Skip Navigation

Speeches of HHB, Lectures on Soil Erosion

Speeches of Hugh Hammond Bennett

Address delivered in connection with the South Carolina Teacher-training program by H. H. Bennett, Chief, In Charge Soil Erosion Investigations, Bureau of Chemistry and Soils, U.S. Department of Agriculture, at Spartanburg and Clemson College, November 4, 1932; Columbia and Rock Hill, November 5, 1932.

Lectures on Soil Erosion: Its Extent and Meaning and Necessary Measures of Control

In order that the thought of the farmers of the country may be more sharply focused upon the widespread evils and cost of erosion and the necessity for its better control, it is going to be necessary for those of us who enlist in this undertaking to orient ourselves with respect to:

  1. the fundamentals of erosion processes,
  2. how erosion impairs and destroys farm land,
  3. the geographic extent of erosion,
  4. the technique of erosion-control methods, and
  5. the best methods for presenting these facts to those who must do the actual work.


The engineering phases of the problem are to be discussed by members of the staff of the Bureau of Agricultural Engineering.


Part I
The Problem

This country is standing at the crossroads with respect to the problem of conserving its agricultural lands. There is immediate and vitally important need for renewed nation-wide effort towards this end. Without further delay the attention of the country must be focused more acutely upon the seriousness of this situation, which if permitted to continue will mean inevitable shrinking of our most fundamentally important asset, the soil, to a point of grave national danger. Unrestrained erosion, rainwater running wild, is rapidly eating into the heart of immense areas of farm and grazing lands. While we are not likely to starve any time soon because of this evil of erosion, farming is going to be pushed down more and more toward the level of pauperized agriculture resulting from greatly depreciated crop yields, as the consequence of increased subsoil farming enforced by erosion-unless in the future vastly more of prevention is done than in the past. Many farmers already have really reached that level.

In closing our eyes, or in never opening them, to the evils of excessive erosion, we are to a degree that is becoming increasingly unsafe, foolishly despoiling the sloping lands of the country by permitting the loss of the richest part of the soil, the humus-charged topsoil, and after that the better part of the subsoil. Let's keep in mind that with many types of soil, this thin surface covering is the very substance of the land, insofar as relating to a good crop production. Those areas of erosive sloping land comprise something like seventy-five percent of the arable land of the country. In the Piedmont region of South Carolina, which embraces almost half the area of the state, I think we can safely put the area of erosive rolling land at about 85 percent of the region. All of this is subject to erosion in some degree, while being used for the clean-tilled crops, as cotton, corn, potatoes and peaches.

Erosion Began With The Clearing Of The First Sloping Field.

Our national habit of ignoring the significant fact that every heavy rain of summer time, and many of those of winter, takes its toll from the surface of the ground, its usury of the thin, humus-charged top layer developed by natural soil-forming processes through ages of rock decay and soil building, began with the clearing of the first sloping area in the tidewater country of eastern America This wastage has continued almost without let or hindrance over far too great a proportion of the country for national safety. It has become a country-wide menace that must be opposed now with all our fighting capacity, otherwise the difficulty of control will become greater and greater as the devastation digs deeper into the more unproductive and unstable subsoil, and continues to reach out over an ever expanding area, giving us more and more subsoil farmers with little opportunity, generally, to wrest a satisfactory living from such stubborn material.

This past summer I had the opportunity of looking over some of the ground first cultivated by white man in this country, near the site of the first permanent settlement at Jamestown, Virginia. In this region the flat river benches and bottoms remain much as they were when those adventurous settlers, coming over on the God-speed, Sarah Constant and Discovery, began clearing the plains between the lower James and York rivers, three and a quarter centuries ago. Much of the sloping areas of the higher watersheds between the major streams entering Chesapeake Bay is still predominantly poor because of the wide-spread erosion that began three hundred years ago. Ancient gullies are to be seen there, some partly healed by reestablished forest, others still deep enough to furnish indisputable evidence of the truth of an early observer who, speaking of the rolling areas near the birthplace of America, said: "... farm after farm ...worn out, washed and gullied, so that scarcely an acre could be found in a place fit for cultivation."

This recent experience in the Virginia Tidewater country took me back twenty-seven years to the time when a soil survey was being made of Louisa County, in the same state, by W.E. McLendon, of Bishopville, South Carolina, and myself. After arriving in that old Piedmont country, once represented in the House of Burgesses by Patrick Henry, a letter was received from our Chief, the late Dr. Milton Whitney, saying that, in addition to the task of classifying and mapping the soils, he particularly wished, we would exert every effort to ascertain the cause of the reputed poverty of the lands of Louisa County. After some weeks in the field we were puzzled over learning that wherever the land had never been cleared a good depth of mellow loam or sandy loam topsoil was invariably present; whereas; in practically every sloping field which had been in cultivation long enough for the stumps to have disappeared neither loam nor sandy loam was found as a rule, only red clay loam and clay loam. In time we came to realize the significance of this:

That the original soil had been removed, down to or near the clay that underlies practically all of the fifty million acres comprised within the Piedmont region, extending from Alabama to New York City.

It was at that stage of scientific enlightenment that my thoughts went back over another span of years to a time when I was digging markers with a hoe at one end of a home-made, wooden horse which my father was using for laying out terrace lines on his rolling farm in south-central North Carolina. At that particular time, his laconic reply to my inquiry as to why this was being done had not impressed me with much vividness. He had said: "We are doing this to keep the soil from washing away." Subsequently, however, that answer came to mean very much to me.

Soil Erosion The Most Serious Continuing Farm Problem.

Following the discovery in Louisa County, it became a part of my duty, as I saw it, to learn more about this process which could so completely change the character of the land. Studying the soils in detail in nearly every county of the Cotton Belt and in numerous other counties, as well as in the various countries of America north of the equator; two definite conclusions were reached: First, that soil erosion constitutes the most serious continuing farm problem in the United States; and, second that no other modern nation of the Western Hemisphere, north of the equator, is wasting; its agricultural lands as rapidly as the United States, even though farming has been going in the countries to the south much longer than in this youthful land of ours. Some time ago it was concluded, also, on the basis of the available information, that no nation or race within historic time has been so wasteful of its agricultural lands as we of the United States. To be sure, vast areas have been laid waste in China, Persia and other old countries, but those countries used their lands for thousand of years, whereas we have used the oldest of ours for only about three hundred years, the greater part for only about forty to eighty years.

A recently evolved theory, based on considerable evidence, is to the effect that the ancient Maya of Peten, Guatemala, deserted that once densely populated area because of excessive washing of the soil, coupled with the choking of the navigable lakes and connecting waterways by deposition of erosional debris. Although large areas of tropical America are highly resistant to erosion because of the physical characteristics of the soils, erosion is very destructive in some parts of these warm regions, especially on the more youthful limestone and schist-derived soils. Along the southeastern border of Peten I have observed erosion on limestone clay that resembles the red limestone lands of northwestern Arkansas, northwestern Georgia, eastern Tennessee and other parts of the United States. It is quite possible that the same thing took place on an extensive and disastrous scale in the little State of Peten, where eight million people once dwelt.

Erosion an Old Problem.

We have been living with erosion for a long time. Mankind has combated it for unnumbered ages. We have opposed it to some extent in this country; indeed, we have fought it rather valiantly in some localities, notably in the southeast, and recently across the Mississippi in Texas, Oklahoma and Arkansas. However, the real task lies ahead.

Five years ago, A. T. Strahorn, of the Bureau of Chemistry and Soils, found in Palestine olive trees a thousand years old still living on walled terraces that were constructed during the Dark Ages. On some of the islands of the Mediterranean bananas are being produced on terraces of the European type which are supposed to have been built before the time of Christ. The aborigines in parts of the Philippines, the Ifugaos and Bontoes, live on rice grown on terraced strips of mountain slopes, the beginning of whose construction antedates history. Washington, Madison, Jefferson, Edmund Ruffin and others clearly recognized the evils of the process. Washington gave up the growing of tobacco, planted clover and practiced crop rotations on his Mt. Vernon estate in order to check erosion. You can still see old gullied areas on the lands that once were his. Some of these are found in close contact with venerable cedars that must have been there in Washington's time.

In 1913 Jefferson, writing about his farm in Albemarle County, Virginia said: "Our country is hilly and we have been in the habit of plowing in straight rows, whether up or down hill, in oblique lines, or however they lead, and our soil was all rapidly running into the rivers. We now plow horizontally following the curvature of the hills and hollows on dead level, however crooked the lines may be. Every furrow thus acts as a reservoir to receive and retain the waters, all of which go to the benefit of the growing plant instead of running off into the streams. "

Long ago, Ruffin, one of the best agriculturists the nation has produced, said that washing of the land was more destructive than all the direct damage done by crops. But Washington's rotations, Jefferson' s contour plowing and the terracing (the American type of terrace) which had its beginning about a century ago somewhere in the Southern Piedmont, failed to stop the wastage; rather, these methods were employed by too few or with too little efficiency and persistence to exert any tremendous effect, except in some localities. The vast majority of farmers went ahead cultivating unprotected slopes until the soil was swept off or until the fields were spotted with comparatively infertile clay exposures, then cleared new land. This practice continues to some extent even today.

