N. B, The authors regret thac many errors escaped correction in the final proof-reading, and also that earlfer editing of the typescript failed to delete fractured syntax, dangling modifiers, other objectionable items, or to repair faulty grammar and mis- placed punctuation. The reader's attention is called to the following partial list sf errors in the hope of clarifying the intended meaning of the topics discussed:
Po X In the captions far Figures 52 and 53 and at several places in the report Horner Ford appears as Horner Foard in error, p. xvi Second pa.ragraph, line 6, change "acquifers" to "aquifers".
Map following p. xx Creek in lower left of map is Horner Ford not Harper Ford.
Po 2 Line 3, insert a cormna after "'channel". p. 42 Second paragraph, Pine 2 from bottom, insert semicolon after the word "error" in place of coma. p. 48 Hyphenate 'Thirty-year" in title of Table 5 and correct last column heading to Ratio of Net Change: Total, changing acres to %. p. 85 Caption for Figure 26, first paragraph, last line substitute "than" for "then". p. 104 Line 9 should read "..., mountain hardwoods, Merrick and Johnson (l952)." Line 10 delete "and'bat start of line.
Third paragraph, line 3, insert "comparing" after "even".
Line 2 of footnote should begin ""iclearing land, ..."
Line 13, change "or" to "of".
Second paragraph, line 5, insert "stand" after "aspen".
Line 13, change "absorbed" to "adsorbed".
Picture is on its side and reference to Figure 60 in caption should be to Figure 59.
Line 3 from bottom, insert "preceding" at end of line.
Line 3 from bottom change "oft'to "or".
N. 3. The authors regret thar many errors escaped correction in the final proof-reading, and also that earlier editing of the typescript failed zo delete fracsured syntax, dangling modiffers, other objectionabPe items, or to repair faulty grammar and mis- placed punceuatfsn, The reader's attention is called to the folPowing partial Efsr of errors in the hope of clarifying the intended meaning af the copies discussed:
Po X in the capcfons for Figures 5.2 and 53 and at several places in the report Warner Ford appears as Rorner Poard in error, p. xvi Second paragraph, line 6, change "acquif ers" to "aquifers".
Map following p, xx Creek in lower left of map is Hsrner Ford not Harper Ford.
Line 3, inserc a coma after "chailneIs',
Second paragraph, line 2 from bottom, insert semicolon after the word ''errorq' in place of crsm.a.
Hyphenate "Thfrty-year" in title of Table 5 and correct last colurm heading to Ratfa of Net Change: Total, changing acres to %,
'\ Caption for Figure 26, first paragraph, East Pine substitute v'than'q for 'ktherr",
Lfne 9 ehculd read ",,,, mountain hardwoods, Merrick and Johnson $19%2)," Line 96 delete "'and" at start of Sine.
Third paragraph, Pine 3, insert "'comparing" after "even".
Line 2 of footnote should begin "in clearfng land, ..."
Line 3.3, -hange ""ors'to r9~frr0
Second paragraph, Xime 5, insert "standrs after "aspen".
Line 13, change "absorbedqBto
Picture is on its side and reference to Figure 60 in caption should be to Figure 59.
Lfne 3 from bottom, insert "preceding" at end of line.
Lfne 3 from bottom change "sf" tts "or"',
THE UNIVERSITY OF NORTH CAROLINA WATER RESOURCES RESEARCH INSTITUTE
Office of the Director 124 Riddick Building North Carolina State University Raleigh, North Carolina, 27607 Telephone: 919:755-2815
September 5, 1972
TO: Wi3Ok1 IT MAY CONCERN
lc'. FROH: David H. Howells, Director , , ,j (.-!.,, ,+,i? +
SUBJECT : Institute Report No. 71--"Effects of Land Use on Municipal Watersheds," by Drs. T. Ewald Maki and William L. Hafley, School of Forest Resources, North Carolina State University
The attached report covers a three-year study of the effects of land use on the Lake Michie watershed and its effects on water quality, quantity, and timing of flow. Lake Fiichie is the sourco of water supply for the City of Durham. It will be of interest to planners, engineers, foresters, munici- pal officials, and regulatory agencies concerned with water supply and pollution control.
The management of municipal watersheds to maxin~izewater quality and quantity is an essential part of the water supply process. The recommendations on pages xix-xx warrant the attention of all persons with related responsi- bilities.
Attachment
THE UNIVERSITY OF NORTH CAROLINA is comprised of the sixteen public senior institutions in North Carolina
UNG-WRRI- 9 2 - 7 1.
EFFECTS OF LAND USE ON NUNICLPbL IqATERSHZDS
By T., E, Makf Behenck Professor a£ Poresery School ot Fetest Rzssurces Narrh Carohinh Ssace Unfversfry RaPefgh, North GaroElna 27607
W, L, Hairley Associate Brofesssr of Forest Blonetry and Statistics Departments sf Foresrrg and StarLstics Norzh CaseHfna State Uaxversfty Raleigh, Nc~%h Casobina 27601
The work upon which this publication is based was supported in part by funds provided by ehe Offxca of Water Resources Research, Department of he Interior, chrsugh the Water Ressurees Research Instlaace of the Universxty of Nsreh Carolina as authorized under the Warer Resources Research Act of 1964.
Project No, B-OEO-NC Agreement No, 14-01-0001-1934, P, P, 1969
The authors are deeply indebted ta many indivfduals and agencies far invaluable assistance and cooperation on various phases of this project, Major conrr~butionswere made by
Graduate Research Assfstants Davfd D, Chapman, Soop 8; Faber,
Harry J. van Lsoek 111, and Charles P. Wflliams ehrough research on specific phases of the project, and to Graduate Srudents
Dale K- AElison, Setyono Saserasumarto and Thsmas 1, Vermillion for data compilations and analyses; by School. Poresc Manager
Laurens G. Jervis fcr reservoir sedimentaefon probfng and stream gaugmg of small watersheds; by Professor J, 0. Lamxi for assistance in securing necessary aerial photo sets and for critical review of the manuscript; and by Professor
Charles W. Welby of the Department of Geosclences :or gufdance and help in csnducelng and rnterpretfng the seismic survey of
Lake Mfchie.
Further, we offer graceful acknowledgmene to Professors
Charles B. Davey and D, L. Holley for a searching review sf, and valuable suggestions sn ehe manuscript, Professor D. B, J.
Steensen for suggestions during the PnirfatFnn of the project,
Professor J, W. Duffield for his photographic csntriburions in addftion to critical review of the manuscript; Wade Brown,
Director of the Department of Warer Resources, City of Durham, and to his staff for invaluable cooperation in supplyikg
background infsrmarion and data on the operation of the
water resources facility; the Recreation Department of the
City of Durham for making equipment available for reservoir
surveys; and the Soil Testing Division of the North Carolina
Department of Agriculture for chemical analyses of soil
samples. The assistance of Professor David H. Howells,
Director and Mrs. Kfnda Kiger, Administrative Assistant
of the Water Resources Research Institute in facilitating
all aspects of this Investigation was invaluable,
Finally, without the dedicated help and cooperation of
all who were associated with the project, especially at the
typing and reproduction stage, the compfetfon of this report would not have been possible. ABSTRACT
A study of ehe Lake Miehfe Watershed, the ssurce of water for the mvnxcipalfzy of Dxrham, Norch Carolina, is reporcede
The study was concerned with evaluating the effects sf fand use on water quality, quantity and timxng of flow durnng the
43-year period since the establfshmen~sf the Pske,
The characterfstrc csmposftion af the solls and the vegetarive cover and %hen influence on the hydrology of the besm are discussed,
Through the interpretat~snsf aerial photographs,
Agrieulrurai Census Data and other historic doeumenzs, ehe nature sf fand use changes wfthfn the watershed have been traced, An fn~erpre~atasnsf the influence of these land use changes fs offered.
Current status of the watershed is detailed and recommendations presented for improving the quaEfzy of the water and the general hydrologic eharaeteristics of rhe watershed.
The study highltghts the problems encountered by a
Water Resources Department that must manage water avafXab$lity on a watershed where LC is not the major Land om-ier. TABLE OF CONTENTS
RECOMMENDATIONS a = XZX
CHAPTER 1
INTRODUCTION . , , . . , , . . , , . . . 1
General . s*-,&oe -*seo8 1
Idhat makes Chis watershed of anterest znd concern?, 2
CHAPTER 111
CONDITION OF THE WATERSHED PRIOR TO 1926 31 TABLE OF CONTENTS (csnrinued) Page
The role of forest fires 35
Resnlts 42
What do these data rmpfy about the watershed , 47
WATER QUALITY, QUANTITY, AND TIMING 59 About wacer quality . . 59 General, . 59
The water ssurcq and the nature of data , 50
Results and analysis of data , 61
Zlmitarfons ro quality control , , , . , . About quantity or yield . 84 General. . 84
Flat River discharge , , 88
About timing and ySeEd 9k
Generalo 94
Station locations and available data 94
Analysis sf stream-flow trends , 96
Prsepeets for POW-flow augmentation a 101 TABLE OF CONTENTS (continued)
Trade offs and economic implications . . 104 The forest floor and infiltration capacity . . 104 Is pine to hardwood conversion justified? , . 106 Are pines devoid of hydrologically beneficial characteristics? ...... El3 What price hardwoods , , , . . . , . 116 What about grass as an alternative . . . . 122 Some examples of stormflow quality and timing . 925
CHAPTER VI THESEDIMENTISSUE ...... 132 General. . , ...... , 132 Recent probings and observations...... 133 Seismic probing of Lake Michie , ...... 148 Character of surface sediments in reservoir . . . 165 A look at the road system . , . . . , . . . 170 Direct estimates and observation of road runoff and sediment transport ...... 174 Reduction of pollution from roads . . . 176
CHAPTER VII FUTURE OUTLOOK FOR THE LAKE MICHIE WATERSHED . . . . 185 General, , , . , , . . , . ,185 Trends in rural land use , , . . , . . 185 Trends in forestry , . , . , , , . 187 TABLE OF CONTENTS $ecarinued)
Page
Urban growth and rurban aneroachment. a 19 E
Management alternatives . a 19 3
EXPERIENCES IN THE MANAGEMENT
OF MUNICEPAL WATERSYEDS. a a 198
Generalz $*. . a 19 8
The Elf Whitnay Forest a a 19 8
The Amherst and Mount dslyoke Watersheds a 19 9
The WaynesviPle Municipal Watershed , 201
SearcLees Cedar River watershed . a a a 20%
Other examples . * 204 LIST OF FIGURES
Page
1, Map of the Warerehed . . . . , . . . . : . 8 2. Evaporatfbnvs.Precfpitatfon . . 22
3. Virginia Pine CoIonfzarion , , 23
4. Upland Hardwood Stand . , 25 5, Natural LobSollyPine Stand. , . . 26 6. Natural Virginia Pine Stand . , . 28 Abandoned Farmstead , ,, . . ForestRecovery fremErosion . Land Use Perceneages . . . . . Major Land Use Trends . . Post Logging Conditions . . , . . .
Residual of Pine Removal by High-Grading . a Weathered Road Cut , . . , . . . Eroding Drainage Dirch . . . . . DirectDrainage to Live Tributary . . . Plot of Turbidity, 1931 - 1968 ......
Monthly Water Consumption, 1931 - 1968 , a
Purification Effort in Lbs. of Alum per PO00 Gallons.
Purification Effort in Lbs. of C1 per PO00 Gallons Farming Practices ...... Hogwallow , . , ......
Cattle Watering Access , , . . . a Resfdual Stand After Harvest of Pine . . . . 24, Waste Disposal in Drainage Ditch . , . 8f
25, More Waste Disposal a , , a a 82 .
26, Flat River with Storm Flow , . 8 a a . 85
Streamflow inCFS-Days, 1926 - 1968 . a a
PfneRootSystems, . . a
Weirs on Sub-basins of the Watershed. * a
Hydrographs Demonstrating Stormflow Recovery , a . Comparison of Hydrographs - Forested vs, Non-forested
Headwater Mud Ba~fn . . a a
Shoreline at Bahama Bridge . . a a 34. 1968 Probing Profile , . . 136 35. Flat River at Head of Lake Michie - Low Water. 138
36, Sediment Bar at Headwaters of Lake 139 37, Drfed Mud Flat at Mouth of Dial Creek . . 142 38, Dried Mud Flat at Mouth af Muddy Creek . . 3.42 39, Dial Creek Mud Basin , . . . 143 40. Lsw Water Vfew from Bahama Bridge. . . . 144 41. Drfed Mud Flat at Headwater, , . . . . . , . 146
42, Uncharted Island 147
43. Profile Locations , . . . I 150 44, Profile of Traverse Line P . , . 151 45. Profile of Traverse Line 2 , . . . . 152 46. Profile of Traverse Line 3 . . 3.53 47, Profile sf Traverse Line 4 . , , . 154 48. Profile of Traverse Line 5 , . L55 a 49. Profile of Traverse Line 6 ...... 156 50, Profile of Traverse Line 7 . , . . . , . . . 157 51. Profile of Dial Creek Probing ...... , 158 52. Profile of Horner Foard Mouth ...... 158 53. Profile of Horner Foard Upstream ...... 139 54. Profile of Right Fork of Dry Creek , . , . . . 159 55. Profile of Right Fork of Dry Creek at Confluence with Left Fork , , , , , . . . . 160 56. Profile of Dry Creak Downstream from Confluence . . 161 57. Profiles of Right and Left Forks of Rocky Creek . . 162 58. Road Drainage Practices ...... 171 59. County Road After Maintenance Activity . . . . 172 60. Drainage Ditch from Newly Maintained Road . . . . 173 61. Sediment Transporting Drainage ...... 180 62. Forest Access Road , , , , , , , . . . 182 63. Forest Stream , , , , , ...... 183 64. Population Growth, Durham County and City . . . . 188 65. Population Growth, Person County and Roxboro . . . 189 66. Renovated Wooded Area , , , . . , . . , . 210 LIST OF TABLES
Page
Scream Mileage of the Watershed , ,
Precipitasfsn in 10 Weetest and Dryesr Years
Farm Psnd Trends . a a a a
e Trends , e e . Land Use Changes on she Dial Creek Watershed
Ranges af Monthly Turbidity Data a -
Chemfsal Characteristics Observed a a a
KonthPy Discharge of Plat River . a a
e from Warershed in CFS per Square Mile
Stand Table for Upland Hardwood vs, Virginia Pine
Reservoir Storage Capacf~yData . a - Average Sediment fn Selected Drainage Sasins
Analysis of Sediment Samples . a a
Comparison of Sediment and Forest Floor Sample Analyses a a
M~rphologyc~f a Filter Strip . a a . SUMNARY AND CONCLUS EONS
An investfgatfsr; was undertakeiz, beginning in the summer of 1968, to determine effeees sf land use on water quality, quantity and nmfng of flow from the Lake Miehie Wazershed which 2% the soutee cf water for %he municnpaiity of Durham,
North @ars1;naC The watershed wrnprfsee -168 sqasrs miles,
Paeed by 296 mIPes of stream channels cf PSar BSver and fts
reservoir behind an 81-fssc high dam campicted in 1926 near
Changes in land use vere deserrnfned chlefEy from interpre- taefon sf six eats sf aerfal photographs, beginning with ehe ffrst flfghr in 1938, and endmg in 1976 A study in 1936-33 by the Soil Gonservatim Service (Martin and Bass, 1940) of erosion and refa~edland use condftfons furnished valuable base point fnfsrrnattm for rhfs inves~igation, Enformation from the aerfal phsrographs was suppfemenzed by intensive ground reeonnalssance sf portions of rhe drainbge basin and by Agrtcultural
Census Data for Person Csun~y,North Carolina. Extent sf reservofr sedirnen~ationwas determined by direct probing and by a seismic survey, Water quality was assessed from Eongeime records rnafn- calned by the Department of Warer Resources of Dnrham, and trends fn streamfPsw were evaluated from records sf Flat Rlver and DfaP
Greek dfschasge since 1925 gsthered by rhe USGS, The inves~igatfsahighixghts the probierns encauneered by a
Water Resources Department thee must manage water avafiability,
on a watershed over which Lk does not exereioe aoncrol, Sugges-
tions are offered for allegialfng the eoadd%8ons rhae now exist,
The prinefpal findings from thfs fnveseigatian can be
tabulae@$under five headings: land use changes, water quality,
drainage basfn dischargep the of BUT~BOW~and the sedimentation
Land use ehtmges:
1, Since the :nM-1920's there has been a net increase of
almost 15 percenr in the acreage of fouesc land wh%ch now
occupies 6%,5 percent of the 167,200 acres of land within the
drainage basin, However, mast of zhfs increase occurred prfor
to 1938 when the first sat of aerfal photos 05 the watershed
were taken,
2, Cultivated acreage exhiblts deeadal fluctuations but
by 1970 it occupied nearly 25 percent of the waeershed, Pasture
and PdPe land have declined slightly to about 16 pbrcent of the
entire drainage basfn,
3, Road mileage has fncreased 15 percent in 40 years during which mileage of paved roads tripled, The surface sf the exlstfng
road system now preem-pcs over two thousand acres of the watershed. 4. Farm ponds have increased from essentially none in 1930 to 620 in 1970, or approximately four ponds per square mile of watershed. Total area in farm ponds exceeds 1,270 acres. 5. Changes in the proportion of different uses do not reflect the actual activity that has faken place during the period of study. As an example, in the 3,384-acre Dial Creek drainage, the pine acreage dropped from 1,115 acres in 1940 to
596 acres in 1970, suggesting a net change of 519 acres.
Actually, 714 acres moved out of the pine type and 195 acres moved into it, resulting in alterations involving over 900 acres of pine forest type. It would appear that activity rather than net change in land use should be a major consfderatfon in the evaluation of the hydrologic condition of watershed lands, 6. Also in the Dial Creek watershed, the hardwood forest type increased from 535 acres in 1940 to 1,121 acres In 1970 showing an increase of 603 acres in this type, but only 17 acres maved out of the type, indicating the basic stqbilfty of this type under the existing mismanagement of forest lands within the drainage basin. In the absence of meaningful forestry effort, except on the Hill Forest of the School of Forest Resources, the forests on the Lake Michie watershed are gradually being converted from pine type to junk hardwoods. Water Quality:
5 1. During 40 years, no trends were detected in the levels
of pH, dissolved oxygen, alkalinity, nagnesium, sodium and
potassium.
2. Calcium, sulphates and chlorides did show a slight
increase in level as did hardness, which exhibited an increase
in variability aver time.
3. Turbidity has tended to decrease, a decrease which
has been accompanied by a considerable reduction in variability
over time. The suggestion is that there has been a stabilization
of conditions on the watershed reflecting improved farming
practices and a larger proportion of the watershed under forest
cover. This trend is evident despite the continuing sediment
transport into streams from the road system.
4. Road maintenance practices, solid waste disposal,
current level of livestock husbandry, and general public use
contribute substantially to degrading the quality of water.
Water Quantity:
1. Mean discharge by months over a 43-year period has ranged from 2,000 -cfs in the low month of October to 7,700 -cfs in the high month of March. These values correspond to mean daily flow rates of 64 -cfs and 248 -cfs, respectively. 2 Tho Pswesf mean daiiy dsscharge sf 4 occurred in
January 1934; and the highes~mean daily dfsckarge sf 760 -cfs occurred in January 1936<
3. Of rhe 131 b%lXlon gallons of wafer thar Ps delivered as precfpitarnon to che 107,208 zcres of watershed sn an average year, nearly 36 billion gal.1sns show 3s ourflow in izhe average year, Under normal csndfrions, it ns estimated that about three-fourths sf the a~nuaPprecapltat-rsa 1s Poet rhrough evapotranspiration, is stared fn dsep acquifers, or leaks out of the basin-
4, Even during the drfesr years on xeeovi sfnae the early
P920fs, rhe discharge from she P%8c River drainage basfn above
Lake Mfehfe,coupled with che exlstfng resesvofr storage capacrty of over 1l,000 acre-feez, has stayed far in excess of Durham's needs, and should remain adequate for several decades under present or Pmprsved water management,
I, The over-all trend In discharge from the Plat River drafnage fn the Bast four decades had been $omward, and the amplitude of peak high and low flows has been reduced,
2, hugrr,eneatfon of 2sw flow on an annual basis 4s possible eo rhe exten"sePf nine inches sf extra wtaer; however, the current level of needs of the municipality relative to the current level of outflow does not call for such augmentation.
Any vegetative manipulation for low-flow augmentation would influence the annual yield, much of which would be lost in overflow.
3. Under a program of organized forest management, an annual yield of one billion gallons of extra water could be realtzed while reaping the benefits of about half a million dollars of stumpage receipts for the watershed landowners £ram annual harvests of wood from an estimated area of 1600 acres each year, 4. For improving percolation, deep seepage and detention storage, retention a£ pines on the watershed rather than allowing conversion to hardwoods is favored because of more effective amelioration of subsoil by the root systems of the former species group,
The §.e.diment a t ion f ssue :
1. Based on measurements of sediment deposits in the reservoir, the rate of sediment production from the Lake Michie basin has averaged better than 266 tons per square mile of watershed per year, or very nearly the same rate as the ''muddy" Mississippi. This rate is based on direct sediment measures, hence, is exclusive of material last in aver f low. 2. Since completion of the dam in 1926, between 10 and PI pereent of the original capacfty of the reservoir has been lost to sediments,
3. The road system appears to be the prfncfpal source of current pollution from sediments.
4. Sediment polEutSon eould be reduced considerably by fmproved road maintenance and soil stabilization practices, especially with regard to drainage. RECOMMENDATIONS
To Improve water production and water quality from the
Flat River drainage basin, or from any comparable watershed, the measures listed below would be required. It is recognized that the nature of the measures has been influenced by the specific pattern of land ownerships within the Lake Michie watershed, but the measures would apply equally to other
municipal water systems where control of land use resides beyond the authority of the water manager. Several recomrnen- dations are appropriate to any watershed, while others apply directly So watersheds that have ownership patterns similar
1 to that of the Lake Michie watershed. 1. Improve road maintenance practices, stabilize
cuts and shoulders and direct road drainage through filter strips rather than directly into a stream.
2. Obtain control over livestock access to streams, particularly to prohibit the establishment of hog wallows astride live streams.
3, Develop controlled waste disposal sites at safe distances from drainage channels of the stream system.
4. Encourage land use conversion practices that take maximum advantage of soil improvement provided by decaying tree stumps which aid in deeper percolation of water into soil. 6 5. Acquire, or obtain control of, well-vegetated filter strips adjacent to all live streams and the lake. 6. Establish stands of loblolly and shortleaf pine along the lake shore to create a more trafficabfe margin around the lake.
7. Encourage better forestry practices on the part of the land owners as a means of obtaining favorable hydrologic conditions as well as creating an economic resource for the land owner.
8. Hire a professional forester to promote forestry activities on the watershed and to act. in an extension capacity to the land owners,
9. Construct a stilling pond at the head of the lake to control further movement of sediment deposits into the lake.
Chapter I
INTRODUCTION
"Search for TRUTH is in one way hard, and in another easy, for it is evident that no one can master it fully, nor miss it wholly, but each adds a little to our knowledge of Nature, and from all the facts assem- bled, there arises a certain grandeur."
---8ristotle (Circa 330 B.C.)
Municipalities in the Piedmont province of North Carolina, in general, are heavily dependent on reservoir storage for their warer, the supply of which is maintained by streamflow from gpeclfic drain- age basins or watersheds. The quality, timing and amount of streamflow issuing from given watersheds, thus, should be a focal point of concern, for these elements of streamflow are influenced by the use, treatment and types of management of all lands within the watersheds,
Streamflow, of course, is maintained by precipitation, It is recognized that of the total precipitation fafling on a watershed, a certain amount returns to the atmosphere directly through evaporation or is used by the vegetation; a portion is stored within the basin, at least temporarily; depending on geologic formations, a portion may be transferred -vfa deep seepage to another basin; and the remainder reaches the channel of the stream mainly either by overland flow (i.e., surface runoff), by subsurface flow (i.e., flow through the soil mantle) or both,
Stormflow, or the observed increase in volume of flow associated with a given storm, is always a resfdual, something less than gross precipita- tion. What happens to the total precipitation received on a watershed, how much of it reaches the stream channel, how soon, and in what condi- tion actually depends on many things; on the amount and intensity of the stom, on season of year, on antecedenE weather and soil ccndltions,
on the shape and size of the wakershed, on kind and depth ot sosE, on
kind and position of aquifers intersected by the stream channel cn types
and condition of vegetative cover, and an intensity axd t.yFes 0T land
use, which determine the condition of the land within the watershed,
Man exercises direct control only on rhe charac.ter an6 cc)ndition
of the vegetation and the character and condftion or' the land s:~~faee
resulting from his allocarion of uses, yet thfs limited contrd can
influence streamflow, espeeiaHiy its quality, in a major way, Be can,
and does, alter the character, amount, and condition of vegeta~ionby
cutting timber, clearing forests, burning, and planting d%fferent crops;
he can, and does, modify the surface of rhe land by plowing, cultivating,
grazing of livestock, constructing roads and buildings, erecting im- poundments, and through related activities, So from a hydrologic. and water production standpoint, it is relevant to examine the extenc to which man's activities and use of the land within a wzterehcd can be directed and modified to improve quality of streamflow, conrrol cimfng, and increase the quan~ity,and in these contexts go determine what effects the uses of land have had on the water supply of a municipal-
ity. The present study of effects of land use involves the watershed of Lake Michfe which is the principal source of water for the City sf
Durham, North Carolina.
What makes this watershed of interest and concern?-- First, in history, physiography, land use, and in past planning, fc 1s rather
typical of watersheds for many Piedmont municipalities. Zr is also large enough to be significant, comprising over 107 ehousand acres, or nearly 168 square miles,
Prior to incursion of white settlers, about 1740, the land within
the watershed had been subjected eo a variable buc unknown amount and
extent of disturbance by Indian occupancy and use, Today in many fields within the basin it is still not unesmon to find arrowheads and other
artifacts attesting to use sf the land under a different culrure, and
quite likely under a less violent treatment of the resources sf forest,
soil and water.
Since 1940, the early prevailing practice of field rotarion has
resulted in most of the Lake Michie watershed lands being cleared for
field or pasture use, or both, at Least once. Crop rotation and ferti-
lizer use have reduced the intensity and extent of field rotation, but
have not eliminated the practice completely, as will become abundantly
clear later from an examination sf changes in this drainage bash within the Past half century,
Advanced plannfng for augmenting Durham" water supply started
even before the turn sf the century, but the decision to construct a
dam on the Flat River creating Lake Mfchie was not reached until shortly
after 1920. The planning was typical of the process engaged in by many other munfcfpalities~ Careful consideration was given to inputs of time and of large amounts of capied for engineering works, but the land that was to provide for water production appears ro have remained out of focus, Water is a free good, at Peast according to the riparian
doctrine, and the concern during planning has been merely to make sure that the strearnfLow is satis5actoey in qualfey and adequate in quantity,
So, starting with known and anticipated needs of consumers, Durham built a system to satisfy those nerds, a system that. went all the way hack to the land, bur wirhout incorporaring the basin lands as a part of the system. The p~sslbleeffects af actfvities on the watershed and changes in land use appear not to have been speciffcally evaluated; to be sure, an early concern about the extena: of sedimentation was expressed, but after the initial ten years of operation, the siltation was adjudged not serious. Up to now the system has worked in an emfnently success- ful manner.
