Allan A. Andales

Elevated CO2 further lengthens growing season under warming conditions.

Observations of a longer growing season through earlier plant growth in temperate to polar regions have been thought to be a response to climate warming. However, data from experimental warming studies indicate that many species that initiate leaf growth and flowering earlier also reach seed maturation and senesce earlier, shortening their active and reproductive periods. A conceptual model to explain this apparent contradiction, and an analysis of the effect of elevated CO2—which can delay annual life cycle events—on changing season length, have not been tested. Here we show that experimental warming in a temperate grassland led to a longer growing season through earlier leaf emergence by the first species to leaf, often a grass, and constant or delayed senescence by other species that were the last to senesce, supporting the conceptual model. Elevated CO2 further extended growing, but not reproductive, season length in the warmed grassland by conserving water, which enabled most species to remain active longer. Our results suggest that a longer growing season, especially in years or biomes where water is a limiting factor, is not due to warming alone, but also to higher atmospheric CO2 concentrations that extend the active period of plant annual life cycles.

Incorporating tillage effects into a soybean model.

Crop growth models can be useful tools in evaluating the impacts of different tillage systems on the growth and final yield of crops. A tillage model was incorporated into CROPGRO-Soybean and tested for conditions in Ames, IA, USA. Predictions of changes in surface residue, bulk density, hydraulic conductivity, runoff curve number, and surface albedo were consistent with expected behaviors of these parameters as described in the literature. For conditions at Ames, IA, the model gave good predictions of soil temperature at 6 cm depth under moldboard (R2=0.81), chisel plow (R2=0.72), and no-till (R2=0.81) for 1997 and was able to simulate cooler soil temperatures and delayed emergence under no-till in early spring. However, measured differences in soil temperature under the three tillage treatments were not statistically significant. Excellent predictions of soybean phenology and biomass accumulation (e.g. R2=0.98, 0.97, and 0.95 for pod weight predictions under moldboard, chisel plow, and no-till, respectively) were obtained in 1997. More importantly, the model satisfactorily predicted relative differences in soybean growth components (canopy height, leaf weight, stem weight, canopy weight, pod weight, and number of nodes) among tillage treatments for critical vegetative and reproductive stages in one season. The tillage model was further tested using weather and soybean yield data from 1995 to 1997 at Nashua, IA. Tillage systems considered were no-till, disk-chisel+field cultivator, and moldboard plow+field cultivator. Predicted yields for the 1996 calibration year were within 1.3% of the measured yields for all three tillage treatments. The model gave adequate yield predictions for the no-till (−0.2–3.9% errors), disk-chisel (5.8–6.9% errors), and moldboard (5.5–6.1% errors) tillage treatments for the two years of validation. A sensitivity analysis showed that predicted soybean yield and canopy weight were only slightly sensitive to the tillage parameters (less than 3% change with 30% change in tillage parameters). The model predicted lower yields under no-till for nine out of 10 years of weather at Ames, IA, primarily due to delayed emergence. Yield under no-till was higher for one of the years (a drought year) when no-till had better water conservation and negligible delays in emergence.

Evaluation of GPFARM for Simulation of Forage Production and Cow–Calf Weights

Abstract A modeling approach that assesses impacts of alternative management decisions prior to field implementation would reduce decision-making risk for rangeland and livestock production system managers. However, the accuracy and functionality of models should be verified before they are used as decision-making tools. The goal of this study was to evaluate the functionality of the Great Plains Framework for Agricultural Resource Management (GPFARM) model in simulating forage and cow–calf production in the central Great Plains. The forage production module was tested in shortgrass prairie using April–October monthly biomass values from 2000 through 2002 for warm-season grasses (WSG), cool-season grasses (CSG), shrubs, and forbs. The forage module displayed excellent (99% explained variance) agreement in the 2001 calibration year in tracking growth and senescence trends of WSG and CSG, which constitute the vast majority of the aboveground biomass. Less agreement (35%–39% explained variance) was observed for shrubs and forbs. The model-explained variances of biomass in 2000 and 2002 (verification years) were 80% for WSG, 67% for CSG, 78% for shrubs, and 82% for forbs. Further development is needed to improve predicted plant response to environmental stresses. The cow–calf production module was tested in northern mixed-grass prairie using June–November monthly average cow and calf weights from 1996 through 2001 for March-calving, moderately stocked Hereford pairs. Overall, GPFARM performed well and tracked cow (81% explained variance) and calf (94% explained variance) pre- and postweaning weights. The GPFARM model has functional utility for simulating forage and cow–calf production with satisfactory accuracy at semiarid-temperate sites, such as southeastern Wyoming and northeastern Colorado. Continued development will focus on improving plant response to environmental stresses and testing the models functionality as a decision support tool for strategic and tactical ranch management.

