SOME HYDERQLQGIQ CHARAC‘FERISTICS ”4 THE UTAH fiURlPER TYPE 9F HQRTHERN AREZQRA The“: for the Dogma of DH. D. MICHIGAN STATE UNIVERSETY Clarence McClelEand Skau 1960 SOME HYDROLOGIC CHARACTERISTICS IN THE UTAH JUNIPER TYPE OF NORTHERN ARIZONA CLARENCE McCLELLAND SKAU sSubmitted to the School for Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 1960 C?ZUO/7 5/2. 'r/ [a ’1 ii .HMucmOIpmmH .cozcwo ugh 10km mm scam ohm mpon .unwfi» may 0p :cmm soamehow xu0H MCwuudh may an umwmooH ma .moznmu xomho xmo Op awaken 0:» .mnocmm .hmHHm> mcho> mg» cuzfl mcwmo .mmum Acmanmov umummhu sachn 05a ounfl :mwhw Ho mcofiuowm .muoHn Haunmsfihmaxm may Mo mEom mo 3mfi> wswommlnpSOm < "momfiameCOpm iii ACKNOWLEDGEMENTS The author wishes to extend sincere thanks to Dr. Donald White who, as major professor, superVised the organization and writing of this study. The author is deeply indebted to the Rocky Mountain Forest and Range Experiment Station, both for permitting official work to be used for this study, and for its technical supervision. Sincere ap- preciation is personally extended to Raymond Price, Director; Marvin Hoover, Division Chief of Watershed Management; Dr. Jacob Kovner, Station Statistician; Mrs. Geraldine Peterson, Assistant Statisti- cian; and Mr. Earl Aldon, Project Leader. Appreciation is also extended to other members of the Committee, Drs. T. D. Stevens, V. J. Rudolph, W. D. Baten, and Prof. I. F. Schneider, for their valuable comments. The author would like to thank Mr. Milo James, State Soil Scientist, Soil Conservation Service, for identifying the soils on the study plots. iv VITA Clarence McClelland Skau Candidate for the degree of Doctor of Philosophy Final examination: December 18, 1959 Dissertation: Some Hydrologic Characteristics in the Utah Juniper Type of Northern Arizona Outline of Studies Major subject: Forestry Minor subject: Soil Science Biographical Items Born, May 13, 1928, Detroit, Michigan Undergraduate Studies University of Michigan, Literature, Science and Arts, 1946 - 1950. B. A. 1950. Michigan State University, Dept. of Animal Husbandry, 1951 - 1953. B. S. 1953. Graduate Studies Michigan State University, Dept. of Forestry 1953 - 1959. M. S. 1956. Experience: Forestry Aide, Research, U. S. Forest Service, 1954 and 1956; Research Assistant, Mich. State Univ. Agr. Exp. Sta., 1955; Graduate Teaching Assistant, Mich. State Univ., 1955-1957; Research Forester, Rocky Mountain For. and Range Exp. Sta., 1957-1959. Member: Xi Sigma Pi Society of American Foresters American Geophysical Union Arizona Academy of Science ABSTRACT This study is part of a research effort designed to obtain basic information relating to watershed management, with particular reference to the Beaver Creek Watershed Project in Arizona. The objectives of this study were to make a preliminary determination of the disposition of precipitation falling in stands of Utah juniper (Juniperus Utahensis (Engelm.) Lemm.) growing on a fairly average range of sites on the Beaver Creek Watershed, and to explore some of the factors con- trolling the disposition process. Other objectives were to provide_ complementary information regarding soil characteristics and root dis- tribution of Utah juniper in this area. The study area within the Beaver Creek Watershed lies on the Mogollon Rim at an elevation of about 5,500 feet. Average annual tem- perature is about 55 degrees, and average annual precipitation believed to be 16 to 18 inches. Topography is rolling to hilly and occasional steep-walled canyons are cut by ephemeral streams. Parent rock is Ccomposed of heterogenous volcanics. Nine study plots were located adjacent to three small watersheds now undergoing calibration in the Utah juniper type on the Beaver Creek Watershed. Data was collected for one year beginning in April, 1958. The soils on all plots were identified as belonging to the Springerville series. Average soil depth for this series is predomin- antly between 30 and 50 inches, although variations from 6 to 94 inches were found on the study plots. These soils are very hard when dry, and very plastic and sticky when wet. An important feature of vi this series is the high content of montmorillonitic clay, about 65 per- cent, which results in a comparatively large water-holding capacity of about 2% inches per foot of soil, with a pronounced swelling and shrinking accompanying the moisture changes. The consequent crack formation is believed to greatly increase the infiltration capacity of the soils. A study of the root distribution indicates that the great bulk of roots is located within the upper 3 feet of soil. Some roots, however, are known to occur at depths of at least 15 feet. Large roots are rather uniformly distributed throughout the profile, while a concentra— tion of fine roots occurs in the upper 18 inches of soil. An average of 15 inches precipitation fell on the nine study plots during the year. From hydrograph records on the three adjacent water- sheds, an average of about 0.20 inch occurred as surface runoff. A similar small amount was estimated to have entered the underlying basalt formation as percolation. A11 percolation and a disproportion- ately large share of surface runoff originated on only two plots, both having soils about 6 inches deep, little vegetation and slopes of about 6 percent. The balance of total precipitation, about 97 percent, eva— porated from either soil or leaf surfaces. With a net reduction in soil moisture storage at the end of the year, average total evapotrans— piration exceeded average total precipitation by about 3 inches. An attempt was made to characterize the density of Utah juniper stands in relation to their influence on the disposition of precipita- tion. 0n the nine study plots, eight stand density characteristics, dealing largely with estimates of crown volume and tree number,were vii measured. All stand characteristics turned out to be highly correlated with percent crown volume as measured by a spherical densiometer. Only interception was directly influenced by stand density. 0f the eight characteristics, crown density as measured by the spherical densiometer proved to be best related to the amount of precipitation intercepted. This relationship, when based on individual observations, was inversely related to the square of percent crown density. The per storm relationship between gross and net precipitation was found to be linear, and gross precipitation was found to be much more important for predicting net precipitation than stand density. The rate of soil moisture depletion was found to be fairly con— stant between 65 and 5 percent of available soil moisture, and to be uniform throughout the upper 30 inches of soil. During periods of de— pletion, amounts of soil moisture withdrawn in either 24 or 30 inches of soil were nearly equal for all stand density classes. It was shown that total evapotranspiration of soil moisture was directly related to average soil depth and net precipitation. Average soil depth was a better predictor of total evapotranspiration of soil moisture because it reflected differences in net change in soil moisture storage and be- cause it indirectly influenced net precipitation through its influence on stand density. ACKADWLEU VITA . . ABSTRACT LIST OF T. LIST OF F CHAPTER II. R III. I IV. TABLE OF CONTENTS ACKNOWLEDGEMENTS VITA . ABSTRACT . LIST OF TABLES . LIST OF CHAPTER II. III. IV. FIGURES INTRODUCTION AND OBJECTIVES REVIEW OF LITERATURE . DESCRIPTION OF THE STUDY AREA Location . Climate Vegetation . Geology and Physiography . Soils Land Use History . Description of Plots EXPERIMENTAL PROCEDURES AND RESULTS General Stand Description Spherical Densiometer Measurements “Line Intercept Method Plot Tally Method Results viii Page iii iv xi xii 32 32 32 39 40 40 43 43 50 50 50 53 53 54 54 Chapter U12 So In V, Chapter Utah Juniper Root Distribution . Results Soil Properties Interception . Results Stemflow . Interception Throughfall as a Function of Gross Precipi- tation . Stand Density Measures in Relation to Inter- ception . . . . . . . . . Throughfall as a Function of Gross Precipi- tation and Stand Density . Surface Runoff, Depression Storage, Subsurface Flow and Percolation Soil Moisture Percent Moisture Bulk Density . Soil Depth . Inches of Water Results Relation of Stand Density to Water Use Evapotranspiration of Soil Moisture Check Plot V. SUMMARY AND CONCLUSIONS General Soils ix Page 54 56 59 69 72 72 72 75 77 79 84 92 93 93 93 95 95 95 99 103 105 105 107 Chapter Chapter VI. VII. .Eflge Roots . . . . . . . . . . . . . . . . . . . . . . . . . 107 Stand Density . . . . . . . . . . . . . . . . . . . . . 108 Surface Runoff, Depression Storage, Subsurface Flow and Percolation . . . . . . . . . . . . . . . . . . . . 108 Stemflow . . . . . . . . . . . . . . . . . . . . . . . . 109 Interception . . . . . . . . . . . . . . . . . . . . . . 109 Soil Moisture and Stand Density . . . . . . . . . . . . 110 Evapotranspiration of Soil Moisture . . . . . . . . . . 111 The Check Plot . . . . . . . . . . . . . . . . . . . . . 111 LITERATURE CITED . . . . . . . . . . . . . . . . . . . . 112 APPENDIX . . . . . . . . . . . . . . . . . . . . . . . . 120 1. Methods of Determining Rates of Evapotranspira— tion . . . . . . . . . . . . . . . . . . . . . . . . 120 2. Techniques Used to Determine Soil Properties . . . . 130 3. Mbnthly Soil Moisture (Percent by Weight) Values . . 131 3a. Procedure Used to Correct Percent Moisture for Rock . . . . . . . . . . . . . . . . . . . . . . . . 134 4. Bulk Density and Related Soil Properties . . . . . . 136 4a. Procedure Used to Compute Average Bulk Density for Each Plot . . . . . . . . . . . . . ... . . . . 139 5. Average Soil Depth . . . . . . . . . . . . . . . . . 144 6. Inches of Water Stored in the Soil by Sampling Date . . . . . . . . . . . . . . . . . . . . . . . . 145 7. Values for the Water Balance Components . . . . . . 151 Tables 1. Tables 10. 11. LIST OF TABLES Average precipitation records for Ash Fork, Payson Ranger Station and Cibecue A surface cover survey Summary of stand description inventory data . Average values of soil physical properties on the study plots Soil cracking survey Total stemflow and interception during the period of study . Analysis of covariance for regressions of throughfall and gross precipitation per storm . Regressions of stand measures of density and percent interception Multiple regression analysis of throughfall, gross precipitation and the square of percent stand den— sity as measured by a spherical densiometer . Surface water movement, excluding storm number 18, for the period of study The depletion curve xi Page 36 48 55 62 67 73 76 78 80 90 97 Figure Frontis Figure LIST OF FIGURES Frontispiece: An aerial view of the study area, with four 10. 11. 12. 13. 14. 15. 16. 17. 18. plots outlined by subsequent cabling The Beaver Creek Watershed A modified Washington State College trapezoidal flume and housing for the water-stage recorder Weather station in the Utah juniper type Relation of Beaver Creek to the Salt River Location map of the study area The study area Record of an intense summer storm from the weather station . Heterogenous volcanic formation near the study area . Heterogenous volcanic formation near the study area . A Light density plot A Medium density plot A Heavy density plot Depth distribution of Utah juniper roots using two different methods . . . . . . . . . . . . A representative profile of the study plots showing a concentration of roots in the upper three feet of soil . . . . . . . . . . . . . . . . . . . . . . . . Range of soilfiprofiles found on the study plots Soil cracks at the surface on plot 3 Light Soil cracking to the CCa horizon Diagram of the self-swallowing action in clay soil xii Page 33 34 34 38 41 42 45 46 47 57 58 64 65 66 68 Figure 19. 20. 21. 22. 23. 24. 25. 27. 28. Stemflow collars in place on two average size junipers Interception of snow by Utah juniper Net precipitation as a function of gross precipitation and the square of percent crown density as determined with a spherical densiometer Net precipitation related to the square of percent crown density as determined with a spherical densiome- ter . A 2 x 3 foot subplot used to measure surface water movement on a Heavy density plot A 2 x 3 foot subplot used to measure surface water movement on a Light density plot Influence of juniper leaf litter on surface water move- ment . . . . . . . . . . . . . . . . . . Influence of rock on surface water movement Soil moisture sampling with a Veihmeyer tube Cumulative disposition of precipitation in Utah juniper - 1958 to 1959 — average of all plots xiii Page 71 74 81 83 85 86 88 89 94 100 INTRODUCTION The burgeoning industrial, agricultural and population growth of Arizona depends upon a corresponding development of its water resource. Most of Arizona's growth has taken place in the desert valleys. The city of Phoenix, for example, has nearly doubled its population in each of the last two decades, while the number of irrigated acres has in- creased from about one-half million to over one million during the same period (U. S. Dept. Commerce, 1957). Since precipitation within these valleys is scant, about six to ten inches annually (U. S. Dept. Com- merce, 1956), and ground-water supplies are being consumed more rapidly than they are being replenished (U. S. Dept. Interior, 1956), the in— czreased supply of water must come from other sources. Recognizing the problem and its urgency, the Arizona Watershed I’rogram was started early in 1956 as a cooperative undertaking of the Eitate Land Department, the Salt River Valley Water Users' Association, 21nd the University of Arizona. One of the first efforts of this group vvas concentrated on alleviating the water supply problem to the city of I?hoenix, located in the Salt River Drainage. Dr. George Barr of the IJniversity of Arizona led a group of technical consultants who examined ‘the Salt River Drainage and its tributaries, an area of about 12,000 square miles (Barr's: 31., 1956), in an attempt to formulate guide lines for increasing the yield of surface water. The conclusions and recommendations of this group were published by Barr St El. (1956). Cooper (1959) using some data from this report made an economic study of multiple land use on the Salt River Watershed. The relative importance of the major land uses was ranked in descending order as: water, recreation, timber and grazing. One of the conclusions of the Barr group was that the most rapidly feasible method of increasing surface water supplies would involve vegetation manipulation. This conclusion was based on the knowledge that vegetation markedly affects surface runoff through its effect on evapotranspiration and soil condi- tions, and that the effects are logically more pronounced as annual precipitation increases. With this report furnishing the initial impetus, a citizen's group, the Arizona Water Resource Committee, was formed to offer fur- ther guide lines for increasing surface water supplies to the Salt River Drainage. This citizen's committee is distinct from, but works in cooperation with, the Watershed Division of the State Land Depart- Inent. With state and federal support, federal funds became available 1:0 the U. S. Forest Service to expand watershed management activities 1 n Arizona. The federal appropriation provided for two thingsl: (1) a large- sscale pilot project, now called the Beaver Creek Watershed Project, Yvas to begin on the Coconino National Forest with the purpose of test- :lng effects of intensive land management practices on surface yields (of water, and (2) research, under supervision of the Rocky Mountain iForest and Range Experiment Station, to better guide the action pro- gram, and to gather basic information for watershed hydrology. 1Aldon. 1959. Unpublished file material of the Rocky Mountain Forest and Range Experiment Station. W1 W2 0‘; The Beaver Creek Watershed is about 275,000 acres in size. Within this drainage basin are two sub-drainages, the Wet and Dry Bea- ver Creek Watersheds. The pilot project's action program is concen- trated on about 30,000 acres of the Wet Beaver Creek Watershed, which includes three cover types: Utah juniper (Juniperus Utahensis (Engelm.) Lemm.; alligator juniper (Juniperus deppeana Steud); and ponderosa pine (Pinus ponderosa Laws.). Research of a similar kind is being conducted in all three cover types (See Figure l). The land management practice selected for the Utah juniper type was cabling. Arnold (1955) had shown that this practice increases grass production. In this operation the ends of a thick wire cable are attached to pins on two Caterpillar-type tractors. Moving parallel to each other about 50 to 100 feet apart, the 300 foot cable, forming an arc, is dragged along the ground. As junipers are encountered by the cable, they are uprooted and left on the ground. To test the effect of this practice on surface yields of water, the Rocky Mountain Forest and Range Experiment Station recommended the use of calibrated watersheds. The site for three contiguous water- sheds, 250, 310, and 330 acres in size, were chosen in 1956. Stilling wells, water—stage recorders and flumes were installed during 1957 (see Figure 2). A network of meterological instruments was also es- tablished within the boundaries of the three watersheds during this period (see Figure 3). After calibration at least one of the three watersheds will be cabled in a manner similar to that now taking place over the Utah juniper type on the Beaver Creek Watershed as a part of nun-on "ocu- IEAVER CREEK WATERSHED :c'rm-o - . on. non-u. ... .... ~ ff . _ 4. \ , -.-. Figure l. The Beaver Creek Watershed. Research and treatment are concentrated on W.S.A. , Wet Beaver Creek. Figure 2. A modified Washington State College trapezoidal flume and housing for a stilling well and water-stage recorder on one of the calibrated watersheds in the Utah juniper type. F 1m Figure 3. Weather station in Utah juniper type, showing recording and standard precipitation gages and housing for hygro- thermograph and maximum and minimum thermometers. FT the pilot project. One of the two remaining watersheds will be held as a control to evaluate the effect of cabling on the yield of surface water. In addition to their part in evaluating the pilot project, the meterological network and stream-gaging devices will furnish basic information concerning precipitation, temperature, humidity and stream- flow characteristics. It was felt, however, that additional basic in- formation was needed in relation to other factors which might affect water yields. The initiation of long-term projecusin this regard could not be immediately begun since so little was known of the hydro- logic characteristics of the area. Rather, a short—term project em- bracing many aspects of basic hydrologic research was undertaken on a plot basis to provide a portion of the initial data to be used in de- fining the area problems and orienting the direction of continuing re- search. These short-term, basic studies constitute the subject matter of this thesis. They include measurements of evapotranspiration, inter— ception, surface runoff and changes in soil moisture storage. These components of the hydrologic cycle were then related to stand density after a suitable method of describing Utah juniper stands was obtained. Other measurements were made of the ground surface, soil profile char- acteristics and root distribution of Utah juniper. Data was also taken from the meterological network and the stream-gaging devices. It should be stressed that no attempt was made to directly evaluate the large-scale pilot project treatment program outlined earlier in this section. 