Thomas L. Thurow and Justin W. Hester
Introduction
Increased dominance of juniper in what had previously been grasslands or savannas substantially alters the fate of precipitation on rangelands. This has very important ramifications because water is a direct or indirect limiting factor to all aspects of production on semi-arid regions. Therefore, understanding how juniper effects water movement through rangelands is necessary to understanding the ecology of the site. The impact of juniper encroachment on recharge of streams and aquifers is also a very important consideration since rangeland watersheds are the primary source of water for much Texas (TWDB 1990). The objective of this paper is to provide an overview of the various routes of water movement through rangeland and then explain how an increase or decrease in juniper cover alters the amount of water flowing through each of these routes.
Concepts of Water Balance
The hydrologic cycle is the process of water movement through the environment. The primary avenues of inflow (precipitation), transit (interception, stemflow, overland flow, infiltration, percolation) and outflow (evap-oration, transpiration, runoff, deep drainage) of water movement are outlined in Fig. 1. Basically,
Water inflow = water outflow + storage. The presence or absence of juniper will influence the transit and outflow paths that water will follow.
Water enters a system in several ways. Precipitation is the major input for most areas. The amount, intensity, duration and seasonal distribution of rainfall are important components of an area’s hydrology. Also, water can flow into one area from another, either over the soil surface or below the soil surface. Dew (water vapor condensation on vegetation or soil) is generally a negligible portion of water inflow except in coastal or montane regions.
As precipitation reaches the earth’s surface, it either strikes objects such as vegetation, litter or rocks, or falls unimpeded to the ground. Interception is the term used to describe raindrops striking something before reaching the soil. Interception is important because much of the energy associated with the falling raindrop is dissipated as it strikes and object and thus the erosive force to the soil is reduced. Raindrops striking bare soil are by far the most important mode of soil erosion. Therefore it is very important that bare soil be minimized through maintenance of vegetation cover.
Another reason interception is important is because a significant portion of annual precipitation may never reach the ground, instead adhering to the vegetation and being lost directly to the atmosphere via evaporation. The water-holding capacity of the intercepting object is governed primarily by the characteristics of the storm (e.g., intensity, wind speed) and by its surface area and physical characteristics (e.g., rough or hairy leaves). The percentage of the rainstorm that never reaches the soil is termed interception loss.
A portion of the intercepted water runs down the plant stems (stemflow) or may drip from leaves and fall to the soil (throughfall), resulting in a redistribution of water reaching the surface. This redistribution of water has important ecological implications because some areas under the tree, such as around the base of the trunk, receive substantially more water and other areas under the canopy may receive substantially less water than indicated by the annual precipitation data. Such a redistribution of water would give the tree an advantage that would come at the expense of plants trying to grow under the drier portions of the canopy.
The water reaching the surface either runs off the site, infiltrates into the soil or is stored in surface depressions and eventually evaporates. Infiltration is the process by which water moves into the soil. Water generally moves into the soil quite rapidly at first, and then the rate of entry slows down as the soil becomes saturated. When the soil surface reaches saturation, infiltration stabilizes at a rate called the terminal infiltration rate. This rate is dependent on a variety of soil and vegetation characteristics. A low terminal infiltration rate is an indication of poor soil structure. Soil structure describes how soil particles are arranged. Soil particles are held together in individual clusters called aggregates. The size and amount of soil pores (air spaces between aggregates) is a function of soil texture and degree of aggregation. The porosity of the soil determines the rate of infiltration. Vegetation and litter protects the soil surface from the energy associated with direct raindrop impact, which can destroy soil aggregates. Decaying matter provides the building blocks which help make strong aggregate bonds. When stability of the soil aggregates is poor, clay particles are readily detached from the aggregate. This causes the aggregate to break apart and thus create smaller size pores. Also, detached clay particles can clog the soil pores resulting in a thin, sealed layer at the soil surface that substantially reduces the infiltration rate. This is a common way that soil crusts are formed.
“A stop-gap approach often used in an attempt to manage crusted soils is to concentrate grazing so that the soil surface is disrupted by hoof action. Livestock trampling does indeed break the crust, incorporate mulch and seeds into the soil, and aid seedling emergence. However, this result is short-lived because the subsequent impact of falling raindrops re-seals the soil surface (i.e., the unstable soil pores will become plugged) after the first several minutes of an intense rainstorm. To effectively address a soil crusting problem, livestock grazing systems must concentrate on addressing poor aggregate stability which is the cause of the crusting. Livestock grazing systems that promote an increase in plant and litter cover and an increase in organic matter produce the only lasting effect in reducing soil crusts” (Thurow 1991).
Some of the water which does not infiltrate into the soil may be temporarily stored in soil depressions. This surface detention is a function of the roughness of the soil surface, slope of the surface, soil texture and structure, and soil depth. Runoff begins when the amount of water at the soil surface exceeds the infiltration rate and storage capacity of depressions.
