Location: Agricultural Systems Research Unit
2013 Annual Report
Dwindling water supplies for irrigated crop production is the most limiting factor facing agriculture in the world today. In the evolving scenario, there is a need for making agricultural water use more efficient by bringing in up-to-date science based technologies in the irrigation field. Innovative decision support systems developed with reliable cropping system models help in the efficient allocation of limited water resources, and its management by the farmers of the region. However, adequacy of the agricultural system models for this application, especially in limited irrigation water management, depends upon accurately simulating the imposed soil water stress effects on crop growth and yield. Inadequate crop growth simulations and the need for enhancing the water stress quantifications in many cropping system models including RZWQM2 have been reported in the literature. In this study, we investigated three progressive modifications of the current water stress (WS) factor in RZWQM2 (WSI1, WSI2, WSI3) for simulating response of corn to different levels of water and compared with the use of current factors. WSI1 was a modification of SWFAC factor for photosynthesis related processes in RZWQM2 using the daily potential root water uptake (TRWUP) calculated by Nimah and Hanks (1973) approach; WSI2 was WSI1 with accounting for stress due to additional heating of canopy from unused energy of potential evaporation; and WSI3 was a modification of TURFAC stress factor in RZWQM2 for leaf expansion that also included the canopy heating effect as in WSI2. These factors were evaluated on the data for corn grain yield, biomass, soil water and canopy cover from multiple water-level experiments conducted at Greeley, Colorado from 2008 to 2011; irrigated and rainfed corn at Akron, Colorado; and irrigated corn at Zaragoza, Spain and Gainesville, Florida, on different soil types. Out of the three water stress factors tested, WSI2 was found to be better than others in simulations of corn grain yield, biomass and LAI.
2. Impacts of initial soil water and level, and interval and method of irrigation on corn N uptake and water production functions in Colorado.
With competing demands for fresh water from various sectors of the burgeoning human enterprises, increasing productivity of the water allocated for irrigation is important for sustained food security on the earth. In the evolving agricultural scenario, deficit irrigation is emerging as a better management practice to increase crop water use efficiency. However, the field experiments conducted on optimization of limited water irrigation for agricultural production represent only the soil and climate of the location and time of the experiments. For timely evolution of limited water management practices, it is paramount that the field experiments are integrated with robust agricultural system models for extrapolation of results across locations leading to development of new management practices for enhanced agricultural production with less water consumed. In this study, we used the RZWQM2 modified with Nimah and Hanks (1973) approach for water uptake and WSI2 for water stress quantification, to simulate corn at Greeley, CO and integrated with the long-term weather records at the location and developed mean crop water production functions (CWPFs) as functions of evapotranspiration (ET) and applied water. We further explored ways of enhancing the crop water use efficiency of available soil water and applied water, with minimum N leaching. At varying initial soil water levels (average measured, and 50, 75 and 150 % of average measured), irrigations were applied to supplement rainfall in order to meet the potential crop ET demand at 100% to 0% (full irrigation to dryland) levels. The interval of irrigations varied from 3, 6, 10 and 14 days. These irrigation intervals encompass the intervals with drip, sprinkler, and furrow or flooding methods of irrigation. The results of the study will be used to develop site-specific recommendations for efficient water and N management.
3. Climate change impacts in agriculture (The Coordinated Climate-Crop Modeling Project C3MP)
The Coordinated Climate-Crop Modeling Project (C3MP) is an initiative of the Agricultural Model Intercomparison and Improvement Project (AgMIP). C3MP is coordinated from the NASA Goddard Institute for Space Studies in New York City, the site of the AgMIP coordination Office. All AgMIP work on crop models is coordinated by the AgMIP crop team, led by Dr. Kenneth J. Boote (University of Florida) and Dr. Peter Thorburn (CSIRO, Australia). The AgMIP seeks to mobilize international crop modelers for a coordinated investigation of climate vulnerability and climate change impacts on agriculture. Collaborations among crop modelers will provide new estimates of how agricultural production will be impacted by climate change and will help assess the consistency of these estimates across climate and crop models. The C3MP participants were invited to contribute results from their own site-calibrated crop modeling experiments using a set of common climate sensitivity tests. C3MP results will contribute to wider assessments undertaken by AgMIP to provide consistent (across crop models), timely and relevant information about climate change and agriculture for decision-makers. Participating in the first phase of the C3MP, using RZWQM2 we simulated the range of irrigated corn production responses at Greeley and Akron, CO to climate change (changes in CO2 concentrations, temperature, and precipitation). The climate change data used were designed to efficiently sample the uncertainty space in projected temperature, water, and carbon dioxide changes in the 21st century [global climate model (GCM) outputs contributed to the Fifth Coupled Model Intercomparison Project (CMIP5), IPCC AR5- to be released in Sept. 2013]. Simulations were conducted using two types of climate inputs. One is observed climate data for the historical baseline period 1980-2010 at Greeley and Akron. The second type of climate input was the s-Merra dataset developed by C3MP to provide a uniform dataset for testing at all sites across the world that also can be used where daily climate data are not available. Simulation results were communicated to C3MP Coordinators for analysis and comparison with other locations for general and relative responses.
4. Application and improvement of GPFARM-Range model to simulate interactions among soil water content, forage production, cattle weight gains, and stocking density
System modeling can help analyze the effect of climate variability on forage growth and cattle production after careful calibration of the model. In this project, we tested the GPFARM –range model for simulating soil water dynamics and long-term (1982-2012) forage production and cattle weight gains under different stocking densities in a semiarid northern mixed-grass prairie at Cheyenne, Wyoming. The interactions between forage production and cattle growth across various climate conditions were explored. The GPFARM-Range model has not been tested for long-term (multi-year) simulations of forage production and cattle weights and their interactions across various climate conditions. In this project, we first calibrated the model for simulating annual peak standing crop biomass (PSC) using observed PSC data from 1982 to 2012 at Cheyenne, WY. The model gave good predictions of PSC with root mean square error (RMSE) of 356 kg ha-1, Nash-Sutcliffe Model Efficiency (E) of 0.64, and index of agreement (d) of 0.9. We further tested the model for predicting cow-calf weights during the same period. The model predicted cow-calf weights very well with RMSE value of 21.8 kg and E of 0.99. Current results suggest that the GPFARM model can be used to simulate long-term forage production and cow-calf weights in semiarid northern mixed-grass prairie. For both measured and simulated data, the cow and calf weight gains did not seem to be affected by PSC values. This result suggests that forage production was not a limiting factor to cow-calf weight gain under the current stocking density. We also used the GPFARM steer model to simulate steer growth at three stocking densities from 1982 to 2012 at Cheyenne, WY (8 steers/40 ha; 8 steers/24 ha; 8 steers/18 ha). Both the measured and simulated PSC values showed a decrease in forage production with increasing stocking density. Both measured and simulated steer weight per steer showed no statistically significant difference among these three stocking rates. However, both measured and simulated steer weights (kg head-1) did show an increasing trend with decreasing stocking density.