fahrenheit degrees for conditions of double CO, would result in an estimated sea-level rise of from 0.5 to 4.5 feet. In the Delaware River Estuary, a rise of 2.4 feet would cause the saltwater front to move about 8 miles farther upstream.

Both the rise in sea level and the saltwater movement could have serious implications for the continued availability of surface and ground water in the area of the upstream saltwater movement.

Ground-Water Modeling

Three ground-water models will be used to assess the effects of sea-level rise

and the resulting changes in salinity on ground-water availability. The final calibration of these models is near completion. An available model of the aquifer system near New Castle, Del., was used for sensitivity analyses. The model results indicate that this semi-confined aquifer system is sensitive to the flooding that would result from a rise in estuary levels. Because of the presence of a confining unit under the estuary that restricts the movement of saline water into the aquifer, a sea-level rise of 5 feet alone would not result in a significant change in recharge of saline estuary water into this aquifer system. A combination of flooding and a sea-level rise of 5 feet, however, would triple the amount of saltwater recharged to the aquifer system because the flooding would extend beyond the confining unit. A second ground-water model will focus on potential changes resulting from sea-level rise on saltwater intrusion in an unconfined aquifer system near Cape May, N.J. A third model will assess the effects of the extended movement of saltwater in the estuary on saltwater intrusion and ground-water availability in a multiplelayered confined-aquifer system near Camden, N.J.

Results of simple analyses of the potential global warming on water resources in the Delaware River basin suggest serious implications for future availability of water-related resources. In this humid, temperate climate, where precipitation is distributed evenly throughout the year, decreases in snow accumulation in the northern part of the basin and increases in evapotranspiration throughout the basin could change the temporal distribution of runoff and reduce streamflow by as much as from 9 to 25 percent unless precipitation increased. Also, ground-water recharge of saline estuary water in one aquifer near New Castle, Del., could double with a sea-level rise of 5 feet. USGS scientists will continue to refine and complete the models and scenarios in order to quantify more accurately the effects and associated risks of the various potential climatic and sea-level changes on streamflow and ground water in the Delaware River basin.

Investigation of Water Quality, Bottom Sediment, and Biota Associated With Irrigation Drainage in the Western United States

By Herman R. Feltz, Richard A. Engberg, and Marc A. Sylvester

In response to concerns expressed by the U.S. Congress over contamination at the Kesterson National Wildlife Refuge in California, the Department of the Interior (DOI) started a program in 1985 to identify the nature and extent of irrigation-induced water-quality problems that might exist in other western areas where the DOI has responsibility. The DOI formed a Task Group on Irrigation Drainage, an interbureau group chaired by the USGS. Initially, the Task Group identified 19 locations in 13 States that warranted reconnaissance-level investigations (fig. 1). These locations relate to three specific areas of DOI responsibility:

irrigation or drainage facilities constructed or managed by the DOI, national wildlife refuges managed by the DOI, and other migratory-bird or endangered-species management areas that receive water from DOI projects. Nine of the 19 locations were selected for reconnaissance-level investigations during fiscal years 1986-87. Study teams composed of three scientists, one each from the USGS (team leader), the U.S. Fish and Wildlife Service, and the Bureau of Reclamation, were formed to conduct the investigations at each location. Surface and ground water, bottom sediment, and biota were investigated at each location. Reports for completed studies are shown in table 1. Reconnaissance-level investigations were started in fiscal year 1988 at the remaining 10 sites identified by the Task Group. Reports for these studies will be published in fiscal year 1990. A 20th study for the Pine River area in southwestern Colorado was added in fiscal year 1989 (fig. 1).

In the first nine study areas, analyses of water, bottom sediment, and biota. sampled were evaluated against Federal and State water-quality regulations and criteria, baseline data for adjacent areas, and other guidelines that might be helpful in making assessments of adverse

effects on fish, wildlife, and humans. Water samples were analyzed for major ions, nutrients, selected trace elements, radiochemical constituents, and for pesticides at sites where they were used. Samples of bottom sediment and biota were analyzed for selected trace elements and, at some sites, pesticides.

Because some constituents exceeded water-quality regulations or criteria in samples from four of the first nine study areas, the DOI Task Group made the decision to proceed with detailed studies of these four areas. These areas and the constituents of concern are given in table 2. Slightly elevated levels of some constituents were found in water, bottom sediment, or biota in some of the five remaining areas, but the levels were not considered of sufficient concern to recommend detailed studies; however, some level of long-term monitoring may be initiated at those five sites.

The four detailed studies are oriented toward meeting two goals: (1) to confirm that irrigation-induced water

quality problems exist and (2) to provide the scientific understanding needed to develop reasonable alternatives that will mitigate or resolve identified problems. Within this context, the working objective for a detailed study is to determine the extent, magnitude, and effects of contaminants associated with agricultural drainage and the sources and exposure pathways that cause contamination where contaminant effects are documented.