Water which infiltrates into the soil will either be eventually returned to the atmosphere via evapotranspiration, seep laterally until it re-emerges from the ground or seep into an aquifer. Evapotranspiration consists of two separate processes: transpiration and evaporation. Transpiration occurs when water vapor is released to the atmosphere through permeable membranes of living organisms such as the pores in the leaves of plants. The amount of transpiration is largely dependent on the surface area of leaves which contain these pores, called stomates. Leaf area and transpiration rate vary between different types of vegetation. Water loss per unit area of leaf tissue of herbaceous vegetation such as grass or forbs is usually less than from trees or shrubs. This is due to physical and physiological differences which determine growth dynamics and the efficiency of water use. Evaporation occurs when water vapor enters the atmosphere directly from the soil or from ponded water. Evaporation from the soil is influenced by soil texture, vegetation cover, and climatic factors.
Deep drainage is water movement down through the soil, below the rooting zone of plants. The amount of deep drainage depends on the amount of infiltration, the evapotranspirational demand, and the soil transmission characteristics. Caliche layers or hardpans, which are cement-like layers in the soil, can reduce the amount of deep drainage to almost nothing.
EFFECTS OF JUNIPER INCREASE OR CONTROL ON HYDROLOGIC PROCESSES
The degree to which juniper can influence water distribution over a landscape can be conceptually evaluated by juniper effects on the hydrologic cycle. The vegetation and soil characteristics of a site are the prime determinants influencing how water is partitioned to each of the potential pathways of water movement. Understanding how juniper can affect these factors is essential to anticipating how water outflow will respond to the increase or control of juniper.
Interception loss
An increase of juniper cover on rangeland can reduce the amount of precipitation that reaches the soil surface. The extent to which the type of canopy cover influences the amount of precipitation that reaches mineral soil is illustrated in Fig. 2. The interception loss associated with the canopies of redberry juniper (Juniperus pinchotii) and Ashe juniper (Juniperus ashei) was 25.9% and 36.7% of gross precipitation, respectively (Hester 1996). Ashe juniper has a much denser canopy and thus has more surface area on which precipitation can adhere and then be lost to the atmosphere via evaporation. The interception loss of the redberry juniper canopy was similar to the 25.4% interception loss from the canopy of live oak (Quercus virginiana) (Thurow et al. 1987). Most of the small rainfall events (<5mm) do not reach the litter layer because of water retention by the foliage. Once canopy storage capacity was reached, percent throughfall increased by way of increased drip loss.
Water that passes through the canopy must also pass through the litter layer prior to entering the soil. The deposition of leaves, twigs, and branches, coupled with the absence of fire and the resistant nature of the litter to decomposition results in a large accumulation of litter under juniper trees. This litter has a high capacity to retain water falling through the tree canopy. The amount of interception loss associated with the litter layer is considerably greater for redberry juniper (40.1%) and Ashe juniper (43.0%) than previous studies on western juniper species, 2-27% (Young et al. 1984), and live oaks, 20.7% (Thurow et al. 1987). This difference is mainly attributable to the greater amount of redberry juniper and blueberry juniper litter build-up.
As a result of interception loss via the canopy and litter, only 20.3%, 34.0% and 53.9% of annual rainfall reaches mineral soil under the canopy of Ashe juniper, redberry juniper and live oak, respectively. This is a stark contrast to the 81.9% and 89.2% of annual precipitation that reaches the soil under bunchgrass and shortgrass cover, respectively. Clearly, the type of vegetation cover, and the associated evaporation loss of intercepted water, strongly influences the amount of water that reaches the soil. This has important ramifications for stream and aquifer recharge because as vegetation cover changes from grasses to trees a greater proportion of precipitation will leave rangeland via evaporation and therefore less precipitation is available for producing herbaceous forage or for deep drainage or runoff.
Stemflow is precipitation that has adhered to the vegetation cover and channeled by the branches to the trunk of the tree. Stemflow does not begin from either redberry or Ashe juniper until precipitation exceeded 3 mm. Stemflow volumes for redberry and Ashe juniper were 7.1% and 5.1% of precipitation, respectively. These values are similar to previously obtained results on juniper in other regions. Stemflow accounted for 9% of annual precipitation in the pinyon-juniper region of Arizona (Collings 1966). Alligator juniper (Juniperus deppeanan) and Utah juniper (Juniperus osteosperma) yielded less stemflow (1-2%) due to shaggier bark (Skau 1964). Trees with smoother bark with wide funnel-shaped canopies tend to have the greater amount of stemflow (Scatena 1990). Other factors which influence stemflow are leaf type and position, total branch area, and branch position and angle (Návar 1990). With that in mind, these tree characteristics explain the significant stemflow differences per canopy area between redberry and ashe juniper. First, the bark of redberry juniper was somewhat smoother than the bark of Ashe juniper. Furthermore, redberry juniper on average is a multi-stemmed tree with fairly straight, highly inclined branches. In contrast, ashe juniper is often a single-trunked tree with a large canopy comprised of many horizontally to downwardly-declined branches, a canopy architecture that will cause many drip points.