11 logic 1 Beaver provid lens a minati densit of sit tors w to pro root c OBJECTIVES The general objective of this study is to provide basic hydro- logic data in the Utah juniper type, with particular reference to the Beaver Creek Watershed Project in Arizona. This initial data will provide a portion of the information needed to define the area prob- lems and to orient the direction of continuing research. Specifically, the objectives were to make a preliminary deter— mination of the disposition of precipitation that falls in varying densities of Utah juniper stands growing within a fairly average range of sites on the Beaver Creek Watershed and to explore some of the fac- tors which control the disposition process. Secondary objectives were to provide complementary information regarding soil properties and root characteristics of Utah juniper. LITERATURE REVIEW The term "watershed management" implies doing something to a wa- tershed to achieve a certain purpose, and in this sense is an ancient practice. With the crystallization of the so-called "scientific method" it is possible to improve the efficiency of watershed manage— ment because the things done to a watershed are based upon understand— ing the physical principles which govern the behavior of a watershed. Physical principles governing the behavior of water have been studied in the laboratory. Heats of vaporization, kinetic motion, hydraulic flow, and a host of other properties have been studied and formulated for water. When this knowledge is carried to the field, however, it fails to provide an ability to accurately predict water behavior on a given watershed. Such a result is understandable in view of the com- plexity of factors operating on water. Watershed management research, then, must concern itself with devising methods of measuring the net effect of the physical complex, and synthesizing these measurements into empirical concepts which can be used to predict water behavior on a specified watershed. Knowing something of the climate, vegeta- tion, soils, and geology of an area, we measure the amount of surface runoff, evapotranspiration, percolation and other components of the "Hydrologic cycle" and relate them to the above factors. From these relations we attempt to arrive at concepts which enable us to predict the behavior of water on any watershed. As Kittredge (1948) points out, there are instances of watershed management research which date back many centuries. Authors sometimes 10 place its real beginning with a study of the effects of forest vegeta- tion on streamflow begun by Engler in 1890 in the Emme Valley, near Wesen, Switzerland (Burger, 1929a, 1929b). Research since that time has been summarized by Zon (1927) and Coleman (1953). It is not the purpose of this review to improve on their work, but to cite those studies which are pertinent to the objectives of this study, with par- ticular emphasis on research in the Southwest. The Water Balance Method. As McIlroy (1957) points out, all forms of the water balance method make use of the following basic equation: P = E + O + D + ASW P is precipitation E is evaporation O is surface runoff D is subsurface drainage [\W'is change in soil moisture For determinations of evapotranspiration on a large scale, the catchment area balance sheet is used. A natural drainage area, rangng: in size from about one to over several hundred thousand acres, is chosen. Precipitation is measured by suitably placed gages, and sur- face runoff measured as inflow to either natural and artificial reser- voirs or by stream-gaging devices. For long time studies, [5W’can be neglected; for short time studies, it can be determined by one of the many soil sampling techniques. D is usually estimated since its di- rect measurement is difficult. For small areas, an impermeable sub- stratum is desirable to eliminate this factor. E can be obtained by ll subtraction. An example of the catchment area balance sheet is fur- nished by Kohler (1957) and by Hoover (1944). Detailed instructions for computing some of the individual components have been compiled by Johnson and D113 (1956) and by Nash (1958). Wicht and Schumann (1957) discuss standards to use in the selection of experimental catchment areas. Where highly controlled measurements are desirable, or where men and equipment are lacking to sample large areas, lysimeters are used. These devices are essentially contained blocks of soil equipped to separately weigh the in-coming and out-going water increments. Re- views of the literature pertaining to lysimetry are provided by Kohnke, Dreibelbis and Davidson (1940) and Harrold and Dreibelbis (1955). The shortcomings may be summarized: (l) evapotranspiration is related to the size area of the surrounding crop given the same wa- ter treatment, (2) downward water movement is retarded by the air—soil interface for shallow depths, (3) soil structure is unnatural, and (4) root growth is inhibited. The alundum tension lysimeter developed by Cole (1957) is designed to provide a tension sufficient to overcome restrictions in water movement, and at the same time keep the soil un- disturbed. The comparatively hugh, 18x50x3-6 feet deep, Base Rock 1y— simeters at Sierra Ancha (U.S.D.A. 1953) rest on impermeable quartzite bedrock. Here the soil was enclosed in place. A rather specialized and widely popular form of lysimetry involves Thornthwaite's concept of potential evapotranspiration. The following quotation is taken from Thornthwaite and Mather (1955): 12 "One cannot determine the amount by which precipitation fails to supply water needs for crops without knowing what these water needs are. In order to make a map showing the distribution of water deficiency (the amount by which preci- pitation fails to supply sufficient water) it was first nec- essary to make a map of water need. This most important climatic element was defined as the amount of water which will be lost from a surface completely covered with vegeta- tion if there is sufficient water in the soil at all times for use of the vegetation. It was called potential evapo- transpiration." The method is essentially empirical, relying on mean air temperature, latitude and length of day to compute a heat index (I) roughly corre— lated with the amount of solar energy (Thornthwaite and Mather, 1957). Thornthwaite recognized the Oasis effect and provided buffer zones against it. Several serious flaws appear in the Thornthwaite approach. In the first place, the relationship between mean air temperature and solar heat energy is used as a constant. As Thornthwaite himself points out (1955): "However, since a higher proportion of the net radia- tion is spent on heating in arid areas, there is even with temperature a lack of conservatism which will result in some error in computing potential evapotranspiration." Second, the relationship between mean air temperature and poten— tial evapotranspiration is, according to Halstead and Covey (1957), ....complicated by the fact that actual evapotrans- piration tends to lower the maximum and mean temperature.... Third, potential evapotranspiration was supposed to be a climatic parameter independent of vegetation type. Conflicting reports have emerged. Both Mather (1954) and Blaney (1952) suggest that potential evapotranspiration is related to vegetation type. Penman (1949, 1951), however, supports Thornthwaite's point of view. Kohler (1957), as a 13 result of findings at Lakes Hefner and Mead, proposes the use of po- tential evapotranspiration to mean: " evaporation from a free—water surface of ex- tended proportions, but independent of any heat storage effects." Further, he states: "These two experiments proved rather conclusively that monthly evaporation from an existing reservoir can be reliably determined by the application of an energy budget, and that the day-to-day variations in evapora- tion are in accord with an empirical mass-transfer equation." To compute actual evapotranspiration by the Thornthwaite method, a procedure is used involving measurement of precipitation, assump- tions regarding surface runoff, percolation and soil moisture avail- ability, and evaluation of soil texture and root depth (Thornthwaite and Mather 1957). One of the most persistent problems involved in short time inves- tigations is that of adequately determining soil moisture, particul- arly over large areas. A number of techniques, as outlined by McIlroy (1957), are available for this purpose, but extensive replication is needed for even comparatively small areas. One of the most common and most accurate is gravimetric sampling. The procedure is, however, time consuming, laborious, and destructive of the soil body. For con- tinuous recording of soil moisture in place, the variation in electri- cal resistance with changes in soil moisture is commonly measured, using porous blocks embedded in the soil. All varieties of porous blocks have at least one major limitation. They may lack long term stability; display hysteresis effects; show dependence on temperature or salinity; measure limited ranges of soil moisture; are difficult to l4 embed properly in the soil. Another frequently used device is the ten- siometer, which measures soil moisture directly with the use of a porous bulb filled with water and connected to a manometer. This de- vice is subject to temperature fluctuation; measures only on the wet end of the soil moisture range (about 0.8 atmos.); requires good con- tact; and is subject to air leakage. A less frequently used technique employs measurement of differ— ences in the rate of temperature rise when heat is supplied to the soil, or else the vertical heat transfer of the soil is measured di- rectly by heat flow plates similar to net radiometers. A description of the technique is given by Porter (Halstead, 1954). A calibration is required for each soil, since the heat transfer is dependent on the particular thermal properties of the soil. According to Baver (1956), the flow of heat will increase with increasingly dark color, increasing amounts of coarse particles and water content (specific heat), and in- creasing compaction and water content (conductivity). A good contact is necessary for this method as the probes and plates are rather sen- sitive. The neutron scattering device is now becoming popular because it offers the advantages of rapid field measurement without additional laboratory work; it is factory calibrated; and is independent of the effects of temperature and salinity. Some of the disadvantages are high initial cost, rather elaborate safety precautions, contact prob- 1ems with the liner for the test hole, and variations in volume of soil sampled with changes in moisture content. There is also some difficulty in obtaining values for small intervals (Anonymous, 1959). 15 Precipitation. McDonald (1956), in a preliminary statistical analysis of selected Arizona precipitation records, found that the year to year variation in winter precipitation (November to April) exceeded the year to year var- iation in summer precipitation (May to October). Conversely, spatial variation was greater in summer than in winter. He points to the un- certainty of annual precipitation means in Arizona with the remark that stations with many decades of records had "95 percent confidence half- widths fully 10 percent of the mean." Secular trends, as studied with lO-year average-moving plots, were very pronounced. Stand Description. The importance of crown volume of trees in relation to hydrologic phenomena has generally been recognized. Early methods largely em- ployed some version of the "crown projection" technique, a laborious procedure at best, involving hand drawings of the vertical projection of the canopy to the ground. Recent techniques have done much to im- prove the efficiency of this method. Spurr (1948) outlines the methods with which aerial photographs now permit rapid and wide-scale deter- minations of crown diameters and crown closure to obtain stand density. Two methods of rapidly obtaining an estimate of crown volume from the ground are now available. Lemmon (1956) developed the "spherical den- siometer", a pocket-sized, convex, chrome mirror with a grid incised on the surface. Evans and Coombe (1959) have developed a type of pin- hole camera which reduces a photograph of a hemisphere to a flat plate. 16 On this photograph are imposed the track of the sun for any given day and a grid of concentric circles for each 30 degrees of altitude and radiating lines for each 30 degrees of azimuth. Other methods have been used to measure crown volume. Cooper (1957) modified the Bitterlich technique to measure crown volumes of shrubs. Cable (1958) related surface area and weight of fasicles of ponderosa pine. When this relationship is combined with estimates of fasicle weight per tree, the total surface area of fasicles for a stand.may be computed. Yamaoka (1958) related total leaf weight and tareast-height diameter to obtain total transpiration for a stand of (2rypomeria japonica. Canfield developed the Line Intercept method pri- rnarily for herbaceous species, but this method can be adapted to mea- surements of crown volume (1942) . Several authors (Bradshaw, 1943; Howell, 1941; Reveal, 1946; and lierman, 1953) have studied the characteristics of juniper stands, but cuily from the viewpoint of wood production. Reveal's (1946) data show tJiat even there a poor correlation was reported between tree sizes and 'hnaturity classes.‘ Irrterception. To date, no information has been published concerning interception in Utah juniper. Early interception studies in other species were sum- marized by Horton (1919). He states that the "amount of interception is primarily a function of stor- age capacity of the plant surface, the duration of precipita- tion, and the evaporation rate during precipitation." Interception by trees was 0.02 to 0.07 inch per storm. The percentage of precipitation intercepted per storm is larger the smaller the storm. l7 Depending on species and stand density, interception values varied from 15 to 80 percent of the total yearly precipitation. Species with needle-shaped leaves intercepted relatively greater amounts of preci- pitation than broad-leaved species. Stemflow accounted for l to 5 percent of total precipitation. Wicht and Schumann (1941) presented an historical review. They emphasized some of the limitations of the early studies, such as fail- ure to record stemflow, failure to sample randomly, and failure to use a sufficient number of gages. These experiments were conducted on gray poplar. The results of this work confirmed the principles re- ported by Horton. In addition, it was suggested that condensation in- creases throughfall values, and that open gages on paired experiments be placed at canopy level. Wilm and Niederhof (1941) conducted interception studies on mature lodgepole pine stands. Net rainfall increased by 0.10 inch for every foot increase in the average size of canopy openings up to about 18 feet. Net rainfall was a straight-line function of gross rainfall for all stand conditions. Grah and Wilson (1944) demonstrated three phases of leaf-surface detention: transitory storage, or water which drops off under still air conditions; conditional storage, or water that falls off when wind disturbs the leaves; and residual storage, or water removed only by evaporation. The work of Kittredge, Loughead and Mazurak (1948) with Canary Island pine indicated that stemflow was apparently not directly re- lated to crown-length density, tree height, basal area, or crown area; 18 however, it did tend to increase as tree height relative to adjacent trees increased. Total stemflow for the plot was approximately 1 per— cent of total precipitation, but as much as 13 percent was recorded for an individual storm. Stemflow commenced when precipitation for a given storm reached 0.20 to 0.25 inch. Interception was greatest on the leeward side of a tree; increased with an increase in wind velocity; and decreased with increased distance from the bole. Johnson's work (1942) with a forest of young ponderosa pine indi- cated that there was no significant difference in interception between seasons when amounts were adjusted to those expected for average gross precipitation. Hamilton and Rowe (1954), working in Chaparral in California, found 5 percent total net interception in buckeye-ceanothus-oak type, 8 percent in ceanothus—manzanita, and 11 percent in oak-ceanothus. The buckeye-ceanothus-oak type was partly deciduous. They state: "Stemflow was influenced to a considerable extent by the branching habit and character of the bark of the vari- ous shrub species." Shrubs having a smooth bark and upright stem had stemflow of nearly 11 inches out of 38 inches total precipitation. Shrubs with rough bark and a spreading branch habit yielded but 2 inches of stemflow out of 27 inches total precipitation. 1 Aldon and Curtis studied interception according to stand density of ponderosa pine, where basal area was used as a measure of stand 1Aldon, Earl F. and W. R} Curtis. 1958. Unpublished report of Rocky mountain Forest and Range Experiment Station. 19 density. They found that between 11 and 25 percent of the summer thun- dershower type of storm is intercepted by pole—size stands. Pole stands of from 48 to 90 square feet basal area per acre intercepted less precipitation than stands with higher basal area. Stemflow was found to begin when 0.20 inch precipitation had fallen in a single storm. Total stemflow varied from less than 1 percent to over 8 per- cent of total precipitation depending upon stand density. Evapotranspiration. The term, evapotranspiration, is customarily defined (van der Bijl, 1957) as, "The combined direct evaporation of water from a plant, soil or water surface and transpiration of water from plants." Physical Principles Governing Evapotranspiration. The following ac- count is taken from Daniels and Alberty (1956) and Haltiner and Martin (1957). Since transpiration represents but a special form of evapora- tion, the kinetics involved are the same. Kinetic theory postulates that molecules of any substance are in constant motion when their tem- perature is above absolute zero. This is the so-called Brownian mo- tion. As the amount of heat imparted to molecules increases, so does their velocity increase. With an increase in velocity comes an in— crease in momentum. Consider a body of water: at a given temperature the average velocity of the water molecules is constant. However, a distribution of velocities exists such that a certain portion of the fastest moving molecules will have a momentum large enough to overcome the attractive forces holding them in a liquid state. As these molecules escape, the 20 mean velocity of the remaining molecules decreases. resulting in a cooling effect. Heat is usually supplied from the surrounding medium, the mean velocity increases, and an equilibrium is established. If additional amounts of heat are supplied, as from solar radiation, the rate of escaping water molecules increases and a new equilibrium is reached. Few molecules can escape the earth's gravitational field, so the picture is essentially one of a closed container. As the number of water molecules leaving a water surface increase, the number of colli— sions in the atmosphere increase. The net result is a number of water molecules having a distribution of velocities and moving in random di- rections. Those molecules with the lowest velocities which strike most perpendicularly to the body of water are most apt to remain. When an equilibrium is reached where the number of in-coming and out-going molecules is equal in unit time, it is known as saturation vapor pres— sure. As the temperature of the system increases, the saturation va- por pressure increases. The Clapeyron Equation (Daniels and Alberty, 1956) describes this relationship. Dalton showed that each kind of molecule exerts a partial vapor pressure independent of other types of molecules. The particular equilibrium obtained for water molecules, then, will be independent of other gases in the atmosphere, and depen- dent only on the temperature of the system. When the number of out- going water molecules exceeds the number of in-coming water molecules to a water surface, evaporation occurs. At this time the actual vapor pressure of water is less than saturation vapor pressure. 21 Factors Affecting Evapotranspiration. According to Halstead and Covey (1957), the available heat energy at the air-earth interface is used in one of three ways: (1) to heat the air, (2) to heat the soil and vegetation, and (3) to evaporate water. The energy used to heat air, soil, and vegetation is usually referred to as "sensible" heat, while heat energy used to evaporate water is referred to as "latent" heat of vaporization. They go on to state the apportioning of the available heat energy depends upon: "(1) the temperature distribution in the soil and air and water vapor distribution in the air, (2) on the availa— bility of water for transpiration and evaporation, (3) on the thermal properties of the soil, and (4) on the rate of movement and mixing of the air in and over the plant cover." Of the four factors affecting evapotranspiration, only (2), the avail- ability of water for transpiration and evaporation, is particularly pertinent to this study. The other factors are intimately related to methods of determining rates, rather than amounts, of evapotranspira- tion. They are discussed in this connection in Appendix 1. An immediate question presents itself: how is the amount of available heat energy determined? For theoretical background the reader is referred to Johnson (1954). In summary, he shows that the heat energy available for evaporation comes from two primary sources, net radiant energy and energy supplied by the transfer of sensible heat from warm air masses. Net radiant energy is the difference be- tween incident solar radiation plus diffuse sky radiation minus re- flection of short wave radiation plus long wave back radiation. The amount of solar and diffuse radiation is fairly amenable to direct measurement, and tends to be constant when differences in day length 22 and cloudiness are removed. The amount of short wave reflection de- pends on surface properties, and is referred to as "albedo". For av- erage surface conditions, the value 20 percent is used, but reflection may vary from over 80 percent for fresh snow to less than 10 percent for either a dense forest or water surface (Landsberg and Blanc, 1958). The amount of long wave back radiation is shown to be fairly indepen- dent of surface properties, and is primarily dependent on mean air temperature, degree of cloudiness and atmospheric humidity (Penman, 1949). The problem of availability is almost entirely concerned with the effects that plants and soils have in modifying the flow of water to the atmosphere. The question may be asked, how does the physiology or morphology of the plant modify the material transport of water to the atmosphere? Theories of the movement of water through a plant are re- viewed by Meyer and Anderson (1954), Richards and Wadleigh (1952) and Bonner (1959). The general theory is that the amount of water trans- ported to the atmosphere in unit time is a function of the driving po- tential divided by the resistance. In electricity this concept is known as Ohm's law, and in plant-water relations as Van den Honert's law. It is convenient to consider this process in two stages, within the plant, and from plant to atmosphere. The driving potential within the plant is usually referred to as the diffusion pressure deficit (DPD) gradient, and is described by Meyer and Anderson (1954): "Whenever evaporation is occurring from the walls of the cells into the intercellular spaces, the diffusion- pressure deficit of the water in the mesophyll cell walls increases. Such cell-wall diffusion-pressure deficits are primarily imbibitional in origin. Water therefore moves 23 into the walls from the adjacent protoplasm, resulting in turn in a movement of water from the vacuole into the pro- toplasmic layer. The resulting increase in the diffusion- pressure deficit is in turn propagated to all parts of the cell. Within the lamina of the leaf, gradients of diffusion- pressure deficits, gradually increasing in magnitude from cell to cell in the direction in which water is moving, are established between the xylem ducts and the cells from which evaporation is occurring. Water therefore moves from a gi— ven vessel or tracheid into adjacent cells, which results in the development of a tension in the water column occupy- ing that element of the xylem. Concurrently tensions are similarly developed in other xylem ducts within the leaf. The diffusion-pressure deficit of the water in the xylem elements will be increased by the amount of this tension. Water under a tension (negative pressure) of 13 atm., for example, has a diffusion pressure just 13 atm. less than that of pure water at the same temperature which is not un- der tension." ' It is seen that the DPD gradient within the plant is expressed in terms of Dixon's "cohesion of water" theory, where tension, or "nega- tive pressure" applied to a contained water column is transmitted equally to all parts of the column because of the cohesive attraction of the water molecules therein. It is further shown that the DPD in the leaf mesophyll cells seldom exceeds 50 atmospheres, so that the total gradient from soil to mesophyll cell cannot exceed this amount. Resistance to viscous flow within the plant is comparatively small, so that there is no real problem of water availability within the plant. Bonner then shows that resistance to diffusion within the plant tissue is uniform, so that the final expression for diffusion within the plant follows Fick's