[Show abstract][Hide abstract] ABSTRACT: Ensembles of process-based crop models are increasingly used to simulate crop growth for scenarios of temperature and/or precipitation changes corresponding to different projections of atmospheric CO2 concentrations. This approach generates large datasets with thousands of simulated crop yield data. Such datasets potentially provide new information but it is difficult to summarize them in a useful way due to their structural complexities. An associated issue is that it is not straightforward to compare crops and to interpolate the results to alternative climate scenarios not initially included in the simulation protocols. Here we demonstrate that statistical models based on random-coefficient regressions are able to emulate ensembles of process-based crop models. An important advantage of the proposed statistical models is that they can interpolate between temperature levels and between CO2 concentration levels, and can thus be used to calculate temperature and [CO2] thresholds leading to yield loss or yield gain, without re-running the original complex crop models. Our approach is illustrated with three yield datasets simulated by 19 maize models, 26 wheat models, and 13 rice models. Several statistical models are fitted to these datasets, and are then used to analyze the variability of the yield response to [CO2] and temperature. Based on our results, we show that, for wheat, a [CO2] increase is likely to outweigh the negative effect of a temperature increase of +2°C in the considered sites. Compared to wheat, required levels of [CO2] increase are much higher for maize, and intermediate for rice. For all crops, uncertainties in simulating climate change impacts increase more with temperature than with elevated [CO2].
Full-text · Article · Dec 2015 · Agricultural and Forest Meteorology
[Show abstract][Hide abstract] ABSTRACT: Decreasing water availability for rice based systems resulted in the introduction of water saving production systems such as aerobic rice and alternate wetting–drying technology. To further improve resource use efficiency in these systems, water management should be attuned to critical growth stages, requiring accurate prediction of crop phenology. Photoperiod-sensitivity of aerobic rice genotypes complicates the estimation of the parameters characterising phenological development and hence impairs predictions. To overcome this complication, we followed a two-step approach: (1) the photoperiod response was determined in growth chambers, through a reciprocal transfer experiment with variable day length, conducted at a fixed temperature, and consecutively, (2) the temperature response was studied by combining the obtained photoperiod parameters with data from field experiments. All four aerobic rice genotypes tested exhibited strong photoperiod-sensitivity. Durations of basic vegetative phase (BVP) i.e. when plants are still insensitive to photoperiod, photoperiod-sensitive phase (PSP), and post-PSP (PPP) varied among genotypes. The temperature response of the genotypes was explored by combining phenological observations in the reciprocal transfer experiment with observations in two field experiments. The temperature range in the field experiments was too narrow to obtain convergence to a unique set of temperature response parameters, regardless whether a bilinear or a beta model was used. Sensitivity analysis however provided clear arguments in support of the recent doubts on the validity of a commonly used set of cardinal temperatures for rice phenology. Using standard cardinal temperatures, the rate of development at temperatures below 31 °C was overestimated. This finding stresses the need for experiments on rice phenology under a wider range of temperatures.
Full-text · Article · Jan 2014 · Field Crops Research
[Show abstract][Hide abstract] ABSTRACT: Given the need for parallel increases in food and energy production from crops in the context of global change, crop simulation models and data sets to feed these models with photosynthesis and respiration parameters are increasingly important. This study provides information on photosynthesis and respiration for three energy crops (sunflower, kenaf, and cynara), reviews relevant information for five other crops (wheat, barley, cotton, tobacco, and grape), and assesses how conserved photosynthesis parameters are among crops. Using large data sets and optimization techniques, the C(3) leaf photosynthesis model of Farquhar, von Caemmerer, and Berry (FvCB) and an empirical night respiration model for tested energy crops accounting for effects of temperature and leaf nitrogen were parameterized. Instead of the common approach of using information on net photosynthesis response to CO(2) at the stomatal cavity (A(n)-C(i)), the model was parameterized by analysing the photosynthesis response to incident light intensity (A(n)-I(inc)). Convincing evidence is provided that the maximum Rubisco carboxylation rate or the maximum electron transport rate was very similar whether derived from A(n)-C(i) or from A(n)-I(inc) data sets. Parameters characterizing Rubisco limitation, electron transport limitation, the degree to which light inhibits leaf respiration, night respiration, and the minimum leaf nitrogen required for photosynthesis were then determined. Model predictions were validated against independent sets. Only a few FvCB parameters were conserved among crop species, thus species-specific FvCB model parameters are needed for crop modelling. Therefore, information from readily available but underexplored A(n)-I(inc) data should be re-analysed, thereby expanding the potential of combining classical photosynthetic data and the biochemical model.