Although the amount of stemflow may seem small, it may be extremely ecologically significant. At the base of the trunk, stemflow is responsible for the redberry and Ashe juniper trees receiving 470% and 462%, respectively of annual precipitation. This additional available water channeled to the base of the tree via stemflow was documented to be taken up by the concentration of fine roots adjacent to the trunk in a study of western juniper (Juniperus occidentalis) (Young et al. 1984). This harvested water supply may carry with it high nutrient concentrations leached from the canopy. Hence, given the water constraints in the semi-arid Edwards Plateau region, additional nutrient-rich water may competitively favor the growth and further establishment of juniper species.
Infiltration
Infiltration rate is dependent on a variety of vegetation factors, especially type and amount of vegetation present. Because cover values are generally greater on woody dominated site, infiltration rates are often observed to be highest under trees and shrubs, followed in decreasing order by bunchgrass and shortgrass sites (Blackburn 1975, Wood and Blackburn 1981, and Thurow et al. 1986). Cover breaks the erosive force of raindrops and the litter buildup obstructs runoff. The litter also contributes to building better soil structure which maintains large stable pores in the soil through which water can pass. For these reasons, infiltration rates were higher underneath the canopies of Ashe juniper than either bunchgrass or shortgrass sites (Hester et al. 1997). These results are comparable to studies on pinyon-juniper rangelands where infiltration rates were strongly related to vegetation cover and runoff originating underneath the shrub was either non-existent or markedly less than runoff originating on an area dominated by grass.
These data appear to contradict the common observation that runoff and erosion often increases when juniper invades a pasture. How can this occur if the area under the juniper is characterized by an improved infiltration rate (i.e., reduced runoff rate)? In a grazed pasture with shallow soils juniper can out compete grasses in a for limited resources (Dye et al. 1995, and Hester et al. 1996). Couple this with an increase in grazing pressure because livestock are utilizing a diminishing forage base, and it is understandable why an increase in juniper often goes hand-in-hand with a decrease in protective cover in the tree interspaces. As the amount of bare soil exposed to raindrop impact increases, there is an increased likelihood of degradation of soil structure and detachment of soil particles. Indeed, it has been documented that most runoff and sediment produced from pinyon-juniper rangelands originates from interspaces (Bolton et al. 1991).
Herbaceous Production
Understory plants may be negatively affected by juniper-induced reduction in light, soil moisture, and soil nutrients. Increases in juniper density and size has the effect of reducing understory plant cover and productivity, with desirable forage grasses often being most severely reduced (Eddleman 1983).
Juniper is an evergreen with extensive lateral and deep roots, and it has physiological adaptations which enable it to extract water from very dry soil. In addition, juniper also has a dense mat of fibrous roots at the soil surface. These traits make juniper a very strong competitor with grass for water, both underneath the canopy and in the tree interspace. Numerous studies have documented significant reductions in grass production due to increased woody dominance underneath the canopies (Engle et al. 1987, Armentrout and Pieper 1988, Ko and Reich 1993, Dye et al. 1995, and Hester 1996). However only Dye et al. (1995), and Hester (1996) noted reductions in grass productivity in the tree/shrub interspace. This may be due to the more shallow, xeric-type of soils that were present at these two study sites. When juniper roots grow into the spaces between trees to compete for water and nutrients, this competition, especially when combined with grazing pressure, puts extreme stress on the grass. This can explain why grass in the juniper interspaces may often be sparse, lack vigor and are difficult to reestablish as long as the juniper is present. Another competitive advantage of juniper is that it is not as quickly affected by drought as herbaceous species because the trees have a deep root system which gives them access to a water source that the herbaceous species cannot effectively tap.
Evapotranspiration
In a Texas shrubland water balance, evapotranspiration is by far the greatest route of water outflow, accounting for 80-95% of water loss (Weltz 1987, Carlson et al. 1990). Evaporation from the soil is generally limited to the top 12 inches unless there is severe cracking of soils. Trees and shrubs such as juniper create a microenvironment under their canopies which tends to reduce evaporation from the soil surface due to shading and protection from wind, etc. However, in the interspaces between juniper, where more bare ground may be exposed due to severe competition between juniper and herbaceous plants, evaporation from the soil will be greater than if a good herbaceous cover existed. As previously discussed, intercepted water adhering to a plant or litter will return to the atmosphere via evaporation. Juniper vegetation and litter provides much more surface area for water to adhere on than grass, therefore a much greater percentage of precipitation will be lost to evaporation in a juniper woodland than in a grassland. Evaporation of intercepted water occurs so quickly that the amount of transpiration reduction in a plant is negligible.
The amount of water lost from a system via transpiration is dependent on the amount and type of vegetation present. Juniper has a very extensive rooting system compared to herbaceous plants, and therefore when it invades a site, it has access to a greater volume of soil water. It is an evergreen, and thus also has the capability to transpire water yearlong. Many aridland shrub species have physiological or physical adaptations which allow them to withdraw water from much drier soils than can grasses. Therefore, juniper can continue to remove water from the soil long after grasses have gone into a drought or temperature induced dormancy.
