Historic, archived document Do not assume content reflects current scientific knowledge, policies, or practices. rea 764M Ute 3 A HYDROLOGIC MODEL OF ASPEN-CONIFER SUCCESSION IN THE WESTERN UNITED STATES sae? Wena, Ag oes ee Et yaa r fA a age Ag Richard A. Jaynes mw -— a. ‘ v e USDA Forest Service Research Paper INT-213 INTERMOUNTAIN FOREST AND RANGE EXPERIMENT STATION FOREST SERVICE U.S. DEPARTMENT OF S Oma AGRICULTURE 2 KT THE AUTHOR RICHARD A. JAYNES, formerly a range technician on the Aspen- Mountain Grassland research work unit at Logan, Utah, obtained a B.S. in Range Science from Brigham Young University, and an M.S. in Watershed Science from Utah State University. Currently he is enrolled in law school at the University of Utah. ACKNOWLEDGMENTS This study has largely been supported by funds from the Forestry Sciences Laboratory, Intermountain Forest and Range Experiment Sta- tion, Logan, Utah. The author would like to acknowledge the assistance of Dr. Gerald F. Gifford, Dr. Kimball T. Harper, Dr. Richard H. Hawkins, Mr. Robert S. Johnston, and Dr. Walter F. Mueggler. USDA Forest Service Research Paper INT-213 August 1978 A HYDROLOGIC MODEL OF ASPEN-CONIFER SUCCESSION IN THE WESTERN UNITED STATES Richard A. Jaynes INTERMOUNTAIN FOREST AND RANGE EXPERIMENT STATION Forest Service U.S. Department of Agriculture Ogden, Utah 84401 CONTENTS Page INTRODUCTION sc: cona: ciate a Sp oeivay hota aier outs yom lems ian by aaerech aie genre eta 1 DEVELOPMENT OF ASPCON) go iiss cctea te comet aepien ceves Melee toner ciate 2 TRANSFER FUNC TIONS oi. 5 tice Veet rantae va eal Meh aes, oes tek Sane Mla reriecaee 4 MODE CALIBRATION? sxe rage eagrta «hel ime ana ane ret eae er tee eee 8 PREDICTED HYDROLOGIC IMPACT-OF SUCCESSION! fa cascu en sens. ale SUMMARY AND CONCLUSIONS © area 206) bide aitoenpe rie he oceune: getter 15 PUBLICATIONS CPE TD) so: “ecru keaeiteess workte sreyiatrd shatiterowel Seach aorta ta tal te seen 15 RESEARCH SUMMARY Hydrologic impacts of grass-forb to aspen to conifer succession in the Rocky Mountain area are simulated by means of a fundamental model. Model algorithms representing hydrologic processes are sensitive to vegetational changes within the subalpine vegetation zone. Reductions in water yield are predicted as the vegeta- tion on a small Utah watershed proceeds from a grass-forb type to aspen to conifers. Streamflow changes are largely attributable to an interaction between seasonal consumption for each vegetation type and the influence of vegetation type on snowpack. The model synthesizes present understanding and provides a framework for future watershed research. INTRODUCTION Forests of quaking aspen (Populus tremulotdes Michx.) are considered to be predom- inantly subclimax plant communities in the Rocky Mountain Region (Mueggler 1976; Bartos 1973). Mature aspen forests are most often replaced by evergreen conifers (Abies spp., Ptcea spp., Pseudotsuga spp., or Pinus spp.) unless some form of major disturbance occurs such as fire, disease, or clearcutting. ._ When an overstory is thus destroyed, prolific root sprouting of aspen generally is initiated and aspen regains dominance on the site. In many areas where natural fires have been curtailed and logging has not occurred, former aspen stands are now dominated by coniferous species. More than 4.1 million acres of commercial aspen forests (Green and Setzer 1974), and possibly an additional 1.5 million acres of noncommercial aspen lands, exist in the Rocky Mountains. Resource Managers are concerned that succession of sizable portions of these forests to conifers will have adverse impacts on the water, wildlife habitat, and livestock forage values of the aspen type. Because water 2S a critical resource in the West, it 1S imperative that we accu- rately assess the impact that succession from aspen-to-conifer may have on water yield. The concept of ecosystem: hydrology assumes complex interactions between the ecosystem and the hydrologic cycle, and that a change in one component should effect a change in the other (Huff 1971). With regard to transpiration, for example, Satterlund (1972) cited several studies that suggest '"'...the ecological principle that vegetation replace- ment by better-adapted species will continue until all favorable niches are occupied...." He concluded that ''...it appears likely that maximum rates and amounts of transpiration during the drying cycle occur under climax vegetation." It has been shown that western aspen may be expected to transpire 3 to 4 inches more water from a 6-foot soil profile than a grass-forb community on a comparable site (Johnston 1969). Douglass (1967) stated that many forest hydrologists believe well- stocked forests use the same amount of water regardless of tree species when end-of- season soil moisture deficits are examined. However, he pointed