DOCTORAL DISSERTATION SERIES titir(ffANG£S IN SOME \IE6L7A1I6N.SOifA. m m sutFACLtwtiffatmaausMA a m tm p 3 m m A m s y m s i C V r T M M b S U B S B jM iN T M O M TM N fA U W fi author HUBEI! EARL D U UNIVERSITY. MICHIGAN S fjflj coLL DEGREE PL D. S ______ date PUBLICATION NO. i ipipip I UNIVERSITY MICROFILMS A UU Alin A A k i i r U I A A U / &5Z CHANGES IN SOME VEGETATION, SURFACE SOIL AND SURFACE RUNOFF CHARACTERISTICS OF A WATERSHED BROUGHT ABOUT BY FOREST CUTTING AND SUBSEQUENT MOUNTAIN FARMING by Robert E. Dlls A DISSERTATION Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science In partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY DEPARTMENT OF SOIL SCIENCE 1952 ACKNOWLEDGMENTS All of the hydrologic data used herein was taken from the records of the Coweeta Hydrologic Laboratory near Franklin, North Carolina. This laboratory is a branch sta­ tion of the Southeastern Forest Experiment Station, Forest Service, United States Department of Agriculture, with which the writer was associated during the sunnier of 1951. The writer wishes to express his appreciation to Dr. Charles R. Hursh, Chief of the Division of Forest Influ­ ences of the Southeastern Forest Experiment Station, for his assistance in outlining the problem and for his continued interest and constructive criticism. Special acknowledgment is made to Edward A. Johnson, Technician-in-charge of the Coweeta Hydrologic Laboratory, for his guidance and help with many phases of the work, par­ ticularly the hydrologic analyses. Jacob Kovner, of the Coweeta Hydrologic Laboratory contributed valuable help in statistical techniques and in the collection of soil samples. Acknowledgment is made to Dr. Lloyd M. Turk, Head of the Soil Science Department, Michigan State College, and Chairman of the writer’s graduate committee, and to Dr. A. Earl Erickson and Dr. Eugene P. Whiteside of the Michigan State College Soil Science Department for constructive criticism and assis­ tance particularly in the soil analyses. Dr. Whiteside further contributed many valuable suggestions on the organize- ill tlon and content of the report. Appreciation is expressed also to Dr. H. G. Wilm, Chief, Division of Forest Influences; Forest Service, United States Department of Agriculture whose discussions provided the writer's original spark of interest in the field of forest influences and watershed management. Robert Earl Dlls candidate for the degree of Doctor of Philosophy Final Examination: Dissertation: Changes in Some Vegetation, Surface Soil and Surface Runoff Characteristics of a Watershed Brought About by Forest Cutting and Subsequent Mountain Farming. Outline of Studies: Major Subject: Soil Science Minor Subjects: Geology, Forestry Biographical Items: Born - July 6, 1919* Miami aburg, Ohio Undergraduate Studies: Wittenberg College, 1937-1941 Colorado A & M College, 1946 Graduate Studies: Colorado A <5c M College, 1947 Michigan State College, 1948-1952 Experience: Member: Member, United States Army Air Force, 19421946; Research Assistant in Forestry, Colorado Agricultural Experiment Station, 1946-1947; Timber Management Assistant, U.S. Forest Service, 1947; Instructor, Forestry, Michigan State College, 19471949; Assistant Professor of Forestry, Michigan State College, 1949-1952 Beta Beta Beta, XI Sigma Pi, Society of American Foresters, Soil Conservation Society of America, American Geophysical Union, Soil Science Society of America TABLE OF CONTENTS Title ACKNOWLEDGMENTS Page . . . . , ii PURPOSE AND S C O P E .................................. 1 PAST W O R K .......................................... 4 THE COWEETA HYDROLOGIC L A B O R A T O R Y ................. 17 THE LITTLE HURRICANE W A T E R S H E D ..................... 22 Land Use H i s t o r y .................................. 22 Geology and P h y s i o g r a p h y ............ 26 C l i m a t e .......................................... 28 S o i l s ............................................. 31 V e g e t a t i o n ........................................ 32 HISTORY OF THE E X P E R I M E N T ......................... 37 Instrumentation - I n s t a l l a t i o n s ................. 37 Precipitation ................................ 37 Stream F l o w .................................... 39 Soil L o s s e s .................................... 41 Ground W a t e r .................................... 42 Period of Standardization, 1934-1939 ............ 42 Clearing Operations, 1940 43 Mountain Farming Treatment, 1941-1931 46 CHANGES IN SOME BIOLOGIC AND EDAFHIC CHARACTERISTICS OF TEE WATERSHED AS A RESULT OF LAND U S E .......... 75 Vegetation Changes ............................... 75 vi Coppice Forest .................................. 75 P a s t u r e s ........................................ 79 Abandoned Cornfield 79 ............................ Trout H a b i t a t .................................... 82 .................................. 85 Soil L o s s e s .................................... 88 In fi 1 tra t i o n .................................... 92 Mechanical Analysis 95 Edaphic Changes ........................... Aggregate Analyses .............................. 98 Soil Organic M a t t e r ........................... 102 P o r o s i t y ........................................ 107 Volume W e i g h t .................................. Ill Permeability .................................... 114 HYDROLOGIC D A T A .................................... 118 Precipitation .................................... 118 Stream F l o w ...................................... 122 Definition of Storms .............................. 131 Classification of Storms ......................... 132 Distribution of S t o r m s ........................... 3.33 CHANGES IN RUNOFF CHARACTERISTICS ................. 136 Percent of Precipitation Appearing as Surface Runoff .................................... 136 Area of Watershed Contributing to Surface Runoff 141 Changes in Flood Peaks ........................... 144 Flood Peak M a g n i t u d e s ......................... 144 Flood Peak F r e q u e n c i e s ......................... 1^1 vii Distribution of SurfaceRunoff .................... 160 S U M M A R Y ............................................. 171 Biologic Changes .................................. 171 .................................. 173 Runoff Changes .................................... 178 PRACTICAL IMPLICATIONS FOR LAND U S E ................ 180 LITERATURE C I T E D .................................... 184 A P P E N D I X ............................................. 188 Edaphic Changes PURPOSE AND SCOPE With the increasing demands made on our water supplies within the past few decades has come the realization that fundamental research concerning this natural resource which is basic to our national eoonomy is woefully lacking. Be­ cause the water resouroe is so closely linked with climate, it was the consensus of opinion for many centuries that man could alter it no more than he oould alter the weather. This is not entirely true, for, in addition to climate, the available water supplies may be affected by the vegetation and by soil factors. Through his use of the land, man exerts a very significant Influence on both the vegetation and the soil. As a result, he also modifies the water re* source, but the nature and extent of this modification has been a subject of much speculation and controversy. Because of the lack of hydrologio data, it has been im­ possible in the past to establish a scientific basis for the management of water as a natural resouroe. Currently, an increasing demand for such information is being made by many public and private interests. pendable supply of clean water. adequate, pure water supply. Industry requires a de­ Municipalities demand an Many public and civic agencies require information for flood control programs and power projects. The recreation and tourist trades lean heavily upon the nation's water resource. Fish end wildlife in- terests are dependent upon clear, cool streams for the pro­ duction of fish and game. In recent years, many ambitious projects have been Inaugurated; new faotories have been built, cities have doubled their facilities for supplying water to an increas­ ing population, sports and recreation areas have been de­ veloped— all at tremendous expense and all making tremendous demands on the local water resouroes. If the water supply is found to be adequate, clean and pure, suoh ventures prosper. Unfortunately, many of these efforts have been hampered by muddy streams and unexpected stream behavior. When these occur, the land-use pattern of the watershed in question is immediately examined in order to looate the source of the difficulty. Frequently, the watershed will not be entirely in natural forest, but will show a mixed pattern of usage. logged. Small areas may be farmed, grazed or This immediately oocasions muoh heated controversy as to just which area is the cause of the trouble. When mountain farming is one of the factors which appears in a mixed land-use pattern, the decreased value of the development often has been attributed primarily to this practice. Just how deleterious the cultivation of small patches of steep land may be has been the subject of muoh speculation. Thousands of acres of steep forest land have been cleared for use as oropland or pasture in the Southern Appalachians• To farm such land successfully requires great skill and care. 3 Many authorities olalm that muoh of It should never be farmed at all. It Is common knowledge that individual farmers may "wear out" many such mountain farms In a lifetime. It Is the intent of this dissertation to determine the effects of clearing and cultivating steep forested slopes on certain surface runoff characteristics as well as to study some of the resultant biologio and edaphic changes In the watershed. Numerous studies have been made on cultivated and for­ ested watersheds and indirect comparisons made therefrom. To the writer's knowledge, however, there has been no report in which a forested watershed has been calibrated, eleareut and eultivated and a direct comparison made. In this study, oarried out on the Little Hurricane Watershed on the Coweeta Hydrologic Laboratory in Macon County, North Carolina, the forested watershed was calibrated from 1934 to 1940. In 1940 the area was eleareut and from 1941 to date has been subjeoted to mountain farming typical of the Southern Appalachian region. As nearly as possible the land has been treated as though a mountain family lived near the stream and tended the area to make its livelihood. If the effeots of elearlng steep forest land on the hydrologlo behavior of a small watershed can be adequately determined, it should serve as a guide to the land-use ques­ tions on larger drainages and basins. 4 PAST WORK Numerous investigations have been undertaken in many localities throughout the country for the purpose of measur­ ing runoff and erosion. Many of these studies, however, have been confined to cultivated areas and others have been made on a small plot or lysimeter scale. The literature has be­ come so voluminous that no attempt is made here to review it all. Instead only selected representative projects whioh pro­ vide a particularly pertinent background to the present study will be cited. In no oase, however, has the writer found referenoe to a forested watershed being calibrated, out-over and put into agricultural land use. Since 1930 the United States Department of Agriculture has established 19 soil conservation experiment stations in­ cluding numerous cooperative projects with state Agricultural Experiment Stations. Similarly, the Forest Service, United States Department of Agriculture maintains 14 stations where researoh in watershed management is currently being conducted. In addition, at least nine other watershed researoh centers are conducting studies under the jurisdiction of other federal and state agencies including the Corps of Engineers in the Department of the Army, the Weather Bureau in the Department of Commeroe, the Geological Survey in the Department of the Interior, the Tennessee Valley Authority and the New York and Michigan State Departments of Conservation. Following, a list 5 of watershed research centers in the United States prepared by Frank: and Netboy (14) is presented. WATERSHED RESEARCH CENTERS IN THE UNITED STATES1 (AS OF JANUARY 1, 1950) U. S. Department of Agrloulture Forest Service (primarily in forest, brush, or range areas) Sierra Aaeha, Globe, Arlz. San Dimas (southern California), Glendora, Calif. Continental Divide, Fraser, Colo. (Bureau of Reclamation, Department of Interior, oo-operating on snow-cover re­ lations phase). Front Range, Woodland Park, Colo. Western Slope, Delta, Colo. Coweeta Hydrologic Laboratory (southern Appalachian Moun­ tains ), Dillard, G a • Boise Basin, Boise, Idaho Buckeye, Athens, Ohio Delaware Basin, Bethlehem, Pa. Central Piedmont, Union, S. C. Great Basin, Ephraim, Utah Wasatch, Farmington, Utah Tallahatohle, Oxford, Miss. Mountain State, Elkins, W. Va. Soil Conservation Service (in agricultural cultural areas) ___________ , ____ Coshocton, Ohio Watkinsvllle, Ga. Edwardsville, 111. Guthrie, Okla. Lafayette, Ind. Waco, Tex. Iowa City, Iowa Blacksburg, Va. Boonsboro, Md. Chatham, Va. College Park, Md. Staunton, Va. East Lansing, Mich. LaCrosse, Wis. Hastings, Nebr. Fanmlmore, Wis. Ithaoa. N. Y. Department of the Army and Department of Commerce Corps of Engineers, in co-operation with Weather Bureau Central Sierra Snow Laboratory, Soda Springs, Cal. Upper Columbia Snow Laboratory, Marias Pass, Mont. Willamette Snow Laboratory, Blue River, Ore. lyrank, B. and A. Netboy, Water, Land and People Knopf, Ino., New York, 331 pp., 1950. 6 U. S. Department of the Interior Geological Survey Central New York, Albany, N. Y. (in co-operation with New York State Department of Conservation) Green River, Tacoma, Wash. (in co-operation with city of Tacoma, Wash.) Tennessee Valley Authority Chestuee Creek, Athens, Tenn. White Hollow, Norris, Tenn. Copper Basin, Copper Hill, Tenn. Henderson County, Tenn. Stations at Statesville, North Carolina (5) and Watkinsville, Georgia (32), both on the Piedmont, are engaged in cropping and erosion control measures, and measure runoff from small plots, lysimeters and field watersheds. Included in the studies conducted at Statesville are plots on two wooded watersheds. A comparison of land use practices at the former station indicates decreasing soil losses in the following order: fallow, continuous cotton, rotation (cotton and corn with winter cover crops), grass, woods burned annually and unburned woods. In surface runoff the trends are the same except that the burned woods area yields a higher percent of precipita­ tion appearing as surface runoff than does the grass area. In the case of the fallow area, over 17i percent of the pre­ cipitation appears as surface runoff as compared with 0.7 percent for the unburned woods. The first experimental watershed project of the Soil Conservation Service was established near Coshocton, Ohio in the Muskingum Watershed Conservancy Distriet (25, 32), Intensive studies are being made there on the effects of land use and erosion-oontrol practices on the conservation of soil and moisture and on flood flows for 44 complete water­ sheds supporting various cover types. These watersheds , range In size from three to 4*600 acres. An analysis of soil water relationships on four small watersheds at Coshocton was made by Dreibelbis and Post (11) in 1941* A comparison among a wooded, pastured and two oultivated watersheds all on similar soils showed a much lower volume of surface runoff for the wooded area. On the wooded watershed only 0.11 inch or 0.2 percent of tlie precipitation ran off compared with 0.60 inch and 1.4 percent for the pastured area and 6*35 inches or 15*0 percent runoff for one of the oultivated areas. A 250 acre experimental tract near Zanesville, Ohio, including three gaged watersheds, was established in 1933 to study the effect of land use on runoff and erosion. in this study is a 2 .23-acre wooded watershed. Included For the five-year period from 1934-1938, Borst and Woodburn (8) noted the average soil loss from this watershed as 0.017 tons per acre per year. The average annual runoff was noted as .1246 inches which amounts to approximately 0*34 percent of the average annual precipitation. Hear LaCrosse, Wisconsin (17) a 160-acre tract contains three gaged watersheds: a pasture cleared of timber, a grazed 8 hardwood forest and a typical ungrazed woodlot. An analysis of eight intense storms ocourring in 1935 indicated that about and 3 percent of the preolpltation appeared as surfaoe runoff on the timbered-grazed and eleared-grazed watersheds respectively while on the ungrazed wooded area runoff ooourred only twice and then in quanti ties so m a l l as to be insigni­ ficant. The same trends were indicated for soil losses frosi the watersheds. An experimental watershed project for the Blaoiclands region has been established on the Brazos Drainage Basin near Waco, Texas (32) where erosion, land use, hydrologle and soil data are being studied on thirty watersheds. in these studies are several wooded plots. Included An analysis of soil losses and surfaoe runoff indioate similar results to those obtained at Statesville, N. C. Plots on virgin woodlot yielded only 0.122 percent surface runoff and 0.002 tons per acre soil loss oompared with nearly 30 and 10 percent surface runoff and 65 and 23 tons per acre per year soil loss for fallow and continuous ootton plots respectively. A project designed to study the effects of land use and cultural practices on surface runoff was established in 1940 jointly by the Soil Conservation Service and the Purdue University Agricultural Experiment Station at Lafayette, Indiana (4) • Inoluded in the twenty gaged drainage areas ranging in size from two to four and one-half acres are two wooded watersheds. Up to 1949, however, no treatment of the wooded watersheds had been attempted. A 9 A similar study involving two small agricultural water­ sheds and one wooded watershed was initiated in 1940 by the Soil Conservation Servioe in oooperation with the Michigan State College Agricultural Experiment Station near East Lansing, Michigan (15» 27). This project is unique in that the majority of the installations on the two oultivated water sheds are designed to record the results dlreotly on one master recorder and switch panel. The primary objective of this Investigation is the study of the hydrology of farm lands under winter conditions of anow-cover and frozen soil. Results of the investigations to date indicate a marked difference in soil losses between the wooded watershed and the two agricultural watersheds. Total soil losses from the wooded watershed for a ten-year period amounted to only 64 pounds as compared with many tons from the two oultivated watersheds. Similar differences in surface runoff were noted except under conditions of frozen ground and snow cover. A commercial olearout treatment was applied to the wooded watershed in 1951, however, results of this treatment will not be available for at least five years. Experimental watershed studies by the Porest Service, United States Department of Agriculture, are being conducted at six of the Porest Experiment Stations: Southeastern, California, Southwestern, Rocky Mountain, Intermountain, and Northeastern (32)• 10 At the Southeastern Station, hydrologlo studies are being made at the Bent Creek Experimental Forest near Ashe­ ville, North Carolina, the Calhoun Experimental Forest near Union, South Carolina, and at the Coweeta Hydrologlo Labora­ tory near Franklin, North Carolina. Nearly all of the water­ shed work, however, is eurrently being conducted by the Coweeta station (30). In addition to the study covered in this dissertation, researoh projeots include the determina­ tion of the effects of the following treatments upon water yield and water quality: all major vegetation, (a) permanent eomplete removal of (b) temporary eomplete removal of all major vegetation, (e) removal of riparian vegetation, (d) local logging praotlees, (e) woodland grazing, (f) removal of understory vegetation (laurel and rhododendron), (g) temporary defoliation by gas, and (h) forest fires. The results of these investigations are summarized be­ low: (a) permanent oomplete removal of vegetation increases water yields by 17 area inches annually, (b) temporary eom­ plete removal of all major vegetation inoreases water yields by approximately 17 inches and this increase becomes pro­ gressively less as the vegetal cover increases, (e) the re­ moval of riparian vegetation tends to eliminate diurnal fluctuations in stream flow, (d) local logging practices, particularly poorly located and constructed logging roads, effect a marked increase in erosion and stream turbidity, (e) woodland grazing brings about a marked inerease in over- land a t o m runoff and aroslon and shows that ths oattls grazed on ths watershed failed to thrive (21), (f) the re­ moval of an understory of laurel and rhododendron effeets an increase in water yield of approximately three area inohss per year, (g) preliminary observations indicate that temporary defoliation of vegetation by gas may be used as an emergency measure in extreme drought periods to reduce transpiration and thus inorease water yields and (h) no significant change in streamflow resulted from a forest fire, probably because the soil under the litter layer was very moist and new leaf growth and sprouting helped to pro­ tect the soil from rainfall impact before intense storms occurred. The major work center for the California Forest and Range Experiment Station is the San Dimas Experimental Forest near Los Angeles. Projeots are under way here to study the disposition of rainfall as influenced by watershed conditions including vegetation, soils, geology and topography; and to develop methods of watershsd management, including the treat­ ment of areas denuded by fire, to assure maximum yield of usable water and satisfactory regulation of flood runoff and erosion. Installations inolude 17 watersheds, 18 experimen­ tal plots and 26 large lysimeters. Forest influences and watershed management investiga­ tions at the Southwestern Forest and Range Experiment Station are oarried out on the Sierra An oh a Experimental Forest near 12 Globe, Arizona. Work projects there are designed to deteiv mine the Influence of vegetation (forest, evergreen shrub, and range) on stream flow, water uses, water losses, erosion and sediment production. Gaged watersheds, plots, and natural lysimeters are utilized. In addition to the Sierra Ancha Experimental Forest, experimental plots are located in representative areas throughout the Salt River Watershed. Plot studies on range land on the Sierra Ancha station demon­ strated that ungrazed range land with good plant cover pro­ duced higher water yields and much greater soil losses than overgrazed range with poor ground cover. (31)* Hydrologic investigations of the Rocky Mountain Forest and Range Experiment Station are carried out at the Fraser Experimental Forest near Grand Lake, Colorado; the Manitou Experimental Forest near Colorado Springs, Colorado; and at the Western Slope Research Center near Delta, Colorado. At the Manitou station, studies are being made on the influence of grazing, timber cutting, and revegetation of depleted watershed lands upon water supplies and more particularly upon erosion and sedimentation. Experiments at the Fraser Experimental Forest are designed to show the influence of lodgepole pine and spruce-fir forests and of the cutting of this timber upon the yield of water largely from stored snow. At the Western Slope Research Center major effort is devoted to the analysis of range and watershed problems for drainage basins of western Colorado. Small grazing and reseeding 13 projects hare been ••tab11shed and plana apa being drawn for studies of the effects of vegetation and grazing on infil­ tration and erosion. At the Int e m o u n t a i n Forest and Range Experiment sta­ tion (V) tests are under way to study the effeets of forest, brush and herbaceous plant cover in natural,depleted and restored oondition on the infiltration, storage, fertility, biologj and stability of forest and range land soils; to determine land use practices for stabilizing eroding water­ shed soils and for maintaining soil stability under the im­ pact of grazing, logging and other wildland uses. Studies are being conducted on ooarse, granitic soils of southwest Idaho; various soils on steep slopes of the Wasatoh Mountains in northern Utah; and on heavy limestone soils on the Wasatch Plateau in central Utah. A surface Infiltration study made on the Uinta National Forest la Utah in 1951 indicates that Infiltration rates averaged from 5 to 50 percent loser on grazed sites than on ungrazed areas. Storm runoff from the grazed plots ranged from 50 to 100 percent more than on the ungrazed areas and soil losses on the grazed plots averaged six to eight times more than on ungrazed areas (35). At the Northeastern Forest Experiment Station, a study was initiated in November, 1 % 8 on the Lehigh-Delaware Ex­ perimental Forest (2S) of about 1800 acres to detemlae the influence of the present scrub—oak cover on runoff. After a a period of calibration it la planned to convert the cover from sorub-oak to a better foreat type by forest management and protection Measures and to evaluate the effect of these changes In cover on runoff and ground water. The earliest hydrologic Investigation In this country oonoerned with the Influenoe of forests on streamflow and run­ off was Initiated in 1909 by the U. S. Foreat Service and the TJ. S. Weather Bureau at Wagon Wheel Gap, Colorado. Bates and Henry (6) reported In 1927 that the cutting of forest cover Increased the total annual water yield, Increased water yield from snow and produced Increased erosion. They further Indicated that the results were not too conclusive due to porous soils, thin original cover and prollflo sprout­ ing of aspen. One of the earliest Investigations was that Initiated by Ramser In 1917 near Jackson, Tennessee. He worked with six watersheds varying in size from 1.25 to 112 acres, five of which were In nixed land use and contained forest cover varying from 14 to 55 percent. Ramser reported In 1927 (24) that forest oover has a decided Influence In reduolng the rate of runoff from a watershed except when antecedent rainfall has been high in which case the influenoe is slight. In 1932 a study was started by the Geological Survey, United States Department of Interior in cooperation with the State of Hew York Conservation Department to determine the influence of reforestation on stream flow in state forests 15 in central New York. Submarginal lands were purchased and planted to coniferous tree species. Ayer (2) reported in 1949 that up to that time practically no significant change in runoff had been effected. One of the most n c e n t reports is that of the White Hollow Watershed in Union County, Tennessee published by the Tennessee Valley Authority in 1951 (29)* The 1715-acre White Hollow Watershed was set aside for watershed studies in 1936. Following acquisition, watershed management inoluded extensive erosion-control operations and tree planting. The study shows the following changes in surfaoe runoff and other hydrologic characteristics as a result of 15 years of improve­ ment and management: (a) The improvement in forest cover which occurred resulted in greater watershed protection with­ out measurable decrease in water yield. (b) There was no shift in the seasonal runoff pattern as a result of land-use changes. (o) No measurable change took place in the total quantity of evapo-transpiration plus other losses. Apparently, since a greater density of vegetal cover must be supported by greater water use through transpiration, balancing factors were in operation. (d) Peak discharges during the summer season were markedly reduced. Reductions in winter peak dis­ charge rates were not appreciable. (e) The greater part of the peak discharge reduction occurred in the first two or three years of Investigations, smaller reductions contin­ uing after that time. (f) Modification of summer peak 16 disoharges were so great that the frequency of peaks during the latter jeara was muoh less than during the earlier years, (g) The time distribution of surfaoe runoff was materially changed. Surfaoe runoff discharge was prolonged to produoe a more sustained flow, (h) Comparison of sediment reoords based upon manually oolleoted samples during early years with reoords obtained during the past year (1950) by means of an automatie sampler shows elearly that there has been a eery material reduotion in sediment load during the 15-year period of observations. It Is apparent, after a review of the literature, that a direct comparison of the results of this study with any previously reported is virtually impossible. Many studies have demonstrated that watersheds or plots with undisturbed forest oover yield less surfaoe runoff and produees less soil loss than grazed or burned forests, pastures and crop­ lands. In few oases, however, have attempts been made to show changes in surfaoe runoff except in terms of total surfaoe runoff expressed as a pereant of the precipitation. The study made in this dissertation is unique in that the watershed was eallbrated under forest oonditions, the forest oover was removed and land use praotloes then applied. In addition, an adjaeent watershed with similar oharaeterlstios was maintained in continuous forest cover thus providing a further control. Consequently, an opportunity was provided to study more detailed ehanges in surfaoe runoff. THE COWEETA HYDROLOGIC LABORATORY With the recognition of the need for additional raaaaroh la watershed management eaaia tha realization alao that tha selection of altaa for such raaaaroh would ba complicated and dlffloult. Foresters, hydrologists and aAgiaaara ooatributad rigid apaolfloatloaa which had to ba fulflllad if tha fladiaga wara to ba valid and ef wore than local significance. One area whloh mat avary important requirement waa a 5,600-aere tract ia tha Wantahala Mountalna of waatara Worth Caroliaa. Thla tract, aatabllahad ia 1933 hy tha United Stataa Forest Service, la now internatlonally known aa tha Coweeta Hydrologic Laboratory. Several faotora oombina to make tha area ideal aa a natural laboratory aultable for fundamental hydrologio raaaaroh. Rainfall ia high, averag­ ing 72 inches par year, and ia rather uniformly distributed throughout tha year. Beoauae of tha frequency of atorma and tha uniformity of tha atorm pattern, it ia poaalble ta obtain valid reaulta in much shorter time than in an area of leas praolpltatlon. Approximately 96 percent of tha precipi­ tation occurs aa rain so there is little snow to complicate tha studies. Seepage losses are virtually eliminated aa tha deep and porous soils of tha area are derived from weathered granite• Topographically, this particular section is also ideal in that its steep slopes and sharp ridges foxm natural boundaries for tbs many small drainage basins— eaoh an in­ dependent hydrologic unit— necessary for research of this type. Blevations vary from 2,200 to 5,200 feet within the boundaries of the station. Although over half of the Coweeta area was cutover 25 years before the government acquired ownership, land use praetlees have altered the character of the forest it­ self very little. A dense mlxed-hardwood forest, typieal of mueh of eastern United States, is predominant at Coweeta. The outover lands support second-growth forest and the re­ mainder of the land is in old growth. Chestnut was formerly the major species but has been wiped out by the blight. largest part of the forest is now in oaJfc-hloJcory. 15 pereent is in core hardwoods: The Another yellow poplar and northern red oaic intermixed with hemlooJc along the streams. Sugar maple, yellow birch, beech and pitch pine occur occasionally at the lower elevations. Because of the similarity of this area to many other parts of the country and beoause of the favorable pattern of precipitation, data derived from studies on the Coweeta tract can be applied elsewhere. Consequently, research conducted here in water behavior and management has national as well as regional and local signifloanee. Figures 1, 2 and 3 give the looatlon, drainage pattern and individual drainage areas respectively of the Coweeta Hydrologie Laboratory. Figure 1. L o c a t io n of th e Cov/esto h y d r o l o g i c L t . b o r r .t o r y . iis. N i H ^AHAlA b- «/> COWEETA HYDROLOGIC / ^ i milM La b o r a t o r y NORTH C A R O LI HA C tO R Q IA B Lv/ E R ID G Mountain CVty Figure 2. DRAINAGE PATTERN Coweeta Hydrologic Laboratory (Dryman Fork excluded) Albert NW" CHAIMS Ficken* Mosm 1 TBS LITTLE HURRICANE WA TE RSH ED The Little Hurricane Watershed, designated as drainage No. 3, is located in the Ccweeta Hydrologlo Laboratory, Maoon County, North Carolina. Its loeation within the Coweeta area is shown in Figure 3- The waters of Little Hurrieane Branch flow into Shope Creek and thence into Coweeta Creek which is a tributary of the Little Tennessee River. The watershed eontains 22.79 aores and assumes the shape of an isoseeles triangle. tially southeast. The aspeot or exposure is essen­ Figure 4 shows an overall view of the watershed, and Figure 5» «• *ep of the Little Hurrieane Drainage• Land Use History Arlor to 1857 the area was included in the lands of the Cherokee Indians and used primarily as a range for live­ stock. In order to improve the quality of gracing the Indians practiced spring and fall burning of the woods. By eliminating the undergrowth and litter, both the Indians and the livestock oould find nuts and acorns more readily. Furthermore, burning the woods was thought to eliminate milk-sick, an ailment of stock that evidently eaused much concern among the Indians as well as the white settlers who followed because it not only killed the stock but was con- Figure 4. The Little Hurricane Watershed. Fif-ure 5 LI'TTLi. ilUF.riC^Ni; KATiFbhEL Dreina-.e No. 3 22.79 Acres Coweeta Iiydrolot’l c Laboratory CHA INS 3190 25 traeted by human beings as well. It was believed that stoek contracted the disease in the dark, damp coves and that burning would remove the cause. The Cherokees were removed from this seetion of the oountry in 1837 by the federal Government and were plaeed en the Qualls Indian Reservation* In 1835 s hurrieane is reported to have levelled all the timber in the Little Hurrieane and the Hurrieane drainage adjacent to it; henee the name of the watersheds. From 1835 to 1857 uhite settlers pushing into this region grazed the drainage area to some extent and praetloed semi-annual burning mueh as the Indians before them had done. In 1857 the seoond-growth timber on the lower ten aeres of the water­ shed was eleared for farming and the area was cultivated until 1887* This ineludes nearly all of the areas vhieh are now the lower pasture and abandoned oornfield. The yields became so low that the fields were then used only for grazing until 1900. In 1901 the area was included in the land purehased by the Nantahala Company, a land speculation group. From 1901 until 1940 "third growth" timber, largely of the oak-chest­ nut and cove hardwood types re-established itself. the dominant trees were 18-20 inches in diameter. By 1934 The chestnut, however, had dropped out because of the blight. Only the best quality oak, chestnut and yellcm-poplar was logged from the second growth forest adjoining the old 26 field in 1914* The operation was handled by the Gannette Brothers according to the texma of the Ritter Lumber Company whieh had aequired the traet in the meantime. The remaining trees were left unharmed exoept for the damage ooeaaioned by the logging. The U. S. Forest Service aequired the area in 1923 and it became a part of the Nantahala Rational Forest. In 1934 the drainage was inoluded in the area sot aside as the Coweeta Ex­ perimental Forest. This name was officially changed to the Coweeta Hydrologic Laboratory in 1949. After a period of standardization or calibration starting in 1934* the watershed was ale area t in 1940 preparatory to the "mountain farming" treatment. Figure 6 shows cutting operations by CCC anrollees during the winter of that year. Geology and Physiography The Little Hurricane Watershed lies in the Blue Ridge province of the Southern Appalachians • The underlying rock is the Aroheaa Carolina gneiss and schist. The thickness of this formation, which was enormous, was greatly Increased by oomplex folding. As a result of folding and the absenoe of open faults and fractures, there is little likelihood that continuous channels exist whieh would permit the sub­ terranean escape of water through the rock. The parent material weathers to form a relatively deep sell mantle with bare outcrops of rock appearing only on the 412541 28 steeper slops* at high elevations• Two small outorops ooour ring on ths upper slopes of the drainage are shown In Figure 4. The topography of the area Is steep and rugged* The mean sea level elevations range from 2,425 feet at the base or weir to 3,124 feet at the top* The dlstanee from the base to the top Is about one*third mile. The land slopes are quite steep with north-south averages 46 pereent and east-west averages 58 pereent* The mean slope for the water­ shed Is 51 percent and the range Is from 10 pereent near the bottom to nearly 80 pereent at the head of the drainage* The drainage pattern of the Little Hurrloane Branch Is dendrltle, the stream channel Is T-shaped and the slopes are eoneave, all Indicating the youthful stage of the stream. The permanent stream channel Is 436 feet long with a drop of 65 feet* The average stream gradient Is 14*9 percent* The ground water table Is only slightly less steep than the general slope of the land surface, and at four observa­ tion wells ranges from 8 to 16 feet below the soil surface* Climate2 The climate of the Coweeta area Is characterized by moderate temperatures and abundant rainfall* The mean annual ^All ellmatlo values given here are based on 15 years reeerd at weather station #1 (Headquarters area) Coweeta Hydrologle Laboratory, U.S.F.S* 29 temperature is 55°F. and the normal frost-free season ex­ tends from April 17 to October 2 3, a period of 189 days. The are rage temperature during the growing season is 65°F. Heeordings of 90°F. are rare and auaaer nights are oool with alnlauas averaging 58°F. The three eoldest Months, Deoeaber, January, and February average 39°F. Periods of sold weather with teaperatures below 20°F. are short in duration. The highest and lowest recorded teaperatures are 94°F. and -15°F. respectively. The average annual rainfall over the Coweeta Hydrologlo laboratory is 77 inches. Figure 7 shows the rainfall distri­ bution pattern for the Southern Appalachian Region and indi­ cates that rainfall is well-distributed throughout the year. For the past 15 years, preoipltation has averaged 3.2 inches in October, the driest aonth, and 7*2 inches in ICaroh, the wettest aonth. The greatest aaount of rainfall is received in the southwest portion of the area and the least in the northeast corner. The dlfferenee between these two zones is about 20 inches a year. The average Monthly evaporation, measured by a standard U. 8. Weather Bureau evaporation pan, varies froa 0.98 inches in Deosaber to 4*10 inohes in May. The average total evapora­ tion for the year is 33*56 inches, or 2.80 inohes per aonth. Cliaatie suaaarios indicating aean, aean aaxiaua, as an alniaua, absolute end absolute alnlaua teaperatures as well as evaporation rates froa a free water surface by 1 ^617677334 INCHES 55-40 Figure 7. MEAN ANNUAL PRECIPITATION 40*50 50-60 60-70 m m o to Coweeta ffcrdrologlc Laboratory n*ur J 31 years and months art given la the Appendix. Preeipitation summaries are given for the little Hurrieane Watershed in the aeetion on hydrologio data, page 120. Soils The soils on the watershed are derived froa Arehean granite gneiss and sehist. The parent roots weather to form a relatively deep soil mantle. A eolluvial fill which is more than 20 feet thiet oeoars on the lower portion of the drainage. On the upper slopes the soil mantle 5-10 feet in thieicness. from Two root outcrops and evidenoe of an old landslide are present. The lower pasture land shows the effeots of former eultivation more noticeably than the area reeently eropped. Zxeept for the eolluvial fill at the base of the water3 shed the soils are classified as Porters loam (10) • Xn the eolluvial fill they are Porters loam, eolluvial phase. The surface soil of Porters loam ranges from 6 to 12 inehes in depth and eonslsts of mellow and friable brown loam. The subsoil, to a depth of 20-28 inehes, is a red to reddishbrown, friable and crumbly elay-loam. Below this is a reddish- brown mixture of olay loam and disintegrated roolc. 3The entire watershed was mapped by Devereux ei al in 1929 as Porters stony loam. After the examination of numerous soil profiles throughout the area and more recent descriptions of the Porters stony loam, Porters loam and Porters loam, eolluvial phase, it is believed that the above elasslfloatlon is more nearly correct. m Porters lotH| eolluvial phase, is similar In eolor to typical Portara loam but is much deeper and in soma plaoaa praotioally no difference axiats batwaan tiia surface soil and the subsoil. This soil contains a fairly high, eon tent of rock fragments which have rolled down from the mountain sides. Aoeordlng to Devereux at al (10) Porters loam Is eonsldarad as one of the batter agricultural soils of the oounty. If it oaauplad more favorable relief probably all of it would be cultivated, but under existing oondltions only a small part is in sueh use. The principal orop is corn and yields range from 15 to 40 bushels per aere. Cabbage, pota­ toes, snap beans, and pumpkins do well also. Porters loam is one of the good pasture-grass soils of western North Carolina. Soils of the eolluvial phase are used for the pro- duotion of eorn, cabbage and potatoes, and the yields are about the same as those obtained on typical Porters loam. Vegetation Previous to olearlng, the primary forest vegetation oonslsted of second-growth forest of the oak-chestnut type, with oove-hardwood and yellow plne-hardwood types on smaller areas. Figure 8 indicates the type map made in 1934* The scale for the merchantable timber out on ten acres below the L contour trail during elearouttlng operations in 1940 is ------4 s u r v e y by A. *. Radford and S. D. Marshall, August, 1940 From the files of the Coweeta Hydrologio laboratory, USFS. 33 given b el air (Scribner Decimal C rule with, allowance for defeet): Speciee Board Feet Pliok pliTe — Tellow poplar 7,750 Black oak 2,310 White oak 420 Chestnut 250 Basawood 590 Rad oak 56< >60 Total (141 logs) - 13,926 To determine what herbaoeoua eorer cornea la naturally following clearing, an observational survey was Made of the ground ©over in August, 1940. Speoiea were Identified and napped aoeording to relative preponderance• The following speoiea were identified: 1. Herbaceous weeds: Caealia atrlplleilfolla Havener!a oiliaris Viola papilionaeea Ambrosia artemlslifolla - Pale Indian plantain Inng-spur orchis Wood violet Ragweed, hotweed, bitterweed Joe-pye weed Zupatorlum purpursum Aoalypha vlrglnlea Three-seeded mercury Ozalls strleta - Wood sorrel Pigweed, tumbleweed Ameranthus hydridus Tellow foxtail, Setaria lutescans pigeon grass Pan!cum lanuglnosum Switch grass Xleoeharis obtusa - Spike rush Green bulrush Soirpus strovlrens Soirpus eyperinus Wool grass Phegopteris hezagonoptera - Broad beech fern Pterldium latiusculum - Bracken fern Dandy or thin-leaved Kyllinga pumila sedge Ludwlgia alt# m l folia - Seedbox (false loose­ strife ) Spanish needles Bldena bipinnate Beggar* tick Bldens frondosa Common or broad­ Plantago major leaved plantain Trautvetterla earollaensls- Tasael-rue or false bugbane 34 Clmleufuga raetaoia Bupatorium urtieaefolium - BlaeJc snaJceroot, blaefc o oho ah - White snaJceroot 2• Robiaia pseudoeaela, Sassafras Tarifolium, Smilax glauea and Vitis bioolor seedlings* 3. Vitis bioolor-Robinia psaudoaeaaia seedlings* 4- llriodeadroa tulipifara, Coraus florida, and Aeer rubrum trldens saadllags. 5* Aeer rubrum trideas, Diosproa Tlrglnlana, Sassafras Tarlfollum seedlings* 6. Sassafras Tarifolium* 7* C o t er herbs and seedlings of trees: Folystlohum aerostloholdes- Dagger fern Adiantum pedatum - Maiden-hair fera Phegopteris hexagoaoptera - Broad beeeh fera Dryopteris aoTeboraeeasis - New TorJc fera Saagaiaarla eaaadeasls - Bloodroot Liriodendron tullplfera - Tallp poplar, yellow poplar Cora as florida - Flowering dogwood 8. Rhus eoppallna, Rhus typhina and herbaeeous weeds. 9. Oxydendrum arborsum, Roblnla pseudoaeaela, Carya s p ., Castan#a dentata seedlings sprouts (Originally an oak-ehestnut type)• Figure 9 shows the post-ole arlag rage tat ion nap. I Figure 8. LOC/iL FOREST TY^KS (Manned in 1934.) Vv'atersiied \To. 3 - Little Hurricane Branch Coweeta Hydrologic Laboratory 10 15 Legend 1 2 4 10 15 17 °itch nine-acarlet oak-cheatnut oak Scarlet oak-che:; tnut oak-bl; ck oak Scarlet oak-black oak-white oak Red oak-cheetnut oak-scarlet oak Yellov; roplar-red oak-hickory Yellov: ooplar-rec muole-white oak Scale: 1 inch z 330 feet Figure 9. ■°OCT CLEARING VEGETATION ’vL'.'D* Watershed No. 3 - Little Hurricane Branch Coweeta Hydrologic Laboratory Scele: 1 inch Legend 1 2 3 4 5 6 7 8 9 Herbaceous weeds Robinia oseudoacacia, Sassafras albidum, Smilax glnuca End Vitis bicolor seedlings Vitis bicolor and Robinia oseudoacacia seedlings Liriodendron tulipifera, CornuE florida and Acer rubrun tridens Acer rubrum tridens, Diosnyros virginiana, S a s s a fra s albidum Sassafras albidum Coverherbs, cove tree seedlings Rhus copallina, Rhus typhina and herbaceous weeds Oxydendrum arboreura, Robinia pseudoacacia, Carya sr>. , and Castanea dentate (originally oak-chestnut) Survey made August, 1940 by Albert E. Radford and Eugene D. 