Optimizing Water Use for Three Old World Bluestems in the Texas High Plains

Final Report for GS02-012

Project Type: Graduate Student
Funds awarded in 2002: $10,000.00
Projected End Date: 12/31/2004
Matching Non-Federal Funds: $6,000.00
Region: Southern
State: Texas
Graduate Student:
Major Professor:
Dr. Vivien Allen
Texas Tech University
Expand All

Project Information


Agriculture in the Texas High Plains is challenged by rapid depletion of ground water. Warm-season grasses offer opportunities for grazing but information is needed on comparative water use efficiencies. Three old world bluestems (Bothriochloa caucasica, ‘Caucasian’; B. ischaemum, ‘Spar’; and B. bladhii, ‘Dahl’) were grown under dryland and low, medium, and high irrigation levels to determine water use efficiency, yield, and nutritive value during 2001 to 2003. Amount of water applied in the high treatment was 100% replacement of potential evapotranspiration (PET) minus precipitation. Medium and low treatments were spaced evenly and calculated as 66 and 33% of the high treatment; the dryland treatment received no irrigation (0%). In 2003, Caucasian was more water use efficient than Spar; no differences were observed among species in 2001 and 2002. Caucasian and Dahl consistently outyielded Spar by about 30%. Maximum yields resulted from a high irrigation level but forage nutritive value was higher under low irrigation. These data provide information for optimizing water use, nutritive value, and for selecting an optimum between irrigation water invested and total nutrient yield.


Water use efficiency (WUE) of forage grasses is crucial in determining species suitable for implementation into forage/livestock systems in semi-arid environments. Previous researchers have defined WUE as dry matter (DM) yield per ha divided by the amount of water added to achieve that yield. Other authors have inferred WUE from plant-physiological parameters such as gas exchange and leaf water relations. Old world bluestem species are widely grown in the semi-arid Texas High Plains but information is lacking concerning their water use – yield relationships under semi-arid conditions when compared at one location.

Previous research has suggested a 10 to 20% higher productivity of Caucasian bluestem compared with B. ischaemum types. The more recently released old world bluestem WW-B. Dahl showed higher DM yields under dryland conditions than a variety of warm-season grasses in central Texas. With rising concerns regarding sustainability of agricultural systems dominating the Texas High Plains due to a declining supply of irrigation water, introduced warm-season grasses such as Bothriochloa species may offer alternatives for designing viable crop/forage/livestock systems. A variety of old world bluestem species, primarily B. ischaemum types, were grown widely on Conservation Reserve Program (CRP) land to minimize soil erosion and degradation. Previous research showed that Dahl can serve as a key component in functioning crop/forage/livestock systems while reducing water and fertilizer needs. This species has been shown to support higher animal gains than B. ischaemum types (unpublished data, Texas Tech University). Also, Caucasian is perhaps more cold tolerant given its origin and could be an alternative for areas further north in the Southern High Plains.

Declining water reserves in the Ogallala Aquifer require research efforts to find solutions for alternative agricultural systems. The Southern High Plains of Texas may be environmentally well suited to establish forage-based systems while reducing overall water use. However, little is known regarding the water use of old world bluestems grown under conditions of the Texas High Plains. Thus, our objective was to test Caucasian, Spar, and Dahl regarding their WUE, DM yield, and nutritive value.

Project Objectives:

The overall objective was to determine forage growth and nutrient yield per unit of added water (water use efficiency) for three warm-season perennial grasses in Southern High Plains.

Specific objectives included:
To determine the influence of dryland, and low, medium, and high irrigation levels on dry matter (DM) yield, water use efficiency (WUE; kg DM ha-1 mm-1 water), nutritive value, plant morphology including percentage live/dead and leaf/stem ratio of Caucasian, Spar, and Dahl bluestems.


Click linked name(s) to expand
  • Vivien Allen


Materials and methods:

This research was conducted for 3 years, 2001 through 2003, at the Texas Tech University Field Research Laboratory located in north east Lubbock County (101° 47’ west longitude; 33° 45’ north latitude; 993 m elevation). Soil at the research site was a Pullman clay (Fine, mixed, super active, thermic Paleustoll) with a bulk density of 1.2 Mg/m3 for the 0.0-0.3 m3 soil layer and 1.49 Mg/m3 for the 0.3-1.0 m layer (Baumhardt et al., 1995). The area has a semiarid climate with an annual mean precipitation of 470 mm and an average air temperature of 15.5 °C. The landscape was nearly level with 0 to 1 % slope (USDA, 1979).