out that patterns of soil moisture depletion for hardwoods and for conifers are quite different. Because hardwoods begin transpiring later in the growing season than conifers, more water may drain. through hardwood soil profiles early in the season. Thus equal soil moisture deficits under hardwoods and conifers may not represent equal amounts of transpiration. Urie (1967) studied the net ground water recharge under hardwood and conifer stands in Minnesota. He found that the net annual water yield to ground water reservoirs from hardwoods exceeded conifers by 2.6 inches. This difference was associated with a greater snowpack under hardwoods and a longer transpiration season for conifers. He found that when transpiration and ground water recharge were combined, the conifers consumed 5.7 inches more water than hardwoods on comparable sites. In a Colorado study, Dunford and Niederhof (1944) concluded that, from the stand- point of net water available for streamflow, aspen is probably superior to conifers. A most meaningful insight to this problem was provided by Swank and Douglass (1974) who observed a 20 percent reduction in streamflow 25 years after a hardwood stand in North Carolina was converted to pine. Such a study is needed in the West to more accurately define the actual changes in watershed hydrology where aspen-conifer succession is occurring. In the absence of such research, a watershed hydrologic model based on recognized hydrologic processes and utilizing appropriate data from past studies and modern computer technology may provide useful insights. Such a hydrologic model may be of particular value in identifying critical research needs. A major purpose of hydrologic simulation modeling is to realistically and pre- cisely represent a system (a series of processes) with a network of mathematical expressions (Riley and Hawkins 1975). Models are comprised of coefficients, structure, and initial conditions that interact to manipulate each piece of input data to produce a desired output. Before a model can be deemed acceptable, it must be properly iden- tified and formulated, calibrated to mimic observed system behavior, and verified through repeated testing. Simulation models integrate the effects of a variety of subprocesses in order to provide for maximum utilization of a given information base in terms of predictive capability of system performance (Riley and Hawkins 1975). The purpose of this study is to formulate a structural watershed hydrologic model that will integrate available knowledge relevant to the hydrologic impacts of aspen to conifer succession. Although Leaf and Brink (1975) have written a rather sophisticated subalpine hydrology model, a fundamental model sensitive to aspects of the hydrologic cycle that may be influenced by vegetation changes would be useful. The model described in this report begins to satisfy that need. DEVELOPMENT OF ASPCON The model describing the hydrology of aspen to conifer succession (ASPCON) consists of a series of moisture storage compartments connected by transfer equations that sys- tematically deal with each set of input data (fig. 1). As moisture enters and interacts with a watershed, a certain amount is lost to the atmosphere via evapotranspiration, while the remainder may become streamflow or percolate deep into the soil. Obviously ASPCON can only be as valid as the assumptions that were made as the model was constructed. Literature pertaining to hydrologic behavior of grass-forb, aspen, and conifer ecosystems was carefully reviewed; only key references are cited. The model's transfer equations were derived from research findings that varied widely in location and purpose and, therefore, often were not directly applicable. Consequen- tly, many water movement equations must be considered educated guesses. This lack of information points out the need for definitive research that directly relates to the critical hydrologic problems associated with aspen-conifer succession. ASPCON is a deterministic, lumped-parameter model. The watershed is treated as a single series moisture storage ''tank.'"' Model coefficients related to watershed char- acteristics represent averaged values. The model calculates weekly water budgets throughout 1 water-year (Oct. 1 to Sept. 30). System input includes only precipitation and average weekly air temperature. The transfer functions for moisture routing within the watershed are described below in the sequence of ASPCON's algorithmic logic. Weekly precipitation Average weekly Channel interception (2) air temperature QCHP Interception Snow oo Rain Interception loss (5) ee loss (4) e SINT A (9) | Evaporation (6) Melt_inflow » eee fe z Snowpack sree i) melt (8) | moisture (10) Overland flow (11) Transpiration (12) QXS, QOF TRAN Evaporation (13) RVAP _aturation Deep percolation (15) Field capacity Interflow (14) Soil moisture storage SM Groundwater storage (1) ama wilting point Groundwater flow (17) Subsurface runoff Deep seepage (16) Channel routing (18) SEEP. Streamflow Permanent Surface runoff Figure 1.