'Marshall. H I S T O R Y OF THE E X P E R I M E N T Instrumentation - Installations Precipitation. Precipitation or recharge to the water­ shed is measured by three standard rain gages, numbers 16, 20 and 67* Gages 16 and 20 have been in operation contin­ uously since July 4, 1934* Gage 67 was installed on June 9, 1940 and has been in continuous operation since that date. Previous to the installation of gage 67, measurements from standard rain gage 21 were also applied to the area. Until June 9» 1940, rainfall intensities were measured by recording rain gage 1, located on the adjacent drainage area No. 7* Gage 10 has been used since its installation on June 9* 1940. Recording rain gage 10 and standard rain gage 67 are looated adjacent to eaoh other in order to provide a check against the acouracy of the former. Figure 10 pic­ tures recording rain gage 10. Charts are changed at least once a week on the recording rain gage and it is completely serviced and checked at least once eaoh year. The standard rain gages are read following each storm or as nearly so as possible. Standard rain gage data are summarized and tabulated by months, hydrologio seasons, calendar years and hydrologic years. Data for the three gages servicing the watershed as well as the weighted areal precipitation is given in the Appendix. Figure 10. Recording rain gage used for rainfall intensity deterainations 39 To compute the total araal precipitation of the water­ shed, the Horton-Thie sen Maaaa method (19) of weighting preoipitation la uaad. Figaro 11 ahoira tha gaoaatrio division of tha watershed for thaaa oaloulations, as wall as tha leeation of tha diffarant installations. Praoipitation intensitias are computed froa tha reoording rain gaga eh arts. These data are than tabulated on praoipitation intensity records and oorraotad to agree with tha reading of tha standard rain gaga loeatad adjaoant to it. Sinoe tha area of tha watarahad is only approximately 23 aerea, a single raeording rain gaga is used tor intensity determinations • Stream flo w . To measure tha disoharge from tha watar­ ahad a 90-degree V-notch weir with a 35i inch blade, similar to that pictured in Figure 10, with a continuous water stage recorder, was installed on July 5, 1934* Tha stilling basin was 6*10" x 4 ’11" x l v5"» with wooden walls and bottom and a log wall betwoen tha silting and stilling basins* In order to measure a wider range in streamflow and to accommodate greater debris loads, anticipated as a result of tha watershed treatment, tha 90-degree V-notoh oontrol was replaeed with a Columbus CIA deep notch oontrol in December, 1939. trol. The Figure 14 pictures the latter stream con­ In addition, a concrete stilling well was installed. eapaelty of the blade on the Columbus control is 2,500 o.s.m. Figure 11. Rain Gage Service Areas. Little Hurricane Branch Cov.-eete Kydrologic Laboratory Legend > > i I — % — Standard Rain Guges Recording Rain Gage Groundwater Y-------------------------CHAIN / / / r2 + 2 5 • ZO 41 To convert the streamflow data Into usable form, the head and time readings are taken from the water level charts and then converted Into volume values, 1. e . , cubic feet per second (o.f.s.) and cublo feet per second per square mile (o.s.m.). used as well. For special storm studies Inohes par hour is Volume data are, as in the case of precipita­ tion data, summarized by days, months, hydrologic seasons and years as well as by individual storms. For special storm studies volume values are plotted over time to give the storm hydrographs. Water stage recorders are completely serviced and over­ hauled at least once per year. Recorder charts are changed at least once per week and frequently following major storms. Sample calculations and recording forms for both preci­ pitation and streamflow data are given in the Appendix. Soil losses. In August of 1941 the concrete debris basin shown in Figure 17 was constructed to measure soil losses from the watershed. The design of the debris basin was based upon Stokes Law of the settling velooity of parti­ cles in a liquid. The basin oonsists of three individual basins each with five baffles. The basin obviously catches all but the finest materials as evidenced by the absence of sediment in the stream channel below the basin. The debris basin is cleaned each spring i f required or more frequently in seasons with major storms producing heavy soil losses. The water is diverted around the basin from the 42 weir spillway permitting continuous streamflow measurements. Samples of sediments from eaeh basin and within eaoh baffle are taken for volume determinations. The sediment is per­ mitted to dry, then it is measured and removed. From the dry weight of the samples and the volume of silt applying to eaoh sample an estimate is made of the dry weight of the sediments trapped 1a the basin. Prior to the installation of the debris basin estimates of soil losses were made from samples taken from the silting basin. Ground w a t e r . To study fluctuations in the ground water surface, four wells were Installed during the summer of 1941* Daily measurements of the water elevations in the wells were made until November 1, 1942. At that time water stage recorders were installed on wells 1 and 2. Weekly measurements of the water levels in wells 3 and 4 were made until April 24, 1944 when, due to a shortage of funds and labor, water level readings in these two wells were discon­ tinued • Beoause of the restrieted soope of this dissertation and the time limits imposed, no attempt was made to include ground water studies. Period of Standardisation, 1934-1939 On July 3, 1934 o 90-degree Y-noteh stream control was put into operation along with tmo standard rain gages 43 (16 and 20). Figure 12 piotures a 90-degree V-notoh weir similar to that originally installed on this watershed. A recording rain gage of the float type (1) and another standard rain gage (21) were installed on Oetober 18 in the adjoining drainage (Hurricane or area No. 7). A survey of the area was made during the sinaner of 1934* On August 3, 1939 the V-noteh weir was removed and on December 20, 1939 * modified Columbus type 1-A deep-noteh stream oontrol was installed in order to measure a greater range of flows and to aooaomedate greater debris loads. The itaTl mum rate of runoff measured before clearing was 110 o.s.m. following a rain of 4*67 inches in November of 193&. The maximum intensity of this rain was 2.12 inches per hour. Before clearing, the rate of sediment movement, based on accumulation in the weir pond, was 914 pounds per day. Clearing Operations, 1940 Logging and clearing operations were started in Novem­ ber, 1939 and were completed in July, 1940. Figures 6 and 13 show the watershed in two stages of olearing. The mer­ chantable timber on the area below the eontour trail was sold to a looal firm and logged to simulate local praotloes. The balance of the watershed was oleared by CCC enrolees. On the lower portion, stumps over 16 inches in diameter were pulled and the brush was piled and burned around the larger Figure 12 o 90 V-notch weir installation on Ccweeta Watershed No. 1, similar to that originally installed on Watershed No. 3. Figure 13. Clearing operations nearly complete, spring 1940. 46 stumps whioh remained. On the upper portion the trees were out and the slash soattered over the ground to form a mulch. About two aores of the area above the contour trail was burned over by a forest fire which occurred on July 2, 1940. A recording rain gage (No. 10) and a standard rain gage (No. 67) were installed near the center of the drainage area on June 1 9 9 1940. Mountain Farming Treatment, 1941-1951 19 4 1 . In the spring of 1941 an area of 5*6 aeres was plowed using a bull tongue or single foot plow. Figures 14 and 15 show the eornfleld during and following plowing. The field was planted to Hiokory Ki n g oorn with no fertili­ zer used. It was cultivated by hoe during May and June and the crop of 132£ bushels was harvested in November. A concrete debris basin was put into operation on August 28 and four ground-water wells were installed during the summer. Soil mantle depths were measured on cross- drainage lines and volume soil samples were taken from eight selected pits. Figures 14 and 15 give views of the Columbus CIA deep notch weir as well as the concrete debris basin. 1942. The field was plowed again with the single foot plow in 1942 and Hickory King oorn planted. wet summer the corn was worked only once. Because of a Windstorms blew down much of the corn. The yield was 84 bushels or 15*3 bushels per acre. The watershed was fenced during the summer with stook fence and a hog-proof fence was put up around the cornfield. Figure 14. Flawing cornfield for first crop of oorn, April 1940. Figure 15. Cornfield In foreground, following plowing with bull tongue or single-foot plow. 49 Of the 17 acres outside the cornfield about 10 acres was too rough or brushy for pasture and was permitted to grow up Into oopploe forest. One hundred pounds of Cherokee pasture seed mix was sown on the remaining 12 acres. Compe­ tition from trees and shrubs prevented a good oatoh of grass. Cattle Were alternated between this pasture and the adjacent wooded watershed (No. 7). In all there were 336 animal days of grazing on watershed No. 3* Herbaoeous vege­ tation was estimated to be 80 peroent utilized by September 21 and evidence of trampling was oonsplouous. 1943. The debris basin was oleaned on May 28. This wa the first measurement of the eroded material trapped by the debris basin since it was placed in operation on August 28, 1941. Water was diverted around the basin on May 4 and by May 28 the material had dried so that it could be shoveled out. Before cleaning, 200 C.C. samples were colleoted from each baffle and the total volume of silt measured. From the dry weight of these samples and the volume of silt applying to eaoh sample, an estimate was made for the dry weight of the silt trapped in the basins. A total of 437 cubic feet or 13,928 pounds of dry soil material was removed. Pro­ rated over the period from August 28, 1941 to May U, 1943* the soil losses amounted to 22.7 pounds per day. The cornfield was again plowed and this year planted to Golden Prolific corn. The oorn grew well except on the red clay “scalds" on the ridges of the portion below the recording 50 rain gaga. TJia oorn on the steepest slopos suffered from waoiling during thunder storms. The oorn yield was 14*6 bushels per acre. On pasture, eight animals were grazed for 67 days. The sprouts below the oontour trail were out baeJc in August. Tho portion below the recording rain gage was picked almost elean but sprout growth on the extreme upper slopes was untouehed. One hundred pounds of a mixture of Italian Bye grass and Bed Top were sown over the pasture in March. The season of 1943 was mar iced by abundant rainfall in early summer. Particularly Important were very intense thunder storms occurring during June and July. The first notable washing of the field ooourred on June 13 when 1.10 inches of rain fell, largely within a 25 minute period. This rain fell at a rate of approximately 5 inches per hour for a five min­ ute interval and 4 inches per hour for a ten minute period. On June 14 another storm almost identical in rates and amount occurred. As evidenced in Figures 18 and 19 these two storms removed large volumes of soil from the field. The weir basin was completely choked but records of peak disoharge were obtained. Less severe thunderstorms were common in the following days. Another very severe storm occurred on July 5 after five days of light showers. This rain again caused great erosion and gave the highest peak discharge recorded for this area. I Figure 16. Columbus CXA deep notcb. weir. (Debris carried down by storm of June 16, 1949) Figure 17. Concrete debris basin and Columbus deep notch, control. Figure 18 View of weir and ponding basin fter storms of June 13 and 14» 1943* 54 The erosion and discharge of Araa 3 for the** storm* was greatly in axeass of tJhat for any other Coweeta area. Tho groat accumulation of silt and debris made It Impossible to seoure oampletely satisfactory records for the falling stages of the stream dlsoharge. However, the rising and peah stages were obtained. As a result of the erosion uhloh took place during June and July, the debris basin was again cleaned out on Septem­ ber 15-17. Tho water was diverted on September 8. The results Indicated the enormous Increase In erosion In 1943• showing soil losses amounting to 1732 cubic feet or a total of 7 9 #®58 pounds for the period from May 28 to September 8. This amounts to 627 pounds per day. Because of reduction In personnel It was not possible to continue dally observations of the well elevations. After November 1, 1942 the depth to ground water was measured weekly. Water stage reoorders, however, were Installed on wells 3-1 und 3-2. 1944. The cornfield was permitted to grow up Into we and shrubs due to laek of funds and labor. Cattle, however, were excluded from It. A drift fence was built just below and to the right of the recording rain gage to control grazing and prevent concentra­ tion on the lower portion of the pasture. Cattle were alternated between the two portions (designated upper and lower pastures) • Five head of oattle were kept continuously on the area from May 8 to September 22 for a total of 685 1 55 animal days. Nearly all available forage was consumed in the lower pasture and the soil surface was severely trampled and compacted. The remaining perennial herbaceous vegetation was considerably reduoed by frost-heaving the following winter. On the cornfield the most abundant weedy species which became established were ragweed and cinquefoil. were abundant along the fence line. Blackberries Seedlings and sprouts ef leeust and yellow poplar also became established. Only on the "scalds" and in the washes foxmed in 1943 was any soil washing observed. Weekly readings of the water level in the observational wells 3-3 end 3-4 were discontinued April 24* 1 945. Cultivation of a corn crop was postponed for an ther season because ef shortage of funds and labor. The shrubs and woody growth which came in on the area in 1944 continued to develop since livestock again were excluded. This cover apparently afforded considerable protection against soil washing as none was noted on the area itself and very little silt was transported into the stream channels. The "scalds" which developed during cultivation were still quite apparent but showed some evidence of healing. 19A6. On April 16 the debris basin was again cleaned. This measurement applied to the period of September 8, 1943 to April 16, 1946, durisg which the field was not in cultiva­ tion. The net volume ef the sediment was 971 eubio feet and the dry weight was 44#165 pounds. For the 952-day period Figure 19- View or lower portion of cornfield after storms of June 13 and 14, 1943* 57 this is a rats of accumulation in ths basin of 37 poonds par da y. The oornfisld was olsarsd of woods and woody vegetation and ploorod again. Hybrid U. s. 13 yollow oorn was plantod without fertilizer. Figure 20 shows brush burning and plow­ ing in April of 1946. bushels per aero. to the area. orop, also. Tho harvest was 65 bushels or 11*6 This oorn was apparently poorly adapted Door and ground hogs took a hoary toll of tho Figure 21 pioturos tho eornfleld and lower pasture in September, 1946. The fenee between Areas 3 and 7 was opened and the eattle were alternated between the upper pasture of Watershed 3 and nil of No. 7. both areas. There was a total ef 1,377 days of grazing on Observations indicated that two-thirds of the use was on Area 3 and 85 pereent of tho food eamo from that aroa. On May 7, ono hundred pounds of lespedeza seed were sown on the lower pasture and the upper pasture below the eontour trail. The extremely dry weather in August apparently elimina­ ted most of the seedlings established from this planting. No intense thunderstorms ooeurred on tho aroa and no severe soil washing was observed. 1&2L. Tho dobris basin was oloanod in April to deter­ mine soil losses during erop year 1946. Tho total not volume of sediment was 383 eublo foot with a dry weight of 13,507 pounds. Tho average dally sediment accumulation was 37 pounds per day. I 440594 Figure 20. Brush burning and plowing in preparation for fourth corn crop, April 1946. Figure 21. View of watershed in September 1946. 60 During tills year, the lower slopes of the cornfield were plowed with a hillside turning plow and the upper slopes with a single foot. Fertilizer was used for the first tisie. The total amount used was 800 pounds of 4-10-4. Hickory Kin g corn was planted again and the yield was 90 bushels or 16 bushels per aore. Three head of oattle were kept oontinuously on the watershed pasture area from May 26 to September 29 for a total of 588 days of grazing. Woody sprouts wore out baok in February on the lower pasture. 1948. The sediment in the debris basin was measured April 1 3 9 1948. This measurement applied to the period starting on April 2, 1947* The basin held 426 oubie feet of material whieh had an estimated dry weight of 16,186 pounds representing a soil loss of 42 pounds per day. The oornfleld w a s again plowed, this year with two one-horse side-hlll turning plows and planted to Hlokory Kin g oorn; 800 pounds of 4-10-6 fertilizer were applied as evenly as possible. After the corn w a s planted, a very dry period prevailed until the latter part of June. The corn, however, made good growth and at no time did the leaves indicate that the crop was suffering from insufficient moisture. The total yield of 80.6 bushels or 14.4 bushels per acre is lower than that of 1947 even though this was a bumper oorn crop year elsewhere in this area. crop. Figure 22 shows the harvesting of 1he 1948 eora Figure 22. Harvesting 1948 corn crop, yield 14.4 bushels per acre. 62 An averagt of four hoad of eattlo grazed continuously on tho two paoturoo and woro giv on supplemental food tho loot two wo oh a on tho aroa* Tho a t o m produolag tho maximum disohargo for tho year ooeurrod on August 2 whan a peak of 141 e.s.m. was roeordod. Tho hydrograph rofloetod every burst of proolpltatlon by haring sharp rises, peaks and rooosslons, Indieating tho sensitivity of tho watorshod. Ororland flow from tho oorn fiold to tho erook ooeurrod July 12, August 2, November 19 and November 29* Observations made during tho above and other storm periods indieatod that most of tho turbidity ineroaso earns from tho eattlo paths loeatod near tho eentor of the watershed. Tho good wood cover that beeame established after the last cultivation of the eornfiold in July apparently kept tho soil from being displaood at a rapid rata. Sxploratory infiltrameter ring tests indieatod tho in­ flue nee of trampling on tho pastured portion of tho watershed when it took 2 area inches of water 20-30 minutes to seep into the soil. On Deeember 10 a thermograph, with one thermocouple in the stream and one in tho air, was installed 10 feet above the log coping at the approaoh end of the ponding basin. 1949. A cleaning of tho debris basin was made April 8 before planting and again following tho July 10 storm. Soil losses amounted to 32,879 end 185,675 pounds respectively or 91 and 1,998 pounds per day. 63 The oorn field was again cultivated in 1949* Hiokory King corn was planted and fertilizer applied for the third time. The soil displacement that had taken place sinoe clearing had exposed large numbers of rocks whioh accumulated on the surface. Fourteen truckloads of medium-sized rook equalling 21 tons were removed. After the oorn was planted in April, numerous intense storms occurred that rilled the soil. The 67 percent above normal rainfall kept the ground vary wet. Following a storm of May 22, replanting of some areas of the field was neoessary in order to get a stand of c o m . Rilling of the steeper portions after the May 22 and 30 storms extended several inches below the plow line and a cultivation whioh followed was not able to obliterate the rills. Figures 23 and 24 show the e o m f i e l d following the May 22 storm. The yield of one bushel per acre is in line with yields along the Little Tennessee River which were one-sixth to one-tenth of those ef 1946. Figure 25 summarizes the earn yields for the seven years of reeord. An average of four oattle grazed oontlnuously on the pasture areas from May 9 to October 10 and were rotated by two week Intervals between the upper and lower pastures. The animals were given supplemental feeding after August 1. Several severe storms occurred on Watershed No. 3 during this year. The outstanding storm was that of July 10 when a reeord peak of 1,850 e.s.m. was recorded. Xt is estimated Figure 23. Appearance of cornfield following May 22, 1949 storm. Figure 24. Rilling in cornfield following storm of May 22, 1949. Yield In buahele P«r 67 that 76 tons of soil and debris came off the drainage In 30 to 40 minutes during this storm. Figures 26, 27 and 28 picture the debris and sediment earrled down by this storm. Fifteen storms were reported with flows greater than 15 e.s.m during the year. Overland flew from the eornfleld to the ereek ooeurrod during all major storm porlods. Observations during storm periods ladloated that highly turbid water flowed through the eornfleld gate and into the main ehannel. Tiro gullies were formed In the two natural swales during the May 22 storm and were enlarged during the June 16 and July 10 storms to a length of several hundred feet and a width of 4 feet with holes up to 15 inehes deep. Overland flew with water running down these gullies Into the permanent ehannel oeeurred during eaeh following major storm. Observations again this year indleated that most of the turbidity, during all but the major storm periods, earns from the oattle paths looated near the eenter of the watershed. A surf see Infiltration study was made In July. The stream and air temperature thermograph. Installed In Decem­ ber, 1948, was operated all year and discontinued on Decem­ ber 31, 1949* Stream profile surveys made in July from the confluence with Shops Creek Indicated considerable variation In temperature along open stream beds, due to shading by forest and brush. Figure 26. Rock deposit in ponding basin carried down by storm of July 10, 1949. Note inlet to stilling well kept clean in order to get record of runoff. Figur e 27. Rock and soil moved in July 10, 1949 storm Figure 28• Sediment carried into debris basin by storm of July 10, 1949* 71 1950. Oa A p r i l 8 of t h is j s a r th# d s b r is b a s in was again c le a n e d shew ing s o i l lo s s s s from July 13, 1949 to April 1 0 , 1950 u 42,330 pounds or 150 pounds p a r day. Ths c u l t i v a t i o n o t oars 1a th s f i e l d was dlssontinuad. Th# o r i g i n a l o b je c t iv e w as t o s ro p t h i s s to o p laAd in seoo rdaaoo w it h p r e v a ilin g m o u n ta in a g r i c u l t u r a l p r a c t ic e s , 7■ 1 T continuing t h e 4e ;rep o ln g u n t i l y ie ld s w e re b e lo w those on other e r o m p d f i e l d s i n t h i s s o o tio a . . In l i s a oT c u l t i v a t i o n aad c ro p p in g , th e a r e a , along with the a p p e r a h d lO w e r p a s tu r e s , was g ra h sd u n d e r a rotatiaonal g ro w in g s p s to p *^ B e fo re g r a z in g s t a r t e d oa the abandoned fls & ft a s u r fa c e i n f i l t r a t i o n der test) whs i n i t i a t e d s tu d y (1 0 iaoh cylin­ s o a r th s e n t i r e c u lt iv a t e d area and this was re p e a le d a f t e r each p e rio d o f r o t a t io n a l grazing. The t o t ^ j r u i a g use o f th S w a te rs h e d f o r 1950 consisted of ■*> ■ - -a 881 an1an1 The c a t t l e w ere p e r io d ic a lly rotated be- tween th o p e m n n e a t pas t a r e s a a d t h e o ld c o r n f ie ld . F ig a ro 29 p ic t a r c s one o f th e g a l l i c s i a th e abandoned o o m f l e l d a lo n g w it h th o v o lu n te e r v e g e ta tio n 4 a August, 1 9 5 0 . 