second law. The chief barrier to transpiration occurs from leaf to atmos- phere. Milthorpe and Spencer (1957) have conveniently expressed this problem by the following formulas: ‘TI'D(e1 - ea), where T is the amount of water transpired, D is the R \flliIJ S f a; 24 co-efficient of diffusion, e1 and ea are the partial pressures of wa- ter vapor between the leaf and ambient air, and R is the resistance. R = d + , l , where d represents the exter— Sc + SID/[1 + SD (1/sI + 1/sw)] nal resistance encountered by the bulk diffusion of vapor away from the leaf and Sc, SD’ 81’ and SW the conductances (reciprocals of re- sistance) of the cuticle, stomata, substomatal cavities, and cell wall which constitute the various parts of the paths of flow of vapor from liquid surface to the surrounding air. Since the DPD gradient from leaf to air may easily exceed 1,000 atmospheres, it is evident that the resistance may be nearly a million times that encountered within the plant. Conductances of the substomatal cavities and cell wall are so large that they will have little comparative effect on the amount of water transpired. The effect of cuticle is imperfectly understood, but, since less than 10 percent of total transpiration usually occurs through this layer, its importance is minimized. Conductance of the stomata therefore assumes major importance in considering the availa- bility of water for transpiration. This subject is treated in detail by Meyer and Anderson (1954) and need not be reiterated here. The work of Milthorpe and Spencer (1957) casts additional light on the regulatory function of the stomates: "Stomatal movement was found to exert a large control- ling influence on the transpiration rate, whereas water content has an extremely small or negligible effect. An approximately inverse linear relation between transpira- tion rate and logarithm of resistance to viscous flow through the leaf is believed to be the resultant of an in- verse curvilinear relationship between the diffusive con- ductance of the stomata and log. leaf resistance and the decreasing difference of vapour pressure arising from the \Ii II. 4 25 higher transpiration rates with increasing stomatal conduct- ances. Nevertheless, the relation demonstrates that the transpiration rate is influenced by the degree of stomatal opening throughout its entire range." A serious problem inherent in most methods used to determine eva- potranspiration concerns the availability of soil moisture. Is water equally available to plants from field capacity to wilting point? If not, then how can this relationship be described? This question was brought to a head in a report by Veihmeyer and Hendrickson (1955), which invited discussion from other authorities in the field. Unfor- tunately, no general agreement was reached; however, certain foci of disagreement emerged. They pertained to: the suitability of various techniques for determining water loss; the measurement of small changes in soil moisture near the wilting point; the influence of dif- ferent soil types on the amount of water held at various tensions; the effect of soil diffusion pressure deficit (DPD) as related to DPD at the leaf-air interface. Smith (1959) attempted to answer the question: "which method of potential evapotranspiration estimation, combined with which theory of the variation of actual evapotranspiration with soil moisture, gives results which are closest obtained by observation?" The relationship between potential and actual evapotranspiration var- ies with assumptions made in regard to the influence of available soil moisture on the rate of transpiration. Thornthwaite's method uses the assumption that the above relationship varies linearly with the amount of available water in the rooting zone of the crop. Veihmeyer's as- sumption is that, between field capacity and wilting point, soil mois- ture has virtually no effect on the rate of transpiration. Penman's 26 method uses an assumption intermediate between Veihmeyer's and Thornthwaite's. Smith concludes: "It has been demonstrated that the moisture status of the soil cannot be estimated on the basis of using figures of potential evapotranspiration alone. It is necessary to account for the fact that actual evapotranspiration is not always equal to potential evapotranspiration. The theory that most closely fits the observed facts is that of Thornthwaite." Richards and Wadleigh (1952) have intensively reviewed the lit- erature on soil moisture availability, and no attempt will be made to improve on their work. Perhaps the most significant development since then concerns the work of Philip (1958). Reviewing Walters and Wadleigh’s work, Philip (1958) states that Wadleigh had contended that the presence of solutes in the soil water contributed to the DPD against which the plant had to work; while Walter held that the pre- sence of solutes was of no importance. Philip (1958) showed that both were right depending on the circumstances. Under conditions of high soil moisture stress and even moderate transpiration, much of the fi— nal transfer of water from soil to root takes place as vapor across a narrow gap surrounding the root. The narrow vapor gap acts as a semi— permeable membrane eliminating the solutes as a factor in the soil-to— root DPD gradient. Under conditions of low soil moisture stress and low rate of transpiration, transfer takes place as liquid water, and solutes do contribute to the soil-to-root DPD gradient. Methods of Determining_Evapotranspiration. The review of methods available for determining evapotranspiration follows the outline of MCIlroy (1957), with explanations and evaluations taken from original Papers whenever possible. The methods available fall into four groups 27 and are based respectively on determining: a. The water balance of the evaporating system. b. The upward flow of water vapor in the air layers near the ground. c. The energy available for the conversion of liquid water into vapor. d. A combination of (b) and (c). The water balance method has been reviewed previously. Methods (b), (c), and (d) pertain more directly to determining rates of evapotrans- piration, rather than amounts over a longer period of time. For this reason they are given a detailed treatment in Appendix 1. Evapotranspiration Studies Pertaining to Southwestern Species. Roeser (1949), using a simple type of lysimeter in Colorado greenhouse experiments, compared use of various species of tree seed- ling. Tree seedlings ranked according to increasing water use were: limber pine, ponderosa pine, Douglas fir, and Engelmann spruce. McGinnies and Arnold (1939), employing the same technique with grasses in Arizona, found that water use of perennial grasses was rela- tively uniform. He also found that trees and shrubs were not as ef- ficient users of water as perennial or annual grasses. Price (1958) cites the work of Rich (1952): "At the Sierra Ancha Experimental Forest in central Arizona from 1936-1939 water use of several plants was de- termined from lysimeters. Perennial grasses grown in ly- simeters under natural rainfall at an elevation of 2,500 feet used 92 percent of the precipitation, winter annuals used 98 percent, and 89 percent of the precipitation was lost from bare soil by evaporation. At an elevation of 5,100 feet comparative water use was 81 percent of the precipitation for grasses, 84 percent for evergreen shrubs, and 78 percent loss from bare soil. 28 "Watersheds in the mixed grassland-chaparral zone (elevation 3,800 feet) in central Arizona have evapo- transpiration losses ranging from 94 percent to 98 per- cent of precipitation; grassland-chaparral watersheds (4,500 - 4,900 feet), from 90 percent to 95 percent; forested watersheds (elevation 5,500 — 7,800 feet), from 77 percent to 90 percent, depending on depth of soil and slope. "Consumptive use during the summer period was about the same as precipitation and was related to moisture available. Little or no water other than surface runoff was yielded to streamflow during the summer. Water use during the winter depended on growing conditions. Sur- plus water first satisfied the soil-moisture deficit and the balance was yielded as streamflow. Type of vegeta- tion cover and its consumptive use in the spring growing season is believed to affect the amount of water yielded." Commenting on the San Dimas lysimeters in California, Price (1958) states: "Sierra Ancha studies have been verified from results of the San Dimas lysimeters in California (Colman and Hamilton), 1947. Woody species utilized all available moisture during the long,dry summer characteristic of that climate. Grass did not use appreciable moisture be- low 3 or 4 feet. During the winter season, pine and grass utilized the water more rapidly than scrub oak. In the years before the lysimeters were planted to woody species, they were sown to annual grass, which gradually changed to a grass-forb cover. While the soil-water loss under the annual grass did not extend below 4 feet of soil, the water in the lower soil layers was depleted when the stand of summer-growing forbs appeared (Colman), 1953." Blaney, Taylor and Young (1930) measured evapotranspiration in the chaparral type in California. They found that "A seasonal rainfall of less than nineteen inches is usually consumed by the brush cover before a portion of it reaches the ground water." and that grasses and weeds consumed only 10 to 12 inches of precipita- tion. 29 Dortignac (1956) studied relative water use of pinyon pine and Bouteloua gracilis grass in the Manzano Mountains of central New Mexi- co at an elevation of about 7,500 feet. Soil moisture was determined by electrical resistance blocks. When winter precipitation was suffi- cient to extend to limestone bedrock at 30 inches, evapotranspiration by pinyon pine exceeded that of blue grama. Evapotranspiration Related to Stand Density. Studies relating evapotranspiration to stand density are compara- tively few in number. Usually, stands of different species are com- pared. Goodell (1952), reporting on the effects of thinning lodgepole pine stands at Fraser, Colorado, states: "The evaporation and transpiration losses of moisture from the soil were found to be unaffected by the thinning treatments. This result is in support of the conclusion reached in the earlier study of harvest cuttings in mature lodgepole pine (Goodell and Dunford, 1948). In that study it was apparent that the effects of the cutting on autumn soil moisture were produced not by decreasing the soil- moisture losses but rather by increasing the rainfall reach- ing the soil. In the present study, the net rainfall was unaffected by the treatments and no influence was found on the soil moisture of late summer." Douglassl, in a study of thinning loblolly pine in South Carolina, found daily evapotranspiration rates to vary with basal area per acre. Daily evapotranspiration from April to September was 0.172, 0.146, and 0.133 inches for stands having 150, 75, and 45 square feet basal area per acre respectively. 1Douglass, J. 1958. Unpublished report. Southeastern Forest Experi- ment Station. .... 30 Lull and Axley (1958) found total evapotranspiration from stands of pine poles, pine saplings and reproduction to be nearly the same in five feet of soil from April to November. Moyle and Zahner (1954) studied soil moisture depletion by a young, even-aged hardwood stand and an all-aged cull stand of hardwoods in southeastern Arkansas. No differences in soil moisture use were mea- sured. Root Characterization. According to Furr and Reeve (1945), "While the distribution of roots varies greatly with species and soil, the concentration of absorbing roots is typically greatest in the upper part of the root zone and near the base of the base of the plant, and decreases with soil depth or distance from the plant." Stephenson (1935) discusses the distribution of roots of orchard trees growing on fine clays. He notes that root penetration is accomplished mainly through worm holes, insect burrows, cracks and old root channels, and that most of the roots are confined to the surface where soils are more permeable and better aerated. Where open spaces are found, such as cleavage surfaces of rock or soil, the roots will proliferate ex- tensively. Kramer (1946) found that suberized roots were able to absorb water, and pointed out that when soils were too dry for root elongation this process might be of major importance for plant survival. Breazeale and Crider (1934) have shown that "A plant which has a tap root growing down into a moist subsoil, with its lateral roots in a surface soil which has been reduced to wilting point, is able to draw a certain amount of the water from the subsoil and exude this water HE: 31 into the surface soil and keep the soil which is in direct contact with the feeding roots at near the wilting percent- This phenomenon may enable a plant to tide over peri- fl age . ods of moisture stress. Working with palo verde and mesquite in Arizona, they show that this process may be reversed, enabling roots to elongate in soils apparently at wilting point. According to Johnsenl, junipers have both lateral and tap root de- velopment. Young juniper have a pronounced tap root development; the lateral system becoming prominent with age. Laterals originate from tap roots at a depth of several inches below the soil, move downward and then outward to distances of over 100 feet in mature ,trees. Feeder twoots concentrate under tree crowns and at the end of the laterals. Insterals are deepest in coarse soils, however, in rocky soils they may penetrate the crevices to unknown depths. Johnsen has photographed Jlxniper roots 14 feet down a rock crevice. \ l Johnsen, T. 1958. Personal communication. DESCRIPTION OF THE STUDY AREA Location. The Beaver Creek Watershed is a tributary of the Verde River Wa- tershed, which in turn is a tributary of the Salt River Watershed (See Figure 4). The study area is located within the Beaver Creek Watershed in the northeastern corner of Yavapai County, Sec. 1, R6E, T15N and Sec. 31, R7E, T16N, in north central Arizona along the Mogollon Rim. It is approximately 60 miles south of Flagstaff1 and 30 miles southeast of Sedona (See Figures 5 and 6). Climate. Long-term weather records are not available for the study area. U.S. Weather Bureau records (U.S. Dept. of Commerce, 1958) of precipi— tation are shown for stations at Ash Fork, Payson Ranger Station and Cibecue in Table 1. These stations are located to the northwest, south and southeast of the study area along the Mogollon Rim by distances of 50, 35, and 95 miles respectively. They were chosen because they straddle the study area, had comparatively long-term records, and were similar to the study area with respect to cover type, elevation and annual temperature (See Table l). Precipitation probably averages between 16 and 20 inches per year on the study area. Records of precipitation taken from the network of 1 Location of the Research Center from which this study was conducted. _. ,_ .— --__,_——— Figure 4. State of Arizona -- relation of Beaver Creek Water- shed to Salt River. Dark bar in east-west position is the Salt River; light bar in north-west position is Verde River; Beaver Creek Watershed is shown as irregularly darkened area leading into light bar just after it slants to the northwest. 33 n). "A". u .I! I III. PEI-I‘M“, U i .; 34 Figure 5. Location map of study area. U.S. 66 FlagSta:f/\_\./\’\ U.S m 89A o B p- Mormon 8 Lake (I: rt \_ g Sedona //’ -‘\\’, w a» f’ "I a f" 2 ., :3 j / Stgneman ake C, l X Study O)Happy Jack Area ,8 1 R5 ..Beaver Ck. G Ranger Sta. "~ ( '¢ \~fi a C ./ SA _ f—J’ij Boundary of Beaver Creek Watershed 'HES xf’ 35 Figure 6. The study area. Mulican Canyon to Mulican's lL 1H D D Ranch __— ‘- r‘flé'" 5mi. L,‘ 1M ‘ Stream Gage . Recording PP'I‘. Gage + Weather Station 0 100 Ft. Sq. Plot /I\ —-— - —.._ Ephemeral Stream Channel HE! Table 1. Records for Ash Fork, Payson Ranger Station, Cibecue - Average Precipitation Payson Month Ash Fork Cibecue Ranger Station (inches) (inches) (inches) January 1.00 1.51 2.03 February 1.19 1.72 2.21 .March 1.05 1.80 1.79 April 0.94 0.99 1.12 May 0.31 0.44 0.36 June 0.45 0.58 0.59 July 1.75 2.42 2.37 August 2.48 3.24 3.57 September 1.47 1.92 1.98 October 0.69 1.24 1.04 November 0.60 1.37 1.45 December 1.27 1.69 2.00 Annual 13.20 18.92 20.51 Years Annual .Station Record Elevation Temperature Cover Type (ft.) (°F) Ash Fork 43 5140 54 Pinyon-juniper Cibecue 24 5300 54 Pinyon-juniper Payson R.S. 50 4850 53 Pinyon-juniper 36 HES 37 gages on Beaver Creek in the Utah juniper type show 27, 19 and 13 inches for 1957, 1958, and 1959 to October 31 respectively. The pre- cipitation occurs as short-duration, high intensity summer storms (See Figure 7) originating over the Gulf of Mexico, and as one to three day winter storms of low intensity originating over the Pacificl. Maximum lO-minute intensities for storms larger than 0.30 inches averaged 2.21 and 0.44 inches per hour for summer and winter periods respectively during the period of study. The distribution of the storms is worthy of comment. Periods of 75, 42 and 41 consecutive days with virtually no precipitation oc— curred. On the other hand, about 45 percent of total precipitation for the study period occurred in a 31-day period from September 7 to October 6. One storm during this period accounted for about 25 percent of total precipitation. Local ranchers report that snow is infrequent and of short duration on the ground. TWO years of personal observation support this view. The author also observed that soil freezing is sporadic (hiring the winter, never penetrating deeply or lasting for more than a day or two. ____ 1Curtis. Unpublished Report. 1958. Files of the Rocky Mountain For- est and Range Experiment Station. Figure 7. Record taken from recording precipitation gage near the study plots. An intense summer storm. 38 -' IE5 39 Average monthly temperatures shown below are taken from the hygro- thermographs records at the Utah juniper weather station: Year Month _1 _2 _§ _4 5 6 7 8 _9 10 ll 12 Ann. 1957 * 58 72 76 72 69 56 42 42 1958 40 44 42 50 66 74 78 76 68 56 45 45 1959 40 39 46 57 Ave. 40 42 44 54 62 73 77 74 68 56 44 44 56 *Instrument installed. The maximum and minimum recorded temperatures were 104 degrees and 6 degrees recorded in June, 1958, and November, 1958, and February, 1959, respectively. Relative humidities of 5 to 15 percent are not uncommon during the usually dry spring and fall. Vegetation. Utah juniper occurs as the primary overstory species on some nine million acres of land in Arizonal. On the Beaver Creek Watershed it occupies an elevation zone roughly between 4,500 and 6,000 feet. Above it lies the alligator juniper type, and below it the semi-desert grass— lamui type. It is found primarily in association with pinyon pine (Pinus edulis Engelm.) at the higher elevations, and with chaparral species, primarily turbinella oak (Quercus turbinella Greene), at the lxnver elevations. 0n the small, calibrated watersheds adjacent to the study plots, Utah juniper occurs as nearly a pure type. 1Jameson. Unpublished file material of the Rocky Mountain Forest and Range Experiment Station. HES 40 Geology and Physiography. The 1928 Geologic Map of the State of Arizona shows the entire Beaver Creek Watershed as an area of heterogenous Tertiary and Creta— ceous volcanics. Detailed geologic maps are not available, but Figures 8 and 9 show the heterogenous character of the volcanics within 4 miles of the study area. The underlying rock formations range from red, porous cinders to bluish, dense basalt. Soil pits on the study area invariably "bottom out" on a basalt layer. A topography map of the Camp Verde Quandrangle published by the [LEL Geological Survey in 1936 shows the land surrounding the study area to be rolling to hilly with steep-walled canyons cut extensively across the Mogollon Rim. The study area is located on a broad, rolling ridge top between these deeper canyons.(See Frontispiece) at an elevation of about 5,500 feet. A11 streams along the Mogollon Rim within the gen— eral area of the study plots are ephemeral. seas A soil-vegetation reconaissance survey of the Beaver Creek Water- shed was initiated in 1958. The Utah juniper type was mapped in 1959. Generally, the soils may be described as fairly shallow, 12 to 48 inches deep, with large local variations in depth. Rock, of pebble to ‘boulder size, is common in the surface foot of soil. The top inch is silty clay loam, grading into clay below. Montmorillonitic type clay [aredominates, as evidenced by pronounced swelling and cracking with changes in soil moisture. The soils on the study plots were tentatively :identified as belonging to the Springerville series.1 :IMilo James, State Soil Scientist, Soil Conservation Service MgHeS. View taken at a cinder pit within 4 miles of the study area. Soils of the Springerville series are underlain by an irregular layer of dense, bluish basalt which is underlain by compacted, buff—colored cinders. Rocks point up the diffi- culty of sampling for moisture or depth. 41 Figure 9. View taken within 100 yards of Figure 8. Under- lying rock formation is composed of porous, red— dish brown cinders. 42 \\~. . tuti. J.— 43 Land Use and History. The general area of the study plots remained "open" land until the establishment of the Coconino National Forest in the early 1900's. Since access roads are few and far between, the primary land use was winter range for cattle. As the Verde Valley became more settled, a small demand was made upon the area for fence posts and fuelwood. With the gradual opening up of the area for the Beaver Creek Watershed Pro— ject, an abrupt increase in the number of deer hunters occurred. The area is now part of an allotment for winter cattle range. Description of Plots Location. The plots are located between the weather station at Watershed #3 and the recording gage on Watershed #1, a distance of ap- proximately 3 miles (See Figure 5). Plots are located just off the small watersheds, but within 1/8 mile of the road connecting the flumes of Watersheds #1, #2 and #3. Number and Size of Plots. Nine main plots 100 feet square and (Inel 25 x 100 foot check plot were established. The nine main plots ‘were divided into a grid with 100 equidistant intersections. Numbered stakes were driven into the ground at each intersection, making 10 rows of 10 stakes and forming a 10 x 10 foot grid. A similar proced— 1ire was used for the check plot. All Utah juniper was cut on and 1Because of the limitations of time and man-power, it was impossible ‘to clear three areas for check plots as originally planned. It was :felt that a smaller check plot was adequate since the variability of nuost factors was appreciably decreased by removing the trees and shrubs. 