Page images

soil is softer, lighter even than newly plowed loam. Forest humus does double duty to mankind. It leads water underground and it feeds the growing trees. It is a natural asset of incalculable value.

A forest has still other means of retarding run-off. Water and snow, falling on the leaves and branches of trees, are either evaporated or delayed in reaching the ground. The water evaporated may be as much as one-third of the total precipitation. More important, however, is the influence of forests in retarding snow melting: A well-stocked forest moderates the extremes of temperature; it is warmer in winter and cooler in summer than open land. As spring comes on, snow melting may begin earlier in the woods, but it usually lasts from four to eight times longer than on open ground. Moreover, as the mellow soil of a forest is likely to remain unfrozen or to freeze less deeply than soil in the open, it absorbs more of the snow water; soil in the open, freezing more freely and deeply during winter, permits snow water to rush off more rapidly into the streams. Spring freshets from melting snow are an important source of river floods. By delaying snow melting, feeding part of the snow water into the ground, and thus prolonging the period of run-off, forests tend to reduce flood crests and to equalize stream flow.


Normal erosion is a geological process, and as long as water flows over the surface of the land it will continue its ceaseless work of carrying the mountains down to the sea. But with her kindly mantle of forest and grass nature protects the land from the gnawing

tooth of running water, makes the soil fertile and life-giving, and | restrains the levelling of the hills to an orderly and imperceptible process.

Abnormal erosion as a man-made process is very different. When the soil-binding plant cover is broken, the earth is exposed to the insatiable cutting power of water. In this day of the steam shovel, one can easily observe abnormal erosion at work in fresh cuts, fills, and embankments. Until vegetation reclothes the soil, every violent rainstorm washes quantities of mud into the streets, sewers, and drainages; but from grass lawns near by only clear water flows.

Erosion can often be observed on abandoned farm land. If the ground is level, the insidious effects of "sheet" erosion are not very obvious; but if the ground slopes even a little the erosive action of water leaves a striking mark. Most of the water, instead of sinking in or trickling over the surface, rushes swiftly down the old furrows or makes little gullies of its own. Gathering speed and force, it rapidly cuts deeper and carries an ever-increasing load of soil.

În denuded or depleted forests + a similar but less obvious process goes on. It is less obvious because most denuded forests reestablish on the land a vegetative cover of some sort. But such areas, unless protected from fire and promptly restocked with young trees, lose much of their value as holders of water and soil. In the fire-swept denuded forest, the mineral soil, already scarred and broken by log

[ocr errors]

9 Normal erosion is described here in its most common and usual form. There are, of course, numerous instances of violent and destructive erosion, as in the case of cloudbursts or landslides, which are due wholly or primarily to natural causes.

* In this discussion, the terms “ forest denudation and “deforestation" are applied to methods of cutting timber, often accompanied by fire, which largely destroy a forest cover and prevent its replacement for a long period of time.

ging, becomes more and more exposed to the erosive force of water. (Fig. 3.) Fire consumes the logging débris and the protecting cover of leaf litter and humus, weakens and thins out the soil-binding grass and other vegetation, and by impoverishing the soil and exposing it to packing and drying reduces the chance of reestablishing a denser cover.

Ravines are apt to form in the old logging roads. Gullies eat back into the bare slopes. The exposed surface is washed by every rain. A balance of nature that had existed for ages is overthrown and the soil of the hills takes its way to the sea. If fire is kept out, however, so that shrubs, briars, weeds, or young trees spring up in abundance, the natural protection from erosion afforded by a cover of forest may gradually be restored.

It is not necessary to remove all of a forest to accelerate erosion. Forests that have been repeatedly burned and badly thinned, but that are clothed with green foliage, may have the appearance of being uninjured. Even a fire that kills no mature trees almost uni


F-175334 FIG. 3.- When the forest is destroyed by logging or fire, the soil, no longer protected

by the mat of decaying leaves and wood, is swiftly gullied by rains. Water rushes off the slopes in torrents, filling the streams and raising their flood crests with a heavy load of soil

formly destroys the underbrush, grass, and leaf mulch, and bares the mineral soil. Farm woodlands are frequently not only burned but overpastured. Such disturbances of the plant cover invite erosion in proportion to their intensity.

