|
|
|
|
|
|
|
||
|
|
||
|
|||
1999 STEEP III Final ReportRESEARCH PROJECT TITLE: Rotation designs for direct seed cropping systems INVESTIGATORS:
PROJECT OBJECTIVES:
KEY WORDS: Alternative crops, no-till, water use efficiency, crop rotations STATEMENT OF PROBLEM: Crop rotation designs for direct seed systems have not been extensively developed for the annual cropping region of the Pacific Northwest. Attempts to use conventional crop rotations in no-tillage systems have been largely unsuccessful due to agronomic problems with crop growth, weeds, diseases, and equipment (i.e. drill) performance. For example, a crop rotation of winter wheat, spring barley, and spring pea or lentil requires the establishment of spring crops into relatively large amounts of surface residues in two of three cropping seasons. The residues create a shaded, cold, wet environment that adversely affects shoot and root growth, enhances the potential for adverse weed pressure, promotes soil-borne diseases, increases soil compaction, and reduces drill performance. Furthermore, the rotation emphasis on spring crops does not take advantage of the greater yield potential of fall sown crops in the high rainfall zone. A re-examination
of crop rotation design for continuous direct seed systems may offer methods
to alleviate or avoid many of the constraints of presently used rotations.
Large amounts of cereal residues retained on the surface can benefit the
production of Austrian winter pea and winter lentil (Huggins and Pan,
1991). But what about other winter crops such as canola/ rapeseed, or
spring sown crops such as corn, canola, yellow mustard, linola, millet
and safflower? Little is known about positive or negative effects that
these crops may have in cereal-based direct seed rotations. ZONE OF INTEREST: Field studies will focus on the high rainfall, annual cropping region and modeling efforts will be applicable to the high and intermediate rainfall areas. ABSTRACT OF RESEARCH FINDINGS: Winter wheat residue is a formidable barrier to direct seed rotations in the high precipitation zone of the Palouse. We initiated two studies in the spring of 1998: one focusing on crop modeling of 10 different spring crops following winter wheat, and the other comparing no-till with conventionally established winter and spring crops after winter wheat. The crops were: spring and winter canola, yellow mustard, hard red spring wheat, winter wheat, proso millet, pinto beans, safflower, soybean, linola, winter and spring pea, and winter and spring lentil. Parameters for the Cropsyst model were developed for ten spring crops grown in 1998. Grain yields were similar between no-tillage and conventional tillage in 1999 for linola, safflower, soybean, spring canola, spring pea, and yellow mustard, while yields of spring lentil, spring wheat, dry bean and millet were greater under conventional tillage. The reported results are preliminary and no conclusions can be made at this time. RESULTS AND INTERPRETATION: In 1998, two field studies were initiated at the Palouse Conservation Field Station near Pullman, WA to evaluate winter and spring crop performance following winter wheat under no-till conditions. The crop modeling study consisted of 10 different spring crops: canola (Sunrise), yellow mustard (Tilney), hard red spring wheat (WB 926R), peas (Columbia), corn (Pioneer 3970), proso millet, dry beans (Bill Z pinto), soybeans (Monsanto Roundup Ready), safflower (S-208) and linola (989) on 50 by 30 foot plots. A no-till double disk drill (Fabro, Inc.) with an offset, leading disk, starter and deep band fertilizer capabilities (all 7.5 inch spacing), and a cone seeder was used to no-till seed into standing Madsen winter wheat stubble (grain yield of about 85 bu/ac) in 1998. In 1999, the Cross-slot drill (notched coulter, inverted T slot, 8 inch spacing, banded fertilizer) was used to seed all spring crops. Data for calibrating the Cropsyst model was collected throughout both growing seasons. Daily air temperature, precipitation, global radiation, net radiation, wind speed, and relative humidity were collected by an on-site weather station. Daily seed zone temperature and water (2 inch depth) and root zone water (0 to 5 feet by 1 foot intervals) were monitored with thermocouples and water content reflectometers in each of the ten crops. Crop development, aboveground biomass accumulation, and leaf area index were assessed throughout the season. A combination of hand sampling and a plot combine was used to harvest grain yield. The second study consisted of seven spring crops in 1998 and 14 winter and spring crops in 1998-1999. The crops were: spring and winter canola, yellow mustard, hard red spring wheat, winter wheat, proso millet, pinto beans, safflower, soybean, linola, winter and spring pea, and winter and spring lentil seeded both no-till and conventionally (fall moldboard plow, spring disk), following winter wheat (Madsen). The fabro drill was used in the spring and fall of 1998 and the cross-slot drill in the spring of 1999. Seeding dates and rates, and fertilizer applications used in the tillage comparison and crop modeling studies for the 1998-99 season are in Table 1. Table 1. Agronomics of alternative crops used in modeling and tillage studies.Table 1. Agronomics of alternative crops used in modeling and tillage studies.