Runoff
Although the litter layer markedly reduces the amount of water reaching the mineral soil, it does have some beneficial aspects as well. Due to the large amounts of organic matter and cover, it contributes to improved soil structure. This increases the infiltration rate capacity of the soil below the canopy. The thick litter layer often associated with juniper also minimizes evaporation loss from soil below the canopy and obstructs runoff that originates from interspace areas. This greater infiltration rate capacity enables the area under juniper trees to accept water inflow faster than the tree interspace which usually has a much lower infiltration rate capacity. Therefore, juniper groves will harvest water flowing off of interspaces. This gives the tree a further competitive advantage. This also explains why runoff yield from a pasture may not change as juniper density increases. Rather the water is simply redistributed within the pasture since water may flow from the interspace until it encounters a higher infiltration capacity of soils beneath a tree.
When the juniper tree is cut and removed, the soil structure, and the associated high infiltration rate, may be maintained for over a decade (Hester 1996). This explains why the area near the dripline usually has substantially greater forage production for many years after the tree has been cut. It also explains why runoff will not necessarily dramatically increase once juniper is removed. Rather, the water continues to infiltrate at high rates into soils previously ameliorated by junipers, thereby increasing deep drainage potential.
Deep Drainage
Juniper affects the deep drainage on rangelands because of its effect on all the other water balance components: 1) a large portion of precipitation never reaches the soil due to interception loss, 2) juniper extracts most of the water that does enter the soil to meet its transpiration requirements. The combination of less water entering the soil and strong ability by the juniper to extract water means that little water has a chance to drain beneath the root zone. Therefore, invasion of juniper on large areas that were once primarily grassland has strong implications for recharge of aquifers. It is a common occurrence in the Edwards Plateau of Texas to have seeps and springs stop flowing in conjunction with increases in juniper cover.
Erosion
Erosion reduces soil productivity and increases non-point source water pollution. Semi-arid rangelands have erosion rates which depend on the interactions of vegetation, soil, and storm characteristics (Bolton et al. 1991). In particular, the amount and type of vegetation present on rangelands has considerable influence on the amount of runoff. Sediment production is closely related to amount of runoff, and is therefore significantly correlated with vegetation parameters. Because juniper canopies are dense, they protect the soil directly beneath from raindrop impact and prevent detachment of soil particles. The heavy litter layer underneath juniper also impedes overland flow, thus reducing the transport capacity for sediment. But because juniper is highly competitive with interspace grasses, an increase in juniper cover often leads to increased exposure of bare soil. Any time bare soil is exposed, the potential for erosion is increased.
EFFECT OF JUNIPER REMOVAL ON WATER BALANCE
Rangeland management practices which affect vegetation cover and composition can affect both on-site and off-site water availability. There has been considerable speculation that brush management on rangelands can enhance water yields for off-site use. Theoretically, this could be accomplished by replacing deep-rooted species (such as shrubs and trees) with shallow-rooted species that consume less water (Davis and Pase 1977). The question of how much additional water could be made available through widespread shrub control is still unanswered for many rangelands. Hibbert (1983) estimated about 0.3 inches of additional water yield could be expected in Arizona for each 1 inch of annual precipitation in excess of 15 inches following elimination of woody species. This would occur from increased subsurface flow and ground water recharge. Data on the potential of brush control to increase water yields in Texas is scarce. Hibbert (1979) speculates that the following conditions should, in general, be met if brush management can be expected to increase water supplies for an area such as Texas: 1) annual precipitation should be greater than 18 inches, 2) brush removed should be replaced with grasses that use less water, and 3) replacement species should be shallow-rooted, deciduous, or have a low biomass.
Results of studies in Arizona and Utah indicate that brush management in the pinyon-juniper regions has a low potential for augmenting water supplies (Hibbert 1983, Blackburn 1983, Belsky 1996). In Utah, Gifford (1975) estimated the water balance for juniper-infested rangeland and for rangeland that had been cleared of juniper by chaining and managing the debris. Although there were differences in the water balance between these sites, there was no appreciable change in deep drainage or runoff, although management of debris did influence runoff on one site. The effect of juniper invasion and removal on the Texas rangeland hydrologic balance may be quite different from the studies cited above because Texas rangelands receive more rain and more intense storms which are thus more likely to cause runoff or deep drainage.
Development of ecologically sound land management practices requires a clear understanding of how potential practices will affect the hydrology of the site. Conversion of one vegetation type to another by brush removal is not the only consideration regarding water yields and water quality from rangelands. The removal method used can substantially affect the hydrology of the site as well.
Any disturbance which transmits energy to the soil, either through direct traffic of machinery or through the forces of raindrops, disrupts soil structure and reduces infiltration. Conversely, any practice which increases the standing vegetation and litter decreases runoff and sediment production. Therefore, the relative hydrologic impact of any brush control technique depends on the answer to several questions (Buckhouse and Bolognani 1982):
How severe is the soil disturbance as a result of this practice?
How successful was the response of herbaceous vegetation following the practice?
How effective was the control method in removing the brush species?
What was the impact of the practice on litter and ground cover?
How long ago was the practice implemented?
The effects of compaction generally are reduced over several years as the soil expands and contracts associated with wetting and drying, freezing and thawing, etc. Established vegetation may become more or less productive over time, depending on the adaptability of the species and subsequent management.