Observing the more conspicuous manifestations of soil rushing the gullies that scar so many slopes, farmers outside the Cotton Belt have occasionally undertaken some small measure of control, usually in the form of dumping brush, rocks or earth into the ravines. With respect to the other type of soil washing, the less conspicuous process of sheet erosion, which is the real giant in the camp of soil-impoverishing agencies, most farmers living beyond the border of the Cotton Belt, and far too many living within the Cotton Belt, have been as soundly asleep as if anaesthetized. In the uplands of a few of the southeastern states the problem has been recognized and rather effectively opposed, locally, for nearly a century through the building of really efficient soil saving terraces. Unfortunately, too many upland farmers either have failed to recognize the slow depletion of their sloping fields or for one reason or another have failed to take issue with this most relentless enemy to continuing soil productivity. Many have failed to maintain their terraces, or the terraces were not of proper size or grade to begin with.

In most parts of the nation the problem has been almost completely overlooked. In some states where erosion is about as bad as anywhere, I have gone year after year checking over soil surveys of county after county without seeing one single terrace or any other effective measure for slowing down wastage of sheet erosion. As stated, an occasional farmer was found who had undertaken at a late stage to check his gullies with brush or stones; but even these rare efforts generally had the appearance of half-hearted interest, or the brush or stones were improperly laid and did little or no good. On the other hand, I have seen on almost countless rolling farms effective measures for speeding up the process, such as the vicious habit of running rows up and down the slopes, a practice based on the erroneous conception that ever crop row should be laid out in as nearly a straight line as possible, regardless of the consequences. This last error of tillage is a costly practice that the average South Carolina farmer is not guilty of.

Cause of Lack of Interest.

There was, necessarily, something back of this general failure of farmers and agricultural specialists to recognize the evils of unrestrained soil wasting. As I see it, the failure was due, in a very large degree, to naked ignorance. Somehow the notion got into the heads of too many specialists, as well as into the pages of publications, .that soil erosion was an evil confined to the Southern States, to China and other far-away places. An unfortunate idea, this. From the standpoint of the nation it helped to keep on the blindfold too long. Some specialists, unaware of the facts, even went so far as to explain why northern soils do not erode. If they had only looked about the countryside with seeing eyes they might have educated themselves and made a good case of the question: Why do some soils of both the North and South erode much more seriously than others? On the whole, there has been more erosion in the Southern States because less grass has been grown and because the soils are not so well protected, in winter, by freezing and with snow blankets. But this does not mean at all that the problem is not serious in many parts of the North. As a matter of fact, the loss of 19 1/2 tons of soil per acre by a one inch rain, up near the Iowa-Missouri line, in 1931, has not been equaled by any measurement made anywhere else, to date, at least not on land so gently sloping as that was, the gradient being 8 feet in a hundred, or 8 per cent. And the loss of 26 tons of soil per acre from eroded land, along with 4l per cent of the rainfall, from 1 4/5 inches of rain falling on land of 16 per cent slope, during August this year, in west-central Wisconsin, completely smashed all records.

Undoubtedly there is a feeling on the part of numerous people that erosion belongs to the natural order of terrestrial dynamics. It does, that is, normal erosion does; but not the kind I am talking about-man-induced erosion, of which more will be said later. There has, been, then, too much of the point of view that erosion is a necessary evil that must be borne with Our thought in this direction has been clouded by the inaccurate conception that our good farm lands were inexhaustible and limitless. We began thinking this way when this seemed to be a fact, when a comparatively small population looked out upon the fertile lands of the eastern empire of hardwoods and beyond over vast expanse of prairie and plain. We felt the same toward our timber supply and the great herds of buffalo which were slaughtered for their hides. But we were wrong. Our better lands have been under cultivation for some time. We can not extend this area, soil building is too slow; but we are permitting it to be reduced-by excessive soil washing.

Enormous Area Ruined By Rainwash.

We have found that not less than 17 1/2 million acres of formerly cultivated land have been destroyed in this country by gullying, and deep washing, or so devastated that farmers can not afford to undertake its reclamation. In addition, about 4 million acres of bottom land, formerly cropped, have been rendered essentially useless by overwash, or by increased overflow resulting from choked stream channels. But these losses, although comprising a larger area than the total extent of arable land in Japan proper, are almost insignificant in comparison with the vast area subject to serious sheet, washing-that slow type of washing which steals a layer of soil with each successive rain.

Sheet Erosion.

It is unfortunate that most people conclude that where ever gullies have not developed there is no erosion. This is wholly incorrect. It is unfortunate, also, that very few thoroughly understand the earmarks of sheet erosion, even some of our agricultural specialist have not seemed to sense the viciousness of this form of soil impoverishment. In order to grasp the full significance of the evil process, it is necessary to know how to identify soil types and to compare the successive layers of soil of a given area with the corresponding layers of another area, in order that what takes place in an eroding field can be precisely checked against what is taking place on the same soil type, occupying the same slope, or different slopes, in other fields, as well as in areas that have never been plowed or caused to wash by repeated burning and over-grazing.

Sheet erosion is that phase of land washing which removes a thin covering of soil from large areas, often entire fields, more or less uniformly during every rain, producing runoff. The process goes on slowly, so slowly that its effects are frequently unobserved until spots of clay and even rock begin to make their appearance over sloping fields, at which stage, unfortunately, it is usually too late to fully repair the damage, since it is entirely impracticable to haul soil back into a field once it has departed on its journey to the sea.

Sheet erosion pursues its course in the direction of soil depreciation, and even soil destruction, with unremitting persistence, taking its toll of topsoil with every rain heavy enough to cause water to run downhill, across unprotected cultivated slopes. We all really understand this, though many of us have not thought very much about it. We know that rainwater flowing out of fields devoted to clean-tilled crops is always muddy, never clear. And when we think about it a second time we come to appreciate the significant truth that this muddy water is discolored with soil material. It is red, yellow or black, according to the color of the soil across which it has flowed. Simple mathematics tells us what must eventually happen where the process is not effectively opposed: The thin covering of surface soil whittled away, down to much less productive subsoil.

Some soils wash much faster than others, and the steeper the slope the more rapid the progress of erosion, under ordinary conditions. Some gravelly soils and porous clay lands (such as the Davidson soil of the Piedmont) are so absorptive of rainwater that erosion may proceed even slower than on slopes not nearly so steep. Usually, however, the steeper the slope the greater the speed of runoff. It has been said: "The power of moving water to scour or loosen soil. particles varies as the square of the velocity, and the power to transport this material varies as the velocity. "2 This, of course, refers to ideal conditions. The actual field conditions are generally highly varied, complicated by numerous factors. we are not so much concerned with the preciseness of rules as with the practical facts. Our measurements of soil losses at the erosion experiment stations are showing that erosion is most severe on the steeper slopes, with the predominant soils of the country. It so happens, however, that thus far erosion is proceeding faster from an 8 per cent slope of Kirvin fine sandy loam than from a 10 per cent slope of Nacogdoches fine sandy loam, with the same rainfall and treatment, at the erosion station, near TyIer, Texas. We can for the moment, nevertheless, dismiss these exceptions as representing local soil conditions that call for special treatment.

Time Required to Build the Topsoil.

According to some of the quantitative measurements made at the erosion stations, nature requires not less than 400 years to build one single inch of the topsoil of some of our important types of farm land. This appears to be true, for example, of the very extensive soil, the Shelby loam, occurring over the rolling parts of the Corn Belt, in northern Missouri and Southern Iowa. This we have learned by separately measuring the rate of soil losses under continuous corn production and under a continuous cover of grass, assuming that land supporting a good grass sod closely approximates the conditions under which nature developed the mellow surface layer of the Shelby loam soil, which was largely a grass-covered prairie type in its virgin state. It takes just seven years under continuous corn farming, in northern Missouri and Southern Iowa to wash off one inch of the Shelby loam soil where the slope is about four feet in a hundred, and only one year to remove this same depth of surface material from land sloping a little more than eight feet in a hundred. In other words, on land which originally produced in the best years upwards of 75 bushels of corn per acre, man is now allowing to go to waste in from one to seven years that which took nature not less than 400 years to build. The average depth of soil on the average slope of the Shelby loam region is about seven inches; the clay beneath, which is reached by erosion proceeding under the present system of farming in from about seven to fifty years, produces in a good year only about 20 bushels of corn per acre, and often nothing in poorer crop years.

In 1931, at the Bethany Erosion Station, good uneroded soil, broken out of bluegrass sod five years previously, produced without fertilizers 51 bushels of corn per acre; whereas, severely eroded soil which had been cultivated 40 years produced less than l4 bushels per acre. It is perfectly obvious that no business could very long withstand such evil treatment. It is not being withstood; I have recently seen farms on the Shelby loam which were abandoned because of erosion. In one instance, the owner, when warned that if he did not bestir himself to stop the soil washing he was going to ruin his farm, replied that his farm owed him a living and, that he was going to get a living out of it. He failed to do that, however; he lost his farm and is now living on another man's land. The farm was so deeply washed that it will not now produce grass; it is a weed farm, without present value.

Probably in the Piedmont region it took even longer for nature to build up the topsoil. I would not be in the least surprised if it turned out that not less than 1,000 years at least are required by nature to build an inch of good, rich Piedmont topsoil. Measurements now under way will soon give us some light on the subject.

Importance of the Topsoil.