Existing stands of WW-B. Dahl, Caucasian, and WW Spar old world bluestems were used. There were three replications of each forage treatment in a complete randomized block design. These forages were established in 1996 in paddocks (0.07 ha per replication). The three forages were replicated 3 times in a complete randomized block design. Water treatments included ‘dryland’, and ‘low’, ‘medium’, and ‘high’ irrigation levels. These treatment levels were placed in sequence to minimize micro-climatic alteration among experimental units. Amount of water applied in the high treatment was 100% replacement of potential evapotranspiration (PET) minus precipitation. Medium and low treatments were spaced evenly and calculated as 66 and 33% of the high treatment; the dryland treatment received no irrigation (0%). Each irrigation treatment plot was 10 by 15 m.

In June 2001, plots, except those assigned as dryland, were equipped with 10 lines of 16-mm irrigation tubing produced by Eurodrip (San Diego, CA) with a 457-mm emitter spacing and a delivery rate per emitter of 1.52 L h-1. The distance between irrigation tubes was 1 m. An appropriate pressure of approximately 207 kPa to run the irrigation system was maintained by pressure valves, which were connected to flowmeters that allowed exact monitoring of the amount of delivered water. Water treatments began in spring with emergence of photosynthetically active tissues and ended in autumn with occurrence of the first frost, with the exception of year 1 that began 21 June 2001, following installation of the irrigation system. Twice a year, 60 kg N ha-1 was applied to all plots. Other nutrients were applied according to soil test recommendations.

Measurement of WUE required calculation of PET, the amount of water input (mm) throughout the growing season, precipitation, soil moisture, and above-ground biomass expressed as DM yield (kg ha-1). Daily PET values were obtained from weather data using the Penman-Monteith equation (Allen et al., 1998) measured at the Texas Agriculture Experiment Station, which was located approximately 6 km in a south-western direction of the research site. Amount of precipitation was measured with 2 gauges installed approximately 150 m apart at the south-west and north-east corners of the research site and averaged. Precipitation was used to calculate replacements for the high treatment adjusted by amount of rainfall during each irrigation period of 1 week. Irrigation started at the same time of the day with application length chosen according to amounts of water needed to replace deficits for the different treatments. Each year, soil moisture was measured gravimetrically at the beginning of April and end of October to determine soil water depletion. One soil sample to a depth of 1 m in 20-cm increments was taken from each plot with soil augers specified for dry soil (JMC Backsaver Handle, Clements Associates, Inc., Newton, IA; Eijkelkamp Agrisearch, Giesbeek, The Netherlands). Gravimetric soil samples were placed in plastic bags, sealed, and stored in a cooling room with a temperature of 4 ºC until further analysis. Soil moisture was determined gravimetrically. Approximately 35 g of soil was oven dried at a temperature of 105 ºC for at least 24 h or until constant weight was reached and weighed (± 0.0001 g) again to obtain gravimetric water content. Volumetric water content values were obtained using soil bulk density values given by Baumhardt et al. (1995), which were measured at this site.

Above-ground biomass was determined by clipping three, 1.0 x 0.5-m quadrates randomly from within each treatment area in mid-July at hay cut stage and at the end of the growing season to estimate total seasonal DM yield and WUE. Forage from the entire experimental area was removed immediately after determining amount of the above-ground biomass in July by harvesting with a tractor-mounted shredder. All harvested forage was removed as hay. Standing biomass in autumn was determined as described above but was not removed March prior to spring growth. Stubble height after harvest in July and March was about 8 cm. Biomass plant samples were weighed (± 1.0 g) in the field immediately after clipping, oven dried (at 55º C) and weighed again after drying to constant weight to calculate percentage moisture and total above-ground biomass production. Biomass for the growing season was expressed as DM yield ha-1.

Water use efficiency was calculated by using measured total seasonal DM yield during the growing season divided by the amount of water used to obtain this yield. The water amount was calculated by measuring irrigation water applied throughout the growing seasons plus precipitation and soil water depletion. Soil water depletion was estimated by subtracting water remaining in the soil profile to a depth of 1.0 m at the end of the growing season from soil water content at the beginning of the season at the same depth. Precipitation that occurred within 2 wks before final above-ground biomass harvest was excluded from the soil water balance; likewise, any precipitation that occurred between biomass harvest and time of soil sampling was also subtracted. Thus, WUE was calculated with WUE = D/W, where D = total seasonal DM yield in kg ha-1, and W = (precipitation + irrigation + soil water depletion) in mm measured for each irrigation level and dryland during the growing season. Additionally, the slope of WUE was calculated resulting from an increase in total seasonal DM yield per increment of irrigation level. Also, forage mass based on dry matter was determined in May, June, August, and September by clipping two, 0.5 by 1.0-m quadrates randomly from within each treatment. Drying procedures were similar to those used for determination of WUE and total seasonal DM yield.

