Productivity of Diverse Pastures in a Wet vs. Dry Year

Jim Gerrish and Matt Sanderson1

Cool-season pasture species vary in their ability to remain productive during dry summer conditions. Productivity of sixteen grass and grass-legume mixtures varying in degree of species complexity were compared in two years of markedly different precipitation levels. Forage production in 1999, when growing season precipitation was 40% of normal, averaged 65% of production measured in the more favorable year of 1998. Range in variation in 1999 yield as percentage of 1998 among mixtures was from 57 to 78%. In 1999 tall fescue (Festuca arundinacea Schreb.) receiving 120 lb N/acre yielded 78% of 1998 yield but 90% of the production occurred before July 5. Mixtures containing birdsfoot trefoil
(Lotus corniculatus L.) as the primary legume provided the best pasture performance in terms of both total annual yield and seasonal yield distribution. Both binary grass-birdsfoot trefoil mixes and a complex eight-species mix provided greater mid to late-summer forage yield than either tall fescue or smooth bromegrass (Bromus inermis Leyss) receiving N fertilizer.

Introduction: Livestock producers relying on pasture to provide most of the nutrients for their stock sometimes face the challenge of drought. Across a large part of the Central US, 1998 and 1999 were very contrasting growing seasons. While 1998 was an excellent growing year with adequate rainfall distribution throughout the growing season, 1999 was the driest June through September period recorded in 25 years of weather records kept at the University of Missouri-Forage Systems research Center and, on the average for the state of Missouri, June through December, 1999, was the driest seven-month period of the 20th century.

An ongoing study comparing forage yield and quality of 16 different pasture mixtures varying in species complexity from one to eight has provided a comparison of pasture performance between a near normal precipitaion year and a very dry year. Our objective was to compare total yield and yield distribution among the pasture mixtures in the two contrasting years.

Material and methods: Sixteen grass or grass-legume mixtures were established in 50 ft X 50 ft plots in a completely randomized design with four replications. Mixtures varied in degree of species complexity. Base grasses were tall fescue (TF) and smooth bromegrass (SB). These two species were established as monocultures and one or the other were present in all mixtures. Grass components were seeded on September 5, 1994 on a prepared seed bed. Grass seed was broadcast and rolled in with a cultipacker. Legume components were frost seeded on March 10, 1995. Red clover (Trifolium pratense L.; RC) and birdsfoot trefoil (BFT) establishment with frost seeding was very good while alfalfa (Medicago sativa L.;A) establishment was very poor. Due to the near absence of alfalfa in the designated plots, the plots originally intended as TF or SB +A were reassigned as grass + 0 N treatments for comparative purposes. Legumes were seeded first as binary mixtures with each base grass and then in combination with one another within each base grass. More complex mixtures were created by adding orchardgrass (Dactylis glomerata L.;OG), timothy
(Phleum pratense L.;T) and big bluestem (Andropogon gerardii Vitman;BB) to the tertiary base grass-legume mixtures. Thus, the most complex mixture contained eight seeded species. Each mixture was formulated to provide 80 seeds/sq-ft with the base grass always providing 50% of the seed in mixtures and all other components equally represented. The one exception was the eight species mix where all components were allocated in equal amounts.

Swards were uniformly managed to encourage establishment during the 1995 growing season and were mechanically harvested in 1996. Tall fescue and smooth bromegrass monocultures received 120 lb N/acre annually as three 40-lb applications applied in March, June, and September. For interpretation of all of the figures in this report refer to the reference number beside each pasture mixture.


Ref. No.Tall Fescue Base                Ref. No. Smooth Bromegrass Base
1)TF + 120 lb N/acre                     9) SB + 120 lb N/acre
2)TF + 0 N/acre                         10) SB + 0 N/acre
3)TF + BFT                              11) SB + BFT
4)TF + RC                               12) SB + RC
5)TF + A +BFT                           13) SB + A + BFT
6)TF + A + BFT + RC                     14) SB + A + BFT + RC
7)TF + A + BFT + RC + OG + T            15) SB + A + BFT + RC + OG + T
8)TF + SB + A + BFT + RC + OG + T       16) SB + A + BFT + RC + TF + OG
                                             + T + BB

In both 1998 and 1999, individual plots were grazed on an as-needed basis whenever a particular plot reached 8 to 10-in. mean sward height. Prior to grazing, six 2.7-ft2 quadrats were clipped from each plot. At the same site, sward surface height was measured and the species composition was visually estimated. Plots were grazed with six to eight steers for four to seven hours to remove approximately 50% of the forage biomass. Residual forage was measured after each grazing event. Total annual forage yield and yield by month were determined.

Results and Discussion: Rainfall in April and May, 1999, was excellent and temperatures were favorable for very rapid pasture growth. Pasture growth rate and forage yields early in the season actually exceeded the rates and yields measured in 1998. Rainfall for the next three months was less than 40% of the long term norm. Pasture mixtures which remain productive under these latter conditions should allow producers to better weather the economic and biological stresses of drought.

In 1998 average sward height at turn-in was slightly greater than our target of eight to ten inches (11.0 in.) while in 1999 the average sward height was slightly less than the target
(7.75 in.). Herbage mass at turn-in was similar for both years with 2415 and 2315 pounds/acre for 1998 and 1999, respectively.

