|
|
|
|
Xuejun Dong, Paul Nyren, Bob Patton, and Anne Nyren,
NDSU Central Grasslands Research Extension Center, Streeter, ND 58483
In the Missouri Coteau of North Dakota, the rolling landscape and stony soils limit tillage but make the area suitable for cattle grazing, a major source of income (Barker and Whitted, 1993; Nyren, et al. 1996; Patton, et al. 2001). To increase animal productivity, it is important to understand how animal intake is related to the rest of the grassland ecosystem. We use a physiology-based ecosystem model, the Hurley Pasture Model (Thornley and Verberne, 1989; Thornley, 1996; Thornley, 1998), combined with systematic field measurements, to explain mechanisms of grazing-induced grassland changes that have an impact on management (Figure 1). This study will be used to establish practical guidelines for long-term range management in the Missouri Coteau area.
In this project, two problems are addressed. The first is the analysis of management practices such as grazing and fertilization to increase economic gains and ecological sustainability considering dynamic soil-plant-climate interactions. This analysis can be used to explore better management solutions for different soil and climate conditions, which are useful for the managers of the grasslands and yet not apparent without system analysis. The second is the effect of climatic drought on plant water use and tissue water status. In the Missouri Coteau, climatic drought strongly limits forage production (Biondini, et al. 1998; Kirby, et al. 1999), and this effect takes place through a series of physiological processes among which is the drought-related decrease of plant water status and water use.
The importance of system modeling for range management was discussed by Dong et al. in the 2001 CGREC Annual Report. One of the strengths of ecosystem modeling is that this experiment-and-measurement-based research tool has the potential to explore possibilities and problems that are difficult to address with conventional field or experiment testing but are relevant to management. Different issues can be considered under the ecosystem context: fertilizer applications, effects of drought, introduction of new species to the grasslands, etc. In these efforts, the role of computer and mathematical modeling is to provide “…a powerful means of integrating several items of knowledge at one level to describe responses at a higher organizational level. Such integration may be valuable for the purpose of research, leading to understanding, or for the purpose of crop management when the models have a practical predictive value. ” (Thornley and Johnson, 1999).
The Hurley Pasture Model and its Usefulness in the Missouri Coteau
Models have been used to study different problems in this part of the Northern Great Plains. For example, Biondini (2001) used a three-dimemsional model to study root resource use in heterogeneous soils; and Patton et al. (2002) used Neural Network to predict seasonal dynamics of forage quality based on climate data. These models and the research schemes have their own well-defined objectives and strengths. For the current proposal, we need a research framework that is simple enough to handle the soil-plant-animal production system as a whole, but is dynamic enough to include major soil-plant processes. We chose the Hurley Pasture Model (HP model) based on the following considerations:
(a). The scale of a physiologically based ecosystem model such as the HP model is relevant to the CGREC grazing systems. Under grazing, the most important process is the partial removal of the photosynthetically active shoots of plants and subsequent physiological and morphological adjustments by plants in response to this mechanical impairment. These dynamic changes can be described using the HP model. The causal-relationships could be helpful to understand the ecological mechanisms of why a moderate utilization of rangelands is frequently recommended (Barker and Whitted, 1993; Nyren, 1996; Biondini, et al. 1998; Patton, et al. 2001; Richardson, 1996).
(b). The model is theoretical at the whole system level but it relies on field data to describe processes such as photosynthesis and nitrogen uptake. This makes the model computationally realistic and eliminates the possibilities that many levels of description might dilute the main cause-effect connections. In addition, this allows new experimental results to be incorporated into the model.
(c). The HP Model is based on well-accepted eco-physiological principles and its robustness and potentials have become evident through applications to a wide range of ecological problems (Thornley and Verberne, 1989; Thornley, 1998 and references therein). For example, it allows a variable shoot-root partitioning and divides plant biomass into structural (such as cellulose) and storage (such as soluble carbohydrates) fractions. This facilitates explicit treatment of plant shoot-root interactions and a description of the physiological mechanisms plants use to tolerate drought stress.
The ultimate objective of this study is to explore the dynamic relationships between grassland production and environmental and management variables. Specifically, it is comprised of three research objectives: (1) To measure specific eco-physiological features of nine range plants and soil characteristics for the understanding of important plant-soil processes and for the estimation of some parameters for the Hurley Pasture Model. (2) To test the Hurley Pasture Model for use in the management of grassland including animal stocking density and timing and quantity of fertilizer application under different soil and climate (light, temperature) conditions in the Missouri Coteau of North Dakota. (3) To test a water simulator (Thornley, 1996) for use in the description of plant water status and water use for a closed grass canopy approximately corresponding to moderately grazed and idled pastures in response to climatic drought in this area.
The work for the first objective began in 2001. Work continues on the second objective dealing with the whole model. The main ideas to carry out the third objective were outlined in the 2001 CGREC Annual Report by Dong et al.