--Flowehart for the succession hydrology model (ASPCON). (Numbers in parentheses refer to definttion gtven in text.) TRANSFER FUNCTIONS' 1. Calculation of initial ground water level (GWL, in) from baseflow. At the beginning of the water-year (Oct. 1) average streamflow for the last rainless week of September is used to define the initial GWL. Initial GWL is the quotient of stream baseflow (in) divided by a ground water recession coefficient (AGW, in/in). 2. Channel interception (QCHP, in). The amount of moisture falling directly into the stream channel is defined as the fraction of the total watershed area consisting of surface water or saturated streambanks (ACHP, in/in) multiplied by the precipitation input. The value for ACHP may be determined from an areal map of a watershed. 3. Precipitation type. Form of precipitation is determined by using a routine similar to the model developed by the Army Corps of Engineers (1956): If TEMP < TMIN, RP = 0.0 If TEMP > TMAX, RP = 1.0 If TMIN < TEMP < TMAX, RP = (TEMP - TMIN)/(TMAX - TMIN) where: TEMP is mean weekly air temperature (°F), TMIN is a critical minimum temperature, below which all precipitation is snow, TMAX is a critical maximum temperature, above which all precipitation is rain, -and RP is the fraction of input; morsture that falls as rain. 4. Rainfall interception loss (RINT, in). Vegetative canopies are known to inter- cept and retain a fraction of rainfall that is ultimately evaporated back to the atmos- phere. The amount of rainfall greatly influences the amount of net moisture (moisture entering the soil) for individual storms; estimates of yearly interception losses are as follows: grass-forb, 9 percent; aspen, 12. percent; and conifer, 20 percent (Helvey 19 Johnston 1971; and Verry 1976). The fraction of moisture received as rainfall that may be considered interception loss is assumed to be an average, weighted by areal cover of each vegetation type, of three rainfall interception storage coefficients (GSTR, ASTR, and CSTR, in/in). 5. Snowfall interception loss (SINT, in). Researchers have many different opin- ions about moisture loss from intercepted snow in coniferous canopies (Satterlund and Haupt 1970; Miller 1962). Estimates of the magnitude of such losses generally range between 6 and 10 percent of total snowfall (Anderson 1969). The amount of snowfall interception loss from leafless aspen is assumed to be relatively minor. The fraction of snowfall that becomes interception loss is defined in ASPCON simply as the weighted average of two interception loss coefficients, SNA (aspen) and SNG (conifer), with respective values of 0.01 and 0.07 in/in. The interception loss of snow by the grass- forb type is assumed to be zero. | Numbers correspond to the items presented in figure 1. 6. Snowpack evaporation (SVAP, in). Doty and Johnston (1969) found evaporative losses from snowpacks in winter as follows: open ground, 0.05 in/in, under aspen, 0.034 in/in, and under conifers, 0.026 in/in. A weighted average of three snowpack evapora- tive loss coefficients, GSV, ASV, and CSV, is assumed to be the fraction of snowfall that is evaporated during the year. 7. Snowpack accumulation. Research suggests that vegetative canopies influence snowpack in western watersheds (Gary and Coltharp 1967: Thies 1972: Dunford and Niederhof 1944; and Meiman 1970). Accordingly, snowpacks in the model are accumulated differently for each vegetative type. Snownpack accumulation is assumed to be a fraction of total net snowfall for each community type (99 percent in grass-forb areas, 106 per- cent am aspen areas, and 95 percent.in conifer areas). This approach is simple yet does provide for a redistribution of snowfall within the watershed that is consistent with field observations. 8. Snowpack melt. Just as snowpack accumulation patterns vary between watershed cover types, the timing and rate of snowmelt may also be expected to change as a func- tion of vegetative succession. Snowpack ablation may be expected to begin first in an open area and last in a coniferous forest. Snowmelt rates should be about the same for open and aspen areas but significantly slower for coniferous tynes (Thies 1972; Federer and others 1972). Snowmelt in ASPCON is indexed by mean weekly air temperature in a manner similar to the Army Corps of Engineers (1960) model. Figure 2 shows that, for each vegetative type, the amount of snowmelt is a function of a melt rate coeffi- cient, GMC, AMC, and CMC (in/°F wk), and a base temperature coefficent, GBASE, ABASE, and CBASE (°F), for grass-forb, aspen, and conifers, as well as mean weekly temperature CE). 