1951. On April 11 the debris basin was eleaned and yielded 2 1 ,3 6 1 pounds of sediment. This is an aoouaulation of 85 pounds per day for the period beginning April 8 , 1950* Both pastures aad the abandoned e o m f i e l d were grazed on a rotational grazing system. A vegetation surrey was made of the entire watershed on August 6. During August, the shrubs, weeds and tree sprouts Figure 29Gully in abandoned cornfield, August 1950. 73 were out book, ataoked and burnad in an effort to Improve the paatura conditions. Tho Maximum flood peak for the aeaaon to September 1 ooourred on July 15 when a peak of 126 o.s.m. was reoorded following a brief, intenae thunder a t o m . Overland flow waa obaerred in the abandoned eornfield and particularly in the lower portion of the lower paature during several atorma produelng flowa of leaa than 5 o.a.m* F ig u re 30 ahowa th e o o n tr o l p lo t a e a ta b lia h e d in th e abandoned e o r n f ie ld aa w e ll aa a te o k a o f bruah o u t o f f th e o ld e o r n f ie ld . Some o f th e tr e e a in th e r i g h t fo re g ro u n d had a tta in e d h e ig h ta o f seven f e e t a in e e th e f i e l d had been abandoned i n 1949* Infiltration ring atudiea were eontinued on the old eornfield aa well aa on the other aegmenta of the waterahed. Figure 30. Fenced control plots in abandoned cornfield, August 1951 • Note stacks of brush out in an attempt to improve the grass cover. i CHANGES IN SOME BIOLOGIC AND EDAFHIC CHARACTERISTICS OF THE WATERSHED AS A RESULT OF LAND USE Vegetation Changes To observe the changes in the vegetal cover on the water­ shed brought about by forest cutting and mountain farming treatment, a vegetative survey was made on August 6, 1951. Due to the heterogeneous mixing of the species through the area, no attempt was made to oonstruot a type map. In lieu of a type map, the vegetation was observed and identified and its relative abundance was noted according to the differ­ ent land use elements. Figure 31 indicates the different land uses in 1951* Coppice Forest. forest are noted. In Figure 31 two areas of coppice Coppice forest wA n was out In 1939-1940 and has not been out back or sprouted. Coppice forest "B" was cut at the same time but was sprouted back in 1941. The forest and herbaceous cover of the two coppice areas was so similar in size, density and species composition that they were treated as one. Since a large portion of the tree stems were of sprout origin the area was classified as coppioe forest. Except along the ridges, the forest cover was so dense that a ground cover of shrubby and herbaceous vegetation was almost exoluded. Along the ridges and In a few openings, greenbrier and wild grape Ttgarm 31• Land Oa* Coaplaxaa, 1991 Littla Hurrioana Vatarahad Oaaaata Hjrdvologle Laboratory w !; !w tUm~Us* M k* Ctmce *aa*sr 77 w*re still abundant* On the slopes, however, these speeies were muoh less oowmon although plant remains indioated that they had been quite abundant there until a few years before. As indieated above, speeies were so mixed that no attempt was made to make a type map. All speeies listed below with the exoeptioa of blackjack oak and eastern hem­ lock ooourred throughout the two eoppiee areas. Blackjack oak wras found onlj on the ridges and one hemlook was found just above the eon tour trail in eoppiee area "B*. Along the ridges, piteh pine, white oak and chinquapin were more common and tulip poplar less abundant than on the lower slopes• Rhododendron and mountain laurel occurred rather commonly near the ridges and to a lesser degree on the slopes. In the adjacent undisturbed lb rest areas a moderate understory of these speeies ooeurs. Rhododendron and laurel appear to be slow in re-establishing themselves. The dmalnant trees were approximately 3 to 3*3 inches in d.b.h. and 12-15 feet in height. The forest tree speeies as well as the shrubby and herbaceous vegetation observed along with their relative abundanee are listed below: Forest Tree S u e d e s '* Tulip poplar Dogwood Liriodendron tuliplfera L. Oornus florida L. Abundant 3411 scientific names from gray's Manual of Botany, Eighth Edition, edited by M. L. Fernald, American Book Company, N. T. 1632 pp. 1950* 7* Sweet hiekory Chestnut Red Maple Northern red oak Blank loonst Pltoh pins Persimmon Sassafras Sourwood Chestnut oak Blaekjaek oak Snarlst oak Blank oak Chinquapin Tallow blroh Bittnrnut hiekory V114 plan Hawthorn or thomapplo White oak Staghorn sumae Spleebush Bastnrn henloek Carya glabra Mill. Castansa dontata (Marsh.) Borkh. Aoer rubrum tridnns Wood, Quorous rubra Tar. borealis (Mlehx.f) Jarw. Robinia psnudoaeanla L. Pinus rig Ida Mill. ““ Dlospyros virginiana L. Common Sassafras albldum (NuTt.) Hens. Oxydendrum arbomun CL.JPCl quareus prinus L. Qua roue marylanXiea Mueneh. Qua reus sossinsa Muanek. quareus yelutina tan. Castansa puaila (L.) Mill. Moderately Cm— on Betala lutas Mlehx.f. Carya ewySifamie nreiig-) t Prunus anariaana Marsh. Crataegus sp. L. Quereus alba LT Rhus typhina E. Benzoin aostiials (L.) Naas. Tsuga sanadensis (L.)Carr7 Rare Shrubby and Harbaaaous Tanatation Blaakbarry Highbush blueberry Mountain laurel Rhododendron Wild grape Greenbrier Rattlesnake weed Wood Sorrel Maidenhair fern Braeken fern Trillium Star flower Ox-eye daisy Daisy flsabane Tellewfringed orchis Lopseed Wood violet Rubus sp. L. Common Vaeoinium atrooeum (Gray) Heller Kalmla latifolla L. Rhododendron maximum L. Yitis bieolor Le Conte. Smllax glauea Walt. Hieraoium vanosum L. Moderately Conaon Oxalis striata L.q” Adlan tun pedatum L. Pteris aquirina IT Trillium ereetum""!.. Trisntalis borealTs L. Chrysanthemum leueanThsmnm var. pinnatifidum Loeoe and Lamotts Erigeron sp. L. Habenaria eilTaris (L.) R. Br. Rare Phryma leptostaeha L. Viola papillonaoeao~"’ Pursh. 79 Pastures,. The vegetation of the upper and lower pas­ tures and of the old oornfleld is quite similar. The vege­ tal eover observed, along with its relative abundance in both the abandoned cornfield and the two pastures is listed below. There is no apparent relationship to the vegetative types mapped in 1940. Nearly all speeies listed oeeur through­ out the three areas. Tree seedlings and sprouts observed in 1951 which were not indicated in the 1940 survey inolude hawthorn, swamp willow, yellow oak, blaokjaok oak, green ash, spioebush, yellow biroh, blaok walnut and butternut. The black walnut and butternut were probably carried into the area by squirrels or groundhogs. All species observed in 1940, however, were present on the area in 1951* A muoh more marked change took place in the shrubby fluid herbaoeous vegetation. Of those species observed in 1940 only nine were found in 1951: switch grass, broad­ leaved plantain, Joe-pye weed, ragweed, wood sorrel, spike rush, green bulrush, wool grass and yellow foxtail. On the other hfluid, a total of thirty-three speoles were identified which were not listed as present in 1940. It is signifi­ cant, too, that many of these invading speoles, such as yarrow, smartweed, mullein, Cfluaada thistle, nettle and purslane are usually assoelated with land abuse. ihanflowd Cornfield. As indicated previously, the species eceqposition of the abcuidoned oornfleld is virtually i i 80 the same as that of ths two pasturss even though it had been only two years sinos the area was cultivated. The vegetation density of the abandoned cornfield as well as of the two pasture areas varies from practically zero on several of the "scald" areas to about thirty to forty percent ground eover on the best areas. The two control plots indicated in Figure 31 were fenoed off in 1949 following the harvest of the last c o m crop. Cattle have been excluded from, these two plots and they have not been sprouted bach. The vegetation here is nearly the same as that of the cornfield in speoles compo­ sition. In the lower plot (Ho. 1) blackberry and wild strawberry are the two most abundant speoles. The ground eover in this plot is nearly oomplete but in spots consists solely of wild strawberry. In the upper plot (Ho. 2) the same speoles are represented but the ground cover varies from 65 to 80 peroent. It is apparent that in both plots tree species will soon take over. Numerous stems, particu­ larly in the lower plot, are over 6 feet in height. Shrubby H erbaceous Vegetation Species Abundance* Co— fiw Vmmm_________ Scientific Name_____ Cornfield Pastures Blackberry Mullein *l*gen& VA A MC Bubus sp. L. Yerbasoum thapsus I. for speoles abundance: - Very abundant — Abundant - Moderately oommon VA A C - Co m on R — Rare Abs - Absent VA C 61 Yarrow Strawberry Wild grape Achillas milliforlum L. c Pragarla Virginians Duchesne. YA Yitis bioolor La Conte* "" c C A C Horsemlnt Crab grass Switch grass Rad baneberry Canada thistle Monarda punctata L. Digltaria sp* Heist. Panloum sp. L. Actaaa rubra"(Ait.) Wllld. Clrcium arvense (L.) Sc o p . C A MC Aba Abs c C C R R Greenbrier Broad-leaved plantain Ox-eye daisy Sallax glauca Walt. A A C C PoJceberry Boneset Plantago major L. Chrysanthemum leucanthumum ▼ar. pinnatifidum Leooa A Lamotte. Phytolaeoa decandra L. Supatorlum perfollatum L. MC A C C C C Leafcup Gentian Star grass Selfheal Tick trefoil Polymnla uvedalla L. C C Sabatia angularis T l . ) Pur ah. C MC Hypoxis hirsuta (L.TCovllle. Common along stream Prunella vulgaris L.2 C MC Desmodlum panioulaTum (L.)DC. C C Joe-pye weed Soar tweed Ragweed Horse nettle White clover Supatorlum purpureum L. Persioaria hydropoper"~L. Ambrosia artemisllfolia L. Solanum oarolinenseL. “ Trlfollum repend L. MC C A C MC MC MC C C Abs St. Johns wart Beggars lice False pennyroyal Evening primrose Wood sorrel Pyperleum perforatum L. C Lappula Virginians (L.)Greene C Isanthus braohiatus TL. C Oenothera biennis L. A Oxalis striota L. " MC MC MC MC MC R Lopseed Spite rush Green bulrush Wool grass Yellow foxtail Mountain laurel Agrimony Purslane Yellow-fringed orchis Partridge pea Phryma leptostachya L. MC R Sleocharis ob tusa " (Wllld.) Schultes Abundant along stream Soirpus sp. L. Common along stream Scirpus oyperinus (L. )Eunth. Common along stream Setaria lutesoens Weigel MC MC Ealmla latifolia L. Agrimonia sp. L. Portulaea oleraeea L. Abs MC MC MC R MC Habenarla oillaris (L.)R.Br. Abs Cassia faseloulata Mlohx. YA MC YA 82 Forest Tree Seedl Inga and Sprouts Speeies Co— on Name Black locust Staghorn sumao Tulip poplar Persimmon Sassafras Scientific Name Abundanoe Cornfield Pastures Robinia pseudoacacia L. Rhus typhina L. ~ Liriodsndron Tulipifera L. Diospjrros Tlrginlana L. ”* Sassafras albidum (NuTt.) Hass* Dwarf sumac Rhus ooppalina L. northern red oak Querous rubra Tar. borealis (Mich, f.) Hess. Hawthorn Crataegus sp. L. Dogwood Cornus florid a""L. Sweet Hiekory Carya glabra MlTl. Green ash YA YA YA A A A YA YA C A C G A MC A C C MC A MC Fraxinua pennsylTanica far. subintegerrima (Vahl.) Fern. MC R Salix sp. L. Along stream channels Pinus rigida Mill. MC C Swamp willow Pitch pine Chestnut or yellow oaik Spieebush Querous prinus L. Benzoin aestiraTe (L.) Hees. Chestnut Yellow biroh Red maple Black walnut Butternut Castansa dentata (Marsh. )Borkh. Be tula lutes MichxTTf1. Acer rubra trldens Wood. Jugl&na nigra L. Juglana clnera”!.. C MC MC R R MC Abs MC R R Soarlet oak White oak Blackjack oak Blaok oak Onerous Quereus Ouereus $uercus MC R MC R MC R MC R eoeoinea Muench. alba L. mary land lea Muench. ▼elutina Lam. MC MC Trout Habitat As a result of the mountain farming treatment, the Little Hurricane Branch was changed considerably as a trout stream. Trout are known to be able to withstand a wide range of acidity, alkalinity and earbon dioxide tension. R MC 83 They are, however, quite eeneitiTe to changes in turbidity, sedimentation end stream temperatures. Turbidity teats comparing the water e f the Little Hurrieaae Branch with that of Bee Branoh, from a forested watershed, were made from 1946 to 1950. The results of these tests are shown in Table X on the following page whieh shows the average monthly turbidity values in parts per million for the two streams. The 17. S. Publio Health Ser- viee Standard for turbidity for potable water is ten parts per million. From the results of this study it is apparent that the waters of the Little Hurrieaae Branoh fall below this standard from April to Oetober inclusive. The increased sediment load carried by the Little Hurricane Branch following forest cutting and treatment is shown in Table XXX. Trout are especially sensitive to thermal fluctuations in their environment, particularly when these changes occur near their upper limit of tolerance. and Heedham*s (23) wXdeal Temperature Limits for Trout*1 are given below. TABIX XX IDEAL AMD MAXIMUM TEMPERATURES FOR TROUT ________________________ LAFTER HBgDHAM) _ _ Ideal Tern- Maximum TemSpeeies perature in perature in _____________________________________________ degrees F.degrees F. Rainbow (Salmo gairdnerii) Eastern Brook (Salvelinus fontinalis) Brown or Loch Levan (Salmo trutta) 70-80 83 66 75 70-80 81 TABU I COMPARISON OP TURBIDITT VALCIKS IN PARIS H R MILLION 1RQM A FOREST AND A MOUNTAIN FARM STRXAM Covoota N o * 3 - Littlo Hurrloaao Branch, Mountain Farm Stria* loar Jan. Fob. Marob 5 14 6 5 6 4 8 8 5 4 Aforago 8 5 6 1947 1948 1949 1950 Aar. May Jiao July Aug. Sopt. Oot. N ot 9 18 11 11 17 28 10 23 16 26 16 20 23 22 13 13 11 12 20 17 10 14 11 12 10 6 12 19 19 22 12 16 12 9 2 2 2 Covoota No. 34 - Boo Branoh, Foroot Stroaa 1946 1947 1948 1949 1950 2 3 4 1 3 4 3 2 3 3 5 1 3 5 7 3 5 10 Aforago 3 3 3 4 6 6 5 5 5 6 8 8 7 2 5 5 4 3 16 6 3 4 3 4 5 5 7 4 7 3 4 3 85 During 1948-1949> a stream tempsrature study vaa eondueted by Greene (16) at the Come eta Hydrologlo Laboratory in whloh the Little Hurrieane Branoh was compared with a stream from a forested watershed. The results of this study are summarized in Figure 32* Prior to the treatment of the watershed, loeal inhabi­ tants had observed the presenoe of trout in tho stream. In 195L* however, no trout were seen by the writer, nor were there any reports of trout having boon observed by personnel of the laboratory within the past few years. It appears evident then that the ehanges in sedimenta­ tion, stream turbidity and stream temperature brought about by the cutting of the forest and the subsequent mountain farming treatment, have virtually rendered the Little Hurri­ eane Branch barren as a habitat for trout. Xdaphie Changes Many factors, such as climate, vegetation density, type of vegetation, slope, geologlo substrata, land use practices and the physical oharaeteristles of the soil combine to de­ termine the stream flow oharaeteristles for a given water­ shed. The amount of preoipitatlon that goes into etreon­ flow is to a large degree determined by the physical proper­ ties of the soil. The oharaeteristles of the plow layer of the soil is strongly influenced by the vegetation it supports. For this reason, any practices which change the nature of 40 HAY JUN Figure 32. Weekly maximum temoerr.tures for the farm and forest stream, (after Greene) OCT NOY *7 the vegetation may Ia turn bring aboat ohanges Ia the phyeleal eharaeteriatlea of the eurfaee soil aad consequently Ia stresmflow. Aa a resalt of the treatment applied to the watershed, the natural equilibrium among the vegetation, the soil and surface runoff would necessarily have suffered dyaamie ohaages. Consequently, a study of the changes in surfsee runoff oharaeteristles would be incomplete without an in­ vestigation of the changes in the physical oharaeteristles of the surfsee soil. One of the most obvious expressions of soil changes is in the degree of soil erosion or in the determination of soil losses following the treatment of the watershed. Measurements of the soil losses from the watershed have been made since the initiation of the projeet and are summarised in this section. Another measure used to determine gross changes in soil oharaeteristles, partieularly as they affeet water relations, is that of ehanges in their infiltration rates. Exploratory tests on infiltration were made using the oylinder ring test method in 1949 and 1950 and are suauaarlxed in the following pages. During the summer of 1951 a large seale infiltration study of the entire watershed was initia­ ted. The results of this investigation, however, will not bo available for several seasons. 88 In orAtr to determine more detailed ehtages In tiie phy­ sical oharaeteristles or the surfaoe soil, soil oore and saek samples were eolleoted for analysis. Cylindrical oores three Inches In diameter by three Inches long and approximately one-pint seek samples were oolleoted from the 0-3 and 3-6 lneh layers from six sites. Five samples were eolleoted from eaoh of the following sites from both layers; tiAdistarbed forest (from eontrol plots In adjacent water­ shed, same soil type), oopploe forest, upper pasture, lower pasture, abandoned oornfleld and eontrol plots within the old oornfleld. The sampling plots were randomly seleeted from numbered grid oross-seetlons except from the eontrol plots In both the undisturbed forest aad the abandoned oornfleld. The latter were stratified at right angles to the eontoor. From the saok samples meohanloal analyses and organic matter determinations were made. The oore samples were used to measure permeability, oaplllary, non-oapiliary, total porosity, volume weight and aggregate analyses. The results of these various determinations are presented In the following pages. Soli L o s s e s . Am Indiested previously, soil losses since 1941 have been measured using the specially designed debris basin. Prior to the installation of the debris basin, soil losses were measured from deposits Ia the silting basin. 89 In M aking these measurements, the v a t c r is allowed to paaa over the w a i r blada before it ia diverted into a trough which by-passes the dabria basins, tJh.ua providing a oontinuoua raoord of atraam discharge. Samplaa of two hun- drad oubio oantlmatara of aadimant ara taken froai each baffla w i t h i n each of tha thraa debris basins after tha sediments have dried. Following volume me a sur aments in aaeh b a f f l a , tha debris is removed. Tha total soil lasses ara than computed from tha dry weight of tha samples and from tha volume measurements applying to aaeh division of tha basin. Tha total soil losses measured from tha inception of tha experiment on July 3, 1934 to tha measurement made an April 11, 1951 nre summarised In Table III. It is apparent that soil losses have increased tremendously following the treatment of the watershed. The actual increase amounts to over 12 times as much soil loss per acre per year. Be­ fore the Installation of the debris basin (watershed in forest cover until the w inter of 1939-1940) the total soil loss amounted to 1,081 pounds per acre er 134 pounds per acre per year. A portion of this amount can be attributed to the treatment since the debris basin was not installed until nearly two years after the el ear out ting of the water­ shed w as started. Since the Installation of the debris basin, an average loss of over 1,900 pounds or nearly 1 ton per acre per year has been measured. The maximum soil Table 3. t£ Summary of Soil Losses. Period No. days Treatment 7 - 3 - 3 L /8 - 2 7 - U 1 2557 Forest cover and initial treatment 8 -2 3 -U 1 /5 -U -U 3 5 - 5 -l* 3 /9 - 3 - l* 3 61*1* 126 C o m * pasture C o m * pasture 9 - 9 -U 3 A -1 6 - U 6 l*-1 7 -l» 6 /3 -2 8 -l* 7 3 - 2 9 -U 7 A -1 3 -U 8 l* - ll* - l* 8 / l*-8 -l*9 l * - 9 - l * 9 / 7 - l l -1*9 118$ 31*5 381 359 93 Fallow, pasture Com* Com* Com* Com, pasture pasture pasture pasture Total loss dry weight in pounds Average loss pounds per day Average loss pounds per acre per year 21*,637 9 .6 1 5 3 .7 1 3 *9 2 8 7 9 ,0 5 8 Average 2 1 .6 627.1* 1 1 8 .1 31*5.7 1 0 ,0 1 *3 .9 1 ,8 9 0 .6 1*1**085 3 7 .2 5 9 5 .5 1 3 *5 0 7 1 6 .1 8 6 3 2 ,8 7 9 1 8 5 ,8 7 5 3 9 .2 1*2.5 9 1 .6 1 ,9 9 8 .6 2 1 0 .9 6 2 7 .5 6 8 0 .0 1,1*66.1* 3 1 ,9 9 5 .1 3 ,3 7 7 .7 1 5 5 .6 8 5 .5 1 0 0 .0 2 ,1 *9 1 .9 9 3 6 .5 1 ,6 0 0 .8 Average 7 -1 2 -L 9 A -1 0 -5 0 I4 -1 1 -5 0 /U -1 1 -5 1 272 365 pasture Pasture 1*2,330 2 1 ,3 6 1 Average Total soil loss 8-2 8-1 *1 to ii- 1 1 - 5 1 Average loss 8-28-1*1 to l * - l l - 5 l 1*1*9,209 pounds 1 *9 3 3 pounds per acre per year 91 loss was measured In 1949 when preolpltatlon ranged wall above average. For th.