The Missouri Coteau grassland supports a diversity of plant species with high forage values, but for practical reasons, we selected for field measurement four cool-season grasses: Kentucky bluegrass (Poa pratensis), western wheatgrass (Agropyron smithii), intermediate wheatgrass (Agropyron intermedium), tall wheatgrass (Agropyron elongatum); a warm-season grass, blue grama (Bouteloua gracilis); two legumes, alfalfa (Medicago sativa) and yellow sweetclover (Melilotus officinalis); a forb such as stiff goldenrod (Solidago rigida); and a shrub buckbrush (Symphoricarpus occidentalis). In our application we plan to use the following strategy:
a. Special emphasis will be given to the field measurements for some model parameters. According to a previous study (Thornley, 1998), the plant sub-model is more changeable than the soil sub-model. So our “key parameters” in the plant sub-model refer mostly to photosynthesis, water relations and nitrogen uptake. At the CGREC, a rain-out-shelter facility (Nyren and Patton, 2001) has been set up and a study on the impact of drought on plant community dynamics is under way (Kirby, et al. 1999). This facility will be used in field measurements.
b. Initially, we will use the simpler version of the HP model to study the soil-plant-animal interactions. The core part of the model changed only slightly from its first version to the latest version. We will use this model as a tool to explore better management strategy for grasslands and for qualitative predictions.
c. The water sub-model will be considered separately. This portion of the HP model runs with a shorter time interval (projecting the initial values 20 minutes into the future and then repeating the process until it reaches the desired end point versus a one-half day interval for the first version of the HP model) and sometimes becomes unstable under very dry conditions. So, in our initial application of the model it is wise to treat this sub-model separately.
d. Due to difficulties in finding species-specific parameters, the model will only consider cool season species. The warm-season grass species will not be considered in the modeling exercise of this particular study.
Plant, soil and animals are major components of the grassland ecosystem. In this study, the plant component is considered in detail and the majority of fieldwork will focus on this component; the soil component is considered in some detail and the field measurement will only address selected variables; the animal component is the most simplified. We will use a graphic simulation program STELLA (High Performance System, 1998) for the computer implementation of the study.
We measured water potentials of two perennial grasses, Kentucky bluegrass and western wheatgrass, under two treatments with cattle grazing and a control, using a pressure chamber (Plant Moisture Stress Equipment, Inc. Corvallis, OR).
During the field season, we took the following measurements: 1) Specific leaf area (SLA, ratio between leaf area and leaf dry mass) for dominant plant species in the grazing pasture, 2) rate of leaf gas exchange (carbon dioxide and moisture), (3) leaf water potential (a measure of plant water status). Only leaf water potential is summarized in this report.
This variable is required in the HP model to simulate plant growth and to calculate plant water use. It involves measuring projected leaf area and usually an area meter is used in the measurement. The accuracy of these meters themselves can be 0.1%. However, this high precision can be meaningless when compared with the much larger errors arising from the preparation of leaf samples, especially when leaves are not laid flat. This problem becomes even more serious when measuring the usually narrow folded leaves of native grasses. It is not surprising that actual measurement of leaf area in some grasses such as Kentucky bluegrass is sometimes avoided (Day and Detling, 1994; Knops and Reinhart, 2000). We tried to overcome this difficulty and accurately measure the leaf area of native grasses by laying the leaves flat on clear laminate paper, scanning the image into the computer and using an image analyzer to calculate leaf area. Though time consuming, increased accuracy is the main advantage of this method. Resulting data will be analyzed later this year. We expect that this measurement will provide insights to the “workings” of the grasslands under grazing pressure, in additional to its uses in the modeling work.
Plant leaves have tiny pores on the surface called stomas which control gas exchange (water vapor, and carbon dioxide) between the inside of the leaves and the outside air. These activities are the bases of survival and growth for plants. Measurement of gas exchange will provide an estimate of several parameters of plant photosynthesis. Also the values will be used as indicators of plant physiological vigor. The results will not be discussed in this report.
To predict the daily water potential wave, the HP model needs to calculate both the pressure and osmotic components. However, the range of physiological viable potentials is predominantly governed by pressure potential and it changes in an exponential way but the underlying mechanism is not well understood. Thus experimental data is required to estimate the amount of water loss or gain with a unit change in leaf water potential (Nobel, 1999). The osmotic potential changes linearly with leaf water content; it governs total leaf water potential only when and after the leaf continues to lose water and become wilted. The two parameters of this linear function also need experimental determination (See Dong and Zhang, 2001). We have some preliminary analysis of the water potential data and will discuss our findings in more detail in the following section.
Plant-water Relations and Implications to Range Management
Water is not just an ingredient of plant structure. Water serves as a medium for biochemical reactions to take place. It is also the main carrier for the transportation of materials both within plants and between plants and soils and helps maintain plant temperature, etc. Plant water status is usually expressed using a pressure unit: water potential (bar or MPa). Though this measurement is only an approximation of real water status in plant structures (Roderick, 2001), its use is widely supported. This pressure unit serves as a common measure to unite two major aspects of plant water status: the turgidity and solution concentration of leaves. Thus for grasses and other plants of low-stature, total water potential is the sum of a pressure potential (a non-negative value describing average turgidity of leaves) and an osmotic potential (a non-positive value describing average solution concentration within leaves). Increasing the pressure potential, decreasing the osmotic potential, or both, can increase the plants’ capacity to survive drought stress (Niinemets, 2001).