GPFARM plant model parameters: Complications of varieties and the genotype x environment interaction in wheat

The USDA–ARS Great Plains Framework for Agricultural Resource Management (GPFARM) decision support system was developed to assist Great Plains producers in making economically viable and environmentally sound strategic plans for whole farm and ranch systems. A major user requirement for GPFARM is to supply the default plant parameters required to simulate crop growth. Developing this plant parameter database is difficult because varietal differences, caused by a genotype by environment (G . E) interaction, increases parameter uncertainty and variability. This article examines species–based plant parameter sets for simulating winter wheat (Triticum aestivum L.) yield responses, explores the significance of the G . E interaction on simulating varietal grain yield, and investigates whether simple adjustments to a species–based plant parameter database can improve simulation of varietal differences across environments. Three plant parameter sets were evaluated against observed yield data for six locations in eastern Colorado: (1) the Default parameter set used best estimates from EPIC–based plant parameter databases, (2) the Dryland Agroecosystems Project (DAP) parameter set further calibrated the default plant parameters against observed yield data for Colorado, and (3) the Theory parameter set modified DAP parameters based on whether irrigated or dryland conditions were simulated. The Theory parameter set simulated yield the best when pooling varieties over environments and locations. However, no parameter set could simulate all the different varietal yield responses to environmental conditions (irrigated or dryland) due to the diverse G . E interactions. The Theory parameter set best simulated the wheat variety TAM 107 across diverse locations, with little bias for either irrigated or dryland conditions. Simple adjustments to a few plant parameters based on whether dryland or irrigated conditions were simulated improved the species–based plant parameter approach used in GPFARM. However, until a better mechanistic representation of the G . E interaction is incorporated into existing plant growth models, opportunities for improving yield response to environmental conditions and management will be limited.

A smartphone app to extend use of a cloud-based irrigation scheduling tool

An irrigation scheduling tool was created on a cloud-based server.Key stakeholder members requested a phone app to increase ease of use.WISE will benefit agricultural producers, irrigation managers, research scientists. Irrigation in Colorado, a headwaters state, is crucial for viable agricultural production; consequently, with the foreseen population growth, there will become a greater demand placed on precious water resources. Technology must be adopted and embraced as part of the solution to water shortage. Researchers at Colorado State University have created an online evapotranspiration-based irrigation scheduling tool called Water Irrigation Scheduling for Efficient Application (WISE) that uses the soil water balance method and data queries from Colorado Agricultural Meteorological Network (CoAgMet) and Northern Colorado Water Conservation District (NCWCD) weather stations. To expedite and mobilize required user interaction with the software interface, a smartphone app has been developed that allows users to quickly view their soil moisture deficit, weather measurements, and the ability to input applied irrigation amounts into WISE. Potential users: agricultural producers, irrigation managers, and research scientists, will benefit from this app as it allows lite access to the tool from any location within a cellular data network. Technology such as the scheduling tool and smartphone app, when adopted within Colorado and the western United States, allow irrigators another tool to better utilize water resources.