It seems strange that we have paid so little attention to the vital importance of the topsoil, the thin humus layer, charged with decaying vegetable matter, containing the bulk of available plant food and representing the abiding place of incredible hosts of beneficial micro-organisms. The soil immediately beneath the leaf-mold of forested areas often is higher in content of phosphorous and lime than the layers below, it is often neutral or even alkaline where the subsurface material is strongly acid. This is due to basic constituents brought up from below by plant roots and concentrated in this surface layer through the medium of decomposing leaves, and grass. Our experiment at the erosion stations are showing not only large decreases in the productive capacity of land following the washing away of soil, but even impairment of the quality of some of the products grown. Cotton, for example, has tested much weaker in strength where grown on eroded soil, as compared with that produced on uneroded soil Immediately alongside; and the content of oil in the cotton seed much lower-according to results obtained at the Oklahoma erosion station.

At the Western Kansas Station, on a 5-per cent slope, the 1931 loss of rainwater falling on native sod was. 05 per cent of the total precipitation, while the corresponding loss of soil was .0025 tons per acre; whereas, the losses from clean-tilled Kafir were 11.79 per cent of the rainfall and 20.85 tons of soil per acre. In other words, native sod held back 236 times as much of the rainfall as Kafir (following wheat) and 8,340 times as much soil. From wheat grown on slightly eroded soil, the losses were 2.79 per cent of the rainfall and 27 tons per acre of soil; whereas, from wheat grown on severely eroded land (desurfaced down to the subsoil), 15.43 per cent of the rainfall was lost and 3.40 tons of soil per acre. This means that in the growing of wheat in western Kansas eroded land (of the extensive Colby silt loam type) is losing 5 1/2 times as much of the rainfall and 12 1/2 times as much soil as land still retaining a considerable part of the original topsoil. Grass closely clipped to simulate an overgrazed condition lost 6 times as much water and 12 times as much soil as the normal prairie sod. Wheat on normal soil (not deeply eroded ) produced 26.7 bushels per acre, as against only 5 bushels on severely eroded land.

It is more pertinent to the farmers of the Piedmont of South Carolina, however, to know that at the erosion station near Statesville, North Carolina, located on red clay loam, of which there is much in this state (287 thousand acres of the same class of land having been mapped in Spartanburg County alone), the average loss of soil and water in 1931 amounted to 14 tons per acre and 12 per cent of the total rainfall, respectively, from cotton; and 65 tons per acre and 26 per cent of the rainfall, from bare ground; whereas, the corresponding average losses from lespedeza were only 1.5 tons of soil and 10 percent of the rainfall. Now, from exposed subsoil of the same original type, receiving the same rainfall and cultural treatment, the corresponding losses, in 1931, were, from cotton, 17 tons of soil per acre and 11 per cent of the rainfall. Thus we see that erosion speeds up as the soil is whittled off, and that land planted to cotton erodes 9 times faster than where it is planted to lespedeza, and at the same time loses more water.

Erosion Speeds Up.

At the Arkansas-Louisiana-Texas Sandy Lands Erosion Station, Tyler, Texas, in 1931, on Kirvin fine sandy loam having an 8-per cent slope, 16.6 per cent of the rainfall was lost as runoff from land planted to corn; from cotton the loss was 13.4 per cent; and from Bermuda sod, 2.5 per cent. From deeply eroded areas, with the clay subsoil exposed, the average loss by run-off from three cotton plots was 17.6 per cent of the rainfall. The corresponding soil losses from the same plots were as follows, expressed in tons per acre: 20.7 from corn land, 15.8 from cotton, and .5 from Bermuda sod. The deeply eroded areas planted to cotton lost an average of 55 tons per acre, or more than three times as much as was lost from the topsoil. Thus, it is seen that on this soil both erosion and runoff are highest on land used for corn and cotton, and that they are very low from grassland. The very much higher loss of soil from the subsoil plots shows that erosion in this region speeds up as the process continues.

Effect of Length of Slope.

At the Bethany, Missouri Erosion Station land planted continuously to corn lost soil, in 1931, from a plot 146 feet long on an 8-per cent slope, at the rate of 104.71 tons per acre, along with 28.22 per cent of the precipitation as runoff. One one-inch rain, on the 5th of June, caused a soil loss from this plot at the rate of 19 1/2 tons per acre. The corresponding runoff was 39.7 per cent of the precipitation. A 73-foot plot immediately alongside of the 146-foot plot lost soil at the rate of 84.08 tons per acre, with a corresponding runoff amounting to 30.10 per cent of the rainfall. The rain of June 5 removed soil from the shorter plot at the rate of 15.7 tons per acre, with a 49.5 per cent rainfall loss. In other words, length of slope here considerably affects the rate of soil loss. At the Central Piedmont Station in North Carolina the losses of soil from a 10-per cent slope were as follows: 14 tons per acre in an area 145 feet long; 12 tons from an area 72 1/2 feet long; and 16 tons from an area 145 feet long. thus the loses do not conform on these markedly different soils. (All these areas were protected, so that there was no intake or loss of water above the catchment basins.)

Annual Soil Losses.

A few years ago, before we knew very much about the enormity of the problem, it was estimated that erosion was washing out of the fields and pastures of this country not less than 1,500,000,000 tons of soil annually.3 This estimate astounded a good many people. Now, on the basis of measurements at one of the recently established soil erosion experiment stations, it is indicated that in 1930, which was about an average year from the standpoint of seasonal condition, 16,534,800 acres of the rolling Red Plains region of Texas and Oklahoma lost 440 million tons of soil. On the 10th of May, 1930, one 5-inch rain in the Black Belt of Texas washed off the rich black topsoil from slopes of only 4 per cent at the rate of 23 tons per acre. This rain, according to this measurement, took a toll of not less than 100 million tons of soil from the sloping part of this famous cotton area. In other words, when we get down to the task of quantitative measurements, we begin to see that here is a problem whose destructive potency not only greatly exceeds all previous ideas concerning it, but comes close to exceeding the possibilities of human comprehension.

We have pointed fingers of warning toward China as a terrifying example of wasted agricultural lands. The devastation in that old, old country has indeed been appalling. Clearing the timber from the uplands and cultivating the stripped slopes without any protection, the soil has washed away from millions of acres of once productive fields; hideous gullies and even canyons have dug deeply into the devastated areas. Hordes of human beings, leaving the erosion-destroyed slopes of the Celestial Empire, have densely concentrated upon the flat valley lands. There, every available foot of ground is used for cops. Even the roots of crops are dug for fuel, and all organic refuse is scrupulously saved for fertilizer. The older fields of the alluvial plains are fertilized with fresh alluvial deposits collected in pits during periods of overflow. After four thousand years of building levees and digging canals, the Yellow River of China, which in that country is known as the "Scourge of the Sons of Han," broke over its banks in 1877 to drown a million human beings. In 1852 this Titan, in mighty flood, changed its channel, sweeping northward to enter the Yellow Sea 300 miles beyond its former mouth. This sea, a part of the Pacific Ocean, derives its name from water colored yellow with the debris of erosion, brought down from the still wasting slopes far up the valleys.

The plant food removed from the fields and pastures of America every year by erosion greatly exceeds that removed by the crops harvested. That taken by crops can be restored in the form of fertilizer, but that taken by erosion can not be restored, because this malevolent process takes the whole body of the soil, plant food and all. Land impoverished strictly by plant food depletion, as sometimes results from continuos growing with the clean-tilled crops, is not worn-out land; the only worn-out land that we have is that which has been so badly washed that it would be entirely futile to undertake it reclamation.

Soil Depth.

Many people have the idea that the soil (as distinguished from the subsoil) is much deeper than it really is. On examining 172 soil samples collected from 34 states and representing, very largely, important upland types it was found that the soil depth as recorded averaged only 9 inches. Many of our most important types of farm land range from only about 3 to 7 inches in depth of topsoil. From some of those we are losing this invaluable surface layer within 10 to 40 or 50 years of cultivation, depending on the character of the soil, the slope and the crops grown. When this layer is gone, the farmer's principal capital is gone. The material below is too poor in countless instances for profitable production, even when prices are good. We cannot build the topsoil back after it is gone, nor can we haul it back into the fields. We can increase the productivity of some eroded land, to be sure, by growing the soil-improving crops, liming and fertilizing; but we cannot replace the soil that nature put there. By terracing and growing grass we sometimes catch material washing down from slopes above still retaining a covering of soil, but that is not what I have reference to.

The Federal Land Bank of Houston, Texas, lends on the basis that the top six inches of soil represents the farmer's principal capital. If the bank discovers that a farmer who has been granted a loan is permitting his fields to wash at a rate faster than six inches in thirty-five years on a thirty-five year loan, foreclosure proceedings ensue. Very few foreclosures have been necessary, however, because if the farmer does not know how to build terraces, the bank sends out its soil-conservation expert and shows him how.

Erosion Surveys.

Some of you may have heard about the findings of the Oklahoma soil erosion survey, completed in 1930. It is such a terrifying example of unnecessary soil depletion that the results will bear repeating. Of approximately 16 million acres in cultivation in that state, 13 million acres were found to be suffering from severe erosion, 5,726,000 acres having reached the stage of gullying. Of 1,694,000 acres abandoned in the State during the past few years, 1,359,000 acres were abandoned because of erosion.

Two years ago the bureau of Chemistry and Soils made a general survey of seventy counties lying in the Brazos River watershed of Texas, finding over one large section within this watershed 88 per cent of all the cultivated area either gullied or subject to wasteful sheet washing, the area thus affected amounting to 6 million acres.