Since I926 when Lake Michfe began supplying water for the City of
Durham, several evenrs and forces have influenced the nature and mag- nitude of changes occurring on the watershed. There was a major dfs- ruption in economic activity induced by the Great Depression, assumed to have been accompanied by relatively extensive colonizatfon of fields by Virginia pine, particularly the badly worn-out ones. There was a major war from 1941 ta 1945 accompanied by accelerated timber cutting under the stimulus of the Timber Production War Project which unques- tionably had some impact on the forested acres of the watershed. The
Soil Conservation Service came into being in the early years of the
New Deal; since I935 it has had an increasingly larger impact on land use, principally in the improvements it has effected in field and pasture practices. From 1933 to 1941 the Civilian Conservation Corps and the Works Project Administration were also active throughout the nation, but because of rhe dominantly private ownership of lands within the Lake Mich5.e watershed, these agencies saw only limited use within it, and their impact is assumed to have been very small,
The Agricultural Conservation Program (ACY) and the Soil Bank Program have influenced land use within ehe watershed, but with very minor exceptions, their effect on forestry and forest Ian$ use has been negligible,
As the prfnci.paP 'source of water for the Cxty of Durham, the
Lake Michfe watershed, perhaps, has been more a topie of general inferest than of genuine esneern either for immediate water needs ar for perpehuatfon of an ample future supply, The large storage
t capacity sf the reservoir and the abundant basin outflow have doubtless conrributed to any complacency which may have existed in the past. Over a 40-year period, the diurnal stream discharge of
Flat River and its tributaries as recorded imsledfately below the
Lake Mfchfe dam has averaged nearly 160 Qcu'ofc feet per second), equivalent to more than one hundred million gallons per day.
In the meantime (since 19261, the population of Durham has tri- pled, probably exceeding PO0 thousand by now (95,40 in 1970), but it is still consuming not much more than ten percent of the average stream discharge.
Prior to 1926, Durham supplemented its water supply by pumping from the natural flow of Flat River. Without a srorage facility, it became clear that the natural flow would not suffice during drought periods when the daily discharge could fall far below ten -cfs, To pro- vide adequate storage of water and to generare power for rnunicfpal use, constructLon of the 81-foot dam was begun in June 2924 and completed in May 1926 (Anon. 1966). The 507-acre lake with an initial storage capacity of 12,631 acre-feet (equivalent to about 4.13 billion gallons) attained spillway level in December of the same year. Since that time,
Durham has not suffered from water shortage; in fact, during droughts, as in late summer of 1968, she has supplied water out of her "largesse" to neighboring communities that were in distress.
With this general historic background as a frame of reference, land use within the Lake Michie watershed will be related fn a more specific way to water quality, water quantity, and timing of flow.
In planning this study, it was postulated that detectable changes in water quality, as well as in timing and quantity of outflow from the upper Flat River basin could be related to changes in land use of the terrain above Lake Michfe. It was also postulated that there had occurred rather a clearcut trend from predominantly agricultural use to forest and that the trend would be reflected in the existing character of the vegetative cover on the watershed. As the results will show, neither hypothesis turned out to be completely tenable. hong ehe sajior physfcal features of any watershed that are rele~
vant in an evsPuacron e? its wazer production potentialities are its
size and shape, stream pattern, length and arder, topography, geology
and soils, the character, amount and disrribution of precipitation
that falls on ii, the conditaon, extent and iloristfc composition of
the native vegerstive caver, and extent of land use for farming and
other pursuiLs. AIL of these features aa~d actfvikies have some bearing
on the amount, &im:ir?g, anid quality of streamtlow ulrfmately reaching
the reservof r, wcfc!: in chis case is Lake MichPe.
Size and The Lake Michie watershed, located in the north------"A- sha~e,
central part of North Carolina, is approximately 168 square miles in
extent, comprLslng roughly 107,208 acres. The headwaters of Flat
Rfver and its tributaries whch generate the streamfPow for the re-
servoir form a Sroad, fan-like area, some .6.wen%y-fivemiles at its widest psfnt and absuz twelve miles in depth, naurowhg abruptly to
a "corridor" less thsn Pom miles wfde along the lower three miles
of the Flat Rxver channel above the irnpo~ndment (Figure I). Less
than fifteen percent of she wauershed area is in the "corridor",
leaving over ninety thousand acres in the broad, fan-like portion
of the wzters'rred. 'E'he Isgdroiogic signiffcatace sf this shape resides
in the fact thac a very large proportion of the precipitation ulti- mately reaching the charnels of rhe system falls on land at a sub- stantial distance from irs desfred seorage point, and is forced to
traverse over and through a variety of condieions before flowing into
the reservoir,
Partern of the stream system. The geologrc structure and origi- P naP land form have influenced the development of a deadritic pattern
for the stream system above the reservoir. The domfnsnt features of
the system above the lake are the 9.7 miles of the main channel of
Plat River, which is a resrrlt of the convergence of the North Fork
and the South Pork of Plar River, each fork extending over 18 miles
toward zhe upper reaches of the water divide sf the basin (Table 1).
Thus, there are some forty-five mfles of main channel and about 83 miles of channel in named creeks and branches. In addition, there
are over 968 rafl.es of mapped channel in ffrsc and second order streams which do not show names on any avaflable map, In brief, channel mfle-
age in perennial streams of the headwaters af Flat River and its trib- utaries adds up to more than 296 miles, i,e,, fn streams above the dam.
Obviously, with this mileage in perennial streams, the basin is
laced and thoroughly dissected so that the maximum distance that water needs to flow before reaching a 'Qiive" channelkin any season of the year probably does not exceed one-half mile. The total mileage fndi-
cates that for every square mile of watershed lands, there is an average of 1,765 miles of perennial stream channel,
The entire stream system in this drainage basin has reached a
fairly "steady state." Headward erosion of channels is negligible.
Mass wastage is likewise nos pronounced as can be dearly observed by the infrequent occurrence of toppled crees along the banks of
the main streams, even after high stormflows. Hence, any major Table 1. Stream mileage of named channels in the Flat River drainage above the Lake Michie dam.
Name of Stream Length fn Miles
Flat River (main channel) South Fork of Flat River North Fork of Plat River Deep Creek Richland Creek Bushy Fork Creek Aldrfdge Creek Efttle Creek Rocky Fork Branch Lick Creek Byrds Creek Dial Creek Ellis Creek Chappels Creek Quarry Creek Burton Branch Mud Branch Rocky Creek Dry Creek Dry Creek (No. 2)
Total mileage - named streams Minor perennial streams
Total stream mileage sedimentation, or sediment pollution affecting water quality of the reservoir must be attributed to sources other than the natural forces which might be expected to continue enlarging the channels of the system.
Geology and soils. A11 of the watershed lies within the Fied- mont provfnce. The predominant soils have been derived from light- colored Carolina slate, with soil types bearing the familiar series names of Georgeville, Herndon, Orange, and Alamance. This soil group comprises over fifty percent of the watershed. Next in prominence are the Appling, Cecil, Colfax, and Durham sofPs which are derived from
coarse-grained granite and other crystaElfne rocks, and occupy over 18
percent of the area" These sofPs are confined to the northeastern part
of the watershed. Daivfdson, Iredell, and Mecklenburg soils comprise
another 18 pereenr of che area, and are found in the northwestern part
of the watershed. These ssFls are derived from diorite, gabbrs, and
other basic crystallme rocks. Over 12 percent of the watershed soils
fall in the Pelena and Wilkes series and are derpved from mixed acid
and basLe crys2alline rocks, including aplites, diorites, gabbros, and
granites, The Helena soils are the more abundant in this group and
are regarded as excellent for tobacco where they occur on rolling, well-
drafned topography. On steep terrain, under tobacco culture they have
undergone moderate ro severe erosion. Alluvial soils, predominently
in the Cangaree serfes, occupy only 2.9 percent of the watershed
(Perkins, ~~-82,~1924; Jurney, -et a1 9 1928). This low percentage of
alluvial soils reflecrs The narrow stream bottoms and flood plain lands
in this part of the Plat River drainage basin,
A mere listing of soils in series groupings does not provide a useful basis for evahuazing the hydrologic response of the watershed.
What is needed is a detailed classification of the drainage basin into
land capability classes whfch take into account not only the spatial
distribution, topsoil thickness, slope position and range, but also
the storage capacity for water. In brief, a hydrologic grouping of
sofls reflecting the surface moisture capacirtes and posftion on the landscape is required. Musgrawe (1955) devfsed four soil groups on the basis of runoff potential, using the final constant or minimum rate of infiltration of the soil as the criterion. He defined this potential as the rate of moisture intake after prolonged wetting. Subsequently, the Soil Conservation Service has adopted the Musgrave groupings as a basis for watershed planning (England, 1970).
It is possible to make some preliminary estimates of the basic infiltration rates for the soils in the series identified above, Eight of the 14 major soil series above are classed in hydrologic soil group
I1B 11 which includes sofls havfng moderate infiltration rates when tho+- oughly wetted, being moderately deep to deep, generally well-drained, with fine to moderately coarse textures, and having a moderate rate of water transmission. For these soils the final constant infiltration rate ranges from 0.15 to 0.30 inches per hour. Soils in 5 of the above listed series (Helena, Iredell, Mecklenburg, Orange, and WiLkes) fall in hydrologic group "C", with infiltration rates ranging from 0.05 to
0.15 inches per hour, and with a high runoff potential. One series, the Congaree, falls in group "Dff which include chiefly clay soils with a high swelling potential, a very slow rate of water transmission, and with the final constant infiltration rates of wetted soil being in the range of 0 to 0.05 inches per hour.
It should be emphasized that the cited infiltration rates for these hydrologic soil groups serve only as a rough approximation of the rates expected on lands under full forest cover; nearly two-thirds of the Lake
Michie watershed is under some form of forest cover having an organic layer (forest floor) that absorbs the energy of raindrops and signifi- cantly modifies the character of the surface soil beneath it, making it more porous. Seventy-two percent of the entire watershed area has
soils that fall in hydrologic group "B", with only three percent in class
"DM, the remainder being in class "C". In brief, from the basic stand-
point of soil resources and their Bnherent original depth and infilra-
tion capacity, the Lake Michie watershed has as good a potential for
water production as is possible in the Piedmont province.
Character, amount and distribution of precipitation. Essentially
all precipitation fallfng on the Lake Michie watershed comes in the
form of rain. Total snowfall may average as much as seven inches per
year, but a snow blanket usually melts within a few days. In the last twenty years the longest continuous period of a complete snow cover on
the watershed lasted approximately three weeks; obviously, snowpack is
not of hydrologic significance here.
Precipitation records for the watershed, particularly of the type
that provide data on intensity and duration of storms are inadequate,
Records are available on a daily total basis from only three stations,
two actually on the watershed. If ane observes the longtime annual
precipitation averages for a number of stations that "blanket" the
watershed, one gains some idea of the magnitude of differences among them.
Station Total precipitation per year Inches
LakeMichieDam ...... 43.44 Rougemont (on the watershed) ...... 43a92 Durham, N. C...... 42,65 Oxford Tobacco Research Station ...... 45,36 Henderson, N. C...... 45.58 * Raleigh,N, C- ...... 45.96 . Chapel Hill, N. C, ...... 45.79 The annual average differences among these stations are not large,
but in summer months when a considerable portion of total rainfall comes
in the form of convectional storms, large variations among stations do
occur. Thus, it is not unusual to observe very heavy showers recorded
at the Rougemont station at the headwaters of Flat Rfver while little
or no rainfall is recorded for the station at the Lake Mfchie Dam.
On the basis of the seven stations it is assumed that approximate-
ly 45 inches of water falls on the 107,200 acres of drainage basin each
year; annual precfpitation is equivalent to 4,824,000 acre-inches, or
roughly 131 billion gallons per year. This amount is well distributed
whether the years are wet or dry (Table 2).
In a study of the hydrologic response of a watershed, it may not be particularly useful to look at average values alone. Yet it may be
informative to compare the 10-year averages for the wettest and driest
years shown in Table 2 with the 45-year average values computed from
records kept at the Oxfprd Tobacco Research Station. (Hardy et ax.,
1967) The station is located only 15 miles from the Lake Michfe water-
shed and has approximately the same elevation above mean sea level as
the average elevation of the drainage basin, roughly 500 feet. Even
Precipitation Departures by Quarters for the Water Year Period Longtime 10 wettest 10 driest Average Years Departure Years Departure Inches Inches Inches Inches Inches Oct .-Dec. 8.98 11.68 +2.70 7.39 -1.59 Jan. -Mar. 10.46 12.43 4-1.97 8,04 -2.42 Apr .-June 12.15 13.14 +O. 99 10.21 -1.94 July-Sept. 13.77 18.61 +4.84 9.05 -4.72 Annual 45.36 55.85 +lo. 49 34.70 -10.66 Table 2, Precipitation in the LO wettest and rhe 10 driest years since 1926, by quarters per water year, at Rougemont, North Carolina.
Quarters of Wacer Year
Oct .-Dec. -Jan. -Mar, Apr .-Jim. Ju1.-Sept. Total Year Inches Inches Inches Inches Inches
14.94 18,81 PO. 56 13-33 12.14 8.52 10.52 10.01 13.94 11.50
Mean of wettest years 11,68 12.43 13.714 18,61 55.85
9.69 10.75 9.73 10060 LO. 44 9.88 6.65 7.69 13.48 13.22
Mean of driest years 7.39 8.04 10.21 9.05 34,70
Driest in % of wettest 63.27 64.68 77.70 48.63 62,13
"Durham station record; all other data from Rougemont station, the departures from longtime averages do not reveal new informa-
tion, but they serve to emphasize that the critical period in water production from such a drainage basin is the 4th quarter
of the water year, i.e., the high water consumption months of
July, August and September. In wet years, the excess water is
not likely to create serious problems, unless for example, such
large amounts (25+ inches) in July-September, 1945, happened to
fall in 2 or 3 intense storms creating violent stormflows that have the energy to deliver accumulated debris and move substan-
tial quantities of bedload from the channel into the reservoir.
The surplus water is simply "water over the dam."The dry years
are the ones of greater concern and the July through Septeriber
period on this watershed appears to be the critical time when high evapotranspiration losses coincide with heavy municipal
drafts on stored supplies. As an example, the annual precip-
itation in 1968 and 1967 was identical, 36.05 inches, but in
1967 the July through September rains amounted eo 11.21 inches
in contrast to only 4.85 inches in 1968 when the water level
in Lake Michie fell to a level of 9 feet below the dam crest,
To be sure, the low rainfall in 1967 may have accentuated the
severity of the drought in the following year, but it was the
late summer-early autumn rainfall deficits that resul~edin
desperate shortages of water at Chapel Hill and other neigh-
boring Piedmont communities. Durham, on the other hand, contin-
ued to enjoy ample supplies. In brief, the rainfall amount and distribution on the Lake Michie watershed over the past 4 1/2 decades suggests that water production is likely to continue to be ample for ~urham'sneeds for some time into the future.
The temperature factor. From a wa,tershed management stand- point, interest in temperatures centers largely on the timing and occurrence of minima, inasmuch as freeze-free length of the growing season is dependent on how low the temperatures may drop, and when.
The occurrence and frequency of temperatures low enough to freeze soil surfaces also determine the importance of this factor in dis-
Lodging soil through alternate freezing and thawing, or preventing in£iltration through actual freezing of the soil surf ace. Tempera- tures low enough to cause substantial freezing of soil surfaces at this latitude and altitude (36"+N lat. & + 500' above m.s.1.) need to remain at or below 20°F for a day or longer.
Longtime temperature records for the Lake Michie watershed are not available, but extrapolating again from the Oxford Tobacco
Research Station data, we note that the average freeze-free growing season may be assumed to be about 200 days. So far as natural vegetation is concerned, it is perhaps more significant to consider the dates when temperatures as low as 24°F occur in the spring and in the fall. The average date of the last occurrence at 24°F in the spring at the Oxford Station is shown as March 12, and the first occurrence in the fall, as November 20. Large variations from these average dates may be expected, and indeed, do occur.
Temperatures low enough to kill tender shoots on yellow-poplar wholly non-functfonei ~1%betore rbs. lasr we& in Ocrober; also, needles from the crop of cke prevm~kyear or rwu cn the native pEne specfes wfll have staprteci ti fdi. These :henoS.sgicaP evencs have a bearing on the intercep~ionand evapstransplration funcrisns in the hydrulogie cycle, k~ccznfluence the proporti~nsf gross precfgitatfon uirmataly reaching scream channels.
The sther major ir,fl.urnces of rsnperature xinfrna sn amount: and qualiry of screamflow wperase thzougla the freezing of soil surfaces, thus cutting off inffltratzon, and ice crystal dislodging sf soil from bare surfaces, The influence of froten soil surfaces on the Lake Michie wa~ershed is perhaps fairly siPghz since the watershed 1s twn thirds rorest and isw Cemperatuzes are not sus- tained long enough to affect surfaces protecaed. by forest floor,
The influence of dislodged soil is more Impautan~due to rhe extensive bare mad cuts and occasional bars rields, No records for the number of days of prolonged freezing are available for the watershed; records at the Oxford Station show that annually there are 69 days, on the average, when minimum temperatures fall to 32' or lower, Casual reconnaissance over field and forest soils within ~hewatershed during the past twenty years suggests that bare soil surfaces may stay frozen intermittently for as long as 20 to 30 days in an '"averageF" winter, and soils under forest floors do not freeze at all,
Evaporation. In considering evaporation, whether from water surfaces, soil, trees, or other vegetation, direct solar radiation is the only significant source of energy. On a free water surface, evaporation accounts for about forty percent of the total radiation received thereon. Practically all of rhe high energy radiation received by water is absorbed, hence evaporation losses from a water surface represent the maximum possible under natural conditions,
From weather records maintained at Rothams~ed,England,
Penman (1948, 1949) and Penman and SchofiePd (1951) demonstrated that the correlation between mean daily temperatures and &Sly evaporation was very poor. Monthly means, however, did exhfbft significant correlation, emphasfzing the fact that annual varfa- tion in evaporation closely follows the solar radfatlon cycle,
Thus, one might expect evaporatfon Posses to vary with latitude, and indeed, they do, In the Province of Ontarfs, Canada, approx- imately 15 inches annually escapes into the atmosphere from a free water surface; in the tropics, similarly, about 60 inches
(Stone, 1952). In Lake Michie, the annual loss is approximately
39 inches (Turner, 1966).
On the basis of this 39-inch estimate of evaporation it may be of interest to observe the order of water loss annually from
Lake Michie and from the farm ponds on the watershed. The lake
surface when at spillway level occupies about 507 acres, Annual water loss of 39 inches amounts to 1648 acre-feet, or about
536,920,000 gallons. The 620 farm ponds within the watershed have a surface area of 1272 acres; assuming the same rate of
loss from them as from the lake, about 1,347,068,000 gallons represents the annual evaporation loss from ponds, Ignoring
evaporation from streams, the total loss from water surfaces within the Lake Miehie system annually amounts to about le88 billIon gallons, or enough water to supply the needs of Durham
for over three months, assuming a daily requirement of 28 million
gallons. Annual loss from Lake Michie alone equals 13 percent of
the original storage capacity of the reservoir,
Viewed against a reference frame of the total precipitation
falling on the 103,200-acre watershed annually, the I,&% billion
gallons lost from water surfaces is only about 104percenr of rhe
131 billion gallons received In an "average" year, However, it
is estimated that of the annual total received, North Carolfna
loses about 75 percent to all forms of evaporation, the main
source being co transpiration by vegetation (Hardy and Hardy, 1991). On the Lake Michie watershed, evaporation losses would thus amount
to some 98 billion gallons each year, leaving a residual of about
33 billion gallons for other disposition. The 1.88 billion gallons
Lost from open water surfaces represents 5.7 percent of the amount available for streamflow and storage.
Prom records available for a sfngle Poeation in Lake Mfehie
(Turner, 196&), it appears that during a 3-year period (1962-1964) the average monthly evaporation exceeded the recorded precipitation in all the months from ~Iprilthrough Occober (Figure 21,
Native vegetation types* In general terms, the upland forests of the Lake Michie watershed can be classed as elther (a) pine,
(b) pine-hardwood, and (c) hardwood forest types. These type names do not coincide with the terminology used in Forest Surveys (Knight and McCPure, l966), but they should suffice for later discussion of any hydrological significance of forest types and silvicultural treatment of them. In fact, for this purpose it will suffice to consider only two major forest types, namely, areas occupied pre- dominantly by pines, and those occupied predominantly by hardwoods.
The extreme variability of mix tn the pine-hardwood type makes this simplification desirable,
The dominant pine species an the watershed is Virginia pine
(Pinus virginia Miller), a species occupying nearly eight hundred thousand acres of woodland in North Carolina (Slocum and Miller,
1953). An aggressive pioneer species, it has quickly colonized fields and pastures whenever previous use of them has ceased and wherever a seed source has been available (Figure 3). Usually Average Monthly Precipitation ------Average Monthly Evaporation
Figure 2. Evaporation vs. precipitation on a monthly basis, Lake Mfchie, Bahama, North Carolina (From Turner, 1966). Figure 3. The aggressive nature of Virginia pine as a "colonizer" without peer is illustrated by this photo of a gravelly light loamy soil on which field use has ceased by stages. The oldest stand is in the right background where use was abandoned some 15 years ago; next, in the left background, is a strip abandoned some 10 years ago; a 5-year fringe of pine borders the older stands. The foreground shows part of the field rep- resenting final capitulation by the farmer a scant 2 years prior to this photo. Pines have already be- come established at initial densities of 5,000 or more per acre, though the seedlings are not yet visible above the weeds. Here is a classical example of even- aged "management" following clearcutting that took place several decades ago, when the land was cleared or re-cleared for farming. The subsequent reforesta- tion has occurred naturally through chance phenomena. Virginia pfne occurs in pure stands buc both shorrleaf pfne and
Poblolly pfne may be found in lfmfted numbers as associates, usually in a dominanr position in the canopy. The latter two pine species are found occurring also in pure stands within the watershed, but sn considerably more limited scale,
The upland hardwoods are a variable mixture of white and red oaks, yellsw-poplar, hickory, sweetgum, red maple, plus a number of mfnsr species. In the best upland stands, the white and red oaks compr%se about 50 percent of the basal area, yellow- poplar is next with about 20 percent of the basal area, followed by hickory witfa 8 to l0 percent (Figure 4).
Tn the pine--hardwood type, the same hardwood species listed above comprise usually 50 percent or more of the number of stems in the stand, with pines varying somewhere berween 28 80 48 percene, and wieh their numbers diminishing with each "selective'" cut, since pines are more in demand, and less defectfve than up- land hardwoods.
From a hydrologic standpoint, it is important to understand the gradual change in the vegetation as field or pasture use is abandoned. When a full stand of pines becomes established on such an area, the weed and grass stages succumb to the shading, copious litter fall, and other forces of competition after crown closure of pines occurs. For several years, understory shrub vegetatfon is essenttally absent, the soil being covered mainly by pine fitter and the developfng forest floor (Figure 5). Figure 4 One of the better upland hardwood stands on the Lake Michie watershed, with oldest trees now about 70 years of age. Use of this land for field crops was abandoned over a hundred years ago, and the hardwood stand which became established under the pines eventually emerged to a dominant position as the pines were removed by 11 selective" cutting or high-grading. The stand contains a few veneer-grade yellow-poplars, and some white oaks as the one in the foreground, which qualify for flooring, furniture stock, staves, and related uses, but the basal area of the total growing stock is less than 100 sq. ft. per acre. -25- Figure 5. A natural stand of loblolly pine, plus an occasional Virginia pine, which became established on this old field site 21 years prior to this photo. The stand at the stage shown contains over 2,100 cubic feet of usable wood (peeled estimate) per acre, and a sub- stantial amount of forest floor, but a shrub under- story has not yet emerged. Land that has been in eonrinuous use for field crops or
pasture and meadow for a dozen years or longer, sr at least long
enough for the rootstocks 02 hardwoods previously occupying the
site to have died, will not readily become "re-colonized" by a
hardwood understory, provfded especially that the pfne overstory
fully occupies the site. In about three decades, depending partly
on the quality of the site, a shrub understory begins Lo develop,
and rapidly increases in size and floristic complexity (Figure 6).
When a mature stand of mixed pfne and hardwoods, or of pine with an established hardwood understory, is harvested, the resur-
gence of hardwoods, principally originating as stump sprouts, root
suckers and established seedlings, is truly impressive, This re-
surgence emphasizes the difficulty of forest renewal for the purpose of producing stands dominantly of pine. As an example of the dimensions of this problem, the total number of seedlings and saplings was counted on a 2-acre tract from which a mixed
stand of hardwoods and pine had been clearcut, The area is on a west slope, with an average gradfent of abouttenpercent, and situated about forty feet above the Flat River floodplain, It
is land on which the most recent cycle of cultivation ended
shortly after the middle of the nineteenth CEZ~UP"~. Four years after planting the tract to Loblolly pine at 6' x 6bparing, an inventory showed that in additfon to the loblolly the following species were present, which had attained groundline diameters ranging from 0.1 to 3 inches: Figure 6. Virginia pine became established on this old field site 42 years prior to this photo. The hardwood understory is prominent, with occasional yellow-poplars and sweet- gum nearly twenty feet in height. Dogwood, red maple, hickory, red and black oaks are also abundant, but mainly in the 3 to 8-foot height range. The annual addition of litter with a higher content of calcium, potassium, magnesium, and etc. is beginning to modify the character of the forest floor which in the first three decades consisted chiefly of pine needles, twigs, and sloughed-off bark. Species Number of stems per acre
Sweet gum Oaks (mainly red and black) Red Maple Flowering dogwood Hickories Holly Yellow-poplar Black cherry Redbrnd Redcedar MiscelEanesus Total per acre
Nearly ninety percent of B;he non-pine species were in the 0.1 to l.O-inch diameter class at groundline, but a substantial number of yellow-poplars, oaks, and sweetgum had attained groundlhe diameters of two and three inches.