Full-text · Article · Jan 2012 · Journal of Experimental Botany
[Show abstract][Hide abstract] ABSTRACT: An ecophysiological model was designed to quantify variation in potato
(Solanum tuberosum L.) canopy dynamics. Model algorithms describe the build-
up phase, maximum cover phase, and decline phase based on five parameters: t m1
(transition from accelerating to diminishing growth during the build-up phase, t 1
(end of the build-up phase when canopy cover attains its maximum level v max ), t 2
(end of the phase of maximum cover), and t e (end of the crop cycle). Parameter
values were estimated for 5 cultivars and 100 individuals of an F1 population using
data collected in 6 field experiments. The model successfully described differences
in canopy dynamics among individual genotypes across environments. Model
parameters were used to derive several secondary variables: P 1 , P 2 and P 3 (length
of the three phases) and A sum (area under the canopy cover curve reflecting the
crop ’ s capacity to intercept incoming radiation). P 1 was relatively conservative,
but P 2 and P 3 varied greatly. There were negative correlations among P 1 , P 2 , and
P 3 , suggesting that genotypes with slow canopy build-up had a relatively short P 2 ,
but a relatively long P 3 . Furthermore, P 2 was short when v max was below 100%,
whereas it could be (much) longer when v max reached 100%. Later genotypes had
a higher A sum because they had longer P 2 and P 3 . Total biomass production
depends on the absorbed photosynthetically active radiation, which is proportional
to A sum . Potato models predicting yield could use our approach to improve
estimates of canopy light interception under diverse environmental conditions.
Our approach also allows identifying potato canopy characteristics relevant for
Preview · Article · Sep 2010 · European Potato Journal
[Show abstract][Hide abstract] ABSTRACT: We investigated the potential of a model-based approach to assist in the genetic analysis of the environment-sensitive, quantitative crop trait canopy cover in potato (Solanum tuberosum L.). We used a model based on beta functions to analyze the genotype×environment interactions related to the dynamics of canopy cover. The model equations describe three phases of canopy growth: build-up phase (P1),
maximum cover phase (P2), and decline phase (P3). The model has five parameters: t m1 indicates the transition from accelerating to diminishing growth during P1, t 1 marks the end of P1 when canopy cover attains its maximum level v max , t 2 marks the end of P2 when canopy cover starts to decline, and t e represents the end of the crop cycle when canopy cover has declined to nil. Values of these parameters were estimated for 100 individuals of an F1 population, their parents, and five standard cultivars differing in maturity type, using data collected in six field experiments. The model successfully described differences in canopy dynamics among individual genotypes across environments. Model parameters were used to
derive several secondary variables: DP2 (duration of P2), DP3 (duration of P3), and A sum (area under the canopy cover curve reflecting the crop ’ s capacity to intercept incoming radiation). The length of P1 (i.e. t1 ) was relatively conservative, but D P2, and DP3 varied greatly. Later genotypes had higher A sum because they had longer DP2, and DP3 . Genotypic and phenotypic variance components of the F1 population were
estimated for all traits across environments and almost all of them proved significant (P<0.01). For most traits, genetic variability and heritability were high. There are opportunities, therefore, for future potato breeding programmes to exploit the genetic variability available in the F1 segregating population and to select for highly heritable traits in order to improve radiation interception efficiency.