This sort of thing is to be found in numerous other localities, as in northern Missouri, southern Iowa, southern Indiana and Ohio, southwestern Wisconsin, Kentucky, Tennessee, parts of Kansas, Nebraska, California and other states. In many of these regions even the topography of some localities has been changed. Gullies have ribbed thousands of slopes than once were smooth and rounded. Unproductive clay sticks out in countless places.

Effect of Soil Variability on Moisture.

Not only does soil character enormously affect the rate and nature of erosion, but it markedly affects also the moisture retentiveness of soil. For example, at the beginning of the season May 13, 1931, virgin soil of the Vernon fine sandy loam at the Red Plains Erosion Station in central Oklahoma, planted to cotton, contained approximately the same amount of moisture as the desurfaced clay subsoil; i.e., 15.08 per cent and 15.38 per cent, respectively; but at harvest time, September 1, the uneroded soil contained 5.91 per cent of moisture, while the exposed clay contained 7.12 per cent. The fact that the virgin soil produced 162 pounds of cotton per acre, as against 98 pounds for the eroded soil, shows that moisture consumption by the much heavier vegetative growth on the virgin soil was much greater than on the clay. Moreover, a smaller proportion of the moisture rigidly fixed in the smaller interstices of the clay particles was available to the plants than of that contained in the larger pore spaces of the fine sandy loam topsoil. In other words, the moisture efficiency of a soil is, in a large measure, a function of the structural efficiency of a soil. Structural efficiency depends on a number of factors, such as texture, vegetable-matter content and structural porosity.

In contrast to the moisture behavior on Vernon fine sandy loam, it is interesting to observe the corresponding behavior on the soil and subsoil of the Houston black clay in central Texas, in 1931, under similar conditions of precipitation (very low rainfall). In this instance, the uneroded clay soil at the beginning of the crop season (corn in this case), on April 10, contained 23.5 per cent of moisture, as against 16.8 per cent in the adjacent desurfaced, humus-free chalky subsoil. The moisture in the normal soil continued to exceed that in the exposed subsoil throughout the growing season, until about the time the corn began to mature, when the moisture content of the normal soil dropped below that of the subsoil, or to 13.8 per cent as against 15.9 per cent. The fact that the uneroded soil produced seven times as much corn as the exposed subsoil undoubtedly explains the reversal of the moisture situation toward the end of the season, when the heavier growth called for a larger moisture supply. Here, too, structural efficiency of the soil obviously plays an important role with respect to moisture efficiency, but the process involved is somewhat different from that in the case of the Vernon fine sandy loam.

Erosion Markedly Affects Vegetation.

The results of careful vegetation surveys on equal areas of eroded and uneroded Houston black clay in central Texas have shown extraordinary changes with respect to native flora, transmutations due entirely to the effects of stripping off the topsoil. The virgin grassland vegetation on uneroded soil, in one representative instance, covered 94 per cent of the surface, and 93 per cent of the plants were grasses. On severely eroded soil, originally of the same kind and occupying about the same slope, which was cultivated and then abandoned, 60 per cent of the surface was bare, and of the plants present only 1 per cent consisted of grasses, while 39 per cent consisted of woods.

Similar surveys in Oklahoma, Kansas, North Carolina and New Jersey have shown in every instance much the same thing; namely, a complete upsetting of the natural vegetative conditions by erosion. We see conspicuous examples of this all over the eastern part of the country: pine forests, sassafras and persimmon thickets, smilax vines, poverty grass and golden rod replacing magnificent forests of hardwoods.

Man-Made Erosion.

The erosion that we have been emphasizing, as previously stated, is of the man-induced kind, the abnormal kind, as distinguished from normal or natural erosion, under which the removal of surface soil goes on exceedingly slowly, due to the favorable conditions of ground stabilization, such as is established and approximately maintained through the instrumentalities of vegetative cover, the relatively high absorptive capacity that goes with normally developed soil porosity and the restraining influences of gentle slopes or angle of repose. Vegetation and soil porosity (such as nature establishes), separately and collectively, exert a tremendous opposing force to transportation of soil material by running water, slowing down surface removal to a point where equilibrium between rate of erosion and soil building from the parent materials beneath is almost, but not quite, established. Had these counterbalancing natural processes attained equilibrium some eons ago, we would not have throughout the world nearly so many valleys and glens and canyons. We would have instead vastly more flat land of a poorly drained nature, or what perhaps would be worse, enormous stretches of severely leached soil such as are known as laterites.

Let's consider for a moment the powerful effect that the ground-cover of forest leaves has upon soil absorption of rainfall. At the erosion station in central Oklahoma the forest-litter was burned from a measured area of post oak-black jack timber in the spring of 1930. Another area of the same size, immediately alongside was left undisturbed. In May of that year, during a period of almost continuous rainfall, the runoff from the unburned plot was at the rate of 250 gallons per acre, while that from the burned plot was at the rate of 27,600 gallons per acre. The excess of runoff from the burned area over that from the unburned area, plus the water-holding capacity of the leaf-litter covering (16.7 tons per acre), was approximately 90 tons per acre. Thus, the effectiveness of a thin cover of leaves as a protection to the soil is seen to be far greater than the mere capacity of the leaf covering to absorb water. The leaf cover obviously functions to send clear water down into the soil rather than the muddy water that flows over unprotected burned areas, which latter chokes up the pore spaces developed through natural processes of soil building (as holes formed by decaying roots and by insects and worms, and the openings that normally go with a mellow humus-charged soil).

And so, when we remove the trees, shrubs and grasses and plow up the ground we lay bare the soil to the wrath of the elements. Erosion is vastly speeded up; and we find that instead of being the "one immutable, unchangeable permanent resource," it has been described, the soil is one of the most destructive [destructible] of our great natural resources, save such things as coal and petroleum whose utilization in the scheme of civilization calls for their complete destruction by fire. The soil is not consumed as by burning, of course; but when it is washed out of the fields it is destroyed, insofar as having value for the farmer who has lost it. And again, as a matter of emphasis, it should be remembered that soil can not be built back in the ordinary life of a man. It can be improved by growing soil-building crops, by use of manure and fertilizer and with the aid of terracing and strip-cropping, but it can not be restored very easily, if at all, to the original condition established by nature.

Deposition of Eroded Matter.


Beyond the washing of soil from sloping areas, the products of erosion go to fill reservoirs, stream channels and to cover fertile bottom lands with relatively infertile sand. Some of the deposits laid down from stream overflows are beneficial, but often they are not because the alluvial soils are already highly productive and need no additional sediments. A bad feature of sedimentation is that sand is assorted from the waterborne material and laid down over good soil or deposited in channels to make streams more susceptible to overflow.

The Mississippi River carries into the Gulf of Mexico every year 418 million tons of solid material, aside from enormous quantities of dissolved matter. The streams from the Piedmont of South Carolina are doing precisely the same thing, though on a smaller scale, of course. But this is not the full measure of the evil of this traffic in farm land, by any manner of means. Very much more of the material washed out of the fields every year is temporarily stranded en route to tidewater; along lower slopes and in the channels and over the bottoms of every river, creek, branch and drainageway.

The Nation's Most Priceless Asset.

Our soil is our most priceless asset, and is likely to continue to be. Marvelous as have been the discoveries of chemistry, it appears that there yet remains to be manufactured upon a purely synthetic basis one single major food product. It is unsafe to make predictions relating to future discoveries, but all evidence points to the probability of the soil continuing to be the primary source of the principal food and raiment of mankind. At any time the land may be called upon to produce largely increased tonnages of commercial cellulose and even fuel for our internal-combustion engines. It is, therefore, an economic necessity and a patriotic and moral obligation to preserve this absolutely vital resource-this product that can not be entirely be rebuilt within the ordinary span of an human life.

Better Land Utilization.

Some very extensive soils are inherently poor. These are rapidly reduced to essential sterility by erosion. Other soils ranging from fair to excellent in productivity are quickly reduced to a condition of poor land if not protected against the wasting force of erosion. More and more we must strive for the maintenance of the virgin fertility of our farm lands, or that part of it still remaining as topsoil. Indeed, the necessity to conserve the soil is acutely vital to the welfare of the state and nation. From now on you are likely to hear more and more about better land utilization, the use of the soil more in accordance with its natural crop adaptations and fitness for cultivation, and the handling of it more in accordance with the peculiar needs of a particular type of farming. Marginal and submarginal land, that is to say, poor and very poor land, apparently have about had their day as crop-producing land, for the while at any rate. While failing to provide an adequate living for those operating on them, they have, nevertheless, in the aggregate contributed largely to the stores of unsold crops. There are some instances where poor soils can be farmed with fair success, but to a large extent such lands constitute a millstone about the neck of those who farm them. Cheap production calls for good land; the high-cost producer of crops is likely to be left too far outside the limits of stable existence for comfort.

How Crop Yields Have Held Up.