44 within 100 feet of the check plot boundaries to eliminate live roots, and the plot was then cleared of slash. Stand Composition and Density. Species of trees or shrubs other than Utah juniper did not exceed 10 percent of the crown density or number of individuals growing on each plot. These were usually turbinella oak or pinyon pine. The nine main plots included three plots each of three stand den- sity classes as measured by a spherical densiometer (See Stand Descrip- tion for an explanation of the spherical densiometer). They were: 10 to 25 percent, 35 to 45 percent and 55 to 65 percent (See Figures 10, 11 and 12). Exploratory work with the spherical densiometer indicated 65 percent maximum crown density in the study area. A crown density of 10 percent is about the minimum that occurs uniformly over an area of 100 feet square. In selecting plots for this study, the densest and sparsest, naturally occurring, typical stands of plot size were chosen and inventoried first. The intermediate density class was then es- ‘tablished with density about midway between the densest and sparsest. Surface Cover. Unusual surface cover conditions were avoided, such as concentrations of herbaceous vegetation, surface rock or slash :from fence post cutting. Jeep roads, well-used cattle trails or former «camp sites were excluded. Surface cover was inventoried by stretching £1 tape along the rows of stakes on each plot and recording observations lit whole foot intervals. One thousand observations were thus taken on (each of the nine main plots. When juniper leaf litter was encountered, :its depth was measured to the nearest 5 inch. Results are shown in Treble 2. Note the fairly regular increase in percent litter and HE! 45 Figure 10. A light density plot (9.9 percent). 1E! 46 Figure 11. A medium density plot (39.8 percent). Note pre- cipitation gage in left foreground. E I I Figure 12. A heavy density plot (58.6 percent). 47 HE: 48 Table 2. Surface cover survey. Average Litter Depth Bare 1 Where Over Plot Rock Ground Litter Misc. Sampled Plot percent of area -——————inch 1L 39.4 53.0 2.0 5.6 0.38 0.008 2L 38.2 54.0 2.9 4.9 0.74 0.021 3L 43.6 42.2 11.9 2.3 0.79 0.094 Ave. 40.4 49.7 5.6 4.3 0.041 1M 47.1 42.8 8.7 1.4 0.66 0.057 2M 36.5 45.1 16.4 2.0 0.82 0.134 3M 26.9 58.6 12.0 2.5 0.64 0.076 Ave. 36.8 48.8 12.4 2.0 0.089 1H 41.0 35.0 18.6 5.4 0.97 0.180 2H 36.9 46.3 16.6 0.2 0.75 0.124 3H 24.3 46.4 ' 26.7 2.6 0.64 0.171 Ave. 34.1 42.6 20.6 2.7 0.158 lIncludes soil cracks, herbaceous vegetation and large branches or fallen tree trunks. 49 average litter depth over the plot with an increase in stand density. JUniper litter does not occur over the entire plot, of course, but was expressed on a plot basis to permit comparison. Soils and Slope. Between and within plots, soils were as uniform as possible with regard to rock content and soil type. Soils are dis- cussed in detail in a separate section. It was felt desirable to locate the plots on well-drained slopes not to exceed 10 percent. Levels were run on each plot with a transit and rod. Percent slope is shown below: Percent Percent Plot Slope Plot Slope 1L 4.8 3M 3.1 2L 5.7 1H 5.8 3L 3.1 2H 5.4 1M 6.6 3H 2.6 2M 4.5 Ck. 2.3 EXPERIMENTAL PROCEDURES AND RESULTS General. The disposition of precipitation was determined according to the "water balance" method which uses the formula: Ppt. = SRO % DS / SF % P / I / / Et(S) / [13M, where Ppt. — precipitation SRO - surface runoff DS - depression storage . SF - subsurface flow P - percolation to groundwater I - interception ET(s) - evapotranspiration from the soil ASM change in soil moisture Surface runoff, depression storage, subsurface flow and percolation constitute minor components in the disposition process, and are dis- cussed together. Precipitation and interception likewise form an in- dividual study. Change in soil moisture and evapotranspiration from the soil are treated together. Methods used to describe stand density are discussed in a separate section, as are the characterization of Utah juniper root distribution and soil properties. Stand Description. The presence of vegetation on a given area influences either di- rectly or indirectly all of the variables in the "water balance" equation. The greater the mass of vegetation, the greater the expected irufluence. The problem lies in finding some easily measurable vegeta- tive characteristic as an index of this influence. The vegetative 51 characteristic most likely to influence the disposition of precipita- tion in juniper stands is total crown volume per unit area, rather than the more conventional measures of usable wood volume or compon- ents thereof. Both volume and surface area of material in the crown are good indexes of leaf litter and organic matter contributed to the soil. Leaf litter and organic matter in turn influence infiltration and surface runoff. In addition, the volume or surface area of leaf material is presumably a good index of the amount of rain or snow in- tercepted and the amount of water transpired by the plant. Crown volume likewise influences micrometeorology, which in turn influences the amount of water evaporated. Some direct measure of the total crown volume or leaf surface per unit area therefore seems the best way of describing stands for this study. Actual weighing of leaf material, or making volume or surface area determinations are extremely time consuming methods, and would destroy vegetation on the plots. Cooper's development of the Bitterlich method (1957) is unsatis- factory because of the wide-spreading crowns which restrict the visa- bility of the observer in the denser stands. Canfield's Line Intercept xnethod (1942) offers promise since it is a standardized technique which can.be applied to Utah juniper, and it permits fairly rapid field work. 11 simple count of number of individuals per unit area by height classes :is likewise a standardized technique and permits very rapid field and « (affice procedures. Number of individuals per unit area multiplied by :rverage crown length, although not a recognized technique, would permit rapid field work and would also reflect differences in age, size and carowding effects among the various stands of plot size. 52 Lemmon (1956) recently developed a pocket-sized spherical densio- meter for "determining from a point the relative amount of light that is cut off by specific areas of the forest overstory." A convex mirror of highly polished chrome is fitted in a recessed box 3 1/2 x 3 1/2 x 1/8 inch. A cross-shaped grid containing twenty-four 1/4 x 1/4 inch squares is incised on the mirror surface. The observer visualizes four equidistant dots in each square, and records the number which occur in the reflection of the overstory. Each dot represents nearly 1 percent. Four measurements are taken at each selected spot, one measurement fac- ing each cardinal direction. Crown density is computed according to the following formula: % crown density = 100% - (Tot. No. Dots Not Occupied) x 100% (100 Loc. x 4 Readings x 96 Dots/Reading) The height at which the densiometer is held depends upon the purpose of the observation. In field trials, using different operators and dif- ferent instruments, there was no significant difference among measure- lnents of overstory density. The spherical densiometer appears to offer good.promise of characterizing stands of Utah juniper on plots as they influence the disposition of precipitation. To test the suitability of these various techniques, the density (pf the plot-size stands of Utah juniper was determined by the following lnethods and the results compared with each other: 1. Percent crown density using a spherical densiometer (%CD-D). 2. Percent crown density using the Line Intercept method (%CD—L). 3. Number of trees per plot higher than 3, 6 and 10 feet. 4. Total length of live crown per plot on trees higher than 3, 6 and 10 feet. 53 Stand density measures were tested against differences in inter— ception and soil moisture change, which, outside of evapotranspiration from the soil, were the major components in the disposition process. When differences in precipitation are held constant, differences in evapotranspiration from the soil are largely a function of intercep— tion and soil moisture change, so that a stand density measure related to interception or change in soil moisture would also be related to evapotranspiration from the soil. Results of these analyses are dis— cussed in the'appropriate sections. The best stand density measure would be the one best able to predict the amount of precipitation in- tercepted or the amount of soil moisture used. If more than one stand density measure had nearly equal predictive value, preference would be given to the most efficient one for field and office work. Spherical Densiometer Measurements. A spherical densiometer was lised to measure crown density at each of 100 points on the 10 x 10 foot grid on each plot. The instrument was held at a height of 10 inches above the ground, the height of the precipitation gage used in the in- ‘terception study. Four readings per location were taken, one facing in.each cardinal direction. Line Intercept Method: Lines were run down the ten rows of satakes established on each plot. Live crown projection over the tape :vas recorded to the nearest foot. A 12-foot pole was used to determine :1 vertical line from the crown to the tape. Holes or irregular crown edges were measured if they exceeded one foot in length along the tape. 54 Plot Tally Method. A tally on each of the nine main plots was made of the number of individuals by height class and the length of live crown of each individual. Results. The characteristics used to describe the density of plot-size stands of Utah juniper are summarized in Table 3. Correlation co- efficients of percent crown density by the densiometer (%CD—D) with all other measures are shown at the bottom of the table. The strong correlation makes use of two or more measures unlikely in any multiple regression with interception or soil moisture. It is worth noting that correlation coefficients improve with increase in tree height and increase in tree height associated with length of live crown. Utah Juniper Root Distribution. Utah juniper, like any other plant, will use soil moisture from soil zones influenced by its roots. Proper understanding of the trans- pirational use of water therefore requires knowledge of the extent to vvhich roots are distributed throughout the soil profile. Of primary :meortance to watershed management is a knowledge of the root—depth distribution of the various plant components on the watershed. Assum- ing a non-restrictive soil depth, a deep—rooted species will transpire Inore water than a shallow-rooted species. The amount of water trans- ;xired will affect the amount of water remaining for surface runoff or percolation . Two methods were used to determine depth distribution of Utah ;hiniper roots. In the first, soil profiles along the recently 55 Hm. mm. om. mm. om. mm. mm. .mooo .huou o.mm a.mm a.pom o.aem m.maa m.mm m.vm m.mm .m>< m.mm m.om van was vmm we «a mm mm H.mm >.om saw no» mma an mm as an m.mm m.mm sum fine oom mm on we ma a.am H.Ne o.eem m.mve m.oae m.mm a.mm o.mv .m>< m.mm m.mm .va wmm one mm mm He 2m m.am m.mm one omm awn mm on an 2m m.Hv m.me mmm mmm woe mm mm mm 2H m.vH 6.62 a.na m.omH a.mmH m.m m.fiH o.em .o>< n.2m a.mm ”ma ems «mm 0H 6H mm Am m.m m.m mm Hm 40H m m 6H Am m.mH «.mH He mm maH m m an AH .wm .am .pm .5« .pm .pu unwound unmoama .5“ ca .au m .5“ m .pm ca .5“ 6 .am m pamoampsu mafia a0u050amson can? nonwam mmoae mo uon cask Mona”: .oz mafia: ho>ou saoao mom numsoa cacao m>aq Hauoe uoam hem mmoae .oz .uon .aaae maoaem>cs :ofiunflaomwu cemum Ho sausasm .m can“? 56 constructed Black Canyon Highway were freshly exposed and root counts made by size and depth distribution. The area of investigation lies about three miles west of the study plots. The vegetation is similar to them and the soils belong to the Springerville series. In the second method, five randomly selected soil moisture sam- ples were saved from each six-inch depth interval on each of the nine study plots. The five samples were combined, reduced to a slurry by adding water, and washed through a 2 mm screen. Only a very few of the finest rootlets passed through the screen; the majority were re- tained. “Separation of rocks and roots was accomplished by air drying the roots, submerging both in water, and screening off the roots as they floated to the surface. This process achieved almost 100 percent root recovery. The roots were then oven-dried and weighed. Results. The distribution of roots with depth using the soil profiles was computed and the average number of roots per square foot is.plotted against depth in Figure 13. From the second method, per- cent oven-dry root weight x 1,000/oven-dry weight of soil is also plotted against depth in Figure 13. These values are averages for plots with an average soil depth of at least 36 inches. It is apparent that the great bulk of roots is concentrated in ‘the upper 36 inches of soil (See Figures 8 and 14), and that a concen- ‘tration of fine roots occurs in the upper 18 inches of the soil pro- file. Larger roots appear to be quite generally distributed within 'the upper 36 inches. The concentration of roots so near the surface czan probably best be explained in relation to the frequency of soil Inodsture recharge. There is no water table to encourage deep root soil depth (inches) 12 18‘ 24‘ 30'‘ (O O) I A N .b on 54‘ 60« 66‘ ‘727 78h 84m. 57 number of roots per square foot (average) 90.1 percent oven dry root weight/oven dry-soil weight x 1000 (average) 1 2 3 4 5 6 7 / / / o / / / l / / Root distribution using / ° soil moisture samples X / _ __ __ Root distribution using soil profiles ° " / / O I / Figure 13. Depth distribution of Utah juniper roots using two different methods. .5 1.0 1.5 2.0 2.5 3.0 Figure 14. , A representative profile of the study plots. Note the large concentration of roots in the upper three feet of soil. Refer also to Figure 8. 58 ‘ES 59 penetration, and soil moisture measurements reveal a rather consistent lack of water below 36 inches. By way of comment it has been the author's experience to find Utah juniper roots occupying rock crevices over 15 feet below the ground. Johnsen1 reports Utah juniper roots as deep as 50 feet. Soil Properties. One of the most commonly found soils in the Utah juniper type on the Beaver Creek Watershed belongs to the tentatively correlated Springerville series.2 The soils on the experimental plots have been identified as belonging to this series3. A tentative description of the Springerville series follows4: Springerville Series The Springerville series consists of brown, moderately well- drained to well-drained Grumusols that occur in the Reddish Prairie soils zone. They have developed in place over basalt lava and basalt (rinders and bombs. They occur on smooth to slightly irregular surfaces tinder a sparse cover of vegetation including western wheatgrass, west— :ern yellow pine, juniper, and some bromegrass. These soils are associated with, and are similar to the McNary series, differing principally in having brown rather than dark-gray A1 1 Private communicat ion . 2Private communication with members of the soil survey party, Soil (konservation Service, which mapped this area in 1959. 3In collaboration with Milo James, State Soil Scientist, Soil Conserva- tion Service, U.S.D.A. 4Nationa1 Cooperative Soil Survey, U.S.A. iE! 60 horizons. Springerville soils are associated also with those of the Show Low and Elledge series but differ from them greatly in parent material and horizonation. The Show Low series is classified as Red— dish Chestnut, and the Elledge series is a Planosol. Soil Profile: Springerville Clay (virgin). All 0-2 inches Brown (7.5YR 4/2) to dark-brown (7.5YR 3/2) clay; strong very fine granular structure; hard when dry, friable when moist, very plastic when wet, very sticky; few plant roots; noncalcareous; pH paste, 6.6; pH on dilution, 7.0; abrupt wavy boundary. 1; - 4 inches thick. A12 2-17 inches Brown (7.5YR 4/2) to dark-brown (7.5YR 3.5/2) clay; weak, crude, very coarse prisms break to strong very coarse and coarse irregular angular blocks; extremely hard when dry, very firm when moist, very plastic, very sticky when wet; abundant roots; few very fine tubular pores; noncalcareous; pH of paste is 6.5 and, on dilution, 7.4; clear wavy boundary. 10 - 17 inches thick. A13 17-30 inches Brown (7.5YR 4/2) to dark brown (7.5YR 3.5/2) clay; strong very coarse and coarse irregular angu- lar and subangular blocky structure; extremely hard when dry, very firm when moist, very plastic, very sticky when wet; common roots; many slickensides at angles of 5 degrees to 45 degrees from the horizon- tal; noncalcareous; pH of paste 6.5, on dilution, 7.4; clear wavy boundary. 9 - 15 inches thick. (a33 30-43 inches Brown (7.5YR 4/2) to dark-brown (7.5YR 3.5/2) clay; with common medium and fine white mottles due to lime segregations; massive; extremely hard when dry, very firm when moist, very plastic, very sticky when wet; few very fine roots; very strongly calcareous; pH of pasua7.7, on dilution 8.3; abrupt wavy or irregular boundary. 10 - 20 inches thick. llr 43 inches Hard, weathered basalt. Range in Characteristics: Depth to bedrock ranges from 18 to 60 irushes but is dominantly between 30 and 50 inches. Occasional to many basalt cobbles occur on the surface and in the profile in places. Sur- face cracks § to 15 inches across and 15 to 20 inches deep develop when 61 the soil is dry, but when the soil is saturated these are difficult to locate. Clay and stony clay types have been reported. Drainage and Permeability. Runoff is very slow and permeability through the profile is very slow. When dry, water enters through cracks. Vegetation. Occasional yellow pine, juniper, blue grama, western wheatgrass, and some black sage. Use. Most of this soil is used for range. A limited area is cul- tivated and used for growing vegetables and small grain. Distribution. This soil has been mapped to date in the vicinity of Show Low, Arizona. Series Proposed. Navajo County Soil Survey area, Arizona, 1957. Springerville is the name of a town in Apache County, Arizona. Type Location. in the NE} of the NW}, SW} Sec. 24, T9N, R22E. Arizona. Determinations were made of color, pH, texture, upper plastic limits, bulk density, moisture at 20- and 60-cm tension and wilting point. Results are shown in Table 4, and techniques used described in .Appendix 2. Profiles of three experimental plots are shown in Figure 15 to demonstrate the variability in depth, stoniness and C horizon *which may be encountered in the field.. With summer and winter wet seasons of limited precipitation, two alternate periods of wetting and drying occur each year. Since most <>f the clay fraction is montmorillonitic, swelling and shrinking occur (on.a pronounced scale. Figures 16 and 17 illustrate the cracking which 62 a Hmlwa o.om m onion n.nm menwm ma manna m.om ma amuse m.«m mauvm em «Mica b.mH ea ooumv H.am emuma Hm manna «.mH m «mime m.Hm «Hum Hm hasbitv. m.Ha a «mine «.mv one 8323 do a muses do a. 335$ 66 a emcee do a aonssz mason waded“; monssz eased ofiumcaa some: m e\n as «\o mam o m.m-m.a H.w menom NH V\m 0» M\v mum NH m.wnm.b o.m mmnvm ma V\v 0» «\m mwm ma m.wlo.b b.h VNINH w «\m m>m.b use m\e Ow «\m awn w b.buo.b m.h «Hum m «\m mam.b use m\v 0» «\N mam m o.mum.b th one moaaemm Aaaoam smfiecoxv aoHoo moaaasm cases an nonasz amnesz m Nbuom o.mc vb.vm m.o> mmumfi N.HN mane o.m wvumm ma vbuvo N.bm venom h.¢b mmuma w.HN NHno m.b omuvm ma mbuow m.oc ownmm v.0b mmlma «.mm OHIO v.6 vNINH HH outmm m.vm Nhnmm e.hm mmumm H.0N «Hue m.m male ma mmnmv o.om «atom N.mm vmnmm v.8N mate v.m muo psooava .a«I moHaEam as as as as canon honesz emcee moo.» emcee moosw. emcee moo.um.o oases m.ouo.m aeso teem seam «Ham seem annexes .muoHa modem so mofipaoaoaa Hfiom mo mosaa> omaao>< .v manna 63 v m.mm m.» v.mv n.0m N.om o.bm v.nm m.mm Hv.HImN.H mm.H wvlwm NH m.mm m.v m.wv o.vm v.8v m.mm o.mv v.mm me.~nmm.~ mm.H mauvu ha o.vm 0.0 «.5v «.mm N.mv h.mm N.om m.mm ev.~lmm.a . mm.a «NINA m m.mm v.8 m.mv h.mm N.bv m.mm ¢.om b.mm H¢.HI¢N.H Hm.H «Hum h v.Nm m.b c.8v H.wm m.me h.mm m.Hm m.Hv mm.Han.H VN.H mum ea m.mm H.mH m.wv m.v¢ v.0m m.mv m.mm N.Hm NN.anmm.o NH.H «no A! R u. dd m. 5.0 R 5.0 R oo\msw .sH moaasmm moaom " meson >m " 3m >m " an >m " an amass " canon aonasz .aao " .amouaoz sofimamh sonom scamsob EUION sofiumHSHam hawmson xasm fiooH u woman whom .Apossfipsoov .v canoe 64 Clay~fi (3 Clay —9 Basalt t) i 30-36" Layered Basalt Intercalated with Clay Basalt Boulders Cindery Clay § \1 7 Basalt 2-48" Basalt m Figure 15. Range of soil profiles found on the study plots. IE! Figure 16. Soil cracks at the surface on plot 3 Light before soil pit was dug. July 15, 1958. 66 Figure 17. Soil cracks extending 24 inches deep to Cca horizon of Springerville profile. 