9. Channel inflow from snowmelt (QMCH, in). Part of each increment of snowmelt may be expected to occur on saturated soil adjacent to stream channels and, therefore, to readily enter the stream channel. The fraction of snowmelt thus contributing to streamflow is equivalent to the product of the amount of snowmelt and a melt inflow coefficient (TMCH, in/in). TMCH functions similar to ACHP and may be estimated from an areal map of a watershed. 10. Active moisture input. The term ''active moisture" is defined as the sum of nét weekly rainfall and snowmelt. Active moisture is capable of entry into the scil system (depicted as the large "'tank'"' in fig. 1) for subsequent evapotranspiration, deep percolation, or direct contribution to streamflow. Y = = a fra] = = ro) = YN Figure 2.--Snowmelt funettons for the grass-forb, aspen and conifer > 30 % 4 © 50 55 60 ee HOE HEE: MEAN WEEKLY TEMPERATURE (°F) 11. Overland flow when infiltration rate is exceeded (QXS, QOF, in). The model provides for calculating overland runoff when active moisture input exceeds infiltration capacity (FI, in/wk). This condition may occur when the soil is below saturation (QXS, a rare occurrence on subalpine watersheds) or when the soil is saturated (QOF, which occurs primarily during the spring snowmelt season). Because the model is incremented on weekly intervals, QXS cannot be estimated accurately. Consequently, the infiltration capacity is set at a sufficiently large value to preclude any QXS. The model may be set for any desired increment period, which could make QXS a more important hydrologic factor: 12. Transpiration (TRAN, in). The model treats evaporation of water via plant stomates (transpiration) and evaporation of moisture from the surface soil as two dis- tinct processes. To reflect the differences between grass-forb, aspen, and conifer communities that are suspected to infiuence TRAN, the following relationship is assumed: TRAN = f (potential evapotranspiration, seasonal plant activity, plant rooting depth, community crop coefficient). a. Potential evapotranspiration (PET, in) is calculated according to the model described by Blaney and Criddle (1962). b. Plant activity index (PAI). Although aspen and conifers have been shown to be comparable in terms of end-of-season soil profile moisture content (Brown and Thompson 1965), there is little direct research that describes relative year-round- consumption patterns. Several researchers have found that conifers may actively trans- pire water at times of the year when deciduous tree species are dormant (Swanson 1967; Owston and others 1972; Smith 1975; and Urie 1959). Accordingly, a plant activity index (PAI, fig. 3) is defined as that fraction of peak activity that a plant community - CONIFER PLANT ACTIVITY INDEX j=) GRASS -FORB ASPEN MONTH Figure 3.--Plant activity index for the grass-forb, aspen, and contfer types. Figure 4.--Effect 1.0 of limiting soil motsture on potenttal evapotransptratton. 0.5 ACTUAL ET/PET PERMANENT 0.5 x FIELD FIELD WILTING POINT CAPACITY CAPACITY may reach when water is not limiting growth. The PAI is thus defined to reflect the week-to-week influence of day length and soil temperature on a plant's ability to transpire water. AVCOGreCk1oOn 1s appived to PET to account for the effects of limiting soi] moisture on transpiration. The relationship outlined in figure 4 adjusts PET according to the following rule: PET' = PET x (SM-PWP)/AWH where: SM is volumetric soil moisture content (in), PWP is water content at permanent wilting point (in), and AWH is one-half of the profile's available water or the difference between the water content at field capacity (FC, in) and PWP. The adjustment of PET for limiting soil moisture is made according to a model by Hanks (1976), which is similar to the approach taken by Leaf and Brink (1975). c. Plant rooting depth (RDP). The capacity of different plant communities to occupy the root zone and the differences in mean soil depths for different watersheds are reflected in a plant rooting depth coefficient. The RDP is defined as that fraction of the total available rooting zone in the soil profile that contains 90 percent of all live plant roots. d, Community crop coefficient. (CC). The crop coefficient 1s included. in the: model to reflect differences in consumptive use rates of water by different vegetation types when all other factors are held constant. The grass-forb community is given the value of 0.9. Although forested communities may be expected to transpire greater amounts of water than nonforested areas, it is questionable whether crop coefficients for aspen and coniferous forests should be different. Unlike coniferous forests, aspen forests gen- erally have a highly productive understory which contributes to transpiration losses. However, coniferous forests have a larger leaf area index and increased quantity of aboveground biomass than do aspen forests. As a result of the above mediating consider- ations, the crop coefficients for aspen and conifer types are set at 1.25. Watershed transpiration loss is weighted according to areal vegetation composition and is calculated as the product of PET', PAI, RDP, and CC values. 