Grass in Farm Waterways

DENSE cover of grass or other close-growing plants is nearly always the best and cheapest protection for farm waterways— the draws and other channels through which excess surface water flows off farm lands into streams. Unless these channels are well protected against erosion, the action of flowing water can turn them into gullies.

Any farmer who has vegetation-lined natural waterways on his land will find it profitable to maintain the vegetation within them and keep them from being damaged by farm implements, livestock, or rodents.  Whenever necessary, the waterways should be fertilized and reseeded.  It is much cheaper to improve and maintain natural waterways than to build and maintain artificial ones.

Artificial waterways should be established on farm land only where adequate natural waterways are not available. If a farmer lays out terraces that cannot drain into a natural waterway on his own land, he may be able to extend them to one on the land of a neighbor.

To be safe from erosion, a waterway must have vegetative or mechanical protection, or both. Also, it must be large enough not to be flooded by the greatest flow of water that usually comes off the land draining into it. The most commonly used mechanical device for protecting a waterway is the check dam—a small dam of concrete, stone, wood, or other material. A series of check dams in time transforms the slope of a waterway into a succession of steps with low risers and long, almost flat treads. At times of storm flow, water falls abruptly over each dam but flows much less rapidly than it would have if dams had not been used. Thus the danger of soil erosion is reduced.

Although for many years farm terraces have commonly drained into vegetation-lined natural waterways, until rather recently a farmer who established an artificial terrace outlet or other waterway usually depended on check dams alone to control erosion within it. Research in the use of vegetation for lining artificial waterways began with the establishment in 1929, at Guthrie, OKla., of the first of 10 soil-erosion experiment stations. In a terrace outlet with a 2½-percent slope, concrete check dams 6 inches high were used and Bermuda grass was planted. At the station at Bethany, Mo., an outlet with a 6-percent slope was protected by burying 2- by 12-inch planks on edge at vertical intervals of 2½ feet and sowing seed of lespedeza, White Dutch clover, Kentucky bluegrass, and redtop.

Both methods proved effective. Later results at Guthrie indicated that in the gently sloping outlet a good growth of Bermuda grass, once it had become established, would have controlled erosion satisfactorily even without the concrete check dams, which served mainly to prevent erosion during the period required for the establishment of an adequate growth of vegetation. Also, it has since been found generally possible on moderate slopes in most localities to dispense with plank checks such as were used in the waterway at Bethany. One way of doing this is to prepare the vegetation-lined outlet before constructing the terraces.  Another is to divert the flow from the terrace outlet until vegetation has become well established.

When these waterways were built we did not know much about how fast water could flow through waterways lined with various kinds of vegetation without injuring the vegetation and scouring the soil. In 1935 the Soil Conservation Service established an outdoor hydraulic laboratory near Spartanburg, S. C., primarily for study of the flow of water in vegetation-lined channels. Such study was begun later at McCredie, Mo. At each of these places, test channels were prepared and different plants common in the locality were grown in them and exposed to flows of different velocities. Thus we have found out a good deal about the rates of flow to which different kinds of vegetative channel linings can be exposed without damage. A farmer who is planning to establish a waterway can now learn from a Department technician what sort of vegetative lining he should use, and whether he should use check dams or other mechanical protection. The needs are determined by the slope and size of the waterway, the character of the soil, and other local conditions.

For vegetation-lined channels in which water flows only part of the time, permissible velocities of flow (that is, the highest rates of flow found to be safe) are greater than for those in which water flows constantly. Where flow is intermittent some scour can be permitted; usually any part of the surface that becomes slightly eroded heals over quickly or the damage can be repaired between rains. The velocities listed here as permissible, therefore, include some that may cause slight scour.  For waterways lined with most kinds of vegetation, higher velocities are permissible when the plants are green and uncut than when they, are dormant or dead, or have been cut short. Dense, tall vegetation retards water flow, and thus reduces the capacity of the waterway.  At the Spartanburg laboratory, comprehensive waterway experiments were made on Cecil sandy clay loam with Bermuda grass, common lespedeza, sericea lespedeza, kudzu, and a mixture of redtop, orchard grass, common lespedeza, and Italian ryegrass. A few-tests were run on centipede grass, Sudan grass, and a mixture of Dallis grass and crabgrass. Seeding of Dallis grass in one of the channels resulted in a poor stand, so that crabgrass came in thickly. The experiment was continued because the same thing is likely to occur in field practice.

Bermuda grass offers almost as much resistance to injury by flowing water when dormant as when green. Bermuda grass was injured less easily by the flowing water if it had been cut frequently during the season than if it had recently been cut for the first time.

Common lespedeza, an annual legume that is very widely useful for forage in the South and grows satisfactorily as far north as southern Pennsylvania, was tested in these four conditions: Uncut green vegetation, cut stubble, fall dead stubble, and spring dead stubble. The uncut green lining was, found to protect a waterway fairly well, permitting a velocity of about 54 feet a second. When the green vegetation had been cut, a flow of more than 7 feet a second caused very little damage. In the fall, the dead stubble endured a flow of about 41, feet a second. By spring, however, the dead plants had rotted at the base and were very easily broken off by flowing water, so that any flow of more than about a foot a second caused damage. The low protective value of the dead stubble in the spring makes common lespedeza unsuitable to be used alone for lining waterways. When seeded together with grasses such as orchard grass, Italian ryegrass, and redtop, this plant does help protect waterways.

Mixed grasses are good for linings that are to be grazed, because they provide variety in the forage. The tests made with the mixture of redtop, orchard grass, common lespedeza, and Italian ryegrass showed that this mixture offers excellent protection. For channels lined with this mixture, the permissible velocity is believed to be about 6½, feet a second in the summer and after cutting in the fall, and 5 feet a second in early spring.