In addition to the 3 major upland types described above, there is a fourth forest type that deserves mention although it occupies only a very minor portion of the forest acreage on the watershed. It is the bottomland hardwood type, with principal species being yellow-poplar, green ash, American beech, sweetgum, river birch, and bitternut and mockernut hickories. Among prorni- nent subordinate species are eastern hophornbeam, redbud, and
flowering dogwood. From a land and forest management standpoint, an extremely vexatious component of the understory vegetation is the massive jungle of honeysuckle that pervades and covers long stretches of the bottomland sites; it deforms and strangles seed- lings and saplings of valuable hardwood reproduction. Forest renewal of bottomland hardwoods is well-nigh impossible without firs^ eliminating or knocking back this persistent weed. Should use of herbicides, such as 2-4-5, T, be outlawed along channel margins, the feasibflity of future perpetuation of an acceptable level of growing stock consisting of valuable bottomland species looks dim, Chapter III
GONDITSON OF THE EATERSHED PRIOR TO 1926
General, The ecology sf man residing and toiling on the lands of this
watershed has been as dynamic as the ecology of the vegetation, native
or introduced. Signs sf impermanence are not difficult to find; some-
times they stand ous fn bald outline in rhe stark desolation of
abandoned homes and farmsteads (Figure 7), Elsewhere, the story is written in code that a century or more of weather and vegetation have
failed to obliterate complete%y, An old tree or two, spreading and
gnarled, and oue of context to che natural forest vegetation envelop-
ing them, suggest existence of some previous farmyard, A pile of hewn stones may mark a homeelte where once men and women lived and loved, and children played. OccasisnalSy , anh far from well-traveled paths, one may find a stone, marking the grave of some pioneer, perhaps long forgotten, and no longer graced by wreaehs of commemoration.
As revealing as these signs of past Lffe and land use within the watershed may be, their hydsohgie significance Is small compared to the impact resulting fro= man's past tilling of the soil. Often this impact is recorded adso in code, requiring some degree of unraveling to interpret its meaning (Figure 8).
There is no comple ely satisfaceory account of the condition of the
Lake Michie watershed lands prior to 1926. The most detailed information on erosion and related land use conditions is found in the report on the soil conservation survey made sf the watershed in 3.936 and 1937 (Mattin and Bass, l940), This report rcf1ec.t.s the hpact of many decades of land Figure 7. Abandoned farmstead on the Dial Creek watershed, a tributary basin of Flat River, -- a symbol of impermanence of man's husbandry of land and soil. It reflects the stirring drama of life itself, but not an elegant epitaph for the pioneer men and women who wrote the story with their lives. Figure A completely healed gully, about 2 to 3 feet deep, within a hardwood stand which emerged as the dominant plant cover on this old field site when the pine stand that had "colonized" the field initially was evidently clear cut about 70 years ago. It is assumed that the pine stand accomplished most of the soil stabilization; at any rate accelerated erosion has been held in abeyance for several decades, and the former gully would very likely go undetected were it not for the evi- dence of very poor growth of hardwoods established in the bottom of the "gash. I' use prior to the date of the surveye For example, the aurhcrs repor7: that nearly two-thirds of the 62,817 acres of vxxlland occupying the watershed at that time were moderately eroded, meaning shab on shfa acreage from 50
so 75 percent of the topsoil was gone, Jt is revealing to note their failure to comment that the ropsail. loss occurred durfn the yeaks the
land had been under cuP$iva.tfsn of field crops, fn pascure use, or bosh,
Under the woods operations then prevailing, or even under 2usrendy employed operations, it would bc imp~ssibPeto sustagn &he degree af cop- soil. 1.06s reported fer the ewoodlond areas by Marein and Raso &P9liO),
The same aurhors stsce char moderate severe, or bJery severe eessian has affected 99 percent of the cropland (38,101 acres), more rhan 98 percent of the idle land (2,991 acres>; and 81 percent of the pastuxe lands (2,647 acres), They define ehe severe and very severe erosien classes as soils with 75 percenr or more sf the osfgfnal topsofi, m all topsoil and s e upper subsoil removed, and lower subss~iexposed or removed, respectfve%y, Of the 107,200 acres in the water hed, only 205 acres were designated as suffering no apparent erosion, The surhsr~ poscufate that the total erosional debrfs produced in ehe warershed up to the date of their survey, beginning from the time of agricultuiak use, was approxfarntely 38,850 acre-feet. Much of ehe debrea did nsc move very far, accumulasfng mainly on the lmer slopes and narrow valleys, In the period from 1926 so 1936-37, they estimated the average annual deposits
fmco rhe reservoir to amount 60 about 45 acre-fee&, with additf~nalma-
terial lose in overflow amounting ro 7c acre-feet annually, One csuid conclude from these preirmxaary estimases that sedfnenracion was no^ adjud ed serious in rbe early yeam sf fife sf the Lake Mfchfe
reservoir, What has not been sva%uatcd is she de pee to which the wdde-
spread Pose of topsoil has ~nfluenced ~heklydrslogie re6 watershed soils and the pscentials for crop production, So far as Forest production Fs concerned, the 50 Ce 75 percent reduceion in topaoib depch, according to CoPlr 9194 ) and others, su gests a volume yield loss of 25
to 50 percent for Soslolly and shostieaf pine ar a e 30 years, Sin~ilar
Issses could be expecied for other comere~allgvaluable species, al- though critical site qualicy id~~formationon then is Paekin the economic fnipl~caeianssf soil io s cannoc be ignored sfmp9y b,&cause reservoir sedimenta.r;$on9s jud ed to be of minor consequence.
The role sf forest f1r5$z Tor che fsras",ed psctfon of the waeershed, a brief review of past Pire hfsrory is relevane, Piedmont forests, in general, have nor been ued by sever2 wild fEres, Among several fac~orschat have tended to keep fire Sssses small and fire occurrence infrequent is the ownership pacrern which breaks up forest holdings into small, seatrered blocks in which fires, ~f they start, can be Localized,
Unlike rhe Coastal !?.Lain, more chan a half sf the forest is in hardwoods which develop less fl.ammable fuels, Sr is also possible that farmers may have exercised greater care in zhefr field and fenceYEne burnfngs, and have not permitted such ffres eo spread so adjofnfng woodlands,
Nevertheless, evidence in ehe form of basal sears and multiple stems in older hardwood stands on ehe Lake Mfchie watershed suggest chat ffres have occurred fn %he woodlands in past decades, Cruikshank $1940) found that upland hardwoods in che Piedmont were burned more severely than any
other forest type in this physiographic province. Evidence of fire was present on 44 percent of the type acreage, with light to medium damage
occurring on nearly half of the burned acreage (Cruikshank, 1940),
No separate record of fire occurrence for the lake Michie watershed
is available, but in the 27 years while Durham County has been partici- pating in the State Forest Service cooperative forest fire program,
there have been 1387 reported fires, averaging 5,7 acres per fire. The largest total annual acreages burned are 1217 acres in 3.951 and 1154 acres
in 1959. Person County, which contains the bulk of the watershed, did not
come under the cooperative fire control compact until 1966, so only a six- year record of fire occurrence is available; in that period a total of
107 fires have been reported, burning over 640 acres. Average fire size has been 7,7 acres, From a watershed management standpoint, up to now wildfires cannot be considered as a major factor influencing water pro- duction or quality,
In rhe future, expected intensification of forest management practices throughout the watershed undoubtedly will involve more wfde-
spread use of fire as a silvfcultural tool. Prescribed burning would be necessary in the natural regeneration of hardwood stands after clear-
cuttfng in order to induce sprout regeneration at groundling. This practice reduces the incidence of butt rot in the maturing stand. Perf-
odic use of fire is also needed in stands of shortleaf and loblolly pine
ro reduce accumulation of fuels and as a measure to hold back the develop ment of hardwood understory, keeping it small enough to be manageable when the time again arrives for renewal of pine, In Virginia pine because of its thin bark, fire cannot 'be used at intermediate stages, but at time of stand renewal, after dearcutting, use of fire would be essential. In none of these instances, would pre- scribed use of fire significantly frnpafr the hydrologic response of the land, If a19 woodlands were restored KO acceptable levels of production, the annual acreage subjeered to prescribed burning would not exceed two to three thousand acres, assuming the area of woodland in the watershed would rentafn about what ~t is today, Chapter IV
LAND USE CKANGES
General. As mentioned earlier, a major objective of this
study was to evaluate the influence of changes in land use on
the quality, timing and yield of water from the Flat River basin
+ starting from 1926 when the Lake Mfchie impoundment began func-
tioning, To detect possible trends in Land use, chief reliance
was placed on interpretation of sets sf aerial photographs which
were available for the watershed for the years 1938, 1941, 1951,
1955, and 1967. During the study, an addfrional flight was made
in 1970 to determine whether any recent, sfgnificant changes had
occurred.
As a first step, it was necessary to establish a set of land
use categories that could be unambiguously identffied on the
existing sets of photos. The quality of photos was not uniform,
and it was at once apparent that a hfgh degree of refinement,
especially for interpreting differences in native vegetation types,
was not possible on a consistent basts from one set to the next,
Moreover, much refinement was not deemed necessary to achieve the
desired objectives. For example, it was recognized that there
were at least four major forest types on the watershed, namely,
pine, pine-hardwood, upland hardwoods, and bottomland hardwoods.
It was also known that the pine-hardwood mixtures exhibited
extremely high variability in species mix; hence, it was not
practical to establish a pine-hardwood category, The bottomland hardwoods comprised such a minor portion sf the tctal forest area,
that a separate category fo~It was not provided eithera In brief,
only two categories of forest were selected in she examination of
the entire watershed: (a) areas predominantky in pine and (b)
areas predominantly in hardwoods. (Is a special study of land
use on the Did Creek watershed comprising only 3,9r05 acres,
a more detailed classification of forest types was used;
reference to this study wlil be maie In a later section @f this
report), The use categories selected for the. entire watershed study were:
1. Forested land
a. Predominantly pine, ineaudkng also pure stands of pine primarily representing more recently colonized old fields, as well as the few existing planted stands,
b. Predominantly hardwoods, in which the pine component ranged from 5 to about 45 percent.
2, Cultivated land
3. Pasture and idle land
4. Other land a. Roads
b. Streams and ponds
c. Residential and industrial
Procedure. Since all of the several flight sets had covered more than the 168 square miles of watershed, the pertinent prints were pulled from each set, including the peripheral ones In which at least 50 percent of the photograph was within the watershed boundary, Alternate photographs along each flight line were then selected for sampling. A 5 x 5 grid of points was placed over a given contzct print, centered by reference to fiducial marks, resulting in the sampling of 25 points per photograph. Land use was identified for each point, basing classification on the predominant use within a fifth-acre circle centered on any spe- cific pofnt, For each set of photographs, the predominant land use was observed and recorded for the entire series of sample points, and from them the percentage of land use for each cate- gory was computed. The number of individual sample points for each set of photos, beginning with the 1938 set, was 1275, 1225,
1250, 1405, 1300, 1425, and 1233, respectively. The percentages are represented in Figure 9 to provide a visual impression of the relative magnitude and trends in each use category. The fluctuations within each use category are the result of sampling error as well as variations over time.
This procedure left undetermined any possible changes which may have occur.red between 1926 and 1938. A diligent search of archives in Washington, Do C. failed to unearth any photo flights of the watershed prior to 1938. It appears that no aerial photos of the State of North Carolina had been taken before that year. A search for alternatives to aid in tracing the land use pattern exfsting on the watershed before 1938 resulted in a decision to rely on Agricultural Census data. Since the major porrion of the Lake Michie watershed lies within Person County, Hardwood Forest
Pine Forest
Cultivated Land
Pasture and Idle
,Roads // Residential --#Water
Figure 9. Percentage of land in each land use category. the census data for that county were selected for study on the
assumption that the County data reflected, wfth sufficient fidel-
ity, changes occurring within the watershed,
On the basis of the tensus daza, the percentages of land in
each use category, as previously aefined, was determined for
Person County, tracmg fr by decades back to 1920, The percent-
ages were then plo~zedon s graph together with the percentages
estimated from the photographs to observe flow well the Lwo sets
agreed (Figure k In the coxbmed graph, the Agricultural
Census data appear trr agree wfrh the alr pl-~otoestmaces, within
reasonable Efmits 3E samplfng error, thus it does nor appear
unreasonable co fft a trend line thraugn the points.
Results, Based on tierial photo estfmares, rhe hardwood
forest area clearly dominates land use within the watershed,
since 1938 it has occupied from about 35 percent to about 41
percent of the total land area (Figure 10), If one ignores
the somewhat anomalous drop indieaeed by the 197'0 estimates, the
trend in hardwood acreage is upward, Even ehough the aerial
photos do not provlde details of trend toward conversion to
hardwoods, evidence of it is wldely available throughout the watershed (Figure ll),
Cultivated land acreage exhibits periodic fluctuations,
and has increased in comparison to pasture use, but overall
agricultural use has declined since 1926, partly due to the
sharp early drop in pasture and idle land acreage, The Forest
hotographs-of the Lake Miehfe watershed.
Based on Agricultural Census Data of Person County.
Figure PO. Percentage sf land in the major land use categories showfng trends over the, Figure 11, Left: Portion of a stand from which a mature overstory, predominantly Virginia pine was "clearcut" three years ago. Logging slash and emerging brush obscure the stumps of mature trees. The stand is now primed to develop into a hardwood type of "green junk." Right: Interior of a similar stand as described above, showing condition eight years after logging. Debris from logging has essentially broken down completely, and the pine stump in the foreground is in late stages of decay. No pine reproduction is evident. Both photos clearly illustrate how the conversion to hardwoods, in the absence of planned forest renewal, takes place, .proportion of land predominantly in pine forest shows a slight but
expected downward trend, ignoring the anomalous jump in 1930.
Where all forested acreage is combined, the trend is upward.
Within the use category termed "other", two sub-categories,
namely ponds and roads, were of sufficient interest to warrant
obtaining more information about them than was possible to
secure with the sampling procedure described for the general
land use phase. Stream channel mileage was assumed to remain
at a stable level, but substantial changes in farm pond numbers
and road mileage since 1926 were suspected.
From the 1970 set of photos, every third photograph along
every other flight line was selected as a sample unit on which
a complete enumeration of all farm ponds was made, These ponds,
identified as to size and number, were then traced back on each
earlier set at the same locations within the watershed to escimate
the approximate time when the ponds appear to have been developed.
Of the 620 farm ponds observed in 1970, only two were in existence
in 1938 (Table 3). The enumeration and dating also showed that
the major activity in farm pond construction began in the early
1950's progressing at an accelerated rate, with a better than
500 percent increase in the last 15 years.
For the assessment of changes in road mileage within the
watershed, aerial photographs were unsatisfactory, mainly because
it proved too difficult to differentiate accurately among stan-
dards, even as gross as paved roads versus those not paved. Table 33 Three-decade increase in number of farm ponds on the Lake Michie watershed, based on enumeration sf iden- tified ponds on successive sets of photos,
1938 1941 1951 1955 1960 1967 1970 Estimated number of 2 2 35 158 312 538 620 ponds
Aecsrdingly, the Scare Highway Department records for Person
Counry were searched for road census data corresponding to the years for wh2eh aerial photos were available, Data were available 6nPy from 9941 to the present, and again it was assumed that changes taking place in the County as a whole refleeted changes of similar nature within the watershed, To partly check
ity sf this assumption, the road network strictly wfthfn the watershed was surveyed for 1971, and was found to fnelude 88 miles of paved roads and 135 miles of roads not surfaced sr paved; these mileages correspond to 39.5 percent and 6005 percent, respec-
tively, This reconnaissance does not deviate greatly from the
Highway ~epartment's1970 data for a11 sf Person County for which ehe cerresponding perceneages were 44.1 percent paved and 55,9 percent not paved (Table 4). One may conclude that the Person
County percentages can serve as rough indfeators, representing
the percentage changes in corresponding years for the road system within the Lake MichPe watershed,
Since 1941, the paved road mileage of Person County
has more than rripled, while the tora1 road mileage has increased
only &out 15 percent over the same period, Haever, mre than half of ~e roads within the county are not paved. 'Moreover, many miles of
paved roads have raw road cuts, with both shoulders and ditches
undergoing accelerated erosion.
Table 4, Trend in Person County road mileage and standards over a three-decade period,
Year R.oads Not Paved Roads Paved Tst al- Miles Pct, Miles Pet, Miles 1941 480,6 82.9 98.3 17.1 578.9
What do these data imply abou~the-- watershed? Neither rhe
census data nor those derived from general interpretation of the
aerial photos tell the entire story, They give only the net
change which has taken place within the period in question, The net change has been large enough to alter the character of the watershed. The best estimates suggest that since 1926, total
forest area has increased by about 15 percent moving up from about 47 percent of the total area of the watershed to more
than 62 percent. Field and pasture acreage during the same perfod has shrunk about 10 percent, However, perfodic activity rather than net change has, doubtless, had a far greater impact on the hydrologic condition of the watershed. To obtain an estimate of total activity would have required an in-depth study of a sub- stantial sample of the same specific points not only on each sec of photographs but on every photo of each set, This approach proved to be beyond the resources available for the general phase of the investigation, Nevertheless, an insight of the exaenr and nature of activity was obtained by a more detailed study dfrected at %Re Dial Creek basin which appears to be fairly represeneative sf orher trfbutaries within the drainage basin (Van Loock, l9?2),
The Dial Greek study illustrates clearly that net change in land use may be considerably less than total change, hence fails to reflect the extent of activity as well as the signi- ficance of rhe changes (Table 51,
Table 5, Thirty year changes in Band-use classes on the Dial Creek watershed, showing ratfo of net changes to total change (after Van Loock, 1932).
Tot a1 Net Land I940 1970 Net Movement * Total Change Use -Area -Area Change In Out Change Total Class Acres Acres Acres Acres Acres Acres Acres
Pine 1,115 596 -519 195 714 909 57% Pine-hardwood 613 613 0 385 385 3 70 0% Hardwood 535 1,121 4-586 603 17 620 95X Cropland 369 535 -234 113 351 468 50% Pasture 95 195 +I00 161 61 222 45% Idle 123 I62 + 39 151 112 263 15% Miscellaneous 134 162 + 28 45 17 62 45%
Total 3,384 3,384
*Movemene represents acreage moving fnto a specific land-use class, or moving out sf it,
By means sf joint and conditional probability techniques,
Van Loock estimated the land use movement, traced the sources which produced the sb5eeved net ehange and determined the resulting
change eccnomicsiLy 1s th& 114 acre decrease from the pine type
existfng 5.n 13W; thBs aitritlon was partly off set by the gain of
195 acres back to psne, ~argelythrough colonfzation of old fields
and Ldle Zmd by Vxeglnsa pine, Both changes have some hydrologic
elgniiicance; the de;ay:bng Fine st~mpson the cutover area improve
percolation of r:**.1a9and C~Eregeneration of old fieLds co pine
improve inff%trariorn, and in a11 probability reduce averland flow
and soil movement,
Tbe pfne-shazdwc:>d type provides the most striking ilkustrarlon
of the dlfferenie between net change versus activity* The acreage
in 1970 was the sane ar in 194 hence no net change, Yet the
Dame amount sf land (3% acres) moved our of the type as came into
it from other ases, ?"sr iirgest share sf the ehange took place
through srl.ecrive cutt%ng 94: pine changing 284 acres from pine-
hardwo~dmixture m'co "pt~are" hadwood. A major reasan why net
change in thls type :-tawined zero was the 290 acres which was eon- vezted from pine to pina-nardwood, largely again through selective
cutting ar high-grad2ng of pine, effecting release of understory
Ezard9~0ods The we1 ~.ncwnmsural srlccession of pine types ts
pine-hardwood and kl:taliy ro hardwood alone has been greatly aceel-
erated by the sekeer~vecztting practices in the DIah Creek basin
as well as 0-2, the tnt~~eLsPe P'iichie watershed (Figure L2), Figure 12. Present condition of a forested tract that has been periodically high-graded of the more valuable pine component. On the left center is a stump of a freshly- cut pine; on the right is a decaying stump of a pine cut about eight years ago. The gradual conversion of a former pine-hardwood forest to a hardwood type is now apparently complete. Area is in the Dial Creek sub-basin of the watershed. Since the large Increase an hardwcod type acreage was wholly expected in che Dial Creek basin a major point of interest concerns its stability, It is evident chat the hardwood type resists change; in 30 years, only 17 ecres in the Dral Creek basm was converted from hardwoods to other uses, As an example, new ground for field or pasture fs more likely to be dcveloptd trom land occupied by pine stands; certainlg zt is easier ro converr pine types to field use. En thfs instance, a total cr 653 acres came into the hardwood type from other uses, 94 percent of ir being csn~sfbu~edby former pine and pine-hardwood rypes,
The shift from pme to hardwoods involves two pofnts sf hydrologic significance. Land that remins Psng under a hardwood forest cover develops a forest floor and surface sol1 character- is~icswhich are optimum fcw fnfflerarisn cf rain, If, also, the underlying sail is deep and has the capabrlity to develop structure, then detention storage of the surface sosl 5s maximized on land occupied by well-stocked hardwood stands, However, on rhe Lake
Michie watershed, maxnmrzation of infi.ltratmn dues not appear to be a limiting critfcal need. Any well-stocked foresr cover, whether pine, pine-hardwoods or hardwoods alone, provides floor and soil surface conditions for adequate infnltration, Afrer approaching equilibrium conditions attaxnabhe under a given floristic composi- tion of the vegetative cover, detenrlon seorage of rain, deep seepage, and related phenomena will depend more on geologic struc- ture, land form and soil characterisrrcs rhan on any peculiar merits of vegetatfve proreetion. Anether major pcant of hydrologic signLf~canee~nvolves
ineer-reiationshrps of evspsrransplrarion and ~neenceptionwith
ner. yield of wster- in rhe Lake Mschze watershed, the deciduous
hardwood complex suffers a drastza reductfon fn rranspfrarfon
from about November to April, Th~sreductton exsends from the
time leaves die naturailg sr are killed b) rrost untfl new foliage
unfolds in tne sprfng, When follage dies and falls eo rhe forest
floor, the process or xntercepeim also ss greatly reduced, Both
the diminished evapotranspa.-sacfc)n 2nd reduced nncer~eptionsub-
stantxallp increase the prapscti.cn sf grass precipitarion avalE-
able for augmensang the basin streamflow, TE must be strongly
emphasized rhar the increased steemtlow wmes in lare arr~mn, winter, md early sprang msnrshs at this latatude (circa 36%) when consilner needs are generally at their lowest, 3 the storage
system 5s not designed to catch and hold the augmented ysePd with-
fn the xmpouadment, rhen the extra water fs just so much more water
over the dam.
In the contexe sf impxovement of the hydrologic condatisn of
the watershed, a third atem needs men~isn,and for the Lake Michie
drafnage basin, it nay be as sngnffieant as elther of the two al-
ready descrfbed, It is simply the rncceased forest acreage, re-
gardless of irs deteriorated eonditlan with respect to the quality
level sf growing sesek, No chinking mdivzdual should question
that land under forest cever, albeit green junk, offers one of the best rypes of prorecrion for the soil; such cover also produces
Two additl.onaX deveYopmeis_ts on the Lake Michfe waze.ti.:zm.* % sfnee is26 are af s~ffi~-&nt interest meneii-jn, here a$ ;;ar:,.- eluding irems on @heuse sf land, One is the ~~rmnendousixreaae in nunabers sf farm ponds, mainly within the two decades,
'The 1292 acres of water surface in the 620 farm ponds 'hav.e 2 negative .i,nfi,:lence water ,>rfelds, Mosr. sf .-=hem1i.e ii:: ~1.e open swil; uil,sl~sd.ed.&hey absorb nearly all cf " -13.",.gh eriei'gy raddstio:r fallfkng +-'& hence Losses ghe pond snrfaces
y= c3.9 += <-. 0. Ypi638snt,.-..- . :xa,;ci.t:j.PUiZipesslbje f-%a111iaaeU~ai:t e-t~qy...-.~..i rb~bta.~:i ..*~.. &E a:: anmcial. ba3Fs the amcsuntg to about Po32 bj.j.,ll;cyi, g~j.?,,:~ys
Al.thaugln, this 1'3~sappears fnpleseivi;, it fs ---,?cry 1L&el,ji muc! 1.
.,". more than affset by the beneffts derived from the ponds, :.'!.ey serve far :reehargfng ground water supplies, Depending soii.,e:xhsxt
0x1 design of ourf$.ow of a?.l-;p.lus w.ater, they Taay ca"r,r;bi and .';to:t3 sion-j2in .- O ,dnr a~nounts; of storn-ir"3.0~,the reducticrr. of wfiich ~73~3.d be impor~antespecl ally during j.ntense, smmer convectionai shsweys, yhgfr illany gmellf~yvalses a,re alscj :i.rn.;c>y?>an,c,>l.fi::~~~~g6 . .- their function as a scurce af water for irrigation, !:hereby conserving ground water supplies,
The ocher iten wonteh ccmncnt as a final point Is the naz::,re of the road mlleage within the watershed. Tn r:he present. stage of road development, the influence of rhe road system must be considered as largely negative because ir: degrades water gurii.ty,
From air photo analysis, it is estimated that about 2,000 acres of watlersh.ed are preempted by road surfacesi Tdhethel- paved or not, infiltration on this acreage is zero for all practical purposes, Assuming that of the total precipitation sf some
45 inches per year, at least some 36 inches occur in the form of sharp, convectional showers or intense, heavy cyclonic storms, the road surfaces would shed some 6,006 acre-feet of water per year that has nowhere else to go but downhill, Some portion of this 1,95 billion gallons flowing downhiSE every year does scour out road ditches and ultimately reaches a stream channel with a variable and unknown quantity of sedi- ments,
As a single example of the magnitude of sediment trans- port possible via a road ditch, a sample of such effluent was caught during a July storm of more than five inches at the point of flow fnto the Flat River channel. The 1,650 cubic centimeter "water" sample contained 96 grams of sediments, oven- dry. Without timing total flow, it may be ldle to speculate beyond this polnt, but suffice it to say that the indicated volume was eaugh"cnstantaneously, and suggests a transport capacity of a ton of sediments suspended in a 5,086-gallon flow.
There is abundant evidence on the watershed to indicate that the road cuts, ditches, and shoulders contribute much of the color in the streamflow moving into Lake Michie (Figures 13, 14, 15).
