The Effect of Grazing Intensity on Rangeland Hydrology

Chad L. Engels, Graduate Student, NDSU Department of Civil Engineering


Table of Contents



Grazing management is a tool that can be used to increase the amount of soil moisture that becomes available for plant use (Rauzi, 1963). The reason for this is that grazing affects plant material, both standing crop and litter, which has a pronounced affect on infiltration (Rauzi and Smith, 1973). Secondly, grazing can affect the surface porosity of soils due to trampling and “ vegetative root thinning”, thus changing the runoff potential of the rangeland (Van Haveren, 1983). When runoff occurs, its erosive nature transports soil nutrients from the site and can even reduce A-horizon (topsoil) depth, which decreases the amount of water that can be stored in the soil profile (McGinty et al., 1991). The A-horizon is rich in organic matter, giving it the ability to hold more water than deeper soil horizons. Erosion research has shown that loss of soil reduces crop yields due to a decrease in the amount of water available to plants (Schertz et al., 1989). Also, the USDA (1989) reported that on shallow soils with sloping terrain, erosion may completely destroy crop productivity. When the ability of a soil profile to store water is reduced, plants run out of water sooner, increasing the frequency and severity of droughts (McGinty et al., 1991). This situation adversely affects vegetative growth and therefore has a direct effect on the productivity of rangelands.



The purpose of this study is to define the hydrologic characteristics of Missouri Coteau rangelands as influenced by two levels of grazing. It has long been known that intensive grazing has a negative effect on the hydrology of a rangeland, however the degree of difference between grazing treatments has been found to be variable. There is currently no significant record of infiltration and runoff data in relation to grazing on Missouri Coteau soils in North Dakota. Due to the intensive ranching of this area, the need to develop an understanding of this relationship is evident from a production standpoint. The data collected will serve as a supplement to the existing Central Grasslands Research Extension Center (CGREC) database, and bring further insight to range ecosystem dynamics.


In this study, runoff and infiltration rates from moderately and extremely grazed pastures are compared. Soil erosion is quantified simultaneously. Collection of infiltration, runoff, and erosion data is achieved through the use of a rainfall simulator and single-ring infiltrometer. Physical soil and vegetative characteristics are measured as well. The soil characteristics investigated include bulk density and porosity. The vegetative characteristics studied include pounds per acre (lb/ac) of above ground biomass and percent surface cover. By relating the physical soil and vegetation characteristics of a rangeland watershed to runoff and erosion potentials, it becomes possible to make grazing management decisions that lead to water conservation and support long-term profitability for ranchers.


The CGREC grazing intensity trial area is a random configuration of 12 pastures consisting of three replications of four grazing treatments. The trial area is made up of three replication pastures of lightly, moderately, heavily, and extremely grazed treatments. The three moderately grazed pastures and three extremely grazed pastures were utilized for this study. During the summer of 1999, one rainfall simulator site was chosen within each moderately and extremely grazed pasture at the CGREC grazing intensity trial area. Three replications of each treatment was the result. In this study, moderate grazing is defined as that stocking rate which leaves 50% (1,650 lb/ac) of an average year’s above ground biomass remaining at the end of the grazing season. Extreme grazing leaves 20% (505 lb/ac). The pastures were stocked in mid to late May and have been continuously grazed into autumn for the past ten years. Rainfall simulator sites were chosen based on similar soil, topography, and plant community characteristics. Single-ring infiltrometer data was collected during the summer of 1998 from existing silty and overflow range study sites. One single-ring infiltrometer measurement was recorded from each site (two sites per pasture), resulting in six replications of each treatment.

The rainfall simulator utilized for this study was on loan from the U.S. Department of Agriculture - Agricultural Research Service (ARS), courtesy of Dr. John E. Gilley, Lincoln Nebraska. A 12 ft by 25 ft runoff plot was constructed within the center of each rainfall simulator setup for measurement of infiltration, runoff, and erosion. Each rainfall simulation consisted of a “dry run” followed 24 hours later by a “wet run”. The duration of each simulated rainfall event was 60 minutes and the intensity was 2.0 inches per hour (cm/hr). Erosion from the plots was determined by collecting runoff samples from the flow discharge exiting the flume. Single-ring infiltrometers were constructed from steel barrels of 22 inches in diameter. The barrels were cut to heights of approximately 11.8 inches and inserted into the ground roughly 2.75 to 3.15 inches. Infiltration measurements were typically recorded for a period of five to six hours during a single run.

Soil surface bulk density was measured using a cylindrical hammer-driven coring device within the rainfall simulator runoff plots. Surface soil porosity was determined indirectly by converting the bulk density values into porosity values. Vegetative cover and biomass was also measured within the rainfall simulator runoff plots. Cover was measured using a 10-pin vertical point frame. Biomass was measured in lb/ac using a 2.7 square-feet clipping frame.


Results and Discussion

The primary soil concern in pasture settings is compaction and pore space reduction. To determine if stocking rates affect the soil, bulk density and porosity measurements were recorded. The soil data obtained from the rainfall simulator plots is found in Table 1. The mean bulk density is 0.94 and 1.05 grams per cubic centimeter on moderately and extremely grazed pastures, respectively. The mean porosity is 64.6 and 60.3 percent on moderately and extremely grazed pastures, respectively. Obviously the soil that is more porous will be less dense and vise versa. As shown by the 95% confidence intervals, the means are statistically different. Thus stocking density does indeed alter the physical properties of soil, and therefore may explain differences in infiltration and runoff rates.