The plants in this mixture give better all-year protection because they have different periods of flush growth and dormancy. In early spring, when the redtop and orchard grass were dormant and the new lespedeza seedlings had not yet appeared, the Italian ryegrass was green and offered good protection.

Moreover, orchard grass and redtop grow in clumps. If either of these grasses were grown alone in a waterway, flows would tend to be rather irregular. The other plants, growing among the clumps, made the lining more uniform, with the result that it had a more satisfactory influence on flow. When flows are deep enough to submerge the plants entirely, the effect of any unevenness of the channel lining largely disappears.

It is believed that under northern conditions equally satisfactory protection would be given to waterways by the Cornell utility mixture, that is, timothy, redtop, Kentucky bluegrass, Canada bluegrass, Mammoth red clover, Alsike clover, and Ladino clover.

The test channels were free from sharp changes in grade, alinement, and cross section, and had _dense, uniform linings of kudzu. Under less favorable conditions, these velocities would be too high. When kudzu was dormant, the loose mulch of dead leaves, stems, and vines gave very little protection.

Kudzu is suitable for use in large natural waterways or gullies where reduction of cross section by its heavy growth is not objectionable.

Centipede grass, a heavy sod-forming grass adapted to the Southeast, competes successfully with other vegetation but can be killed easily by plowing. Therefore it offers promise as a channel lining to be used by farmers who object to the vigorous spreading characteristics of Bermuda grass. No tests were made on cut centipede grass, because this grass does not grow tall and is seldom cut. The permissible velocity of flow for channels lined with centipede grass and having slopes of less than 10 percent was 9 feet a second when the grass was green and 8 feet a second when it was dormant. Centipede grass cannot be grown successfully farther north than South Carolina.

Sudan grass, a fast-growing annual, was tested as a temporary cover for waterways from which flow cannot be diverted while perennial plants are becoming established. For a Sudan grass-lined waterway having a 3-percent slope the permissible velocity of flow is about 4 feet a second when the grass is full grown and green and about 3 feet a second when the grass is dead.

An objection to using Sudan grass as a channel lining is that during dry years it crowds out any permanent vegetation seeded with it, so that the channel has very little protection after the grass dries or is harvested, and it is difficult to establish permanent vegetation before the following spring.

Sudan grass is grown successfully as far north as the northern boundary of South Dakota.

Dallis grass and crabgrass offered rather poor protection in comparison with other grass mixtures. For a channel lined with these grasses and having a 6-percent slope, permissible velocity of flow is about 3½ feet a second. Dallis grass is adapted to southern conditions only.

At McCredie, Mo., tests have been made on two kinds of channel lining, Kentucky bluegrass and a mixture of timothy and redtop, in channels in Putnam silt loam.

Kentucky bluegrass occurs throughout the northern half of the United States, except where the climate is too dry. The results of the tests showed that the permissible velocity of flow for a channel of 4-percent slope lined with a poor 1-year-old stand of Kentucky bluegrass was only 3 feet a second. After another year, however, permissible velocity for this channel had increased to 7 feet a second. In any locality, the permissible velocity for a channel lined with Kentucky bluegrass is governed by the density of the stand of grass that the soil can produce and maintain. It may be as low as 1 foot a second or as high as 8 feet a second.  Kentucky bluegrass that is thick and long enough to shingle the channel surface completely can withstand velocities much higher than 8 feet a second, but the possibility that such a stand will deteriorate makes it unjustifiable to use this grass where velocity will be high.

Timothy and redtop make a good grass mixture for our purpose in northern climates. A channel lining of timothy and redtop withstood a velocity of 7 feet a second at the end of a year and about the same velocity after 2 and 3 years. Only moderate scour resulted when this lining was exposed to a flow of 8 feet a second.

For waterways on soils that are sandier and more easily erodible than the Cecil sandy clay loam at Spartanburg or the Putnam silt loam at McCredie, permissible velocities of flow are lower, of course. Channel vegetation must be carefully maintained if it is to be dense, uniform, and free from weeds. Any damage such as may be caused by rodents, livestock, or farm implements should be repaired immediately. Long-continued heat or drought often reduces the density of vegetation in waterways, with the result that parts of the soil surface are left thinly covered or entirely bare, and thus are likely to be eroded even when velocities of flow are less than those classed here as permissible.

In considering the recommendations I have made, the farmer should bear in mind that the test channels were carefully constructed and were free from irregularities and that nearly all the linings were in uniformly good condition. For waterways of irregular cross section or waterways in which the vegetative lining is expected to become less dense as it grows older, the permissible velocities are lower.

THE AUTHOR
C. E. Ramser, research specialist in hydrology of the Soil Conservation Service, grew up on an Illinois farm, was graduated from the University of Illinois, and has devoted the past 33 years to engineering research in the Department of Agriculture.  His contributions to agriculture won him the 1944 award of the John Deere medal of the American Society of Agricultural Engineers.

FOR FURTHER READING
Cox, Maurice B.: Tests on Vegetated Waterways, Oklahoma Agricultural Experiment Station Technical Bulletin T-15, 1942.
Palmer, Vernon J.: A Method for Designing Vegetated Waterways, Agricultural Engineering, volume 26, pages 516-520, 1945.
Smith, Dwight D.: Bluegrass Terrace Outlet Channels, Agricultural Engineering, volume 24, pages 333-336, 1943.
Smith, Dwight D.: Bluegrass Terrace Outlet Channel Design, Agricultural Engineering, volume 27, pages 125-130, 1946.