Evaluating the ability of four crop models to predict different environmental impacts on spring wheat grown in open-top chambers
Highlights
► Crop models were tested against complete data sets with different CO2 and H2O regimes. ► The crop models adequately simulated the open-top chamber experimental conditions. ► Dynamic interplay was simulated best using the Farquhar photosynthesis sub-model. ► Simple photosynthesis models increased specific assimilation rates under elevated CO2.
Introduction
Plants assimilate carbon from atmospheric CO2 via photosynthesis. From a pre-industrial level of approximately 280 ppm (Siegenthaler et al., 2005), recent CO2 concentrations now amount to 390 ppm (Tans, 2009) and are predicted to yield 450 ppm in the year 2030 (Alcamo et al., 2007, OECD EO, 2008). Because CO2 is not substrate saturated at current atmospheric CO2 concentrations, CO2 enrichment commonly boosts crop yields of C3 cereals, such as wheat, and thus increases food and feed production. Throughout the last two decades, this view was supported by various experiments using climate chamber, open-top chamber (OTC) and free-air-CO2-enrichment (FACE) technology (Amthor, 2001, Fangmeier et al., 1999, Högy et al., 2009, Wullschleger et al., 1992). Although most studies have demonstrated positive effects on the aboveground biomass production of wheat under CO2 enrichment (Ewert et al., 2002, Fangmeier et al., 2000, Poorter et al., 1996), several studies have suggested a reduction in the crop biomass under elevated CO2 due to interactions with other environmental factors (Long et al., 2005, Long et al., 2006). Long et al. (2006) and Schimmel (2006) summarized that CO2 fertilization effects on plant production have been overestimated as they account for up to 13% of wheat grown in FACE experiments in contrast to the 31–36% observed in chamber-based studies. Nevertheless, Ziska and Bunce (2007) have shown that differences of the CO2 responses on plant production between several experimental systems were less significant if the data were normalized to the different levels of CO2 enrichment. Moreover, elevated CO2 inhibits the nitrate assimilation from soil (Bloom et al., 2010) that may be largely responsible for CO2 acclimation and, in turn, decreases in grain protein quality (Högy and Fangmeier, 2008). Furthermore, CO2 fertilization effects on plants using the C3 metabolism are only expected if the environment is not limited by temperature or water supply. Extreme weather events, such as the prolonged periods of limited water supply predicted by climate scenarios for Europe (McGregor et al., 2005, Parizek et al., 2004), could interact with CO2-induced impacts on crop growth in the future.
Throughout the last decades, various crop models have been developed that describe physiological processes and the cycling of water and nutrients in terrestrial agro-ecosystems. Model types can be characterized as static or dynamic, deterministic or stochastic and empirical or mechanistic (Dent and Blackie, 1979, Thornley and France, 2007). To date, model development for complex system simulations aims to be dynamic rather than static, deterministic rather than stochastic and mechanistic rather than empirical (Wang, 1997). Empirical models are direct descriptions of measurements, and they define the characteristics of a system in a simple way. At some points it is inevitable to use empirical sub-models, including when the physical processes are not well understood. In crop science, the typical use of empirically determined development stages and the thermal time leads to a principal empirical model approach. The advantages of empirical models include the little effort needed for calibration and their robustness. A mechanistic model would encompass the necessary physics, chemistry and physiology to describe the crop growth processes from seed initiation to senescence. Mechanistic models offer more options to improve the system and to understand processes and their interactions. They are therefore commonly favored for modeling climate change (Thornley and France, 2007, van Wijk et al., 2002). In practice, simulation models are mechanistic to varying degrees, with inevitable empiricisms built into the sub-models.
Cross comparison between models is a well-established tool for model evaluation (Ewert et al., 2002, Jamieson et al., 1998, Jamieson et al., 2000, Porter et al., 1993, Priesack and Gayler, 2006, Rastetter, 1996, van Wijk et al., 2002, Yin and Struik, 2010). Cross comparison shows the quality of model components and indicates which components need more intense scientific attention. Moreover, these comparisons assist the model user to evaluate modeling outputs, particularly when the models are used for regionalization or for climate change simulations.