The air temperature and precipitation throughout the growing season are shown in Figure 2. Results of plant water potentials along with major weather data are summarized in Figure 3. Under non-stress situations, plant water potential undergoes a daily change. It is highest near predawn, drops gradually with sunrise and reaches minimum level at noon or early afternoon. As evaporation decreases water potential remains more or less stable increasing through the evening hours until mid-night or early morning when water is recharged and plant water potential is equilibrated with soil water potential of the root zone. This equilibration suggests some sort of correlation, if not absolute equality, between soil water potential and predawn plant water potential. This is supported by field measurements (Sala, et al. 1981). The physiological importance of this correlation could be interpreted as: the higher the correlation, the more likely the water from a particular soil layer contributes to plant shoot or leaf water potential. Figure 4 shows the relations between predawn leaf water potential (Figure 3) and soil water content (means shown in Table1). Regression lines are used where significant linear relations exist.
In the literature, several studies conducted in the Northern or Central Great Plains have shown that heavy grazing, when compared with non-grazing or light grazing, leads to a better water availability both in the soil and the remaining plants. This difference is a result of conservation of soil moisture due to reductions in transpiring leaf area and plant water consumption in heavily grazed pastures compared with lightly grazed pastures (Archer and Detling, 1986; Svejcar and Christiansen, 1987; Day and Detling, 1994).
The results of our field observations showed a different story. While western wheatgrass did not show dramatic change during the very dry season of 2002, water potentials of Kentucky bluegrass responded sensitively to water availability. A better water status in Kentucky bluegrass was observed in the exclosure where the top 6 inches of the soil had considerable moisture content even in the very dry months of May and June (precipitation of these two months was about 78% and 31% lower, respectively, than the long-term means (see Table 2) and Figure 5 for soil moisture data). In pastures of both moderate and extreme grazing, water potentials of the tillers were extremely low in May and June, with tillers drying out during the measurement. Analysis of soil–plant–water relationship suggests that (a) tillers of Kentucky bluegrass at all three grazing treatments appeared to rely on soil moisture in the top 6 in. depth (Figure 4A); (b) in addition, tillers of plants in moderate grazing may also rely more on a deeper layer (6-12 in.) for water than those in extreme grazing (Figure 4B). This is consistent with the observation that plant rooting depth under moderate grazing could reach below the 6 in. soil depth, while rooting depth under extreme grazing or non-grazing had little growth below 6 in. depth (Richardson, 1996).
Although Kentucky bluegrass plants in the exclosure had the most favorable water status from our field observation, grazing rather than idling is a correct way to maintain a high forage value for the Kentucky bluegrass-dominated grasslands (William Barker, personal communication). Heavy or extreme grazing, on the other hand, could exacerbate the drought stress to the plants. As a result, moderate grazing may be an appropriate choice from a soil-plant-water relations point of view in this particular grassland. But the intensity of grazing may need to be lowered (not known just from this study) in times of drought, because Kentucky bluegrass tillers have a very limited capability of drought tolerance. In this plant, water use is rather opportunistic and leaves begin to fold inward under moisture stress of about -23 to -25 bar. Leaf gas exchange rates (photosynthesis and transpiration) drop following the folding of leaves (data not shown in this report). On August 27, 2002 for example, when leaf water potential dropped to about –50 bar, measured photosynthesis became negative and stomas closed. A week later, following a major rain event, physiological rates increased to the highest possible values for this plant. Because of this opportunistic mode of water use, we suggest that soil evaporation rather than plant transpiration is the major atmospheric moisture output in the Kentucky bluegrass-dominated grasslands under a drought situation. This may be one of the reasons why a different trend was observed in our study site as compared with the published results by Archer and Detling(1986), Svejcar and Christiansen (1987) and by Day and Detling (1994). Another consideration is the type of grazing animal. Heavy grazing with cattle can decrease the soil infiltration rate by increasing soil bulk density (Engels, 2001), while grazing by the prairie dog (Archer and Detling, 1986; Day and Detling, 1994) might improve soil moisture due to increased infiltration through the prairie dog burrows.
In western wheatgrass, however, there was only a slight response in water potential to drought, and there was no significant linear relationship between predawn water potentials and soil water content in the extreme grazing and exclosure treatments; however, in moderate grazing, a linear relationship exists for the top 6 inches of soil (Figure 4C). This suggests that western wheatgrass may use soil moisture from a wider range of depths. Also the fact that this plant has a rigid leaf blade suggests a high capability of drought resistance. Further study using the automatic-rainout-shelter (Nyren and Patton, 2001) is required to see how the plant responds to more severe drought.
We take this opportunity to thank the CGREC staff for their continued support. Also, we’d like to thank following people for support and participation in this work: Drs. Ed Deckard, William Barker, Don Kirby, Mario Biondini, David Hopkins and Jimmie Richardson at NDSU Fargo campus and Mr. Brian Kreft at the CGREC. We appreciate the help from Bethel Zurmiller at CGREC for measuring leaf area.
| NDSU Central Grasslands Research Extension Center |
| Home | 2002 Annual Report |
