Features Drainage Management Systems
P in surface runoff

Patience is wearing thin in some communities because of water quality problems linked to the loss of crop nutrients from agricultural land and other sources. The city of Des Moines, IA, is moving ahead with a lawsuit against three of the state’s highly agricultural counties that manage drainage districts traced to high concentrations of nitrates in the Raccoon River, a major source of drinking water for the city.

Des Moines officials claim tile drains should be regulated as point sources under the Clean Water Act. Nonpoint pollution sources now make up the largest source of total phosphorus (P) leading to excessive algae growth and disruption of fisheries in Lake Erie.

Much of the most productive farmland in Ontario, Michigan and Ohio is tile drained. While drainage has economic and environmental benefits, under some conditions excess nutrients from commercial fertilizers and livestock manure can quickly move to tile drains through preferential flow paths and thereby escape from the field.

Merrin Macrae, associate professor at the University of Waterloo in Ontario, opened a conference sponsored by the Michigan chapter of the Soil and Water Conservation Society and Michigan State University Extension in East Lansing on March 4, entitled A Matter of Balance: Systems Approaches to Managing the Great Lakes Landscapes. Her presentation summarized the results of work with farmers in southern Ontario in tracking nutrient cycling and water quality impacts from surface runoff and tile-drained lands to Lake Erie tributaries.

Southern Ontario farms are largely in corn-soybean-wheat rotations similar to many farms in Michigan and Ohio. Macrae’s research group worked with several progressive farmers (ANSWERS group) and had help and funding support from the Ontario Ministry of Agriculture, Food and Rural Affairs (OMAFRA) to establish year-round sampling sites to measure runoff of both overland flow and tile drainage. The soils were silt loam, clay or sandy loams. Soil test P averaged 14 parts per million (ppm), ranging from five to 25 ppm with a tile depth of three feet and spacing ranging from 30 to 45 feet. Various low disturbance tillage practices were used. Commercial P was subsurface banded and poultry litter was used at one location.

In Ontario, precipitation is fairly well distributed throughout the year with no real predictable wet or dry periods, but runoff is highly seasonal. Very little runoff occurs during the growing season unless there is a huge storm; most rainfall is stored in the soil profile and is used by the crop. Most runoff occurs during the non-growing season. Typically, 80 percent of runoff occurred between October and June, with most of that at snowmelt in March. Because P loss was proportional to runoff, most nutrient loss occurred in the non-growing season, with most losses during spring snowmelt. During the growing season, the nutrient loss was linked to runoff following heavy rainfall events, so maintaining crop residue cover and other erosion prevention BMPs was important.

Nutrient loss from tile drains has come under close scrutiny in recent years. In Ontario, Macrae measured much higher P concentrations in the surface runoff than in the tile drain effluent. Compared to tile drainage, surface runoff was a much “hotter” source of P. At one site, surface runoff was only a minor contributor to the total runoff volume, but it was a major contributor to dissolved reactive phosphorus (DRP) loss. Tile drainage contributed most of the total runoff but overland flow carried most of the nutrients. Most of the total annual P loss was particulate P carried by sediment in runoff during snowmelt.

The timing of nutrient application and soil test P was important. When commercial P or poultry litter was applied in the fall, autumn rains led to spikes in tile drainage P. Late fall and all applications on the frozen ground should be avoided. Very early fall or spring applications carry less risk.

No-till cropping is associated with better soil aggregation and structure, but with little tillage nutrients can concentrate near the surface where they are at risk of loss to overland flow. No-till soils tend to have greater macropore development, which can enhance rapid movement of nutrients to the tile drains. Periodic tillage may break the flow paths and mix the nutrients more evenly throughout the root zone and thereby reduce soil test P at the surface. Subsurface placement of fertilizer P will help reduce surface stratification. Cover crops, buffer strips, grass waterways and other forms of vegetation to slow water movement and retain sediment are proven methods for reducing sediment and nutrient loss. There is some concern in northern locations that some winter-killed covers can release soluble P but the risk differs by species.

On a watershed basis, current estimates are that Michigan, Indiana and Ohio farmers in the Lake Erie watershed are losing about 1.5 pounds per acre of P per year. Logic dictates losses would be similar in the Saginaw Bay watershed. The goal for Lake Erie is to reduce losses by about 40 percent, to less than one pound per acre in the western basin. In Ontario, Macrae measured consistent field losses on the cooperating farms of less than half a pound per acre per year, even in wet years. The low losses were achieved with strategically low soil test P and subsurface banding or immediate (before rainfall) incorporation. Crop rotations, cover crops after wheat, residue management, rotational or minimum tillage after wheat, no nutrient applications in late fall or on frozen ground and the use of multiple, bundled BMPs held crop nutrients in the root zone for crop use.

The challenge of keeping nutrients in the root zone year-round highlights the importance of using multiple barriers with stacked and/or bundled BMPs to stabilize cropland and prevent runoff. Monitoring and sampling throughout the non-growing season, and particularly during snowmelt and early spring, is critical for truly understanding the timing and quantity of nutrient loss. Most runoff and nutrient loss occurs at snowmelt. Overland flow tends to be the “hottest” source of nutrient loss so select BMPs that slow the water down. Avoid late fall nutrient applications and all applications on frozen ground. Spring applications carry less risk. Banding carries less risk than surface broadcasting, but when nutrients are broadcast they should be incorporated before rainfall because the most loss occurs with the first rainfall events.

This article was originally published by Michigan State University Extension. Reprinted here with permission.