Techniques used to control brush on rangelands include: 1) mechanical, including root-plowing, chaining, shredding, raking, or hand-slashing; 2) chemical; 3) prescribed burning; and 4) biological, such as using goats and sheep to heavily graze browse species. The following is a summary of effects of common methods of juniper removal on site hydrology. Note that many of these studies primarily involve small plot responses. Few studies at a watershed scale of resolution have been conducted to date.
Mechanical Control
Mechanical methods of control for juniper generally include hand-slashing, chaining or cabling, and bulldozing. These are listed in the order of increasing degree of disturbance to the site. Hand-slashing is used for newly invaded sites and involves removal of the top growth only with minimal physical impacts to the soil and associated vegetation. Chaining results in moderate soil disturbance and creates pits where large brush species have been uprooted. Associated species may be bent or crushed but usually recover quickly. Bulldozing uproots the entire shrub or tree, and is generally used on mature trees or those juniper species which re-sprout. Soil disturbance is extensive locally, with destruction of soil structure and compaction. Large pits are created where the root mass is removed.
On pinyon-juniper rangelands in New Mexico, sites were clear-cut and slash was either removed, scattered, or burned (Wood and Jared 1991). Storms greater than 1 inch produced runoff following this control method. Runoff and sediment production from the slash-scattered treatment was less than when slash was removed. The slash-burned treatment caused the greatest amount of runoff and erosion, particularly in the year following the burn. The slash-scattered treatment was most desirable because it reduced surface runoff and erosion by providing the best vegetation and litter cover which protected the site from raindrop impact and provided an obstacle to overland flow.
Schott and Pieper (1987) examined the recovery of the vegetation community following cabling a pinyon-juniper rangeland in New Mexico. As mentioned above, the pattern of vegetation recovery and the types of vegetation and biomass produced following juniper removal have an important influence on both the short- and long-term hydrology of a site. The rate of recovery and extent of biomass produced was largely dependent on soil depth. However, the following pattern of succession was the same regardless of soil depth or site differences:
1) the first community to form following disturbance was composed of annual plants,
2) perennial grasses, forbs, and half-shrubs then began to move into the site and became dominate, and
3) juniper and other shrubs began to encroach and gradually took over the site.
Because cover on sites dominated by annual plants varies seasonally, the soil was well covered during portions of the year but quite bare during other times. Therefore, runoff and erosion were likely to increase directly after cabling because of exposure of bare soil on these annual plant communities. As perennials moved onto the site, vegetation cover increased and became more stable over time, but the patterns of water use also changed. The success of a brush removal technique is often judged by the extent to which reestablishment of the brush species is hindered.
Blackburn (1983) summarizes several studies conducted on watersheds in the pinyon-juniper region of Arizona. On one site, chaining removed 25% of junipers with no difference in subsequent water yields between chained and natural sites. On another site, juniper was either cabled or hand-slashed, with approximately a 100% kill rate. The pits created by cabling helped to capture overland flow, so that runoff was less on cabled than on hand-slashed sites. Total runoff was only slightly increased for treated sites as compared to untreated sites. Quality of water was slightly greater on hand-slashed than cabled sites because there was no soil disturbance.
In Utah, several studies evaluated the effects of chaining and debris management on site hydrology (Williams et al. 1969, Gifford et al. 1970). The treatments included chained with debris windrowed, chained with debris left in place, or untreated. Sites were seeded with perennial grasses following chaining. The chained and windrowed sites produced 1.2 to 5 times more runoff and 1.6 to 6 times more sediment than the undisturbed sites. The chained with debris-left-in-place sites produced equal or less runoff and sediment than the undisturbed sites. Windrowing the debris required additional use of mechanical equipment (greater soil disturbance) and also exposed large areas of bare soil. This caused infiltration to be lower on windrowed sites. On rangeland sites in Nevada, there was no difference between chained and seeded sites and undisturbed sites 5 to 11 years after treatment (Blackburn and Skau 1974).
Prescribed Burn
Fire is a natural component of Texas rangeland ecosystems and is a cost effective tool in controlling brush (Scifres 1980). However, storms that occur prior to vegetation regrowth can have a profound impact on soils and associated herbaceous vegetation. Under certain conditions which are not well understood, burning can increase the water repellency of soils inhabited by plant species which contain waxy coverings on their leaves (e.g., live oak (Quercus virginiana)) or volatile oils (e.g., juniper), and soils that have a high organic matter content (DeBano 1981). Burning will also reduce the cover of vegetation and litter needed to protect the soil from raindrop impact and to slow runoff. Therefore, the greatest impact on site hydrology usually occurs in the period following the burn when vegetation has not had an opportunity to reestablish a protective soil cover. Prescribed burns are different than wildfires, since they are applied under very specific climatic conditions and at specific times of the year in order to provide effective control of brush with minimum impacts on desired associated vegetation and soils.