13. Evaporation of rainfall from surface soil (RVAP, in). The model allows for a portion of rainfall to be evaporated from the surface soil. Generally in these forests, rain that falls during the growing season readily evaporates after each storm and seldom contributes to soil moisture recharge. A function was synthesized to reflect this phenomenon: AK = RAIN/PET If AK > 1, then AK is assigned the value of 1 RVAP = RAIN - (RAIN x AK) where: RAIN is net rainfall (in). The value of RVAP is then subtracted from the soil moisture content. As a consequence of defining RVAP as a function of rainfall amount as well as PET, significant amounts of rain are evaporated from the soil only during the growing season. 14. Soil profile interflow (QF, in). When soil moisture is above the water con- tent for field capacity (FC, in), moisture may move laterally through the soil profile until it reaches the stream channel. Soil moisture in excess of field capacity is multiplied by an interflow coefficient (FQF,; in/in) to define interflow. 15. Deep percolation (QI, in). The quantity of water that percolates through the soil profile and enters the ground water reservoir is calculated similar to QF except that a deep percolation coefficient (FK, in/in) is applied instead of FOQF. 16. Deep seepage (SEEP, in). A portion of the water entering the watershed may leave the area without contributing to local streamflow. In other words, a fraction of moisture is routed via deep seepage into aquifers. The deep seepage storage compart- ment receives moisture when the ground water level reaches a certain maximum (TOP, in). When this maximum is reached, the ground water level is multiplied by a deep seepage coefficient (DPSP) to calculate the amount of water added to SEEP. 17. Subsurface flow from ground water storage (QGW, in). The amount of water entering the stream channel from the ground water reservoir is defined as the product of the ground water level and a ground water recession coefficient (AGW, in/in). 18. Channel routing of flow. Moisture for streamflow that is generated by the model may be expected to experience a timelag before passing through the gaging station at the mouth of the watershed. Therefore, the model provides for fractions of generated runoff to be delayed up to 5 weeks. ASPCON computes weekly and yearly water budgets by summing all components of streamflow, evapotranspiration, and changes in soil moisture and ground water storage. MODEL CALIBRATION The model was calibrated for an ''average'' water-year on the West Branch Chicken Creek Watershed (CCW), Davis County Experimental Watershed in Utah. The present vegeta- tion status on the 217-acre CCW is approximately 20 percent grass-forb, 78 percent aspen, and 2 percent conifer (Johnston and Doty 1972). A total of 47 inches of precipi- tation fell during the modeled year, of which 11.6 inches was rain and 35.4 inches was 1 Figure 5.--Mean weekly temperatures and aN evapotransprration POTENTIAL y) v for the Chicken Creek ) \ Watershed. EVAPOTRANSPIRATION ¢ POTENTIAL EVAPOTRANSPIRATION (IN/WK) MEAN WEEKLY TEMPERATURE (°F) snow. Average soil profile depths to limiting horizons were assumed to be 5 feet. Aver- age weekly temperatures and potential evapotranspiration fluctuated according to the patterns shown in figure 5. A series of annual hydrographs for observed CCW streamflow were analyzed and the model coefficients in table 1 were adjusted until a predicted hydrograph was produced that agreed closely with past watershed behavior. During the Calibration’ process, the*onilly coefficients to be adjusted were those coefficients not easily estimated from a knowledge of watershed characteristics but to which the model Tsesensitive.. Mable 2 presents the values for model- coefficients set according to, the best available knowledge from the literature. The purpose of this calibration procedure is not to model CCW, but to develop a reasonable point of reference against which hydrologic changes attributable to vegetation changes may be estimated. Once an acceptable hydrograph was obtained, all coefficients except the vegetative cover parameters were held constant throughout the remainder of the study. Thereafter, the areal cover of vegetative types on the watershed (CVG, CVA, and CVC, table 2) was sequentially altered to simulate the entire grass-forb to aspen to conifer sere. Water- shed response to relatively wet and dry years was examined for five different vegetative combinations by increasing or decreasing the amount of annual precipitation. Input: wet year = 58.8 in (125 percent of normal), dry year = 35.3 in (75 percent of normal), and -drought@year = 23.5) in,.