67 may penetrate to the C horizon. Figure 18 illustrates the self- swallowing action which ensues. Soil cracks on the study plots were first observed to open during May. In order to gain some estimate of the extent of the cracks, a study was initiated in mid-July when soils were extremely dry and cracking most pronounced. A tape was laid along the ground down the 10 rows of stakes, so that 1,000 feet of line per plot was sampled. Whenever a crack intersected the line, its width was measured, and the location of the crack with respect to its occurrence in bare soil, next to a rock, or under litter was recorded. Cracks finer than 1/16 inch were not measured, nor could any estimate be made of cracks not reaching the surface. Results are summarized below in Table 5. Plots 1 and 2 Light are omitted because their shallow depths do not permit a good comparison. It is worth noting that surface cracks larger than 1/16 inch are not found on plots 1 Medium and Heavy. Percent soil moisture on July 22 is included in Table 5 to suggest the reason for this difference. Table 5. Soil cracking survey. No. Ave. % Soil Moisture % Occurrence Jplot Cracks Width 7-22 Soil Rock Litter In. 1M 0 21.8 1H 0 21.3 2H 54 .80 18.8 50 2 48 2M 209 .74 17.6 49 20 31 SM 227 .62 17.0 74 10 16 3L 249 .65 15.2 68 23 9 3H 297 .71 14.5 51 13 36 a b c Dry Soil After Wetting Wet Soil Boulders and Lime.Concretions o / Soil Crack ———> O C? C) ””,;; 1 Q M Line of Stress C: ‘—_""’ -—————<> Caliche Basalt a. Cracks have developed to the caliche layer. b. The cracks are beginning to close, with pressure due to swelling building up. The stresses leave a year-long record by the forma— tion of "slickensides", or smooth cleavage planes. Soil clods will break along these planes. The nearer the surface, the more inclined. c. The crack has closed, but because coarse particles have washed in, will open along this weaker area next year. Rocks have been car— ried to the surface over many years. Figure 18. Self-swallowing action of Springerville clay. 69 A comparison of percent litter (from Table 1) with percent cracks occurring under litter is also included in Table 5. The disproportion- ately greater percent cracks under litter suggests that this zone, di— rectly beneath the crown, is more completely dried than that in the open. Such a result would be expected in accord with the usual distri- bution of tree roots. The high percent of silt plus clay at the soil surface is very conducive to rapid sealing and low infiltration rates. This condition is exaggerated by summer storms of relatively high intensity and by the fact that vegetation cover is scant. The high percent of less than .002 mm clay and very low percent of non-capillary pores are good in- dicators that the soil is comparatively impermeable to downward or lat- eral flow of water. The soil cracks, then, seem to be the main avenue of rapid moisture penetration and subsequent movement of water within the soil body. That moisture can penetrate rapidly was demonstrated by the period September 12 to October 6 when about 6 inches of precipita- tion fell on the small, calibrated watersheds. This period included two storms, about 3.50 and 1.00 inches respectively, with maximum 10- xninute intensities exceeding 4.00 inches per hour. Surface runoff lmeasured at the flumes of the watersheds averaged only about 0.22 inch during this period. Interception. Part of the precipitation falling on a stand of Utah juniper never :reaches the ground. It is retained by the leaves and bark of the plants axui is then evaporated or absorbed by the foliage. This part of the 70 hydrologic cycle is called interception, and the amount of precipita- tion reaching the ground is referred to as throughfall. When the stor— age capacity of the leaves and bark is exceeded, water runs down the branches and trunk to the ground. This portion is referred to as stem— flow. The objectives of this study were: 1. To estimate throughfall per storm as a function of gross pre— cipitation per storm. 2. To determine which stand density measure is best related to total percent interception, and to define the nature of this relationship. 3. To estimate throughfall per storm as a function of gross pre- cipitation per storm and the best measure of stand density. The method used follows the formula: Net interception = Total precipitation - (Throughfall / Stemflow) Measurements of precipitation, throughfall and stemflow were taken on the nine plots. Precipitation in the open was measured by two gages per plot. Open gages were so located that there was no foliage or other obstruction within 45 degrees of the vertical line above the gage. They were located as close to the plots as such natural openings permitted, and were placed on opposite sides of the plot. Throughfall was measured by four precipitation gages per plot. These gages were placed at randomly selected locations, and moved to new random locations after each storm greater than 0.02 inch as mea- sured in the open. All precipitation gages were set with the top of the collection funnel level and 10 inches above the average ground surface (See Figure 11). Stemflow was measured on three trees per plot by means of collars. (See Figure 19) fitted to the stems. The individual trees were Figure 19. Stemflow collars in place on two average size junipers. Five gallon cans were used to col- lect stemflow. As much as 25 gallons were collected from individual trees after a storm. 71 72 representative of those found on the plot as regards size, configura- tion and distance to adjacent trees. The following equation was used to convert pounds of stemflow from three trees per plot into equival- ent inches of depth: Inches precipitation per plot = .000006407 x number trees per plot x pounds stemflow from three trees. Results. Stemflow. Total stemflow for each plot is summarized in Table 6. Because of the relatively small contribution of stemflow, it was omitted from further interception analysis. It is interesting to note that stemflow begins somewhere between 0.25 and 0.30 inch rainfall. No stemflow was measured after snow storms,even those exceeding 1.00- inch water equivalent when temperatures the following day were well above freezing. Figure 20 shows interception of snow. No snow re- mains on the trees although several inches are on the ground." The more rapid loss of snow from crowns can be logically ascribed to greater evaporation opportunity through increased wind movement and lower albedo of dark foliage. Some snow also blows off. It is also worth noting that the values of 0.25 to 0.30 inch for the point of stemflow beginning are higher than those generally re- ported by Kittredge (1948). This is probably attributable to the in- creased storage capacity created by shaggy bark. Interception. Gross precipitation, interception and percent crown density using the spherical densiometer (%CD—D) were tabulated by plot, by storm and by individual precipitation gages. Thus, there 73 Table 6. Total stemflow and interception during the period of study. Percent Gross Stemflow Inter- Percent Plot Precipi- Stemflow of Gross ception of Gross _‘ tation Precipitation Precipitation in. lb. . in. in. IL 14.89 1,424 0.15 1.0 0.96 6.4 2L 15.65 644 0.09 0.6 0.28 1.8 3L 14.29 1,040 0.18 1.3 1.43 10.0 Ave. 14.94 1,036 0.14 1.0 0.89 6.1 1M 15.80 578 0.18 1.1 2.88 18.2 2M 14.83 871 0.44 3.0 2.53 17.1 3M 14.39 335 0.09 0.6 2.87 19.9 Ave. 15.01 595 0.24 1.6 2.76 18.4 1H 15.79 423 0.22 1.4 3.07 19.4 2H 15.03 549 0.57 3.8 3.58 23.8 3H 14.22 619 0.46 3.2 3.06 21.5 Ave. 15.01 530 0.42 2.8 3.24 21.6 Ave. 14.99 720 0.27 1.8 2.30 15.3 Figure 20. Interception of snow by Utah juniper. Notice that no snow remains on the trees, while sev- eral inches is on the ground. No stemflow occurred from this storm. 74 75 'were 136 measurements of interception per plot, 408 per density class, and 1204 total. The data show that Utah juniper intercepts amounts of precipita- tion according to the density of the stand. These amounts were totaled for the period of study. They were then averaged by stand density classes and for all densities combined as shown in Table 6. Throughfall as a Function of Gross Precipitation. Individual re- A gressions of throughfall (Y) over gross precipitation per storm (X) were computed for each plot. They are listed below with appropriate correlation coefficients: . Correlation Plot Regression Coefficient 1L 3? = .9445x - .0093 .997 2L Y = .9907X - .0050 .999 3L Y = .9723X - .0293 .996 1M Y = .9281X - .0513 .988 2M Y = .9488X - .0440 .993 3M Y = .7962X - .0020 .990 1H Y = .9088X - .0509 .991 2H X = .7568X - .0005 .995 3H Y = .8897X - .0434 .996 Analysis of variance was used to test variance due to regression. The "F" values for variance due to regression were all significant to 0.1 percent. The "a" term was never significantly different from zero. An analysis of covariance was run on the individual regressions, first collectively, then by density classes, and finally within density classes (See Table 7). Individual regressions were then plotted by density classes. It was observed that nearly all of the variation in throughfall could be accounted for by regression. Since each of the . . -IWIIJ. a £4... . I [I’m]; 76 Table 7. Analysis of covariancel: Using regressions of throughfall over gross precipitation per storm. Source d.f. . £32}: MSq. F Between A11 Plots Within 1206 21.1634 0.0175 . Reg. Coef. 8 2.6643 0.3330 l9.03** Common 1214 23.8277 0.0805 Between Light and Medium Plots Within 808 13.2584 Reg. Coef. 1 0.5380 0.5380 31.46** Common 809 13.7964 0.0171 Between Medium and Heavy Plots Within 808 19.3997 Reg. Coef. 1 0.1113 0.1113 4.62* Common 809 19.5110 0.0241 Within Light Density Plots 1 vs 2 9.34** 1 vs 3 3.09 Non. Sig. 2 vs 3 1.36 Non. Sig. Within Medium Density Plots 1 vs 2 0.56 Non. Sig. 1 vs 3 14.18** 2 vs 3 25.13** Within Heavy Density Plots 1 vs 2 21.18** 1 vs 3 1.97 Non. Sig. 2 vs 3 37.06** 1After Snedecor. 77 :regression equations describes a "tight-fitting" curve, small differ- ences in slopes of the curves were predictably shown to have signifi— cance when tested by covariance. An apparently puzzling problem arose when differences within den- sity classes Medium and Heavy were found to exceed differences between them. Inspection of the plotted regressions showed plots 3 Medium and :2 Heavy accounting for the large differences within density classes. From Table 6 it is readily apparent that differences in percent crown density within a stand density class do not provide an explanation. The explanation was provided by examination of the storm-size classes sampled. No storm sizes between 1.17 and 3.35 inches, and only two storms of about 1.00 inch, occurred. Much of the total variation in throughfall could be accounted for by variation in these large storms. Plots 3 Medium and 2 Heavy intercepted appreciably more precipitation during these few large storms than the other plots. Additional samp- ling of large storm sizes would probably "iron out" the slope differ- ences within the density classes. Determining the Best Stand Density Measure in Relation to Total Percent Interception. Total percent interception was computed for each of the nine plots by the formula: Tot. % Interception = ( 1 - Tot. Throughfall ) x 100 ( Tot. Precip. ) Regressions of total percent interception over each of the eight stand density measures used on the nine plots were computed. They are listed in Table 8 with appropriate correlation coefficients, "F" values due to regression, and standard errors of regression. 78 .mososs o.nH* me.o Ho.m mH.me no. as.e x xmmo.o w soon as so>o moose mo nausea :soso o>AH fleece .m mm.o mm.m m~.mu om. me.e x xaao.o a some m hose moose so nausea cacao o>sfi Hence .s am.o ss.m ma.mm mm. Hm.m x xemo.o w some n pose moose mo names” nacho o>sH sauce .8 om.o om.m oo.em Hm. ms.v \ xme.o w some oH sons seems: some» soossz .m om.o Hm.m oo.mm am. mo.e \ me.o a some o sens seem“: moose sonasz .e ma.o ss.e mo.mH Hm. oH.m x xen.o w some n sane seams: moose sonssz .m mm.o Hm.m m.NN hm. om.H \ xmm.o w Apaooaovafi essay avenues cacao paooaom .N mm.o sm.a m.es no. am.o x xmm.o a. Asoeoaoemsoec sefiosoo geese scouted .2 cos“ R :m: a scammohmom oHQaHam> usopsoaoeau enhm mnlmmma hon aoamm cuccssum .aofiuaooaopaa uaooaoa use mufimsoe mo moasmmos psmum mo mcofimmoawom .m canoe \11 IIt.:i|...n._ .. 79 Table 8 shows that the stand density measure best related to total percent interception is percent crown density as measured by the spheri— cal densiometer. Total length of live crown for respective tree heights is consistently better related to total percent interception than number of trees. However, number of trees higher than 10 feet shows a better relationship than total length of live crown on trees higher than 3 and 6 feet. Percent crown density as measured by the Line Intercept method shows no advantage over number of trees higher than 6 feet. In terms of inches of water, the standard errors for the various stand density measures for the year of observation are listed in Table 8. For practical purposes the number of trees higher than 10 feet could most easily be field sampled with little loss of precision as compared to percent crown density by the spherical densiometer (%CD-D). Whether the relationship between number of trees higher than 10 feet and total percent interception remains the same for years having different preci- pitation patterns is a moot point. Mast likely the relationship will change, but only to a small degree. Throughfall as a Function of Gross Precipitation Per Storm and the Best Stand Density Measure. A multiple regression was computed using the following variables: X1 = Gross precipitation per storm. X2 = (%CD--D)2 at the random location where throughfall was measured for each storm. Y = Throughfall measured at each random location per storm. Results and tests are shown in Table 9 and plotted on Figure 21. While the relationship between gross precipitation and throughfall is linear, 80 Table 9. Throughfall as a function of (%CD—D)2 and gross precipitation per storm. Multiple Regression Equation. 9 = 0.026 / 0.907x1 - 0.208x2 A Y = Estimated Throughfall X1 = Gross Pgecipitation per Storm m= I 0 : E . I 6 7*. 0 «p I a t 3 ; Q) 6" g . C) 2 I s .51 m L .4i 0 3‘, Q ' 0 .2 (3 C) .17 \9 C) ‘G I 7 f- 1 —~--'r '- ‘ r".——“~‘r———"’t - l O! 00 10 . 20 30 40 50 60 70 80 90 100 percent litter Figure 25. The influence of juniper leaf litter on surface water movement. Storm of September 12 to 13, 1958 3.35 to 3.92 inches. 88 Ilitlvlull .blltbu'r..u. ... 1.41 1.3< 1.2+ 1.1A 331.0‘ .1: 0 I: "‘ .9. I p 5 E .8 (D > O E .7. H G) +3 a a .6“ 0 O (n ‘H s. .5- :3 U) .4J .3i .2~ .11 Figure 0 G> <30 0 G 67 G 09 000 0 o 0 G (9 00 C) C) 0 O O O O o O . . g r ‘ 10 20 30 40 50 60 70 80 90 100 percent rock 26. The influence of rock cover on surface water movement. Storm of September 12 to 13, 1958. 3.35 to 3.92 inches. 89 90 Table 10. Surface water movement excluding storm number 18, for the period of study. Number Storms Plot Surface Water Producing Surface Average CD—D Collected Water Movement Soil Depth in. equiv. in. % 1L 0.88 7 6.0 16.2 2L 1.20 5 6.0 9.9 3L 0.10 1 21.1 23.7 Ave. 0.73 4.3 11.0 16.6 1M 0.64 3 48.0 46.5 2M 0.56 3 32.6 39.9 3M 0.16 2 48.0 39.8 Ave. 0.42 2.7 42.9 42.1 1H 0.24 1 48.0 58.6 2H 0.14 1 37.6 60.7 3H 0.14 1 30.0 56.9 Ave. 0.17 1 38.5 58.7 .. . AI‘I “ear. 91 surface runoff as measured by the water~stage recorders on the small, calibrated watersheds. Surface runoff from these watersheds averaged a total of about 0.22 inch during the entire period of study, 0.11 inch occurring during sampling periods 9 and 10 (September and October) The value 0.11 inch was assigned to the Medium stand density class, typical of the stand density on the small, calibrated watersheds. The value 0.11 inch was then multiplied by the appropriate index of either 1.5, 1.0 or 0.4 to obtain realistic surface runoff values for periods 9 and 10. Light density plots 1 and 2 were given individual treatment for September, 1958. Since the soil depth of these plots is about 6 inches, and they were very dry prior to storm number 18, they could absorb only about 1.50 inches of water. Precipitation was 3.36 and 3.92 inches on Light density plots 1 and 2 respectively for that storm. Therefore, about 1.86 and 2.42 inches were unaccounted for on these plots. Two recording precipitation gage charts both show that the storm lasted about fifteen hours, and that about one-half of the total amount occurred as high intensity bursts, likely to cause water to run off a saturated soil. Accordingly, one-half of the 1.86 and 2.42 inch amounts was assigned to surface runoff, the other half to percolation. Depression Storage. This quantity was negligible for the year, since the plots were located on uniform slopes permitting fairly rapid drainage. Observations taken on the plots immediately after storms showed no free water surfaces of any size. Subsurface Flow. Since total flow on the small, calibrated water- sheds was only 0.22 inch, and since the bulk of it occurred as sharp 92 peaks following two intense summer storms, subsurface flow was assumed to be negligible. Percolation. Some percolation to groundwater may have occurred during the year through large soil cracks or otherwise. The soil never reached field capacity on any of the sampling dates, and was very dry during most of the year. It is very doubtful that appreciable amounts of water could have percolated to the underlying basalt formation. Soil Mbisture. The importance of soil moisture can hardly be overstated in rela— tion to watershed management. The soil profile has a definite storage capacity for water, and the amount of surface runoff for a given storm will be governed in part by the degree to which this capacity is used prior to the storm. The amount of available water in the soil profile will also govern the amount of water evaporated from the soil and transpired by plants. The degree of susceptibility to compaction and erosion at a given time is influenced to a great degree by the amount of water in the soil. To account for the disposition of precipitation using the "water balance" method, and to explore the nature of soil moisture trends, periodic measurements of changes in soil moisture were made. Inches of water stored in the soil at any sampling date was computed by the gen— eral formula: Inches of water = Pw x Pa x d- 100 Pw is percent moisture, oven dry basis Pa is bulk density d is soil depth in inches 93 Percent Moisture (PW). Percent moisture samples by six-inch in— tervals were taken monthly from April, l958,to April, 1959,at four randomly selected locations on each plot (See Figure 27). Field samp- ling and laboratory procedures are well standardized. In order to sample to parent rock it was often necessary to reject two to ten samples, in which case samples were taken at locations resembling the original random point. Appendix 3 shows average Pw by sampling date, by plot and by 6-inch depth interval. Appendix 3a outlines the pro- cedure used to correct Pw for rock included in the samples. Bulk Density (Pa). Bulk density samples were taken by the follow- ing 3-inch depth intervals to bedrock: 3 9-12 21-24 33-36 45-48 57-60 6 15-18 27-30 39-42 51-54 0. 3- Samples were taken at 4 representative plots, two pits per plot. Field sampling and laboratory procedures are described by Garey (1957) and Hoover, st 21. (1954). Bulk density values and supplementary data for the 4 plots are shown in Appendix 4. The procedure for computing aver- age bulk density values by 6-inch depth intervals for each plot is dis— cussed in Appendix 4a. Soil Depth (d). Average soil depth to parent rock for an entire plot was computed from soil moisture samples taken with the Veihmeyer tube. Only samples which gave every indication of striking parent rock were used. Even these can be misleading, so the sampling error would tend to produce an average plot depth less than actual. Appendix 5 shows the number of samples, range of depth, mean depth, confidence limits and standard error for each plot. 94 Figure 27. Soil sampling with a Veihmeyer tube in a dry clay soil using a soil jack. Separate tube points were used depending on soil moisture, depth and stoni- ness. When the soils were dry, as they usually were, the tube could be driven about 1 inch for every blow with the 15-pound driving hammer. 95 Inches of Water. An actual example of the inches of soil water computed for plot 2 Medium density on April 22, 1958, is shown below: Average Average Average Depth moisture Bulk Density Mbisture % in. O— 6 25.6 1.15 1.77 6-12 34.5 1.23 2.55 12-18 31.7 1.28 2.43 18-24 31.8 1.32 2.52 24-30 32.3 1.35 2.62 30-32.6 33.5 1.37 1.19 13 08 Total Appendix 6 shows inches of soil water and differences in inches of soil water by sampling date, by plot and by 6-inch depth interval. Results. Relation of Stand Density to later Consumption. An attempt was made to relate some stand density measure to soil moisture use. The logical periods between sampling dates to compare were those in which water was readily available to the plants. It is generally known that when soil moisture closely approaches wilting point further withdrawal is exceedingly slow, and differences in soil moisture use due to dif- ferences in plant density are not apt to be measurable. Determination of the point at which soil moisture depletion rate decreases markedly was of primary concern. The period from April 24, 1958, to July 23, 1958, was selected as suitable for studying depletion rate. Only 0.60 inch of rain fell during the 89 days of the test per- iod. Only plots having an average depth of over 30 inches were se- lected. Average percent moisture was first computed by 6-inch intervals for sampling dates 4, 5, 6, and 7. Six plots, 1, 2 and 3 Medium and 96 Heavy density, were used for depth intervals from 6 to 30 inches. Plots 1 Medium and Heavy density were rejected for the O-to 6vinch interval because they showed excessive sampling variation. Values are shown in Table 11. These values were then plotted to obtain a field estimate of wilting points. Table 11 also shows average percent mois- ture at fiéld capacity, average bulk density and average available water-holding capacity by 6—inch intervals. Using the estimated wilting points, average inches of available water was then computed for each of the four sampling dates. Average percent available water of average total available water-holding capa— city was computed for each sampling date by 6-inch intervals and plotted. It was observed that the rate of available water withdrawal is nearly constant from about 65 percent down to about 5 percent, and nearly equal for all 6—inch intervals. The shape of the depletion curve is generally in agreement with Penman's concept (Smith, 1959). The selection of periods which offered the best chance for mea— suring differences in soil moisture use between stand density classes presented some difficulty. The winter periods were rejected because of the traditionally low evapotranspiration rates. Differences would be too small. The September and October periods were rejected because they were periods of soil moisture accretion. The periods from April 24 to June 20, 1958, and March 3 to April 23, 1959, seemed ideal because: available soil moisture was above 5 percent of total; no surface runoff occurred to complicate matters; precipitation was scanty; no precipita- tion occurred at least 10 days before the sampling dates, thereby 97 o mH nH mm mm on em ontvm o NH NH mm hm mm mm vmlmH o mH mH am pm mm mm mHINH o oH oH mm mm pm on «Him 0 H H mm an em mm mno mHsh mosoammmHn mash mocoammmHn has ooqouommHQ .aq< huHoaamo mafiuHomlhoaaB oHanHw>< Hence no noun; oHanHo>< accused ommao>< oo.o mH.o we.o mb.o omnvm oo.o mH.o om.o mw.o vNumH oo.o mm.o mm.o mm.c mHINH oo.o «H.o mv.o mm.o «Hum oo.o mo.o mm.o mo.H one mHsh 0:35 Nmm. .aQ< nouns oHanHa>< monocH mwmuo>< m mH m.HN «.mm o.mm NN.H mm.H m.vm m.mH omnvm v.mH m.HN m.mm H.om mm.H hm.H m.mm v.mH vmlmH b.wH v.HN «.mm N.om N¢.H mm.H m.mm b.mH mHnNH «.mH o.om N.vm o.Hm ov.H mN.H v.0m N.mH «Hum m.HH «.mH N.om o.bm vo.m mH.H ®.ov w.HH wuo .Illllllldcmouom monocH unwound monocH mHSh mash has .AQ< mpHomqmo muHoaan psHom 38 @233 333733: .33an 3on 33H; ouaumHoz oHanHw>< tzm an an spawn owmuo>< Hmuoe mmmso>< mwoao>< onaumHoS ousumHoS owoao>< mmasm>< .o>p=o coHuoHamo one .HH oHnt minimizing sampling variation; because of the small amount of precipitation. 