If the suspended material is not considered objectionable, one might find a redeeming feature fn the rapidity with which water is t~ansportedInto stream channels the road system before it can be lost through evapotranspiration, It could be a factor in low-flow augmentation, -5.5- Figure 13. Road cut on rural road No. 1571 showing typical dislodging of soil material during alternate freezing and thawing, and continual sloughing into the road ditch. Photo taken February 26, 1971. Figure 14. View of eroding road shoulder and ditch on NC road No. 1571. Even with moderate gradient, as here, the amount of flushing of loose soil material and the scouring of ditch bottoms is quite substantial. Photo taken February 26, 1971. Figure 15. Direct drainage from NC road No. 1120 into the South Flat River, creating a fan-like deposit which is continually erodLng into the stream. Quality sf water 9s of primary concern to a muniefpa%2ty,
regardless of the source of supply, When the source b8ppens to be stored water in a reservoir, in wEtf ch the supply 1s ~ia~ntafned by the outf%aw of s~~mrbasin, the qaanttcy and timrag also can become matters of high interesr and gra7e concexn, Qualify, quantiry, and kining are rfs basic ae"L&ib~fce~ok wsflerq T4anV8 aetfvf tiea can affecc cash of these a%~ribrrres m some sleu~e9ons ar at some time 2% a WE^ and ko a degree to inzpraoe them; in other places or at other times he can exert infl.wnce on chern resalting in decidedly deleterious effects& In ~blrsstudy, we shall examfne to what extent these attributes have been altered by changes fn land use on che watershed of Lake MicRie,
---About Flat er ~ua9ity-- General, Depending on intended use, eke tern, water quality, can have a varfeey of meanings, or be defined in different ways,
Here rhe concern centers sn potability, primarily for culfnarg purposes, and the inputs necessary to bring rziw water to a state or condition suitable and safe for the consuming public, Adequaee avai%abXfty is not enough; %he concern musk be for availability of water in a state or csndicfen requiring mtnimum costs of treat- men t ,
Changes in land use taking place an a warershed eoncefvable can have a marked fnfluence on the availability and condition of
-59- calcium, magnesium, ~odfum,potassfurns SULFU~,and chlorsne; the results are recorded in mflPigrarns per liter,
ather records kept at the West Durham plant include the amount of water treated, the amount used for generation of electricity-,1s and the amount of flocculant, alum or ferrisol, and of chlorine used in the treatment process,
To assess variation in water qualicy, the analytical results from the determinaefons made intermiceently throughout the month were selected for stat2stical treatment using monthly averages, maxima, and minima. Data on r.hc mineral consrituents derermfned daily every third month were also used .Ln this study, but sfnce only rhe average values had been kept on these analyses, no measure of daily variation was possible, Finally, the data on water crearment were analyzed to determine what influence the changes in water qua1.i.t~had exerted on the treatment effort,
Resu3.t~of analysis of data. The data as recorded represent time series which usually consist of an underlying trend, a seasonal variation and a random varf at f ore. The primary concern of this study was to determfne if in fact an underlying trend does exist in the data and to evaluate its significance in management consfderations of the water system,
- The impoundment at Lake Mlchie was initially designed to generate electricity for Durham's munfcipal needs; now it is use.d to generate just enough power to run the pumps, It was recognized that over the period sf data zolleclrsn modiffcations have occurred in the analytical and measurement techniques used. In evaluation of the data, points of change were noted and the literature investigated to determfne the influence of the changes on the value of the data, Where nrsui- fications were known to affect the recorded data, an adjustment was applied so as to make the total time series as corisistefie as possible.
Results of the entire analyses on water qualfty appear too voluminous to include in this report. Accordingly, the average turbidity per month has been plotted as representatfva of xhe type of data derived from water quality analyses (It;"i.g~ire;.&I
Similarly, the average maximum and minimum values of turbidity have been tabulated for two 24-month periods (1932-33 and 1966-
67) as an illustration of the change in variability that has occurred (Table 6).
Several features are evident in Figure S6. One, there is a definite seasonal pattern, correlated directly with the seasonal pattern of rainfall and hence runoff. Two, there is an observable trend, particularly in the first 15 years, suggesting thac changes in land use activity over the period have resulted in greater stabilization of the watershed. Three, the variab$%l.ty as indfcated by the reduction in range between monthly maximum md mfnfmda has been considerably reduced. This last observation is consfsten%
Table 6. Range of rurbfdfty in ppm SfO,, for two 2-year
Dafly Turbidity -PPm Sf02 -Date -Max. Mfn, -Date Mo, Yr. Mo. Yr.
1-32 140. 80. 1-66 2-32 100, 70, 2-66 3-32 150. 80. 3-66 4-32 200, 100, 4-66 5-32 130, 70. 5-66 6-32 70" 40. 6-66 7-32 273, 50. 7-66 8-32 120, 70. 8-66 9-32 90, 50. 9-66 10-32 80, 50. 10-66 11-32 80, 70. 11-66 12-32 75. 60. 12-66 1-33 75, 65. 1-67
2-33 7D 0 65. 2-67 3-33 80. 65. 3-67 4-33 150. 50. 4-67 5-33 400. 90. 5-6 7 6-33 300, 120. 6-67 6-33 140, 90. 7-67 8-33 135, 95. 8-67 9-33 105, 70. 9-67 10-33 90, 40. 10-67 11-33 55. 40. 11-67 12-33. 55. 50. 12-67 with the negative trend in turbidity as a further reflectfon of
the stabilization of the watershed, reducing the amount of
erosional debris from it.
Observations obtained from plots and analyses of the remain-
der of the chemical characteristics are summarized in Table 7.
The table indicates the existence or non-existence of trends and
notes changes in variability over time, The change in turbidity
and in silica could be attributed at least in part to changes in
land use and to improved soil husbandry measures employed on
farms, Other observable trends could possibly be attributed to
more abundant application of commercial fertilizers in recent
years, heavier L5dng of pastures, occasional use of calcium
chloride to reduce road dust, and the more assiduous de-icing
of roads and bridges, thaqgh ascription of the trends to these
sources is admittedly tenuous.
In evaluating the purification effort, one encounters a
difficulty because of change in philosophies and techniques of
water treatment that has occurred after the data collection began.
For example, prior to 1949, chlorine was added to the water from
the settling basins along with flocculant and then also added to
the finished water. It was found that the "primary" chlorine was
being wasted. Further, it was determined that larger dosages of
"ffnal" chlorine were more effective in the control of bacteria
and were not detrimental to drinking water. Hence, there was a definite shift toward using more chlorine in the treatment process. Zm~e7: Sumwiry of observgifons of chemical eharacterfstfcs obeafned from water aarnpies analyzed at the West Durham TEeaement Plant,
Trend Variabf lfty (Max. - Min,) Slight negative Csnsiderabfy re- duced over time Slight positive Increase over time None No perceptible
None No perceptible change None No perceptible change Ncae
S?.fght negative
None
Slight pssitfve
None
None
Slight positive
Slight positive In order crs compensate for thss chmge, separhre segsi-essbx equations 2v0iobng che water quallsy vaal~blesmeasured weee fitted to the last 15 years of ,~sSngpounds of f9oecuBant and pounds uf chlorine as rhe de~endentva~abks. These equa- tions were applied ta sr;e earlier data co predict how much fPocsuPanc and chksrine weald d~sa~xteen us~dif iamxit treat- mene psacfices had be~nin effecr frsm he beginanng, Sxgure 14 prowides a plot sf wacer eonsumptim. Pits January 1931 to April
1968 while Figures 18 and 19 prc~vjdeplots if the resl;lrs of the adjuseed purifieetfon effrtt pea 1000 gaEPuns or water created,
These figures indieate that ~h~hik?there has been an increase in water ~~i~sump%i.on,there has been a S~-fin?"Ldeecresse ID the purification effort necessary cwer ibk jL?--y'ea~perid represented by the daea, (Cereafn sf rhe quality characteristfcs inportant in rhe fffced equatmn were not coll.eeted before m~d1934,)
The notion of purfficatxon efforr has been used to evaluate trends in rhe system rather; than dolh.rb per unxf since It elfmi- nates the problem of depreciation of Eaciiities, rriodPf~catfons of manpower, changing costs sf mater%aPs, etc, It is a direct measure of the influence of changes of water qwility o-~ertime and Ic relaces only eo those changes, Hswebes, one faetsn "La% the pre- diction procedure does not aecutnr for is decrease Pn varfabilfsy, mafnly in turbfdity, over tame, The predaetisns for adjustment of purifieatfon effcrr were based sn aonthly averages of the quality cha~acter9stfcs~The influence sf greater veria5~4ffyfrom day eo
day in the earlier years suggests that even & grearer dffferenrxal
In rnanskaly purifi~atimcrfort: between zhcse ysars and cusrrenr years would have been indicared if the varfabilfty could have been
raken into accouat .
to LSmitations *-- control Any watershed not owned, managed, and cerr~roiledby rhe operating autkor5.1:y is subject C'-o the vicfssitbdanaus rnlluenees generared by human occupancy and use of the drainage bassn- The Departmz:~~of Water Resouxces sf the City of Dt~rhamhas uovnrnal equfty in rhe lhnd around the 507- acre Lahe Ffchte, exazending to the 350-f ss~contour xntemal; f0e, , ten feet above the eresr sf the dam, Because of the variable land form arnd associated gradients, the caatwr Xnterval cireuar,seribes an indeterirtfnate acreage of land, a ragged fringe of terrain around the lake, Beyond thfs fringe 19es the balk of the 168 square miles af watershed, on which reside some P2,000 people on farms and In smaEP, rural communitfes, Thzs pspulatfon estimate excludes the south portfan of the Crty sf Rsxbaro which lies within rhe water- shed.
Lake Mishfe has a perimeter of approximately 15 miles, If one makes the liberal assumption that the 350-foot contour intertal takes in a 18-chain (660 Pee) wfde strip of Band along the lake shoreline, the total area comes to aboar 1,200 acres. The boundary of this acreage is not surveyed, marked? nor posted, Even a casual reconnaissance reveals shar many people use the periphery of rhe lake Intensively eo draw on, or eo renew their spiries in different ways and fljcrns. A canarkably rich variery sf artif scrts ~:swmainly
of aluminum, rubber, and plastLc abound in the woods around the
laka, but e- en mure z-xoris i-terns, as an s'eighteea-7nour" girdle
draped unezrernoriiocelp ~sr.che bushesphave been lefr behfnd. No qaeskian that this reservoir has great appeal to pesple who are dram there t3 rwreare, and to renew or consume thefr spirits.
Intensive as thla pexZpRezr3. activity appears to be, it probably
and ft influences outflow from the watershed in a number of ways.
Occzisionaily, one stiU. finds relatively primitive modes of tillage and minimal efforts ac follswfng proven soil conser~atio~zpractices
(Figure 28). Over zhe years) the Agricultural Extensisn Service and the Scil Conservarlon Service have done a very commendable job on stimulating land owners to employ improved farmXag practices,
The improvements are seflecred in the current level of soil manage- ment, fse~l-crd~ngthe use cf cover crops, strip croppi.ng, permanent cover on the steeper slopes, and in Peavizlg filter strips sf grass or native vegetation along streams, around ponds, and between fields and roads. It does appear that such sedimentation as is generated Figure 20. Left: A field about 114-mile from the west shore of Lake Michie being prepared for sowing in late April. The use of horse or mule-power for tillage is a rarity on this watershed. Grass and weed filter strip along margin of field at extreme right helps to trap sediments carried by surface runoff from the cultivated area. Right: View of part of same field shown on the left. The grass and weed filter strip is not continuous, and in places the land has been tilled to the road ditch. In the past, the road ditch, along with accentuated overland flow from the road crown, has transported considerable quantities of sediments toward the lake. Photos taken April 25, 1972. by the ruzreas farmlng practices 5s do-m 60 tolerable levels so far as streamflow from the upper Fia~R%vm draina
There are other farm practices which are likely to adversely
affect water qualf~y, Narfonwlde, rheme is a rendcnsy for many
Lake Mfchle watershed, At any pate various studies, as reported
impressive, and some of the constxtuenice eventually reach reservo2rs
in eencentrarfons high rnovgh to be undesirable, Similar sratemenss
are likely applicable to pesticides, but in neither case is che
serioueness of contamination, eutrophication potenefal, and ocher
facets known specifically for chis drainage basin, 811 we can
surmise at thls Btagc is that fertilizer and pesticide movement
is as conzplex as zhe soils which make up the warershsd, The
variabflfcy in sodl properties likely to affect movement and leaching
loss has been dearly emphasized by the basic study of Lutz (1969)
The use sf tributary streams to provide hog wallows fs a
common practice on the watershed (Figure 21)- The wallows are
usually situated ae the heads of small, perennial streams, ox
astride of them, but fn any event En a position co contribute
to the pollution sf rhe stream and the reservoir, a form of Figure 21. One of numerous hog wallows on the Lake Michie watershed. This one is situated on a tributary of the left fork of Dry Creek, only six-tenths of a mile from the lake. Vigorous growth of young hardwoods species along the stream bank reflect the enriched habitat from which they are drawing their sustenance, including the "fortified" water. Photo taken May 1972.
Figure 22. Typical stock watering place on a tributary of Flat River. A former brush filter strip has been essen- tially eliminated by intensive trampling and browsing, and the stream bank is subjected to almost continuous "tillage" that moves the soil material into the channel. Photo taken February 1971. The main forms sf pollution from ferrning zcgivity on this watershed, thus, are most lfkely to stem frm livestock grazing, feeding, and watering practfces, as well as from fertilizer and perticlde use, but a critical assessment of the relative impor- tance sf each proved to be beyond %he scope of this study.
As has been mentlosaed esrlier, the dominant vegetatlve cover on the watershed is some farm uf fer~stgrowth, which now occupies nearly two-thirds of the land surface, Current Bogging practices at worst, can have very Latcle rffecc or degiading the water qua- lity, Rarvestfng on most psivaee holdings is prfncipally of two cypes: (a) it either involves a "e.ommer'cial" clearcut, without any follow-up effort coward fores~rerlewal (Flgalre 12, vide sup~s) or (b) it consists of periodic high.-grdlngs, tlrimately leaving the area occupied by trees of no commercial value (green junk)
(Figure 23) or gradual conversion of pine-hardwood types into hardwoods. (Figure 13, vide supra),
Am exception to the prevailing rrecaE-.mer~tsf forest land on this watershed is rhe l480-acre Hill Foresz which straddles Flat
River, with its south boundary being less than a mile above Lake
Michfe. The Hill Forest is operaced by rhe North Carolina State
University School of Forest Resources, and is used intensively for fnstructional and research purposes, It provides a variety of examples of successful forest renewal on the kind of cull woodlands that occupy such extensive areas of the wa~ershed, It also contains demonstrations of siPvicuEtural practices and forest management Figure 23. View of a portion of Dial Creek bottom in which a recent high-grading operation has been completed. The hydrologic condition of the bottomland has remained essentially unimpaired, erosion is absent, but the stand has been reduced, accelerating the the development of a scraggly overstory oi large unmarketable culls, and a subordinate component of hophornbeam, redbud, river birch, and dogwood, mostly deformed, but capable of occupying the land as a jungle of green junk. Photo taken March 1972. operations designed to promote high quality fsrese protarticn on a practical basis, with some early study areas InJtiared about forty years ago, The management activities on this pro>- erty have been supported by stumpage receipts from the operncims that have renovated and improved the growing stock to accepraa~e levels sf qaality,
remainder of the watershed land under forest ccver hsa r;aeio~:i such a minimal. sern'o%ance of f ores;: managemensC cspera~iorns, c: r melding of the woodland potential inro the fam economys rhac ;t does not deserve further elaboration. However it car1 be empila-- sized again that the harvesting practices occur so sporadi~~~lp in both eime and space that any disturbance associared v.iLh rhe operations cannot possibly have a deleterious effect ou tne quality of streamflow. As stressed earlier, the haivesring practfces in the absence of any effort to practice forestry is continuing the process of conversion of pine and pine-hardwood types eo hardwoods, buts the conversion cannot be consfcie~edas conferring any critically needed hydrologic benefits en ;t-tiIs watershed,
Another form of pollution, possibly as yet more & -prjr,exrti;;.; than a genuine threat to water quality steins from tire co,.ym?On practice sf disposing solid wastes mainly along tributarfes in the watershed (Figures 24 and 25). The main source of &his Figure 24. A road drainage ditch leading to a tributary of Deep Creek has become a popular spot for disposal of cans, many of them neither bio nor "chemo"- degradable. Figure 25. The variety of discarded material on the Lake Michie watershed is almost mind-boggling. Seen here is a deposit of refuse that is biodegradable, but some of the paper will blow into the stream before deteriorating. For the time being, the raw road fill and shoulders probably contribute more than the refuse to the pollution of this stream. waste may be the suburban comunlties and the curbanL' establish- ments which do not have access ta an organized scheme of d.lsposal, and possibly an oeeasional. farm tss smaPI. to have surplus land on which ro pile discarded junk, Be that as it may, the stuff is immensely unsightly, and the rich varfery of articles left to
11 gracefPthe stream banks and roadsides, often does coneain items, such as old rnartresses, pfllsws, conrainers of chemicals (pre- sumably all empty), which carmot in any stretch of rhe Smaghatfon impart anrehing merirsriuus to the quality of outflow- Some ftems such as plastic derergenr bottles found beached along the shores of Lake Miehfe may well have been transported from headwater points all rhe mPes dsw'i7.streas-nduring periods of high stomflow,
Finally, so far as water 2 Constituting an area which is chiefly residential buc where same farming is carried on (Webster8s New Collegiate Dictionary). be ascrxbed Lo suspended sediments sranspsreed by flow in road dic~hssleading directly into the stxeamr- After the initial imprcvexenr fn farming practices had cceurred mthfn the first decade; in che hfe sf the reservoir, stksequene increase in road bullcifng and ehr characrer ana frequency of road mainte- nance pz-irises have assured a eoatznuxng supply of loose gar tlcdaee mrcer faeilf rating easy trarrapczt f nsm road shoulders as12 5;tcTizs lnco SCLC~Z;~channe~s~ IL ai.j bcjc*ircmy every storm I-~~kavy eueugh to generate overland fPww iron rsad surfaces pro- duces tnaugh sediment transport to color the srreams; generally, the depth uf color is most pronounced in the nine mikes or so of zhe maxi ~hannePof Plat Rlver (Figure 263, Aboue -Qusn+-fty ---- ----General, As was emphasized in the i:?trodu@tory section of this report, streamflow is rhe prcduct cf precipirat$on, but not all rhe precipitation that falls on any gTven basln flows imme- diately o~tof it, The complex runoff plocess can be thought of as Iravang several flow components- These rEsw iomponents include: (a) channel precipitation which is khe psvtron falling directly meo cke flowing stream; (b) sveriand flc,x .pjn~ckis the portion of nee precipitation that falls to PnflPtsate, thus running ever the sox1 surface fnto the stream chaiurek; -.-Bottom: Flat River at same location as shown in A, following a 5+-inch thunderstorm of great intensity on July 10, 1970. A sample of "water" taken during the storm from a nearby road ditch showed suspended sediments to equal one ton per 5,000 gallons of ditch flow. the surface soil mantle (including root holes, anfmal burrows, structural interstices, and related avenues) that it is considered as part of the stom hydrograph; and (d) baseflow, or groundwater outflow, which fs the component of streamflow or runoff occurring i.e: the channel between storms, and fs the "oldest" water issuing out of any given basin, In upland drainage basins under good forest cover, about $5 percent of rhe total outflow occurs as baseflow, En eastern United States as a whok, about 80 percent of total streamflow is base- flow, and only 28 percent as direct runoff or stomflow which is the sum of three flow components, viz: channel precipitation, overland flow, and subsurface stomflow (Hewleft and Nutter, 1969). In the Lake Miehie drainage basin, the geology, the generally compact subsoils, and the influence of past land use suggest the high probability that direct runoff amounts to a larger proportion sf the total basfn outrflow than would be found in some other physiographic province, especially fn basins under well-s~ocked forest cover, It may be useful at this point to summarize the outflow records from the pasr several decades to provide a more concrete picture of the water production potential and capacity of the Plat River drainage basin. The average discharge rate per monCh 2cr the 44-year period at -I/' These rates represenr sumaations of dafly values for Table 8, Average nanth3.y dfa~har@,based on 43 years of data, of Pfac River in cubic eet per second (cfa) at stacisns above the lake and be QW the dam, respect- ively, wish record lows and h9 s, and year of oc- currenee shown for each, Boeb: USDI-Geologic Survey Surface Water Records, Record --Record Ae etation 900 fr, Month below dam -Low e f s efs P P January 6,%07,20 8,40 February 3,946.22 257,SO (4934) March 8,308,35 P,941,00 (1931) Apr9b 7,574,02 1,290,20 91942) M~Y 3,593,98 778, 00 (1964) June 2,584,28 401.20 (1955) July 2,839,04 4,80 (1966) August 3,143.05, 2.60 (19663 September 2,750,52 9,00 (1943) Qotsbsr 2,085.99 9,30 (1943) Mean Annual Bischarge 54,785,71 mfle below the Bahama Brfdge, 2t is of Interest to compare the yield records from sr wleh those ot the Flat River basin above the reservoir (Table4 )-. Dial Craek watershed is only 4.71 square miles %n extenr, whereas che Sasm above the upstream etatfon on Flat River extends 150 square miles; consequently, che only basis on which to compare yaelds from such disparate areas is by reducing drseharge rates eo rubfz feet per second -40- Table 9 , Comparison of discharge recorded at two peinte on Flat River and near mouth sf Dial Creek in terms of -efsm (cubic feet per second per square mile) based on 43- year msnehly averages. Flat River Station Flat River Statfon U'2-mile upstream Dial Creek 900 feet below dam ef srn -cf sm -cf sm January P,34 1.25 1.03 February 1.81 1,%0 1.66 Apri l. 1,47 1,46 1.48 July "58 .59 September ,52 October .42 Nsverab er ,64 December -.88 per square mile (i,e,, their respective cfsmPs), Data are also shown in terms of cfsm's for the station 900 feet below the dam but it is well to recall that the natural flaw of the Plat River system at this point is obscured and altered to the extent of Durham's draft on the stored water. In comparing the natural flow of Dial Creek and Plat River, it is of interest to note that the yield per square mile during the months of January through Aprfl is greater from the 150 square miles of the Plat River basin than from the 4,31 square miles of the Dial Creek watershed which does not csnt~ibuteto the record of the gauging station above the impoundment, What appears more significant is that the May through August yields per square mile from Dial Creek average higher than the yields from Flat River for the same months. Take the June average, for example, when the yield is 0.09 cfsm's higher from Dial Creek, This advantage may not look very impressive, if we recall that one cfsm is equal to only 0.7013 gallons per minute from one acre. But this amount quickly accumulates with time so that eac.h acre within the water divide of Dial Creek yields some 2700 gallons more than from each acre of the Flat River basin above it for the month of June. Even this differential may not look impressive, but taking the 3384 acres in the Dial Creek watershed, the June yield comes to more than 9,137,000 gallons above the amount that would be available if every acre were performing only on the same level as the acres in the Flat River drainage beyond it, There is no readily available explanation for the observed differential; it may stem largely from major geologic differences between the two basins, and the influence parent materials have had on soil formation, structure and related eharacteristfcs, Within the last 30 years, there has also been about six percent more of the Dial Creek basin under forest cover than is mue for the total Plat River watershed whi& includes also the land in the Dial Creek drainage, It may net be useful ra speculate further why differences in water yield, modest as they are, should emerge; suffice it to emphasize again ehat even small differences in cfsm9s from large basins will ezm~unc to very sizeab3.e volumes of water over a month or longer, The lowest monthly dfseharge flows recorded at the station above Lake Plfchie have been the 21,25 -cfs in September 1968 and 21.36 -efs in November 1933. The next lowest discharge on record is 38,56 -cfs in October 1941. Even the lowest discharge on record for the 43-year stretch would provide more than 400 mfllfon gallons, allowing some loss to evaporatfoa; and this amount mEght suffice to keep the current poptiPation of Durham operating for a month or so without too much stress, Wfth the water production potential of the Lake Mfchfe drafnage basin, Durham, or any other munfcfpaPity with a comparable Piedmont watershed, needs to direct its attention more toward attaining quality of outflow than quantity, General, Fr~mthe sueset it was recognized that the yield from the Lake M~chiewatershed was more than sufficient to supply the needs of the Clby of Durham- Emphasis in the analyses has been directed, eberefore, toward determining rhe influence of accual and poss%ble chsnges in iand use on the timing sf water yield. Flaw daca for analysis have been avaklable from records mafnrafned ac three gauging staeaxis whlch have been monitorfng streamflsw from rhe watershed since completion of cke dam. These statism have been Installed and mainrained by the USGS, Srazion Iocazlfons and available-- data, Major analysis of trends concerned the data fro3 Station No- 2-855 of the USGS series, This statfsn is located on ~heFlat River near Bahama NorLh Carolina, specifically at laeitude 36' -90' -57'' N, longieude 78" -52' -44" W, one half-mile upstream from the head of Lake Michfe, and msnitorfng the flow from a caechment area of approximately 156 square miles- Records are available from 192% to date, but In che analysis only the data from 1926 through 1968 were used because at the time of the analysis later records were not available in usable form. Data were also analyzed of the records for the Dial Creek station (No, 2-860) which Is siruated three-eighths of a mile upstream from Lake IvIichie specirieally at: laeitude 36' -10' -36" N, longftude 78" -51' -24'WW, The catchment area is 4.71 square mfles. At the cxme of the analysis the records covered the period from October 1925 through September 8968, but again only data from January 5926 were included fn the study, The ehird sration fn this invesrigatlon is Noo 2-865 situated on the FSac River, 9 0 feet beam the Lake Mfchfe dam. Its specffic Lueatfon is Patft~ide36" -08' -55" N, longitude 78" -49' -43" W, Re~o~dsfor this station were available from September 5923 throu h September 1966, with the exception of water-year 1960, Srnce the station is located below the dam, It has a slightly larger catchment area, estimated to be about 171 square miles instead of zhe 168 square miles reported for the Lake Mlchie watershed, The station above Lake Michie and the one on Dial Creek are rnsn5toring the natural flows from their respective catchment areas; the starion below the dam records flows from additional trfbutarles of the Flat River drainage basin, including Dry Creek, Rocky Creek, Muddy Creek and other minor streams flowing directly into the lake. But tt should be emphasized that the recorded flow is modified by withdrawals of variable amounts to supply Durham's needs, and, at least in past decades, the power plant has been closed down to conserve water during low flow periods, thus involving an additional aberration from che natural flow sf Plat River, Initially, it was expected chat adequate precipftatlon data would be available khac cod$ be used jointly with the streamflow data as an adjustment factor, Search for dependable data spanning the entire period proved futile; moreover, for stations directly on the watershed, only daily totals were available with no fnfor- matfon on intensity sf storms, Acesrdfngly, an alternative approach to dfreet adjustment was decided on, The data for rhe 42-yeas period were divided into three consecatlve segments of 14 years eaeh. More by desfgn than ehanee, the three segments correspond to three periods durkng which human steupancl7 and ineensfty of activities may have been materfally different on the watershed, The first 94-year period spans the great Depression, the second, the war years (WW IF and Korea), and the thfrd, the post-war or industrfaf expansion years affecting this region, as well as the nation. Analysis of streamflow trends, The underlying trend in streamflow for each 14-year period was analyzed, and the treads from segment to segment were then compared to determhe if dffferences could be detected, In this approach two assumptions, reasonably