Notwithstanding the vast continuing losses caused by erosion, we are not on the verge of a land shortage. In spite of the appalling wastage, we have recently been confronted with the anomalous situation of having on our hands apparent large crop surpluses. Improved varieties of crops, largely increased use of fertilizers and labor-saving machinery, and the abandonment of worn-out land for land still retaining a part of the original topsoil have contributed toward continuing big crops. These improvements speak of progress, in one direction at any rate. They are not given for the asking, however, Machinery does not minimize erosion, unless it is built and used for that specific purpose. It may even cause the speeding up of the process. And it should be remembered that notwithstanding the improvement of seed and cultural methods, the improvement and increased use of farm machinery, the increased use of fertilizers and soil-improving crops and the far-reaching assistance of educational research and extension services, at large cost, our crop yields are not all increasing. In numerous localities the yields have fallen. The average yield of corn for the ten-year period 1871 to 1880, inclusive, was 27.04 bushels per acre, as against 26.13 bushels for the ten-year period 1921 to 1930. The highest and lowest annual yields for the first period were 30.8 and 20.7 bushels per acre, respectively; the corresponding yields for the later period were 29.6 and 20.6 bushels per acre. Wheat yields increased from a ten-year average of 12.4 bushels per acre for the period of 1871 to 1880, inclusive, to 14.33 bushels for the period 1901 to 1910, inclusive; but for the last decade, 1921 to 1930, inclusive, the yield was 14.25 bushels. The average yield of cotton for the ten-year period 1871 to 1880, inclusive, was 186.42 pounds per acre, as against 152.96 pounds for the period 1921 to 1930, inclusive.4 In numerous localities the yields of these crops have declined much more than these national averages, even to the extent of causing much abandonment of land.

The Meaning of Conservation.

Finally the nation is coming to understand the meaning of conservation. Our resources in land, timber, mineral products and wild life were so vast in the beginning, that it was quite natural for us to think of those things in terms of permanent abundance and inexhaustibility. It was quite natural also for us to project the wasteful practices, such as always characterize pioneer conditions, far out beyond the bounds of those primitive times. For many decades conservation was considered a sort of euphonious term, vaguely construed and patiently granted a fitting cloak for those harmless humans who, according to the ideas of the average man and woman, were born to look on the dark side of life.

We are slowly coming out of this hypnotic state of mind. We are beginning to see that the spirit of "let it alone" and of imagined self-sufficiency are erroneous conceptions with respect to land usage, and that stubbornly following their lead can only take us into deeper difficulties. Many of us are no longer inclined to look upon wasted hillsides of shallow soil and exposed clay and gully-destroyed fields as an unavoidable act of nature. Now we are seeing such things in their true inglorious meaning, the product of man's wasteful methods. When we are told that the Grand Canyon of the Colorado had but a trench thousands of feet deep before white man crossed the Atlantic, we immediately answer: "Yes, and it probably took thousands of centuries for the cutting." Just speculate for a moment as to what would have happened if such mighty chasms had cut down at the rate gullies are now being formed in countless fields, such, for example, as Providence Cave in southwestern Georgia, a ravine which has been gouged out to a depth of 150 feet in fifty years, or a 75-foot gully near Ventura, California, from which beans were harvested in 1914.

Topography and General Extent of Erosion In South Carolina.

Having outlined the behavior of erosion, its effects and extent from the national standpoint, let's examine the situation as it relates specifically to South Carolina. The Piedmont portion of the state will be emphasized because it is here that the evil is most costly and extensive.

The Piedmont, together with a comparatively small segment of the Blue Ridge Mountains, comprises about one-third of the state-the western part, or that part lying above the fall line, marked roughly by a border zone through which the waters of the Catawba, Broad, Saluda and Savannah rivers cascade to the lower coastal plain region. This line crosses the state near Augusta, Columbia, Camden and Cheraw. Thus, this upper country of the Piedmont and Blue Ridge Mountains covers part of Aiken, Lexington, Richland, Kershaw and Chesterfield counties and all of the 18 counties to the west of these.

The region is characteristically of rolling topography, with a range of elevation from around 200 to 300 feet above sea level, along the fall line, to around 900 to 1,000 or a little more along the line of contact with the Blue Ridge. As viewed from a commanding position, the general upland level has the character of a plain, with here and there isolated peaks of more resistant rocks (generally quartzite) rising conspicuously above the dominant skyline (such as the peaks of the Kings Mountain group). In detail this great area, although once a true plain sloping gradually toward the sea, has been dissected by an intricate system of rivers and lesser drainageways. Generally, there is a difference in elevation from the crest of the inter-stream divides to the stream bottoms below, ranging from about 50 to 200 or 250 feet. Parts of the region are of relative smooth surface; some of the broader divides are flattish or of table-land character. There are few places, nevertheless, without some slope. It is probably safe to say that in the neighborhood of 85 to 90 per cent of the uplands has a slope of 2 per cent or more. The average slope would run, perhaps around 7 or 8 per cent, or a little more. There are countless slopes having gradients in excess of 15 per cent (and, of course, many that are steeper in the mountainous segment). Many of these steeper slopes are in cultivation or have been sometime in the past. It might as well be stated now that, as I see it, Piedmont slopes steeper than about 12 or 13 per cent generally should not be used for plow crops. This is based on observation of countless examples of badly eroded steep areas, and also on measured losses of soil at the erosion stations. Certain gravelly and stony areas can be cultivated with a fair degree of safety on somewhat steeper slopes, because of their greater stability. Of course, the thick-growing, soil-holding crops, as grass and lespedeza, can be gown on most any kind of slope; but these do not belong in the category of plow crops.

As a matter of emphasis, it might be said again that all sloping areas are subject to erosion when used for the clean-tilled crops. I feel that it is safe to estimate the proportion of south Carolina Piedmont upland subject to erosion, when used for clean-tilled crops for any considerable length of time, as represented by 85 to 90 per cent of the area involved. On the smoother lands the losses take place comparatively slowly, as a matter of course; but they take place, nevertheless, and eventually impair the soil. Erosion begins quite actively within 1 to 3 or 4 years after clearing off the virgin hardwoods on the steeper lands; it begins at once, ordinarily, on reclearing those more sloping areas that formerly were cultivated, and then abandoned and allowed to restock with second-growth pine. It is probably safe to estimate that at least 70 per cent of all the cultivated upland of the Piedmont portion of the state, together with that formerly cultivated and now grown over with pine, has, in the course of time, lost from 4 to 18 inches of soil and subsoil, with countless places where the washing has cut away all the soil and subsoil, on down to soft decomposed rock, or even down to hard bedrock. This loss means more when we come to understand that the average depth of topsoil in the region is not more than about 8 or 9 inches, and that the better part of this surface layer, the part so richly charged with humus, is seldom deeper than about 4 inches.

Piedmont Soils.

A word about the soils of the region: As to types, the soils of the Piedmont of South Carolina are not very numerous; but the variation among these types is decidedly intensive in many places, especially on those slopes where erosion has cut through the top layer of small and large areas, down into the upper subsoil, and, after that, on down into the lower subsoil, finally to the depth of soft, decayed rock, from which these soils have been formed by nature's lengthy process of building. Thus, we find numerous hillsides spotted gray, red and yellow, according to the progress of erosion, and also according to the complexity of the virgin soil.

The principal soils are those of the brittle red clay subsoil group, designated as the Cecil series. These have been derived from granite and related rocks, first through decay of the basal rock and then the building of this material into soil by ages of freezing, thawing, the action of percolating rainwater and the effects of vegetation, microorganisms, earthworms and burrowing insects and animals. The most extensive types are the Cecil sandy clay loam, Cecil clay loam and Cecil sandy loam. Other important groups are the Appling, Wilkes, Georgeville, Alamance and Iredell. These are all fully described in the various county soil survey reports, so that time need not be taken to discuss them here. These reports not only describe the soils but show their extent and location and give their crop adaptations and better methods of use. The names and areas of the different types found in Spartanburg County may be cited as an example:

Name Area in Acres
Cecil sandy clay loam 209,152
Cecil sandy loam 168,832
Cecil clay loam 66,688
Cecil gravelly sandy clay loam 11,392
Cecil gravelly sandy loam 5,696
Cecil coarse sandy loam 2,688
Cecil fine sandy loam 2,304
Meadow 23,488
Appling sandy loam 9,216
Iredell fine sandy loam 6,784
Iredell loam 896
Louisa sandy clay loam 5,184
Louisa clay loam 4,800
Congaree silt loam 3,008
Congaree fine sandy loam 768
Davidson clay loam 1,728
Durham sandy loam 1,280
Porters loam 250
Total 524,162
Examples of Erosion In South Carolina.

As examples of the severest phase of erosion, that is, gullying plus deep sheet erosion, we will probably find the greatest areas in a strip crossing the state through Lancaster, Fairfield and Union counties. As long ago as 1912 a soil survey of Fairfield County showed on the soil map 90,560 acres of land, practically all of which had been tilled at one time or another since the settlement of the country, under the classification Rough gullied land. This was once good land, but the greater part of it had been deeply dissected with gullies, and between the gullies most of the soil, and in countless places even the subsoil, had been washed off at that time. There were scattered areas of still cultivable land between the gullies and along some of the crests of the broader ridges; but such areas were isolated, so that the soil surveyors considered that they properly belonged under the general classification mentioned above. By far the greater part of this area really represented land which had been destroyed, at least in the sense of having further value for crop production by the ordinary farm methods. Indeed, much of the land was so badly cut to pieces that major engineering operations would have been necessary to restore it to a reasonable state of arability.

