Agricultural producers in North Dakota are aware of concerns about degrading water quality, and many of the producers are interested in implementing conservation practices to reduce the export of nutrients from their farms. Producers often implement conservation practices without knowledge of the water quality of the runoff from their farm or if conservation practices they may implement have any effect on water quality. In response to this lack of information, the U.S. Geological Survey, in cooperation with North Dakota State University Extension Service and in coordination with an advisory group consisting of State agencies, agricultural producers, and commodity groups, implemented a monitoring study as part of a Discovery Farms program in North Dakota in 2007. Three data-collection sites were established at each of three farms near Underwood, Embden, and Dazey, North Dakota. The purpose of this report is to describe runoff and water-quality characteristics using data collected at the three Discovery Farms during 2008–16. Runoff and water-quality data were used to help describe the implications of agricultural conservation practices on runoff and water-quality patterns.

Runoff characteristics of monitoring sites at the three farms were determined by measuring flow volume and precipitation. Runoff at the Underwood farm monitoring sites generally was controlled by precipitation in the area, antecedent soil moisture conditions, and, after 2012, possibly by the diversion ditch constructed by the producer. Most of the annual runoff was in March and April each year during spring snowmelt. Runoff characteristics at the Embden farm are complex because of the mix of surface runoff and flow through two separate drainage tile systems. Annual flow volumes for the drainage tiles sites (sites E2 and E3) were several orders of magnitude greater than measured at the surface water site E1. Site E1 generally only had runoff briefly in March and April during spring snowmelt and during only a few large rain events throughout 2009–16. Flow was somewhat continuous at sites E2 and E3 throughout the year during years of increased precipitation, such as in 2010 and 2011. At Dazey farm, annual flow volumes at the most downstream site D3 for 2010–15 ranged from 88 acre-feet (2012) to 12,060 acre-feet (2010). The largest monthly runoff volumes at D1 (most upstream site; combination of data from site D1a [original site] and site D1b [relocated site]) and D3 were in March and April during spring snowmelt runoff and rain events.

At Underwood farm, total ammonia and total phosphorus had the highest concentrations at the most upstream site (U1) and decreased sequentially at sites U2 and U3 downstream. Total ammonia and total phosphorus concentrations at the sites for Underwood farm also generally were higher than measured at sites for the Dazey and Embden farms. At Embden farm, nitrate plus nitrite concentrations were lowest at site E1 (surface-water site) and highest at sites E2 and E3 (drainage tile sites). Nitrate plus nitrite concentrations at sites E2 and E3 also were the highest among all the sites at all three farms. Median total nitrate plus nitrite concentrations for sites E1, E2, and E3 were 0.22, 13, and 10 milligrams per liter as nitrogen, respectively. Nutrient concentrations generally were greater at site D1 (most upstream site) compared to site D3 (most downstream site) at Dazey farm. Higher concentrations at site D1, which is farther upstream and closer to potential sources of nutrients, compared to lower concentrations at site D3, which is farther downstream and receives more runoff, indicates that dilution may be the reason concentrations decrease downstream.

Annual loads for chloride at all three Underwood sites were the greatest in 2011 and the least in 2012, which coincided with years of the greatest and least annual flow volume, respectively. Total ammonia had a similar pattern at the three sites. Nitrate plus nitrite loads displayed a different pattern than chloride and total ammonia, indicating possible different sources. Chloride, total ammonia, total phosphorus, and suspended sediment were transported past site U1 mostly in March and the least from July through October. Monthly nitrate plus nitrite loads had a different pattern than the other constituents, indicating other possible sources such as fertilizer application in the surrounding cropland.

Annual loads for Embden farm were considerably greater at sites E2 and E3 compared to site E1. Annual yields for all constituents also were substantially greater at sites E2 and E3 compared to site E1, mainly because of a combination of higher flow volumes and small contributing drainage areas at sites E2 and E3 compared to site E1.

The greatest annual loads at Dazey farm site D3 for chloride, nitrate plus nitrite, and suspended sediment were in 2010 and 2011, and zero loads were estimated for 2012 because no flow was measured at the site. Mean monthly loads generally were greatest for most constituents in March and April at sites D1 and D3 except for suspended sediment that had the greatest monthly loads in May.

To mitigate runoff and water-quality effects of their operations, the producers implemented various agricultural conservation practices before and during the Discovery Farms monitoring. Even though it was difficult to quantify the effects of the agricultural conservation practices implemented at the farms, the data collected from the Discovery Farms program provided a better understanding of some of the variables that affect runoff and water quality.