Interaction of Simulated Drought and Grazing on Mixed Grass Prairie

Don Kirby, Wendi Rogers (Graduate Student), Paul Nyren, Bob Patton

Table of Contents

Historical Perspective
Study Methods
Expected Results


Drought is a common climatic phenomena of the Great Plains including North Dakota. By examining the last 100 years of climatic data, two and four years of every ten are ecologically and/or economically considered drought years in eastern and western North Dakota respectively. Despite the recent wet climatic cycle in the northern Great Plains, the start of the next drought may only be a few weeks away.

Grasslands in North Dakota are important to the economic well-being of the livestock industry. Nearly 80% of the diets of cattle in this state are made up of forages of which grazed forages comprises the majority of this diet fraction. Controlling drought's effects on plant communities may be tempered through proper livestock management (Olson et al. 1985). The proper management of livestock should optimize forage produced in drought by maintaining plant numbers and structures such as leaves, roots, etc. necessary for growth, and increase the amount of effective precipitation for plant use through soil surface management (Rauzi and Smika 1963). Livestock management should also allow a more rapid recovery of plant communities hence forage following a drought event.

The objectives of this research are two-fold. The first objective is to simulate drought on properly grazed rangeland and measure the effects on herbage production, basal plant cover and broad-leaved plants. The second objective will be to monitor the recovery period of the affected plant community to pre-drought conditions. The simulated drought period will be two consecutive years followed by an expected recovery period of two or more years.

Historical Perspective

Drought is defined as a 15% to 25% or greater deficit in the annual precipitation (Society for Range Management). Written evaluation of drought's effects on rangeland plant communities in the northern Great Plains began in the 1930's. All reports for the last 70 years have only examined natural droughts or drought periods with no control over environmental inputs. Therefore, each evaluation of drought has reported on a unique event that may or may not be representative of what is occurring throughout or between regions.

The best known chronicle of drought in North Dakota was maintained by Dr. Warren Whitman (Whitman et al. 1943). He noted that in 1934 and 1936, summer precipitation averaged 10.4 cm compared to 23.5 cm for the ten year period 1930-1939. In 1933, needle-and-thread and prairie junegrass occupied greater area and were more dominant in the vegetation composition than throughout the drought period (1934-1936) and did not fully recover until 1938. Sandberg's bluegrass was infrequent prior to 1934 but increased from one plant-clump per quadrat in 1933 to 17 plant-clumps per quadrat in 1938. Blue grama was the most severely affected of any of the major grass species in the 1930s drought. Its cover in 1937 was only one-third that of 1933; but by 1940, blue grama had recovered to pre-drought cover.

At the Livestock and Range Research Station near Miles City, Montana, two of the four driest summers experienced occurred between 1927 and 1934. In 1934, the summer rainfall was 8.9 cm compared to the preceding 57-year average of 23.6 cm. The area of blue grama on 55 individual m2 quadrats averaged 1542-cm2 in 1933 and decreased to 389-cm2 by 1935. Buffalo- grass experienced a decrease in average cover from 876-cm2 to 185-cm2 between 1933 and 1935. Sandberg's bluegrass increased in area from 4-cm2 in 1933 to 118-cm2 in 1935. Due to its quick growth after the last frost, maturing early in June, and going into dormancy by early summer, Sandberg's bluegrass was able to utilize snowmelt and early spring precipitation to out-compete warm season grasses for resources (Ellison and Woolfolk 1937).

Drought conditions between the fall of 1987 and summer of 1989 decreased plant basal cover on rangeland being grazed at both Dickinson and Streeter, ND locations (Kirby 1994). Additionally, herbage production in 1988 averaged 50% or less of the 9 and 11 year averages at the two locations. Kirby suggested that annual changes in species composition, plant basal cover and herbaceous production were influenced more by the amount and timing of precipitation than by stocking rate and grazing system employed.

Under grazing pressure, Biondini and Manske (1996) in western North Dakota reported that rainfall was more important than a grazing system in controlling measurable ecosystem-level variables. Total basal cover under moderate grazing, approximately 50% intensity, tends to be stable and unresponsive to stresses such as grazing or drought. The average annual net primary productivity in 1988 declined 46% with the decline continuing into 1989 on shallow and silty range sites. Cool season grasses and sedges decreased while warm season grasses and forbs increased under the climatic stress.

Study Methods

The study will be conducted on the Central Grasslands Research Extension Center 8 miles northwest of Streeter, in south central North Dakota. Study sites will be located on section 14 (T138N R70W), a long-term grazing intensity study initiated in 1988. The station lies within the mixed grass prairie (Whitman and Wali 1975). Important grass species include western wheatgrass, Kentucky bluegrass, blue grama, needle-and-thread, prairie junegrass, and threadleaf sedge. Some important forbs include white sage, fringed sage, silver-leaf scurf pea, red false globemallow, and goldenrod species. Western snowberry is the dominant low shrub on pastureland.

A grazing intensity study was begun in 1988 on mixed-grass prairie that had a history of intermittent light grazing for ten years. Grazing intensity treatments were defined as the amount of aboveground standing biomass left at the termination of the grazing period. For this research project only the moderate grazing intensity will be sampled. Each year the moderate treatment was grazed between 86 and 181 days beginning in mid-May. Average aboveground biomass remaining at the termination of annual grazing periods averaged 50% for the moderate grazing intensity.

Automated rainout shelters will be installed to control the amount of precipitation received on rainfall treatments. Rainfall treatments will consist of 60-65% of the long-term average (drought), 100% of the long-term average (average), and the natural amount received annually (natural). Three replicate shelters, 3 x 6 m will be randomly assigned to silty range sites for the drought and average rainfall treatments. The shelters will be on rails equipped with electric motors and wet sensitive apparatus, that when wetted will activate the motors to move the shelters across sampling plots. To prevent lateral water movement, each rainout shelter boundary will be trenched to 60 cm depth, lined with plastic and refilled with soil. Drought conditions will be simulated for two consecutive years for the region. Plots subjected to the drought and average treatments will be drip irrigated at two-week intervals using the long-term bimonthly recorded precipitation for the site.

Prior to the grazing season, portable enclosures will be located within each rainfall treatment to prevent livestock grazing. In early July herbaceous production will be estimated by clipping quadrats from each replication of the three rainfall treatments. Herbaceous biomass will be separated into cool and warm-season grasses, grass-like species and forbs. Basal cover will be sampled using the ten-point frame technique (Society for Range Management 1986). Cover estimates will be by major grass species, grass-like, forbs, litter and bare ground. Density of forbs will be determined using 0.25 m2 and nested frequency quadrats. All forbs will be identified and counted.

Expected Results

Drought is a recurring event on the grasslands of the Great Plains. It creates short and long-term losses in forage and livestock production. Initially we will study the effects of controlled drought on moderately (proper) grazed rangeland. We will note changes in community production, structure and function and relate these to potential ecological and economic gains or losses. This community will also be examined to determine how quickly ecological and economic functioning of a properly grazed drought stricken plant community occurs. Future research will be conducted using like methods to examine poorly managed grazing (overgrazed) practices as a contrast to the present study

Dr. Donald Kirby
Animal and Range Science Department
North Dakota State University
Box 5727 State University Station
Fargo, ND 58105

Paul E. Nyren, Director
Central Grasslands Research Extension Center
4824 48th Ave SE
Streeter, ND 58483


1999 Annual Report Directory