Modeling of Full and Limited Irrigation Scenarios for Corn in a Semiarid Environment

Population growth in urbanizing areas such as the Front Range of Colorado has led to increased pressure to transfer water from agriculture to municipalities. In some cases, farmers may remain agriculturally productive while practicing limited or deficit irrigation, where substantial yields may be obtained with reduced water applications during non-water-sensitive growth stages, and crop evapotranspiration (ET) savings could then be leased by municipalities or other entities as desired. Site-specific crop simulation models have the potential to accurately predict yield and ET trends resulting from differences in irrigation management. The objective of this study was to statistically determine the ability of the CERES-Maize model to accurately differentiate between full and limited irrigation treatments in northeastern Colorado in terms of evapotranspiration (ET), crop growth, yield, water use efficiency (WUE), and irrigation use efficiency (IUE). Field experiments with corn were performed near Fort Collins, Colorado, from 2006 to 2008, where four replicates each of full (100% of ET requirement for an entire season) and limited (100% of ET during reproductive stage only) irrigation treatments were evaluated. Observations of soil profile water content, leaf area index, leaf number, and grain yield were used to calibrate (2007) and evaluate (2006 and 2008) the model. Additionally, ET and water use efficiency (WUE) were calculated from a field water balance and compared to model estimates. Over the three years evaluated, CERES-Maize agreed with observed trends in anthesis date, seasonal cumulative ET (Nash-Sutcliffe efficiency ENS = 0.966 for full irrigation and 0.835 for limited irrigation), leaf number in 2007 (ENS = 0.949 for full irrigation and 0.900 for limited irrigation), leaf area index in 2008 (ENS = 0.896 for full irrigation and 0.666 for limited irrigation), and yield (relative error RE = 4.1% for full irrigation and -3.4% for limited irrigation). Simulation of late-season leaf area index in limited irrigation was underestimated, indicating model overestimation of water stress. Simulated cumulative ET trends were similar to observed values, although CERES-Maize showed some tendency to underpredict for full irrigation (RE = -7.2% over all years) and overpredict for limited irrigation (RE = 12.7% over all years). Limited irrigation observations showed a significant increase in WUE over full irrigation in two of the three years; however, the model was unable to replicate these results due to underestimation of ET differences between treatments. While CERES-Maize generally agreed with observed trends for full and limited irrigation scenarios, simulation results show that the model could benefit from a more robust water stress algorithm that can accurately reproduce plant responses such as those observed in this study.

Whole-System Integration and Modeling Essential to Agricultural Science and Technology for the 21st Century

Abstract In the 21st century, agricultural research has more difficult and complex problems to solve. The continued increase in population in the developing countries requires continued increases in agricultural production. However, the increased use of fertilizers, pesticides, and water required for the new higher yielding crop varieties has been causing environmental problems. Excessive leaching and runoff of agricultural chemicals are seriously affecting the quality of both the groundwater and surface waters. Increase in soil salinity, decline in soil organic matter, and increase in soil erosion remain the major problems in intensively farmed areas. Even the air quality is being affected. At the same time, market-based global competition is challenging the economic viability of traditional agricultural systems. Global climate change will pose additional challenges. The solution or mitigation of these changing and multiple problems will require continual improvement or changes in management and selection of dynamic cropping systems using a whole-system approach. Therefore, synthesis and quantification of disciplinary knowledge at the whole-system level is essential to meeting these challenges. The process-based models of agricultural systems provide such a synthesis and quantification for evaluating the effects of varying management practices, crops, soils, water, and climate on both the production and the environment. These system models will greatly enhance the efficiency of field research for developing sustainable agricultural systems, serve as guides for planning and management, and help transfer new technologies to various conditions of developing countries. Current state of the system models and their applications for these purposes are reviewed, and advancements needed in models to improve and extend these applications are presented.

Modification of a soybean model to improve soil temperature and emergence date prediction.