98 interception effects were also minimized In the first analysis, only the Medium and Heavy density plots were compared, appreciable soil depth. since there remained but one Light density plot with any Analysis of variance showed no significant difference between the Medium and Heavy density plots for either period. The following figures indicate the size of the differences: Depth Period inches 0-30 4/24 to 6/20/58 0-30 3/3 to 4/23/59 Total Average Inches Water Days Used Per Plot Medium Heavy 58 4.02 3.81 48 1.39 1.31 fiié 5.41 5.12 Average Inches Water Used Daily Per Plot Medium Heavy 0.069 0.066 0.029 0.027 0.051 0.048 In the second analysis a comparison was made of inches of soil moisture withdrawal in 24 inches of soil for the same sampling periods. In this instance, only single plots at replication 3 were compared. These plots are located within 100 yards of each other; their soil properties are nearly identical; and they all received the same amount of precipitation. Values for the 18 to 21.1 inch depth interval on plot 3 Light density were extended proportionately to 24 inches. Re- sults are shown below: Average Inches Depth Period Days Water Withdrawn inches Light Medium Heavy 0-24 4/24 to 6/20/58 58 2.98 3.19 3.23 0-24 3/3 to 4/23/59 48 0.99 0.99 0.99 Total 106 3.97 4 18 4.22 Average Daily 0.037 0.039 0.040 It is apparent that total, period and daily differences are so small as to be negligible, and may easily be attributed to sampling variation. 99 Apparently evaporation alone during the April to June and March to April periods was able to keep pace with combined evaporation and transpiration of soil moisture in 24 to 30 inches of soil. Soil crack- ing would certainly contribute to making such a supposition possible. Evapotranspiration of Soil Moisture. Evapotranspiration of soil moisture was computed from the formula for the "water balance" method. Values are summarized in Appendix 7 and average values for all plots shown in Figure 28. Inspection of the data showed that values ranged from 12.74 to 18.17 inches for the year. Since these differences existed, the fac- tors responsible for them were studied. The largest and most impor— tant differences in factors associated with differences in evapotrans- piration from the soil were stand density, net precipitation and average soil depth. Interception differences were shown previously to be a function of stand density and differences in net change in soil moisture storage are shown to be largely a function of average soil depth by the following regression: Y = 0.22 / .093X The regression equation is significant to 1 percent; the regression coefficient significant to 0.5 percent; and the standard error of re- gression 0.62 inch net soil moisture change. There were obvious dif- ferences in surface water movement and percolation between the plots, but these differences were both small and, in a sense, artificial, hence were not included in the analysis. The selected factors presented a rather puzzling picture at first because they were interrelated. The following correlation Eggclu “33.89885; £3893: .1..\III\ §§Bg< 0— 30008.— EEBEEEBEHBEHEEBUEEE in: «tom 101 coefficients suggest the nature of the complexity. Values used for evapotranspiration from the soil and net precipitation were one year totals. x1 Variable x2 Variable "r" Value Et(s) %CD-D .5350 Et(s) Ave. soil depth .8447 Et(s) Net precipitation -.2417 %CD—D Ave. soil depth .8100 %CD—D Net precipitation -.8067 Ave. soil depth Net precipitation -.6548 In order to test the independent effect of percent crown density using the spherical densiometer (%CD—D) average soil depth and net precipitation on evapotranspiration from the soil, a multiple regres- sion was computed with evapotranspiration from the soil as the depen- dent variable. The following regression coefficients were obtained: '1 b" X1 %CD-D -.00105 X2 Ave. soil depth .13655 X3 Net precipitation .77485 A "t" test showed that the regression coefficient for percent crown density by the spherical densiometer (%CD—D) was not significant. This result is not surprising in view of the fact that only 15 inches precipitation occurred during the year of study. Little chance was given for an expression of soil moisture demands by different stand densities. Indeed, evaporation alone from bare soil could well exceed 15 inches. 102 A new multiple regression, using only average soil depth and net precipitation, was then computed with evapotranspiration from the soil again as the dependent variable: Q = 1.15 / .1360X1 / .7819X2 The standard error for the whole regression was 0.68 inch, with an "F" value of 27.8, significant at 1 percent. The "t" values for regres— and b sion coefficients b 2 were 7.23 and 3.28, significant at l per— 1 cent and 5 percent respectively. Partial correlation coefficients were then computed between aver- age soil depth and evapotranspiration from the soil, holding net precipitation constant, and between net precipitation and evapotrans- piration from the soil, holding average soil depth constant. The par- tial correlation coefficients were .9358 and .7698 respectively, both significant at 1 percent. It appears then that average soil depth is a better predictor of evapotranspiration from the soil than net precipitation. This conclu- sion makes sense when the effect of average soil depth on net precipi- tation is considered. As shown previously, the "r" value for percent crown density by the spherical densiometer (%CD-D) and average soil depth is .8100; e.g., the deeper the soil, the greater the amount of crown volume. And, the greater the volume of crown, the less the net precipitation - "r" equals -.8067. The indirect effect of average soil depth on net precipitation is given by the "r" value of —.6548. Thus, net precipitation is dependently related to average soil depth. It should be stressed immediately that the conclusion, "Average soil depth is a better predictor of evapotranspiration from the soil 103 than net precipitation.", applies only to situations in which other factors are held nearly constant between sites, and in which full ex- pression of soil moisture use by different stand densities is re- stricted. It is quite likely that the above conclusion applies to total precipitation below 15 inches, since soil moisture use according to stand density would probably be restricted, and since differences in net precipitation between sites would be correspondingly smaller and have less effect on evapotranspiration from the soil. It also seems logical that the relationship between soil depth and evapotranspiration from the soil depends to a great extent on dif— ferences in net soil moisture storage between the first and last deter- minations - the greater the difference, the greater the effect. However, given the situation where first and last readings were equal, and all differences in evapotranspiration from the soil due to differ- ences in net precipitation, soil depth would still be important due to its indirect effect on net precipitation. The Check Plot. Average inches of water in 30 inches of soil was computed for the Medium and Heavy density and check plots at replication 3. Initial inches of water on June 20, 1958 in the Medium and Heavy density plots was adjusted to that of the check plot. The check plot was cleared in mid-June when nearly all soils were slightly above wilting point. By April 23, 1959 an average difference of 3.40 inches was measured be- tween the check and Medium and Heavy density plots. This difference very likely would have been accentuated had not the winter period 104 between November and March been very dry. Less than 4.00 inches of precipitation occurred between November 1 and March 31, the usual period of soil moisture accretion. Soil moisture samples were taken at replication 3 on August 11, 1959 to see if the difference in water storage remained. Approximately 4.00 inches of precipitation occurred between the April and August sampling periods. The check plot had a thin cover of herbaceous vege- tation at the time of the August sampling. Only a difference of about 0.30 inch of soil moisture separated the check plot from an average of the Medium and Heavy density plots on August 11. Soil cracking was observed to begin in early May, and it is quite possible that evapora— tion alone could account for the disappearance of soil moisture at depths of 30 inches. l1..ll.l!.. u. b. SUMMARY AND CONCLUSIONS The demand for increased water supplies for the rapidly expanding population centers in the arid valleys of Arizona is intense. One possibility for increasing water supplies involves management of the upland watersheds which contribute surface water to the arid valley reservoirs. Federal money was appropriated for both a pilot project and a research program on these upland watersheds. This study is a part of the research program, conducted by the Rocky Mountain Forest and Range Experiment Station of the U.S. Forest Service. It is designed to obtain basic information relating to water- shed management, with particular reference to the Beaver Creek Water- shed Project. The study is exploratory in nature, since so little was known of the hydrology of the area. Results will furnish a portion of the information needed to define the area problems and to orient the direction of continuing research. Objectives were to make a prelimin- ary determination of the disposition of the precipitation that falls in Utah juniper stands growing within a fairly average range of sites on the Beaver Creek Drainage, and to explore some of the factors which control the disposition process. Secondary objectives were to provide complementary information regarding soil properties and root charact- eristics of Utah juniper. General. The study area lies on the Mogollon Rim at an elevation of about 5,500 feet. Topography is rolling to hilly with occasional steep-walled 106 canyons cut by ephemeral streams. The underlying rock formations are composed of heterogenous volcanics, ranging from red, porous cinders to bluish, dense basalt. Nine study plots were located adjacent to three watersheds, 250, 310, and 330 acres in size, now being calibrated in the Utah juniper type on the Beaver Creek Watershed. Datawmae collected for one year beginning April 24, 1958. A check plot was also established and data collected beginning June 20, 1958. The plots and small watersheds are enclosed by a network of pre- cipitation gages established in 1957. Records from these gages show that most precipitation occurs as high intensity, short duration sum- mer storms and low intensity winter storms of one to three days. These records agree with general findings throughout Northern Arizona. Hygrothermograph records within the same network show an annual aver- age temperature of abOut 55 degrees, with extremes of 6 and 104 de— grees.. Relative humidities of 5 to 15 percent during the usually dry spring and fall are not uncommon. The storm distribution during the period of study was very uneven. Vegetation on the plots is preponderantly Utah juniper (gunigerus Utahensis (Engelm.) LemmJ, with occasional clumps of turbinella oak (Quercus turbinella Greene) and individual pinon pines (Pinus edulis Engelm.). Ground vegetation is sparse; rock. bare soil and juniper leaf litter making up about 35, 50 and 15 percent of the surface cover respectively. Slopes varied from 2 to 7 percent. 107 Soils. The soils on all plots were identified as belonging to the Spring- erville series. The series description is given for an eastern part of Arizona where the average soil depth ranges from 30 to 50 inches. Soils on the experimental plots averaged slightly over 30 inches, but a range of 6 to 94 inches was encountered from individual soil moisture samples. The soils on the plots contain about 65 percent montmorillonitic-type clay; wilting point is about 19 percent and field capacity about 37 percent soil moisture; bulk density averages about 1.33 and total available water-holding capacity is about 2.50 inches per foot of soil; average pH is about 7.5 and color of the subsoil is reddish brown (5YR 4/3). These soils are very hard when dry and very plastic and sticky when moist. The montmorillonitic-type clay, with its 2:1 lattice structure, produces a pronounced swelling and shrinking of the soil body with changes in soil moisture. The periodic swelling and shrinking leads to a self-swallowing action, thought to be responsible for the concen- tration of rock in the upper foot of soil and on the surface. The shrinking action can also produce large cracks which, beginning to show in late spring, can occur to the depth of the soil body. These cracks appear to be responsible for greatly increasing the infiltra- tion capacity of the soils. Roots. A study of the rooting habits of Utah juniper indicates that the great bulk of roots is located within the upper three feet of soil. 108 Some roots, however, are known to occur at depths of at least fifteen feet. Large roots are rather uniformly distributed, while a concentra- tion of fine roots occurs in the upper 18 inches of soil. Stand Density. In order to relate the disposition of precipitation to stand den- sity, some measure of stand density was required. The stands on the nine plots were measured by eight different methods, relating primarily to crown volume and number of trees. A strong correlation was shown to exist between percent crown density as measured by a spherical densio— meter and all other methods. This strong correlation eliminated the need to test all stand density measures against the factors in the dis- position process. Surface Runoff, Depression Storage, Subsurface Flow, and Percolation. Depression storage, subsurface flow and percolation were found to be negligible factors. Due to the limited data, an intensive analysis of the surface runoff component was not warranted. The plot data do show that a disproportionately large amount of surface water movement occurred on plots having shallow soils, about 6 inches deep, fairly steep slopes, 5 to 6 percent, and scanty vegetation. Other data suggest that surface water movement seldom occurs where juniper leaf litter covers the ground and that it increases with an increase in percent rock cover on the surface. Streamflow records from the three, small watersheds averaged about 0.22 inch surface runoff for the period of study. During a three-week period over 5.50 inches or precipitation occurred, including two fairly intense storms of about 1.00 and 3.50 inches. This period produced the 0.22 inch surface runoff, indicating that the soil cracks, apparent at 109 the time, were responsible for increasing the infiltration capacity on the seemingly impermeable soils. Stemflow. Stemflow contributes a relatively small amount of water to the soil, about 2 percent of total precipitation. It begins to occur when rain storms exceed 0.25 to 0.30 inch. The shaggy bark of Utah juniper is probably responsible for these comparatively high values. No stem- flow was recorded after snow storms, even those exceeding 1.00 inch water equivalent. Most likely the increased opportunity for evapora- tion and the fact that some snow is simply blown off the trees accounts for this phenomenon. Interception. During the course of 34 storms, over 1,200 randomized measure- ments were taken of net precipitation. Results showed that intercep— tion accounted for an average of 15 percent of gross precipitation. This amount varied from 2 to 24 percent, depending on stand density. Of the eight stand density measures used to describe the plots, percent crown density using the spherical densiometer showed the best relationship with total percent interception. For practical field purposes, however, a simple count of the number of trees higher than 10 feet appears to be as satisfactory for predicting interception. Regression analysis showed that the per storm relationship be- tween gross and net precipitation was linear. Covariance analysis of these regressions showed significant differences in net precipitation between averages of the Light, Medium and Heavy density plots. 110 A more intensive analysis showed that net precipitation per storm was both a direct, linear function of gross precipitation and an in- verse function of the square of stand density as measured by the spherical densiometer. Gross precipitation was a much more important factor in predicting net precipitation. Soil Moisture and Stand Density. The inches of water stored in the soil was determined by monthly periods on the nine plots from April, 1958 to April, 1959 and on the check plot from June, 1958 to April, 1959. Differences in inches of soil water storage were computed between months by 6-inch intervals for each plot. Several attempts were made to relate soil moisture use to stand density. Since differences in soil moisture use would be masked if soil moisture was near wilting point, the shape of the depletion curve was studied. Soil moisture at each sampling date was expressed as per- cent available water of total available water—holding capacity. It was shown that from about 65 percent down to 5 percent the rate of avail- able water withdrawal was nearly constant and was nearly equal for each 6-inch depth interval down to 30 inches. Two drying periods, when soil moisture was available above 5 per- cent and other conditions were favorable, were selected for their ability to show differences in soil moisture use between different stand densities. These periods covered a total of 106 days. Average inches of soil moisture withdrawn daily from 30 inches of soil showed no real difference between Medium and Heavy density plots during either lll period. A similar comparison was made for the Light, Medium and Heavy density plots at replication 3 where all other site factors were nearly identical. No differences in soil moisture use in 24 inches of soil were noted during either period. Evapotranspiration of Soil Moisture. Evapotranspiration of soil moisture was computed for each plot from the "water balance" formula. Yearly values ranged from 12.74 to 18.17 inches. To explain the differences between plots, the effects of stand density, net precipitation and soil depth were studied. These factors were mutually interrelated, but when their independent effects on evapotranspiration of soil moisture were identified, it was found that differences in stand density had little effect, while differences in soil depth were more important than differences in net precipitation. Soil depth was important because it reflected the comparatively large differences in net change in soil moisture storage and because it indirectly influenced net precipitation through its influence on stand density. Stand density was not important because prolonged periods of little available water and cold weather restricted its full expression. The Check Plot. At the end of a lO—month period the check plot contained about 3.40 more inches of soil moisture in 30 inches of soil than the Medium and Heavy density plots at replication 3. By August, 1959, however, this difference had almost completely disappeared. About 4 inches of rain had fallen from April to August, 1959. LITERATURE CITED Anonymous. 1959 Instruction book - model 2800 portable scalar. Nuclear- Chicago Corp. Chicago, Illinois. 23 pp. Arnold, J. F. and W. L. Schroeder 1955 Juniper control increases forage production on the Fort Apache Indian Reservation. Rky. Mt. For. & Range Exp. Sta. Station Paper No. 18. 35 pp. Barr, G. W., 3: El. 1956 Recovering rainfall. Dept. Agr. Econ., Univ. of Ariz. Tucson. 218 pp. Baver, L. D. 1956 Soil physics. 3rd ed. John Wiley and Sons, Inc. New York. 489 pp. Berger-Landefeldt, U. 1956 Determination of evapotranspiration by the exchange method. Processed paper. Files of the Rky. Mt. For. & Range Exp. Sta., Flagstaff Center. Blaney, H. F. 1952 Consumptive use of water. Trans. Amer. Soc. Civ. Engin. 117: 949-973. , C. A. Taylor, and A. A. Young 1930 Rainfall penetration and consumptive use of water in the Santa Ana River Valley and Coastal Plain. Cal. Dept. Pub. Works. Div. Water Resources Bull. 33. 162 pp. Bonner, J. 1959 Water transport. Sci. 129: 447—450.. Bradshaw, K. R. and J. L. Reveal 1943 Tree classification for Pinus monophylla and Juniperus Utahensis. Jour. For. 41(2): 100-104. Breazeale, J. F. and F. J. Crider 1934. Plant association and survival, and the build-up of moisture in semi-arid soils. Univ. Ariz. Agr. Exp. Sta. Tech. Bull. 53. Tucson. Burger, H. 1929a Wald und Wasserhaushalt. Schweiz. Ztschr. f. Forstw. 80: 38-44. (Eng. Trans., Forests and the water regime. U.S. For. Sen, Div. of Silvics, Trans. No. 102, by Albin Mier, Jan. 1935) 113 1929b Einfluss des Waldes auf den Wasserabfluss bie Landregen. Schweiz. Ztschr. f. Forstw. 80: 196-199. (Eng. Trans., In— fluence of the forest on the runoff of general rains. U.S. For. Ser., Div. of Silvics, Trans. No. 101, by Albin Mier, Jan. 1935) 1933 Waldklimaflagen, Schwein. Centralsaat f. Forstl. Verschsw. Mitt. 18: 1-192. Cable, D. R. 1958 Estimating surface area of ponderosa pine foliage in Central Arizona. For. Sci. 4(1): 45-49. Canfield, R. H. 1942 Sampling ranges by the line interception method. Southwest- ern For. and Range Exp. Sta., Processed Paper. 28 pp. Cole, D. W. 1957 An alundum tension lysimeter. Contribution from the Soils Dept., Wis. Agr. Exp. Sta., Madison.. Processed Paper in the files of the Rky. Mt. For. and Range Exp. Sta., Flagstaff Center. ' Colman, E. A. 1953 Vegetation and watershed management. Ronald Press. New York. 412 pp. and E. L. Hamilton 1947 The San Dimas lysimeters. Cal. For. & Range Exp. Sta. Res. Note 47. 33 pp. Cooper, C. F. 1957 The variable plot method for estimating shrub density. Jour. Range Mgmt. 10(3): 111-116. Daniels, F. and R. A. Alberty 1956 Physical chemistry. 2nd ed. John Wiley and Sons, Inc. New York. 671 pp. Dortignac, E. J. 1956 Water yields through watershed management. N. Mex. Water Conf. Proc. 1956: 69-97. Evans, G. C. and D. E. Coombe 1959 Hemispherical and woodland canopy photography and the light climate. Jour. Ecol. 47(1): 103-113. 114 Furr, J. R. and J. O. Reeve 1945 Range of soil-moisture percentages through which plants undergo permanent wilting in some soils from semi-arid irri- gated areas. Jour. Agr. Res. 71(4): 149-170. Gary, C. L. 1957 A field technique for bulk density sampling of soils. Univ. of Ark. Agr. Exp. Sta. Bull. 582. 10 pp. Goodell, B. C. 1952 Watershed-management aspect of thinned young lodgepole pine stands. Jour. For. 50(5): 374-378. Grah, R. F. and C. C. Wilson 1944 Some components of rainfall interception. Jour. For. 42(12): 890-893. Halstead, M. H. 1954 Final report. Micrometeorology of the surface layer of the atmosphere. The flux of momentum, heat and water vapor. The Johns Hopkins Univ. Lab. of Climatology VIII: 233-361. . and W. Covey 1957 Some meteorological aspects of evapotranspiration. Soil Sci. Soc. Amer. Proc. 21(5): 461-464. Haltiner, G. J. and F. L. Martin 1957 Dynamical and physical meteorology. McGraw-Hill, Inc. New York. 470 pp. Hamilton, E. L. 1954 Rainfall sampling on rugged terrain. U.S. Dept. Agr. Tech. Bull. 1096. 