Plant material for forage nutritive value was sampled monthly within each species and water treatment plot replication from May through October. Thus, a total of 36 plots were sampled 6 times over the growing season. From each plot, 6 random forage samples at each sampling dates were taken at a clipping height of approximately 8 cm and composited in a paper bag for drying. Samples for determining nutritive value were placed immediately in a dryer and dried at 55º C until constant weight was achieved. Samples were then removed and ground in a Wiley mill (brand name and location) through a 1-mm screen and ground forage samples were placed into plastic bags (Whirl-Pak, Nasco Fort Atkinson, WI), holding approximately 50 g of ground material and stored at room temperature for further analysis.

Analysis for forage nutritive value included neutral detergent fiber (NDF), acid detergent fiber (ADF), crude protein (CP), and total non-structural carbohydrates (TNC). These parameters were predicted by creating a regression equation based on the near-infrared absorption patterns (near-infrared absorption technique, NIR) of all forage samples from each date of collection and a calibration set of the entire sample population analyzed by means of conventional wet chemistry procedures. Fiber components such as NDF, ADF, hemicellulose, (NDF-ADF), cellulose, and lignin were determined based on standard procedures described by Van Soest (1963), and Goering and Van Soest (1970). Percentage CP was estimated according to A.O.A.C. (1995). Dry matter digestibility (DDM) was determined as an estimate of ADF [DDM% = 88.9-(0.779 ADF%)] according to Linn and Martin (1989). Percentage TNC of samples was analyzed based on the procedure described by Mounsif (1986) with the following modifiactions. The spectral absorbance of glucose standards of 0.1, 0.2, 0.3, and 0.4 mg ml-1 was measured with a spectrophotometer (Bausch & Lomb, U.S.A.) and a regression model was built based on the resulting spectra of these standards. Ground forage samples were weighed as close as possible to 0.5 g and boiled in a reflux apparatus with 60 ml of 0.2 N HCL for two hours. Samples were then cooled to room temperature and filtered into a 100 ml volumetric flaks. Distilled water was added to a volume of 100 ml, from which 1 ml was taken and diluted with 4 ml of distilled H20 before mixing on a vortex mixer for 30 seconds. Form this solution, 1 ml was taken again and mixed with 10 ml of anthrone solution (330 ml conc. H2SO4, 5.0 g thiourea, 0.25 anthrone) and placed on a heater block set at 97º C. Samples were heated for 15 minutes, then removed and cooled to room temperature. Absorbance of these samples was read together with the prepared glucose standards at a spectrometer setting of 612 nm. The content of TNC of each sample was estimated with the regression equation according absorbance of the glucose standards.

Effect of species and water treatments on response variables were determined by analyzing variances within and among treatments (Steel and Torrie, 1960). The Proc GLM procedure of SAS (SAS Institute, 1998) was used to calculate the slope of WUE and to determine linear, quadratic, or cubic effects of water treatments on RWC, canopy light interception and reflection. The Proc GLM procedure was also used to analyze species effects on forage mass. The Proc Mixed procedure was used to determine species effects on total seasonal DM yield. This procedure was also used to determine to effects of water treatments on WUE within each species.

Research results and discussion:

All collection of data and analysis of samples has been completed. Furthermore, the graduate student on this project defended successfully his research and anticipates graduating in May 2004. Results of our research were also presented at the Crop Science Society of America Annual Meeting in Denver, CO, in November of 2003. Findings and conclusions from this research were partially published in progress reports that were presented at our yearly field days in summer of 2002 and 2003. Manuscripts will be drafted in spring of 2004 to prepare and submit publications to scientific journals.

Results from the first year of investigation (2001) suggested that Dahl bluestem (Bothriochloa bladhii) and Caucasian bluestem (B. caucasica) were similar in WUE expressed as slope (19.4 and 18.5 kg ha-1 mm-1, respectively). Spar (B. ischaemum) generated 14.5 kg ha-1 mm-1, but differences with other species were not significant. In 2002, data suggested that Caucasian bluestem generated 19.5 kg DM ha-1 mm-1 that was similar to Dahl with 19.3 kg ha-1 mm-1 and Spar generated 16.3 kg ha-1 mm-1. In 2003, differences (P < 0.05) between Caucasian and Spar bluestem were observed. In that year, Caucasian generated 16.9 kg DM ha-1 mm-1, Spar accumulated 7.4 kg ha-1 mm-1 and Dahl 12.3 kg ha-1 mm-1 (Fig. 1).