Recent studies have shown that the CROPGRO-Soybean model does not predict soil temperature very well in Iowa. This typically gives errors in predicted emergence date, which translates to errors in timing of development and biomass accumulation during the remainder of the season. In order to improve the model, an energy balance-based soil temperature model was integrated into the soybean model and compared to the original soil temperature model, which was driven primarily by air temperature. In the new model, temperature at the soil surface is estimated from the basic energy balance equation at the air-soil interface and the soil temperature profile is calculated using the one-dimensional heat flow equation. The model was calibrated using five years of bare-soil temperature data measured at an experimental farm in Ames, Iowa. Validation of the new model using five additional years of bare-soil temperature data from the same location gave slightly better predictions of soil temperature in the top 5 cm (RMSE = 3.0, R 2 = 0.86 for validation years), and responded better to surface perturbations than the original model (RMSE = 3.2, R 2 = 0.80 for validation years). Under bare soil conditions, the new model generally gave lower RMSE and higher R 2 values compared to the old model at all soil depths. The models were also compared for accuracy in predicting emergence date. Experimental data on soil temperature and emergence for soybeans planted on weekly intervals over an eight-week period were used to test the models. The new model gave excellent predictions of emergence, with an average error of 0.6 day for the eight weekly experiments. The old model had an average error of one day. Under cool conditions, the new model gave more accurate predictions of emergence dates. However, under warm periods, both models typically gave the same accuracy, and were within about one day of the measured emergence date.

EVALUATING GPFARM CROP GROWTH, SOIL WATER, AND SOIL NITROGEN COMPONENTS FOR COLORADO DRYLAND LOCATIONS

Alternative agricultural management systems in the semi-arid Great Plains are receiving increasing attention. GPFARM is a farm/ranch decision support system (DSS) designed to assist in strategic management planning for land units from the field to the whole-farm level. This study evaluated the regional applicability and efficacy of GPFARM based on simulation model performance for dry mass grain yield, total soil profile water content, crop residue, and total soil profile residual NO3-N across a range of dryland no-till experimental sites in eastern Colorado. Field data were collected from 1987 through 1999 from an on-going, long-term experiment at three locations in eastern Colorado along a gradient of low (Sterling), medium (Stratton), and high (Walsh) potential evapotranspiration. Simulated crop alternatives were winter wheat (Triticum aestivum L.), corn (Zea mays L.), sorghum (Sorghum bicolor L.), proso millet (Panicum miliaceum L.), and fallow. Relative error (RE) of simulated mean, root mean square error (RMSE), and index of agreement (d) model evaluation statistics were calculated to compare modeled results to measured data. A one-way, fixed-effect ANOVA was also performed to determine differences among experimental locations. GPFARM simulated versus observed REs ranged from -3% to 35% for crop yield, 6% to 8% for total soil profile water content, -4% to 32% for crop residue, and -7% to -25% for total soil profile residual NO3-N. For trend analysis (magnitudes and location differences), GPFARM simulations generally agreed with observed trends and showed that the model was able to simulate location differences for the majority of model output responses. GPFARM appears to be adequate for use in strategic planning of alternative cropping systems across eastern Colorado dryland locations; however, further improvements in the crop growth and environmental components of the simulation model (including improved parameterization) would improve its applicability for short-term tactical planning scenarios.

Optimization of irrigation scheduling using ant colony algorithms and an advanced cropping system model

Abstract A generic simulation-optimization framework for optimal irrigation and fertilizer scheduling is developed, where the problem is represented in the form of decision-tree graphs, ant colony optimization (ACO) is used as the optimization engine and a process-based crop growth model is applied to evaluate the objective function. Dynamic decision variable option (DDVO) adjustment is used in the framework to reduce the search space size during the generation of trial solutions. The framework is applied for corn production under various levels of water availability and rates of fertilizer application in eastern Colorado, USA. The results indicate that ACO-DDVO is able to identify irrigation and fertilizer schedules that result in better net returns while using less irrigation water and fertilizer than those obtained using the Microsoft Excel spreadsheet-based Colorado Irrigation Scheduler (CIS) tool for annual crops. Another advantage of ACO-DDVO compared to CIS is the identification of both optimal irrigation and fertilizer schedules.