Crop modeling study: Two different approaches can be used to calibrate a crop simulation model. One approach is to modify crop-input parameters until simulated data match measured data. In the other approach, crop parameters are obtained directly from measured data. The last approach is the method we are attempting and is more appropriate since the coefficients of the basic growth relationships that the model tries to simulate are obtained directly from measured data. The basic growth parameters required for the CropSyst model that we are obtaining from measured data are the biomass-transpiration coefficient (Tanner and Sinclair, 1983), biomass-light (photosynthetically active radiation--PAR) coefficient (Monteith, 1977), specific leaf area, stem/leaf partition coefficient, extinction coefficients for total radiation and PAR and thermal time requirement. The biomass-transpiration coefficient will be calculated by integrating weather, crop and soil water data. The stage of active growth at which the crops are not water stressed will be used for the calculation. The same stage of active growth will be used to compute the biomass-light (PAR) coefficient. This coefficient will be estimated using measured radiation data and data for the fraction of radiation intercepted by the canopy. Extinction coefficients will be computed using LAI and intercepted PAR sampled during the growing season. Specific leaf area and stem/leaf partition will be calculated from LAI, total dry matter and partition data. Observed phenology will be used to compute thermal time requirement.All crops except soybean reached maturity before fall frosts in 1998. Dry bean stands were extremely poor in 1998 because of cold temperatures during establishment so no data were collected. All crops were successfully established in 1999. Total dry matter (TDM), leaf area Index (LAI) and PAR intercepted by the canopy were measured over time during 1998 and 1999. Table 2 shows TDM, harvestable dry matter (HDM) and Harvest Index (HI) in 1998. Grain yields and water use efficiency of the 1998 crops were reported last year. Table 2. Total aboveground dry matter (TDM), harvested dry matter (HDM) and harvest index (HI) measured for yellow mustard (YM), spring canola (SC), linola (Li), spring pea (SP), safflower (Saf), spring wheat (SW), soybean (Soy) and corn (Co) at the Palouse Conservation Field Station, Pullman (WA) during 1998.
Figure 1 shows TDM and LAI evolution during the growing season for yellow mustard. Similar information was obtained for the rest of the crops.
Figure 1a and b. TDM (1a, left) and LAI (1b, right) development of yellow mustard at the Palouse Conservation Field Station, Pullman, WA, during 1998. Figure 2 shows PAR intercepted by the crop canopy 1998 and Figure 4 shows soil water depletion to a depth of 1.5 m during the 1998 growing season for each crop. Lightning damaged some of the sensors used during 1998. Gravimetric samples of soil water content were collected to complete soil water monitoring.
Figure 2. Fraction
of PAR intercepted by the canopy of the crops that completed maturity
at the Palouse Conservation Farm, Pullman (WA) during 1998.
Crop, weather, light interception and soil water data were integrated for the computation of key model parameters. The parameters were calculated directly from measured data. Table 3 shows the list of key growth parameters for CropSyst obtained from the 1998 measured data. There were problems measuring leaf area of Linola. The leaves of this crop are very small and the samples taken were not large enough to obtain good measurement of leaf area and related parameters (extinction coefficients for PAR and global solar radiation, stem/leaf partition coefficient and specific leaf area). Therefore, model parameterization using 1998 data for that crop will be attempted by matching simulated and observed data. Table 3. Estimated
CropSyst crop parameters. Parameters calculated from measured experimental
data at Pullman (WA) during 1998.
Biomass-light (PAR) coefficient (Monteith, 1977) with no water stress was calculated using daily PAR and the fraction of PAR intercepted by the canopy. Figure 4 (left) shows the relationship between measured LAI and noon PAR intercepted by yellow mustard during 1998. Values measured at 10-day intervals were integrated over the growing season. From this relationship, noon extinction coefficient for PAR (K PAR) was calculated. Daily extinction coefficients for total global radiation (K Total) were calculated using canopy descriptors presented by Campbell and van Evert (1994) from noon K PAR. The parameter x in Table 3 describes canopy structure. The higher the value, the more horizontally the leaves are distributed in the canopy. Finally, daily PAR intercepted by the canopy was computed using data of measured LAI, K PAR and daily global solar radiation measured by the weather station. Total dry matter accumulated until the occurrence of soil water stress was used for the computation of the biomass-light (PAR) coefficient. The date when water stress occurred was calculated as the point when the soil water to a depth of 0.6 m decreased below 0.6 of the available soil water. Figure 4 (right) shows the relationship between unstressed total dry matter accumulation and intercepted PAR for yellow mustard.The slope is the biomass-light (PAR) coefficient. Similar procedures were performed for all the crops with reasonably computable data.