# 93 day period from April 9 to July 11, a total loss of 185>*75 pouads or a n average of 1*998*6 pounds, oas ton, per day was measured for th« en­ tire area. As iadieated previously, am estimated 76 tons of soil and debris earns off the watershed in a period of 30 to 40 minutes during the storm of July 10, 1949 wh o n a reeord peat of 1,850 e.s.m. w as reeorded. Figures 24* 25 and 26 indieate the magnitude of soil losses resulting from this storm* These data indieate that the greatest soil losses were sustained while the eornfield was under cultivation* From September, 1941 to September, 1943 the average soil loss amounted to nearly 1 ton per aere per year* Follow­ ing this period, the eornfield wa s permitted to lie idle until the 1946 season. The pastures, however, were grazed* A deerease to approximately 600 pounds per aere w a s noted for this period. oultlvated. per year. From 1946 to 1949 the oornfleld was again Soil losses mounted to over 2& tons per acre After the 1949 season the eornfield was eoaver­ ted to pasture. A deerease in soil losses was noted from August*1949 to April, 1951 to 1,600 pounds per aere per year. If the losses from the eornfield and the lower pasture eould be analyzed separately it undoubtedly would be found that by far the greatest pereentage of the total loss was 92 contributed by these two areas. Evidence indicating this is found in the numerous scalds in the cornfield and the lower pasture in whioh the topsoll has been completely removed exposing the red subsoil. The analysis above also offers further proof. On several ooeasions during the 1951 season, storms which produced peak discharges of less than 5 o.s.m. were observed to produee turbid overland flow in the abandoned eornfield and in the lower pasture. Xn ten years of field observations during and following storms, overland flow or evidence of overland flow has never been observed in the ooppioe forest area. Infiltration. A surface infiltration study made on the watershed in July, 1949 demonstrated how the small, heavily trampled area of the lower pasture might be the source of more than half of the total storm runoff. To measure the infiltration rate a steel oyllnder ten inches in diameter and approximately six inches in height was driven into the ground for a depth of several inches. Two area inches of water were poured into the cylinder and the time required for the water to disappear from the soil surface was noted. The values thus derived were converted to inches per hour infiltration. Figure 33 shows the equip­ ment used to determine the infiltration rate in this manner. The results of this study are indloated as follows: 93 Infiltration Rates* Land use Average rate of Infiltra tlon In inches per hour Channel area Lower pasture Upper pasture Cornfield Capplee forest •0 0.56 3.00 4*00 6*00 Before erasing started on the abandoned oornflald In 195® n similar Infiltration study was made over the entire area and was repeated after each period of rotational graslng. Bata In the following table show that It does not require many animal use days to deerease the ability of the soil to talcs in water* Changes In Infiltration Rates Following Qraz**»* Rata la Xnohss per hour 3*02 1.35 0.62 Period of Use Before grazing After 13 animal use per After 30 animal use per days sere days sere A similar surfaee Infiltration study was Initiated in 1951* The entire watershed was gridded and monthly Infil­ tration tests were made during the summer months on more than one hundred plots* These observations are to be con­ tinued for several seasons and the data are to be used as the basis for a special Investigation on Infiltration. unpublished data from the files of the Coweeta Hydrologle Laboratory Figora 33 • Equipment uaad in malting infiltration maasuramants. 95 Mtoiiaflltal i f l t l w l i . Tii« objeotive of a mechanical analysis Is to determine the aizo distribution o f the indi­ vidual particle* within tiia soil. preoood by soil texture. These roaulta m a y bo oz- Soil texture, other things being equal, influences tho amount of aurfaee area of the eoil partioles, whloh in turn affects the water holding oapaoity of a soil. Retention storage, or water held in the oapillary pores of nULneral soils against the pull of gravity, is greater in silts and elays sines the surface area exposed to water is many times greater than in sands. On the other hand, a high eon tent of sand frequently provides greater opportunity for the development of large non-oaplllary pores and thus increases detention storage. The American Society of Civil Snglneors (1) indicates that clay can hold 9 times as much water as fine sand against the force of gravity, but this varies greatly w i t h the hind of clay. The reten­ tion storage capacity in inches depth of water per foot depth of soil for fine sand is given as 0.5 as compared with 2*5 for silt loam and 4*5 for clay. Xt is apparent that the mechanical composition of soils and consequently soil texture is of hydrologle significance. In preparing the samples for the mechanical analysis, the larger clods a n d aggregates were broken down and the entire sample placed on a 2 mm. sieve. The weights of the material w h i c h remained on the sieve after shaking, largely pebbles, rock fragments and ooneretlons were noted. From these weights and fr cm the total weight of the sample, per- oentages were oaloulated and are shown in column 2 of Table IV. T o determine the meehanieal composition of the portion of the samples passing through the 2 mm. si ere, the Bouyouoos (9) hydrometer method of meohanieal analysis was employed. Table IV summarizes the results of these analyses. From these data it appears that the treatment applied to the watershed has not effected any large scale changes in the meehanieal composition of the less than 2 mm. fraction of the soil. There is little variation indicated in either soil layer among the average values for each site. The undisturbed forest and the oopploe forest show the highest content of sand and the oopploe forest has the lowest values for fine d a y . The pastured areas are lowest in sand con­ tent and highest in fine d a y . found in the fine d a y The greatest variation is content in the 3-6 inch layer. The highest value of 24.5 percent was found in the lower pasture samples a n d the lowest of 10.7 percent in the oopploe forest, a difference of 13*8 percent. In view of the dls- oussion above the two forested areas should shew higher detention storage a n d lower retention storage than the pastured areas and the cornfield should be intermediate be­ tween these extremes. However, if the materials greater than two millimeters have an influence similar to sand then the undisturbed forest might be in a less favorable situation slnoe it TABIZ 17 SUMMARY 07 RESULTS 07 MECHANICAL ANALYSES Percent of Soil layer and sample land uao greater than 2 mm. Sum of sand and eoarsar fractions In pareant Mechanical composition of less than 2 m , fraction Peroent Percent Peroent Feroent sand silt clay fine day 0-3 lnoh layer Uadlaturbad foraat Coppice foraat Upper pasture Lower pasture Cornfield Control plots 3*6 9.3 7.7 7.3 8,6 7.3 81,7 91.4 99.6 97.7 99.1 91.0 66.2 ;( 7.3 64.2 . I 63.5;£ 4.8 61.8 , £10.8 64.0 I 5*9 62.4 2 5.2 18.0 £ 2.9 23.0 2.4 6.0 15.3 16.9 4.7 21.4 2 2.5 20.4 2.3 3.7 £ 3.0 3.7 4.7 1.7 2.8 1.3 1.9 3.8 5.2 1.7 1.9 12.1 4 4.3 9.8 i 3.6 17.5i 7.9 16.6 % 7.7 12*9 15 5.2 14.4 1 3.7 £ 4.5 82.8 97.4 95.8 90.4 95.6 89.9 62.5 j£ 9.3 64.2 ,I 3*1 56.0 ■ £ 1.9 55.6 ;£ 6.4 61.9 £ 3.9 59.1'£10.5 17.5 2 2.9 19.0 5.2 17.0 3.8 17.6 7 3.1 22.0 4.2 18.8 3.7 3.0 £ 2.3 6.1 4.3 3.6 1.2 2.3 7 1.1 2.2 1.5 4.1 4.7 17.0 1 7.9 10.7 1? 1.6 23.4 1I 5.1 24.5 1J 6.4 13.9]I 4.5 18.0 1? 6.9 15*5 £ 27 .2 2 36.1 35*9 £ 35*1 28.6 7 7 2 2 2 7 2 2 2 7 7 2 3-6 lnoh layer Undisturbed forest Coppiee forest Upper pasture Lover pasture Cornfield Control plots 20.3 33*2 39*6 34.8 33*7 30*8 7 9.1 5.7 2 8.6 7 3.1 212.8 7 2 2 2 2 2 2 2 2 93 contains muoh lass of this coarser fraction than any of the other plots as shown in oolumn 3 of Table XT. The undis­ turbed forest samples show the lowest values for the sum of the sand and material larger than 2 mm. for both soil layers. In the surface layer the highest values are indi­ cated for the upper pasture and the cornfleldv 99*6 and 99*1 respectively, as eompared with 81.7 percent for the undisturbed forest. In the 3-6 inch layer the highest values were found for the coppice forest, upper pasture and the abandoned cornfield. According to the United States Department of Agricul­ ture, textural classification, the samples from the undis­ turbed forest, oopploe forest and the cornfield fall into the sandy loam class. The upper and lower pasture samples are sandy clay loams while the oontrol plots show sandy loam in the surface samples and sandy clay loam in the sub­ surface layer. The small degree of variation noted in the mechanical analyses would appear to substantiate the classification of the fine portion of the surface soils of the watershed into one textural class, loam (according to more recent standards, this would be classified as sandy loam). Aggregate Analyses. One of the primary objectives of an aggregate analysis is the determination of the extent to which the finer mechanical separates are aggregated into coarser fractions. An aggregate analysis thus provides a measure of soil structure. Soll-water relations and aera­ tion conditions ara b o t h strongly Influenced by soil struc­ ture • For example, porosity, a i r eapaelty, water bolding capacity, volume wslgbt and p s m s a b l l l t y to w a t e r and air are Influenced by soil structure. Tbe total percentage of aggregates or "state of aggregation" as suggested by Barer (7), gives a good Indication of tbe arodibility of soils. If, for example, tbe state of aggregation Is blgb, 1. s., tbe soli eontalns a blgb eontent of water-stable aggregates, susceptibility to erosion Is considerably lower tban In tbe ease of low aggregate stability. In tbe latter ease tbere Is little binding together of tbe partieles Into granules, consequently falling raindrops and surfaee flo w tend to disperse tbe soil. Under sueb condi­ tions tbe soil takes up w a t e r more slowly and Is obviously blgbly erosive. Tbese two eondltlons are blgbly unfavorable from a bydrologlo viewpoint, botb from tbe standpoint of flood control and w a t e r s h e d management for Inereased wa t e r yields• In M t f i e g aggregate analyses of tbe samples used In tbls study Yoder's employed. (37) dunicer or wet-slevlng m e t h o d w a s Tbe fire oven-dry oore samples from each site and eaeb layer w e r e mixed an d two eomposlte samples extrac­ ted. Aggregate analyses were then made on tbe eomposlte samples. A suumary of tbe results Is presented In Table V. Y a n a t i o n s w i t h i n the samples were very small except for tbe 100 surfaoe a oil samples representing the eopploe Tor eat and the cornfield. The results g i v e n in Table V Indioate that the treat­ ment of the w a terahed haa effected a marked change in the stability of aoll aggregates and consequently, aoll structure. Xn both the aurfaoe and aub-aurfaee layer a, bat par- tloularly in the a u r f a e e , the undlaturbed foraat ahoira a coneiderably higher content of w a t e r atable aggregatea greater tha n 2 mm. in diameter than all other aitea exeept the eopploe forest. Ev e n here in the aurfaoe la y e r a differenee o f o v e r 1 0 peroent exlata. Theae differeneea are particularly significant w h e n one oonaidera that the un­ dlaturbed forest contained much smaller quantities of par­ ticles greater than 2 mm. in diameter than the other sites studied. Theae values are somewhat lower in the 3 - 6 inch layer than in the 0-3 inch layers of the forested areas and the upper pasture, however, t h e y do not show so m u c h varia­ tion in the ease of the upper pasture as in the forested areas. The differences between the two layers in the control plots, the cornfield and the lower pasture are not marked, probably due to the fact that these areas have been culti­ vated. A n analysis of the degree of aggregation in the fine earth material is presented in column 3 of Table VI. The undisturbed forest shows an even higher aggregation here than in the oase of total w a t e r stable aggregates. In the TABU V STOttAHT OF THE RESULTS OF AGGREGATE AIALTSES Slit Ataraga Paroant of Aggrsgatas and Partide a bj Siza Classss (in milllastara) .105 .25-.5 Orar A. -.25 1.-2. .5-1. 2.-A. .10} Surfaoa layar. 0-3 iaahas Uadisturbsd foraat Cappias foraat Uppar pastors Lovar paatura Carafiald Control plots 85.52 73.20 64.10 53.92 59.00 46.24 7.76 6.30 10*50 12.72 12.22 13.76 3.26 4.96 7.00 9.92 7.00 9.36 1.40 3.84 5.66 8.16 6.26 8.96 1.36 2.48 4.40 7.18 5.62 6.80 1.12 2.54 2.36 3*86 3.20 5.64 2.56 4.68 5.98 4.24 6.70 7.24 6.68 11.50 12.30 11.68 12.48 16.24 3.74 6.76 10.98 7.90 6.26 12.32 3.76 5.68 9.04 7.76 6.48 11.50 3.94 4.72 7.54 7.92 4.56 10.08 2.64 2.10 3.90 4.26 2.62 4.72 5.40 5.00 7.66 3.76 3.44 5.08 Uadisturbad foraat Goppioa foraat Uppar paatura Lowsr paatura Cornfisld Control plots 73.84 63.54 48.58 56.72 62.16 40.06 TOT Sub-surfaoa layar. 3-6 lasbss 102 0-3 ineh layer the undisturbed forest shoes ov e r 25 pereeut higher aggregation than all other sites exeept the eoppiee forest where the difference is approximately 16 peroent. In the 3-6 inoh layer only the undisturbed forest is markedly better aggregated than all the other areas. The relative trends for both layers are quite similar to the values for total wa t e r stable aggregates over 4 ms* in diameter* From field observations the greatest part of the soil loss from the watershed is obviously from the cornfield and the lower pasture. The changes in aggregation or struc­ ture of the surfaoe soil (0-3 lnoh layer) have no doubt contributed to these differences in soil losses. In the differences between the two pastures this is probably the major factor. The marked changes in soil structure as indicated by aggregation contribute materially to the changes in permeabil­ ity and consequently to runoff characteristics which are noted later in this seetlon. Soil Orff*"** m a tter. The presence of a high content of organic matter in a soil has considerable significance from a hydrologic view. A high content of organio matter has a marked influence on the storage oapaelty that a soil has for water* Detention storage, 1. e«, water detained temporarily in the large non-oapillary pores, is increased by the TABU VI BXOBXX 0? AGGBJGATION OF FIDE XABTH HATJBIAL Soil layer and lead uae Aggregates over 2 m , in peroent (whole soil basis) Priwary parti­ cles oyer 2 an. (whole soil basis) pereeat Aggregates of Aggregates of fine earth over fine earth as pereeat of fine 2 an. (whole soil basin) earth pereeat pereeat (fine earth basis) 0-3 i&oh layer Uadi•turbed foreat Coppioe foreat Upper paeture Lower paatura Cornfield Control plots 93.3 81.5 74.6 66*6 71.2 60.0 13.4 21.4 26.5 26.4 26.0 22.2 79.9 60.1 48.1 40.2 45.2 37.8 92.3 76.5 65.4 54.6 61.1 48.6 80.5 75.0 60.9 68.4 74.6 56.3 16.9 24.9 28.4 25.8 25.2 23.5 63.6 50.1 32.5 42.6 49.4 32.8 76.5 66.8 45*4 57.4 66.0 42.9 3-6 inoli layer Undisturbed forest Coppioe forest Upper pasture Lower pasture Cornfield Control plots 104 inoluaion of organic matter beoauae of it a influenoe on aoll struoture. Decaying roota and greater biological activity alao reault in the formation of the large hydraulic path­ wa y a which channel water through the aoll profile and even­ tually to ground-water • Similarly, retention atorage, that water retained or held in the aoll and made available for plant growth, ia uaually increaaed through the incorporation of additional organic matter. tlve capaolty. Organic matter hae a high moiature-adaorpIn the colloidal atate it takea up aa much as 4*4 tlmea its own weight of water. When decompoaed and mixed w i t h the aoll, it coata the aoil particle a with a gel-like, porous and highly adsorptive substance. Clinging to the particles, this material, in effect, increases their aurfaoe areas and thus their storage capaolty (22). Organic matter, as mentioned above, affects the pro­ perties of the aoll wh i c h in turn exert their influence on the storage and transmission of water. Organic matter aids alao in the formation and maintenance of water-stable aggregates thus reducing soil dispersion and consequent erosion. In this study, a determination w a s made of the organic matter oontained in the samples oollected from the watershed. T m making these determinations, the dry combustion msthod patterned after the work of Sohollenberger (26) was em­ ployed, i. e., measuring the amount of carbon dioxide 1 105 evolved in the combustion of the soil and converting the carbon dioxide oontent Into percent organlo matter using a conversion factor of .471. A s u m m a r y of the results ob­ tained is given in Table VII below. The values for all individual samples are given in the Appendix. TABIX VII AVERAGE CONTEXT 07 ORGANIC MATTER 1RGM THE DIFFERENT LAND USE COMPONENTS OF THE WATERSHED Site Percent Organic Matter Surface Layer. 0-3 inches Undisturbed forest Coppice forest Upper pasture Lower pasture Cornfield Control plots in oornfisld 7.03 8.97 7.62 4.00 4*40 7.28 Sub-surface Laver. 3-6 inches Undisturbed forest Coppice forest Upper pasture Lower pasture Cornfield Control plots in cornfield 4.87 5.51 4.41 2.43 4.58 4*62 In the surface layers, all areas except the lower pasture and the oornfisld show values greater than 7 percent organlo matter. This is probably the result of contribu­ tions to the litter by the slash accumulation and the heavy herbaceous cover following elearouttlng. Similarly, the 106 upper pasture and the control plots la the abandoned corn­ field show a higher content of organlo setter than the undis­ turbed forest. In the ease of the upper pasture, this Is probably- due to minimum usage by oattle, heavy herbaoeous and shrubby ground cover and to the fact that the vegetation Is out bash or sprouted periodically, thus Increasing the amount of litter aeoumulated on the soil surface. The high average value for the oontrol plots In the oornfleld Is apparently the result of two seasons abandonment permitting the development of a good ground cover and consequent litter accumulation. For the surfaoe layer, this value Is considerably higher In comparison w i t h the cornfield whloh hss been grazed since abandonment. In the sub-surface layer It Is noted that the values for the two areas are approx­ imately equal. In comparison, the lower pasture and the oornfleld show values considerably lower than the undisturbed forest and less than half the content of the oopploe forest. The re­ sults of row cropping and over-grazing are thus apparent In the differences In soil organic matter. In the sub-surface layer, 3-6 Inches, all valias■ are lower by one and a half to over three percent than In the surface layer, exoept in the cornfield where the similarity of organic content In the two layers Is the result of their by the reeent plowing and cultivation. The sub-surface layer of all plots, exoept the lcarer pasture, 107 have similar organlo contents. The organlo eontant of tho sub-surface of tJha lower pasture is only about one-half as great as In the other plots. This difference together with the lower oontent of organlo matter in the surface of this area as oompared to the upper pasture Indicates differ­ ences in the soils of these areas that are not due to current management differences. Porosity. Soil porosity Is undoubtedly one of the most significant of all physioal soil properties in hydrologlo studies. Prom a hydrologlo standpointt the primary function of a soil is that of a storage reservoir. This storage reservoir sots In the same fashion as a large dam project. In the ease of floods on dam-protected streams, the flood waters aeoumulate first In the reservoir. After satisfying the initial storage in the reservoir, the water continues to accumulate In the overflow reservoir and is detained temporarily until it can be safely released to the stream below. Following the storm period the temporary storage is released as rapidly as possible until normal storage capacity is reached. The w a t e r left in the reservoir is retained and released slowly for use as irrigation water, for power generation, municipal supply and other uses. A good soil reservoir should act in the same manner. The storage oapaolty of a soil or its pore volume (porosity) is divided into two classes, 1. e«, capillary and nonoapillary. The non-capillary or large pores in the soil 10s o or respend to tho ororflow reservoir 1a tho dam-reservoir system. Tho non-eaplllary poroo ooaslot of thoso spaeoo bo two on tho aoll partloloo or aoll aggregates that aro so largo that absorption and film forooo oannot rotaln all tho wator la them against tho pull of gravity. Thus, wator Is hold In thorn only temporarily, similar to tho overflow resorvolr. Suoh storago Is termed dotontlon storago by tho hjrdrologlst • Tho small eaplllary poroo In tho soil provide tho hydrologlst's rotentlon storago. Wator In tho eaplllary or rotontlon storago reservoir Is h o l d against tho foroo of gravity but Is subjeot to tho pull of evaporation near tho surfaeo of tho soil and transpiration at any depth whoro living roots ooour. Tho wator thus rotalnod In tho soil Is that whleh Is utlllzod by plant growth or Is dlsslpatod from an area by evaporation. This rotontlon storago re­ servoir In tho soil thon nets as tho normal rosorvolr of a rlvor system. To determine tho porosity values for tho samples used heroin tho 3 x 3 lnoh soil eoros wore saturated, weighed and plaeod on a tension table for approximately 24 hours at a tension of 40 oontlmotors. Tho weights wore roeordod and thoso values wore used In determining non-eaplllary or dotontlon storage. After weighing, tho samples wore oven- dried at 105°C. to determine eaplllary or rotontlon storago os we l l as volume weight. Tho average values obtained for 109 non-eaplllary, eaplllary and total pore apaee are presen­ ted In Table VIII. TABUS Till PERCENT CAPILLARY, NON-CAPILLARY AND TOTAL PORE SPACE Site Percent Capillary Pore Voltuue Percent Non-eaplllary Pore Volume Peroeat Total Pore Volume Surfaoe Larer. 0-3 Inebes Undlaturbed forest Copplee forest Upper pasture Lower pasture Cornfield Control plots (oornfleld) 36.1 36.0 39.0 37.9 40.1 20.7 24.5 14.9 14.6 16.0 56.8 60.5 53.9 52.5 56.1 29*2 24.6 53.8 32*6 37.0 37.7 35.4 40.5 22.3 20.1 18*5 15.1 15.6 54.9 57.1 56.2 50.5 56.1 37.9 13.1 51.0 34.5 36.5 38.5 36.5 40.3 21.5 22.3 16.7 14*8 15.8 55.8 58.8 55.1 51.5 56.1 33.5 18.8 52.4 Sub-surfsee Layer. 