Total annual forage yield in 1999 was about 65% of the 1998 forage yield when averaged across all pastures (Figure 1). The first assumption might be that any pasture which provided greater than 65% yield potential could be considered to be lower risk pastures when planning for future drought protection. This might be a safe assumption if the goal were only total forage production. For example, a single-cut hay production system that might have that objective. Total forage yield for the sixteen pastures is shown in Figure 2.

Tall fescue receiving 120 lb N/acre was quantitatively the highest yielding pasture as well as the pasture providing the highest percentage of yield in a poor year compared to the very good 1998 growing season. Statistically, four other pasture mixtures were not significantly lower yielding than TF + 120 N. The common characteristic of the other higher yielding mixtures is that they all contained birdsfoot trefoil as the primary legume. Red clover contribution to pastures was much lower in 1999 compared to 1998. While this may have been a drought effect, it may also have been a reflection of timing in the red clover persistence cycle. In stands managed for natural reseeding as this study is, there are generally two years of strong clover production followed by an off year. Stand measurements made in October of 1999 did find a significant presence of young red clover plants in most plots designated as red clover mixtures.

Smooth bromegrass response to N fertilization was disappointing in both years. While addition of 120 lb N/acre to tall fescue produced almost 3500 lb/acre additional forage per acre compared to 0 N, 120 lb N applied to smooth bromegrass produced only 300 pounds more forage than unfertilized bromegrass in 1999. Even with the excellent growing conditions of 1998, N fertilization produced less than 1000 pounds of additional forage on smooth bromegrass pastures. Obviously, the economics of fertilization of these two species must be quite different. Smooth bromegrass responds much better to N fertilization in hay situations compared to grazing because so much of the biomass is produced above typical grazing heights. Tall fescue, on the other hand, produces an abundance of dry matter in the lower 6 to 8 inches. The economics of N fertilization on grazed tall fescue can be quite good due to potentially high production in relatively short pastures. Coupled with high stock density which can enhance nitrogen recycling in the system, N fertilization of tall fescue looks quite good.

The primary drawback of N-fertilized fescue pastures in dry years is yield distribution. While 1999 total production appeared quite good, the monthly distribution of available forage was a different matter. Mixtures which compared most favorably to TF + N in both total yield and 1999 yield as percent of 1998 yield were those mixtures containing birdsfoot trefoil. Monthly production of TF + BFT compared to TF + N was quite favorable, particularly in late summer (Figure 3). Over 90% of TF + N total yield for the season was produced before July 15. While this may be acceptable for hay production, it is unacceptable for pasture production. The pasture with greatest species complexity, including a C-4 grass exhibited a unimodal yield distribution pattern for the season compared to the typical bimodal distribution curve associated with most cool-season
forages (Figure 4).

Summary: While N-fertilized grass pasture produced the most total forage in a very dry year, 90% of the forage production occurred before July 5. Total season forage yield of grass mixtures containing birdsfoot trefoil were not significantly different from N-fertilized grass but did produce significantly more forage during the late summer period compared to N-fertilized grasses. Mixtures appear to be more drought resistant than grass monocultures.

Figure 1. Total forage yield in 1999 averaged 65% of total forage yield in 1998.
(DMY=Dry Matter Yield)

Figure 1

Figure 2. Total forage dry matter yield in 1999 with least significant difference compared to tall fescue receiving 120 lb N/acre.

Figure 2

Figure 3. Tall fescue + birdsfoot trefoil pastures exhibited significantly more late summer regrowthcompared to tall fescue + 120 lb N/acre during the 1999 dry season.

Figure 3

Figure 4. Pasture mixtures vary significantly in yield distribution through the season with more complex mixtures exhibiting more unimodal yield distribution compared to a simple binary mixture.

Figure 4



Jim Gerrish and Matt Sanderson1

Drought is a problem which livestock producers must periodically face. Dry conditions typically reduce total pasture yield and reduce livestock carrying capacity. We compared productivity of 16 different cool-season pasture mixtures between a season with favorable rainfall for optimum pasture production and very dry year. Treatments included grass monocultures receiving 120 lb/acre nitrogen fertilizer as well as grass legume mixtures. The most complex mixture in the study consisted of eight different forage species including a warm-season grass which would be better adapted to hot, dry conditions. Forage yield in the dry year was reduced by about one-third over all mixtures but individual pasture mixtures varied from less than 25% reduction to almost 50% product ion. Pasture mixtures which yielded greater than the mean of all treatments in the dry year on a percentage basis relative to the favorable year should be those mixtures which are best adapted to withstanding drought stress. Tall fescue fertilized with nitrogen produced the most total forage in both the wet and dry years, but the distribution of yield through the dry weather was very poor. Tall fescue and smooth bromegrass pastures containing birdsfoot trefoil provided the best overall balance of total yield and distribution through the dry season. This research indicates that in cool-season dominant pasture areas, grass-legume mixtures provide more stable forage production compared to nitrogen-fertilized grass.

1 Research Assistant Professor, University of Missouri-Forage Systems Research Center, 21262 Genoa Road, Linneus MO 64653. ( and Research Agronomist, Pasture Systems and Watershed Management Research Laboratory, USDA-ARS, Building 3702, Curtin Road, University Park, PA 16802-3702.

September 26, 2000