At the Sonora Experiment Station, infiltration rates and sediment yield were determined to assess how prescribed burning interacts with different vegetation types to influence rangeland hydrology (Hester et al. 1997). Burning significantly reduced the infiltration rate and increased the sediment yield of juniper, bunchgrass, and shortgrass vegetation types. These changes in hydrology were the result of a loss in protective cover which dissipates raindrop energy and obstructs runoff. Factors affecting infiltration rate included soil organic matter, the amount of bare soil exposed, and aggregate stability. Factors affecting sediment yield were the amount of bare soil exposed, and bulk density. These results correspond with those obtained by Roundy et al. (1978) where burning on gently-sloping pinyon-juniper sites reduced the infiltration rate and increased sediment yield 2 to 3 times on areas where old litter mounds had existed under the juniper trees. Hester et al. (1997) concluded that in order to minimize the risk of runoff and erosion, tree groves should either be burned as spot fires or limit the use of prescribed fires that burn the entire pasture to seasons of the year when the risk of high intensity storms is low.
Mechanical/Burn
Combination control methods are often needed to provide a high kill rate on difficult to eradicate brush species such as juniper. Combination methods may also be used when associated vegetation is undesirable for management purposes. In this case, the brush species and associated vegetation are removed and the site is prepared for reseeding or other cultural practices.
In Utah, chained sites with different debris management treatments (described earlier) were reevaluated 6 years after treatment (Buckhouse and Gifford 1976a,b). Infiltration rates for chained and windrowed, chained with debris left in place, and undisturbed sites were similar. Sites were then burned. There was a significant increase in sediment production directly following the burn, but rate of erosion returned to normal within 9 months. The increased erosion was attributed to decreased infiltration caused by hydrophobic (water-repellant) soils. Water quality was slightly reduced following the burn, but this was a short-term effect.
In Texas, Ashe juniper was bulldozed and then burned on sites with three different slope gradients (Wright et al. 1976). On the level sites, this control method did not affect infiltration, erosion, or water quality. Sites with moderate slopes had greater runoff than undisturbed sites 12 to 18 months following treatment. Amount of runoff from steep slopes did not return to control levels until 30 months following treatment. Vegetation on steep slopes did not develop as rapidly or as uniformly as on level or moderate slopes. Wright et al. (1976) attributed recovery of site hydrology to the revegetation of bare areas where junipers had been removed. Sediment production increased with slope gradient and intensity of the burn. Sediment production was up to 35 times greater in the year following the burn, and did not return to normal levels until 10 years following the burn on the steeper slopes. The source of erosion was primarily the loose soil torn up beneath the bulldozed trees. The rates of erosion and infiltration stabilized once herbaceous cover reached 65%.
Fire provides rapid mineralization and dispersion of plant nutrients. The oxides which form are substantially more soluble, and are easily lost from the site due to leaching of the ash layer or via runoff (Tiedemann et al. 1979). Most nutrients lost from a site, however, are associated with sediment. Loss of nutrients from bulldozed and burned sites (Wright et al. 1976) were generally low, although water quality was somewhat reduced on moderate and steep slopes. Water quality returned to normal levels by 4 years after the burn. Wright et al. (1976) concluded that the natural healing on sites with slopes less than 20% does not degrade the rangeland resource.
Chemical
Depending on the herbicide used, application can affect all the vegetation on the site (non-selective), or it can be targeted to control only a particular species or group of species (selective). It can be applied aerially as a spray, or can be applied as pellets. Application of herbicide does not require any soil disturbance, therefore it is often the method of choice where soils are easily compacted or on steep slopes where access is limited. Picloram or picloram in combination with 2,4-D are the most effective herbicides in controlling juniper. However, use of herbicides has come under public scrutiny in the last decade. Persistence of herbicides in the system varies with the type used. Persistent herbicides may leach into the groundwater or may be carried off-site in runoff. This has raised concern about potential hazards to human health.
In Arizona, herbicide was used to control Utah and alligator juniper, removing approximately 83% of the overstory (summarized by Blackburn 1983). Annual runoff increased on the herbicide-treated site when compared to untreated rangeland, and was greater than from mechanically treated sites as well. However, the water quality was highest from the herbicide-treated site compared to mechanically-treated sites because there was no soil disturbance.
Preliminary Results of Juniper Control Effects on Water Yield At the Sonora Agricultural Experiment Station
In 1988, seven moderately grazed 10 acre watersheds were equipped to monitor runoff from juniper dominated rangeland. The slopes of these watershed ranged form 3% to 10%. The soils are Tarrant silty clays which overlay a fractured limestone substrate. Soil depth ranges from 6 inches to 18 inches.
In February, 1991, after the watersheds were well calibrated with the results of several years of stormflow, all woody vegetation on three of the watersheds was cut with axes and manually carried off the drainage. The purpose of this treatment was to determine, with as few confounding effects as possible, the degree to which brush cover influenced water yield. The runoff from each of the treated and untreated watersheds continued to be monitored through 1993. In addition to this research, concurrent studies used weighing lysimeters to track moisture movement through the soil profile, documented transpiration and stomatal conductance of dominant vegetation types, documented the amount of precipitation loss via interception, and documented infiltration characteristics of dominant vegetation types through the use of a rainfall simulator. A preliminary assessment of the combined results of these studies and their implication for water yield is summarized in Table 1.