The 90,000 acres referred to, however, did not fully cover the extent of erosion in the county. There were numerous other eroded areas included with the better soil types, as shown on the map. Beyond that, 46,650 acres of stream-bottom land were mapped as meadow, which represents land of a swampy or semi-swampy nature, having for the greater part little value except for grazing and the growing of trees. These bottom lands, nevertheless, once were considered the richest of the county.

The work of this survey came under my supervision as a representative of the Department of Agriculture, by which it was made. After a lapse of twenty years I went back into this county, this year, to have another look at the situation-to see what had taken place since the survey was made. There were three roads which formerly I had driven over quite comfortably with horse and buggy. Again I tried these roads in an automobile, last September, but was forced to turn back in every instance. The roads had grown into canyons; trees were growing up and down them in many places. (It is a fortunate thing that nature provided the pine tree for land of this kind, since man insists on creating such conditions. At any rate, the greater part of it is being reforested by natural processes.)

Amidst a maze of ravines I followed another winding road and finally reached an old farmstead. Here a magnificent old mansion was tumbling to ruin. From about its foundations 3 feet of soil have washed away. Of 1,004 acres comprised in this erstwhile farm, not one field remains in cultivation, although practically the entire area once was farmed. I went around over this farm and found not so much as a single acre in one place which seemed to me as worth plowing. The soil washed away long ago, and after that most of the subsoil. Rock was exposed in thousands of places-hard rock and soft rotten rock. Gullies streaked the land in all directions, not only here but through all the surrounding country. I saw the farm burial place on a knoll about one-eighth of a mile from the disintegrating mansion. I walked over to this. It was more of a process of sliding down and crawling out of the intervening gullies. It took over 30 minutes to get there. Samples taken from various places, including one diminutive patch of the original hardwoods, show that the soil has suffered terribly. The humus of the skeleton soils now to be found is too deficient for any chance of good yields. Recovery of such parts of the land as still might be plowed would entail expensive terracing, growing of humus-supplying crops and addition of fertilizers. The land is now best suited to timber, and pine trees are taking it in charge.

After going over many parts of the county and checking the soil map made twenty years ago, it was estimated that the area of essentially destroyed land has increased something like 20 per cent, or at the rate of nearly 1,800 acres a year.

Now let me add here that there still remains much good land in Fairfield County, smooth land, good for crops. I am referring merely to the worst part of the land, and in fairness let me say that there are other counties in other states, both north and south, which contain even a larger proportion of worn-out land. I have in mind a county where a soil survey has just been completed which shows 190,000 acres of formerly cropped land that has been ruined by erosion, and still another with 200,000 acres. As I see it, nothing is to be gained by covering up those facts. We must bring them out in the open as a stimulus to greater efforts at soil conservation.

Extensive Sheet Erosion.

In another county, Spartanburg County, a soil survey has classed 297,216 acres as clay loam, sandy clay loam and gravelly sand clay loam. Examination of the soil in remnants of virgin stands of mixed hardwoods and pine shows the original soil to consist of some 4 to 8 inches of brownish or yellowish, mellow sandy loam and loam. This top layer is gone or largely gone from 297,000 acres. These clay loams are in a sense new soils, representing products of the excessive erosion which has taken place over the cultivated slopes. They are not worthless soils, by any means, for the fortunate reason that the upper subsoil of the great majority of the Piedmont country is still fairly good land, where duly fertilized and properly terraced and rotated so as to prevent washing from extending down into the deeper, less productive subsoil. There are many good terraces in Spartanburg County, and with good farming these lands will last yet a long time. Nevertheless, soil-denuded fields are not so productive as were the fields with the original soil, and they are not so easily tilled. They bake more in dry weather, and corn suffers very much more than it would on the original loam and sandy loam topsoil. This fact is attested by the generally poor crop of this year throughout the Piedmont from Washington, D.C., to east-central Alabama. With few exceptions, land eroded down to clay has produced in this region, this year (a dry year following a wet spring), exceedingly poor corn. On smooth, uneroded land, however, there is in the same region a fair crop of corn. Under certain conditions, cotton also suffers more on these erosion-produced lands.

It should be observed that Spartanburg County has, also, some measure of land where the process of despoliation by erosion has advanced to the point of danger. There are, for example, 209,152 acres of Cecil sandy clay loam in the county. Most of the original forest was removed from this long ago. Now about 60 per cent of the land is under cultivation, the remainder being largely covered with old field pine. Now when we critically examine these old-field-pine areas we find many of them contain gullied or deeply eroded soil where ordinary farm crops are not likely to give a very satisfactory living to those who undertake its cultivation.

These two conditions of sheet erosion and gullying are to be found throughout this great Piedmont country. They are present in every county, indeed in every township, all across the belt from Alabama to within sight of New York City. Much of the abandoned land in the Carolinas and southward was terraced long ago. While the terraces did much good, undoubtedly, I am convinced that many of them were not built in just the right way. At any rate, erosion has gone ahead too aggressively, and probably not less than 500,000 acres of land in upper South Carolina, alone, have reached a stage of worthlessness, or an approximation of that, insofar as safe use for farm crops is concerned. Its best use now is for the growing of trees and grazing.

The appearance of the Piedmont country has changed vastly since the coming of white man with his axes and plows. This change was described in a very interesting manner 30 years ago by Mr. F.W. Taylor and Thomas D. Rice, in the soil survey of the Abbeville Area, South Carolina. They had the following to say about this country around Abbeville, which is representative of a considerable part of the Piedmont section of the state:

"The original forest growth was quite different from the forests of the present time. On the highlands the oak, hickory and chestnut were of large growth and stood far apart. There was no underbrush and the woods were carpeted with grass and the wild pea vine. Along the streams and in the valleys the distinctive growth was willow, beech, birch, black walnut, ash, poplar and gum. The cane also flourished best here, although it often grew upon the higher ground. The cane growth was the standard by which the early settlers estimated the value of the land. If it grew only to the height of a man's head, the land was esteemed ordinary, while a growth of from 20 to 30 feet indicated the highest fertility. Not only the forests, but the cultivated fields as well, present a very different aspect now from what they did after the country was first opened up. It was then new and beautiful and as remarkable for the luxuriant richness of its landscape as it is now for the striking features of its rolling hills and long, narrow valleys. The original forest has disappeared almost entirely, and has been replaced by scrubby oaks, by underbrush, and by short leaf pines of the abandoned fields. The chestnut and chestnut oaks have been dying out for the past sixty years, and the cane likewise almost disappeared."

Deposition of Erosion Products.

We have discussed erosion in the uplands. What has happened to the bottom lands? Already it has been pointed out that the products of soil erosion are distributed from the point of origin all along the journey to tidewater. More than half of the area of the stream bottom land in the Piedmont plateau, not only in South Carolina but for the entire area of 50 million acres, has been buried with sand and mud to depths ranging from a few inches to six feet or more, since the agricultural occupation of the region. Much of this has been classed as meadow in the soil surveys made throughout this great region: a semi-swampy type of land, usually covered with willow, alder, sweet gum, smilax, blackberry and rushes. Of 269,440 acres of alluvial mapped in the stream bottoms of the South Carolina Piedmont, 210,752 acres, or 72 per cent of the total area, has been classed as meadow. W.E. McLendon (a graduate of Clemson College) in the report published in connection with the soil survey of Anderson County, South Carolina, has the following to say of this type of land:

"Some areas were originally a good loam, but have since been covered with loose sand, rendering them almost worthless. Others are quite sandy beneath and loamy on top. The creek bottoms vary more widely still. The original soil was for the most part a brownish or black sandy loam to loam, but the floods of recent years have covered the older soil in many places with loose sand. Now in these areas there is a mixture of loamy and sandy spots, even in very small areas. Generally the narrower bottoms are the sandiest and most subject to changes with every overflow. In the earlier settlement of the section the bottom lands were highly prized for the production of corn and forage crops. Then overflows were not very frequent and rarely destroyed the crops; but as more and more of the uplands were cleared floods became correspondingly more frequent and disastrous, until now the cultivation of the bottom lands is considered very risky. Large quantities of sand have been washed in from adjacent slopes, badly obstructing the streams and rendering the soil of little value."


Part II


The National Program of Soil and Water Conservation

It should be clear from what has been said that we have not made very much progress in this country with respect to the problem of erosion control. Rather, the farmers over the greater part of the country have been pretty soundly asleep about the seriousness of this process of land impairment and absolute land destruction. Generally, too, they not only have done little or nothing in opposition to continuing soil impoverishment by this process, which of all the combinations of soil-impoverishing agencies having to do with the human factor is by far the most serious; but many of them have not even recognized the most menacing aspect of the problem: the effects of sheet erosion. Moreover, there has been actual encouragement of erosion through such unwise practices as the running of rows up-and-down the slopes and the continuous growing of clean-tilled, humus-exhausting crops. Not only this, but until recently nothing was done in the way of persistent experimental effort to determine the most effective methods of control, not to mention the working out of the basic principles of erosion processes or the measuring of the losses. In practically every other field of agricultural endeavor an enormous amount of research work has been done. We have experiment stations in every state, and numerous substations on major types of farm land, and still more experiments in outlying fields. A very considerable part of this work has related to increasing crop yields and the maintenance of soil-fertility through the use of fertilizers, soil-improving crops, manures, crop rotations, better tillage methods and so on. All of this work, indispensable as it is, has no intentional specific bearing upon prevention of soil decline by erosion. This most impoverishing of all agencies affecting the soil, an agency that rapidly steals the whole body of the soil, plant food and all, has been left out of the efforts of mankind in the interest of better land use almost as if it represented something untouchable.