5-6 inebes Undisturbed forest Coppioe forest Upper pasture Lower pasture Cornfield Control plots (eornfleld) Surfaee S o i l . Inches average Undisturbed forest Copplee forest Upper pasture Lower pasture Cornfield Control plots (eornfleld) 110 In terms of total porosity llttlo difference Is notod among the sitos for sltlior the surfaoe or the sub-surfaee lay sr. Higher total porosity values are indloatod In tho surfaoo layor for the two forostod areas, although tho differences are not great, than in the other four areas. More slgnlfleant than total porosity, however, are the percentages of eaplllary and non-eaplllary porosity which determine retention and detention storage, respectively. In the surfaoe soil layers, the lowest values for eaplllary pore volume were found in the control plots in the abandoned cornfield and in the two forested areas. These are the areas and layers showing the lowest sum of sand and greater than 2 mm. particles by meehanieal analysis. Xn the sub­ surface layer the undisturbed forest gave the lowest value, although the dlfferenoes in no ease were greater than 8.2 percent. Apparently the coarser textured sites and layers have higher eaplllary pore space, or moisture reten­ tion. The structure seems to be more important than the texture in these areas. The fact that the cornfield shears the highest values might indicate a favorable influence of cultivation on this property, however, the signlfieanee of differences of these magnitudes is doubtful. Xn relation to surfaoe runoff, the most important values are f o r detention storage or non-eapillary pore volumes. In the surfaee layer the highest values are indicated for the control plots and the two forested areas. In the sub- Ill surfaoe layer the two forested areas show the highest noneaplllary pore volume wh ile the eoatrol plots show the lowest. This w o u l d appear to lndleate that any Improvement In the porosity conditions In the control plots as a result of abandoning the oornfleld have been effected only In the surfaoe layer. The minimum non-oaplllary porosity of any layer Is the limiting faotor la determining the rate at whloh water moves through the total soil. In this case the two forested areas show minimum values of 20.7 and 20.1 while the minimum values for the other four areas range from 13*1 to 15*6, suggesting greater permeability rates for the two forested areas. Although the differences are not marked, the average non-eaplllary porosity values for the 0-6 lnoh layer are highest for the two forested areas. The treatment of the watershed has apparently effected a decrease In non-eaplllary porosity and an Increase In eaplllary porosity. There has been an Increase In the amount and force of precipitation reaching the soil as a result of oleareuttlng the forest. These changes, as well as changes In other physical charac­ teristics of the soil, particularly the aggregate stability and permeability, have resulted In significant modification of the runoff eharaoterlstlos of the watershed. Volume W e i g h t . Volume weight may be defined as the ratio between the dry weight of a given mass of undisturbed 112 soil and its volume. The usual method of determining ▼oluao w e i g h t , or, as it is frequently termed, the apparent spec if io gravity or bulic density, is to diYide the oven-dry weight of the undisturbed soil in grams by the volume of spaoe which this soil oeeupies in cubic oentimeters. The volume weight of a soil is dependent for the most part on structure and organic matter oontent. Ordinarily very oompact soils with low pore volume and low aggregation possess high volume weights. On the other hand porous9 well aggregated soils show low volume weights. Similarly, soils with a high oontent of organlo matter have a lower speolflo gravity as well as a lower volume weight than those with a low content of organlo matter. Forested soils, beoause they usually have a higher content of organlo matter, show lower volume weights in their surfaee layers than grazed or cultivated soils. Consequently, soils which possess low volume weights should show good soil-water relations. Conversely, those with high values for volume weight would tend to show a low infiltration rate, poor aeration and attendant low de­ tention storage oapacity. In determining the volume weights the oven-dry weights of the soil cores in grams were divided by the volume of the oore in cubic centimeters. Where soil cores were not full, volume corrections were made by filling the depressions wi t h sand then measuring the volume of sand utilized. 113 Table IX indicates the average values for the six sites for both the surfaoe and sub-surfaoe layers. Individual values are Ineluded along with porosity determinations In the Appendix. Table IX AVERAGE VOLUME WEIGHT VALUES _____ Average Volume Weight_____ Surface layer Sub-surface layer 0-3 Inches 3-6 Inches Site Undisturbed forest Coppioe forest Upper pasture Lower pasture Cornfield Control plots (cornfield) .88 .82 1.03 1.11 .93 .98 1.05 .98 1.07 1.28 1.06 1.06 The average results given In Table IX Indicate that In the surfaoe layer both the undisturbed forest and the coppice forest areas have more favorable soll-water relations than the other areas. The highest values In both the surface and sub-surface layers were found In the lower pasture. This Is partly due to soil compaction resulting from heavy grazing use, and partly due to the lower organlo matter oontent there. In the sub-surface layer, although the two forest areas show slightly lower values, the variations are asall except for the lower pasture, which has a higher volume weight. All values for the sub-surface layers are somewhat higher than In the surface layer of the same area. The 114 results tend to indicate a poorer physical oondltion of tiie aurfaoe layers In the lower pasture which would substantiate field observations on surfaee runoff conditions. Permeability. The permeability of a soil is ordinarily eonsidered to be the rate at whieh water moves through the soil column. It differs from infiltration in that the latter is eonoerned only with the rate at which water enters the soil and may be concerned with only the immediate sur­ face of the soil. It is evident then that Infiltration and permeability together provide the most important measures of physieal soil oharacteristies from the standpoint of surfaoe runoff phenomena. A soil may have a high infiltration rate and a low permeability rate or the permeability rate may be high and a **surfaoe bottleneck" may be present, giving a low infiltration rate. Xn either case, or if both values are low, comparatively little water can be stored and transmitted through the soil and high surfaee runoff re­ sults • Since soil moisture deficits must be satisfied before water starts permeating or peroolatlng through the soil column, permeability determinations were made on saturated soil cores. As nearly as possible a one-half inch head of water was maintained on the soil core for a period of one hour. The permeability rate was determined by measuring the amount of water which percolated through the soil core in that time. In extremely permeable cores one-half hour was 115 used and the resulting values were doubled* Table X shows the average permeability rates for the six sites. Sinee 100 millimeters of water are approximately equivalent to one lnoh In the 3 x 3 oores, permeability rates were eonverted to inches per hour by dividing millimeters by 100. Figure 34 shows graphloally the permeability rates and the corre­ sponding infiltration rates in inohes per hour. The per­ meability values for individual samples are given in the Appendix. TABLE X AVERAGE PERMEABILITY RATES Site Undisturbed forest Coppioe forest Upper pasture Lower pasture C o m f i e Id Control plots (eornfleld) Aurfaoe layer 0-3 inches 17,112 16,293 2,021 660 1,241 6,503 Sub-surfaoe layer 3-6 inohes 6,417 9,468 1,685 274 888 6,405 The permeability tests show muoh greater difference in the land-use areas than the other physical soil analyses. Both forested areas show very high permeability rates in both the surfaee and sub-surface layers. The uppar pasture, eornfleld and, particularly, the lower pasture sheer very low rates in comparison. The rates indicated for the oon- trol plots approach those of the forest areas whioh appar- Figure 34- Average Permeability and Infiltration Rates in Inches per Hour. Little Hurricane IVatershed Coweeta Hydrologic Laboratory I 1 Permeability (from soil samples 1-collected in 1951) Infiltration (from field tests made in 1949) 20- Undisturbed Coppice forest forest Upper pasture Lower pasture Cornfield Control plots 117 ently ref la eta tiie effeots of taro years of abandonment • By far tlie low©at rates In botii layers are those for the lower pasture, undoubtedly the result of over-grazing with Its consequent compaction. Aeeordlng to Baver (7), Lassen, Lull and Frank (22) and Fletcher (13) permeability or peroolatlon is dependent upon the non-eaplllary pore voluae. As Indicated previously, non-capillary porosity decreased as a result of the treat­ ment to the watershed although the decrease was only in the magnitude of approximately 6 percent. Assuming that non-oaplllary pore space determines permeability, from these data it appears that relatively small changes in this pore volume may effect very marked changes in the perme­ ability rate. The values noted for the infiltration rate in Figure 34 indicate a trend similar to that for the permeability rate. The ohanges indicated for both the infiltration rate and the permeability rate indieate a close relationship with the changes in surfaoe runoff noted in the sections which follow. However the much lower infiltration rates than permeability rates show that the immediate surfaee of the soil is the limiting faotor in moisture detention and is causing surfaee runoff. It is probable that the shipping of the soil oores contributed in some degree to the high permsability values noted. HYDROLOGIC DATA The baaio data la nearly all hydrologlo investigations ar# measures or praoipitatlon or reoharge and streamflow or discharge. The volume and rate of discharge from a given watershed or hydrologlo unit Is a reflection of the amount and Intensity of preolpltatlon which it receives and the characteristics or the watershed. The nanner In which these raotors are measured on the Little Hurricane Watershed and the methods employed in converting the raw data into usable form has been noted previously. Precipitation The total areal preolpltatlon received by the water­ shed, as measured by the three standard rain gages servicing the area, is summarized by hydrologlo years in Table XI. Figure 35 indicates the average monthly areal precipitation as well as the area inohes of stream dlsoharge by months and by hydrologlo seasons. Preolpltatlon, vegetation and soil conditions are all reflected in the dlsoharge curve shown in Figure 33. From January through March precipitation is at its maximum. Evaporation and transpiration are at a minimum and soil moisture conditions are at their peaic. Much of the preolpl­ tatlon which ooeurs during these months filters rather rapid­ ly through the soil reservoir into ground water slnoe soil Figure 35. Average Monthly Arei.l ?reci’>ithtion and Stret-mflow. Watershed No. 3 Cov-eeta Hydrologic Laboratory Inches Streamflow Ground Water Recharge Maximum Evaporation and Traneoiration Soil Moisture Recharge 120 TABLE XL SUMMARY OF ARZAI. PRECIPITATION BY HYDROLOGIC YEARS Hydrologlo jrtar* Precipitation In inches 61.94 84.84 72.85 63.72 74.76 56.79 50.33 67.81 77.70 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 66.64 65 .21 79.13 69*00 71.08 107•18 76.22 Average** 71.54 * hydrologlo year from November to October ** Bata for 1934 and 1951 incomplete moisture deficits ara vary low. Conaaquantly, stream dls­ oharge is at its biggest level. A large portion of tbs water appears as el sen, non-turbid wa t e r fr o m sub-surface and ground w a t e r flow. Tbs majority of tbs mtorms during tbls season produce low intensity preoipitatlon and snow wblob ordinarily does not accumulate to great depths to form flood hazard conditions. Throughout April and continuing into M a y 9 precipitation steadily declines. At the same time vegetation commences g r o w i n g , temperatures Increase and consequently losses to 121 evaporation and transpiration Increase. Starting In May and continuing throughout June and part o f July precipita­ tion Increases rather sharply. Likewise evaporation and transpiration are Increasing and the dlsoharge curve indi­ cates only a slight rise as a result of the inoreased preolpltatlon. Starting In July and continuing through October preolpltatlon steadily deelines as evaporation and transpiration continue to make heavy demands. This is indloated In the dlsoharge curve in that stream discharge is at its lowest ebb during September. October marks the end of the growing season and with it cosies a sudden decrease in both evaporation and tran­ spiration. At the same time the precipitation ourve swings sharply upward. Stream discharge shows a more gradual climb until the soil moisture deficits, resulting from the heavy use by vegetation and evaporation, are satisfied. The cycle is completed in the winter months w h e n soil moisture deflolts are met and ground wa t e r reoharge again oocurs. In studies of individual storms and their effects on stream discharge a measure of precipitation intensities as well as a measure of total amounts are required. Record­ ing rain gages 67 and 1 were used to obtain precipitation intensities. The preolpltatlon intensity values as well as the mass or accumulated precipitation values are taken from the preolpltatlon intensity records (see Appendix) 122 and are shown graphically along with stream dlsoharge for representative storms used In this study in Figures 36 through 44* An examination of individual storms shows that prior to treatment bursts of precipitation usually resulted only in a continued steady increase in stream flow. Following the olearcutting of the forest and the initiation of mountain farming, relatively small bursts of precipitation resulted in immediate and sharp increases in stream flow. The treatment of the watershed has thus resulted in pro­ ducing stream flow of extremely "flashy” characteristics. Stream Flow Gage height readings recorded for the weirs used in measuring discharge from the watershed are converted to cubic feet per second from rating tables calculated for the respective stream controls. Since precipitation is expressed in inohes and in inches per hour, the cubic feet per second values were converted to inches per hour to aid in graphical presentation and analysis. The discharge values in inches per hour (see Reoord of Runoff, Appendix) were plotted over time to give the storm hydrographs for individual storms. Hydrographs for representative stoxms used herein are given in Figures 36 through 44. The average monthly discharge values for the entire period of reoord are presented graphically in Figure 35 and Figure 36. 7.00 is • Precinitation hyetograph, mass pre­ cipitation and stream hydrograph for a unit or single summer storm (6/13/43) Watershed Ko. 3 Coweeta Hydrologic Laboratory ‘ 5«* .V S " P£K H»u*. L.oo WCtiES at o ANO 4*0^ **' I f'A £ C tP irA T io N //* (NCHiS S* 34)*, •«■■ * 5» 0: Zoo P.E. 1.00 .Of■ -PB 1500 tboo ITOO 1800 T ime 1900 2ooo P r e c ip it a t io n in inches M o inches pm Ho u r Tlgan 500 *?fT 4*00 !• ’ 37. P re o lp lta tlo n hyetograph, mss p re c ip ita tio n and etreaa hydrograph fo r a m ltlp le etora (8 /2 /4 8 ) Ifaterahed No. J Ooweeta Hydrologlo Laboratory flgan 38* P re c ip ita tio n hjetograph, oast p re c ip ita tio n and streaa hjdro{*rapb fo r an interm ediate type tto rn (8/1 2/1*3). IHCH03 //«CA Watershed Vo* 3 Coveeta Hydrologie Laboratory C rA TfOM //V if*CH*S AhfO P.E. f4O0 noo TfM£ 1)00 £/oo a n d RUNOFF /N tN C H E S «• % PER 8 HovH .Z5t & D? 8 5.P0 9 8 tucnes « ■ « H o w r % 3H/J. I I I Is fc-.v i 0700 - n *---- a800 • 9 90 looc /too (Zoo Figure UO Precipitation hyetograph, mass precipitation and stream hydrograph for storm of June 15, 1937 (Summer) Watershed No. 3 Coweeta Hydrologic Laboratory £.50 &/**. P r e c ip ita tio n m INCHES and inches HR Hoot 1 2.00 1.00 OJ50 Vicar* 41, Precipitation fcyetograplL, mass precipitation and a l m a hTdrogmph tor storm at A«cait A, 19^3* (Summer typo storm) Watarahad Vo* 3 Oowoota Vjdrologlo Laboratory Vlfur* i»2* PRFCtP{TAT\ort IN iNcuci A*t> tucHesF£* Nw r Precipitation hyetograph, m i s precipitation and stream h y d ro g ra p h for storm of August 2 2 , 1 9 3 5 * (Unit summer storm) Watershed Wo. 3 Ooveeta Hydrologio Lahore,tory 300 Zoo ^ .is ■ m 2(00 2090 T /M £ PnolpltitlOBhyotograph, bbi prtolpltation ani itnam hydrogwph for iton ofSept. JO, 1936 (Intornodiata type iton) H&terihedlo. ) OovMta Itylrologiclaboratory 945 *j*^19* *$.& - Praoipitation, in inohaa par hour for 15 nlzmtaa 149 TABLE X V I CHANGES IN FLOOD PEAKS FOLLOWING FOREST CUTTING AND MOUNTAIN FARMING Inches Per Hour Pre­ cipitation for 15 Minutes 2 3 4 5 Peak c .s.m. 19341939 11 27 43 59 Peak c.s .m. mo1945 42 70 97 126 Percent Increase Over 1934-1939 382 260 226 213 Peak c.s.m. 19461951 27 84 141 197 Percent Increase Over 1934-1939 245 311 328 334 An additional study was mads to siiow the changes in the magnitudes of flood peaks using all storms with maximum 30-minute preoipitation intensities exceeding 0*90 inch per hour. In this study the maximum flood peaks in cubic feet per second per square mile from Watershed No. 3 were plotted against the maximum peaks recorded for the control watershed (No. 2) for the same storms for the 1934-1939 and 1940-1951 periods. Least square linear regressions were calculated and straight line curves fitted to the data. The results of this analysis are presented in Figure 47* The increased magnitude of flood peaks is even more marked than in the case when only unit summer storms were used. Prior to the treatment of watershed No. 3» a storm producing a flood peak of 30 c.s.m. on watershed No. 2 showed a peak of 41 c.s.m, on watershed No. 3* Similarly, a storm produoing a peak of 70 c.s.m. on watershed No. 2 Figure **7- MAXIKDM FLOOD PBAJC RVLATIOV BM KA.TKRSHBD8 VO. 2 AVD 3 W / 3 94o -/9s/ = /O. 1 2 2 - 4 - 7 6 0 4~p°-~ iiooq po&iEs in c*s.n* - aatannaa. so. All etorma with h u e precipitation Intanaitlas orar 0*90 iadhea par hoar Zoo-- 8 21 too - - Zo 4o 6o So Mnxlwin^loo^^>eal^^n o.a.n. - Watarahad Vo. 2 151 produced a peak of about 85 c.s.m. for watershed No. 3. Following forest cutting and mountain farming, the pealcs from watershed No. 3 corresponding to 41 and 85 values for the 1934-1939 period had increased to over 36O and 670 c.s.m. respectively. In examining the individual storms and the resultant flood peaks, the highest flood peak reoorded for the period that the watershed was in forest cover netad was 109 c.s.m. In the eleven-year period following forest cutting and mountain farming 12 floods oeourred which exoeeded this foxmer maximum peak. The highest peak recorded was that of July 10, 1949 which exoeeded 1850 cubic feet per second per square mile. The two highest 5-minute precipitation intensities reoorded during the 17 years of record, however, both occurred prior to 1940 while the watershed was forest covered. From these analyses it is apparent that the treatment of the Little Hurrloane watershed has effeoted striking in­ creases in the magnitude of flood peaks. Flood peak frequencies. Even a cursory examination of the data reveals a marked change in the frequency of floods following forest cutting and treatment of the watershed. To get a quantitative measure of this change, a flood peak frequency study was made. The same standard used in the flood peak magnitude study, 1. e., floods resulting from storms having a maximum 30-minute precipitation intensity of over 0.90 inch per hour, was used as the basis for this 152 study, except that all storms, winter, Intermediate and multiple as well as summer, were Included. The flood peaks for the watershed for the period In whioh the watershed was In forest cover, 1934-1939, and the treatment period, 1940-1951, were classified separately Into 10 c.s.m. groups and arranged In order of magnitude. These values are given In oolumn 2, Table XVII and are presented graphically In histograms of the flood peaks, Figure 4 8 . Mass totals were then calculated In order of descending magnitude. From these values occurrence percentages were oomputed. The mld-polnts of the c.s.m. classes were then plotted against their corresponding percentages on a logar­ ithmic scale and smooth curves fitted to the data to give the frequency ourves In Figure 49* A logarithmic soale was selected for the ordinate In order to emphasize the maximum flood peaks since these are the values which are of greatest Importance In watershed management and flood control work as well as In engineering structures for water and erosion oontrol. It Is apparent from Figure 49 that a decided Increase In flood frequencies has been effected. Assuming an average of 50 storms of flood magnitude In 10 years, 12 flood peaks over 50 c.s.m. could be expected with the watershed In for­ est cover. Compared with this, 23 flood peaks In exoess of 50 c.s.m. should oeeur following forest cutting and subse­ quent mountain farming. On the same basis, at the 100 c.s.m. TABLE XVII FLOOD PEAK FREQUENCY DATA, WATERSHED NO. 3 Flood Peak Classes la Number of Occurrences Totals O•S.II. 19341939 m o1951 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80-69 90-99 100-109 110-119 120-129 140-149 170-179 190-199 390-399 1850-1859 6 6 1 6 1 9 15 6 6 4 6 1 1 2 1 Percentage of Total Occurrences Mass 1 19341939 22 16 10 9 3 2 1 2 5 2 1 1 1 1 m o1951 19341939 m o1951 63 54 39 33 27 23 17 16 15 100 72.7 45.4 40*9 13.9 100 85.7 61.9 52.4 42.9 36.5 27.0 25.4 23.8 9.1 4.5 13 11 6 4 3 2 1 20.6 17.5 9.5 6.3 4.8 3.2 1.6 153 Jl^ore 46v HISTOGRAMS 07 ThOGD HUEl Coveeta Hjrdrolocle Ltbomtoiy Haterahed Ho* 3 (Treated) 2Dt 16- mQ-1831 io ■ 5" 316/6*4 I0r 5- h so □ a /oo flood peak* in o.a.m* — t— t50 Zoo ZSO rigor* 1»9. TRBQPBHCT OT FLOOD PEAXf Vat*rsh*d Vo. 3 OowMta Hydrologic Lal>. 