Table 1. Water balance on rangeland at the Texas Agriculture Experiment Station, Sonora, TX.
| 100% Grass | 70% Grass12% Oak
18% Juniper |
40% Grass24% Oak
36% Juniper |
|
| Rainfall | 22.6 | 22.6 | 22.6 |
| Interception Loss | 3.0 | 6.3 | 9.6 |
| Water Reaching the Soil | 19.6 | 16.3 | 13.0 |
| Runoff | 0.2 | 0.2 | 0.2 |
| Water Going in the Soil | 19.4 | 16.1 | 12.8 |
| Evapo-transpiration | 15.7 | 15.8 | 12.8 |
| Deep Drainage | 3.7 | 0.3 | 0.0 |
| Moderate Stocking (animal units/sec) | 34 | 22 | 11 |
| 3.7 Inches of Deep Drainage/yr = 100,500 Gallons/Ac/Yr | |||
These data indicate that substantial water yield can be achieved through conversion of pasture vegetation from brush to grass dominance. It is notable that there was little increase in runoff from these pastures. This is because the soil under the cut juniper maintained very high infiltration rates after the trees were removed. The soil under the trees had a very high organic matter content (10% to 12%) as a result of decades of litter deposition and decomposition. Adhesive byproducts of organic matter decay are primary constituents that help bind soil particles together and aid formation of water stable aggregates. The porosity of the soil is a function of soil texture and the degree to which the soil is aggregated. Porosity and pore size determine the infiltration rate of a soil. Therefore, the added precipitation reaching the soil as a result of reducing interception losses did not runoff of the pasture but was instead channeled into the soil whenever overland flow crossed an area previously occupied by brush cover. The moderately grazed pastures also had a good herbaceous cover in the juniper interspace. The presence of these grasses and forbs helped to stabilize the site and reduce the potential for runoff and erosion when the junipers were cleared. Concurrent studies being conducted in other regions of the Edwards Plateau indicate that on sites with even shallower soils, degraded herbaceous communities and/or steeper slopes, more of the water reaching the ground after juniper control may leave the site as surface runoff.
Summary
Juniper increase can have a major impact on rangeland hydrology. The presence of juniper alters the amount and distribution of water reaching the soil. Junipers are highly competitive with the understory vegetation for water and nutrients, often reducing the productivity of grasses and forbs and increasing the amount of bare soil. The increase in bare soil, particularly in the spaces between trees, typically leads to increased runoff and soil loss as the juniper infestation increases. The method and degree of juniper removal can significantly impact the hydrology and erosion on rangeland watersheds. The effect of removal method depends on degree of elimination of brush and associated vegetation, steepness of slope, soil type, precipitation characteristics and vegetation recovery time.
Literature Cited
Armentrout, S.M., and R.D. Pieper. 1988. Plant distribution surrounding Rocky Mountain pinyon pine and one-seed juniper in south-central New Mexico. Journal of Range Management. 41:139-143.
Belsky, A.J. 1996. Viewpoint: Western juniper expansion: Is it a threat to arid northwestern ecosystems? Journal of Range Management 49:53-59.
Blackburn, W.H. 1975. Factors influencing infiltration and sediment production of semiarid rangelands in Nevada. Water Resources Research 11:929-937.
Blackburn, W.H. 1983. Influence of brush control on hydrologic characteristics of range watersheds. IN: Proceedings of the Brush Management Symposium, Society for Range Management, Albuquerque, NM.
Blackburn, W.H., and C.M. Skau. 1974. Infiltration rates and sediment production of selected plant communities in Nevada. Journal of Range Management 27:476-480.
Bolton, S.M., T.J. Ward, and J.S. Krammes. 1991. Hydrologic processes in the pinyon-juniper vegetation zone of Arizona and New Mexico. pp. 31-44. IN: Agencies and science working for the future, New Mexico Water Resources Research Institute, November.
Buckhouse, J.C., and G.F. Gifford. 1976a. Sediment production and infiltration rates as affected by grazing and debris burning on chained and seeded pinyon-juniper. Journal of Range Management 29:83-85.
Buckhouse, J.C., and G.F. Gifford. 1976b. Grazing and debris burning on pinyon-juniper sites- some chemical water quality implications. Journal of Range Management 29:299-301.
Buckhouse, J.C., and D.A. Bolognani. 1982. Hydrologic response following rangeland improvement in eastern Oregon. pp.10-12. IN: Research in rangeland management, USDA, Agricultural Research Service, Special Report 663.
Carlson, D.H., T.L. Thurow, R.W. Knight and R.K. Heitschmidt. 1990. Effect of honey mesquite on the water balance of Texas Rolling Plains rangeland. Journal of Range Management 43:491-496.
Collings, M.R. 1966. Throughfall for summer thunderstorms in a juniper and pinyon woodland, Cibecue Ridge, Arizona. Geological Survey Professional Paper 485-B, U.S. Government Printing Office, Washington, DC.
Davis, E.A., and C.P. Pase. 1977. Root systems of shrub live oak: implications for water yield in Arizona chaparral. Journal of Soil and Water Conservation 32:174-180.
DeBano, L.F. 1981. Water repellent soils: a state-of-the-art. USDA Pacific Southwest Forest and range Experiment Station General Technical Report PSW-46.