44 pp. and P. B. Rowe 1954 Rainfall interception by chaparral in California. Cal. Dept. Natural Resources, Div. of Forestry. 43 pp. Harrold, L. L. and F. R. Dreibelbis 1955 Evaluation of agricultural hydrology by monolith lysimeters. U. S. Dept. Agr. in cooperation with Ohio Agr. Exp. Sta. Tech. Bull. 1179. 166 pp. Herman, F. R. 1953 A growth record of Utah juniper in Arizona. Jour. For. 31(3): 200-201. Hoover, M. D. 1944 Effect of removal of forest vegetation upon water yields. Trans. Am. Geophys. Union 26: 969-975. 115 . D. F. Olson, and L. J. Metz 1954 Soil sampling for pore space and percolation. Southeastern For. Exp. Sta. Station Paper No. 42. 28 pp. Horton, R. R. 1919 Rainfall interception. U.S. Mo. Wea. Rev. 47: 603-633. Howell, J. Jr. 1941 Pinon and juniper woodlands of the Southwest. Jour. For. 39(6): 343-344. Johnson, E. A. and R. E. Dils 1956 Outline for compiling precipitation, runoff and ground water data from small watersheds. Southeastern For. Exp. Sta. Station Paper No. 68. 40 pp. Johnson, J. C. 1954 Physical meteorology. Pub. jointly by Technology Press of Mass. Insti. of Tech. and John Wiley and Sons, Inc. New York. 393 pp. Johnson, I. W. 1942 The interception of rain and snow by a forest of young pon- derosa pine. Trans. Am. Geophys. Union: 566-570. v.Karman, T. 1930 Nachr. Ges. Wiss. Gottingen, Mathphysik. Klasse. (Cited from Sutton, 1953) Kittredge, J. 1948 Forest influences. McGraw-Hill, Inc. New York. 394 pp. , R. J. Loughead and A. Mazurak 1941 Interception and stemflow in a pine plantation. Jour. For. 30(6): 505-522. Kohler, M. A. 1957 Computation of evaporation and evapotranspiration from meteorological observations. U.S. Dept. Commerce. Wea. Bur. Prepared for A.M.S. meeting in Chicago, Mar. 19-21, 1957. Processed paper. 10 pp. Kohnke, H., F. R. Dreibelbis, and J. M. Davidson 1940 A survey and discussion of lysimeters and a bibliography on their construction and performance. U.S. Dept. Agr. Misc. Pub. 372. 68 pp. Kolmogoroff, A. N. 1941 Comptn. red. acad. sci. U.S.S.R. 30: 301; 32: 16. (Cited from Sutton, 1953) Kramer, P. J. 1946 Absorption of water through suberized roots of trees. Plant Physiol. 21(1): 37-41. Landsberg, H. E. and M. L. Blanc 1958 Interaction of soil and weather. Soil Sci. Soc. Amer. Proc. 22(6): 491-495. Lemmon, P. E. 1956 A spherical densiometer for estimating forest overstory den- sity. For. Sci. 2(5): 314-320. Lull, H. W. and J. H. Axley 1958 Forest soil-moisture relatnxusin the Coastal Plain sands of Southern New Jersey. For. Sci. 4(1): 2-19. Mather, J. R. 1954 The measurement of potential evapotranspiration. Johns Hopkins Univ. Lab. of Climatology. Publications in Climato- logy VII(1): 1-232. McDonald, James E. 1956 Variability of precipitation in an arid region: a survey of characteristics for Arizona. Univ. of Ariz. Institute of Atmospheric Physics. No. l. 88 pp. McGinnies, W. G. and J. F. Arnold 1939 Relative water requirements of Arizona range plants. Ariz. Agr. Exp. Sta. Tech. Bull. 80. 240 pp. McIlroy, I. C. 1957 The measurement of natural evaporation. Jour. Austral. Instit. Agr. Sci. 23(1): 4-17. Meyer, B. S. and D. B. Anderson 1954 Plant physiology. 2nd ed. D. van Nostrand Co., Inc. New York. 784 pp. Milthorpe, F. L. and E. J. Spencer 1957 Experimental studies of the factors controlling transpira- tion. Jour. Exp. Bot. 8(24): 413-437. Moyle, R. C. and R. Zahner 1954 Soil moisture as affected by stand conditions. Southern For. Exp. Sta. Occ. Paper 137. 14 pp. Nash, J. E. 1958 Determining runoff from rainfall. Pub. of the Instit. of Civil Engineers. Great George St. Westminster, London. 117 Pasquill, F. 1949 Eddy diffusion of water vapour and heat near the ground. Proc. Roy. Soc. Series A 198(1052): 116-140. Penman, H. L. 1949 Natural evaporation from open water, bare soil and grass. Pub. of Physics Dept. Rothamsted Exp. Sta. Harpenden, Herts., England. 1951 The dependence of transpiration on weather and soil condi- ' tions. Jour. Soil Sci. 1-2: 74-89. Philip, J. R. 1958 Plant Physiol. 33(264). (Cited from Bonner, 1959) Prandtl, L. 1934 The mechanics of viscous fluids. W. F. Durand (ed.). Aerodynamic Theory. Vol. III, Division G (Berlin). (Cited from Sutton, 1953) Price, R. 1958 Watershed management research in the Southwest (A literature review). Presented at the 83rd Ann. Meeting of the Amer. For. Assn., Tucson, Ariz. Priestley, C. H. B. and W. C. Swinbank 1947 Proc. Roy. Soc. A. 189, 543. (Cited from Pasquill, 1949) Reveal, J. L. 1946 Single-leaf pinon and Utah juniper woodlands of western Nevada. Jour. For. 42(4): 276-278. Rich, L. R. 1952 Symposium--Consumptive use of water: forest and range vege- tation. Amer. Soc. Civ. Engin. Trans. 117: 974-990. Richards, L. A. and C. H. Wadleigh 1952 Soil physical conditions and plant growth: Agronomy 11. Byron Shaw ed. Academic Press, Inc. New York. pp. 74-254. Roeser, J. Jr. 1949 The water requirements of Rocky Mountain conifers. Jour. For. 38(1): 24-26. Smith, G. W. 1959 The determination of soil moisture under a permanent grass cover. Jour. Geophys. Res. 64(4): 477-483. Spurr, S. H. 1948 Aerial photographs in forestry. Ronald Press Co. New York. 340 pp. 118 Stephenson, R. E. 1935 Root penetration in relation to soil aeration. Report of 50th Ann. Meeting Oregon State Hort. Soc.: 19-34. Suomi, V. E. and P. M. Kuhn 1958 An economical net radiometer. Tellus 10(1): 160-163. (Stockholm) and C. B. Tanner 1958 Evapotranspiration estimates from heat-budget measurements over a field crop. Trans. Am. Geophys. Union 39(2): 298-304. Sutton, 0. G. 1953 Micrometeorology. A study of physical processes in the low- est layers of the earth's atmosphere. McGraw-Hill. New York. 333 pp. Tanner, C. B. 1957 Factors affecting evaporation from plants and soils. Jour. Soil and Water Conservation 12(5): 221-227. Taylor, G. I. 1931 Rapp. Cons. Explor. Mer. 76, 35. (Cited from Pasquill, 1949) 1932 Proc. Roy. Soc, (London), A 135, 685. (Cited from Sutton, 1953) Texas, the A & M College of 1957 Forecasting micrometeorological variables. Dept. of Oceano- graphy and Meteorology. A & M Project 93. 276 pp. Thornthwaite, C. W. and J. R. Mather 1955 The water balance. Drexel Instit. of Tech. Lab. of Climato- logy. Pub. in Climatology VII (1): 1-104. and 1957 Instructions and tables for computing potential evapotrans- piration and the water balance. Drexel Instit. of Tech. Lab of Climatology. Pub. in Climatology X(3): 185-311. U.S. Dept. Agr.. 1953 The Sierra Ancha experimental watersheds. Southwestern For. and Range Exp. Sta. 31 pp. U.S. Dept. Commerce 1957 Statistical abstracts of the United States. Bur. of the Census. U.S. Dept. Commerce 1956 Climatological data of Arizona, annual summary. Wea. Bur. U.S. Dept. Commerce 1958 Climatological data of Arizona, annual summary. Wea. Bur. 119 U.S. Dept. Interior 1953 Water supply paper 1313 and subsequent annual papers. U.S. Geol. Sur. van der Bijl, W. 1957 The evapotranspiration problem. First contribution. Kansas State College. Dept. of Physics. Report 1956-1957. 94 pp. 1958 The evapotranspiration problem. Second contribution. Kansas State College. Dept. of Physics. Report 1957-1958. 82 pp. Veihmeyer, F. J. and A. H. Hendrickson 1955 Does transpiration decrease as the soil moisture decreases? Trans. Am. Geophys. Union 36(3): 425-448. Yamaoka, Y. 1958 The total transpiration from a forest. Trans. Am. Geophys. Union 63(7): 266-272. Wicht, C. L. and D. E. W. Schumann 1957 Experimental investigations of the effects of forests on stream-discharge. Commonwealth Forestry Conference. Australia and New Zealand. 12 pp. Wilm, and Niederhof 1941 Interception of rainfall by mature lodgepole pine. Trans. Am. Geophys. Union, Part III: 660-666. Zon, R. 1927 Forests and water in the light of scientific investigations. U.S. For. Ser. (Reprinted with revised bibliography, from Appendix V of the Final Report of the National Waterways Com— mission, 1912. Senate Doc. No. 469, 62nd Congress, 2nd Session) 106 pp. 120 Appendix 1. Methods of Determining Rates of Evapotranspiration Vapor Flow Methods. A. The Dalton Equation. The first method of determining evapotranspiration by vapor flow ' methods makes use of the basic formula: E = a(1 + bu) (eS - ea) E is the rate of evaporation. a and b are empirical constants. u is windspeed at a given height. es and ea are vapor pressures at the surface and in the bulk air re- spectively. According to McIlroy (1957), "Almost all of the work done in this connection has been on evaporation from water surfaces. This permits eS to be taken ases , the saturated vapor pressure at the surface tem- perature. He then discusses Rohwer's studies, where the constant for a becomes 0.4; the constant for b becomes 0.27; u is surface wind in mph; and eS and ea are expressed in mm/day. Again quoting McIlroy, "Visentini [1936] and others have preferred to use in place of the ground-to-air vapor pressure drop the more easily measured quantityéa — ea, or saturation deficit (s.d.). Hereéa is the saturated vapor pressure at the bulk air dry bulb temperature Ta- This is based on the erroneous belief that the evaporative power of the air is dependent on its 'absolute' dryness as expressed by s.d., rather than its 'relative' dryness as indicated by the va- por pressure difference between ground and air. Although it has no physical justification the use of s. d. as an in- dex of evaporation has apparently given moderately success- ful results in certain applications, .... 121 Van der Bijl (1956-1957) then reviews the work of Albrecht, Haude, Kalweit, Ivanov,_Skvortsov and Poliakov, all of whom apply some modifi- cation of the saturation deficit approach to geographical areas with varying degrees of success. B. The Aerodynamic Approach. The second method of determining evapotranspiration by vapor flow methods is called the "aerodynamic approach." For a more complete theoretical background, the reader is referred to Sutton (1953) and Haltiner, st 31. (1957), since only the main outlines will be presented here. The fundamental problem involves a determination of the turbul- ent transfer of water vapor at a small distance above the ground. Dalton's empirical constant, a, is replaced by an "eddy transfer coef- ficient" and the ground-to-air vapor pressure drop by the specific humidity gradient. To understand why this method is in better accord with physical principles of evapotranspiration, consider the following description of a field situation by Sutton: "To fix ideas, consider conditions near level ground, either bare or covered with short grass, in the afternoon and evening of a clear day in early summer or in the fall, in the circumstances in which the synoptic chart shows only small horizontal gradients of pressure. When the sun is high in the heavens, the temperature of the surface rises considerably above that of the air immediately adjacent to it and the wind is highly turbulent. As the sun sets, the temperature of the ground falls rapidly below that of the air, with the result that the layers of the atmosphere im- mediately in contact with the ground are chilled and become denser than those above. The maintenance of the turbulent state implies that masses of air are being removed continu- ally in the vertical, so that if the fall of density with height is very pronounced, considerable work has to be done in lifting the denser masses against the gravitational field, at the expense of energy of the mean motion. The inevitable result is that, in such circumstances, the 122 turbulent motion becomes less pronounced and may even die away completely. This, in turn, means that the supply of mo- mentum from the free stream to replace that absorbed by the friction of the ground is reduced because of the loss of mix- ing, and the flow, as a whole, tends to settle down to slow motion in parallel layers. Provided that the sky remains clear and that there are no large horizontal pressure gradi- ents, this state of affairs will generally continue until dawn, when the incoming radiation raises the temperature of the ground, and ultimately that of the lowest air layers. The position is now reversed, for the less dense air is be- low, and any tendency to vertical motion is enhanced by the prevailing density distribution. Soon after dawn the near- laminar flow gives place to turbulent motion, which continues throughout the day." ‘ The specific humidity gradient can be measured directly by suit- able hygrometers. The difficulty comes in determining the "eddy trans- fer coefficient." To get the meaning of this term and the theory of measurement, we again quote Sutton: "....much of the early work, and especially that on at- mospheric turbulence, was based on a 'kinetic-theory' model, in which wandering masses of fluid, called eddies, were sup- posed to behave like molecules, ”A first natural step toward a theory of turbulence is to adopt the fundamental ideas of the kinetic theory of gases by expressing the transfer of momentum, or any suit- able, conservative entity, by means of virtual coefficients of viscosity, conductivity, and diffusivity, defined in much the same way as their molecular counterparts." An expression of turbulent flux, independent of any theory of the structure of eddy motion is shown to be: "Mean flux = k ii - pE'w' (3.11) dz k is the appropriate molecular coefficient (viscosity, conductivity, diffusivity. E is the amount per unit mass of fluid cufany transferrable conserva- tive entity. 2 is height, or direction perpendicular to the xy plane. pE'w' is the amount of E transported in unit time through unit cross section of a plane parallel to z = 0. 123 "Thus the mean flux across a plane perpendicular to the z direction depends chiefly upon the existence of a correla- tion between the fluctuations in velocity and in the entity being transferred. If momentum is being transferred, E = u, k =,q, and thus Mean flux :4 fig - pu'w' dz in which the second term will be recognized as the appropri- ate Reynolds stress, the mean flux of momentum being the fric- tional force per unit area, or shearing stress. Matter in suspension, such as water vapor, smoke or dust, implies that E be expressed as grams of matter per gram of air - thus for the diffusion of water vapor E must be the specific humidity, or the ratio of absolute humidity to the density of the (moist) air, so that pE is the absolute humidity, or concen- tration of water vapor (mass per unit volume). "The fundamental problem in the analysis of turbulent mixing is to express (3.11) in terms of the mean entity E and its derivatives. The exchange-coefficient hypothesis is that the term -pE'w' can be expressed as the product of a virtual coefficient of mixing and the gradient of the mean entity dE/dz." For the conservative property, momentum, the analogy between molecular and eddy motion was extended by Prandtl (1934), who developed a "mixing length" term which is analogous to the mean free path of molecular motion, and by von Karman (1953), who was able to express the Reynolds stress in terms of molecular motion. However, Sutton (1953) suggests that the "mixing length" term fails to account for large ed- dies. Taylor's (1932) "vorticity-transfer" theory is discussed along with statistical theories and a "similarity theory" of Kolmogoroff's (1941). These theories are an improvement on the "mixing length" theory, but none of them are adequate as yet. Halstead (1954) intro- duces correction terms in the transfer of momentum for effects of sur- face roughness and variations in temperature with height. 124 When the "eddy transfer coefficients" are expressed in terms of conductivity or diffusivity there is usually an implicit assumption that they may be equated with each other and with momentum. Pasquill (1949) used field measurements to test the aerodynamic theory. He points to the work of Taylor (1931) and Priestly and Swinbank (1947) in relation to the buoyancy effect, and demonstrates that the aerody- namic theory holds only under neutral conditions, or when the lapse rate, or rate of temperature fall with height,is small. It is shown that, in the absence of thermal stratification, the relationship be- tween the fluxes of momentum, heat and water vapor is valid. Under non-neutral conditions the relationship is not valid, but may be mini- mized by taking measurements close to the ground. Berger-Landefeldt (1956) summarizes the limitations of the aerodynamic approach. Besides those already mentioned, he stresses that the accuracy requirements for the instruments are prohibitive using Halstead's assumption of a logarithmic wind profile; that neutral conditions are comparatively in- frequent; and that microclimatic influences, rather than instrumenta— tion difficulties, may be responsible for the general failure of the method. Extensive field trials were conducted during the Great Plains Turbulence Field Program, sponsored jointly by ten universities and five government scientific organizations. The data collected permitted a comparison of the aerodynamic approach with the heat balance approach (discussed later). Site description, instrumentation and field data are presented in a final report (1957) of the A. & M. College of Texas, Dept. of Oceanography and Meteorology. The Johns Hopkins Laboratory of 125 (Ilimatology (1954) also issued a final report on earlier work, which provides an excellent discussion of the instrumentation problem. Van der Bijl (1957-19581 using the field data, made comparisons in evapo- transpiration rates between the methods of Halstead and Suomi (aerodynamic and heat energy approaches, respectively). The field data also permitted him to compare evapotranspiration rates using the for- mulas of Thornthwaite and Lettau. His conclusions are: "The relationship between evaporation as a dependent variable and meteorological parameters as independent var- iables for periods less than a few hours cannot yet be determined. For longer periods the agreement between com- puted and actual evaporation values improves, but the same can be said for the values computed with the help of simple formulas (Thornthwaite's. Author's note.)." C. Direct Eddy Flux Determination. The third method of determining evapotranspiration by vapor flow methods involves a direct determination of the vertical eddy flux of water vapor, which is the component indirectly achieved by the aerody— namic approach. Since McIlroy (1957) was instrumental in the develop— ment of this method, we can do no better than quote him: "At any point in the atmosphere turbulent air movement will normally be present, giving rise to a fluctuating ver- tical component of windspeed. In the presence of a vertical gradient of humidity these fluctuations will be associated with simultaneous fluctuations in moisture content of the air. Rising air will tend, on the average, to be moister or drier than descending air, according to the sign of the gra- dient. This will give rise to a net transfer of water vapor along the gradient. "The mean rate of this transfer, per unit time and per unit area, is known as the vertical eddy flux of water vapor. Letting a bar over a quantity denotes its mean value with respect to time, and a dash the difference between an instan- taneous value of a quantity and its mean value, the eddy flux is given by 126 Ex = (pW)'q' where p, W and q are simultaneous values of air density, ver- tical windspeed and specific humidity at the measuring point. Hence (pW)' represents an instantaneous fluctuation in the rate of upward air flow at the point, and q' an associated fluctuation in moisture content. "Records of pW and q taken at a point near the ground, . will permit the direct evaluation of Ex, which is of course the quantity sought indirectly by the aerodynamic approach. "It is obviously impossible to measure instantaneous values of pW and q, but with sufficiently rapid response in- struments only comparatively high frequency fluctuations are missed, and these make little contribution to the total transfer. "The actual rapidity of response required for satisfac- tory flux measurement depends on the pattern and scale of the turbulence. In general the requirement decreases with height above ground and increases with windspeed. It also varies with the roughness of the surface and with certain weather factors. "At heights of a meter or more a response time of about a fifth of a second appears to be adequate under most cir- cumstances. This means in effect neglecting fluctuations much faster than a second in period. "Such a rapid response cannot be obtained with stand- ‘ard meteorological equipment but is well within the limits of performance of fine sensing elements. However, rather elaborate recording or computing systems are necessary. Apart from this and the degree of sampling variation likely to be encountered, the method is completely independent of the nature of the surface above which the measurements are made. Much further work is required before regular field use of this technique will be possible. However, there is no doubt that it is even now the best available for accuracy and general applicability." Surface Energy Balance. This method makes use of the formula R equals G plus H plus LE, where R is net radiation 127 G is heat flux into the ground H is sensible heat flux LE is latent heat flux of evaporation A more complicated form is given by Tanner (1957) and Tanner and Suomi (1957), in which vertical and horizontal flow are separated, and changes in both sensible and latent heat stored in the "volume" (air, crop and water vapor) are considered. They point to the work of Rider and Robinson showing that if neglected the storage term will only in- troduce an error of about 1 percent during clear days when Et is high, and about 10 percent during sunrise, sunset and nighttime, when Et is low. Ignoring the horizontal flow factor (divergence) may produce errors as large as 40 percent, however. When measurements are taken near the ground or crop surface, the horizontal divergence is greatly minimized. The heat flux into the ground is measured directly by thermometers near the surface and heat plates deeper in the soil. Net radiation is measured directly by the "economical net radiometer" de- veloped by Suomi (1956). 