Å type of erosion that is hastened by deforestation is the caving of stream banks where trees have been cut away. The tangled roots of willows, alders, birches, cotton woods, and other trees and shrubs growing along streams bind the soil of the banks and weaken the cutting power of the current. As a result, well-wooded streams usually flow in clean-cut, narrow channels, whereas deforested streams are apt to eat into the banks, wash the soil into the channel, and form a wide, shallow, meandering bed. (Fig. 4.)

Erosion as a result of the destruction of vegetation cover is not confined to forests. In the grasslands of many portions of the Great Plains thinning and deterioration of the vegetation have been going on for decades. The cause is overgrazing, but the effects on soil and stream flow are essentially the same as the effects of deforestation, though less pronounced.


The striking increase of erosion on watersheds that have been deforested has been observed frequently both in America and in Europe. It can best be studied in small valleys, for there the contributing factors are more easily observed and weighed than in large drainages.

In 1913 fire destroyed the brush cover on a small branch of the Los Angeles River in the mountains of southern California. Careful study led to the conclusion that in the year following the fire 100.000 cubic yards of soil had been washed from an area of 1.2 square miles.

The soil that was washed from a burned hillside of 200 acres in San Diego County, Calif., buried a 12-acre alfalfa field so deeply that no signs of cultivation remained.


(Photo by Thorndyke Saville, North Carolina, Geological and Economic Survey) Fig. 4.—The evil of soil erosion extends down to the main streams, where it clogs the channels, cuts down the carrying capacity, and iurpairs flood-control works. This reservoir, formed by a 38-foot dam built as recently as 1910, has been made useless by silt. Top soil to the amount of 126,000.000.000 pounds is swept into our rivers each year from poorly handled farms, grazing lands, and forests

[ocr errors]

On a tributary of the Sacramento River, where vegetation was killed by smelter fumes, as much as 400 cubic yards of soil per acre has since been removed by erosion.

A flood in a deforested canyon in California carried from twenty to forty times as much sediment as the coresponding freshet in an adjoining forested canyon, and continued muddy after the other streams had cleared.

The effects of denudation of grasslands at high altitudes in the Wasatch Mountains of Utah have been studied by the Forest Service. On a small, badly overgrazed watershed, rainfall, run-off, and erosion were accurately measured for a number of years. Grazing was then excluded and the vegetation allowed to increase to about two and one-half times its previous density. This increase in density decreased erosion by 56 per cent. One storm in the first period washed down 2,354 cubic feet of sediment; a storm of equal intensity and greater duration in the later period washed down only 815 cubic feet of sediment.

67677°-28 -2

Similar examples of erosion following denudation could be cited from other parts of the United States, particularly the southern Appalachian Mountains, and from many other parts of the world. Such erosion has been especially severe in parts of France, Italy, Spain, Greece, Asia Minor, and China.


The swift run-off that follows forest destruction raises the flood level with silt, gravel, and bowlders. It also raises the flood level with an excess burden of water that would under normal conditions be absorbed into the ground. The reduction of this surplus burden of silt and water is a prime objective of flood control.

The influence of deforestation in speeding up run-off has been accurately observed and measured in small valleys. One of the most famous of these experiments was carried on for many years at Emmenthal, in Switzerland, where the run-off from two small valleys, one completely forested, the other only partially forested, was measured and compared. It was found that the stream from the completely forested valley was more uniform in its flow and carried less débris than the other stream. After very heavy rainfall of brief duration the forested stream carried only from a half to a third as much water as the partially forested stream. In the summer the forested stream kept up a higher and more regular flow.