Soil cover is often the most significant determinate for the amount of runoff and erosion one can expect from any given watershed. From this standpoint it is clear that we wish to manage rangelands with vegetative cover in mind. The amount of grass, forbs, and litter measured in pounds/acre from the rainfall simulator plots is in Table 1. There is five times as much grass and 12 times as much litter on the moderately grazed pastures than on the extremely grazed pastures. The mean forb biomass density is statistically indifferent. This trend was repeated on the ring infiltrometer test sites as well. The percent surface-coverage expressed as foliage (living plants), litter (decaying plants), and exposed soil is also in Table 1. It was found that the amount of cover provided to the soil by living plants was statistically indifferent between treatments, about 6 percent for both. However, on extremely grazed plots, 45.1% of the surface area was exposed soil. On moderately grazed plots 0.8% of the soil-surface area was exposed. This significant difference was due to the amount of litter present. On moderately grazed pastures 93.1% of the soil-surface area was protected by litter, on extremely grazed sites this value was reduced to 48.1%. The data shows that while standing vegetation is directly affected by stocking density, it is the decaying litter that is of most concern in regards to soil-surface protection.


Table 1. Mean Vegetation and Soil Values for Rainfall Simulator and Ring Infiltrometer Test Sites

Grazing Type
















(% cover)


(% cover)


(% cover)











Upper 95% C.I.










Lower 95% C.I.




















Upper 95% C.I.










Lower 95% C.I.










* Recorded from 1998 ring infiltrometer test sites.

Runoff, infiltration, and erosion data for the rainfall simulator plots is contained in Table 2. Twenty-four hours after the initial rainfall events, 2.0 inches of simulated rain was applied to three moderately grazed plots and three extremely grazed plots. As can be seen, 0.75 inches of rainfall took the form of runoff on extremely grazed plots on average. The mean runoff on moderately grazed plots was only 0.35 inches of rainfall. Looking at it another way, 82.4% of the total rainfall infiltrates into the soil on moderately grazed pastures while only 63.4% infiltrates on extremely grazed sites. These quantities are reinforced by the runoff initiation times as well. The mean time required for runoff to begin after rainfall starts is 32 minutes on moderately grazed plots and only 9 minutes on extremely grazed plots.


Table 2. Mean Cumulative Infiltration & Runoff and Mean Erosion Rates from Rainfall Simulator Plots














Runoff Initiation*





















% Difference








* Mean length of time from the start of rainfall to the start of runoff.

The mean erosion rate for the entire 60-minute rainfall event shows an even more dramatic trend. Soil loss on extremely grazed pastures is occurring at a rate of 10.3 lb/ac/hr. Compare this to a mean erosion rate of only 0.5 lb/ac/hr for moderately grazed pastures and we notice a dramatic difference. Remember the importance of topsoil as discussed in the introduction in regards to drought.


The actual rate at which infiltration is taking place was measured directly using the single-ring infiltrometer and indirectly using the rainfall simulator. Infiltration rates measured by both methods are found in Figures 1 and 2. Figure 1 contains the ring infiltrometer data. The infiltration rates are expressed as an average value for each treatment fit to a model that describes the infiltration rate over time. This average is displayed in Figure 1 as a 95% confidence interval. This simply means that the infiltration rate for any pasture grazed at the given intensity should fall within the stated range 95% of the time. As can be seen, the range is much greater for moderately grazed pastures than it is for extremely grazed ones. Figure 2 contains the six individual infiltration rate curves for the rainfall simulator runs. As can be seen, one moderately grazed pasture maintained an infiltration rate greater than the 2.0 in/hr rain event that was applied. Again it can be noted that infiltration rates recorded on moderately grazed pastures showed more variance than those recorded from extremely grazed pastures. Do to the scope of this study, only a limited number of replications could be performed. Thus proving a definite difference in the rate of infiltration was not possible. However, there are several indications that result from the data. First, the mean infiltration rate is higher on moderately grazed pastures and lower on extremely grazed pastures. Second, infiltration rates recorded from extremely grazed pastures appear to replicate better than those taken from moderately grazed pastures. This may be due to the consistency of grazing that is seen on overstocked pastures compared to the spot or patch grazing that typically occurs on properly managed pasture. Third, mean infiltration rates recorded from extremely grazed treatments fell within the 95% confidence intervals of mean infiltration rates recorded from moderately grazed treatments. Finally, both methods of measurement yielded similar results.







Grazing intensity appears to influence the physical properties of soil and vegetation. Differences in bulk density, porosity, plant biomass, and vegetative cover were all noted. It was found that decaying litter may be the most critical determinate for soil protection. Also, grazing intensity is directly related to soil erosion rates. Extremely grazed pastures had significantly greater erosion rates than did moderately grazed pastures. Additionally, for a 2.0 in. storm event with a duration of one hour, more water infiltrates into the soil on moderately grazed treatments. This is because the runoff initiation time is delayed on moderately grazed pastures and the average infiltration rate is lower. Individual infiltration rates recorded from each study site replicate much better on extremely grazed pastures than moderately grazed ones. While a study of this magnitude cannot be used to make definite statements supported by detailed statistical analyses due to limitations in the number of setups, it can be viewed as a valid indicator. Therefore, the information yielded by this research is important in that it is a “piece of the puzzle” that is needed in order make sound grazing management decisions.

Table of contents

NDSU Central Grasslands Research Extension Center
Home | Livestock Research | Economics Research | Range Research | 2001 Annual Report