156 level, only 2 to 3 flood peaks from the forested watershed could be anticipated as compared with at least 10 flood peaks In excess of 100 c.s.m. from the treated watershed. To show that the changes are a result of the treatment rather than of climatic fluctuations during the treatment period, the same procedure was followed for the control watershed which was in forest cover for the entire period, except that 5 c.s.m. classes were used to better define a curve in graphic presentation. These results are given in Table XVIII and Figures 50 and 51. The corresponding curves for the control watershed, Figure 51, show that no great change has taken place in the precipitation pattern. Actually, the 1934*1949 period shows higher flood frequencies than the post-treatment period, indicating that possibly, had climatic conditions been even more alike for the two periods, the changes brought about on the treated watershed would probably have been even more marked. Table XIX summarizes the changes in flood peaks at the 1, 2, 5, 10, 20, 50 and 80 percent levels for the Little Hurricane Watershed. This increase in flood frequency, along with the in­ creased magnitude, aids in explaining why channel bank vegetation is being removed and why the vegetation which 8tarts growing on this site is washed away before it has an opportunity to become firmly established. It shows, too, i TABU mu FLOOD P1AK FREQUENCY DATA, CONTROL WATERSHED NO. 2 Flood PtaJc C1&8898 In 0*8 0-5 6-10 11-15 16-20 21-25 26-30 31-35 36-40 41-45 51-55 56-60 66-70 81-65 Number of Oeourrenoes 19341939 6 3 3 2 1 1 1 2 19401951 9 20 14 8 2 1 1 1 1 1 Mass Totals 19341939 21 15 12 9 7 6 5 4 19401951 58 49 29 15 7 5 4 3 2 2 1 1 Percentage of Total Occurrences 19341939 100 71.4 57.1 42.6 33.3 28.6 23.8 19.0 100 84.5 50.0 25.8 12.1 8.6 6.9 5.2 3.4 9.5 1 1 19401951 1.7 4.8 157 > Vigors 5ft. 2*T EI8T0GSAKS GF FLOOD PXAJC8 Oovnata Hydrologic laboratory Vatarahad Ho. 2 (Control) /5 /o- 5 /aT igaa-igg JT-- H-------- — --1-- M O B M O M Q JO St OQfr XXO|«0^«1 «%MMO0 Z *0J| V « (M » W I m u o d o u jo zo is d b a u •ts •**&*£ 160 why fanners of this area, in increasing numbers* are con­ verting their first bottom-lands from truck and corn cultiva­ tion to permanent pastures. TAB 135 XIX FREQUENCY OF FLOOD PEAKS BEFORE AND AFTER FOREST CUTTING AND SUBSEQUENT MOUNTAIN FARMING Percent of Observations Period 1 m 1934-1939 1940-1951 «tH 2 hi« 175 2000/ 5 10 20 50 80 flood peak in c.s.m. exceeds • s e e 135 1000 100 230 75 130 55 100 30 60 20 30 Distribution of Storm Runoff To show the effects of forest cutting and subsequent mountain farming on the manner in which storm runoff comes off the experimental watershed* distribution graphs were made for the two periods: 1934-1939* in which the water­ shed was in forest cover* and 1940-1951> the mountain farming treatment period. The method outlined by Wlsler and Brater (36) was used to prepare the storm runoff distribution graphs. Five storms* two for the "before" period and three for the "after** period were selected for this study on the basis of similarity in storm type and precipitation amount and in­ tensity. The bases of the hydrographs were divided Into three-minute intervals and the total flow and the base flow 161 were calculated Tor these Intervals• Three-minute Intervals were selected to give a total of approximately 20 equal In­ tervals or points for subsequent graphical analysis. Storm runoff was obtained by subtracting base flow from total flow (unit hydrographs) • Storm runoff percentages were then com­ puted for each three-minute Interval. Tables XX through tttv show the calculations for the Individual storms. The values for the "before” and "after" periods were then averaged to give the composite distribution graph data whloh Is shown In Tables XXV and XXVI. These data were then plotted to give the final distribution graphs for the two periods whloh appear In Figure 52. To show the differences In the volume of storm runoff coming off the watershed at maximum flood stage, a time period equal to one-tenth the base of the hydrograph (three-mlnute Interval on each side of the aotual peak) was marked off on the distribution graphs and the percentages for these Inter­ vals were determined by planlmeter. These results are Indi­ cated on Figure 52• It la obvious from these graphs that a definite change has taken place In the distribution of the storm runoff. One notable effect Is the change In the time of concentration (time that runoff began to the maximum peak) for the water­ shed. While the watershed was in forest cover, the peak occurred approximately 35 minutes from the time stoim runoff first began. Following forest outting and mountain farming TABLE ZZ DERIVATION OF DATA FOR DISTRIBUTION GRAPH FOR STORM OF JUNE 12, 1936 111m y r i n ---- 5----- 5“ “ s 4 ------ P “ “ 1 ----3-Minute Total Head Volume In Base Flow Head Volume in Storm Runoff Percent of Volume in Cubic Feet Total Storm Cubic Feet in In./Hr. Interval In In./Hr. Ave. (Interval) Cubic Feet Runoff Inst• Ave• (Interval) Inst. Number 0 1 2 3 4 3 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 .0070 .0082 .0096 .0110 .0124 .0136 .0162 .0194 .0210 •0240 .0274 .0302 .0298 .0210 .0182 .0160 .0144 .0130 •0120 .0100 .0090 Totals Conversion Factor: .0076 .0089 .0103 .0117 .0130 .0149 .0178 •0202 .0225 .0257 .0288 .0300 .0254 .0196 .0171 .0152 .0137 .0125 .0110 .0095 31.43 36.81 42.60 48.39 53.77 61.63 73.62 33.55 93.06 106.30 119.12 124.08 105.03 81.07 70.73 62.87 56.66 51.70 45.50 39.29 1387.23 .0070 .0073 .0076 .0079 .0081 .0084 .0087 .0090 .0093 .0096 .0099 .0102 .0104 .0103 •0102 .0100 .0098 .0096 .0094 .0092 .0090 .00715 .00745 .00775 .00805 .00825 .00855 .00885 .00915 .00945 .00975 .01005 .01030 .01035 •01025 •01010 .00990 .00970 .00950 .00930 .00910 29.57 30.81 32.05 33.29 34.12 35.36 36.60 37.84 39.09 40.33 41.57 42.60 42.81 42.39 a . 77 40.95 40.12 39.29 38.46 37.64 756.48 1.86 6.00 10.55 15.10 19.65 26.27 37.02 45.71 53.97 65.97 77.55 81.48 62.24 38.68 28.96 21.92 16.54 12.41 7.04 1.83 630.75 .29 .95 1.67 2.39 3.12 4 *16 5.87 7.25 8.56 10.46 12.29 12.91 9*87 6.13 4*59 3.48 2.62 1.98 1.12 .29 100.00 Inches per hour to cubic feet for 3 minutes equals 180/.04352 equals 4136. 162 TABUS XII DERIVATION OF DATA FOR DISTRIBUTION GRAPH FOR STORM OF JUNE 15, 1937 8 1 5 6 7 2 3 4 9 Volume in Base Flour Head Volume in Storm Runoff Peroent of 3-Minute Total Head Total Storm Cubic Feet Volume in Cubic Feet in In./Hr. Interval in In./Hr. (Interval) Ave. (Interval) Inst. Ave. Cubio Feet Runoff Number Inst. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 .0022 .0045 .0065 •0080 .0100 •0120 .0135 .0155 .0190 •0220 •0260 .0290 .0290 •0240 .0200 .0150 .0127 .0115 •0090 .0075 .0050 Conversion Factor: .00900 .01100 .01275 .01450 .01725 .02050 •02400 .02750 •02900 •02650 •02200 .01750 .01385 •01210 .01025 .00825 •00625 13.86 22.75 29.99 37.22 45.50 52.73 59.97 71.35 84.79 99.26 113.74 119.94 109.60 90.99 72.38 57.28 50.05 42.39 34.12 25.85 1233.76 .0050 .0023 .0025 .0027 .0029 .0031 .00335 .0037 .00405 .0044 .00475 .00505 .0052 .0052 .00515 .0051 .0051 .0051 .00505 .0050 .0050 9.51 10.34 11.17 11.99 12.82 13.86 15.30 16.75 18.20 19.65 20.89 21.51 21.51 4.35 12.41 18.82 25.23 32.68 38.87 44.67 54.60 66.59 79.61 92.85 98.43 88.09 21.30 69.69 21.09 21.09 21.09 20.89 20.68 20.68 51.29 36.19 28.96 350.32 21.50 13.44 5.17 883.44 .49 1.40 2.13 2.86 3.70 4.40 5.06 6.18 7.54 9.01 10.51 11.14 9.97 7.89 5.81 4.10 3.28 2.43 1.52 .59 100.00 Inches per hour to cubic feet for 3 minutes equals 180/.04352 equals 4136. 163 Totals .00335 .00550 .00725 .0022 .0024 .0026 .0028 .0030 .0032 .0035 .0039 .0042 .0046 .0049 .0052 .0052 .0052 .0051 .0051 .0051 .0051 .0050 .0050 TABLE XIII DERIVATION OF DATA FOR DISTRIBUTION GRAPH FOR STORM OF JULY 11, 1946 8 5 6 7 9 4 Volume in Base Flow Head Volume in Storm Runoff Peroent of 3-Minute Total Head Total Storm Interval in In./Hr. Cubic Feet in In./Hr. Cubic Feet Volume in Ave. (Interval) Cubic Feet Runoff Number Inst. Ave. (Interval) Inst. 2 0 1 2 .0045 .008$ 3 4 5 .0950 6 7 8 9 10 11 12 13 14 15 Totals .0300 •0900 .0675 •0400 .0260 .0150 .0125 .0115 3 .0065 .0198 •0625 .0925 .0788 .0536 .0330 •0205 .0138 •0120 .0100 .0108 .0085 .0075 .0092 •0068 •0060 •0080 .0072 .0064 26.88 81.89 258.50 382.58 325.92 222.52 136.49 84.79 57.08 49.63 44.67 38.05 33.09 29.78 26.47 1798.34 •0045 .0055 .0065 .0075 .0073 .0072 .0071 .0070 .0068 .0067 .0066 .0064 .0063 .0062 .0061 •0060 .0050 .0060 .0070 .0074 .0072 .0072 .0070 •0069 •0068 •0066 .OO65 .0064 .0062 .0062 .0060 20.68 24.82 28.95 30.61 29.78 29.37 28.95 28.54 28.12 27.30 6.20 .45 57.07 229.55 351.97 296.14 193.15 107.54 56.25 4.10 16.49 25.29 21.28 13.88 7.73 4.04 2.08 28.96 1.60 26.47 25.64 25.64 22.33 17.79 11.58 7.45 4*14 406.57 1391.77 100.01 26.88 24.82 1.28 .83 .54 .30 1.65 .12 Conversion Factor: Inches per hour to cubic feet for 3 minutes equals ISO/.04352 equals 4136 . *T9T 1 DERIVATION OF DATA FOR DISTRIBUTION GRAPH FOR STORM OF MAY 22, 1949 8 4 5 6 7 9 Storm Runoff Pereent of Volume in Volume in Base Flour Head 3-Minute Total Head Total Storm Cubic Feet Volume in Cubic Feet in In./Hr. Interval in In./Hr. Runoff Ave. (Interval) Cubio Feet Ave. (Interval) Inst. Number Inst. 1 0 1 2 3 4 5 2 •0080 .0070 •0110 .0190 .0150 •0800 6 7 .0530 8 .0420 9 .0330 .0280 13 14 15 16 17 18 19 20 21 22 23 24 25 Totals .0060 •0060 .1890 .1050 10 11 12 3 .0230 •0200 .0170 .0150 .0125 .0110 .0090 .0085 •0080 .0078 .0076 .0074 .0072 .0070 .0068 Conversion Factor: 28.95 39.29 62.04 204.73 556.29 814.79 326.74 196.46 155.10 126.15 105.47 .0065 .0185 •0160 .0138 76.52 66.18 57.08 •0100 41.36 36.40 33.92 32.67 31.85 .0075 .0075 .0074 .0074 .0073 .0073 .0072 .0071 .0071 30.19 29.37 28.54 •0068 .0095 .0495 .1345 .1970 .0790 .0475 .0375 .0305 .0255 .0215 .0118 .0088 .0082 .0079 .0077 .0075 .0073 .0071 .0069 88.92 48.80 31.02 3248.81 .0070 .0075 .0078 .0080 •0080 .0079 .0079 .0078 .0078 .0077 .0076 .0062 •0068 .0072 .0076 .0079 •0080 •0080 .0079 .0078 .0078 .0078 .0076 .0076 .0075 .0074 .0074 .0074 .0073 .0072 .0072 .0071 .0070 .0070 .0070 .0070 .0069 .0069 .0068 25.64 28.12 29.78 31.43 32.67 33.09 33.09 32.67 32.26 32.26 32.26 31.43 31.43 31.02 30.61 30.61 30.61 30.19 29.78 29.78 29.37 28.95 28.95 28.54 28.12 762.66 3.31 11.17 32.26 173.30 523.62 .13 .45 1.30 6.97 21.06 781.70 293.65 163.79 122.84 93.89 73.21 57.49 45.09 35.16 26.45 18.19 10.75 6.21 4.14 2.89 31.44 11.81 6.59 4.94 3.78 2.94 2.31 1.81 1.41 2.07 1.24 .63 .42 .08 .05 .03 2486.15 99.98 2.48 1.06 .73 .43 .25 .17 •12 .10 .02 Inches per hour to cubic feet for 3 minutes equals 180/.04352 equals 4136. TABIX HIV DERIVATION OF DATA FOR DISTRIBUTION GRAPH FOR STORM OF AUGUST 4, 1949 1 2 3 5 6 1 4 3-Minute Total Head Volume In Base Flow Head Volume in Interval in In./Hr. Cubic Feet In In./fcr. Cubic Feet Are. (Interval) Ave. (Interval) Inst. Number Inst. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Totals .0040 •0080 .0180 .1150 .1500 .0260 .0180 .0145 •0124 .0105 .0100 .0085 .0075 .0065 •0062 .0060 •0058 .0057 •0056 .0055 •0054 .0060 .0130 .0665 .1325 •0880 •0220 .01625 .01345 .01145 •01025 .00925 .0080 .0070 .00635 .0061 .0059 .00575 .00565 .00555 .00545 24.82 53.77 275.04 548.02 363.97 90.99 67.21 55.63 47.36 42.39 38.26 33.09 28.95 26.26 25.23 24.40 23.78 23.37 22.95 22.54 1838.03 .0040 .0048 .0054 •OO64 .0068 •0064 .0063 .0062 .0061 •0060 .0060 .0059 .0058 .0057 .0057 •OO56 .0056 .0055 .0055 .0054 .0054 .0044 .0051 .0059 •0066 •0066 .00635 •00625 .00615 .00605 .0060 .00595 .00585 .00575 .0057 .00565 •0056 .00555 .0055 .00545 •0054 18.20 21.09 24.40 27.30 27.30 26.26 25.85 25.44 8 9 Storm Runoff Peroent of Volume In Total Stoxm Cubio Feet Runoff 24.82 24.61 24.20 23.78 23.58 23.37 23.16 22.95 22.75 22.54 22.33 6.62 32.68 250.64 520.72 336.67 64.73 41.36 30.19 22.34 17.57 13.65 8.89 5.17 2.68 1.86 1.24 .83 .62 .41 .21 .49 2.40 18.44 38.31 24.77 4.76 3.04 2.22 1.64 1.29 1.00 .65 .38 .20 .14 .09 .06 •04 .03 .02 478.95 1359.08 99.97 25.02 Conversion Factor: Inches per hour to oublo feet for 3 minutes equals 180/.04352 equals 4136. 167 TABLE 2X7 COMPOSITE DISTRIBUTION GRAPH DATA FOR "BEFORE" PERIOD - 1934-1939 Interval Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Peroent of Total Storm Runoff Date of Storm " -6/12736 .... .... -6715/57---.29 .95 1.67 2.39 3.12 4.16 5.87 7.25 8.56 10.46 12.29 12.91 9.87 6.13 4.59 3.48 2.62 1.98 1.12 .29 *Flood peaks matched •49 1.40 2.13 2.86 3.70 4 .40 5.06 6.18 7.54 9.01 10.51 11.14 9.97 7.89 5.80 4.10 3.28 2.43 1.52 .58 Average .39 1.18 1.90 2.62 3.41 4*28 5.46 6.72 8.05 9.74 11.40 12.02 9.92 7.01 5.20 3.79 2.95 2.20 1.32 •44 168 TABLE 2X71 COMPOSITE DISTRIBUTION GRAPH DATA FOR "AFTER" PERIOD - 1940-1951 .45 4.10 16.49 25.29 21.28 13.88 7.73 4.04 2.08 1.60 1.28 .83 .54 .30 .12 .04* I .45 1.30 6.97 21.06 31.44 11.81 6.59 4.94 3.78 2.94 2.31 1.81 1.41 1.06 .73 .43 .23 .17 .12 .10 Average .49 2.40 18 .44 38.31 24.77 4.76 3.04 2.22 1.64 1.29 1.00 .65 •38 .20 .14 .09 .06 .04 .03 .15 .75 4.49 18.66 31.68** 19.29 8.41 5.29 3.35 2.22 1.73 1.36 .96 .66 .41 .23 .11 .08 .05 •04 o » f O V j J V t .... oooo 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 i i i• 1H 1 « Interval Number Peroent or Total Storm Runoff Date of Storm 7/HAb 8/4/49 4/22/49 *Values dropped to give 20 Intervals **Flood peaks matched .03* .02* .01* .01* Figure 52 THE EFFECTS OF FOREST CUTTING A:ID SUBSEQUENT 'FOUNTAIN FARMING ON THE DISTRIBUTION OF STORM RUNOFF. Coweeta Hydrologic Laboratory Watershed Mo. 3 32 30 28 26 Mountain Farm (1940-1951) 24 22 20 18 16 U 12 10 A Forest (1934-1939) 8 6 4 2 tS Time Intervals (3 minutes) 20 170 the peak occurred on the average of 15 minutes after runoff began. In a few cases w i t h short, intense summer storms, the time from the start of runoff to the peak was observed to be as little as 10 minutes. It is apparent, too, that the peak percentage itself was more than doubled as a result of the treatment. While in forest cover, the m a x i m u m value was approximately 12 per­ cent*, following cutting and treatment, this value Jumped to approximately 32 percent. Apparent, too, is the change in shape of the hydrograph to an almost needle-shaped storm hydrograph• Significant, too, is the change in the peak percentage based on the maximum discharge for a time period equal to one-tenth the base of the storm hydrograph. For the "be­ fore" period, indicated as "A" in Figure 52, this amounted to 24 peroent of the total storm runoff. As a result of the treatment, the peak percentage "B" in Figure52 increased to 49 percent. SUMMARY For many years it has been a common practice In the Southern Appalachians to clear off the native forest cover on steep slopes and then to attempt to farm the area. In this study a determination of the effects of this use of land on some of the biologio, edaphio and surface runoff characteristics of a 23-acre watershed was made with the following results. Biologio Changes B y olearcutting the forest cover and applying differ­ ent land use praotioes to the watershed a marked change in the vegetation was produced. Nearly half of the water­ shed, approximately 10 acres, was permitted to grow back into natural forest cover. Eleven years after cutting and negligible use by cattle, the dominant sprouts and seedling trees are approximately 3 to 3i Inches in diameter at breast height and 12 to 15 feet in height. By examining adjacent uncut areas and timber cruise data it was found that there is little ohange in species composition exoept for the presence of wild plum, hawthorne and staghorn sumac on the cutover area. These invading species will probably disappear within a few years following complete canopy closure. 172 Shrubby and herbaceous cover In the coppice forest area is very sparse except along the ridges and in a few openings. Many species associated with forest openings were observed although most of these will probably disappear within a few years. In the more dense portions of the coppice area black- berry, for example, has virtually disappeared although there is ample evidence of its former occupation. In most of the adjoining forest area there is a moderate understory of mountain laurel and rhododendron. It is apparently much slower in becoming re-established than many other species since it is fairly oommon only near the ridges in the coppice forest area. In the two pasture areas as well as in the abandoned cornfield, the most marked changes are those in vegetation density and in species composition of the shrubby and her­ baceous cover. On all three areas vegetation density is low, particularly in the lower pasture. Changes i*i species composition were marked by the appearance of such more or less noxious or unpalatable species as mullein, yarrow, Canada thistle, smartweed, nettle and purslane — quently associated with land abuse. all fre­ The forest tree seed­ lings and sprouts observed in 1951 consisted of about the same speoles which were represented on the area prior to clearcuttlng. Invading species included green ash, haw­ thorns, black walnut and butternut. 173 In tne control plots, established in the cornfield at the time It was abandoned In 1949* the ground cover Is nearly complete after only two years. The most abundant species In these two plots are blackberry and wild strawberry. It Is obvious In both plots that tree species will soon take over* Numerous stems In both plots exceed 5 feet In height. In the absence of forest vegetation, stream tempera­ tures lnoreased to the point where they were higher than the maximum limits for trout. Similarly, stream turbidities were lnoreased by approximately three times. Consequently, due to changes, the stream Is no longer suitable for trout habitation. Edaphlo Changes The physical characteristics of the soil Influence the rate at which preelpltatlon enters and Is transmitted through the soil and these factors In turn affect stream flow. A summary of the physical properties of the soils on different parts of the watershed are presented In Table XXVTI. In addition, Infiltration studies were made in 1949 and 1950 by the Coweeta Eydrologic laboratory* The results of the surface Infiltration study made in 1949 showed marked differences in the Infiltration rates of the portions of the watershed used In different ways. In all the forested plots sampled on the Coweeta area, in­ cluding the coppice forest portion of the Little Hurricane TABLE XXVII SUMMARY OF PHYSICAL CHARACTERISTICS OF THE SOIL Site Perme- Organic Sand and Aggregates Volume Total Capillary NonWeight Porosity Porosity Capillary ability Matter Coarser over 4 am. Porosity Content Material Peroent Peroent Peroent In ./Hr. Peroent Peroent Peroent 0-3 inoh layer .88 Undisturbed forest •82 Coppice forest Upper pasture 1.03 Lower pasture 1.11 Cornfield .93 Control plots (cornfield) .98 60.5 53.9 52.5 56.1 36.1 36.0 39.0 37.9 40.1 20.7 24.5 14.9 14.6 16.0 53.8 29.2 24.6 1.05 .98 1.07 1.28 1.06 54.9 57.1 56.1 32.6 37.0 37.7 35.4 40.5 22.3 20.1 18.5 15.1 15.6 1.06 51.0 37.9 13.1 56.8 171.1 163.0 20.2 6.6 12.4 85.0 7.03 8.97 7.62 4.00 4.40 7.28 81.7 91.4 99.6 97.7 99.1 91.0 46.2 4.87 5.51 4.41 2.43 4.58 82.8 97.4 95.8 90.4 95.6 73.8 63.5 48.6 56.7 62.2 4.62 89.9 40.1 65.5 73.2 64.1 53.9 59.0 3-6 Inoh layer Undisturbed forest Coppice forest Upper pasture Lower pasture CornfleId Control plots (cornfield) 56.2 50.5 64.2 94.7 16.9 2.7 8.9 64.0 175 Watershed, the average Infiltration rate invariably exceeded six inches per hour. The values noted for the cornfield, upper pasture and lower pasture respectively in 1949 were 4*00, 3*00 and O .56 inohes per hour. Sinoe the majority of storms on the Coweeta area show precipitation intensities in excess of 0.56 inohes per hour, high surfaee runoff rates would be anticipated from the lower pasture. Similar infil­ tration tests made on the abandoned cornfield immediately before and following grazing in 1950 Indicated that even short periods of grazing eauaed sharp decreases in the in­ filtration rate. As indicated by samples collected from adjacent undis­ turbed forest plots, the permeability of the soil averaged approximately 171 inches per hour in the 0-3 inch layer and over 64 inohes per hour in the 3-6 inoh zone. Tests after treatment indicate a marked decrease in these rates, ranging as low as those in the lower pasture, where the average rates were 6.6 and 2.7 Inohes per hour for the 0-3 and 3-6 inch layers respectively. The peroent of water-stable aggregates likewise showed a marked decrease, from 8 5*5 and 73.8 percent of aggregates over 4 mm. in size in the surface and sub-surface layers for the undisturbed forest respectively, to a low of 46.2 and 40.1 in the control plots in the abandoned cornfield. The soil dispersion resulting from the decrease in aggregation is undoubtedly responsible for increases in soil losses from 176 the watershed and indirectly for the increases in surface runoff. An analysis of the aggregation of the fine earth material indioated similar results. Differences in organic matter content, volume weight and porosity are also evident. The effect of former cultiva­ tion in the abandoned eornfleld is evidenced in that the organie matter oontent of both surfaoe and sub-surface layers are nearly the same. In the control plots in the abandoned cornfield the organie oontent of the sub-surface layer is approximately the same as in the cornfield. However, the two years protection offered the control plots appears to have been sufficient to increase the organic matter oontent ma­ terially in the surface layer. The lowest organic matter oontent in both layers is noted for the lower pasture. Apparently grazing has effeoted soil compaction and thus in­ hibited the incorporation of litter and humus. It is pro­ bable, too, that earlier cultivation of this area may have brought about soil changes in the lower pasture. The coppice forest area shows the highest content of organic matter in both layers. This is probably the result of an accumulation of litter from slash and from a heavy herbaceous cover following clearoutting and the decaying of root systems from the trees formerly oocupylng the area. The greatest differences in volume weight in both layers were found between eoppioe forest and lower pasture, indica­ ting a d o s e relationship between organic matter oontent and 177 volume weigiit. Xn the surface layer, volume weight values range from 0.82 to 1.11. In the sub-surface layer all values are very nearly the same except In the lower pasture where volume weight was 1*28, again showing the effeots of heavy trampling. These values, as well as those noted for organic matter oontent, suggest that the most marked changes In the physical characteristics of the soil occur In the 0-3 Inch layer. In porosity values, a slight decrease In total and noncaplllary porosity and an lnorease In capillary porosity, In comparison with undisturbed forest conditions, Is Indicated. In non-caplllary porosity a decrease of approximately 6 per­ cent by volume Is shown. According to many writers, non- caplllary porosity determines permeability. From the results of these determinations, small changes In the large pore volume, then, may effect marked changes In permeability rates. One of the greatest changes In the soil as a consequence of the treatment Is In soil losses from the watershed. During the calibration period, 1934-1939* and until August, 1941 the average soil losses amounted to about 154 pounds per acre per year. Following the cutting of the forest and the application of mountain f a m i n g praotlces, the average soil losses Increased to well over a ton per acre per year. A small portion of this Increase might perhaps be attributed to a change in the method of collecting so 11 losses. However, virtually all the increase should be assigned to the treatment of the watershed. 178 Cultivation alone appears to be responsible for marked increases in soil losses. These losses Increased sharply for a two-year period following the cutting of the forest cover in which the cornfield was cultivated and the pasture areas were grazed. During 1944 and 1945 the cornfield was protec­ ted and permitted to lie idle while the pastures were being grazed. During this period, soil losses dropped noticeably. Following this period, the cornfield was again cultivated for four years and the soil losses mounted to a high in 1949 or over 2j tons per acre per year. In 1950 and 1951 the cornfield was abandoned for cultivation and grazed along with the two pasture areas and again soil losses declined. Runoff Changes A study made on the Little Hurricane Watershed indicated an average runoff percent of 2.66 for all storms with a max­ imum 30-minute precipitation intensity over 0.90 inohes per hour for the period in which the area was in forest cover. Following forest cutting and mountain farming this was in­ creased to 4.50 peroent. Summer storms alone gave percen­ tages of 1.53 and 4.79 for the before and after periods, respectively. From field observations and from the results of infil­ tration and permeability tests it was felt that the coppice forest area and possibly portions of the upper pasture as well, were not contributing materially to the total volume 179 of surface runoff* Based on ‘ this assumption the runoff per­ centages were calculated on the basis that 6 , 8 , 10 and 12 acres were contributing all the surface runoff. It is highly probable that the small heavily-grazed and cultivated areas are the significant flood source areas and consequently have an effect on total watershed conditions far out of proportion to their actual area* A very significant change in the magnitude of flood peaks occurred as a result of the land use treatment applied. Unit summer s t o m i yielding precipitation at the rate of three inches per hour for 15 minutes, for example, showed an in­ crease in the resulting maximum flood peak from 27 to 84 cubic feet per second per square mile. At the rate of five inches per hour for 15 minutes the increase theoretically would be from 59 c.s.m. to approximately 335 o.s.m. A study of flood peak frequencies for the watershed in­ dicated similar marked changes. Histograms of the flood peaks show that during the period of standardization (land in forest cover), 1934-1939, only three floods occurred with maximum peaks over 50 c.s.m. (three per five years). Prom 1940 to 1951, after clearing and cultivation, 25 floods with maximum peaks in excess of 50 c.s.m. were noted (more than ten per five years). A similar analysis of data from a con­ trol watershed actually indicated lower frequency values for the later period. 