The four crop models in this study were chosen because of the different degree to which they include mechanistic approaches to model genotype-by-environment interactions and because of the different approaches to simulate effects of CO2 on crop growth. The selected models are CERES-Wheat 2.0 (Ritchie and Godwin, 1987, Ritchie et al., 1987), SUCROS2 (Goudriaan and van Laar, 1994, Groot, 1987, Spitters et al., 1989, van Keulen et al., 1992, van Keulen and van Laar, 1982), SPASS (Wang, 1997, Wang and Engel, 2000) and GECROS (Yin and van Laar, 2005). For simplicity, the four models are further abbreviated with CERES, SUCROS, SPASS and GECROS. In SUCROS and CERES, the CO2 response is simply controlled by an empirical increase of the light use efficiency (LUE). The SPASS model assumes a constant initial slope where photosynthesis is entirely CO2 limited with a switch to a horizontal maximum photosynthesis rate, and the GECROS model applies the non-rectangular hyperbolic response to CO2 concentrations of the Farquhar model (Farquhar et al., 1980). The objective of this study was to test the four crop growth models included in the Expert-N modeling package in terms of their ability to simulate aboveground biomass production, grain yield and yield quality of spring wheat under various environmental conditions. For comparison, only the plant models were exchanged, and the models of water flow, nitrogen transport and heat transfer were the same for all four crop models. We therefore analyzed
- (i)
if the impacts of atmospheric CO2 and water shortage on biomass production can be adequately simulated by each of the four different crop growth models and
- (ii)
if a more mechanistic modeling approach for the CO2-induced responses on crop growth is an improvement compared to the established models that include empirical assumptions.
Section snippets
The Expert-N model package
The model package Expert-N was developed to provide models for the simulation of soil-plant-atmosphere systems. The modular design helps to combine simulation models from the available components that were implemented in the Expert-N package. These components include sub-models to simulate soil water flow, soil heat transfer, turnover and transport of soil carbon and nitrogen, soil management and crop growth (Priesack, 2006, Priesack and Bauer, 2003). The output of different crop growth
Simulation results
Fig. 1 presents the simulation results of all four models together with measured data of the reference treatment. Table 3 presents the results of the statistical evaluation of those output variables that were measured repeatedly during the vegetation period for all treatments. Table 4 lists the simulation results and corresponding measurements of model outputs, for which only one measurement at the end of the vegetation exists, and it depicts the values of for these variables.
Discussion
The four crop models were evaluated focusing on three aspects: (A) the overall model performance, (B) the adequate description of the basic processes and (C) the applicability to the different environmental conditions. Furthermore, we analyzed the potential of each of the models for its use to simulate crop growth under climate change conditions at the regional scale.
Conclusions
Four crop models were calibrated to an experimental OTC dataset of different environmental conditions. The performance of the models in describing the impact of the considered effects of water availability and CO2 concentration followed the order SUCROS > SPASS > GECROS > CERES. Hence, the more mechanistic models, GECROS and SPASS, do not generally lead to better model performance when describing the reaction of the plant to environmental conditions.
Nevertheless, this manuscript described the
Acknowledgments
The study was funded by the German Research Foundation (DFG) as part of the Joint Research Project ‘Regional Climate Change’ (PAK 346) at the Universität Hohenheim and the Helmholtz Zentrum München. We thank Dr. Hagen Scherb for his advice on statistical analysis. We thank two anonymous reviewers for their constructive criticisms that significantly helped to improve the manuscript.
References (91)
Effects of atmospheric CO2 concentration on wheat yield: review of results from experiments using various approaches to control CO2 concentration
Field Crop. Res.
(2001)- et al.
A crop moisture stress index for large areas and its application in the prediction of spring wheat phenology
Agric. Meteorol.
(1982) - et al.
Effects of elevated CO2 and drought on wheat: testing crop simulation models for different experimental and climatic conditions
Agr. Ecosyst. Environ.
(2002) - et al.
CO2 enrichment enhances flag leaf senescence in barley due to greater grain nitrogen sink capacity
Environ. Exp. Bot.
(2000) - et al.
Effects of nutrients on grain quality in spring wheat crops grown under elevated CO2 concentrations and stress conditions in the European multiple-site experiment ‘ESPACE-wheat’
Eur. J. Agron.
(1999) The importance of grain or kernel number in wheat: a reply to Sinclair and Jamieson
Field Crop. Res.
(2008)- et al.
Modelling biomass growth, N-uptake and phenological development of potato crop
Geoderma
(2002) - et al.
Effects of elevated atmospheric CO2 on grain quality of wheat
J. Cereal Sci.
(2008) - et al.
Modelling CO2 effects on wheat with varying nitrogen supplies
Agric. Ecosyst. Environ.
(2000) - et al.
A comparison of the models AFRCWHEAT2, CERES-wheat, Sirius, SUCROS2 and SWHEAT with measurements from wheat grown under drought
Field Crop. Res.
(1998)
Simulated nitrogen dynamics and losses in a layered agricultural soil
Agric. Ecosyst. Environ.
The DSSAT cropping system model*1
Eur. J. Agron.
Reliability and input-data induced uncertainty of the EPIC model to estimate climate change impact on sorghum yields in the U.S. Great Plains
Agric. Ecosyst. Environ.
Implications of increased Greenland surface melt under global-warming scenarios: ice-sheet simulations
Quaternary Sci. Rev.
Free-air CO2 enrichment (FACE): blower effects on wheat canopy microclimate and plant development
Agric. Forest. Meterol.