Recognizing the acute need for better knowledge of erosion processes and better methods for erosion control, a national program of soil and water conservation was inaugurated about three years ago, following a survey of the erosion situation in the country, which was carried out by myself. Now we have eleven erosion experiment stations in as many major soil and climatic regions, where through the cooperation of the states and the Federal Department of Agriculture this problem is being vigorously attacked. The nearest one of these to South Carolina is the Central Piedmont Station, 10 miles west of Statesville, N.C., on Highway No. 10.

At these stations it is proposed to try out at the earliest possible moment every promising practical method of slowing down erosion, in addition to measuring the losses of soil, soil fertility and rainwater from different slopes undergoing various cropping and tillage practices. The matter of terracing, as previously stated, will be discussed by members of the technical staff of the Bureau of Agricultural Engineering. I am going to attempt to point out some of the most promising means for control through the instrumentality of soil-saving crops and special soil treatment, based on present findings and indications at these youthful experiment stations.

Strip Cropping.

Strip cropping represents a rather simple method of planting along the natural contours of the land thick-growing, soil-saving crops like sorghum, small grain, clover, lespedeza, velvet beans, grass and sweet potatoes, in strips between clean-tilled crops, such as corn, cotton and Irish potatoes. The purpose of the thick-growing crops, occupying positions corresponding to those of locating terraces, is to slow down the rate of water flowing across the field, thus to cause a part or all of the water-borne soil to be deposited. Now this is almost precisely what happens with terracing. The terrace catches the flowing water, slows down its rate of flow and causes the deposition of part of its suspended load of rich soil material, slowly passing the excess water off to the sides of the field. The strip crops retard the rate of running water and spread the water out over the protected parts of the field so as to save the soil and to cause more of the water to be absorbed by the ground.

Our experiments with this system of cropping have proved highly encouraging since the very beginning. In the great cotton-producing Black Belt of central Texas there has been no appreciable erosion from strip-cropped areas for a period of 20 months, and there was practically none this year at the Red Plains Station in central Oklahoma. The experiments with this system on the Cecil soil at the Central Piedmont Station in North Carolina have been similarly encouraging this year, the first year of installation. Strip cropping is so practical, so cheap and so easy to apply that we are convinced it is going to be extensively used in many parts of the country. Its most promising use is on the gentler slopes. We feel, according to results thus far obtained, that where the slope is greater than about 5 to 6 or 7 per cent, depending on the soil, the character of rainfall and the crops grown, the system may need the support of terracing. On a farm of 6,000 acres in the Texas Black Belt, where every field is now under strip cropping, it has been found that where terraces are needed it is not a very difficult matter to induce tenant farmers to build them. The very fact that there is need for this type of cropping appears to affect the psychology of a farmer to the extent of causing him to begin to think more about the benefits to be gained from giving the land some sort of protection against erosional wastage. This stimulus, further aided by the fact that the strips are laid out just as terraces are and need no additional surveying for the subsequent location of a terrace, is pushing along soil conservation methods in parts of the Black Belt of Texas faster, I believe, than any other method we have presented to the farmers of that region.

Probably the strips of the thick-growing crops should be planted broadcast in the Piedmont country, rather than in rows. I observed this summer a considerable number of fields in the South Carolina and Georgia Piedmont which are being treated in this manner. Whether the strips were consciously planted for the purpose of slowing down erosion or not was not determined, but they were there and there is no question but that they will improve the situation in the fields thus handled. The thick-growing crops that I saw were cowpeas and sorghum mixed, and sweet potatoes. If, after the sorghum mixture is cut for hay, the stubble is left during the winter and early spring, this stubble will probably control the situation quite effectively until next planting time. The strips can be made as wide as one chooses. The wider they are the more nearly complete will be the control. We have found that where the contour of the land is such that the width of the clean-tilled crops between the thick-growing crops varies, the matter of having short rows can generally be avoided, if deemed desirable, by varying the width of the thick-growing crops rather than that of the clean-tilled crops. (Strip cropping is discussed in Leaflet No. 85, U.S. Department of Agriculture, copies of which can be had on application.)

Strip Subsoiling.

We are testing out the efficacy of strip subsoiling, but thus far have not obtained very promising results. This system, nevertheless, is used in certain parts of the bean-growing districts of California with good results. It simply consists of subsoiling the land to 18 or 20 inches along the contours, over strips varying in width from one to several rows, and leaving below this strip an unsubsoiled strip of equal or greater width. This develops a sort of subterranean terrace, but we are not prepared to recommend its use as yet.

Cover Crops.

No argument is necessary to convince any thinking person that it is much better to keep the land covered with some form of vegetation throughout the year than to leave it bare, especially in the more southerly regions where the ground is seldom protected with a blanket of snow. Fields plowed in the fall and left bare for next year's cotton or corn planting are certainly going to wash to some extent during the winter, and they are even more likely to wash with spring rains, if such rains come before planting time, as they so frequently do. Erosion being most destructive during the summer season, the same thing applies with even greater emphasis in relation to summer treatment of the land. We have learned that any kind of vegetation on the land is better than no kind. Our measured losses at the erosion stations show that soil moves off bare slopes much faster than under any other condition. The rate of soil loss at Statesville, North Carolina, from bare ground was, as already stated, 65 tons per acre as against 17 tons where cotton was grown and only 1–1/2 tons where lespedeza was grown. A winter cover of grain, whether of oats, barley or wheat, is mighty good insurance against depreciation of the land by erosion; moreover, a good crop of wheat or oats is a mighty useful thing to have about the average Piedmont farm. We are testing out every legume that we can get hold of, both the winter and summer types, for the purpose of working them more and more into our cropping systems. Thus far lespedeza looks to be one of the best legumes for this region, but cowpeas are still a mighty good crop and so are velvet bean. Soybeans have in some parts of the country a reputation for being not especially effective as an erosion-control crop, but our studies indicate that this is more likely to be true where the crop is grown in rows or is drilled up-and-down the slopes, whereas it is not particularly true where the crop is thickly sown across the slopes.

We have had much to say about soil losses from erosion and nothing to say about soil losses from leaching. The best information available is that rainwater soaking down through the ground carries away by this process of leaching very little of the phosphorous content of the soil and not so much of the potash but the process does remove considerable lime and nitrate nitrogen. Here again cover crops come into the picture. The roots going down into the soil undoubtedly gather in and hold a considerable part of those soluble materials which otherwise would leach down into the substrata. Part of this retained plant food will pass up into the leaves of the plant to be incorporated with the soil wherever the cover crop is plowed under. The plowing under of cover crops, or of any other kind of humus-supplying material, for that matter, is decidedly helpful in slowing down erosion. Every test we have made shows this. The saving is remarkably high in many instances. Take the matter of manuring: Way up in the Corn Belt of southwestern Iowa we found that two rains in August this year washed 11-1/2 tons of soil per acre from unmanured land as against only 4-1/2 tons per acre from land immediately alongside, to which 8 tons of manure had been added. These decaying vegetable materials act within the soil after the manner of a sponge, making them more absorptive and keeping the pore spaces open so that rainwater can better penetrate to desirable depths.

Crop Rotations.

All of our experiments have confirmed the soundness of old ideas on the value of crop rotations. This is one of the methods that nature uses in the prevention of the levelling down of the earth's surface to the condition of marsh land. If lightning strikes a timbered area in the dry season and sets fire to the woods, causing the timber to be burned off, the thing that happens in most parts of the country with the very next growing season is the appearance of a robust crop of weeds, grasses and shrubs. In the Piedmont, out of this preliminary growth which affords a partial protection to the land pine trees spring up and soon establish a good stand of second growth forest. This gives still more protection. Gradually, from this stage of nature's rotation, we see another move ahead, the appearance of hardwoods. Eventually a climax vegetation is reestablished, this consisting of a growth like the original mixed hardwoods and pine forest found in remaining patches here and there throughout the Piedmont region The practice of crop rotations is commendable from every angle of consideration. Nature endorses it, and nature is a pretty good planner of affairs. Even if the rotation serves no other purpose than to increase the organic supply of the soil and to protect the land from excessive loss of moisture, it serves well.

Strip cropping, I believe, is going to encourage the practice of crop rotations. Some farmers will begin with this method by putting in narrow strips of the thick, soil-saving crops, with broad strips left for cotton or corn. This may lead on, I believe it will lead on, to the seeding of broader and broader strips of the thick-growing crops, eventually to give an equal area to the two types of crops. By shifting the position of these occasionally or from year to year, we will have a true rotation, and I am pretty sure we will have, also, satisfactory control of erosion, or a situation very close to that, for the gently sloping lands at least.

There is no need to discuss crop rotations further. We all know what the practice means. We all know that legumes improve the fertility of the soil, restoring in some measures at least one of the important elements of fertility, nitrogen, which is stolen from the land by erosion, by the crops removed and by the never-ending process of leaching.

Soil Manipulation.