(50. percent of normal). Symbol : FI ACHP. TMCH GSTR ASTR CSTR SNA SNC GSV ASV CSV ACCG ACCA ACCC CCG CCA CCC DPG DPA DPC CVG CVA CVC Soil moisture of Soil moisture of Soil moisture of Table 1.--Model coeffictents mantpulated during caltbratton Definition S-ft profile at beginning of water-year S-ft profile at saturation S-ft profile at field capacity Pp wp p Soil moisture of a S-ft profile at permanent wilting point Fraction of soil water (SM>FC) becoming interflow Fraction of soil water (SM>FC) becoming deep percolation Groundwater reservoir recession fraction Deep seepage to aquifers from ground water fraction Maximum ground water level Channel routing coefficients Snowmelt initiation temperature for the grass-forb types Snowmelt initiation temperature for the aspen type Snowmelt initiation temperature for the conifers type Melt rate index for the grass-forb type Melt rate index for the aspen type Melt rate index for the conifer type Critical maximum temperature for precipitation Critical minimum temperature for precipitation Units Simulation value Table 2.--Model coeffictents held constant during caltbrattion Definition Infiltration rate Fraction of precipitation intercepted by the stream channel Fraction of snowmelt readily entering stream channel Vegetation storage of rainfall for the grass-forb type Vegetation storage of rainfall for the aspen type Vegetation storage of rainfall for the conifer type Snowfall Snowfall Snowpack Snowpack Snowpack Snowpack Snowpack Snowpack interception loss fraction for aspen interception loss fraction for conifers evaporation fraction for grass-forb evaporation fraction for aspen evaporation fraction for conifers accumulation factor for grass-forb accumulation factor for aspen accumulation factor for conifers Crop coefficient for the grass-forb type Crop coefficient for the aspen type Crop coefficient for the conifer type Rooting depth coefficient for the grass-forb type Rooting depth coefficient for the aspen type Rooting depth coefficient for the conifer type Fractional area of watershed occupied by grass-forb Fractional area of watershed occupied by aspen Fractional area of watershed occupied by conifers 10 Units in/wk in/in in/in in/in in/in in/in in/in in/in in/in in/in in/in in/in in/in in/in Assumed value 13. 24. PAs 10. 34. 42. PREDICTED HYDROLOGIC IMPACT OF SUCCESSION Predicted weekly water budgets for the CCW were found to reflect complex inter- For example, the upper portion of figure 6 actions among assumed hydrologic processes. illustrates when rain and snow were received on the watershed, the lower portion of the The figure shows how vegetation affects the timing of moisture entry into the soil. timing and amount of active moisture input is a function of snowpack melt rates and 6A SNOWFALL Hittin qin 7 \3 Pas a 0 | > — - 4 oF pas eanent cqunneconaccnceueseunsnen @ "1a, = Thy PRECIPITATION (1N/WK) (SF uw CONIFER g = = Sei Re n eo ean) ke (>) = l 0 (Re Diee a 2 bat Woe A oN Sos 6 RA oe SE 40 MONTH Figure 6.--Precipitatton and active moisture input for the Chicken Creek Watershed. 11 > \S GRASS-FORB \Fn ISU TRANSPIRATION (IN/WK) fan) a MONTH Figure 7.--Weekly transptration patterns for the Chicken Creek Watershed when dominated by: grass-forb, aspen, and contfers. evaporative losses (interception and soil moisture evaporation). Figure 7 presents the patterns of consumptive water use when the watershed is dominated by grass-forb, aspen, and conifer types. Greater consumptive use rates for conifer-dominated conditions may be attributed to the plant activity patterns of evergreen canopies. The combined effects of active moisture input, transpiration, and other components of the hydrologic cycle are reflected in the hydrographs in figure 8. Although the timing and magnitude of runoff under different types of vegetation cover vary substantially during the melt season, streamflow before and after the melt season is similar for all types. Dominance of aspen on a formerly grass-forb watershed causes spring runoff to be delayed slightly with lower peak flows. Spring runoff under conifer-dominated conditions is even further delayed and reduced. 2.0 Figure 8.--Spring runoff hydrographs for the Chicken Creek Watershed when dominated by: grass-forb, aspen, and contfers. STREAMFLOW (IN/WK) Table 3.--Water budget components for an average water-year on the CCW at different stages of suecesston Vegetation!: status 98-1-1 90-9-1 80-19-1 70-29-1 60-39-1 50-49-1 40-S9-1 30-68-2 20-78-2 20-68-12 20-58-22 20-47-33 20-36-44 20-25-55 20-15-65 20-5-75 Inehes Zils Zee 20. 20. 20. 20. 20. 20. 19. 18. 16. 16. 16. Se AL Sy PS NOH WrRNUONWOODONWWNANOWW Inches 25% BBS 22). Zu Ze ZN 20. 