'Van der Bijl (1957-1958) makes a comparison of several types of net radiometers, concluding that Suomi's model could be improved by "providing the radiometer with sensors of larger heat capacity" to obtain more reliable results for one hour periods. Other models are compared, and he recommends the use of standardized testing surfaces. The chief difficulty encountered in the energy bal- ance approach lies in separating the latent and sensible heat terms. This separation is usually performed according to the Bowen Ratio. This ratio was determined for conditions of laminar flow, where the 128 coefficients of heat and water vapor transfer were shown to be equal. As Tanner (1957) comments, however, "There is some controversy regarding both the equality of Kh and Kw (turbulent transfer coefficients of heat and water vapor, respectively) and the constancy of Kh/Kw as atmospheric stability changes (the "bouyancy effect")." He further states: "The main advantage of the energy balance--Bowen ratio method as compared to aerodynamic methods, which require a product of the wind and vapor pressure gradients, is the fact that the measurement of net radiation and soil heat flux is made at the surface and places reasonable limits on the magnitude of the sensible and latent heat flux. The aerodynamic method refers all the measurements to a region in the atmosphere. Both methods include assumptions on the eddy coefficients of transfer but the error in the aerody- namic method is proportional to the errors arising from an incorrect assumption concerning the eddy coefficients whereas the energy balance method is less sensitive.... H Combined Surface Energy Balance and Dalton Equation. This method is associated with the names of Penman (1949) and Ferguson (1952) who independently arrived at the conclusion that, where a relationship could be established between the temperature and vapor pressure at a surface, then the energy balance equation could be com- bined with the Dalton type equation to eliminate both of these varia- bles. Such a relationship is rather readily established for a water surface. According to Penman (1949): " the chief justification of the great attention given to evaporation from open water is found in the oppor- tunity it presents of providing a reproducible surface of known properties. Because of this, it is convenient to ap- proach the problems of the dependence of evaporation from bare and cropped soil on weather conditions through a study of evaporation from open water, seeking an absolute rela- tion between weather elements and open water evaporation, and comparative relations between losses from the soil and losses from open water exposed to the same weather. 129 Evaporation from an open water surface is approximated by a combining of the energy balance and Dalton type equations." For the mathematical treatment and assumptions used, the reader is referred to Penman's original paper (1949). The final expression be- comes E = (HA+ Ea?) (13+ 3*) E is evaporation in unit time. H is net radiant energy available at the surface. [515 dea/dTa, where ea and Ta are sat. vapor pressure and temperature of the air,respectively. Ea is value of E(o) obtained by putting es = ea. 'Y'is constant of wet and dry bulb hygrometer equation. Empirical coefficients, based on lysimeter tests, are then obtained for the relationship between evaporation from open water (Eo), bare soil (Eb) and crop cover (Et). They are shown below: This method has met with the same sort of limited success as Thornthwaite's method because of the large empirical content and simplifying assumptions. 130 Appendix 2. Techniques Used to Determine Soil Properties Soil color was determined from Munsell color charts after wetting the soil to a l : 1, soil to water, ratio. Soil pH was determined with both a Beckman pH Meter and Fisher titrimeter. Soil texture was determined by the Bouyoucos method. Upper plastic limits were determined by the Yoder method. Bulk density was obtained with the sampler and procedure proposed by Carey. Laboratory methods of determining soil moisture at 20- and 60-cm tension are described by Hoover, Olson and Metz. July soil moisture determinations, preceded by 70 consecutive rainless days, were used for an approximation of permanent wilting point. Wilting point was determined at 15 atmos. using a pressure membrane, but consistently yielded results 2 to 5 percent higher than the July readings. 131 m.mN b.NN H.0N m.HN m.NN m.wH m.mH on m.wN w.HN m.mN o.NN H.NN m.mH N.mH VN N.®N m.wN m.bN H.oN w.ON m.mH N.®H NH m.vN m.wN m.wN m.mN H.mN v.vH m.vH NH m.vN d.mN H.om b.NN u.NN H.NH ¢.h mlo m uon b.0N «.mN m.ON m.HN h.mN «.mH h.mH H.mN vNumH H.HN m.mN $.0N v.HN m.mN o.bH N.mH m.wN NHINH H.HN N.mN m.mN v.mH b.5N v.vH m.vH h.bN NHlm N.mH o.HN o.mN. m.HN m.mN m.HH H.oH N.vN mlo Hum uon N.NH N.vH m.mN o.mH H.mH H.mH w.m H.HN one HIN uon m.m m.HH m. m.mH m.mH m. h.mH m.m b.NN mlo HIH uon r Ammnoch e m NH oH w b v canon mmmH mmmH .muoHa.xoo:o can ustH noH osHm> AuanoB >2 usoonoav manumHoE HHOm NHsuzoz .m xHeann< 132 m.mH N.NH N.NH H.NH v.mH b.bH v.wH H.NH v.mH m.mH N.HN we 0.0N o.NH N.0N H.NH v.mH N.NH m.mH b.5H m.>H v.wH v.NH e.mH N.HN Nv N.ON N.NH H.HN N.NH m.mH o.NH o.NH b.wH m.bH o.NH m.mH v.mH m.mN mm v.0N m.mH m.mH b.mH 0.0N N.NH m.mH o.NH w.hH N.NH N.NH e.NN N.NN om H.0N N.NH N.0N N.NH n.0N N.NH N.NH m.mH N.NH o.NH N.NH m.vN w.bN VN n.0N N.NH ®.0N N.NH n.0N m.0N N.NN m.ON m.>H m.bH m.mH N.vN H.NN NH N.0N v.mH m.vN v.mH $.0N m.HN N.vN N.NN N.mH m.mH H.NH v.ON N.NN NH H.0N H.mH H.NN v.mH m.ON N.NN N.HN H.vN N.HH H.NH m.NH h.mH N.mN ono 21m uon o.NN N.hN ®.VN H.mN N.NN m.mN o.mN N.0N N.NH N.NN m.vN m.mm mm N.HN o.vN o.mN N.NN m.bN m.vN H.NN m.VN v.mH n.mH m.0N N.mN N.NN om N.NN m.NN N.VN N.NN o.NN H.mN m.bN o.vN H.NH H.NH H.HN H.vN N.HN vN N.NN N.mN m.mN o.NN N.NN N.NN b.5N o.NN m.bH a.pH b.0N o.vN b.HN NH w.NN m.mN m.bN m.mN o.NN N.NN N.NN b.0m o.bH m.bH e.mH N.NN m.vm NH o.HN N.HN m.om v.mN v.mN N.HN b.mN m.Nm N.NH m.vH o.NH m.HN o.mN mno SIN pon N.NN m.mN N.NN H.HN v.mN o.mN v.NN o.NN o.NN n.0N N.mN N.NN we m.NN m.vN m.mN H.mN m.mN b.mN N.HN b.ON N.NN m.nN v.mN v.Hm Ne o.NN m.VN N.NN .o.mN N.NN H.NN N.NN b.mN o.ON H.NN N.mN N.NN ¢.Hm mm N.NN N.NN e.NN b.mN b.mN N.NN N.NN N.NN N.NH N.HN H.®N N.NN m.om om N.NN m.VN b.mN m.mN m.VN N.NN h.NN N.HN N.ON H.NN b.¢N N.om o.NN vN o.NN o.mN h.mN m.mN N.mN v.mN m.mN N.vN m.ON v.NN N.NN H.NN m.NN mH N.NN o.mN N.NN m.VN H.NN m.mN N.om N.bN N.HN o.NN m.NN «.mN v.mm NH v.0N N.NH m.om N.NN N.NN o.Hm H.NN N.NN N.NH H.oH H.NH m.mm H.om mno SnH Hon Antennae e n a H NH HH oH m m s e m e seams mmmH mmmH .muoHn asHpmz you 03Hm> Auanos mm unmonoav manumHoa HHOm manzos .m xHusmna< 133 N.NH v.wH v.mH N.NH N.NH N.NH o.HN v.mH o.NH N.NH N.HN N.NN on H.NH «.mH v.mH b.wH N.NH N.NH N.HH H.HN m.mH N.oH N.bH v.HN N.NN vN b.mH N.NH N.NH m.NH N.NH $.0N o.NH o.NN m.mH o.¢H N.NH v.HN N.NN NH H.NH N.NH N.ON N.NH N.ON o.vN N.0N N.NN v.mH h.mH m.NH ¢.HN m.bN NH o.NH N.NH e.hN o.NH 0.0N 0.0N o.HN N.NN b.m N.oH h.NH o.NH m.vN muo man uon N.NN N.NN m.vN o.vN b.mN N.NN N.NN N.NN H.mN Nv v.mN N.NN N.NN o.NN m.VN w.VN >.0N H.NN b.HN N.HN N.NN m.NN N.NN mm o.vN m.vN N.NN b.VN o.NN N.NN m.mN H.HN N.HN N.NN m.vN N.NN om b.NN m.mN m.mN b.mN m.VN m.mN N.mN o.NN H.HN N.0N v.mN N.VN o.NN vN H.NN N.NN m.mN o.NN «.mN m.mN o.NN N.HN v.HN m.oN o.NN o.NN m.mm NH N.NN H.NN m.mN N.NN «.mN mnbN w.mN N.NN n.0N H.NH N.NH «.mN m.Hm NH N.NN N.NH N.NN N.VN o.NN v.Hm o.mN b.vm m.ON m.oH v.0H o.NN N.HN mlo muN pon N.vN N.VN N.mN v.vN v.mN N.vN N.NN N.NN N.NN N.NN m.mN w.vN on we N.NN H.mN N.mN N.NN N.NN H.vN H.vN m.NN N.NN N.NN v.mN m.vN om NV o.HN b.mN N.vN N.NN m.NN N.vN N.NN N.NN m.NN v.NN N.NN m.mN N.om mm v.NN v.mN N.¢N N.NN o.NN v.VN o.NN N.NN v.NN N.HN N.NN N.NN m.om on N.HN N.NN b.vN N.NN v.mN N.NN o.vN N.ON o.NN N.HN o.VN N.NN o.HN vN N.NN N.NN b.vN N.NN N.NN N.VN H.NN v.NN N.NN N.NN H.vN v.bN N.NN mH N.NN N.NN m.mN H.mN N.NN m.mN o.NN N.NN n.0N o.NN o.NN N.NN v.vm NH v.NN N.NH m.mm o.NN m.mN N.om v.nN v.mN n.0N N.NH m.mH N.NN N.HN mno :IH uon Hmonoch v m N H NH HH 0H m w h m n v spoon mmmH wmmH .muoHQ z>mmm now osHm> AusmHmB kn unmoamnv oasumHoa HHOm chuzoz .m vacoaa< 134 Appendix 3a. Procedure used to correct percent moisture for rock. Dry screening was impossible, since some rock was crushed by the pestle as readily as the oven-dry clay. About 10 samples per 6-inch depth interval per plot were collected after oven-drying, weighed,and reduced to a slurry. The slurry was then passed through a 2 mm screen, the stones separated and weighed. Percent stone of oven-dry soil weight was computed for'each 6-inch depth interval for each plot. These values were then plotted on graph paper and curves drawn. The curved values are shown on the following page. Curved values were then used to correct Pw for all sampling dates. A check was made by repeating the above procedure for samples taken on March 3, 1959. Agreement between curved and actual values was good; the maximum deviation would change inches of water in a 6-inch depth interval by only 0.20 inch, while the maximum change for total soil depth did not exceed 0.40 inch. Most values were considerably smaller than this, and were not consistently in one direction. Plot no. l-L 1-M 1-H Depth in. 0-6 6-12 12-18 18-24 24-30 30-36 36-42 42-48 0-6 6-12 12-18 18-24 24-30 30-36 36-42 42-48 Rock pct. 18.5 03000030000154 00000000 h‘h‘h‘h‘le'w é 01001010000 Plot no. 2-M Depth 1n. 0-6 6—12 12-18 18-24 24-36 0—6 6-12 12-18 18-24 24-30 30-36 36-42 42-48 Rock pct. 11.0 ~100th 00000 Akkifilfivfimw 00000000 Plot no. 3-L 3-M 3-H Depth in. 0-6 6-12 12-18 18-24 0-6 6—12 12-18 18-24 24-30 30-36 36-42 42-48 0-6 6-12 12-18 18-24 24-30 30-36 6-12 12-18 18-24 24-30 30-36 135 Rock pc HNNN on NNHH h‘h‘h'H N (”0000000 QNNNNQ t. 000101 0010000 000000 136 N.HN N.m v.v v.m «.me v.wm N.NN N.Hv mN.H mmlmm N.NN H.HH v.o H.m m.Hm m.ov m.>m o.Ne oN.H omnbm N.mm w.m o.N N.H m.om N.NN m.Nm v.0v om.H b.mm m.m o.H N.H w.mv m.mm m.Hm m.mm om.H emuHm N.NN N.HH m.m o.m m.mv v.mm N.mn v.mv mN.H H.8m m.N o.H N.H 0.0m H.NN n.Hm N.NN Hm.H manna h.>m m.N N.H m. m.om v.bm N.Hm N.NN mm.H wmumm o.om o.v o.N m.H N.Nv n.0m m.om H.NN Nm.H m.hm >.N N.H m. m.>v N.mm N.Nv N.mm mm.H omubN m.bm v.N N.H o.H o.Nm m.Nv m.vm o.Nv mN.H w.vm N.m v.N N.H H.Nv N.mm o.Nv H.hm Hm.H vNIHN N.NN m.m w.H N.H n.8v H.mm H.Ne N.NN mm.H w.mm N.v o.N m.H N.Nv m.mm m.om H.NN Nm.H m.mm H.v o.N o.H m.mv m.mm o.Hm H.Hv mN.H b.5m m.N N.H o.H v.Hm w.ov o.Nm o.Hv oN.H wHumH o.mm o.m m.N N.H H.Nv N.>m m.om N.NN NN.H b.ww N.HH H.m b.v H.Ne N.bm N.vm o.Nv mN.H N.mm m.v v.N o.N o.Nv N.NN v.0m w.ov vN.H NHIN e.vm m.m o.m ®.N N.om N.Ne N.Nm m.mv vH.H b.vm m.mH m.w a.» N.Nv m.mv H.Nm «.Hm NH.H mum H.Nm m.o h.m N.m H.om N.Nv N.Nm o.mv bH.H w.vm N.m m.N v.N v.Nm m.mv m.mm o.Nv mH.H m.mm e.vH m.m m.m H.Hm H.mm b.mm v.vm mm. N.NN N.bH w.m N.HH H.bv N.Nm N.Nm m.vm mm. m.mm . N.v v.N H.N H.em v.mv n.0m m.om NH.H muo :IH .noa .poa NM elm. NM a. NM mm .mm. .mm maHmoaom muHmoaom Eouom a .pdm sonsoB Eonom :oHHmHSHam huHmsmn manHHHaao anuHHHanzoz somapom pm as stm sumo: uoHa NooH u saanonoa Hausa eHmm amass eHam amass eHom amass .mmHasoaoaa HHOm couaHon can huHmsmv stm .v chstd< 137 N.NN w.b m.m m.N m.mv N.HN m.>v N.NN ov.H H.8m m.N N.H w. H.Nv o.om v.vv v.Hm Hv.H mmumm m.mm m.v H.N m.H o.mv N.NN H.Nv m.vm mm.H N.NN v.v o.N m.H m.mv N.vm o.Nv m.mm em.H omnbN N.NN m.o m.m v.N m.m¢ v.Nm o.Nv b.vm ov.H b.mm m.v H.N N.H b.mv o.mm m.mv N.NN NN.H leHN m.vm H.m v.N N.H N.ov H.mm N.Nv o.bm NN.H m.om b.m N.m N.v N.Nv b.mm m.mm N.Nv mN.H mHlmH N.NN m.oH w.m h.v H.Nv N.NN N.Nm m.mv ¢N.H NHIN m.>m m.N N.H o.H 0.0m o.NN N.Hm o.ov NN.H e.vm m.m w.N N.N o.Nv v.wm N.Nv b.0v NN.H mum «.mw m.oH o.m m.w o.Nn H.Nm o.No v.mm mo.H N.NN N.>H >.N v.m N.NV m.oe m.om N.Nv 0H.H m.Hm e.w m.v >.m o.Nv o.ov m.mm b.mv NN.H muo SIN N.HN N.m v.v H.N v.0m b.mm N.vm N.NN Hv.H mvumv o.Nm 0.5 o.m m.N o.Nv o.mm o.Hm m.bm hm.H Nvlmm H.NN m.w w.m m.N v.mv b.vm o.Nm N.NN NN.H mmlmm N.NN N.m m.m v.N o.Nv m.vm N.Hm N.NN NN.H N.mm w.v v.N N.H N.>v m.vm o.Ne N.NN mm.H omle m.vm m.m N.N H.N m.bv v.mm m.om m.bm -mm.H N.NN N.o N.m ¢.N N.Nv m.mm o.Nm N.NN mm.H vNIHN H.om m.m o.N m.H N.Nv N.NN m.om m.bm vm.H b.Nm m.b h.m m.N H.bv N.mm w.om m.bm vN.H anmH m.mm m.w m.m m.N m.bv w.mm w.om N.NN NN.H NHIN N.>b N.NN m.vH o.NH N.Ne >.mv v.No N.Nm mo.H v.Hw m.mH H.NH N.HH N.Nn H.Nv m.vm m.ow mo.H m.mw b.eH w.b m.m 0.0m m.Hv o.Nm m.mv ON.H muo muN slum. an m. m m m m m a. .9 maHmoaom muHmouom Eouom a .uam :onsmB Eonom :oHumnsawm muHmson mawHHHanD maaHHHamonoz cooaaom an an tzm spawn uon eooH u seHnonom Hence eHam amass eHmm amass eHem noes: .Hpoanpzoov moHpsoqona HHom cmpsHoa use muHmnmU stm .v xHocoad< 138 h.mm m.v N.N N.H b.mv N.ov m.om H.Nv HN.H N.hm N.N N.H m. v.Hv m.om o.Nv b.Hm em.H mmuom m.mm v.v H.N v.H w.mv v.Hm h.b¢ mmNm mv.H m.mm >.v o.N ¢.H v.0e H.NN e.Nv m.mN vv.H omubN m.mm m.o N.N N.N N.Nv H.NN v.mv v.vm vv.H N.NN v.> o.m m.N m.mv N.NN 0.0m m.vm mv.H «NIHN N.HN H.N H.e o.m N.Nv N.NN v.0m N.NN NN.H v.mm m.o N.N v.N m.wv o.vm N.om «.mm NN.H mHumH N.NN N.o N.N o.N b.Nv N.om m.mv N.Nm H¢.H NHIN b.>m N.NH N.m m.v v.ev N.NN m.om o.NN NN.H N.vm m.m N.N N.N m.mv o.mm H.Ne m.bm NN.H h.mm m.v o.N m.H m.vv o.Nm m.mv m.nm NN.H mum N.NN N.oH b.m o.m N.om o.ve m.mm N.Nv vH.H N.NN H.HH v.m m.v m.mv m.bm N.Hm >.Nv 0N.H N.Nm b.> H.v N.N N.Nv H.NN N.Hm m.Nv NN.H muo Houm .eoa .eoa m a m NM mm mm .mm .mm NHHmoaom maHmOAOQ souom é .umm :oncmB Eonom :oHuaazumm muHmson mamHHano mamHHHaaocoz somapom an stm spawn Hon gooH n muHmonom Hauoe oHom nouns UHmm nouns nHmm hopes .ApmssHucooV mepHoaoam HHOm coHMHmu can huquou stm .v chcoqa< 139 Appendix 4a. Procedure used to compute average bulk density for each plot. Bulk density samples were individually corrected for rock by the procedure outlined in Appendix 3a. A second oven-drying was not nec- essary in this case. Bulk density values from plots l-H, 2-H, and 2-M were combined be- cause of similarities in soil properties and actual bulk density values. The check plot appeared to be distinct not only in terms of bulk density values, but also in texture, wilting point, soil moisture at 60-cm tension, and stoniness. This difference in soil properties other than bulk density was common to all the number 3 plots, and was not surprising since they are all grouped at one end of the study area. Bulk density values were then plotted by averages for each 3-inch depth interval sampled. Curved values were then obtained at the mid- point of each 6-inch depth interval as shown and tabulated on the fol- lowing three pages. Plots not sampled were assigned values from plots most resembling them in other soil properties. ‘1 140 mm mm on L H a . xoom Havana onoHHmo 1") EN H [W mmsosH I swamp HHOm an S «H 2 N.H H m a. .1 d .ae xHucoQa< OH.H ON.H O. nu TL H. . w. on H u S T... 1 «A ov.H om.H 141 .mm XHusonn< monosH I spawn HHOm mo om hm an Hm we we Nv mm mm mm on bN VN HN NH mH NH m m M" o . T 1:11-..I¢I...IHJ1II.1 HI fl 1H1--- 1 141 . 111s is :JI-.I-.JIIII .. .-I. , _ 1 l 1 4 00.H Q \i . oH.H \ Q . - on H m. TL .. x. O . D. 2H HTH eoE . m 4 ...... \:\1\ 1 OM.H.M. I g I g In a 1| 0 M amnHmmzmeE ..osH 142 Appendix 4a. Final curved values for bulk density. Plot Depth Curved Bulk Density Value no. in. 3—Check 0-6 1.26 3—L 6-12 1.37 12-18 1.40 18-24 1.43 24-30 1.44 30-36 1.31 3-M 0-6 1.20 Estimated (larger percent litter than 3-0) 12 1.37 18 1.40 24 1.43 30 1.44 36 1.45) 42 1.45) Estimated (no evidence of caliche 48 1.45) from color, pH, texture, plastic limit tests) 3-H 0-6 1.18 Estimated (larger percent litter than 3-M) 12 1.37 18 1.40 24 1.43 30 1.44 36 1.31 2-M 0-6 1.15 2-H 12 1.23 18 1.28 24 1.32 30 1.35 36 1.37 42 1.38 48 1.39 143 Appendix 4a. (continued). Plot Depth Curved Volume Weight Value no. in. l-L 0-6 1.20 Estimated since less litter 1-M 0-6 1.15 1-H 12 1.23 18 1.28 24 1.32 30 1.35 36’ 1.37 42 1.38 48 1.33 54 1.29 60 1.28 66 1.28 144 Appendix 5. Average soil depth. 10 Percent Plot Mean Range Number Confidence Standard Samples Interval Error 22. ‘13. in. ‘in. in. *l-L 3.35 2-8 14 0.7 0.32 *2-L 3.41 2-7 14 .7 .32 3-L 21.1 12-30 45 1.1 .65 1-M 47.7 20-62 38 2.7 1.60 2-M 32.6 19-48 45 '2.2 1.30 3—M 48.4 31-60 42 1.9 1.13 **1-H 48.0 2-H 37.6 22-72 40 2.6 1.54 3-H 30.2 19-37 42 1.1 .68 3-Ck 31.6 24-44 35 1.3 .77 *Because of the difficulty of measuring the depth of soil intercalated with basalt, a depth of 6 inches was assigned to these plots. **Because it was usually impossible to hit bedrock with an 8-foot tube, a depth of 4 feet was assigned to this plot. Soil moisture change be- low 4 feet never exceeded 1 to 2 percent between sampling dates, values well within the range of sampling error. Average soil mois- ture between 4 and 8 feet on April 22, 1958 and April 21, 1959, showed a difference of 0.7 percent. 145 oo.- NH.m mm“: no.1 Hneos no.H NH. NN.H Nm. NH.H NH. Nm.H NH.- ne.H on oo.H oo. oo.H om. ov.H oo.- ms.H no. Nv.H «N NN.H oo.- NN.H No. oo.H oo.: om.H no.1 He.H NH oo.N oH. oo.H ob. NH.H vo.u NN.H oo.u oN.H NH NN.H mN.- NN.H oo.H Ho. no. on. NH.u mo. o-o sonm HoHa mmwn NN.N oH. mo.H- mmumw mmun_ Haeos sH.H Ho. NH.H mN. mo. Ho.- om. No. so. NN.- NH.H 5H.- NN.H H.HN oH.N oH.- mN.N No. me.H so. oo.H oo.- oN.H Ho.- oo.N so.- oe.N NH NN.N NH. mH.N so. NH.H No.- oN.H mm.- on.H Hv.- oo.H No.- NN.N NH oo.H Ho.- HH.N 4N.H so. HH. ob. Ho.- so.H so.u os.H oo.- so.H ono Hum eoHa NN.H oo.- «N.H on. so. so. sN. so.- so. oN.- so. oo.- NN.H ouo H-N eoHa NN.H oo.o- He.H .oN.o NH.H oo.o «N.o oo.o- NN.o oo.o- ns.o oo.o- mo.H ouo HIH eon 0H mmnvfim u 0 mGUEQ u 0 mwocm u N. mQUGQ u 0 mmoflm " m mmoflm u an bN .uooulhmmufinumm .uammulhmmmwn "HN .mz< NN.H VN.H no.H sn.o ns.H sN.H n HV.N oo.N HN.H nH.o oN.o NV.H N nN.N NN.N oN.o oH.o ns.o HN.H H Haven nN.o1 HN.H Vn.o Nn.o1 oN.o1 oN.o1 .m>< NN.o1 NH.N no.o sn.o1 ns.H1 Hn.o- n oN.o1 oN.o so.o so.o1 oN.o1 NN.o1 N 8.? No and 3.? oo.? oN.o1 H 2n Q HN.H N NN.o H a nH.o ns.o .m>< nH.o nH.o n nH.o HN.H N nH.o NN.o H can nH.o oN.o oN.o Ho.o oo.o so.o .o>< nH.o oo.o NN.o no.o oo.o so.o n no.o oH.o oH.o oo.o oo.o Ho.o N oN.o so.o NN.o Ho.o oo.o no.o H .eeH nN.N Hn.n NN.H so.o oo.o Hn.o .a>< so.N Nn.n Hs.H no.o oo.o nn.o n nN.N no.n no.N oo.o oo.o Hn.o N sV.N no.n oo.N oo.o oo.o ns.o H .eaa .aH .aH ..aH .aH .aH .aH HoHo Hoo Hno Hsv Hno Hno sN .eoo nN .eaen HN .ma< NN sHao oN ease NN ss: .02 on on on on on on noHa nN .namn HN .w:< NN sHao oN mean NN as: NN .aa< .nuoHQ Npanou uanH 1 mucosoasoo mocmHnn nouns map How mosHm> .b vacoqa< 152 fin- .m>< .o>< .m>< .o>< .o>< Hn.nH oo.H Hn.o NH.H nn.o nn.o oN.H NV.VH sn.H oN.o nn.o NV.o Vn.o so.N HN.NH Ho.o oo.o VH.H nn.o Hn.o no.o Vs.NH Nn.o Ns.o HN.H Nn.o oo.o ss.o nH.H1 Nn.o1 Hn.o1 Ns.o No.o1 nn.o1 no.o HN.H1 no.o1 oN.o1 nn.o no.o sn.o1 ns.o1 so.o1 VH.o1 oN.o1 oN.o no.o1 Hn.o1 nn.o so.o1 NH.o1 Ns.o1 ss.o oH.o1 oo.o1 ns.o HN.H nn.o Ho.o Nn.o sn.H oo.H Ho.H no.o oo.o nH.o no.o oo.o no.o oe.H. HH.o oo.o sH.o so.o oo.o nH.o oN.o no.o oo.o no.o Ho.o oo.o no.o NN.H no.o oo.o nH.o so.o oo.o oH.o VN.VH ss.o oo.o NN.H nn.o oo.o NN.H oN.VH Ns.o oo.o no.N Nn.o oo.o VN.H nn.nH oN.o oo.o NN.H nn.o oo.o Hs.H NN.NH ns.o oo.o HN.H nn.o oo.o on.H .:H .:H .sH .:H .:H .:H .nH Hey HnV HNV HHo HNHV HHHo NN .ae< on .sss n .as: oN .aao on .ooo H .eoo Hseos on on an on on on on .sas n .ssz oN .aae on .omo H .omo sN .eeo coHnmm NsHHQEmN .HvoscHusoov nuoHQ aancoc HSNHH 1 musocoasoo mostmp swans one How mmsHN> .b chcoaq< 153 sn.H ss.H NN.H Ho.H HN.N nn.N .o>< no.H NH.N HN.N no.H Ns.H ss.o n ns.H no.o NH.H nn.o Ho.N NV.N N on.H no.N nn.N VH.H HN.N Ho.N H Hnoem ns.o ns.N sn.o1 sn.o1 HN.N1 vn.N1 .m>< ns.o NH.N NH.H1 oo.H1 Ns.H1 No.N1 n oN.o nn.N nH.o ns.o1 Ho.N1 oo.N1 N Ns.o sN.N nn.o1 NH.H1 HN.HT Ho.H1 H 2nd. HH.o HH.o .m>< HH.o HH.o n HH.o HH.o N HH.o HH.o H can sn.o nn.o ss.o so.o oo.o NH.o .e>< ns.o NH.H Nn.o Ho.o oo.o HH.o n NN.o so.o ss.o no.o oo.o HH.o N Ho.o ns.o sn.o no.o oo.o sH.o H .naH NN.N ne.n on.H no.o oo.o nn.o .e>< so.N nn.n ns.H no.o oo.o nn.o n nH.N nn.n Hs.H no.o oo.o nn.o N «N.N nN.n NN.N no.o oo.o sn.o H .naa .0H .0H .:H .:H .:H .:H HoHV Hoo an Hso Hno Hno sN .eoo nN .aamn HN .ma< NN sHso oN mass NN sea .02 on on on on on on eoHa nN .eann HN .m=< NN sHso oN ease NN sax NN .aa< coHHom NsHHQENN mHoHa zuHmcoU schoS 1 mucmsoasoo moanmn soon; on» HON mosHm> .b chconq< 154 I tr. I‘ll. ...."m' .NQHHQENm oHDHnHOE HHON :H :oprHnm> mp Nomzmo mH :oHHnHHaHomHQ mo mnooxo :H :oHpasHancmsuoam>NH .o>< HN.NH VN.0 HN.H Nb.0 NH.H HNN.01 NN.H NN.NH HNN.01 HN.H No.0 VN.0 NN.0 NN.H 0N.NH NV.H N0.H Nb.0 VH.N NN.01 V0.N bH.NH 0N.H NN.0 NV.H NV.0 NH.o1 NN.H .o>< NN.V1 VN.01 HN.H1 bN.0 Nb.01 NN.0 NV.01 Nb.N1 bN.0 HN.H1 NN.H NN.01 NN.01 No.01 VN.N1 bN.01 N0.H1 N0.H bN.H1 NN.0 HN.01 NV.N1 H0.H1 NN.01 NN.0 No.0 NH.0 0N.01 .m>< NN.0 NN.0 NN.0 NN.0 .o>< Hb.N bH.0 00.0 NN.0 No.0 00.0 NN.0 00.N VH.0 00.0 NV.0 HH.o 00.0 NN.0 NN.N bH.0 00.0 NN.0 b0.0 00.0 HN.o bN.N HN.0 00.0 NN.0 No.0 00.0 NN.0 .o>< Ho.NH Nb.0 00.0 N0.N NN.0 00.0 HV.H NN.VH Hb.0 00.0 N0.N NN.0 00.0 VV.H NN.VH Nb.0 00.0 N0.N VN.0 00.0 VV.H 0N.NH 0N.0 00.0 NH.N NN.0 00.0 VN.H .sH .:H .:H .cH .cH .zH HVo HnV HNV HHV HNHo HHHV NN .sa< on .aas n .aas NN .aao .oao .oeo H5009 0» OH OH 00 o» 0» 0N .Hn: N .HNZ NN ssh 0N .000 .000 bN .uoo mNoHsom MGHHQENN .HpmscHusoov muoHQ zuHmcoU EDHUNE 1 musmcoasoo mosmHNn nouns on» HON mous> .b chcmqa< 155 at .1'” I'D-'4 . u. 1 .NnHHQEmn oHSHmHOE HHON :H :oHumHHm> an ummsmo wH soHumuHQHomna Ho mnooxm CH coHunsHamcmauoam>mH ss.N NV.H Hn.o No.o No.N os.N .m>< sH.n so.o nn.o no.o mn.H os.N n Ns.o HsH.o1 nV.o ns.o sn.H ss.N N ns.H nn.N nH.H so.H on.N nn.N H anen sn.o1 NN.N ne.o nn.o1 No.N1 NN.N1 .m>< nn.H1 NV.N no.o1 sn.o1 no.H1 HN.N- n oo.N1 nn.V so.o nn.o1 sn.H1 on.N1 N sn.o nn.o ns.o no.H1 on.N1 nN.N- H anV so.o so.o .m>< so.o Vo.o n so.o so.o N so.o so.o H can ss.o NH.H nn.o so.o oo.o sH.o .m>< oV.o nH.H on.o no.o oo.o sH.o n vn.o nn.H ns.o No.o oo.o NH.o N ns.o on.o Hn.o no.o oo.o nH.o H .eaH NN.N nv.n NN.H no.o oo.o Nn.o .m>< No.N nn.n oV.H no.o oo.o nn.o n HN.N sn.n NN.H oH.o oo.o on.o N on.N NN.n NN.N no.o oo.o nn.o H .eaa .aH .aH .aH .aH .aH .aH HoHo HNV Hno Hso Hno Hno sN .aeo nN .aaan HN .na< NN sHao oN ease NN sax .oz on on on on on on noHa nN .aaen HN .na< NN sHso oN ease NN sea NN .sa< coHHmm NGHHQENN .nuoHQ HHHnsoU z>mom 1 mucosoasoo mocmHnn scans on» How mosHm> .b ancoaa< 156 Hn.nH Hs.o NN.H ns.o sn.o so.H ns.o NH.NH nn.o vo.H Hs.o nn.o No.H Hno.o1 nN.VH ns.o so.H no.o ns.o no.o sn.o nn.sH ss.o on.H ns.o no.o NH.H ns.o HN.N1 nH.o1 NN.H1 oo.o no.o no.H1 Ho.o HN.N1 no.o1 oo.H1 sn.o VH.o1 No.H1 nH.H so.N1 oN.o1 No.H1 sn.o no.o1 no.o1 ns.o so.n1 NN.o1 on.H1 oN.H Hs.o nH.H1 NN.o .m>< no.o no.o no.o no.o .m>< os.N NN.o oo.o ns.o HH.o oo.o on.o nn.o HN.o oo.o on.o oH.o oo.o NN.o nn.N nH.o oo.o on.o nH.o oo.o nn.o NN.N NN.o oo.o oV.o oH.o oo.o sN.o .m>< Ho.nH ss.o oo.o oo.N sn.o oo.o NN.H NN.VH ns.o oo.o sn.H Nn.o oo.o HV.H no.nH ns.o oo.o oo.N nn.o oo.o NV.H ms.nH nn.o oo.o nH.N sn.o oo.o NN.H .eH .aH .aH .aH .aH _.aH .aH Hvo HNV HNo HHV HNHV HHHo NN .Ha< on .Ha: n .Hss NN .ase on .omg H .omo Hanos on on on on on on on .Hs: n .an oN .aao on .omo H .omo sN .Hoo mpoHHom NGHHQENN .AcwssHusooV nNOHQ NNHmcmu N>Nmm 1 musmsoaeoo mocmHNn Honda mnu HON mmsz> .b KHUHOQ0< 1'“ .-‘ 11 I 1’" g. D: I"?! 1‘! 0 2 at?