At our Kansas Erosion Experiment Station we have developed a hole digging cultivator, with shovel plows attached in such a manner that they scoop out thousands of holes in the field, piling the dirt along the side of the holes. Each hole then becomes a reservoir for holding rainwater on the land until it sinks into the ground. This machine was developed for use on grain land, especially on fallow land (where summer plowing is practiced in order to conserve more of the rainfall by keeping the weeds killed off). Operating like any ordinary cultivator, this implement digs 10,000 holes per acre. Under ordinary moisture conditions 3 gallons of water are absorbed by each of the holes before they begin to fill up, and then 2 more gallons are required to fill them. Thus, with our first machines we are saving 50,000 gallons of rainwater per acre. At the same time erosion is checked, not only to the extent of this ground disposal of 50,000 gallons of rain per acre, but with slowing down the runoff from the heavier rains by causing the flowing water to zigzag from one hold to another. Out in Kansas where the land was scarified with this machine the loss of rainwater from fallow land, at the time of the last tabulation of results obtained, had amounted to only 1-1/2 per cent of the total precipitation as against a loss of 34 per cent from untreated land immediately adjacent which had been handled according to prevailing practice in the region. This winter the man who made the machine is going to undertake to develop it for use with any kind of crop. It has already worked satisfactorily in wheat fields and I believe it is going to be a serviceable implement in the cultivation of corn and cotton. At any rate, we are not going to permit it to stand as merely a saver of soil and water in the less humid parts of the country, if our best efforts at further development prove of avail.

Gully Control.

In connection with gully control let me say at once that the time to stop a gully is during its infancy, before it has grown into a ravine or canyon. Quite simple methods will go a long ways in the direction of control if this course is followed. We are having almost 100 per cent success with the new type of gully control which we effect with grass dams. With this simple, inexpensive method we have almost completely effaced numerous gullies from large fields and are now getting even better crops where the gullies formerly gnawed at the heart of fields than from other parts of the fields.

The method is as follows: Fill old fertilizer sacks or any other kind of sacks, no matter how rotten they may be so long as they half-way hold together, with grass roots and soil or with soil and grass seed (probably Bermuda grass would be the most effective for this region) and place these across the washes. In the smaller gullies there will be many places where one of these sacks will fill the bill, but usually at least three sacks are needed, and in some instances more than this. A cardinal principle to be observed in all instances is that the sack in the center of the gully shall be so placed as to lie lower than those placed along the slopes of the wash. If this is not done, the water will not concentrate toward the center; it will certainly wash around the ends. I might say that this precaution applies to every other type of dam except those that have overpasses for disposing of the impounded water when it reaches the top of the dam. No matter whether you build a brush dam or a stone dam or an earth dam, or a grass dam, be sure above all else to have the center lower than the sides, after the manner of a V.

The grass grows right through the rotting bags, takes hold of the ground and quickly establishes a stabilized soil condition in that part of the wash. The plan calls for a succession of these dams up-and-down the wash, as is clearly shown in our Leaflet No. 82 (Controlling Small Gullies by Bluegrass Sod).

We are also having splendid results with gully control, where the gullies are larger than those which can be effectively stopped with grass dams, small brush, stone or wire dams, by planting willows in the moist bottoms of the ravine, straight across, and then black locust up-and-down the sides. Plant the willows across the bottom in strips about 3 to 10 feet wide, in the spring at about budding time. There will be little trouble in getting a start if there is any moisture in the bottom of the gully. If not, some sort of small temporary rock dam or brush dam should be installed to collect a moderate amount of soil. The soil will contain moisture and in it willows can be grown effectively. Of course, untimely freshets may, and probably will, wash out some of these plantings, during their infancy especially; but when we come to fighting erosion we must have perseverance if we are to get results. We must maintain terraces if we get results from them; we must repair our houses, bridges and highways if we get best results from them. Indeed, we must keep eternally on guard and on the move if we are to make the most of our lives, including the preservation of our sloping fields.

We are finding on certain soils that black locust grew very rapidly, so rapidly indeed that we are counting on their producing tow fence posts per tree in a period of 10 years, in northern Missouri. Our experiments at the Statesville Station indicate that we can get good results with black locusts in this country, although the trees may not grow quite so rapidly as in some of the other parts of the country. We do not know about this as yet. They are one of the few plants we have been able to get started on deeply eroded Cecil soil, that is soil where, over considerable areas, not a living sprig of vegetation was growing.

We are even plowing down the sides of gullies and planting these to locusts, and with the aid of willow trees planted in the bottom of the ravine we are rapidly building up dams, and this at almost no cost, beyond the time spent on the job. Locust seed can be procured from the forestry departments, and from these an abundant supply of seedlings can be quickly obtained for transplanting.

In connection with gully control, I see little hope of taking charge of such ravines as I ran across in southwestern Georgia this past September. Down there I came to a place where 37,000 acres of formerly tilled land, representing the best type of land in southern Georgia, had been surveyed out as land destroyed by gullying. I think nowhere in the world could a larger man-induced gully be found than one of these. It was 150 feet deep. I talked with a man who went to school in a schoolhouse that once stood in the center of this canyon. If it were there now it would be suspended in the air. The schoolhouse toppled into this yawning gully long ago, and with it has gone the barn from which the water ran that started the gully. A tenant house and a graveyard with 50 graves have tumbled into the chasm also. These buildings and the remains of mortal man have joined the soil debris of 37,000 acres of destroyed farm land on its journey to the caverns of the Gulf of Mexico.

The only possible practical way, probably, to put such gullies under control would be to build diversion dams or terraces about their heads so as to pass the water off in other directions, keep it out of the gully. This has been done in the very region referred to. These are vertical-walled gullies, the kind that grow by the washing out of the looser material of the substrata, to be followed by tumbling in of the soil from above. Most of the gullies in the Piedmont country are more or less V-shaped, and these are much easier controlled than the straight-wall, undercutting type of gully, such as we do find in parts of the Piedmont-in those parts where the substrata consist of soft rotten rock. Such gullies call for special treatment, and I suspect this treatment will call for diversion terraces, especially where the ravines have grown to any considerable size.

Honeysuckle is another excellent crop for gully control. This will not grow satisfactorily on raw clay which hardens and loses its moisture in summer, but it can be started in the same way suggested for willow dams.

Use of Steep Land.

It has already been suggested that we should not use our steeper slopes for clean-tilled crops. If we can not get them into grass, the chances are that nature will take charge of the situation and plant them in pine trees, here in the Piedmont, especially if we keep the fires out while they are young in order to give them a chance. Now these trees are not going to be harvested next year. But they may be harvested as a profitable crop if we live long enough, or our children may take care of the harvesting. In the meantime, they cover up ugly spots on a farm, make an eroded farmstead more habitable and contribute to the general good of the region by keeping soil out of the streams and reservoirs. I saw last summer while in this state a large reservoir, built for the operation of a large cotton mill, which had completely filled up with sand and mud, swept down from the uplands. This process is going on all through the Piedmont region. In all probability the only way that the process of silting of reservoirs can be stopped or materially slowed down will be to quite cultivating the steeper slopes and to get them into trees or grass. In addition to that it is going to be vitally necessary to give better protection to the slopes that we do cultivate. We have an abundance of land still left in this country. There is no necessity for cultivating so much steep areas. The teachers of the country, if they will, can do much toward the encouragement of better land utilization by driving home to the farmers of the regions they serve the thought that in fighting erosion we are opposing the most powerful agency that nature employs in carving out the valleys of the world. Erosion is more powerful in modifying the surface of the earth than earthquakes, volcanoes, tidal waves and all the excavations of mankind since the beginning of history The job of controlling this prodigious implement is going to be difficult enough on our smoother lands. The problem is simply hopeless on the steeper lands except through vegetative means. We must resort to trees, to grass, to all forms of vegetation in this phase of the fight.

Other Lines of Investigation.

It is not necessary to go into further detail with respect to the investigational and demonstrational side of erosion control. We are doing many other things. We are measuring the effects of waste farm and woodland materials in reducing soil losses, such as rotted wheat straw, leaves and pine needles; we are measuring the possibilities of renewing the productivity of land made poor by erosion and the time and cost involved, even though the main point of the national program is to conserve the soil yet remaining rather than reclaim land already ruined.

The educational side of the program we consider to be a matter of major importance. We must depend on people like you to a very large degree for help in this direction. You will see the absolute necessity for this and your obligation to assist in bringing the users of land out of their apathetic point of view toward this pathetic problem. You do not need to be told, I am sure, that the soil is a very sacred thing, a resource that was not given us to waste. You will give of your knowledge, and freely of your advice, to those whom the preservation of this most indispensable asset, this asset that belongs as much to posterity as to ourselves, has been entrusted.


  1. Address delivered in connection with the South Carolina Teacher-training program by H.H. Bennett, Chief, In Charge Soil Erosion Investigations, Bureau of Chemistry and Soils, U.S. Department of Agriculture, at Spartanburg and Clemson College, November 4, 1932; Columbia and Rock Hill, November 5, 1932.
  2. W.W Weir: Soil Erosion in California: Its Prevention and Control. Bul. 538, University of California, College of Agriculture, 1932.
  3. Soil Erosion a National Menace. Circular 33, U.S. Dept. of Agriculture, (1928), page 5, Bennett, H.H. and Chapline, W.R.
  4. The statistical data relating to corn, wheat and cotton are from The Yearbook, U.S. Dept. of Agriculture, for 1900, 1925 and 1930; Crops and Markets, U.S. Dept. of Agriculture, Vol. 7, No. 12; and by memoranda from the Division of Crop and Livestock Estimates, Bureau of Agricultural Economics.


< Other Speeches
< Back to History