The burning of the brush cover in a small canyon in southern California has already been mentioned. After a heavy rain the following winter the run-off from this canyon exceeded that from any other stream which experienced the same storm. Its flood peak reached the unprecedented flow of 700 second-feet per square mile of catchment area.

Several years ago a disastrous flood was caused at Pueblo, Colo., by local cloud-bursts. In canyons that had been largely denuded, the crest of the flood occurred six hours after the beginning of the storm. In the forested canyons the flood crest occurred from 18 to 48 hours after the beginning of the storm and caused much less damage.

That such effects come in part from destroying the absorptive capacity of the forest floor is shown by experiments made by the Forest Service in the mountains of southern California. Within three years after the brush cover on a canyon drainage was destroyed by fire the soil had lost 45 per cent of its water-absorbing power. This loss was due partly to the burning of the litter and humus and partly to the rapid erosion of the topsoil.

The Missouri Agricultural Experiment Station has shown that over a period of six years nearly half the rain falling on a plot of bare uncultivated soil ran directly off into the streams, whereas only a little more than one-tenth of the rainfall ran directly off similar soil covered with bluegrass sod.

In the Forest Service experiments in the Wasatch Mountains an increase of two and one-half times in the density of the grass cover decreased the surface run-off 55 per cent, even though at its densest the grass was less than half of a full cover. From one storm during the early period of scanty cover the surface run-off from the small watershed under observation amounted to more than 15,000 cubic feet; whereas in the later period of denser cover, from a storm of greater duration but equal intensity, the surface run-off was less than 6,000 cubic feet, or 39 per cent as much.

The influence of deforestation in drying up springs is widely established by authoritative examples both in America and in Europe. Springs form a striking and common evidence of the presence of underground water, which is the principal contribution of forests in equalizing and prolonging the flow of streams. The disappearance of springs after deforestation is ordinarily proof that rainfall has run more rapidly from the denuded surface instead of sinking into the ground.


[ocr errors]

The examples cited indicate the effects of deforestation on small watersheds. It is frequently asserted that deforestation has no influence on large rivers because great floods occurred before the forests had been touched by the ax. At the opposite extreme, it is asserted that floods are due primarily to deforestation. It is necessary to use moderation and common sense in approaching this problem. Forests can not prevent the great floods that come from heavy, long continued, widespread rains, especially when combined with the rapid melting of snow. On the other hand, both common sense observation and careful study have shown that forests may reduce the violence of floods and erosion not only in small watersheds but in large river systems.

The difficulty of “proving the case” by exact measurement arises in part from the impossibility of weighing each of the complex factors that influence the behavior of a large river, such as varying conditions of soil, slope, vegetation, cultivation, and climate as well as natural storage in lakes, swamps, and overflow lands. In the absence of exact quantitative proof, we find it difficult to visualize the cumulative influence of thousands of square miles of distant and perhaps scattered forests on the main river channel where the flood rages. When we visualize the forest, we think of an acre or a few acres, and we ask how such a limited surface reservoir can have any mea

rable effect on the giant flood that is bursting the dikes of the river. How much can it hold back from the torrent? What relation can its steadily flowing springs, rivulets, and brooks have to the great flood?

The answer is to be found in the vast extent of the forest reservoir as compared to the relatively narrow channel in which the mightiest river is confined, especially if it has been pinched between levees. It is like an enormous roof drained by a single waterspout. The forests of a great river basin constitute, it is true, only a shallow reservoir but one that may make up for its shallowness by its extent. . If these forests are in good condition, they hold the soil in place; they carry a large proportion of their water underground, pay it out slowly through springs and rivulets, and thus hold back the flood stages of countless brooks, creeks, and little rivers. Now, it is precisely these springs and rivulets, brooks and small rivers that feed the main rivers.

From direct rainfall on their surface and from the run-off of their immediate banks the large rivers receive only an insignificant fraction of their water. The greatest rivers and the most violent floods merely collect and concentrate in one channel the individual contributions of innumerable minor and in


[ocr errors]
« PreviousContinue »