20. al) Lys aye 16. KOK 16. L3% Udi. :Streamflow: : Streamflow: +ASM :Runoff2 WDRFONONWOONDHLUONOW 49, 48. 47. 46. 45. 44, 43. 42. 42. 38. 36. 55! 34. 34. 53\. Via Percent DNWORrFNNWADONUA OND MPrRNONNNNN WW WwW eats isl. ce} te WAM ONION AAWHEANWOrW ~ DOMrWANOL HA HKWHR AO ee oe) BSN Ge © roo oO OQ SHG Oon © - RINT AGWL SEEP -TRAN SINT SVAP Inehes- - --------------- 1.8 6.9 9.6 1.0 0.0 lew, Liss 6.9 9.9 ileal ell Ved 1 649) - 1055 ea aul 6 1.9 6.6 Thi apa Bue 15/6 alte! G6 a7 52 ae ners) 1.8 6.6 127.0 Wee sie 1:5 1.9 6.6 12757 Nee ais 1.4 1.8 66) —1)26:9 1.3 a) 1.4 Then? 6.5: Aisi l 13-3 aS ie.) Wed S79. Se 2 1.4 he) Meas: 1 6 Sop lbh 1.35 a eZ 1.8 So) 16).8 1.6 1.0 are ae Se) oi! Tina 132 alge) | 1.8 4.9 16.9 8 La! eel 16 Sit Al. 1.9 le Ted: 28 ALO 7 54: 2.0 1.9 h.0 lPercent watershed areal cover composed of grass-forb, aspen, and conifer types, respectively. 2Runoff percent is equal to (Streamflow + SM)/precipitation. 3See figure 1 for an identification of alphabetic codes. 4ASM and AGWL represent the net annual change in soil moisture and ground water level, respectively. Predicted annual water budgets for the "average'' year for different combinations In the CCW test area, the principal sere of vegetation types are given in table 3. following burning or clearcutting is visualized as less than 4 years' dominance by a grass-forb type, which is quickly followed by aspen dominance, which in turn is pro- gressively replaced by conifers. erass-forb climax, and thus stabilizes at this level. a position assumed for the watershed on the grass-forb to conifer sere. the sere ls not specified, since this may vary widely from site to site. for QCHP is a constant value (0.314 in) for all conditions. COnO 28 In) tor OMCH, 1.77 sto V.71 in for QGW, and 2.66 to 2.87 in for RVAP. Approximately 20 percent of the area is considered Each line in the table refers to The. length of The value The values for QMCH, QGW, and RVAP exhibited the following minor trends from beginning to end of the sere: 0.29 Several components of the water budget (TRAN, RINT, SINT, and SVAP) exhibited rather consistent trends along the sere: tion change with the timing and amount of moisture input and moisture loss due to evapo- transpiration. net change from the initial soil moisture at the end of the year will affect the follow- ing year's runoff (the soil storage compartment must be recharged prior to the runoff season). different stages of succession are illustrated in figure 9. the watershed, a net reduction in water available for streamflow of 3.4 in has occurred. As the watershed proceeds from aspen to climax conifer conditions, ines. Lost: Annual streamflow under a variety of precipitation conditions Variable substantially along the successional gradient (table 4). to alter the efficiency with which the watershed generates runoff: efficiency accompanies years of below-average precipitation. The other values in table 3 reflect the interaction of vegeta- The value for streamflow plus soil moisture change is presented since The amount of streamflow reduction plus the change in soil moisture for By the time aspen dominates Late an additional 4.6 was found to vary precipitation appears decreased runoff seral stages (conifer dominance) accentuate the reductions in streamflow for relatively dry years. = = 2 STREAMFLOW REDUCTION + ' a a be = a = 2 = Lo SI a = rn = > ra o) = S) ”n ao = ASPEN COVER Ft, Coldinse: Color Gary,-H. bus “and G.>B2‘Coltharp: 1967. Snow accumulation and disappearance by aspect and vegetation type in the Santa Fe Basin, New Mexico. USDA For. Serv. Res. Note RM-93. Rocky Mt. For. and Range Exp. Stn. 5" Ft. -Collinss;: Color Green, A. W., and T. S. Setzer. 1974. The Rocky Mountain timber situation, 1970. USDA For. Serv. Resour. Bull. INT-10.. Intermti" For. and Range Exp... Stn.,, Ogden, "Utah. Hanks, J. R. 1976. USU Soils 566, Physical Properties of Soils Laboratory Materials. Utah State Univ., Logan. Helvey, J. D. 1971. Summary of rainfall interception by certain conifers in North America. In Biol. Effects’ in the’ Hydro. Cycle,.-p:- 103-113. Proc." 3rd" ints Symp. on Forest Hydrology Prof., Purdue Univ., Ind. huts, Di De 1971. Ecological hydrology: ,-in Biol. Effects™in: the’ Hydro:-Gycle; “p? 18-30.) Proc: 3rd Int. Symp. on Forest Hydrology Prof., Purdue Univ., Ind. Johnston, ~R2-S 1969. Aspen sprout production. and water use. USDA. For. Serv.Res. Note INT-389: Intermt. For. and Range Exp. Stn., Ogden, Utah. Johnston, R. S. 1971. Rainfall interception in a dense Utah aspen clone. USDA For. Serv. Res. Note INT-145. Intermt.. For. and Range Exp. Stn. ,. Ogden, Utah. Johnston, R. S., and R. D. Doty. 1972. Description and hydrologic analysis of two small watersheds in Utah's Wasatch Mountains. USDA For. Serv. Res. Pap. INT-127. Intermt. For... and Range Exp, Stim; Ogden, Utah. Leaf, C. F., and G. E. Brink. 