180 One of tJio most significant changes in surface runoff brought about by forest cutting and subsequent mountain farm­ ing is the manner In which the runoff water comes off the watershed. Runoff distribution graphs for the before and after periods show that the peak runoff occurs about 15 min­ utes after the beginning of a storm, since the area has been cleared and farmed. Before olearing, the peak runoff did not occur until about 35 minutes after the beginning of a storm. For both periods the duration of runoff was approx­ imately 60 minutes. Prior to treatment approximately 12.5 percent of the storm flow came off the watershed during the peak 3-minute interval. Following treatment, this value jumped to nearly 32 percent. PRACTICAL IMPLICATIONS FOR LAND USE As a result of increasing population and economic pres­ sures, thousands of acres of steep forest land within the Southern Appalachian region have been cleared for use as pasture or cropland. The advisability of this practice has been extremely questionable from an agronomic and, parti­ cularly, from a hydrologic standpoint. The results of this study show that forest cutting and the application of average to poor farming practices have a very deleterious effect upon the physical properties of the surface soil and increase surface runoff during storms. The common practice in this region is to cut off the forest cover, plow the selected area and plant it to row 181 crops. Within a short period of time after cultivation is begun, the organic constituents and natural structure of the former forest soil begins to breaJc down, leading to soil dispersion and resultant accelerated erosion. In cultivating row crops on steep land the exposed topsoll washes away in­ c r e a s i n g l y with eaoh year and fertility declines until, after a period of approximately 10 to 15 years, yields have decreased to the point where the venture is no longer pro­ fitable . Since the farmer invested a great deal of labor in the original clearing he frequently decides to convert the wornout area to pasture instead of allowing it to return to forest cover. All too frequently the land now becomes over-grazed, resulting in rapid compaction of the already eroded soil due to trampling by the cattle. Infiltration and permeability rates qulokly decline and the soil loses its capacity for water storage. The cumulative effect of these abuses is to increase the volume of surface runoff and to multiply the frequenoy and magnitude of floods many times over that which would have occurred under natural forest conditions. At first glance, it might appear that the injury is confined to an already worn-out area and thus is of minor consequenoe. However, once these floods begin to ooour in increasing number and magnitude, it beeomes clear that the damage is much more extensive and much more serious. A 182 The increased erosion produces sediment which silts in reservoir systems* When such lands are part of a watershed contributing to a municipal water supply, lnoreased turbidity of the streams may result in a necessity for purification. Such flood source areas produce erosion on lower lands which normally would not be subject to serious soil washing. Within the past ten years many farmers in this region, who depended on the cultivation of row and truoic orops in the first bottom lands along the streams for their greatest source of income, have been forced to convert their first bottom lands to permanent pasture which can withstand increased flooding. Results such as these could probably be expected from average to poor farming practices in other areas with similar topography and soils. It might be suggested that better farming methods could alleviate the situation. C o m might be rotated with clover and small grains and the amount of fertilizer used could be lnoreased to advantage. Less damage from trampling would oocur if fences were erected and a more restricted number of cattle permitted to graze. How­ ever, all these improved praotioes are difficult and expen­ sive on suoh steep land. When stream conditions in undisturbed forests are ob­ served, one is convinced that good forests, good soils and good water go hand in hand. Soil conditions under undisturbed forest encourage storage of water and maice possible the 183 oontrol of erosion. It Is reasonable to believe that through, the ages there has developed an harmonious adjustment of vegetation, soil and water. This natural adjustment, how­ ever, appears to be In delloate balance. It Is Impossible to disturb the forest without disturbing this equilibrium. Xt I s obviously impractical, however, to leave all land In forest no matter how excellent the supply of water thus assured would be. Land must be used, but must also be carefully managed In order to husband Its potentialities for human satisfaction. As we come to understand all of the physloal forces which must be Jcept In balance, we will be better able to develop land management practices whloh will permit us to utilize all resources without exploiting any one at the expense of another. L I T E R AT UR E CI TED 1. American Society or Civil Engineers* Hydrology Handbook* Manuals of Engineering Practice No. 28. 184 PP*» 1949. 2. Ayer, G. R. A Progress Report on an Investigation of the In flue no e of Reforestation on Stream-Flow in State Forests of Central New York. United States Department of Interior, Geological Survey in cooperation with State of New York Conservation Department. 185 PP.* 1949* 3* Bailey, R. W., G. W. Craddook and A. R* Croft* Water­ shed Management for Summer Flood Control in Utah. United States Department of Agriculture, Forest Service, Misc. Pub. No. 639. 24 pp., 1947* 4. Bain, F. L. Hydraulic Research in the United States. United States Department of Commerce, National Bureau of Standards, National Hydraulic Laboratory, Volume 13» 220 pp., 1949. 5. Bartel, F. 0. and C. S. Slater. Progress Report of the Central Piedmont Soil and Water Conservation Experiment Station, Statesville, N. C«, 1930-1935* United States Department of Agriculture, Soil Conservation Servioe, ESR-6, 133 PP. (October) 193®. 6 . Bates, C. G. and A. I. Henry. Forest and Streamflow Ex­ periment at Wagon Wheel Gap, Colorado. Monthly Weather Review Supplement 30:1-79. 1928. 7. Baver, L. D. Soil Physics. So n s , I n c ., 39® P P •» 194®. New York: lohn Wiley and 8. Borat, H. L. and R. Woodburn. Hydrologic Studies Compilation of Rainfall and Runoff from the Watersheds of the North Appalachian Conservation Experiment Station, Zanesville, Ohio. United States Department of Agricul­ ture, Soil Conservation Servioe. Mimeo., 13® PP* 193®. 9. Bouyoucos, G. J. Directions for Making Mechanical Analyses of Soils. Soil Science, Volume 22, No. 3> 193®. 10. Devereux, R. E., E. F. Goldston and W. A. Davis. Soil Survey of Macon County, North Carolina. United States Department of Agriculture, Bureau of Chemistry and Soils, No. 16, Series 1929. 21 pp. 185 11* Dreibelbis, F. R. and F. A. Post. An Inventory of Soil Water Relationships on Woodland, Pasture and Cultivated So i l s• Soil Science Society of Amerioa Proceedings, Volume 6 , pp. 462-473. 1941. 12. Fernald, M. L. Gray’s Manual of Botany. Eight ed., New York: Amerioan Book Co. 1632 pp., 1950. 13* Fletcher, P. W. The Hydrologic Function of Forest Soils in Watershed Management. Paper presented at the Annual Meeting of the Society of American Foresters, Division of Watershed Management, Biloxi, Miss., Decem­ ber 1951. 14. Frank, B. and A. Netboy. Water, Land and People. New York: Alfred A. Knopf, 331 pp. 1950. 15. Garstka, W. U. Hydrology of Small Watershed under Winter Conditions of Snow-Cover and Frozen Soil. Transactions, American Geophysical Union, pp. 838-671. 1944. 16. Greene, G. E. Land Use and Trout Streams. Journal Soil and Water Conservation. Vol. 5(3):125-126, (July) 1950. 17. Hays, O. E. and H. B. Atkinson. Hydrologic Studies — Compilation of Rainfall and Runoff from the Watersheds of the Upper Mississippi Valley Conservation Experiment Station, La Crosse, Wisconsin, United States Department of Agriculture, Soil Conservation Service, 1939. 18. Horner, W. W. and C. L. Lloyd. Infiltration Capacity Values as Determined from an Eighteen-month Record at Edwardsvllle, Illinois. American Geo-physical Union, Part 2, 21st Annual Meeting, pp. 522-541, 1940. 19. Hursh, C. R. Outline for Compiling Precipitation and Runoff Data from Small Drainage Areas. United States Department of Agriculture, Forest Servioe, Appalachian Forest Experiment Station, Technical Note No. 34. 59 PP. (August) 1940. 20. Research in Forest-Streamflow Relations. UNAS^LVA, Vol. V (l):2-9. (January-March) 1951. 21. Johnson, E. A. Effect of Farm Woodland Grazing on Water­ shed Values in the Southern Appalachian Mountains. Journal of Forestry 50(2): 1 0 9 - H 3 » 1952. 22. Lassen, L . , H. W. Lull and B. Frank. Some Fundamental P l a n t-Soil-Water Relations in Watershed Management. United States Department of Agriculture, Forest Service, Division of Forest Influences. 75 PP*t 1951. 186 23. Needham, P. R. Trout Streams. Publishing Co., 1938. 24. Ramser, C. E. Runoff from Small Agricultural Areas. Journal of Agricultural Research. Vol. 34(9):797-823. 1927. 25. _____ and D. B. Krlmgold. Detailed Working Plan for Watershed Studies In the North Appalachian Region. United States Department of Agriculture, Soil Conser­ vation Service, SCS WHS#1, 80 pp. (November) 1935. 26. Schollenberger, C. J. Determination of Soil Organic Matter. Soil Science 59:53-56. 1945. 27. Smith, J • L. and Gr. A. Crabb, Jr. Progress Report on the Wooded Watershed: Michigan Hydrologic Research Project. Paper presented at Forestry Section, Mich­ igan Academy of Arts and Science, Ann Arbor, Michigan, April, 1952. 28. Storey, H. C. Forest and Water Research Project — Delaware-Lehigh Experimental Forest. Commonwealth of Pennsylvania, Department of Forests and Waters, Harrisburg, Pennsylvania, 44 pp., 1951* 29. Tennessee Valley Authority. Effect of 15 Years of Forest Cover Improvement upon Hydrologic Characteristics of White Hollow Watershed. Tennessee Valley Authority, Division of Water Control Planning, Hydraulic Data Branch, Report No. O - 5163, 74 PP. 1951. 30. U. S. D. A. Watershed Management Research Coweeta E x ­ perimental Forest, United States Department of Agricul­ ture, Forest Service, Southeastern Forest Experiment Station, 33 PP* 1948. 31. Watershed Research Aids Salt River Valley, UnitedTstates Department of Agriculture, Forest Servioe, Southwestern Forest and Range Experiment Station, 12 pp. 1947. 32. Research Progress Report, Influences of Vegetatlon and Watershed Treatments on Runoff, Silting and Streamflow. United States Department of Agrioulture, Misc. Publ. No. 397, 80 pp. 1940. 33. Quarterly Report of the Coweeta Hydrologic Laboratory, United States Department of Agriculture, Forest Service, Southeastern Forest Experiment Station, Third Quarter, July-September, 1949* Ithaca, N. Y.: Comstock 4 187 34. Quarterly Report of the Coweeta Hydrologic Laboratory, United States Department of Agriculture, Forest Servioe, Southeastern Forest Experiment Station, Fourth Quarter, Ootober-Deoember, 1950. 35. Annual Report 1951 — Intermountain Forest and flange ^Experiment Station, United States Department of Agrioulture, Forest Servioe, Intermountain Forest and Range Experiment Station, Ogden, Utah. 64 pp. 1952. 36. Wlsler, C. 0. and E. F. Brater. HydrologyJohn Wiley and Sons, Inc., 419 PP* 1949* 37. Yoder, R. E- A Direct Method of Aggregate Analysis and a Study of the Physical Nature of Erosion Losses. Journal American Society of Agronomy, 28:337-351* 1936. New York: APPENDIX Title Page 3atalog of Storms Having Maximum 30-minute Pre­ cipitation Intensities over 0.90 Inches per hour • • 189 3ummary of Evaporation by Months and Y e a r s ............ 190 Precipitation Summary, Standard Rain G-age N o . 20 . . . 191 Precipitation Summary, Standard Rain (3-age N o . 16 . . . 192 Precipitation Summary, Standard Rain G-age No. 67 • • • 193 Precipitation Summary, Standard Rain G-age N o . 21 . . . 194 Weighted Areal Precipitation Summary • • • 195 Runoff S u m m a r y ........... 196 3ample Precipitation Intensity Record 3ample Record of Runoff ............... 197 ............................... 198 3ample Calculation, Least Square Linear Regression Sample Calculation, Derivation of Standard Error Sample Calculation, Analysis of Covariance Soil Organic Matter Values Volume Weight Values Porosity Values . .199 . . . 200 ....... • 201 ......................... 202 .................................. 203 ...................... 204 Permeability V a l u e s .....................................2°5 Coweeta Hydrologic Laboratory Southeastern Forest Experiment Station Watershed No. 3 Catalog of Storms Having Maximum 30-minute Precipitation Intensity over > 0.90 Inches Per Hour Peak Runoff 31ST Before 1 Aug. 2 Apr 3 Jun. 4 Jul. 5 Aug 6 Sep 7 Sep 8 Nov. Period 1934-1939 22, 1935 2, 1936 12, 1936 12, 1936 24, 1936 20, 1936 30, 1936 5, 1938 Intermediate Period 9 June 13, 1943 10 July 5, 1943 11 July 30, 1943 12 Aug. 12, 1943 13 Feb. 17, 1944 14 July 11, 1946 15 July 15, 1946 16 Aug. 25, 1947 17 Apr. 8, 1948 18 Aug. 2, 1948 19 Aug. 14, 1948 20 Nov. 19, 1948 21 Nov. 28, 1948 22 May 22, 1949 23 July 10, 1949 24 Aug. 4, 1949 25 Aug. 20, 1949 26 Sep. 6, 1949 27 Sep. 18, 1949 28 Oct. 6, 1949 29 Oct. 16, 1949 30 June 3, 1950 31 Aug. 29, 1950 32 Aug. 30, 1950 33 June 12, 1951 34 July 15, 1951 35 July 16, 1951 36 July 27, 1951 .7494 1.5160 .8305 1.9060 2.4770 3.9050 0-1944 4.5720 6.8000 14♦1600 4.3300 2.4480 2.4260 4.0200 1.7220 5.0200 1.7220 1.8830 1.4260 4.3300 66.0390 4.4830 1.8150 1.7970 3.4010 2.4260 1.4005 2.1800 2.8100 3.2710 1.8780 5.1150 1.5160 4.500 Total Precipita­ tion inohes 15SST •*7grT Time Peak Type .07446 .07129 .03261 .06598 .03614 .08295 .10780 .16995 2035 0210 1450 1600 2010 1510 0105 0255 S M S S s M M M 1.91 4.10 0.91 1.83 0.94 1.85 5.85 4*67 128 .19898 191 .29594 398 .61626 122 .18845 46 .07181 69 .10654 68 .10558 113 .17496 48 .07494 141 .21848 48 .07494 53 .08195 40 .06206 122 .18844 1849 2.86609 126 .19511 51 .07898 50 .07821 96 .13253 68 .10529 40 .06095 61 .09487 79 .12229 92 .14235 53 .08176 144 .22260 43 .06598 126 .19584 1605 1805 1730 1715 2200 1710 1630 1330 0325 0600 1230 1330 0820 2300 1705 1810 0050 0505 1245 1725 0205 0035 2305 1605 1905 1301 1617 1615 S D S M D S S D I) D S M M M M S 1.35 1.29 1.83 2.46 2.24 0.76 1.73 2.33 0.88 3.81 1.23 3.52 3.85 1.17 2.88 0.57 0.35 3.03 1.63 3.33 1.81 1.12 1.73 2.18 2.31 1.75 0.57 — 48 46 21 43 23 54 70 110 s M M M S M M M M S S s 189 Iftorm^De scriptIon ration of Preolp.— Hra.-fro. Min. Intenffotal RB - MP “ St ora 10 Preolp. aitlee n/Hr. Min. From ?o" kin. Claaa Min. 15 ain. 20 aln. 30 mln. 60 mln. oa ¥0 25 25 00 2005 0525 1513 10 325 10 530 2025 1630 0600 0240 550 20 800 425 1850 1830 1900 2000 1705 1900 1530 0905 1000 1545 2015 1930 2322 1915 1805 0205 1603 1635 0515 2100 2100 1210 610 1239 1543 1605 0400 0420 2300 0010 1500 1617 1700 40 1930 1260 0925 193 1200 1435 75 1915 185 1325 770 1710 670 1530 180 250 660 335 2035 0240 1450 1600 2010 1510 0105 0255 65 1605 1805 1730 1715 1700 2200 1650 1710 1520 16 30 1305 1330 0255 0325 0030 0550 1205 1230 212 5 1330 0000 0820 2110 2300 1642 1705 1800 1810 0025 0050 1300 0505 1120 1245 1430 1725 0155 0210 0020 0035 2100 2310 1225 1605 1610 1905 1230 1301 1545 1617 1605 1615 20 1545 1350 0820 1655 5.70 3.90 3.00 5.58 2.64 4.92 2.94 2.70 5.24 2.60 2.80 4.52 4.23 1.95 2.40 4.41 2.25 3.72 3.90 1.50 1.60 3.32 1.58 2.54 2.68 2.12 2.58 2.00 1.86 1.40 6.42 2.64 4.98 2.40 4.64 2.56 3.92 2.24 1.63 3.04 3 .6 6 1.98 3.06 2.16 1.59 2.28 3.00 3.00 3.00 5.10 5.00 1.72 4.50 2.68 4.20 2.20 2.52 2.49 1.08 0.72 1.43 4.11 1.65 2.48 1.42 2.16 1.78 1.38 1.52 2.40 3.70 1.36 2.48 0.94 0.58 1.18 3*46 1.36 1.08 1.02 1.02 0.78 2.2 8 2.20 2.16 1.95 1.65 1.92 1.32 1.34 1.32 3.72 1.98 3.06 3.72 1.26 0.96 2.28 4.44 3.06 2.04 1.44 3.06 2.82 2.36 4.32 3.28 1.16 0.80 1.80 1.68 2.10 1.10 0.68 1.62 3.60 1.96 2.4 8 3.12 2.20 1.68 1.20 2.08 2.60 4.40 1.88 4.12 2.04 2.20 3.78 1.70 I.64 2.60 1.62 1.12 3.36 2.54 2.04 2.60 3.24 4.74 2.28 5.04 2.01 0.84 0.89 1.96 0.90 1.47 1.47 1.09 1.26 1.00 1.13 1.50 0.75 0.76 1.47 2.25 0.70 2.13 1.15 0.83 0.48 0.57 2.47 0.55 0.34 0.69 1.27 0.68 0.80 0.53 0.74 1.17 1.10 1.66 0.56 1.37 190 Goweeta Hydrologic L ab ora to ry United St a te s Forest Service e v a p o r a t i o n p r om u . s . we at he r BUREAU PAN AT WEATHER STATION 'NO. 1 M e No. 5.31 Computed by 0 . Curtis ll/l6/50 ChecKed by B. Rogers 2/51 Evaporation expressed in inches per month 3 1 191 Elevation 2425 Rise_____ 8?2 __ Azimuth___ 2150___ SlopeDistance 2291 7 • le No Co Yv ee f a H y d r o l o g i c L a b o r a t o r y United States Forest Service C o m p u t e d by STANDARD RAIN GAGE NO. 20 3 . 1 3 4 _____ ___ Monthly precipitation in inches Jan ?eb Mar' A p r . May Jun | JTul jAu^ ! Sep T, IOct Nov Qe _H »i ' r, A *+ i .a ^ ~ , ; . Z 5 | 4 W ?-77 !6.3? k h > 2.75 1./< /A35 f?S ■lJo \lt li 74. 743.4.7* m S . M ? 4.H M U 3 S.iz u w :\w\%i? j,% jji ' ■M l :0j4 tll-23 j J £ t s L i y i t a 4 l U < > , f , i 4 !U f J W L.747I/M 4; A * , 5-JT IA«5 !4d * . 5.77 5 % ;i.ZO /./3 ;0-iS i-A /-i?iz.(7|.U<..U«t m ±U7 __ 776 7.^; S37 if8t |?.4( ;7-?4 3.73 Ml /i*2 il41 #44 j m jrrm , I .3.fe|AT7. , . 4 5.S4;2.«I :5.57 io.n /z 72 ; SJfl 2-/7 f iU>oA s.a 'Aci ,'dl 45* : 7-tf.i/f : /.-A?. 3,70 2.34 y m \ s . n |$.4*. s-si ■2,1s y» .J .O .L.. rS;;^s;feiiar 13 3.91 , £zei3.^ 1 31 \i.?$ 11,13 3.M Sp I A/s , w \4H mi ,? t$r. $.sr xui4,u 63? A oo _8.,3q U sI ^ A 4* t1% . % ssjq .v A 14,4? £:rt\MS\iz.41 A 7 {7-5z ill 4.14 j,of A«S.ii7# m AC, d.ltjit- J0J3 .7JC 943 ,UZ \tu4 .U3lI*X fo.rif.il« {■<&,.23+ i n . t-DS 7si 555:Soy[m 4 28!Rf> 921 L;i M l M i _____ c h e c k e d b y _____________ _________ 7.E# 2855 655 170° gJ.erat.lon H»i _ Azimuth Slope Distance 2977 V ie No. ___ Tear Coweetc Hydrologic Lab ora tory United Stotes Forest Service C o m p u t e d by checked b y STANDARD RAIN DAOS NO. L6 Monthly precipitation May ___ ___ __ _________________ inches Jun Jul !Aug a I Sep u s u l 72)2 &Zil liU A M IS? m ^ m 3 1 0 , 1 5 1 , SAB 5-05 \ m !I2I 32pL- 2-4.4 4» A/S-U.98 Jl56.7,8? 2£l :.SS5 117 3,24; 4-24 4.72A M ,),28 1M mxUL m iUQ to it 3.o4 [3,27 1J6 XJ.S OlB 1\Z,?AB 7-?T 4.o4 j0 2 3:o4 $9{ K3t ?£z 4iti 773 4.*3\ j q j I A - n 1 IWSoo 1 $.72 4$i IAS ^ . ;i.g U 2 s $76 ti.52 t m 3-zc i*7. 1AI -+ 195 C o w e e t a H y d r o l ogi c L a b o r a t o r y United States Forest Service le No 3,231 C o m p u t e d by C hec ked by WATERSHED NO. 3 Monthly weighted areal precipitation in inches ter year Jw. /V5 mo mi Fob.| Mar TOTAL m 511 M.% 15.45 U k 8.90 (545 i,is Jul 5-o.jj m !a?3|2.% ! I 1944 - 431|5.o»ji-54 (.41\ l)5]l.t5 ’■IS] 3.to\ I,Si , $ 3 6 3.41 \l.5o \3.& lb.lt 3.09 116 £•». OS. J7S tot 1337;6.55. SOS, 9.9? 451 m 5.7) /./3 j.$l.1S5 llfl H-5o t.% i 4 o 4 | ..... ... 770: 2.53! 948 5.z4 8a4.?.s|> 1 .i2 Z I_ M U9 3.70! 1.9o\SM 5LV) m t 3.03 [1.01 y j2-f % 1 Q \ 1.18 1945... 14*5 Jun May SM {06 t i t 4.6+4.36 7.15 ,1.05 346 . r, 11.4V 4 /I'. t-St I0 .2 Z; LS&;4AZ , tv\ 413 9 W ■£08>4% jkis l it; 11.04 <111. IM . 2.89 i t) 151 \ m 9 1940 Apr 4ns *<(0i/twT74*».?tjT97!%so 1 4-/sj 4.15 ! 3.27 M i ,f9!4/4«t Jsi; S % . 4zf 531; 7471 ' " fl.8Sjiz.o4! 279 7.«[9!34 (.03 ;4 4 4 433[ S-T4 az? 7 .3 0 . 7,931 8-8/; 7.5424/ 155/ .4-77j ft/i/MI. 4®: 5-3zr S/4i+24Jo,?. 2 7,/t c‘s~~•^ July 16 51 V«’=S* ., ■S. Z$X5 July -ie •e'v . ^ee* _>TS CTo Ked L ~ ecc i) 1543 ~3J 15 S O M . 3 L JJ.SS3L /55 9 /60S . 1610^ /7L J tfi5L e J)8 M jf: .39 2.28 PefTCrliS !8)___ -31 \ 33 J Vc* iv„*¥i % -0 .. 2£u AJO .05 .05 j Qq EE M 3 L Rfr.-ii-5Ro- c-3geo^ec.D'o* . PltS . d /ls ___ re' Cof. ctc , “ Hc^r i '"CtiB 9*t Hf i '6' n " ,S' 16 J 2*1 _. -- u ond"^ens'ty *cr3ve"Ti"ne "'envois 20 5C s /f7 1.88 Vzo/s/ -.., y^/s/ 3* htZ l.ll 240 36 .Si, OeCfcCCDf '>ec-?dCy. Do'e Dee. r-te- z. v ---- -- -T - T - ~T T- r 4— i —I —1 — I I — M — t- H 4 - -i vaQ !tVN i N.I 4*v50 ^ 'n.of?? JK fYvSOs 50 C*S?-*fOrvl rfstr> ^ rr^cr>wr>fO — . IlX Ny On . , + ■ ■ I i-. i o: LT 3 ITi uri -» in O 5J 2 ^ 5j o '> & 4 i£> 5*1 ON CsJ'S Oi r>.r4 in< ini*3“^ <®'>*o;o 0;4''^iNo. ' i1® £ ^^"^ 5s O ^i, q q <1!^ ^ s ^ o i Io o 3 ' S A ^ 1 -. Q o < > 3 3 m 5 7 * 1 v j , | O',s. o- rsf -o U S ..,I .50 i;o4‘y-'|CX> Z j? '>0syJSW S s *s N O . tu: !oir3'° |2j, . «s{! ^ S>Jo e oO « 4 -t- 4 -J ^_2liCts. rv^jo^.*'» V-l tvt,rj tio ijl*4 -3 oi * *i » V " * "X_ N_ ”'*N» ^S*- 1— j““7 ] I ii§ M ‘X.^ S. *> 1 1+ WS i i i j ; H r\ l\ L W ; 1i 1 W i "x Vsj N ^ ^ i i ^ q :q 1 ^ 1 xf' S's, NJ. ^x, % ! .r 0 f >lN to tn Nv.! — ■-\- *0 in - V ? ^ sh n ^ 0(> (TNi 'I' tr Sn;sn CD K iol J f- ie 199 lefiST S q u a r e L i N m n % £ G R £ S S JOtf Coweeta Hydrologic Laboratory United States Forest Service tyRTenstieo h/o. z Com puted /9S+-/93? No. by C h e c k e d by /5-MiNvr£ />/r£^yRL OILS S /z j/si ________________ OiOO 8 >: ir m m X {.<>m - s u m tszu Ifxi y M z m A t i z*%£L Z-4Q6J* ;4 DFRIVAT/0 N OF 5 TfiHDfiRD /FRRoR 200 Coweeta H ydrologic L a b o ra to ry Un ted S t a t e s F o r e s t S e r v i c e (FROM Ff6kc£SloMs) mreesHtDWZ )- ie No. C o m pu t ed by ChecKed by - D/LS, A i/C. / ? £ / ____ ._______________ 15'MINVTF /MTtRVAL J S k I* s j-:ia:;ce V-S #3 relation - - 15 minute nrecinltntion rerio : - c iva! ". h ;cirfiiogic Laboratory ■ tt;o States c'orest Service r| ietv:een periods 10£ 0 - 19A 5 o nnn 1 DUS t/si. K'J 0, " 1991 Selected uniform summer storms h ;. v i n r maximum 30 min. precipitation intensity over 0.90 "/hr. ' W !l4(-;}5! w .* '/j % \ 22f. Mt5 rt$052. \uv. Mat ?€k •._•£ h$?s f. 3 iw-m$ MC-iyst Tomi \47/ : \tX$\ £ x.2 £ x-i U itzi >41.0^44 i H M \i * £x* , --i IP : .... £ty J90M H7M- !*95\ ,78/ SI1ill i'j : // ; 25 4- \ x M-£SLL f 3 Afcrf£?;/*rj f * .... t S6-34 i 4- I -V T . TsTfiL - . * H |H f--* - //iW ; I ** 7&2ft, 13 .4 ?' ?£ Soy^c*. „ 54 X2 4.(3 4 /4