Figure 4. Relationship between measured LAI and noon PAR intercepted by the canopy (left) and relationship between unstressed TDM and intercepted PAR (right) for yellow mustard at Pullman, WA in 1998. These relationships are used to estimate the biomass-light (PAR) coefficient (Monteith, 1997). The biomass-transpiration
coefficient (Tanner and Sinclair, 1983) was computed from the relationship
between unstressed total dry matter accumulation and the ratio of daily
transpiration over daily vapor pressure deficit (Figure 5).
Figure 5: Relationship between unstressed total dry matter and the ratio of daily transpiration over daily vapor pressure deficit for yellow mustard at Pullman, WA, 1998.. This relationship is used to estimate the biomass-transpiration coefficient (Tanner and Sinclair, 1983). Daily vapor pressure deficit was computed from data collected by the weather station. Transpiration was computed as the product of potential evapotranspiration and the fraction of global solar radiation intercepted by the canopy. Potential evapotranspiration was calculated with the Penman-Monteith formula using CropSyst. Daily weather data collected at the site was input to the model. Daily global solar radiation intercepted by the canopy was computed using data of measured LAI, K Total and daily global solar radiation measured by the weather station. These preliminary
results indicated that measured field data gave reasonable estimates of
crop parameters. Most of these values are in the range suggested for simulation
of these types of crops with CropSyst (Stockle and Nelson, 1998). However,
because of unknown adaptability of the crops (soybean, corn) to this region
and special characteristics of the crops (linola, canola), better estimation
of the crop parameters is needed. Model corroboration: Four corroboration plots were identified with key crops during 1999. Two plots are located at Sprague and Dusty, WA, respectively. Weather stations were installed in these plots. Initial soil water content and mid season light interception were also sampled at these plots. The other two plots are located at Harrington and Lind, WA, respectively. William Schillinger, Research Agronomist at Washington State University Lind Research Station, will provide soil, weather and crop data collected during 1998 and 1999 from these plots. Assembly of environmental and economic databases: Historical weather and regional soil databases will be used to perform the long-term simulations needed to estimate alternative crop production potential and profitability. Historical weather information from the database NCDC Summary of the Day (EarthInfo Inc., CO) will be used as the base weather data. Weather data from stations inside and near the boundary of the study area will be used as input to a spatial interpolation process. To facilitate accurate interpolation, the selected weather stations will be filtered to assure a minimum number of years of data. The use of stations outside of the study region should decrease boundary errors. The daily weather data from the selected stations will be spatially interpolated to increase data resolution. When simulating regional crop performance, there are two possible options to increase spatial data density when a greater continuity of the outputs is desired. One option is to make model predictions with non-interpolated input data and interpolate the outputs. The other option is to interpolate the inputs first. In the case of the dryland areas of the PNW, where weather stations are not very well distributed according to precipitation and temperature variation the best option is to the interpolate inputs and then perform the simulations. To perform this interpolation and for further use in other applications of this work, a program for the management of inputs, outputs and data analysis will be built using Visual Basic for Applications (VBA) under the environment Excel. This program will be used to interpolate daily historical weather from the weather stations to a regular grid of points that covers our study region. The long-term simulations will be performed at each of these points. The interpolation program will also create the weather files needed by the crop simulation model. Soil information from the State Soil Geographic (STATSGO) database for Oregon, Idaho and Washington will be used in the long-term simulations. The soil information will be integrated with the grid of simulation points by performing overlay operations with a geographic information system (GIS). Using the attribute files of the STATSGO database and the County Soil Surveys (Soil Surveys, USDA-ARC, Soil conservation Service), soil parameter files for the crop simulation model will be created for each simulation point. The geographic information system will be used to link each simulation point to weather and soil data. Soil and weather data bases have been identified and analyzed. The software program for input, output, data management and analysis have been built. The program for spatial interpolation and cross validation of the weather data has been written. Enterprise budgets for each crop will be computed using experimental data from the site at the Palouse Conservation Farm. Executed farm operations and inputs used for the establishment and maintenance of the crops will be used to calculate total variable costs at Pullman according to the computer program MACHCOST of the Washington State University Cooperative Extension. Adaptation analysis: The production of each crop will be simulated at each point in the simulation grid for a period ranging from 30 to 40 years depending on the length of the historical weather record for nearby weather stations. The simulations will be performed with different conditions of initial residue load and initial soil water content. These initial residue and soil water conditions will represent different situations left by preceding crops. Agronomic Information from the calibration and corroboration experiments will be used to classify the alternative crops into their post harvest conditions (e.g. residue levels, depth of soil water extraction). This information will be useful in designing crop rotations. For example, what is the relative effect of following a high water using crop versus a low water using crop? CropSyst outputs will be organized using the same spatial grid as used for the input files and thus forming output maps within the GIS. In addition to simulated yield, we will analyze crop phenology and physical variables related to soil water balance. With these factors we can assess a crop's adaptation to the environment of a given point in our study area. Cumulative probability distribution functions for the outputs will be calculated at each simulation point. This will allow the analysis of the output information both geographically and probabilistically. The yield data produced will be used to perform an economic and region-wide analysis of the profitability and risk of growing alternative crops in the dryland region of the PNW. An interpolating technique will be used to calculate net returns over total variable costs and break-even prices and yields at a region-wide scale. Budgets for different counties and rainfall areas computed by the Washington State University Cooperative Extension will be used to find a regional variation in budget levels. This variation will be described with county coefficients. Each coefficient will represent the fraction of the county budget in relation to the budget calculated at Pullman. The coefficients of each county will be calculated based on published total variable costs for winter wheat with no-till. The coefficients of each county for total variable costs for winter wheat with no-till will be interpolated to each point in the simulation grid. The total variable costs for each alternative crop at each simulation point will be calculated based on the total variable cost of each alternative crop estimated at Pullman locally and affected by the interpolated coefficient. Gross returns will be estimated as the product of simulated yield and market price minus other additional costs. Using yield probability information, the probability of break-even prices and break-even yield will be calculated at each simulation point. This information will be summarized in a thematic map format. This will result in maps of the region summarizing key findings such as the probability of the break-even yield with the current market price for a given alternative crop. No-till (NT) versus
conventional tillage (CT) following winter wheat: Winter crops established
poorly in high residues under NT during the fall of 1998 (Table 4). Dry
soil conditions and residue tucking during seeding by the double-disk
drill created an unfavorable environment for seed germination and establishment.
Cold temperatures may also have limited the fall establishment of winter
canola. By spring, stands of winter peas, winter canola, and winter wheat
were nonexistant. The lack of snow cover during winter, heavy rodent damage,
and fluctuating warm and cold temperatures during early spring limited
winter survival. Adequate stands of winter lentils remained, however,
spring stands were irregular in NT (3.7 plants/sq. foot) as compared to
CT (11.7 plants/sq. foot). Due to excessive residue tucking, we switched
to the cross-slot drill for planting of spring crops. Greater establishment
occurred with linola and spring wheat under CT than NT while spring pea
establishment was greater with NT. Grain yields were similar between the
two tillage regimes for linola, safflower, soybean, spring canola, spring
pea, and yellow mustard. Grain yields of spring lentil, wheat, dry bean
and millet were greater under CT (Table 4). Yield of soybean and dry bean
were limited by early September frost that occurred while the beans were
flowering. Yellow mustard and safflower both performed well under no-till
conditions. Safflower in less water limiting landscape positions remained
actively growing until late fall despite temperatures in the 20's (F). INTERACTION WITH OTHER SCIENTISTS CONDUCTING RELATED ACTIVITY: Interaction of scientists
on this study has been excellent with frequent exchanges of ideas, expertise,
equipment. Greater interaction with scientists at the University of Idaho
and at Oregon State is needed as the study progresses, particularly for
validation of modeling efforts. PUBLICATIONS AND PRESENTATIONS: Preliminary results
were presented to approximately 150 people at the Palouse Conservation
Field Station during the July 8th field day. Crop modeling results were
presented at the American Society of Agronomy meetings in Salt Lake City,
Utah (Nov. 4th, 1999). Table 4. Alternative crop establishment, development, growth, and grain yield under no-till (NT) or conventional tillage (CT) treatments following winter wheat.
|
||||
 Contact
us: Hans Kok, (208)885-5971
|
Accessibility | Copyright
| Policies | WebStats | STEEP Acknowledgement |
||||