tenable, were made: (a) that periods were of suiffcient duration to reduce variarfons in precfpftation among segments to tolerable dimensions, and (b) thar the trends within periods could be considered independent, A fit of linear trends was then made to eaeh segment, As an fllustraefon of the obgerved t~ends~resultsbased on data only from station No. 2-855 are presented here (Pfgure 27). The trend during the first period was one of increasing flow. This trend corresponds to a period of considerable change, Census data on fa~ms,discussed in Chapter IV, indicated some decline fn farming in the ~wenties,but a considerable increase in the acreage of land coming back into farming toward the end of the thirties, It was a time when probably as much as ten to rwelve percent-. of the land in the watershed changed from worn-out fields back ineo young pine, In the initial stages of this conversfon, the areas pre-empted by pines would not yet have resulted in a major transpi- rational draft on soil moisture, but prior to crown closure, &long with weeds and grass, they would have started the improving of the infiltration capacity of the soil surface. Lt was also a time when markets for pine pulpwood began to open up, and this market very likely hastened the conversion of pine and pine hardwood stands to essentially pure hardwoods, It was probably too early for the programs of the Agricultural Adjustment Administration and the Soil Conservatfon Service to have had much impacc on this water- shed, bur all of the factors mentioned, and perhaps others, may have contributed to the observed increase 4n flow from rhe water- shed. We can only surmise, not prove, that combined influence of all of them resulted in an upward trend in outflow from the basin, The overall trend in streamflow over the encire 42-year period indicates a reduction in average monthly flow. There is a distinct difference between the trend in the first period and the Past two (Figure 271, these differences agreeing wfrh the general decline in outflow, Since the three trend lines have been fitted for separate time segments, the line termini need not coincide, It is perhaps relevant to speculate why a decline should have occurred, In pine stands amenable to use of fire, failure to follow - a schedule sf prescrfbed burning is resulting in an ever-thickening understory of hardwoods, comprising both tree and brush species, and brush growth is thickening in all other stands, with the result that Y.osses through interception of rainfall are fncreasfng even though the acreage in pine type is declining. Where clear- cutting of essentially pure pine stands and selective cutting of pine in mixed stands have occurred, there has been an improvement in the hydrologic condition of the soils through gradual decay of the root systems of the cut pineso The decaying stumps and roots of the harvested pines open up massive channels that facilitate infiltration, percolation and deep seepage; the end result is a greater capacity for detentfon storage in the soil, and a larger proportion of rainfall going into deep seepage, thus tending to reduce peaks of stormfPow and maintain higher baseflow levels during interims between storms, Since the dam was completed in 3.926, there has been an increase of about 15 percent in acreage of woodland, Some of the stands have matured in that tfme, some have grown older, and generally the total forested acreage is now making larger drafts on basin soil mofstu~e,especially since there are more acres doing it than four to flve decades ago, Ansther factor that possibly has altered the regimen of Flat River is the fncrease in acreage of farm ponds from essentially zero to nearly 1300 acres, 'Withour doubt, the ponds have helped to reduce peaks of stormflow, but direct evaporation from them has also bee^ a factor as pointed out earlier, The estimates ok evaporarisn losses from ponds made in a preceding sectLon were based on Lake Michie data, but pond losses per unlt area normally would exceed those from a large impoundment because the shalEower depths, less "Lmrbulence and related hccsrs would tend to reduce the energy requirements for evaporation from ponds, In using pond water for irrigation, additional evaporation losses have been accentuated, These factors, and conceivably others, have modified the timing of runoff, bringing extreme highs and Paws closer together, while overail yield has been reduced, On balance, the outcome is not necessarily undesfrable; fn fact, it may be preferable, viewfng it from the standpoint of wzter ava%lability, flood hazard reduction, uniformity sf flow, and related angles, k trend toward stabilizazion of rhe watershed and warev avaiPabfEity has occurred. Here and there, forturtously at zero cost, a forest crop has been regenerated by nature and hzs developed into an economic resource for some landowners within zhe watershed, renewal is parr of the whol~package, estimate for reasanabky weil-watered regions, sub as c5e Lower Piedmont sf North Carolma, wodd be about 0,09 -mhs of water by tree crowns. F'sr forests in eastern 'Jnfted States, a reasonable a ccmplece clearcuseing of ~n entire basln Cansidering that an acre-inch is equivalent to 27,514 gallons, the gain in water may Look, at fizse gksnre, quite impressive, Realistically, what sort sf a task would one face in actempelrag to achieve and e.2 maintain the fndieated increase in streasnflow, Osae wsalld aeed to elearcut all the trees and brush in the fbresr of a gigen basin, and to prevene regrowth, either by p~rfadicmowing, application of herbicides, or both. Water demands fron @he Lake Plichfe Sradnsge basin suggest that resort to sueh a drastic p~ocedureis not only not necessary but surely not sensible, noa or ever, Pf one made rhe Uzspian assumption that someday all forest Band within 'the watershed would be placed under management, one could expect after a fairly long adjustment period, a dffferent picture to emerge, h essential coroPlary to these psstulations is that even- aged management is the appropriate system to adopt for the magex commercial species eomprfsing the forest types in the basin. An additional coroflary, considering the variation in site qualfty and in species growth rates, is that an average rotation length of 48 years would be desirable or necessary with dearcutting as the sensible method of harvesting, On this basis, each year the equivalent sf about 1,600 acres of marketable stands on average site quality Sand would be clearcut and regeneraTed. From this acreage we could expect ro reaPSze the extra streamflow of about nine kehes in the first year, or s~methfngin excess of over 400 million gallons cJ The yleld increase indicated above would 5ssue not from a single basin, but from scattered porr5ons of the landscape rhroughout the watershed, Under an organized system of forest management, the increase would be repeated each year with sue- eeeding composites of 1680-acres subjected to a harvest cut and followed by regeneratton of a new stand, Successful forest renewal on each cieareut portlsn, together with the development of subordlnaee vegetation after the hitial year following harvesr, would result in transpiration and anzerceptisn losses gradualdy buildfng up until possibly by the tenth year the yield increases would have dropped to zero, Some studies have indfeated persisfence sf yield increases beyond two decades (Reinhart, L965), but under Piedmont conditions, a decade would appear eo be ample time fox initially-increased yields to recede to zero, The cumulative effect of a series of suecessfve elear- cuts suggests that an annud yfeld increase of a bfllfon gallons of essentfallg sedfment-free water would be a very conservative estimate of gains in water production under the system proposed. In addition, sfgnificant returns from a -sariable harvest of forest products wcpuYd be realized each year, At current prices, the stumpage retwms from harvestfng 1689 acres of well-stocked marketable timber would amount ts about half a miPEion dsflars. Moreover, with managed forest resources, various amenity values would be developing to levels far superior to chose existing with the current green junk which can be almost Impenetrable because of the mass sf briars, vfnes, and other brush vegeta~ion, Trade-offs and economic i-mplications, Numerous references in the literature emphasfze the superior infiPtration capacity of soils under mature hardwoods, For example, at the Coweeta Hydrologic Laboratory of the U, S, Foresr Service near Franklin, North Carolina, it was observed that an 11-inch storm over a period of 17 hours produced no overland flow from a basin occupied by well-stocked, moclntafzl hardwoods, Merrfck and Johnson (9952), and Mather (1953) in New Jersey found it possfble to apply annually from 400 to 600 inches of wasee water into a hardwood stand without producing surface runoff. The frequent implications from such observations are that in terms of fnffltration the hardwoods are superior to pines, What is generally not emphasized in such comparisons 5s the strong likelihood of a vast difference in the lapsed time that the land has been under the influence of the different forest cover types. The forest floor and inffltration capacity. There are noticeable differences fn che character of the humus type which accumulates under pure pine fn contrast to pure hardwoods. There are differences also among hardwood vegetation types, as, for example, between yellow-poplar and oak-hickory (Henderson, Twenty years after planting of 1obPolPy pine on a cut-over site on Hill Forest in the Lake Michie watershed and preventing development of hardwcsd brush unders-cory by periodic applications of herbicides, the Sitter layer (E-layer) amounted to 8,560 Ib. per acre, and the fermentation (F-layer) and the small amount of humus @&=layer)amounted to an estimated 20,400 lbI per acre, on an oven-dry basis, Where an understory sf hardwoods and brush species had been permitted co develop, the. respective values for L-layer and P- plus H-lagers, were 6,400 Ib. and an estimated 14,100 I'D, per acre. 231 contrast, in the hardwood stand pictured in Pfgure 4, the E-layer anmunted to 4,800 Eb, per acre, and the F-layer, together with some semblance of an H-layer amounted to approximarely 8,200 lb, per acre, oven- dry. Metz --et aZ. (1930) in the Virginia Piedmont showed that the volatfle constituents in the total liosr thar accumulated in 16 years under a Virginia pine planting amounted to an average of 7.3 tons fn canerast to EL tons under companion plots of Loblolly pine on rhe same soil type. Hardwoods, generally, have less aecunulation of organic debris on the soil surf= %han do pines; In brief, the fitter breaks down faster, and becomes incsrpora~edmore readfly, in the surface mineral soil. It is this process of ancorporation over many decades that fmprsves structure and results fn the high infiltration capseity under hardwoods, and makes the hard- wood soil profile appear in this respect superior to pine, In addition to ro%al amounr of floor accumulating under different forest types there are measurable differences in physical and chemf cal character lsries between the -.mull hum~stype which tends to develop under hardwoods and the mor hums cype devalopfng under pure pine stands with varying P intergrades depending on combfnazions of pfne-hardwood mixtures. Likewise, rhe difference between the rnffltration rates sf the two major humus types (mu11 vs. E) being discussed here is quise large, hevertheless, the forest kloozs thac have developed under ttre pine cypes on he Lake iqichie watershed prov5.de adequate procectfsn rs che soil surface and also adequaee mfiftratian of rain for the prevailing storm inten- sities, One may conclude that a rrade-off of pine type to hardwoods on the basis sf need for better PnELlt~ationis not justiffed, even though the conversion is possfble and continues at zero cost. Is pfne to hardwood conversfon justFffed fo~fncreased water yields? Again, it should be emphasized that the con- version from pfne to hardwood can be accomplished at zero cost, if opportunity costs are ignored, In fact, wherever stands include a nominal component of marketable pine, the conversion in Piedmont uplands may occur with substantial, instantaneous returns to the forest swner, In at least one operation, it is known to have been achieved with returns of $300 per acre in stumpage receipts, prfncipally from che few high quality naturalpine stems per acre, and wJth no orher inputs on management. This level of return, of course, was realized only at the instant of harvesting, and does noc take Snto account the losses thar would accrue from subsequent occu- pancy of the lands by hardwoods, regardless of subsequent positive cash returns realized from them. Conversion to hardwoods can be expected to measurably increase water ySel.ds. The increases result from reduczfons in both transpfration and interception losses which on a yearly basis are greater from well-stocked stands of pine. Between pines and hardwoods there are readfly recognized dffferences whfch influence the amount of water Post through evapotranspiration. Hardwoods differ from pines Fn density of crowns, in growth form, in branching habit, in total leaf area per tree, in depth of rooting, in roughness and thick- ness of bark, and in the habit of shedd.fng foliage. Hard- woods are deciduous, most species dropping all leaves each autumn, in contrast zo pfnes on which at least the current year's crop of functioning green needles persfsts while older needles are shed. Wfthin similar sites and climates, these differences in form, growth habit, phenological responses, and related elements have always been assumed to result in substantially greater water consumption and "loss" from growfng of pines rhan of hardwoods. Although hardwoods transpire through the bark and buds during the months i hat rhey remain leafless, the loss on the average woulci appear to be small in comparison to the amount of water transpfred and intercepted by green, functioning foliage of pines, Studies by Kramer (1942) indicated that transpiration declines to about one-fourth of the maximum rate in several pine species when temperatures fall to 32'F. However, it is well to remember that on the Lake Michie watershed in an average "dormant" season only 69 days occur when temperatures drop to 32'F or lower; consequently, an ample number of days remain in those months when hardwood species are leafless while pines retain the capacity to outstrip the deciduous trees in the amount of water transpired and intercepted. This much is surmisable from observation and intuition, not from critical study. Any attempt to apportion or estimate the amount sf water lost through transpiration apart from that lost through interception, even hardwoods versus pines,leads one on difficult ground. The variability and complex interrelatfonships among soils, geologic features, climate and vegetation make almost any generalization an over-simplification. One way to avofd the problem of estimating the losses separately has been to study streamflow in relation to precipitation and manipulation of vegetation. Some @f the more definfkive research caneerning vegetation man~pulatmnon furesced watersheds has been done af th2 Coweeta Hydrologic Laboratory of the U, So Forest Service sn the Appalachian Msunrains of western North Carolzna, As reported by Swank (P968), clearcuti~ingof a mature hardwood forest fo~hmiby regeneration with white pfne resulced in nearly a four-fneh redtletion in stremiflow ten years afcer planting of the p%w, In this 40-acre experirnentd watershed, ehe annals1 water yield from ehe basin affstr erom closure ocsasred and the s%"i becarria: occupied by rhe 10-year pfile picrln$atisn was roughly four million gallons Less ehan it would have been if occupied by the original s~andof hardwoods. Swank indicated chat sereamfPow reductions were greatest during the dormanr season, althoagh even summer months, except August, contrfbuted to he annual decrease in sereamflow, ALSO at the Coweeta BydraPogic Laborarsry, HePvep (19679 showed tha~roeal Fntereeprfon lass fn w'hlte pine increased with age over a range of PO to 65 years, and in pfne srands of every age tested the total Psss exceeded the Posses caY- culated for mature hardwoods. In northeastern U, S,, Mader --er ale (1965) approached the problem sf water loss through study of soil moisture depletfon, and they found depletion ealy m. the season %o be greater under white pine and under grass than under a stand sf oaks; they also found chat re- saturation sf rhe sofL in rhe fall occurred earliest under rhe oak seand, En concsast to fanding a solid basfs for making direct esrimates of rranspira~fsnlasses between pines and hardwoods, there is ns dearth of information on estimated fntereep~fan 1 Losses wh~chhave been stadfed in great detail (Zireke, 1967). In generai, the findrngs on fnterceptfasn conform with chose reported by HkZwey ($967) from she Coweeka researches. How- ever, s~ernflooditferences among species are not so well documented- As might be expected, sternS9ow does vary both ansang specses m3 among tree sizes within speehs. Mader of trees ~ZOT~Ithree inches to 13 inches in DBEE@iameter at breast hefghr), red saks exceeded whfre pine in arnsunt of scemflow over ark ent%re zange of sununer storms from 0,2 inches 90 %,$ Enehrs, In the largest trees and heaviest storms, the difference fn favor of saks was about ten gillons per tree per storm, These ata lend emphasis to the suggestion made by Swank (P9&8j, and many others, thac over 90 percent of th; rainfall fnrercepted by confferous species must be counted as a loss so Ezr as streamflow is concerned, Er has been implfed earlier that even during summer months differences in swface area and character of folfage make it appaar likely that phea intercept greater quantities sf rain than de Rardwosda- The equations developed by Belvey.and Patrfc (1965) p~ovidea bash for estimating interception of 0 about rdne inches for pines, floor, There are very few data to provide a bhjta Sf>: es;i- mating the extenc sf inrereeptfon Sssses by sb5 i~jt~;cr wood forests of the southern Appalachians fnre~:eptim oy Bitter may average as much as three inches annua2iy Fcr ?Ire soil, but obviously antecedent conditions can introduce considerable varfatisn from thls average value, Since forest floor layers under ,gel&-stocked pine srands always exceed those under hardwoods borh in weight and depth, it is safe to postulate that losses Erom interception by organic debris accumtaia~ionsunder pine are measurably greater than losses due es rhtj fl~orlayers mder hard-- woods. At the preserlr stage of knowledge, "Le difference in this type of Loss eancst be sdtisfactorily quaz.tiEPkd, En conclusion, transpiration, canopy interceptmn, and forest floor interception comprise the principal loss cf water to streamflow from forest-covered carchmenks, and the loss from these sources, at feast on an annual basfs, is greater in pine than in hardwood types. The fourth phenomenon in this phase of rhe hydrologic cycle is sternfiow, the process by which water fs eased downward to the soil surface at the base or' the rree where conditions are mssr. favorable for rapid fnfiftration, Even with stemflow, the hardwoods appear superior to pines. Thus, one is led to the conelusion suggested by DeEfs (1967), as well as others, that in terms of water consumptfon and losses to streamflaw hardwoods possess sufficient advantage to make them the preferred species in drainage basins managed for water yields , Ic must be emphasized, however, that the adusanrage sf hardwoods in the growing season is not sufficiently large by any of the criteria considered above to justzfy t~jectzsn of pines, The decision to convert pine types ~c kard~c~ds must be made solely on the premise that the exzsa ws-re;: yields realized from getting rid of pines can be stored as a teage against periods of drought, As a rough guess, cke gain .from converting well-stocked predorninmtly pfrie rypes LC I-7mpscaSLy stocked hardwood stands would amount probably fe $~>.:e5nehes of water annually most of it being realized fsezr; L~Eextri 'Fdormant'beasonyields, Are pines devoid of hydrologically beneficial-. - &arscterLs- tics? In extoPling the virtues of hardwoods in waaer~hed management, it is easy to overlook at least two importazt characterfsties of well-stocked stands of pines, thsy produce annually a copious litterfall ranging, after cralia closure takes place, from 3,000 to 6,000 lb, of a~ren-dry material per acre. Thfs annual "complement of m~iLh"~rm~des a very stable and durable proeeetfon to the safl. ;nd is we subject to drift as are hardwood leaves, For reservsir xargin proteeeion, pines are elearly superior to hardwoods, Second, pines have a deep penetrating root syscepr, that begins to decay when a tree is eut or dies frsm nseoca; causes, The roots of such trees open up massive channels nn izke soil, facilitating infiltration, deep seepage and aeraticin; moreover, the organic matter (celhPosic material) pro~xdeseaargy for soil macro- and micro-fauna and microflora QFfgure 28), Many hardwood species have the capaci.ty ro sprout froin stu~~ps,and roots; this capacity maintains much of the old root syseem living and functional, hence incapable of ameli~raring fhe soil to the same extent and depth as 9s plnea Tr 1s uu,mi~ed, without proof, that two to ~Rreegeneratiom cE pire harvested by clearcutting at 38- to 48-yea-c. fntecvabs wLL3. I:C~);C*LC ihe hydrologic condition of Pfedmont sofls ons side,:&^^ mcre than would hardwoods with a rotacfan long snough to produce hn acceptable volume of usable wood on previously waded upland soi1s. The importance of the soil amelioration ftmc.tion of pine roots cannot be over-emphasized. An average pine 12 inches -dbh has approximately 22 cubic feet of marketabee wood in the bole, but also some four to five cubic feet of woody material in the tap root and main Eatarals BXC~.PII%V~0% the sttannap portion above groundline, As the root system of e growing tree expands, it obviously causes some soil compaction bur also some lifting to loosen parts of the surface profile, The amel~orasfve action of decaying roots of cut or dead pgnes is s@ Large that Figure 28. Top. Portion of the root system of a pine growing on a loamy fine sand to a depth of 4 feet or more. Note the well-developed tap root and the prolifera- tion of laterals. Bottom: Portion of pine root system bull-dozed from silt loam having a hard pan at about 30 inches. Note blunting of tap root as it encountered the pan. Total all diameters 2 1.5 hundred years, structure and in rimb~isf trees per acre above the 3-inch acre, 38 years deer: estab8Psisment; in contrast, che ~;aT.eable hardwood stumpage ac today's prices might amount ac nest ~o as much as $250 per acre, Bcth of chew outcomes are based <'in the physical prod~ecivizgachieved on the me hand by Virginia pfne on an average si~eand by mixed upland hardwoods on some 3f the best existing upland foresr soils for hardwoods, estabffshed by chance natural regeneration at zero cost and wleho~utappPfcaLPon sf: managemens:. As nenefsned, the present level of growing sro& and volume production has been attained in 38 years far the pfne and apprsxfmateSy 70 years for the stand of mixed harrl%sooda, Under these cfreums%ances, many- well-scoeked pfne svands, naeuraEly regenerated but unmanaged would attain a higher level of production than that shown in the exan~pPe in Table 10. Conaelrsely, unmanaged mixed hardwoods could nst be expected to do better than the level shown in Table 10 over anjr exceaded area because of che scarcfcy of suffiesensly ierrile upland hardwood sites, From the forest pradmctfom etandpoF-nt, %he economic advantage in unmanaged stands 3s clearly in favor of pine, the dfscounted present value at six percene, for example, suggesting a differential of 8 to l for pfnc, The discounted differential, sf course, might be csnsideuab3y Less if hardwoods were compared at an earlier age. This cursory analysis suggests that che costs of maintaining junk hardwoods or permitting them to preempt aPY of the 65,000 acres of L,r,-eest %and on the watershed appear to come I~xgkt, "re warer manager should take small comfort from this sort of a trend or development, for the unmanaged jur,k hardwoods do eonsetme essentxaily as much water, or even more than produetfve hardwood fsreses, albeit less than well-stocked stands of pine, On the other side, the individual forest land owner is not concerned about water productfon, hence he would have to look to other justificaelons for action to improve forest produetfon on his holdings, Viewing the situation from the outside, %twould appear that any forest owner who has as much as $60 per acre fn stumpage among the green junk on average pine site quality %and could expect befrer than a six percent return on his investment by harvestfng the marketable material and plowfng the capital back into establishment of productive pine stands, Many landowners have this level of usable wood resourze~iQR thefr existing forese holdings, despite the generally decimated eondieion of the stands as a whole. As Timko (1962) emphasized, forest tracts have more attraccfve uses fox aoae), aad LLie preferences which do not accsmodare wafting far trees cc graz in size and value. Many ocher factors and reasons ZOIEPLC be cited to suggest why land owners are dasesuraged from under- takfng the type of venture suggested in the preceding paragadsin but deeailing them here would go beyond the purgose of ~hks study. Ir should be added, however, khat foresf managemeal requires fafrh ehat this land hzs a futare, and it involves sacrffiee of income new to leave a Eegaey of &alltreEa for the use and enjoyment sf whose who come af%ef US- In summary, the condirion of existing forest cover cn the Lake Mfehie watershed is such that it does lltrle more than hold the land together, but a highly productive foresr would perform an equally satisfactory hydrologic function, while acerafng ocher values, both substantive and intangfble, En cons~derfng forest-covered watersheds, Meadows and Naylor (1969) have indicated that a poinr may be reached when it would pay to use land to produce water rather than to produee timber as the primary goal in land management. It is bellewed that conditions on the Lake Miehie watershed are as yet a very long way from having reached a stage in water produetion to forego he possfbility of managing the forested portions for timber, Solely from the standpoint of improving detention storage or rain, the production of several generations of pine crops woudd appear fully justified. As emphasized earlier, pines have a mare effective capacity than hardwoods to perform deep tillage and to faeflftate nominally deep seepage, There fs no vegrtatfon that can aceomplfsh this function more effeerively ~hanstands of pine, and on rhe eroded Piedmont uplands with compact subsoils, "ti.fIagew wfth pine culture would appear to be the most sensible treatment for watershed lands for several decades ahead. By that cime the sites may be sufficiently improved to grow desirable hardwood species on the uplands, or strains of hardwoods may have been developed that will do better than the native strains available for reforestation at the present time. On thPs water- shed, if demands for water rose high enough to necessftare changes in land use from the present level. of activity, and the present aPPoca%$on of uses, an intermediate stage of permitting the land to revert to hardwoods might suffice before it became necessary to convert woodland to grass. Whar about grass as an alternative? Many studies have indicated that grasses use less water than do trees' Gther herbaceous cover shows the same types of gains; EOL example, Croft and Monnfger (1953) demonscrated in Utah, that replacement sf an aspen with the understory herb vegetation increased water yields by four inches; subsequecL removal of the herbaceous vegetation produced an additional four inches which also became avaflable for streamflow, On the San Dimas Expertmental Forest, Corbett and Crouse (1968) showed that in years sf average precipitatfon (26 inches) an estimated 5.3 inches could be saved by converting from chaparai brush to grass. In the Piedmont, Metz and Douglass (1959) studied mofsture depletion to depths of 66 fnches fn soils under pine, pine-hardwood and broornsedge. The forest vegetation used considerably more water and to greater depths than the more shallow-rooted grass. In a paired watershed experiment at the Coweeta Bydrologfc Laboratory in western North Carolina, Douglass (1967) compared a mountain hard- wood catchment wfth a companion catchment from which the same type of hardwood forest had been cleared followed by planting to Kentucky 31 fescue, He found that by the fifth year afr.er planting the cleared 22-acre watershed to fescue, the total water use by the grass was six inches less than that used by the hardwoods. To achieve the full benefits of grass would requfre using ehe areas for production of hay rather than,for use as pasture, This restriction is suggested by such studies as conducted by Dils (9953) at the Coweeta Hydrologfc Laboratory where one study showed that after 30 animal-use days per acre, the infiltration capacity of the surface soil was reduced from 3.02 inches per hour to 0.62 inches per hour. Prolonged grazing would compact the surface suffLcientPy to induce overland flow. Moreover, large differences between grass versus hardwood cover could not be expected, if high densities of deep-rooted species of grass were employed, and maintained in a vigorous condition tNrough fertilizing. In any event, Sn due time, one might expect a change in the hydrologic character sf the surface soil profiles, possibly accompanied even by reductions in detention storage capacity. Hoover (1950) showed large differences in the detention storage of the surface two feet of soil under an old growth hardwood stand in contrast to soil from an old field abandoned 3% years prior to the sampling. The capacity for the profPles based on his data varied by depth as shown below: Detention Storage of Water Depth from soil surface: P Old Growth Hardwoods Field abandoned 35 yrs. ago. Inches Inches 'Inches 0-6 1.93 0.82 6-12 1,lO 0.12 12-18 0,62 0.08 18-24 -Oa40 -0-09 A11 depths 4.05 a 1,2l These data suggest the slow recovery that is being made by the abandoned field site which remained apparently for an initfal period of five or more years under a cover of weeds and broomsedge, even though eventually a volunteer pine stand dfd occupy the site. Several