1975. Land use simulation model of the subalpine coniferous forest zone. USDA For. Serv. Res. Pap. RM-135.. Rocky Mt.. For: and Range Exp. ..Stn.,,Ft,. Collins5. Colo; Meiman, J. R. 1970. Snow accumulation related to elevation, aspect, and forest canopy. Jn Snow Hydrology, p. 35-47. Proc. Workshop Seminar, New Brunswick Univ., Fredericton. Miller, D. H: 1962. Snow in trees--where does it go? West. Snow Conf. Proc. 30:21-29. Mueggler, W. F. 1976. Type variability and succession in Rocky Mountain aspen. Jm Proc. Utilization and Marketing as Tools for Aspen Management in the Rocky Mountains Symp., p. 16-19. USDA For. Serv. Gen. Tech. Rep. RM-29. Rocky Mt. For. and Range Exp. Stn., Ft. Collins; Colo, Owston, P. W., J. L. Smith, and H. G. Halverson. 1972. Seasonal water movement in tree stems. For. Sci. 18:266-272. 16 Rilevaec bas and Reon Hawiernis'. 1975. Hydrologic modeling of rangeland watersheds. In Proc. 5th Workshop of the U.S./Australia Rangelands Panel, Boise, Idaho, p. 123-138. Sattermliund., De R= 1972. Wildland watershed management, p. 145. Ronald Press, New York. Satterlund? Da Re, wand iH: F. Haupt. 1970. The disposition of snow caught by conifer crowns, Water Resour. Res. 6(2) :649-652. Smith, J. LL. 1975. Water yield improvement research of the Pacific Southwest Forest and Range Experiment Station and its usefulness to wildland resource management. Proc. Lake Tahoe Res. Seminar IV, South Lake Tahoe, Calif., p. 3-24. Swanks We dies and J, BasDouglass,. 1974. Streamflow greatly reduced by converting deciduous hardwood stands to pine. Science 185(4154) :855-859. Swanson, R. H. 1967. Seasonal course of transpiration of lodgepole pine and Engelmann spruce. Im Int. Symp. on Forest Hydrology, p. 419-434, (Sopper and Lull, eds.) Pergamon Press, Oxford. Wowleshy Ia Ike 1972. The effects of elevation and vegetation type on snow accumulation and melt in Logan Canyon. Utah. Utah State Univ., Logan, Master's Thesis, Wrene, IDs lle 1959. Pattern of soil moisture depletion varies between red pine and oak stands in Michigan. USDA Lake States For. Exp, Stn., Tech. Note 564. Wester IDs tale 1967. Influence of forest cover on ground-water recharge timing and use. In Int. Symp. on Forest Hydrology, p. 313-324, (Sopper and Lull, eds.). Pergamon Press, Oxford. Very (Bie) OS: 1976. Estimating water yield differences between hardwood and pine forests: an application of net precipitation data. USDA For. Serv. Res. Pap. NC-128, North Gentral For. and-Range Exp. Stn., St. Paul, Minn. V7 Headquarters for the Intermountain Forest and Range Experiment Station are in Ogden, Utah. Field programs and research work units are maintained in: Billings, Montana Boise, Idaho Bozeman, Montana (in cooperation with Montana State University) Logan, Utah (in cooperation with Utah State University) Missoula, Montana (in cooperation with University of Montana) Moscow, Idaho (in cooperation with the University of Idaho) Provo, Utah (in cooperation with Brigham Young University) Reno, Nevada (in cooperation with the University of Nevada) vy U.S. GOVERNMENT PRINTING OFFICE:1978—777-095 / 22 Jaynes, Richard A. 1978. A hydrologic model of aspen-conifer succession in the west- ern United States. USDA For. Serv. Res. Pap. INT-213,17 p. Intermt. For. and Range Exp. Stn., Ogden, Utah 84401. Hydrologic impacts of grass-forb to aspen to conifer succession in the Rocky Mountain area are simulated by means of a fundamental model. Model algorithms representing hydrologic processes are sensitive to vegetational changes within the subalpine vegetation zone. Reductions in water yield are predicted as the vegetation on a small Utah watershed proceeds from a grass-forb type to aspen to conifers. Streamflow changes are largely attributable to an interaction between seasonal con- sumption for each vegetation type and the influence of vegetation type on snowpack. KEYWORDS: ecosystem hydrology, succession, Populus tremuloides, hydrology model. Jaynes, Richard A. 1978. A hydrologic model of aspen-conifer succession in the west- ern United States. USDA For. Serv. Res. Pap. INT-213,17 p. Intermt. For. and Range Exp. Stn., Ogden, Utah 84401. Hydrologic impacts of grass-forb to aspen to conifer succession in the Rocky Mountain area are simulated by means of a fundamental model. Model algorithms representing hydrologic processes are sensitive to vegetational changes within the subalpine vegetation zone. Reductions in water yield are predicted as the vegetation on a small Utah watershed proceeds from a grass-forb type to aspen to conifers. Streamflow changes are largely attributable to an interaction between seasonal con- sumption for each vegetation type and the influence of vegetation type on snowpack. KEYWORDS: ecosystem hydrology, succession, Populus tremuloides, hydrology model.