Dye, K.L., D.N. Ueckert, and S.G. Whisenant. 1995. Redberry juniper-herbaceous understory interactions. Journal of Range Management 48:100-107.
Eddleman, L.E. 1983. Some ecological attributes of western juniper. pp. 32-34. IN: Research in rangeland management, USDA, Agricultural Research Service, Special Report 682.
Engle, D.M., J.F. Stritzke, and P.L. Claypool. 1987. Herbage standing crop around eastern redcedar trees. Journal of Range Management 40:237-239.
Gifford, G.F. 1975. Approximate annual water budgets of two chained pinyon-juniper sites. Journal of Range Management 28:73-74.
Gifford, G.F., G. Williams, and G.B. Coltharp. 1970. Infiltration and erosion studies on pinyon-juniper conversion sites in southern Utah. Journal of Range Management 23:402-406.
Hester, J.W. 1996. Influence of woody dominated rangelands on site hydrology and herbaceous production, Edwards Plateau, Texas. M.S. Thesis, Texas A&M University, College Station, TX.
Hester, J.W., T.L. Thurow and C.A. Taylor, Jr. 1997. Hydrologic characteristics of vegetation types as affected by prescribed burning. Journal of Range Management 50:In press.
Hibbert, A.R. 1979. Managing vegetation to increase flow in the Colorado River Basin. USDA, Forest Service, General Tech. Report RM-66.
Hibbert, A.R. 1983. Water yield improvement potential by vegetation management on western rangelands. Water Resources Bulletin 19: 375-381.
Ko, J.K., and P.B. Reich. 1993. Oak tree effects on soil and herbaceous vegetation in savannas and pastures in Wisconsin. American Mid-Atlantic Naturalist 130:31-42.
Navar, J. and R. Bryan. 1990. Interception loss and rainfall redistribution by three semi-arid growing shrubs in northeastern Mexico. Journal of Hydrology 115:51-63.
Roundy, B.A., W.H. Blackburn, and R.E. Eckert, Jr. 1978. Influence of prescribed burning on infiltration and sediment production in the pinyon-juniper woodland, Nevada. Journal of Range Management 31:250-253.
Scatena, F.N. 1990. Watershed scale rainfall interception on two forested watersheds in the Luquillo Mountains of Puerto Rico. J. Hydrology. 113:89-102.
Scifres, C.J. 1980. Brush management, principles and practices for Texas and the southwest. Texas A&M University Press, College Station, Texas.
Schott, M.R. and R.D. Pieper. 1987. Succession of pinyon-juniper communities after mechanical disturbance in southcentral New Mexico. Journal of Range Management 40:88-94.
Skau, C.M. 1962. Interception, throughfall, and stemflow in Utah and alligator juniper cover types of northern Arizona. Forest Science 10:283-287.
Texas Water Development Board (TWDB) 1990. Water for Texas today and tomorrow. Document No. GP-5-1. Austin, Texas.
Thurow, T.L., W.H. Blackburn and C.A. Taylor, Jr. 1986. Hydrologic characteristics of vegetation types as affected by livestock grazing systems, Edwards Plateau, Texas. Journal of Range Management. 39:505-509.
Thurow, T.L., W.H. Blackburn, S.D. Warren, and C.A. Taylor, Jr. 1987. Rainfall interception by midgrass, shortgrass, and live oak mottes. Journal of Range Management 40:455-460.
Thurow, T.L. 1991. Hydrology and Erosion. p. 141-159. In: R.K. Heitschmidt and J.W. Stuth. Grazing Management: An Ecological Perspective. Timber Press.
Tiedemann, A.R., C.E. Conrad, J.H. Dieterich, J.W. Hornbeck, W.F. Megahan, L.A. Viereck, and D.D. Wade. 1979. Effects of fire on water. USDA, Forest Service, General Tech. Report WO-10.
Weaver, J.E. 1954. North American Prairie. Johnsen Publishing Co., Lincoln, NE.
Weltz, M.A. 1987. Observed and estimated (Erhym-II Model) water budgets for south Texas rangelands. PhD. Dissertation, Texas A&M University, College Station, Texas.
Williams, G., G.F. Gifford, and G.B. Coltharp. 1969. Infiltrometer studies on treated vs. untreated pinyon-juniper sites in central Utah. Journal of Range Management 22:110-114.
Wood, M.K., and W.H. Blackburn. 1981. Grazing systems: Their influence on infiltration rates in the Rolling Plains of Texas. Journal of Range Management 34:331-335.
Wood, M.K., and N. Jared. 1991. Hydrologic responses to fuelwood harvest and slash disposal on a pinyon-juniper dominated grassland site in the Gila National Forest. pp. 25-30. IN: Agencies and science working for the future, New Mexico Water Resources Research Institute, November.
Wright, H.A., F.M. Churchill, and W.C. Stevens. 1976. Effect of prescribed burning on sediment, water yield, and water quality from dozed juniper lands in central Texas. Journal of Range Management 29:294-298.
Young, J.A., R.A. Evans, and D.A. Easi. 1984. Stemflow on western juniper trees. Weed Science 32:320-327.
Comments: Allan McGinty, Professor and Extension Wildlife Specialist