rotations of pine would undoubtedly be required to restore the hydrologic characteristics of this old field site to the stage found in the old-growth hardwoods, but the pines would improve the detention storage capacity even to a greater depth than was observed in the soil under the hardwoods. If a shallower rooted grass ware employed as a replacement for trees, the expected macro-pores in the surface soil might not differ significantly from that under pine forest cover, but the process would not extend to the depth possible with pines. Development of detention starage to greater depths is a critical need in Piedmont watershed soils, and grasses, which would conserve waterswould not meet this objeetfve. Some examples of stormflow quality and timing. The outflow from the Plat River basin is a summation of discharges from many small tributatirs, named and nameless, as well as from the main channel, in all of which one can observe copious stormfPow at times (cf. Figure 261, and after rainless periods, groundwater depleti.on flows, Where there are channels which are not being used for sewers, hog wallows, livestock watering, or solid waste disposal sites, the water issuing from such isolated basins fn small, clear channels appears eminently potable in quality, As previously mentioned, there are some 296 mfles of perennial stream channel on the watershed, and most sf thfs mileage is under the influence of more than one major land use, A few minor streams or brooks issue from basins complefely under forest cover (Figure 29). The two watersheds in which runoff is being monitored by means of the weirs pictured in Figure 29 have demonstrated very simiPar stormflow patterns. Both watersheds are covered principally by hard- woods, and in both the peak in stormflow is reached in a relatively short time, often in two to four hours after initial rise above baseflow is recorded (Figure 301, There is, sf course, considerable variation dependfng mafnfy on antecedent conditions and intensity of storms, but without data on storm intensity, no critical analysis of the several factors that influence the pattern and magnitude of peak flows Figure 29. Above: a 90" V-notch weir and stage recorder monitor- ing the flow from a 74-acre completely forested water- shed in Block E of the Hill Forest in Durham County, N. C. Except for a nominal quantity of organic debris, largely in particulate form during stormflow, and prob- ably in colloidal form during baseflow, this stream remains clear the year round, Below: A similar in- stallation in Block C of the Hill Forest monitoring the flow from a 102-acre watershed, forest-covered except for a small field at the head of the basin. The small field involves enough soil disturbance to cause the stream to turn "milky" with sediments during stormflow, in striking contrast to the clear flow ob- servable at the weir in Block E (above). has been attempted. Some 30 to 50 percenr of the variation in volume of discharge at point of peak flow appears to be accounted for by a regression using antecedent rairi during three days preceding a given event (AR) and inches of rain during the specific event (R), The eime in hours, ro reach peak flow STP) may very occasionally contribute a little to the significance of rhe regression, as shown below, for an estfmate of discharge (Q) in the watershed monitored by the weir in Block E, rhe top piceure of Figure 29. Q(cfs) = -0.570 -+ 1,115 (.MI) -F 0.738 (Rj + 0.063 (TP), This equation applies to the events which have occurred in the second quarter of the water year (Sanua.ry to March, inclusive); in other quarters, the time to attain peak flow (TP) was not significant. In these small watershed studies, the rapfdity with which many stormflows develop to peak heights has been greater than was anticipated, reflecting at leaat in part the large portion of the precipitation that issues from Piedmont forested basins in subsurface stormflow, since from such basins overland flow is seldom observed, This is not to say that forests have no regulating effect on discharge from these small basins. Ln Figure 31 the hydrographs from two storms have been superimposed on the same scale to illustrate differences in the reeessfon rate of direct runoff from two watersheds, one of which has a forest cover, the other mainly in a well-sodded pasture; the two streams are not in the Lake Michfe Watershed but geologfcally the so%Ls are of the same parent material as in the basin of Dial Creek, a major trfbutary of Flat River. From observation, it is known chat the stormflow from the former, and larger, basin (108 acres) develops as subsurface flow; in the Patter (X05acres) It develops mainly as overland flow since the pasture soil is very compact, The two events recorded in Figure 31 are a rain of 1-06 inches on the evenfng of May 92, and a rain of 1,49 inches, the night of May 15 and early morning of May 16, Essentfally no difference was observed in the time to reach peak flow, but in the pasture the direct runoff receded to or below the antecedent baseflow level fn a few hours, while the stomflow recessfons from the forested watershed remained noticeably above its antecedent level for several days after the storms ended. A major portfon of the 108 acres fn the forested watershed has been out of cultivated row- crop field use for only 33 years prior to the events recorded here, so a dramatic difference fn the behavlor of the outflow pattern could scarcely be expected, It is postulated that after several rotations of pine, detention storage capacity and deep seepage will 5e improved sufficiently to begin showing a significant increase in the time to reach peak flow, a marked diminution of stage heights, and even a slower rate of recession from peak storm flows, Chapter VI THE SEDIMENTATION ISSUE General. In the initial years of the lake, as stated earlier, the sedimentation rate was of concern to the Depart- ment of Water Resources. In 1926, before the reservoir filled up the Water Resources Division of the N, C, .Department of i ConseruatTon and Development esrablished ten ranges for future use in determining sedimentation rates, Profiles were taken across these ranges and the ends of the ranges permanently marked. In October 1930, profiles sf the upper five ranges were re-established by soundings, and later, :in June 1933, new proffles of the upper three ranges were again obtained, No report of the results af the survey of these profiles could be found other than a reference by Eakin and Brown (1939) that they had been made. In 1935, H. M. Eakin, Head, Sedimentation Scudfes, Division of Research, Soil Conservatlon Service directed a survey of silting of reservoirs, and Lake Miehie was one of the reservoirs chosen for study. Profiles of all but two of the original ranges were obtained and direct probing of sediment depths were made, Direct probing was employed, since sounded profiles from this study, plotted against the original profiles, gave results in considerable disagreement with those obtained in the first survey, The possibility of scouring was ackcaowledged , but the discrepancies were of & nature chat s~ggeerederroneous ass5gn- ment of elevacfons in the xfgznal survey as moxe logically the prime cause, In the 1935 btildg of Lake zir"hit3, additlanai ranges were establ%shed, coraling 32 GVZC the entire fake, Prom soundings and direct sedfnars,~neasaemerrts, an esziaa're was derived of A surrmary of the f ludfngs -f;cl.;, cir.i-- ? 535 scudy are included later in Table 11, ---sbse~~ritfsnb- kn the S&ll of 1 exploratory pro'nings ofthe mud 'tbhsi~lcf the ~~perch-zee -fourths mile screfch of the lake were miiile A 3~3lorxgeellate summer and autumn dreughr had brought "Le l&ke level nine feet behw the spillway, leaving long stretches of mud basin and shoreline exposed to faeilftate direct measurement of the depth of deposfts at several poines (Figure 32 snd Sfgure 331, At the heighe of the drought the widrlx of che pool just above the debsrrchure of Muddy Creek had shrunk to Lws feet de its naraswest point, with a maximum depth of orne foot, Atchis stage of ksw cracer it was possible cs make a measurernenr of sedmmenz deprh aersss the entire mid basin, revealing maxinun depths of some 15 feet, and a reduction in etsrag~eapacfryof approxlmatefy 65 percent at thfs point (Figure 341, Figure 32. View of portion of extensive mud basin about one- half mile below head of pool. The mouth of Muddy Creek is in the upper left of photo. At waters edge, the depth of sediments exceed 18 feet and the material in the deposits is fairly uniform in texture, with less than 10 percent sands, about 65 percent silt, and 25 percent clay. The coarser lighter-colored material in the foreground is erosional debris from a road ditch leading into the lake at the east end of the Bahama bridge. Photo taken November 6, 1968. Figure 33. A stretch of exposed shoreline about one eighth mile below the Bahama bridge during the autumn drought of 1968. Normally, the trunk and branches of the tree are under enough water to be out of sight but not necessarily out of contact with propellers of motor boats. At this point, the fine sediment depth at water's edge is approximately four feet. Photo taken November 6, 1968. The preliminary explorations provfded an opportunity to examine many areas that in "average'hater years are not in open view. The original stream bed from the gauging station one half-mile upstream from the head of the lake to the point just below the dam has very little gradient, the former being at 346,85 feet above mean sea level, the latter at 256.60 feet, indicating a fall of only a little more than 90 feet in a distance of approximately 3 lb.2 miles. Nevertheless, while the pool level was dropping nine feet below the spillway, some fresh scouring and movement of earlier deposfts were occurring (Figure 35). Where the seream rounded a sharp benq in the original channel and a large protruding boulder served to aid in reducing rate of flow, a large bar sf somewhat coarse material, along with occluded leaves, had accumulated to depths of 10 to 12 feet (Figure 361, It varied considerably in texture, having about 17 percent sands fn the surface six inches, but about twice as much material in the sand fractions in a zone 24 to 30 inches below the surface. Roughly, 50 percent of the material was in the category of silt. At the surface, and at two to three feet, clay averaged about 24 percent. At the highest points of the bar, (which is scarcely appropriate. to term a "sandbar" since the surface in spots is comprised of materials up to 70 percent in the sflt texture class) w%EPows had become firmly established but would stand in water when the lake is at the dam crest level or higher. Figure 35. During normal lake level or periods of high water, the head of the pool serves as a stilling basin or box into which suspended sediments are deposited. The Flat River streambed at this point has enough gradient that permits even relatively low flows, as shown in this photo, to transport the deposits farther downstream into the reservoir. The first significant stormflow following a low stage period performs a complete sluicing of the remaining erosional debris. The flow at this point is quite thick with suspended silt. Photo taken October 29, 1968. Figure 36. Downstream from a sharp bend in the original stream channel a silt bar has accumulated on the inside of the curve to a depth of 10 to 12 feet. The surface texture is only 17 percent sands, but enough leafy material and other organic debris is occluded in it to keep the surface from the "crevasse" development observed on the surface of mudbasins. Photo taken October 29, 1968. Between the silt bar and rhe shore is a Lbgeon-like area or slack-water where $a-lc or' tsrbclsnce hhs petcnfeted. finer suspensoid material to settle oue, Sand conyefit in the sedi- ments was found to be as Isw as twa ra thiee percent, bur clay in the surface two to three fee^ ranged between 2% and 30 per- cent, at least plasti~enaugh fs show rnodzrate "crevasse" development on drying (Figure f7) i~ srxcuza, me deposits in this area are essensiakley the same as far the ~najsrpartion of the mud basin. All the ninor screams excrpL one, Lsadlng direcrly into Lake Michie are clcgged allmat bantk2,ir wish erosional debris (Figures 38 and 39) .. %he aile excepr-isn is Rocky Branch which showed only about Lwo to three feet of sludge ar maximum depths across the mouth of the stream. TrmedrateLy above the point of debouehure, the other streams had maximum sediaent depths ranging up to 11 feet, At prevaiPing lake levels, thz excenL of sedhentation that has oceurred over several decades is ncs~readiiy spparent. For example, during the 1968 draught, an extensive sedtment bar came fnto view below the Bahama bridge (Ffgure 403, Probed on October 29, 9968, the sediments were found ts be at Peast 14+ feet in depth; precise depth could nor be dererrnined because footing was not stable enough t3 per mi^ exrr~catingthe probe against the hfgh sucsion force 0% the soupy deposits. On parts Figure 37. A slack water area where "fines" have settled out and accumulated to depths of 8 to 11 feet in the period since the dam was built. Clay content of the surface two to three feet ranges from 25 to 30 percent, and the material is plastic enough to shrink on drying and develop the fissures plainly visible here. Photo taken October 29, 1968. Figure 38. Mud basin at the mouth of Muddy Creek as shown from photo taken from a boat in the narrow channel seen in Figure 41. Direct probing showed maximum sedi- ment depths of 11 feet in spots. Figure 44 presents a profile of the lake and sediments along a transect crossing at this point. This photo taken on October 29, 1968. Figure 39. Exposed mud basin at mouth of Dial Creek, Picture taken from causeway separating this pool and Lake Michie on State Road No. 1616. A profile across this stretch is shown in Figure 51. This photo was taken October 29, 1968, Figure 40. Looking south from the Bahama Bridge toward the recreation headquarters site, when the lake was nine feet below spillway crest: i.e., with sur- face of water at 331 feet above mean sea level. Depth of sediment on this bar is at least 14 feet; thus, only the "tip of the iceberg" is ex- posed. Seen in the far background is another bar extending upstream from the island shown in Figure 42 (vide infra), some three to four hund- red feet; photo taken October 29, 1968. of the mud basin where the srr~facehsd ~~lbtQI:x%~nt~Fne.act dry, the footing was adquare to p~iSthe ?robe pipe even frsx depths any of the original x,>s cf the k&e, 11xis essuasd to have been important enough to indsde in the $925 rapographlc mapping, that they may have sriginated as scunip sprouts or root suckers from a previous stand clearsd in preparation for sventua: flooding by the Sake. Incrsmenr cores extracted fmci sev3ra.P rrees inn 1971 showed rfng counts (years) ranghg Erox abod.e. 55 gears to 42 years at breasr height, Probrng halfxay dcross che upstream en island, within the area oceupf ed 'by frees, revealed '%softB' sedi- ment depths ranging frsm two to rkrze feet in ehickness, yer the base of the bole of eke trees appearea to Gave noraal caper, characteristic for the respectave species, at Peast suggesting that no large vo%me sf erosional debris had been recently dumped there. The exfsting species rnlx ireeludes sweecgum, elm, ask, blackgum, red maple and eotronwsod, wick dfametera ar breast height generally ranging from 10 fnches upward, and at least one tree measuring 29 inches -d53, All these apaciaa are narmakly encountered fn Piedrnont bot.afo~ldnC,a, Figure 41. Sun-baked sediments provide a stable footing over an otherwise quick-sand-like ooze. Mouth of Muddy Creek is in background, Figure 45 presents a profile of the basin along the near water's edge shown in this picture, Figure 42. Island viewed from recreation site boat dock during stage when lake level was at or slightly above spill- way height. The bar extending northward from this island, as seen in Figure 40, is out of sight, In the 1968 probing at a point about 300 feet upstream from the tree-covered portion, sediments ranged from five to nine feet in depth. This island is not re- corded on early maps of the lake basin, but was in evidence when the lake filled in late 1926. Tree age and sedfmene depths do not unequivocally establish that the "island" was missed in the fnitfal tapographic mapping of the prospective lake site. However, early oblique aerial views of the lake provide proof ef ies existence, Suffice it to say now, that in the dynamfes and morphology of sedfmentaeion of reservoirs, the island 9n the iast four decades, or so, has influenced the development of the sediment bar extending 300 feet upstream from it, Seismic probing of Lake Mfchie.--- The initial direct probing~,by means of 3/8-inch segnented pipe, were confined principally to exposed portions sf the mud basfn, and to shallow water areas, available for exploratfon during the 1968 drought. The probings covered a screech of reservoir for approximately a mile starting downstream from the "normaPt' head of the pool, In the autumn of 1930, an attempt was made to map the present position of the lake bottom and sedfnlentation depths over the entire lake, including the parts where direct probtng was too difficult or not feasible. For this purpose, high resolution sonar equipment was employed, Although the planned detail in the survey could not be completed because sf boar failure, the entfre expanse of the lake from above rhe Bahama bridge to the dam was traversed, and areas of maximum sedimentation were transected. The survey lines along whfch soundings were obtained are shown in Figure 43. PrcsfLies based on the sonar resolucioa data are shorn in Figures 44 rhrough 50 and are identified by nuslber on the map Pn Figure 43. The percent loss of reservoir storkge volume far each prcfi2-e is given with the figures, ~ddiefonakdirect probfngs and soundings were made fn 2971 and 1972 of the mafn tribu~ariesof Flat River entering dxrectky into the Bake, as, ranges are included on Figure h3, and the profiles su:marfzing the results are shown in Figures 51 ~hrough57, eonffrming what was suggested by viewfog the several debcuchures during the 1968 drought, Data from the 1935 survey md ,he sonar ssundings and dPrecr probings provide a 3asis for esairnating che meal. sedi- mentatfon since I926 and tm average rate of sedimentation in the 33 112 year period since 1935 (Table kk), a few samples from different Pocatfons on the sedin:ent bars and mu?. basin were collected during the low wacer stage in 196 two depths were sampled, namely, the 0 Co 6-inch zone and the 24 to 30-inch zone, Volume weight determinations an the samples showed them to range in weight from about 55 96. per mainly on the praportion sf sands, bur where occluded organic matter was abundant, rhe strmpEes naturally were lighter, To obtafn an estimate of fhe tons of sedhenes irz the lake, an average weight. of 35 Ib. per cbbic foot was used, Figure 43. Map of Lake Michie and its tributaries at the 340 foot contour showing traversed sonar survey lines and probed lines. Numbered lines correspond to lines for which profiles are presented in Figures 44 through 57. The line identified as A in the location at which the profile shown in Figure 34 was observed. Elevation feet 250' I t Horizontal Scale Figure 44. Profile of Traverse Line 1 looking up stream. The north bank of the lake is to the right and the mouth of Muddy Creek would be on the left. The tic-mark along the water surface line (340 ft. elevation) indicates the intersection with Traverse Line 2, and the hash-marked area represents ~he65.2 percent of the area which is sediment. -2 ST- 4 ara (6C C m aJHSW El 2 ma, *e.tvl U1u5b). Figure 51, Profile of direct probing, of sediment basin at debouchure of Dial Creek, line 8 of Figure 43, looking up stream. Bash-marked area represents the 38.8 percent of the area which is sedfment, I roo ' 1 Horizontal Scale Figure 52. Profile of direct probing of Worner Foard at fts mouth, line 9 of Figure 43, looking up stream. The hash-marked area represents the 25.2 percent of the area which is sediment. Figure 53, Profile ef dfrect probing sf Hsrner Foard, line 10 of Figure 43, looking up stream. Bash-marked area represents the 59.6 percent of the area which fs sediment, , 20 HsrfzontaP Scale / Figure 54. Profile of the direct probing of the Right Fork of Dry Creek, line I1 of Ffgure 43, lookfng up stream. Hash-marked area represents the 73.2 percent of the area which is sediment, Horizontal Scale Figure 550 Profile of the direct probing sf the Right Fork of Dry Creek Sine 12 of Figure 43 looking up stream at its point of confluence with the Left Fork. The hash-marked area represents the 40.8 percent of the area which is sedfment. Figure 57. Profiles of the direct probing sf the Left (top profile) and Right (bottom profile) Forks of Rocky Creek, Lines 14 and 15 of Figure 43, looking up stream, The hash-marked area represents the 32.3 and 28.7 pereenz of the area which is sedirne~t,respectFveLy. Table 11. Reservoir storage capacity loss through sedimentation. Data from 1926 and 1935 taken from Eakin and Brown (1939). April Jan, Oct. 1926 1935 -1970 Capacity (acre feet) 12,671 12,276 11,299 Total sediment deposits (acre feet) 395 1,392 Avg. annual deposits ------(acre feet) 45.14 30.82 Deposits per sq. mi/ ------year (acre feet) -269 .I84 Storage capacity ------loss (percent) 3.12 10.83 On the basis of the limited data, the zota1 sediment mass that has been transported from the watershed into the lake exceeds 2.241 million tons. Considering only the period from 1935 to 1970, the accrual of sediments totaled 977 acre-feet in 37.5 years. The annual average rate therefore is more than 44.6 thousand tons, or 2669 tons per square mile per year for this 167.5 sq. mi. watershed. This rate is considerably below the weighted world average rate of 520 tons per square mile per year as reported by Holeman (1968). In comparison with some river systems within conterminous U. S., or bordering it, the Flat River watershed appears as productive of sediments as the Mississippi River (Table 12), sometimes referred to as the "Muddy Mississippi"'. St, Lawrenee Alabama Colmmbia Delaware Potomac Ohio FTfss%ssip~i Mi sssuri Brazos (Texas) Colorado Eel (Seotia, California) Durham's water supply, However, the major streams of the water- shed still are deeply discolored even after a relatively light rain (See Figure 261, making them aesthetically unattractive, and adding to the sludge in the settling basins. In an attempt to pin-point the major source of the sediments producing the discoloration an extensive survey of the watershed was made. This survey lead to the conclusion that the prime source of sediment pollution issuPng from the watershed is the extensive rural road system, primarily unimproved, However, before undertaking to look at the road system and maintenance thereof in mare detail, it may be fnstructfve to examine the nature and composition sf the sediments which appear to have reduced the storage capacity of Lake Mfchie by between 10 and lP percent since 1926. Character of surface sediments in the reservoir. Indict- ment of the rural and predominantly unimproved (%.e,, not surfaced) road system as possibly -the major contributor of sediments into the reservoir may make examfnation of the sedi- ment character and composition of some interest, As was noted previously, the sampling of sediments was confined to zones within the upper 2 1J2 feet of the lake deposits, hence they are likely to have been brought down within the last two decades, Moreover, there appears to be evidence that they are likely to have originated mainly from road shoulders, cuts, ditches, and other parts of the system and not from fields and pastures, Comparing material from the sediment bar near the head of the lake (cf. Figure 36) with material obtained from the sedi- ment bar near the recreation site headquarters (cf. Figure 40) showed considerable variation in the proportions sf different textures between the two depth zones, but the silt texture class predominates in all samples (Table 131, In chemfcal composition, the relatively high orgaaie matter concenrration particularly in the samples from the bar ar the head of the lake must be attributed to the abundance of tree leaves, ground up and whole, that were occluded in ehe silt. En relation to native upland soils calcium appears to be high, but perhaps in the past there has been some appl5catkon of CaC12 to road surfaces to settle the dust, Phosphorus appears to be low, suggesting that fields and pastures may have contributed relatively lfttle material which could settle out with the suspensoid material, Although nitrate nitrogen amounts may not be a good and true estimate even from this limited sample, since handling and treatment of the samples prior to analysfs could have permitted some incubation EO occur, the nitrogen quantities reflected by these samples results from the greaeer quantity of organic matter from hardwood leaves in the materfal obtained from the bar at the head of the lake. Table 13, Textural analysis and chemical composition of samples of sediment bars at two locations and from two dgpths in Lake Michie, Samples collected in October and November, 1968, Sediment Bar Upstream Texture-class, Sediment Bar ac Head of from Island near $oat item or element Lake Michie (Fim -Sampled Zone Sampled Zone Sands - percent 17, 34, 12, 62. Silt - percent 58. 42, 66, 20, Clay - percent 25, 24. 22, 18. 2 / Chemical Analysis- pd' 4,1 4,3 Organic matter - % 2.7 3.2 Calcium - Ib./acre 960. 976, Magnesium - lb./acre 98. 72. Phosphorus - lb./acre lo, 90, Potassium - lb./acre 40. 32, Manganese - lb,/acre 74. 34, Nitrogen - lb. No3/acre 224. 112, Nitrogen - lb. NH4/acre 88. 74. Soluble salts - ppm 1440, 1000, 1/ - Texture and pH determined in laboratories of the School of Forest Resources. -21 All other determinations were inade by the Soil Testing Division Laboratories of the North Caroiina Department of Agriculture. It is also of some interest to compare samples from the fine material which has settled out to build the mud basin of the reservoir, known to exceed 15 to 18 feet in depth in some spots (cf. Figure 44 through Figure 491, with samples from the native woods above rhe high water line of the reservoir, i.e., above the 358-foot contour interval. A major difference fn texture appears in the low proportion of sands in the mud basin, as contrasted to the samples from the woods (Table 14). The silt fraction, obviousPy predominates 3.n the mud basfn per se, but it is high also in the native upland woods samples which were adjudged to be Georgevflle silt loam. Clay would not be expected to sertle out in quantity except as absorbed material; hence, it is not surprising that its concentration is not much different from rhe upland samples. In chemical composition, calcium again is indfcated as being twice as high as in the surface six inches of native upland woods soil. Phosphorus is not more abundant in the sediments than in the upland soils, but potassium, as expected, is less abundant in the sediments than in the surface of the woods sofl which is annually enriched by the litterfall, in this instance primarily by hardwoods. The relatively high quantity of manganese fn the woods soil samples as contrasted to those from the mud basin is sf interest since the Department of Water Resources in Durham experienced considerable difficulty withmanganese Table 14, Textural analysis and chemical composition of samples from two depth zones of mud basin in Lake Mfchie and similarly of adjoining woodland above high water line. Samples collected October and November, 1968, Predominantly mixed Texture-class, Mud basin above Bahama hauhm~d'~shore boveOn Mu west dy item or element Bridge (Figs 6 36 & 10) Creek $~i~.31f Sampled Zone Textural AnalysA-11 Sands Silt Clay -Chemical Analysfs-2 /" Pg 5.1 Organic matter - % 11.7 Calcium - lb./acre 912. Magnesium - lb./acre 139.20 p osphorus - lb./acre 10. Potassium,- 13. /acre 69. Manganese - lb,/acre 25, Nitrogen - 13, No3/acre 32, Nitrogen - lb. NH4/acre 24. Soluble salts - ppm 360, -l/ Texture and pH determined in laboratories of the School. of Forest Resources. -2 1 A11 other determinations were made by the Soil Testing Division Laboratories of the North Carolina Department of Agriculture, accumulations in the water system pipes until a suitable treat- ment was found, These data suggest that ehe manganese issuing from the drainage basin fn solution perhaps continues to move with the water rather than being precipitated fnto the bottom ooze, These daea provide a rough estimate of rhe character of the surface layers of sediments in Lake Mfchie; but they do not imply applicabifiry to the total sediment profile of the entire pool.