A comparison of the wheat crop simulation models AFRCWHEAT2. CERES wheat and SWHEAT for non-limiting conditions of growth
Field Crop. Res.
Nitrogen losses from entire plants of spring wheat (Triticum aestivum) from tillering to maturation
Eur. J. Agron.
Testing the root growth subroutine of the CERES-wheat model for two cultivars of different cycle length
Field Crop. Res.
Growth and yield responses of spring wheat (Triticum aestivum L. cv Minaret) to elevated CO2 and water limitation
Environ. Pollut.
Grain number, wheat yield, and bottling beer: an analysis
Field Crop. Res.
Simulating the effects of elevated CO2 on crops: approaches and applications for climate change
Eur. J. Agron.
Growing wheat in biosphere 2 under elevated CO2: observations and modelling
Ecol. Eng.
The effects of climatic variation in Europe on the yield response of spring wheat cv ‘Minaret’ to elevated CO2 and O3: an analysis of open-top chamber experiments by means of two crop growth simulation models
Eur. J. Agron.
Comparison of different modelling strategies for simulating gas exchange of a Douglas-fir forest
Ecol. Model.
SPASS: a generic process-oriented crop model with versatile windows interfaces
Environ. Model. Software
Crop growth, soil water and nitrogen balance simulation on three experimental field plots using the Opus model – a case study
Ecol. Model.
Nitrogen uptake in relation to water availability in wheat
Plant Soil.
Growth in elevated CO2 enhances temperature response of photosynthesis in wheat
Physiol. Plant.
Models of growth and development of wheat in relation to plant nitrogen
Aust. J. Agric. Res.
Effects of source-sink relations on photosynthetic acclimation to elevated CO2
Plant. Cell Environ.
Application of the CERES-wheat model for within-season prediction of winter wheat yield in the United Kingdom
Agron. J.
Carbon dioxide enrichment inhibits nitrate assimilation in wheat and Arabidopsis
Science
Crop responses to drought and the interpretation of adaptation
J. Plant Growth Regul.
Growth of individual leaves of spring wheat (Triticum aestivum L.) as influenced by temperature and light intensity
Ann. Bot. Lond.
Acclimation response of spring wheat in a free-air CO2 enrichment (FACE) atmosphere with variable soil nitrogen regimes 3. Canopy architecture and gas exchange
Photosynth. Res.
Systems Simulation in Agriculture
Nutzpflanzenkunde
Spezieller Pflanzenbau, 3, neubearbeitete und ergänzte Edition
Advantages of an open-top chamber plant exposure system to assess the impact of atmospheric trace gases on vegetation
Vereinigung für Angewandte Botanik
A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species
Planta
A User's Guide to CERES-Wheat – V2. 10
Modelling Potential Crop Growth Processes
Simulation of Nitrogen Balance in a System of Winter Wheat and Soil. Simulation Report CARBO-TT No. 13
The collection of a minimum dataset and the application of DSSAT for optimizing wheat yield in irrigated cropping systems
Optimizing Nitrogen Fertilizer Application to Irrigated Wheat. IAEA TECDOC-1164
Effects of barley yellow dwarf virus on root and shoot growth of winter wheat seedlings grown in aeroponic culture
Plant Dis.
Cited by (61)
Long-term soil organic carbon and crop yield feedbacks differ between 16 soil-crop models in sub-Saharan Africa
2024, European Journal of AgronomyBayesian multi-level calibration of a process-based maize phenology model
2022, Ecological ModellingModeling the response of winter wheat phenology to low temperature stress at elongation and booting stages
2021, Agricultural and Forest MeteorologyCitation Excerpt :To test whether the prediction error of crop phenological development under low temperature stress can be reduced by replacing different temperature response functions, the performances of six types of temperature response functions were evaluated by incorporating them into the WheatGrow phenology submodel in order to quantify the effects of low temperature stress at elongation and booting stages on elongation-anthesis duration of wheat. The temperature response functions and the cardinal temperature parameters used in these functions are from HERMES (Kersebaum, 2007, 2011), DSSAT-CERES (Hoogenboom and White, 2003; Jones et al., 2003; Ritchie et al., 1985), Expert-N-CERES (Biernath et al., 2011; Priesack et al., 2006), DSSAT-Nwheat (Asseng et al., 2000; Kassie et al., 2016), Apsim-E (Wang and Engel, 1998; Wang et al., 2017), and WheatGrow model (Cao and Moss, 1997; Liu et al., 2018; Zhu et al., 2013), respectively. The temperatures used in temperature response functions of HERMES, DSSAT-CERES, Expert-N-CERES, and DSSAT-Nwheat model are the average daily temperature.
Multi-model evaluation of phenology prediction for wheat in Australia
2021, Agricultural and Forest MeteorologyHow well do crop modeling groups predict wheat phenology, given calibration data from the target population?
2021, European Journal of Agronomy