Maximum DM production in 2001 was observed in Caucasian bluestem under a high irrigation level (100% replacement of PET) with 11.7 Mg ha-1, followed by Dahl with 11.3 Mg ha-1, and Spar with 7.9 Mg ha-1 at the same irrigation levels (Fig. 2). Observed yields under dryland conditions averaged 3.2, 2.4, and 1.5 Mg ha-1 for Caucasian, Spar, and Dahl, respectively. In 2002, maximum DM production was again observed in Caucasian bluestem with 22.3 Mg ha-1, followed by Dahl (21.5 Mg ha-1) and Spar (18.5 Mg ha-1) under the high irrigation. Dry matter yields obtained from dryland sites were 4.6, 3.0, and 4.2 Mg ha-1 for Caucasian, Spar, and Dahl, respectively. In the last year of our research, Caucasian averaged 19.8 Mg ha-1 in dry matter yield under a high irrigation level; Spar generated 11.3 Mg ha-1 and Dahl 12.7 Mg ha-1. Under dryland conditions, Caucasian averaged 4.1 Mg ha-1, while Spar generated 4.0 Mg ha-1, and Dahl averaged 1.7 Mg ha-1.

Concentration of crude protein was higher (P < 0.05) in Dahl bluestem compared with the other old world bluestems tested averaged over all irrigation levels and dryland (Fig. 3). Fiber concentrations in investigated species were affected by irrigation levels. Neutral detergent fiber increased with increasing amounts of irrigation during all three years of research (Fig. 4). Similar observations were made regarding ADF (Fig. 5). Dry matter digestibility was higher under low irrigation compared with all other irrigation levels and dryland (Fig. 6). Total DDM (kg ha-1) was higher under a high irrigation level compared to all other water treatments (Fig. 7). Leaf/stem ratios were affected by water treatments, showing that with increasing irrigation, plant maturity also increased and, thus, leaf/stem ratios decreased (Fig. 8). Live/dead ratios increased with increments in irrigation (Fig. 9).

Participation Summary

Educational & Outreach Activities

Participation Summary

Education/outreach description:

Philipp, Dirk. 2004. Influence of varying replacements of potential evapotranspiration on water use efficiency and nutritive value of three old world bluestems (Bothriochloa spp.). Ph.D. Dissertation, Texas Tech University, Lubbock.

Project Outcomes

Project outcomes:

This research was presented during field days in June of 2002 and 2003 to over 300 producers from the region. The number of participants at these field days increased steadily over the last years and we conclude that producers from the region understand the urgency of developing sustainable farming systems for the region. An ever recurring concern at these meetings was the amount of optimum water inputs, forage nutritive value, and projected profitability of the proposed grass species that are already used in the area. We recognize that with disseminating detailed information on water consumption and the relation to forage nutritive value of old world bluestems, we can enhance the knowledge base of farmers from the surrounding area such that they can make rational choices to conserve water resources. Spar old world bluestem was widely planted on the Conservation Reserve Program (CRP) acres in this region and is, thus, the species that area producers are more familiar with. Our data clearly indicated that either Dahl or Caucasian provided more biomass and higher nutrient yield (kg ha-1) than Spar under any moisture regime but particularly under limited or no irrigation.

Economic Analysis

Dry matter production per dollar (U.S.) invested in irrigation water was greater for Caucasian than for Spar bluestem across all levels of irrigation, but results were inconsistent for Dahl bluestem. A more detailed economic analysis based on nutritive value as well as total DM yield is being explored.

Farmer Adoption

Old world bluestems in the Texas High Plains have been widely adopted by area producers. Those of the B. ischaemum type were planted on a high proportion of the acreage in the original CRP. Since its introduction in 1994, Dahl bluestem has gained rapidly in popularity with area farmers. Results of our research indicated that Caucasian and Dahl bluestem were superior to the B. ischaemum type Spar bluestem. With superior cold tolerance Caucasian may have greater application than Dahl in the more northern areas of the region, but has been less widely adopted to this point. Results of this research provides the basis for producers to make informed choices regrading selection of species for yield, quality, and water savings.


Areas needing additional study

We suggest to focus further research on incorporation of these grasses into grazing systems. Also, the effect of irrigation levels on mineral content should be investigated. Further research is needed to delineate responses to increments of water under more frequent defoliation events. A more indepth economic analysis should be conducted.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.