& Probing clearly indicated presence of varying textural strata from predominantly sands to essentially fine silt with clay. Sampling to characterize the entire profile proved to be beyond the scope of this study. A look at the road system. The prevailing maintenance practice on the road system is to grade and open u~ the drainage ditches of the roads directly into main or feeder streams of the watershed. Figure 58 presents examples of chis practice. The adverse effect sf direct drainage fs intensified by the fact that there has been little attempt, or at besc little success, at stabilization of road shoulders and embankments. Figures 59 and 60 are prime examples of the combined influence of maintenance practices and unstabilized roadsides, A survey of the watershed indicated that there are 36 locations where roads intersect with the perennial streams. Better than 60 percent of these fntersectisns involve unpaved roads; even on the paved roads there are numerous crossings Road ditch mainten- ance allows direct drainage into stream tributaries of sedi- ment-laden runoff from the unimproved road system. After maintenance the crown and shoulders of m .any rural roads remain raw and loose for weeks or months, serving effectively as coloring media for the watershed streams. Figure 60. Ditch drainage from newly maintained road shown in Figure 60. Scouring of road ditches by stormflow generated by every substantial shower adds to the sediments moved into and by the streams. where shoulders and embankments have not been stabilized. Although the entire scream system was not traversed, aerial photographs indicate that direct drainage from cultivated fields into the stream system rarely occurs; instances as are illuscrated in Figure are not of sufficient moment to be considered the only source of soil for coloring of the streams. Direct estimates and observations of road runoff and sediment transport, +%ontributions to pollution of streams and sedimentation of reservoirs by road systems are large, as emphasized in the publication of N. C. Association of Soil and Water Conservation Districts (Anon,, 1971) and it became clear in the early stages of this project that the road system fn the Lake Michfe watershed is no exception. However, a critical study of this problem would have required additional resources, Nevertheless, the problem seemed of sufficient importance to justify some exploration within the scope of this project. A storm in October, 1971, provided an opportunity to collect samples of water flowfng from road ditches directly into a stream. The samples were collected in the cascade from the ditch to the stream so as to include only suspensoid material and to eliminate bed load from the sample. In all, 16 sites were visited and quart containers were used to catch each sample. These samples were then analyzed to determine the weight of sediments in each, The range in oven-dry material over the 16 samples was from .0015 Ibs/gal to .0325 lbsBga9 wich an average of .0084 lbs/ga1, The ditch bottom quality at the points of entry eo rhe stream channels at the IS sites varied from well grassed to bare soil. The time of year could nut be eonsidered optimum for sediment yields from road systems, Dfseker (et al. 1963) reports that maximum sediment yields occur in May and June when high intensity thunder storms flush the accuumulaced sedsments resulting from winter frost heaving* Hence, the estimate of average sediment yield from these samples Ps quite conservative, However, at one site the flow rate was timed, and later che cross section of the dftch was determined, Fkese yielded an eselmate of 25 gallons of water per minute entering the stream, Applying the average of .0084 lbs. per gallon -co the observed flow yields an estimate of .21 lbs, sf sediment per minute, or 12.6 lbs, p er hour. One could extrapolate these figures to tons per year by applying a variety of assmptfons such as the number sf days or hours of rain per year of sufficient amount or intensity to produce surface runoff, and by assumfng the above data represent somewhere near an average of sediment transport by road systems, However, it is not reasonable to assume the above figures to be average; fn face, it has already been 'stated that they are very likely below average for rhe geology and soil variation in che Plar River basin- They are presented merely as a minor example of sedfment pollution that derives from road systems, Reduction of poBEution from...- roads, The bess way to reduce 0r elimfna'ce ~~adsa6a SO~FC~of pollution is to stabilize the beds, shoulders and em?rp~n"Kmenta, However, a less costly alternasive exists, By developfng diversfon ditches at intervals, leading water away from road ditches es spzead it across bands or strips sf vegszatfon before it enrers the srream system, makes it possible es filter our most of ehe particulate matter, A searching inquiry of the most effective vegeration cover is needed, buc observation %n this study indicates that for a moderate slope a one-chain (66 fee^) wide strip performs a filtering function effectively so long as the strxp has a well-established vegetarive cover, Observation indicates that can fmmensely hfnder forest regeneratfon and other sflvi- cultural practices, During the investigation of rhe Lake Miehie watershed, six sites were studfed where filter strips were in existence, some by design and some by acefdenc, The results from one of these sites fs reported here as an example of what can be aceompffshed, This site is located ar the hterseetion of This filter strzp, altbsugh behg rhe most effective of the six observed, probably had recesved the greatest sediment load inasmuch ds che eoad c.unatsuct$sn had been so recent, The poPne is timat It dfd slow dsm the ewauer was thriving on rhe deposits and ;m good shape to again retard the surge of debris fhom ~kenexr scam, In summary, any study ccf the rssd systejn im chis Pfedmsne waeersked leads one to the c0nc4~sfanchar construction and maintenance or" r~a4zoadc ioncrabar:~hta-*ily rowam scream pollution and reservoir sedLrraentacnon, It 1s a193 certain that much of the po8X~taunLs G.VLIS~L.LE, ana 53ne of the measures to reduce degrdation of water qwaisey really do not requfre greater inpats, neseEy Ease senslbZe applicationsa Road Nom 1613 may be used ds aa m.ax~Ee; iL sknrbs the east stde of Lake Michie closely for absur a fuvarh of a mile northward from zhe Kahama bridge, About an eighth mile from rhe bridge, a newly graded road cakes off steeply r-o the right, and has a -serp aecxveiy erodmg dirA which drains through a culvert SeLsw Road No. Poi3 mto a narurai drain leading directly into the Lake (Figure GI), Ns effort apparently was made ro divers the rurrent oL sediment-laden runoff from the newly graded road over a strip of forest floor, Rather it is permizted rs foiPaw ehe natural spring channel to the reservoir. Some or" che sediment dues drop out as the gradienr levels off, but a sizeable alluvial fan is forming on the fake shore,, The above example is only one of rnhny rhat mighe be cited as illustrative of sitxa~imsghi~h er~uld be greatly Figure 61. Looking westward from Road No. 1613 along a natural drain which is transporting sediment from a newly- graded road from which stormflow is led underneath Road 1613 -via a culvert. Diversion to spread the water over an expanse of forest floor would have been possible here with little or no extra cost. improved with lirtle or no added eosr, Other situations can be improved only at some cost, as illustrated by a nomtnal treatment of a forest access road (Hgure 62). Many current practfcee %n highway or bridge maintenance defy rational explanatfan, A good exanple can be seen from the repeated events or happenings at the Bow-water bridge construered in 1969 is a replacement for the old brfdqe on Road No, 1614 crossing Plat River at tshe Slvcunr Forestry Camp, The hfgh stormflow that struck the bridge fn July E978 (See Figure 26) washed the sofl from both ends of the bridge rendering the road ifnpassabfe. The road nnafntanance crews quickly repaired the damage by gouging out about 10 cubic yards of fill from the nearby road cut. In 1971, klgh stsrmflow cleaned out the "soif" rna"ce5al.s from bock ends of this brldge again, and again mafntenance crews repalred che damage, employing the ssme procedure and source of fill as in July 1970. The rains associated with Wurrieane Agnes on June 20-23, 1932, again have inflicted the same damage on this Law-water bridge, which presumably wfll be "repairad'"n the same manner as before. One facet of this repeated damage is that the rip- rapping appears inadequate to withstand the force of storm- flows generated by Plar River. But what seems more important from a water quaPfty standpoint is the repeated sEuieing of the fill used in the periodle repair operations, into the river, and eventually Into the reservofr, It would Figure 62. A forest access road leading up a grade of about 10% has been crowned with a moderate application of crushed rock. The high, well-rounded crown and the naturally-vegetated shoulders and ditches assure that the runoff which develops does not transport sediments into the woodland stream shown in Figure 63. Figure 63. This perpetually clear woodland stream crosses the road shown in Figure 62; it has been observed to remain clear during storms that result in runoff which drains into it essentially free of sediments. appear that those in charge of road maintenance have no compassion for the Durham folks who unfortunately must pay for the folly of such road mafntenance practices. Shkptler Vi6 FUTURE OUTLOOK FdR THE LASE MICHSE WATERSHED General. 'T%rsttghl_nuc this report ft Bas been imgalS.ed or directly asserted thac water ,rzd~crioa 2rom the upper Flat River drainage Lasin is ample Enuuang$gx for ~urharn's needs for decades r*.. c::iiiie' Thq.:; yi-ogrmsis is fir% despite nomfna: changes that :cay o,.cus 5-11 [.he (~$2of ~heland wi"Lifl the watershed. Some cfisxlges a;s a!.ow%y k,zking place now, some trends are evident, ard- ssorne asicaf~~oaal.cilte;atl"ons will, surely take place that will influence rke hydrsiogxc characteristics of the drainage basln, Henze, smnr expected ihanges and trends n~ightbe worth considering at cb?a psine, Trends in ruzal----- land tse, The screage devoted eo aEP farmlrrg operations 1s ez~zcterito remain. ZafrPy skabje for several decades, GerfaanZy frL the pas2 30 years these has been no marked change in net .fa.mSng acreage, There could be sizeable shifts in apeas ii3ed for difierent crops and the proportion devoted to cultivated ~rspsas contrasted to pasture and meadow may change, although any trend In the past 28 years is scarcely discernible. For the rhe bemg, there appears tu be no reason eo expetr the upenzng up of any sizeable areas of "new ground"'; such acci-vf ty as may take place will primarily Lnvolve esrisolldeeicrn of existing patches of fklds, pastures and neado+js, This prognosis should not be eonstrusa as elinharing rhc possfbiPfsy of future expansfon af agricultura; operat?-nns on the warershed. Martin and Bass (1940) j~dgedcnaL less rhan nine percent of the land in the Lake Kiehle dramage bash was too sEeep for row crops. Thus, except fcc these steep areas, and lands $1unsuitable becease of ucirhvorr;bii.- soils or advanced erosion", it is assumed that khe btlk of tcd warerhhed is vulnerable to some future wave of tiilagz. In recent times, the tread has tzen towad larger individual holdings and fewer small fdrms, in the rnafn this trend has been accompanied by greater ~fficieaeyEn farming operations and in the employment of marzagement practices that promote more effective soil and wzzer conservarion, There is no reason to expect any reversal 112 this trend& though ultimately an equilibrium point or a cefling in rhis develop- ment may be reached for ehe terrain and Land resources, Yet, society now is in a state ~f fluxS and "out-migratlornQs from the "inner city" is proceeding rapidly, Frm Durhax northward the rate of migration is well-doeurnenred by the new csnstructfon that springs up along all thoroughfares, gradually but inexorably moving toward the warer divide of Lake Michfe, For the time being, this activity is conffned to zones dong all-weather roads, making it easy to seovs back zo rhe etty where the gainful employment is rnafntained, nation, and the p~pt~la~raiyof $.oxbo~c -hews % sharp upiA7ard trend snm& 1950, The dwfndlsrig fa~mpsydatisa PL~Person County does not necessariiy Zean lesb Earmiag actrvity on the watershed. As implied abdve, xr nrap s~mplymeen a change sf farm size and possibly a -5ange in the nature and qaality of famicg operations, Trends in foress~~Hn the last 34 years, rhe area under some form. of fores~cover has 1n~:ea~edby only about four pereenr, reaching rhe current high of &lightly nsre than 62 percent oE rhe entire warer.-heci, There appears no reason ae ?resent tc expect a Larger psroFoxelon of the lands in rhe drainage basm ea reveit to naraqt begetacaon xnclvding forest, so far as net acreage f..r :,mcd~nc3ds unless some o County of Durham x City of Durham a County Exclusive of Cfty 1940 I966 YEAR Ffgure 65. Twentieth century population growth trends for the city and county of Durham. Q Person County x City of Roxboro r*" g County Excluse~eof Roxboro /# /-- /' YEAR Figure 66. Twentieth century population growth trends for Person County and the City of Roxboro. accommodation is reached between land owners and downstream communities concerning the safeguarding of the quaffty of the water produced, To Jmprove waeer quality, an increase in forested acreage and rhe inst~rueionsf sensible forest practices would be desirable. POP the time being, and for several decades beyond the current one, it is expected that about three-fffths of the land will remain under some form of forest cover or natfve vegetatfon, This prognosis does not rake into aeeovnt the praspect of burgeoning rurban growth or rnovemsnt, nor some major revolution in che agrieulrural sector of rhe region, including reactions rs neo-Malrhusian scares, Some forest land is befng taken out of pstentfak wood production, for all practical purposes forever, through urban sprawl whfch appears to preempt field, pasture and forest use quite indiscriminately. For Durham, the sprawl has not yet come very close to entering the watershed of Lake Mfchie, but if it is intended to maintain this basin as the principal source of water for the municipalfty, crossing of the divide should be discouraged while there is still time for action. With the exception sf the HfYl Forest referred to in Chapter V, forestry practices within the watershed have no where else to move but up, It has been stated earlier that forestry practices and the qualfty of forest management on the watershed lands are minimal. A large acreage at present is seeupfed by green junk whieh may look cornforring to a water manager fro2 the air, but if is junk that 5s using water withsu~developing very much value efrher fn needed forest products or in aesthetics, En the current popular concern for the envffoment and open space, there can coneei.vably develop strong moves to set aside portforis of the watershed Eores~lands for enhance- ment of the kinds of amenity values essociated with forests. Such moves might tend to rnafntafn land in forest cover which might be dfverted otherwise ro differenz non-forest usss, but they would nor enhance forest production, In our opin5on, renovation of rhe existing acreage of green junk into a hochwald condition would be very desirable before any setging aside of land for mere open space takes place, By that time, the public might become cognizant= of the possibf9.ity of enjogring amenity values in cqnsort with a form of foresc production which is compatible with water produetfsn. In the existing forest acreage, aesthetically superior or superlative spots are rare and hard to find, The general trends in 20th century population growch in rhe eovntfes within which the Lake Michie watelshed lies have been alluded precedfng to In paragraphs (cf. Figures 64 and 45), but some derails concernfng the significance of encroachment may deserve addftfvnal comment. Historically watersheds have always been areas of restricted human use and habitation, and Pfmitations have also been placed on the kinds and fntensitfes of use sf reservoir waters. Tn the case of Lake Michie, swimmfng is prohibited and boating must cease at sundown, but beyond these verboten activities, control stops essentfally at the shore. In this study, no attempt was made to identify the orfgin of new residential construction wfrhin the watershed; some of the observed activity doubtless represented replacement of previous dilapidated and abandoned farm homes, but some of it also may have been new migration, a vanguard element in the development of a rurban culture. But even though there may be no major movement of people to inhabit the watershed lands, rurban and suburban developments, where rhe latter are outside the boundaries of municipal services, especially garbage and solid waste collection, will have, and are having, an impact on the watershed. Some discrfmfnatfon, to be sure, in solid waste disposal is currently practfced; much more of the dumping appears to take place near stream channels crossing roads than is observable elsewhere. Woodland roads and trails, of course, are also used by the more hardy and venturesome haulers sf garbage and solid waste, but all-weather roadsides seem to be favored as disposal sires, and same very fmpressive collections are found along roads at stream crossings. The estimated 296 niles of perennial stream channels, equivalent cs P,7& miles of o en, flowing stream per square mile of land increases the potential for contamination of the system from sewage, especxaPPy ff developments occur, as they often may and do, near tributaries of Flat River. Surveillance of sewage disposal facilities in past residential construccisn and in any future developments should become an item of genuine eoneern to the people of Durham. Ffnally, in the context of demographic trends, rurban movement, and suburban growth beyond mmnicipally supplied services, there has ts be a more fvntelkPPgent and sensfble procedure for disposai sf solid wasres. The municipaMties and couneres need to join hands to find effective solutions to this central problem involving both the health of the citfzenry and aesthetics of the rural countryside. It is almost beyond com~rehensisnhsw the present practice is tolerated in frone ef the eyes of the traveling public, It must be that too few people are aware that water runs down- hill. Management alternat%ves. Durham and any other muniei- palJty dependent on the outflow from land predominantly under private control are trapped in an unenviable position. Water in streamflow may be a free good, but it enjoys no freedom from pollution and contamination generated by the users of the land from which the outflow issues. In a legal sense, land, regardless of fts distinguishing atcrfbutes, is the sole property of the owner, and what the owner and user of rhe land does, or fails to do, is outside the rules and desires of the water consumer downstream. In brief, it is illegal to empty one's "honey-buckets" into another man's well, but not to spread the excrement where it may wash into a city's water supply, The more intensive the polluting and contaminating actfvity on the land happens to be, the more the treatment cost required to make the outflow potable. The downstream consumer pays for that treatment, presumably and traditionally "through the nose". Under these circumstances, what are the alternatives for assuring sustained production of reasonably high quality water from this watershed, or any similar watershed, in the future? Without legislation, it would appear that the alternatives are not many, and may be of questionable feasibflfty. One alternative would be outright ownership of the 107,200-acre watershed, assuming ownership means complete control, * as well as exclusion of use except by permission of the water authority, This alternative would be costly today, particularly in comparison to the days when the transaction could have been accomplished with a modest supply of beads and liquid refresh- ments. Complete ownership, or at least complete control, is a feature of several municipal watersheds throughout the country, and the arrangement and operation of them have been eminently satisfactory and successful. Conversion to complete ownership by the Cify of Durham beginning npw could be expected to require several decades. Another alternative is acquisition of a limited acreage of land to serve as filter strips along the channel margins of the 296 miles of perennial streams that lace the watershed. A one-chain-wide stri along the banks of each stream would be sufficient to function as a filtering strip, if well vegetated, preferably with trees. Grazing by lfvestock would have to be excluded to maintain the surface conditions in the strips in a shape to function effectively as filters, and watering of livestock accomplished by piping to areas outside the strfps as recommended earlier, In many miles of existing stream channel, an adequate natural vegetative cover already exists over the required strip width, but considerable planting would be necessary. Total land area needed for this purpose would not exceed 5,000 acres. Limited use of the strips, including forest production on the more accessible stretches would be feasible without reducing the effectiveness of the filtering function and at the same time would help defray the cost of maintenance and protection of the strips. The assumption here is tha~the majority of the land owners would agree to negotiate preferably on a voluntary basis. Another alternative is legislation which would authorize the Water Resources Department co exercise the necessary control to insure mafntenance of water quality both in surface runoff and in stormflow and groundwater outflow from the basin, In practice, this alternative would require also the maintenance, protection and control of filter strips as outlined in the preceding paragraph; included would be the barring of livestock access to the stream channels, drainage of fields across the strips to the streams, dumpfng of refuse on the strips or in the streams and other currently existing practices which degrade the quality of outflow, Obviously, to exercfse this degree of control would require a patrol force on a year-round basis, with power and authorfty to deal with violators. Still another possible alternative is the securing of cooperation from owners and users of the watershed lands. With strenuous efforts at appeals and education and possible institution of tax incentives, voluntary compliance with suggested wishes of the water users might be achfeved. As long as free access exists to the pubPfc at large, any improvement even through voluntary compliance on the part of resident users of the watershed would not be sufficient without securing compliance from transent users of the water- shed lands. For adequate safeguarding of water quality in the future, a combination of acquisition and legislation, appears to be the feasible alternative for this watershed. All of the above alternatives are based on the assumption that improved praccices in road maintenance and drainage would be instituted. Without attention to the road system any improvement achieved through the suggested alternatives would be at least partly nullfffed. Chapfer VIE1 EXPERIENCES IN THE MNAGEmNT OF MUNICIPAL WATERSHEDS General. In the study of the Lake Michie watershed, we have been dealing with land which fur all practical purposes is completely under private control, More specfffcaily, the land is not merely in prfvate hands bue under use, treatment, and management of some 750 to 800 fndividual farm ownerships, averaging four to five fams per square mfle of drainage basin. Thfs diversity of management phflosophies and ownership objectives makes this watershed nor necessarily unique In the Pfedmont provhee, bur unique in comparison to municipal. watersheds elsewhere in the United States, POP decades, many watersheds have been successfully operated unburdened by the diversity of ownership and use which exists wfthin the Lake Miehie water- shed. It may be useful to review briefly, in the csneext of history and land use, experiences encountered in the manage- ment of other watersheds. The Elf W$icney Forest, Thfs watershed is under the sole ownership of the New Haven Water Company which began operating in 1849, initially with not much more land than was required for reservoir sites, Gradually acquisition buflt up the ownership around the reservofrs until by I930 a dozen lakes were all within the company Lands comprising over 20,000 acres in ten separate tracts ranging in size from 245 acres to 10,500 acres (Hawley and Ma~ghan, E930). The ELf P&~xtney Forest suppkfes the water for New Haven, Conneericut, and surrounding eommuniries, the I.arads being wfthirn a 19-qfke radius sf New Haven, Forest rfianegemen t act fvi~ies s;z the company hsPdings began on a smell scde ic 1900, and were exyanded jn 1904 ts the entire Poresrry, Each divfston is placed un2er a foreman with a permanene zrew adeqmte EL^ pcope~Pycaring for the reservoirs and the watershed, Although their major work is chargeable to the water supply, they assist in fire protectl.on, and various other forestry opegca~ioons fncluding the planting of pines in the sprfng season, h slack periods the men also conduct timber stand fmprovenent, suc5 as release, prunfng, and related work. Major products harvested from the tracts are Iurnber, pflfngs, railroad ties, pulpwood and fuel, hb1fc access ro the lands is restricted, by fencing where necessary. The Amherst and Mount Bolyoke Kgssachusetts Watersheds. These two watersheds are examples where municfpaf ownership and control extends over slightly more than 30 percent of the entire Band within the water divides, Major sbjectfves of ownership and management are the producrfon of good water and good wood, Wildlife productasn is incorporased with management operations, an2 hunrlng fs permitted, Limited recrea~ionis also permftted on rhe hhersz watershed, hherst owns 1,854 acres of the 6,000-acre watershed which produces water for 11,000 townspeople and 12,000 college students. HoEyoke owns 4,951 acres of the 15,000-acre watershed which furnfshes water for a city of 53,000 inhabitants (MaeConnelP, l965), Annual logging operations are carried out on che Arnherst watershed, and even though they sgarted wfth understocked stands and previously mutilated forests, the stumpage returns now are more than double the annual expenditures on forestry operations, including inventories, silviculture, planting, and related work. All open areas on the Amherst watershed have been planted but on the Holyoke some areas are yet to be planted. Water improve- ment work has consisted of planting pPnes on rhe shores of the reservoirs and releasing existing na~uralregeneration of conifers from competing brush and low-grade hardwoods, Over 50,000 trees have been planted on shorelines, and both mechanical and chemical control of hardwoods have been used to assist che development of conifer species, principally pines. These two watersheds represent examples of rather intensively managed properties, primarily for the production of high quality water, but where the town fathers have not cringed at the thought of cutting trees, regenerating new stands on cutover areas, and renovating beat-up areas of green junk which uses essentfally the same quantities of water as productive forests from which annual or periodic revenue is derlved. The Waynesvflle North Carolina Municipal Watershed. The town sf Waynesvflle owns 8,244 acres of mountain land which supplies water for izs inhabitants and to neighboring communities of Hazelwood and Lake Junaluska. The property is all in Haywood County and the town oms all of the headwaters of Allen Creek, except for 300 acres of interior holdings (Anon, 1953). Purchase of the property was completed in 1915, and for the next 30 years it was given protection but all cimber cutting was prohfbited, In I945 the town officials decided that the prohibiting of timber harvest was not sensible, so they arranged with the North Carolina Division of Forestry and the TVA Division of Forestry Relations to study the feasibility of managing the drainage basin lands for both water and timber. The feasibility study indicated that conduct of forestry operations could be done without impafrment sf water quality, and the first timber sale was arranged for in 1948. The initial progress was very rapid, and within five years, 3.69 million board feet of timber had been harvested, nettfng the town more than $l00,000 in stumpage receipts. The beneffts of this first phase extended far beyond the value represented by stumpage receipts. An estimated $500,000 was earned by sawmill and woods workers; and an estimated six mil.lion board feet of the lumber was processed by local industries into furniture and other items or used in local construction, with a total end value of approximately $2,000,600. This is a good illustration of compatible dual use of watershed lands, despite che hazards sf steep mountain terrain. Erosion on the WaynesviEle watershed has been minfmized by careful construction of logging roads and careful placement of skid trails, and concinual survefllance of road condition and drainage, Water is diverred so that it runs through natural leaf Eftter and forest floor, reachfng the streams wfth little or no sediment load. A11 open areas have been plansed to get unproductive areas back into forest as quickly as possible. Seattle's Cedar River Watershed, Seattle is noted for good water, and provides one of the truly outstanding examples of dual management of land to supply both water and timber from its 143-square-mile Cedar River watershed, Cedar River has served as a source of water for Seattle since 1901. The water- shed lies on the west slope of the Cascade mountains, wfth elevations ranging from 500 to 5,500 feet above sea level, By mid 1950's the City had acquired 43 percent of the 91,500 acres in the watershed, and eventually will have sole ownership of the entire area. At that time Seattle's Water Department was serving a populatfon of 650,000;the projected capacity of the system is estimated as adequate for the needs of a population of at Peast one million. With the exception of an 8,000-acre tract in rhe vicfnity of Cedar Lake, the city has pursued a polfcy of acqufrfng land only, with timber and logging rights vested in the former land owner, but with the City stipulating restrietfve clauses in the contraess as deemed necessary to safeguard purity and qualfty of the warer supply- Since 1924, Seattle has had a definite policy of forest management fn effect, with a permanent forestry staff employed as a par& of the Warer Department force, The City furnishes Its own faresr fire protection, operating lookouts and suppression crews, and mainmm~ngits own fire-fighting equipment, It has eonsrructed a network of access roads and permanent fire trails. The Water Department has even operated a forest tree nursery to supply ehe seedl~ngsneeded espeefaPEy ro reforest severe early burns which had deforested critical slopes within the basin before the Cfty acquired the land. In 1954 modern headquarters buffdings were constructed at Cedar Falls to house all the forest managemen% and relaeed activit~es on the watershed, In the mid-1350's the Cedar River Logging Agreemenr was negotiated to coordinate the tim3er harvesting operations of the three private ownerships represented on the watershed, this agreement greatly sfmpfffysng the responslbifitfes of sanfeation, fire proteetion and watershed management for the City. Annual eut was limited to 35 mi~lfsnboard feet which is within the sustafned yield capacity of the growfng stock on the watershed, and harvesting is by eleareuttfng in strips or blocks inter- spersed by blocks of timber, The entire operarion and manage- ment fs a truly impressive demonstration of dual use fn which the quality of water IS safeguarded while a very substantial revenue is obtained from the growing forest resource, Other examples, Many other €xamph?S of sWce~sfulwater- shed management could be cited, as, for instance, Bremeston, Washington, illustrating what a small cicy tan do with foresight and planning, The city oms most of the watershed (several thousand acres) and has been curfing a million board feet a year for the past 15 years; ultfmately ie expecrs to harvest five to six million board feee annually as the growtng stock within the basin is developed ro rke full capaciry of the land. As a ffnal example, the water supply of the City of Denver might be cited, Denver's vater aeeumalates on the west slope of the Rockies, and is broughr to che city by trans- mountain diversion through the Moffat Tunnel, Much of the flow csPPected on the west side comes Largely from National Forest lands, This amangement for municipal water represents an example of a city and a public agency cooperaring in the development of a dependable water supply utilfzfng the exfsting forest and range protecefon and management facilfties of the latter to safeguard the source, Chapter IX RECOmENDATIONS In earlier chapters the prospects for fmprovfng waeer yields and camfng have been fndicated, but it has also been stressed thae the most urgenz need on the Lake Mfehie Watershed is to institute measures that will improve the ej~i%TP~yof water, A coroLPary so xmprovement measures is the safeguardfng of the water source from the ever-present threat of concamination, In large measure, eontarnination and pollution are problems com- pounded by rhe mufi~pleownership pattern, but at should be qufckly acknowledged that currently the most pronounced pollution is being generated by actfvirfes of a public agency. To improve water production and quality from ehe Plat River drainage basin we suggest the following steps and measures, without specifae emphasfs on the order of priority: 8, -Mafnsenaace of the road system, Diversion of road runoff into f%leer ser%ps, stabilizatfon of cuts and fills, reduerion in the schedule of periodic ditch pulPEng, and eessarion of the praetice of leading road ditches directly into streams, are amng the ftzns needing immediate attentfon. 