To ensure that all areas having problems related to irrigation drainage in Western States were identified, a comprehensive survey of all DOI irrigation projects and wildlife management areas was conducted in fiscal years 1988 and 1989. Fifteen additional areas that may require additional investigation were identified. These areas are undergoing intensive evaluation of existinginformation to determine whether a reconnaissance-level investigation is

necessary.

Results of the completed reconnaissance-level investigations and preliminary

data from the ongoing reconnaissance investigations and detailed studies provided the following generalizations for areas where problems have been detected:

• Elevated concentrations of trace elements have been detected in several of the study areas, and pesticides have been detected in some of the study areas; • Alkaline, oxidized soils that contain elevated concentrations of trace elements in semiarid environments indicate potential problem areas;

• Selenium, boron, arsenic, and mercury are the constituents found most often at elevated concentrations in water, bottom sediment, and biota in the study areas;

pollution is largely a function of how the agricultural land is managed.

Farmers can use land-management practices that control erosion, increase soil moisture, and reduce the transport of farm chemicals and fertilizer in runoff from the fields. These methods, which are generally referred to as bestmanagement practices, include development of grassed waterways and field borders, strip cropping, contour farming, and crop rotation. In contrast, when traditional or standard land-management practices are used, waterways are poorly maintained, crop production is continuous and without rotation, and the rows are plowed straight without regard to slope or topography.

To better define how agricultural land-management practices affect water quality, the USGS in cooperation with the Guilford Soil and Water Conservation District and the U.S. Soil Conservation Service began a 6-year study in 1984 of four small basins in the Piedmont of North Carolina. The Piedmont, a physiographic province extending from Virginia through Alabama, is characterized by clayey soils, rolling topography, and abundant rainfall. This area was chosen because of the highly erosive nature of the soils and the ongoing local effort to convert existing farmland to bestmanagement practices. Results of this study should be transferable to similar agricultural lands throughout the Piedmont physiographic region of the Southeastern United States.

The study is designed to monitor chemicals applied to the land through farming practices as well as nutrients resulting naturally from atmospheric deposition. It also monitors water quantity and quality of overland runoff, concentrations of chemical constituents percolating through the clay soils in the unsaturated zone, and constituents reaching the ground water. Farmers cooperating in the study are helping keep detailed records of the chemicals applied to their fields and of their farming activities such as plowing. Data collection is scheduled to end September 1990.

Four areas including two row-crop fields, a mixed land-use basin, and a forested basin were selected for study. The row-crop fields are adjacent-one having best-management practices (7.4 acres)

and the other having standardmanagement practices (4.8 acres). The amount of sediment, nutrients, and selected organics in runoff and the volume of runoff were monitored for the two fields as were the nitrate plus nitrite and pesticide content of soil water.

In the mixed land-use basin (665 acres), changes over time in water-quality constituents in runoff are being monitored at a streamflow gage, as standardmanagement practices are converted to best-management practices over the duration of the project. In the forested basin (44 acres), background hydrologic and chemical-quality conditions are monitored. These areas are within a 4-mile radius, and the effects of atmospheric deposition, which is monitored at one of the agricultural field sites, are assumed to be equal among all four areas.

Analysis of surface-water-runoff data through May 1989 indicates that for the two row-crop fields, in general, concentrations of sediment, nutrients, and selected organics in runoff from the field having best-management practices are dramatically lower than the concentrations found in the field having standard practices. A general relation appears to exist for nutrients and sediment concentrations for the four sites and for precipitation. With the exception of nitrogen, the lowest constituent concentrations are measured in precipitation, followed by increasing concentrations in runoff from the forested area, the field site having best land-management practices, and the mixed land-use site, with the standardpractices field site having the highest concentrations. Interestingly, nitrate plus nitrite concentrations were found to be higher in precipitation than those measured in runoff at the forested site, probably because the nutrient was bound up by the forest litter and also used by the plants.

The difference in sediment concentrations between the two agricultural fields is striking. The standard-management practices field had a mean concentration of 11,200 milligrams per liter (mg/L), compared with the best-management practices field mean concentration of 3,230 mg/L. Sediment concentrations for the mixed land-use basin were generally lower than those observed for the agricultural field sites because of the

presence of 14 small farm ponds in the basin that act as sediment traps.

The erosion process, which creates suspended sediment and the resulting sediment yield, tends to sort the soil particles by carrying away the fine silts and clays associated with most of the soil fertility. Sediment yields-the amount of suspended sediment moving past the runoff gage-are consistently higher in the field having standard land-management practices than yields in the field having best land-management practices. During the 1987 water year, the sediment yield from the basin having standard practices was almost 36 tons per acre, compared with 5.4 tons per acre that came off the field having best-management practices.

A seasonal comparison of runoff differences between the two agricultural fields shows that the greatest amount of runoff occurred in the standardmanagement practices field during the growing season (May-September) and in the best-management practices field in the barren season (October-April). This is probably because during the growing season there is more bare, hardened ground, which promotes runoff, in the