2, A method of retfllzing streamflow for waterrng of livestock without direct access to streams has been suggested. A11 hog wallows should be elfmfnated from streams, and location sf hog pens should be designated so that a sufficient fflter scrip separates the pens from perennial streams, . 3. Effective control of solid wasre dfqosal, There is existing legislation agafnst dumping, but it is totally ineffective, Farmers probably have enough space on their holdings where they em discard their own junk, but suburbanites and others resfdfng on small patches of ground and either choosing or forced to live outside a garbage and solid waste compact, are probably now the largese contributors to pollution and desecration of the countryside, The municipality of Durham and the county commPssioners of Person, Durham and Orange counties, and possibly the governing body of Roxbors, the southern edge of which is within the Lake Michie water divide, ought to fake a serious look ae this problem and decide once and for all whether the existing conditfons on the watershed are within ehe tolerance IfmTts of health, environment, and minimum levels sf the quality of life, A prograin to promote public awareness of this problem should be aggressively conducted and designated dumping areas provided 4, Imp~o clearing land. to prepare "new groundqqir is recommended that Bnfiially &he stumps be ieft In the ground and permitred ro deteriorate rather khan blading them out of the ground and "mueklng up" subsoil in the process, By shearing srumps near grsundffne, use of disks, drills, mowers, and other machinery will be feasible while the Land Ss held fn meadow or grain durnng he waf~for stumps to deeeaiorate ensugh to permit phwing, intensive eulizlvaf~on,and relaeed farmfng praetacea, This practice fs more satisfactory in elearing land under pine stands, but it can be apphed co hardwood seands, although fairially it may require appPlcatisn sf herb~cfdes,xf stmp sprouts and root suckers are considered objeceionable, A gesd example or the success of this mode of clearing of hardwoods can be seen in the pine done bank area sf the Sehenck Memorial Forest fn Wake County operated by the School of Forest Resources, N. 6, State out rhe watershed, Ic 1s estfmaeed that in 1926, some 3c3 rnrles of che 15-mile shoreline of ehe lake was In eulbfvared fields; by Pace slxties chis use had shrunk to about two-thirds of a mile of land which did not have an adequate filter strip to stop sedfment transport directly ineo the lake, While this is a desirable state, the Water Resources Department has no protection agafnst reversion of the trend. An aggressive program of securing control, either through purchase or lease arrangements, of the protective strip along the stream as well as the lake should be started at the earliest possfble date. Such a program would insure adequate or acceptable quality of water coming into the reservoir. Concomitant with acquisition or control, a patrol and inspection activity must be organized to assure that the effectiveness of the protective function sf the strip is maintained. Filter strips are not a panacea for stream protection in every instance, as pointed out by Streeby (1971). In the Piedmont watersheds, well-managed and protected filter strips appear to be very effective. Such strips trap suspended sediments carried by overland flow which develops especially from intensively-used pastures as indicated by the longtime watershed research in western North Carolina (Anon., 1970). Moreover, tall vegetatfon,especially trees, are very effective in reduchg stream temperatures which has a favorable influence en he fish population and the water quality as demonstraced by Levno and Rothacher (l967), Meehan (19911, Brown (1971) and many others. Thus, fileer strips can serve more than one purpose, but they are effective only if protected against grazing, wildfires, and ocher inffuznees that compact the soil and destroy the forest floor, 6. Regeneration of pine along lake shores. Replace Virginia pine, hardwoods and existing brush-covered stretches with lobPolly pine or shortleaf pine by plantfng. Both of these pines produce a more copious and durable forest floor ehan does Virginia pine, and the floor can betrer withstand the traffic along the shores, POP shorelines, pines have the advantage that needles do not drift as do the hardwood leaves which float into arhe lake and probably add nothing of value to the quality of the water. Moreover, lobPolly and shortleaf pine withstand prescribed burning, as Virginia plne does not, and periodfcally it may prove desirable to use fire to keep the under- story brush thicket along lake margins under control. The conversion could be done gradually by short stretches of shoreline to reduce the vfsual impact of change to a minimum. 9. Encouragement of sensfble forestry practices on the part of land owners en the watershed. Abundant experience has dernsnszrated that the areas sf junk hardwoods can be zransformed with superlative results (Figenre 663. The emphasis on eonversfon of low-grade hardwood areas to pfne is based on rhe fact that on most of the soil types which have suffered moderate erosion, or worse, during field and pasture use, few hardwood species can be expected az this stage to grow at acceptably ra~idrates or produce high quality wood, The pine species have demonstrated their eapaci~yto grow where hardwosds do poorly. Perhaps equally important is the capacity of pines to rapfdly improve the hydrologic characteristics of the soils as has been pointed out earlier, (cf. Figure 28). So the emphasis in conversion and renovation work should be on renewal efforts with pines. It should not be eonstrued that the hydrologic condition of the land cannot be damaged through the application of fsresr management operations, Careless location af access roads and trafls, indiscriminare skidding, failure to take fneo account differences in slope and soil type Figure66. Renovation of areas of junk hardwoods can often be achieved without excessive expenditures or special equipment. A small amount of marketable material covered the cost of forest renewal on this area; log- ging was. followed by a broadcast burn, then planting in March 1956 at the time the top photo was taken; below is a view from the same camera point showing the area as it looked in April 1972, Yellow-poplar was planted in the best spots and loblolly pine over the remainder of the area. tolerances to ruafflc, especxally In wer-wearher hggmg, and neglect fn applying fcflow-q measures when necessary to stabilize axsearbed soil in critical areas are some cf she n!aj&r factors ehae can reduce the qualiry of outflow from areas harvested and sihseq~entlgiegenereted, It is entirely feasible to carry aut all the necessary forest maaagemenr operations wfehsut adversely affecefng the watershed, as has been demonstrated by Hewlerc and Dsuglass (~9689, for example, in their study sf blending Sorest uses in mauntafnoils terrain, 8. T&etEler or not rhe munizipafity enters an exrensive addirional Land acquisition program, employmene sf a farester in an extension eapaei~yis srrongPy recom.ended. A professional forester fn esntznusus contacr with landowners wfthfn the watershed could stimulate the undertaking sf forest management on the private holdings while engaged in misslonary work to emphasize she cftizen respansrbXXfty rs improve the quality of water productfon* Ae the same time he could advfse how rh%s improvement csuld be achieved with minimum inputs or out-of-packer expendgtures by rhe landowner,while establishing an appreciatmg asset on his land. At the dutset this professional could serve as City Forester for Durham, so that his contributions wpuld be more evident to the taxpayers. 9, Construction of a stiffing pond or basin at the head of che reservoir. Currently the suspenssid material drops out when flow levels off; subsequently, durfng low pool levels it keeps moving farther into the reservoir, and eventually beyond easy reach (cf. Figures 35, 36 and 40). Continuing transport of sediments into the reservoir would justify investment In a stilling basin which would catch the suspensoid material in a concentrated area from which it could be dredged out perfodfcally and set aside to solidify and dry. With proper planning, the material could be marketed as "topsoil" or "top sift" for use in construetion sites where entire reliance is not placed on asphalt and concrete. A modest investment in a stilling basm would appear to be sensible as a measure to slow down future attrieion in reservoir storage capacity. REFERENCES Anonymous, 1953. Waynes7ikie watershed management demonstration. A 1953 progress repart. Town of Waynesville, TVA, and Division of Forestry, N. C, Department of Conservation and Developnent, 10 pp,, illus. Anonymous, 1966. The water supply system of the City of Durham, North CzroE%na, Department of Waeer Resources and Department of Public Works (Unnumbered bulletin), 18 pp,, fllus, Anonymous, l969* Research Triangle Region Development Guide. Research Triangle Regional Planning Commission (unnumbered atlas) 107 pp, , illus, Anonymous. 1970, Watershed Research in Western North Carolina, A study of the effects of agrfcultural covers upon the hydrology of small watersheds, TVA project Author. No. 455.1 Final Report 115 pp., iPPus. Anonymous. L97l, Erosion from public roads in North Carolina N. C, Association of Sofl and Water Conservation DEst, Agr, Ext, Ser,, and Soil Cons. Ser, (unnumbered bulletin), 13 pp., (processed) , BromffePd, Louis. 9947, Malabar Farm. Chapter X: The story of Kemperqs run, p, 218-225, Harper & Bras., New Uork, Brown, George W, 1971. Water temperature in small streams as influenced by environmental factors and fogging. Proc,, Symposium on Forest Land Uses and Stream Environ- ment, Conrinufng Education Book, School of Forestry & Department of Fisheries & Wildlife, Ore, State University, p, 175-181, illus, Coile, Theodore S, 1948, Relation of soil characteristics to site index of P~bloE%yand shortleaf pines in the lower Piedmont region sf Norsh Carolina, Duke University School of Forestry Bul, 13, 78 pp,, illus. Coife, T, S, 1952. Sofl and the growth sf forests. & Advances in Agronomy, IV, p, 329-398, ilPus, A. G. Norman (ed,) Academic Press, N. Y. Csrbetr, Edward S, and Robert P. Crouse. 1968, Rainfa19 interzepkdcn by annual grass and chaparaE,,,Tossee compared. USDA F~restService Research Paper PSW-48, 12 pp., fbbua, Croft, A, R, and L. V, Monnfgerc 195.3, Evap~traaepfra~fon and other water logsea on some aspen forest types in relation to waser available for sfreaa flow, Trans., Geophysical Union %464):563-574, illus. nk, Jn W, 1940, Forest resources of the Fiedmon~ ion of North Ca~oPina~U,S. Depar~mentAgr,, Forest Service Release N3, 6, 55 pp,, 211~~. Delfs, J* 1967, Enzereeptfon and e~emflowIn stands of Norway spruce and beech in West Germany, & William Eb Sopper & Howard W, Lull (ed.) International Symposium on Forest Hydrology. Pergamsn Press, N. Y., po 179-185. Dils, Rabert E, l953* Influence of forest cutting and mountain farming on some vegetation, surface soil and surface runoff cha~ascerist~cs,U, S, Farest Service, Sau~heassernForesr Experiment Station Paper No, 24, 55 pp,, fP%us. Dfseker, E. G., E, C. Rfchardson, and B. H, Hendricksm, 1963, Roadbank erosion and its eanrroL Bn the Piedmont upland sf Georgia. U, S, Dept, Agr, ARS 41-73> 40 gp,, illus, Dsuglass, James E, 1967, Effects sf species and arrangement of faresrs on ewapotranspfrarion, 2 Willfarn E. Sopper and Howard W, Lull (ed.) Incs%Symposium on Foree~ Hydr~logy, Pergamon Press, N. Y, pn 451-461 i%lu$. Eakfn, Henry M, (revised by) Carl B, Brsm, 1939. Silting of reservofms, USBA SofP Cons. Service Tech. Bul,No, 524:31-33, England, C, B, 197Qe Land capabilfry: H hydrologic response unir in agrieulrural watersheds, USDA ARS 41-172, 12 pp., illus, Hardy, APberc V,, Charles B, Carney, and Henry V, Marshall, Jr, 1967, Clfnare af North Carolina Research Stations, N, C. Agr, Exp, Sta, BulYerin 443, 75 pp,, illus, Hawley, Ralph C, and WiEIiam Maughan, 1930, The Eli WhPtney Forest: .A demonstration of forestry practice. Yale University School of Forestry Bul, No, 27, 46 pp,, fllus. Helvey, J. Do 1964, Rainfall interception by hardwood forest litter En rhe southern Appalachfans, U. S, Forest Service Research Paper SE-8, 9 pp,, flEua. Helvey, J. D, 1967, Interception by eastern white pine, Water Resources Research 3(3):723-729, Helvey, 3, D, and J, H, Patric, 1965, Canopy and litter interception by hardwoods of eastern United States. Water Resources Research 1(2):193-206, illus. Hewlett, John D, and James E, Douglass. 1968. Blending forest uses, USDA Forest Service Research Paper SE-37, 15 pp., illus, Hewfett, John D, and Wade L, Nutter, 9969, An Outline in University of Georgia Press, 13% pp., i Plus, Henderson, G. S,, Regina M, Anderson, L, Boring, J. W. Elwood, T. Grizzard, W. Po Harris, A, Hunley, W, McMasters, W, J. Selvfdge, J. L, Thompson and D, E, Todd. 1979. Walker Branch Watershed: A study of terrestrial and aquatic system interaction, Ecological Sciences Division Annual Progress Rpt. ORNL-4759, p. 30-48, illus, Hibbert, Alden R. 3.967, Forest Treatment effects on water yield. In Wil1fam E. Sopper & Howard W, Lull 4ed. 9 Int '1 ~~Gosibunon Forest Hydrology. Pergamon Press, N, Y,, p, 527-543, illus, Holeman, J. E. 1968, The sediment yfeld of major rivers of the world. Water resources research, 4:337-747. Holley, D, L, and W, T. Huxster. 1972. Forest ownership, Tarheel Economfst, Department of Eeonomfcs, N. C. State University, pa 2, Hoover, Marvin D. 1950, Hydrologic characteristics of South Carolina Piedmont forest soils. Proc,, Soil Sci. Soc. of America 14~3.53-358, ilius, Jurney, R. C., S, R. Bacon, and J, J. Morgan. 1928. Soil survey of Person County, North Carolina. USDA Bureau of Chemistry & Soils Bul, 14, 36 pp., illus. Knight, Herbert A. and Joe P, McClure, 1966. North ~arolina's timber. U, S. Forest Service Resource Bul, SE-5, 47 pp., f 1lus. Kramer, P, J. 1942, Species dffferences with respect to water absorption at low soil temperatures. Am. Jour, of Botany 29: 828-832, Levno, A1 and Jack Rothacher. 1967. Increases in maximum stream temperatures after logging in old-growth Douglas- fir watersheds, U. S. Forest Service research Note PNW-65, 12 pp,, illus. Lull, Howard W. 1965. Factors influencing water production from a forested watershed & Proc., Municipal Watershed Management Symposium, U, of Mass., Coop. Ext. Ser. Publication 446:2-7, illus. Lutz, J. Fulton. 1969. The movement and storage sf water in North Carolina soils and the role of the soil in determining water quality. Univ. of North Carolina WRRI Report No. 24, 43 pp., illus. MacConnell, William P. 1965. Management of the Amherst and Holyoke, Massachusetts watersheds. & Proc., Municipal Watershed Management Symposium, Univ. of Mass. Coop, Ext. Ser. Publication 446:58-60. Mader, Donald L., Richard S. Affleck, & Peter S. Machno. 1965. Watershed Management Research Program, Department of Forestry and Wildlife Management, College of Agriculture, U. of Mass. -In Proc., Municipal Watershed Symposium, Coop. Ext. Ser. Publication 446:22-26, illus. Martin, Irving L, and Turner C, Bass, 1940, Erosion and related land use conditions on the Lake Michie watershed, near Durham, N. C. USDA (unnumbered bulletin) 19 pp., illus, 1940. Mather, J. R. 1953, The disposal of industrial effluent by woods irrigation. Trans. Amer. Geophys. Union 34:227-239, Meadows, John C, and Thomas H, Nayfor, 1969, Simulation sf regional economic impacts sf water resources development, an expfsrarory study, Univ, of North Carolina WRRI Report No, 6, 2.6 pp, Meehan, WiYE%am R. 1970, Some effects of shade cover on stream temperatures in Southeast Alaska. USDA Forest Service Research NQC~PNW-113, 9 pp,, illus. Merrick, Elliott and E, A, Johnsan, 1952, Mountain water. Ameriem Forests 58110):30-32 (381, flfens. Mets, Louis J, and James E, Douglass, 1959. Soil moisture deplat%on wder several Piedmont cover types, USDA Forest Service Tech, Bul, Nc, 1204, 23 pp., illus. Merz, Louis J,, Carol G, Wells, and Paul R, Kormadik, 2,970. Comparing the fsreat flsor and surface soil beneath four pine species in the VZr inia Piedmont, USDA Forest Service Research paper SE-55, I pp, Miller, Willfan DD, 1954, Pine or hardwoods? A comparison of growth sates of loblol%y and Virginia pines and upland hardwoods in the Piedman% of North Carolina, Southern Lunberman 188(2355):3%-34, Musgrave, 6, W, 1955, How much of the rain enters the soil? U, S, Dept, Agr, Yearbook 151-159, illus, Penman, H. L, 1948, Evaporation in nature, Reports on Progress in Physics. The Physical Society of Great Britafn, XI:366-388, Penman, H. L. 1949. The dependence of transpiration on weather and soil conditions. Jour, Soil Seienee 1:%4-89, Penman, H, L, and R. H, Sehofield, 1951, Some physical aspects of as~imffationand transpiration, Symposium, Society far Experimental Biology, Gr. Britain, No, v, p, 115-3.29, Perkins, S. O,, R. E. Devereux, L, L. Brfnkfey, S, F. Davidson, and W, A, Davis, B924, SofE survey of Durham County, North Carolina. USDA Bureau of Soils. (unnumbered bulletin) p. 1351-1379, iffus. Reinhart, K, G. 1965. Inereash wazer production ehravgh manipu%atlan of vegemtlon, In Pnoc,, Municipal Watershed Managemens Symposium, U, of M~S,Coop, Ext, Ser, PsnbPicaCion 446:17-21, SEoeum, G, KO and W, D, MITPer, 19 Virginia pine: Reproduction, growth and manag ent on the Hi11 Demonstration Forces, Durham County, Nc C- Agr, Exp, Sta, Tech Bui, No, 100, 5% pp,, illus, Streeby, Larry E, 1971, Buffer srrfps -- same eonsiderations in she dea8aion c leave, Proc,, Symposium ~n Forest Land Uses a~dStream Environment, Continuing Education School of Forestry & Deparsment of Fisheries and Wildlife, Orec Stare UnBv,, p. 994-198, Swank, W, To 1968, The influence of rainfall interception on s~reamf%ow, elemson University WRRI report No, 4:fOP-112, illus, Timko, Mfchael Glen. 1962, Evaluation of volume, growth, quality and value of PobEokfy pine, Virginia pfne, and uphand hardwoods En the Iowcr Piedmont of North Carolina. 61, F, Thesis, School of Forest Resources, N, 6, State UwPv,, 106 pp (Processed), Turner, 9, F,, Jr, 1966, Evnporarfoa study in a humid region, Lake Michic, North Carolinab Geological Survey Professional Paper 2324; 133-150, if%us, Van Looek, Harry J, 697%. Land-use changes and the economic evaluation of nasural Virginfa pfne stands on the Dfak Creak wacarshed, Durham County, North Carolina, M, S, Thesis, School af Foresr Resources, N, C. State Unfversfty PO9 pp, , sblue , QPsoeessed) Zinke, Paul 6. 1963, Foresr fnrerception studies in the United S$a%as- & William En Ssppcr & Howard W, Lull (ed.) Int'E Symposium on Foresc Hydrofo yo Pergamon Press, N, Y., p, 137-161, GEO SSARY Accelerated erosion - Man-caused or man-related erosion as contrasted to geologfc norms. As an example, many streams in expandfngwrbanized areas are now undergofng accelerated erosion of banks as a result of man-related aetivftfes whfch have greatly increased volumes of stormflow, Acre feet - A measure sf volume of water, usually associated with estimates of storage capacity. An acre-foot is equivalent to a level acre covered with water to a depth of one foot, hence equals 43,560 cubfc feet or 325,851.gallons. Acre fnch - A measure of volume equal to 1/12 of an acre foot. It is useful in transfatfng inches of precipitation falling on land, pavement, roofs, etc, into gallons sf water, One acre inch is equal to 27,154 gallons. Basal area - A measure of stand density in a forest expressed in square feet per acre or square meters per hectare. It is derived from measurements of diameters of all trees, or of trees above some minimum diameter, at breast height (4 1/2 feet above the general ground level), converting the fnch measurements to square feet and summing them. Baseflow - The flow issuing from drainage basfns between storms. It 5s so-called groundwater outflow-sustained by aquifers and also by the slow drainage of.saf1 moisture in the vadose zone, or zone of aeration. It is always the 11 oldestffwater that reaches the stream channel since it fs the deferred portion of outflow preceded by surface runoff and subsurface stormflow. Butt rot - A term applied to deterioration of the basal portion of the bole of a tree as a result of heartwood- destroying-fungf.which enter.the tree through basal wounds or often fn sprout stands from the parent stumps. Channel precipitation - Precfpitation which reaches the stream channels dfrectly without detouring over and through the land. In sluggfsh streams in level terrain, rise fn stream fevel is essentially equal to the fnches of precipftatfon, particularly sfnee capillary deficiencies in stream bank soils are likely to be minimal. Pn swift flowing streams of hfgh gradfent, the channel precfpitatfon fs always a resfdual, something much less than the fnches of gross precfpftatfon. - Sesrage sf water fn rhe non-eapKL2.ary pores of the soif, Pn X~DEchannefs, animal and fnseeL burrows, and fn similar erevxces, (Non-capillary pores are pores generafly greater than O0k mfllfmeaer in diameter,) Waher ri deeention storage is only eemporarily held; depending cn the pemeabilfry 3f the surface and subssfl, such water flows IareraZly to stream channels or peresleees downward sas$sfymg eapflfary moisture deffciencfes, with any surplus dtima~elyreaching rhe wacer cable, FPducfal marks - Index marks, usnaBfy four, connected wrch the earnera lens and foemrng fmages on the negacfve, whxh are used to define rhe prfncipfe point, or center, of an aerial phctograph? - The route traversed by an aircraft during a photographic mission, FPoecuPant - The chernfeal added to water fn the pur3ffeatfsn process to cause suspended mater%al to aggregaee into a flocculent of sufficient size to preefpftare out cf the water, The most common fPoceulank. used is al-m, Forest floor - The a~eumulatfonof organlc matter resrfng on mineral soll under forest stands9 Sr comprises dfs~fnguishabPe layers of varyfng ehiekness depencjfng on such factsre as species mix of trees and lesser vegeratfon, elfmate and nfsrsry sf use sr treatment (ice,, grazing, fire, selective loggfng,_or high grading, ete, 1 - The applfcarfan of business mefhods and technreal forestry prfnzipfes to the operation of a forest property Green junk - A derogatory term used for describing the composition sf a forest which fs Heft or develops from a serfes of high-gradangs of the beef specfes and the hnghesr qmlfry Crees, leaving brush, vines, fnferfor specfes, and cnfP (defecrxve) trees to dominate the site, - The growing sroek of a srand of timber fs the sam of rhe vsfme or number of $& the frees, or pare ~f the rrees above some speerffed efze ffmfe, withln rhe area 2n question. The growing stock sf a mwsged forest eonsafniag areas of vario~ssize or age ciasses from seedifngs and saplings to rnaczre pulpwood or saw~inberis the average voHenme based on the summation of all the different areas, HochwaEd - AII old German term applned eo fore~" oorfg5nating from seed. It. Fmplfes forest renewafi In which the species esnposfffon and :iaafiry is under nomrnaliy szrfct eantrsl. Its couneerpert 2s -----Kiede~vaLd, a rerm used fcr denocfng forests of coppice arxgin, meanizg foiesxs which start from ~-Incrernenc cswe - a L,:= ez;rsteed frcm the bole of a tree w5eh a specfal.fy-designed augas, permittfag study of radial growth paerern an3 rate Wsea also for dec~rrnfnfngtree age fr: apeefes which e~lzfbxcdf s~",n~tavrxiial rfzigs; Pnffltrarisn - The prceEss of warer eaterkg the soil surface, Many factors influem? Enfnlfratfon, bur faresr floor 2s a major one became le ebso.bs che energy of rain drops and greaePy helps ro keep pwes in ite waderPymg soihopen, Enffltration capacity af a s-"..ni is the maxfrnm rate atwhieh water can enter t5e boii, Mass wastage - Downslope rucvenitrtk by slumps and slides; a com~onphenomenon in rercr&tl y arbsnf zed areas where new peaks sf stormflow nave developed as a res~~lfof e~rtensrvepaving and eonstrueefsn cutclcg of? hffltratlon sf ,?ah, - Fsxesz ffooc which accurmlztes in pure pine stands or stands predozirtantly fa pme species* - The fosesr Slosr which actarn~latesgenerally under hardwood forests after a sufficlenc lapse sf the, Nes-MaPthus5an - In a den5grkzing sense, a term applied to individuals who appear to be excessively concerned about the early probablllty hat mankind and rhe world wfEl run out of resources, especially food, Raw water - A term used by Che Durham waeer purificaaion plant to dis~ingufsh- the wacer coning from Lake MIchie from that in the settlfng basins located at the purification plant, - A wrn in water P~w,orrginating %n the Code Napoleon, and ckxefly concerned wf~hrhe natural use of wacer by an owner of land on a scream, where use of the water of the stream as far domestsc and household purposes, for drfnknng watec and for watertrig of domestic Enwestoek. Every ocher use, whether by she rlpar~anowner of someone else would appear zs bs classified as araifkctai use, The counter- part sf the riparxan docrrbne 2s the doerrine of prior apprspriatron whxh dcminares water law zn western U,S,; ir recsgn;zes vested r~ghrsand prsvides pmkeerian far rhem, Rusban - d cerm referring ta an area whch is primarrly resid~ntialbur wlthin which farming on a smaE scale rs carried on, SxBv%e~Prsre- The rhesry and prazrere sf conerollfag forest estab%fshmenk, csmposfcfsn and growth. xlso, the ape of producfng and tending a forest, - A term ~sedin ecmparing preduz%;uity sf land rowing, More speeffnedJ.y, ~heterm, site index 1s emp%oyed Pn q~snr~fyingland pradue~fvfryfor cfmber growth, The Tarter term fa an expression of $owl hefgkr, in feet, of the dsmxnant stand rehred go a base age, 30 pears having been used for esseern U,S, species an the past A snre fn6ex of 85 for Virginia ine,wsuld denote that ehe domxnan~sctind would attam 8% feer in 50 years, and for ~h~sspecies, xt would furrher denote a very good ssre- S~and- A stand of t~mberfs an aggregate ef Ecees and assocfsred woody vegeraEmn sufffeaently alike in species eornpssftisn, age, or condstmn to be distinguxahable from adjacent aggregares, Thus, age disefnguishes saplings from sawt~mber,florfstfe csmposstfsn dlstinguiahes yellow- poplar from Vfrginfa pine, and eond~tfondfstfngufskes a mass of green junk from a hoehwald- of oaks- - The d~sehargeof a stream frsm a given storm, shswmg szorir.fI~win @fs plotted over tme in mfnirtee, hours, or okher mits of time; the resulting curve is the hydrograph for t&ac storm, Hydrographs, sf course, may 'be plotted for strearn discharge fndependent of storms, and such graphs prov8de exeeEPent information abour he eondieion of drafnags basf ns , - The value of timber 5s fr stands meuc in rRe woods, based on estfrnates of vsfame and quaBfcy cierfved from timber csuxses ax snvenrorfes, Stumpage prices are fnf%uence&nos only by species, ~tanddensity, and guafxky, but by seraaln, soil, dis.p,arace. frsm mancdfactt-sing plant, accewsibiYSty, and. relared faetors, Subsurface flow - The pomxon of sereamflow reaching the channel so rapidly duziwg and after a given seorrn that it can 5e considered a part of the atom hydrograph, The entfre stormflow 1.~1streams from wall-forested watersheds usually devdops from substirface ffow, w%thout any overland fXow. Timber s%~ndfmprovement --- - BPI r:fit"hf~ng~ not a part of a major harvest felling, msde ducing the fife of a forest stand for the general purpose =f jrnproving rhe srand as regards csmposftnea, condi~lot;, or rate of growth- The series - A cerm applied LO ,a colPestisn of data which describes ~kernsvemewr of some ~haracrerfstfeover time, Watershed managemerit - The ixanagemenz of ~henatural resources of a drainage bash prfmariPp foi the psoducrfon and protection of water supplies and water-based resources, including reduction and esnrrol of accelerated erosfon, reduction sf b flood peaks, prorectfon of aestheefc values, but not excluding any use or activity compatible with necessary safeguardfng of water supplies from contaminaeixm and pollution, Water yea; - "New Year%@'day for the water year, or hydrologic year is October 1, PrequentLy, the water year is divided co quarters, the. period October 1 co December 31 being the time sf SGLI moisture recharge, January E through March 31 being &he period of ground warer recharge and the two remaining quarters bang the permd of maxfmum evapotranspiration and ground weer depletion, -efs - Cubie feec per second, a measure of stream discharge rate. One cu'bfe foot equals TC48gallons, One -cfs equals 646,316 gallons per day $24