No preview · Article · Jun 2010 · European Potato Journal
[Show abstract][Hide abstract] ABSTRACT: Cynara (Cynara cardunculus) is a perennial C3 herb that has its potential as bioenergy crop. This paper aims (a) to derive empirical relationships to predict cynara seed yield per head and per unit area, avoiding laborious extraction of seeds from the complex structure of its inflorescences; (b) to determine the head-weight distribution per unit area, the seed composition and the oil profile of cynara seeds; and (c) to estimate the range of cynara biomass, seed and oil yield in representative parts of Greece. We analyzed 16 field experiments, varying in crop age and environmental conditions in Greece. Seed yield per head (SYhead) can be accurately predicted as a linear function of dry head weight (Hw): SYhead=0.429·Hw−2.9 (r2=0.96; n=617). Based on this relationship, we developed a simple two-parameter equation to predict seed yield per unit area (SY): SY=HN·(0.429·μ−2.9), where μ is the mean head weight (g head−1) and HN is the total number of heads per unit area, respectively. The models were tested against current and published data (n=180 for head-level; n=35 for unit area-level models), and proved to be valid under diverse management and environmental conditions. Attainable cynara seed yields ranged from 190 to 480 g m−2 yr−1, on dry soils and on aquic soils (shallow ground water level). This variation in seed yield was sufficiently explained by the analyses of head-weight distribution per unit area (small, medium and large heads) and variability of seed/head weight ratio at head level. Seed oil concentration (average: 23%) and crude protein concentration (average: 18.7%) were rather invariant across different seed sizes (range: 26–56 mg seed−1) and growing environments.
[Show abstract][Hide abstract] ABSTRACT: Leaf area index (LAI) is an important variable for modelling canopy photosynthesis and crop water use. In many crop simulation models, prediction of LAI is very sensitive to errors in the value of parameter “specific leaf area” (SLA), which often relies on destructive measurements to determine. In this study, we present a model for predicting LAI of greenhouse crops based on the quantification of easily measured morphological traits as affected by temperature and radiation. Our model predicts LAI based on canopy light interception as a function of node development rate along with specific leaf size and elongation rates characteristics defined on a leaf number basis. Growth studies with five greenhouse crops (cucumber, sweet pepper, chrysanthemum, tulip and lilium) were conducted in different greenhouses and different sites during 2003 to 2009. The model was evaluated, in comparison with two commonly used methods for predicting LAI – the growing degree days (GDD) based model and SLA based model, using independent data from other experiments. The coefficient of determination (r2) and the root mean squared error (RMSE) between the predicted and measured values using our photothermal method are 0.99 and 0.95 (r2, RMSE) for leaf number, 0.98 and 0.01 m for specific leaf length, and 0.98 and 0.13 m2 m−2 for canopy LAI. For the GDD-based model, the r2 and RMSE are 0.93 and 4.23, 0.82 and 0.04 m, 0.87 and 0.48 m2 m−2 for the three traits, respectively. For the SLA-based model, the r2 and RMSE for canopy LAI is 0.81 and 1.24 m2 m−2 when using the estimated SLA data as input or 0.94 and 0.25 m2 m−2 when using the measured SLA data as input. So, our model better predicts LAI for greenhouse crops at different latitudes and a range of planting densities and pruning systems. Although calibrations for specific light regime, pruning practices and cultivars are needed, the fact that production conditions in commercial greenhouse production are often well controlled and production practices are often rather standardized implies a general applicability of our model.
[Show abstract][Hide abstract] ABSTRACT: Intimate relationships exist between form and function of plants, determining many processes governing their growth and development. However, in most crop simulation models that have been created to simulate plant growth and, for example, predict biomass production, plant structure has been neglected. In this study, a detailed simulation model of growth and development of spring wheat (Triticum aestivum) is presented, which integrates degree of tillering and canopy architecture with organ-level light interception, photosynthesis, and dry-matter partitioning. An existing spatially explicit 3D architectural model of wheat development was extended with routines for organ-level microclimate, photosynthesis, assimilate distribution within the plant structure according to organ demands, and organ growth and development. Outgrowth of tiller buds was made dependent on the ratio between assimilate supply and demand of the plants. Organ-level photosynthesis, biomass production, and bud outgrowth were simulated satisfactorily. However, to improve crop simulation results more efforts are needed mechanistically to model other major plant physiological processes such as nitrogen uptake and distribution, tiller death, and leaf senescence. Nevertheless, the work presented here is a significant step forwards towards a mechanistic functional-structural plant model, which integrates plant architecture with key plant processes.
Full-text · Article · Mar 2010 · Journal of Experimental Botany
[Show abstract][Hide abstract] ABSTRACT: This paper aims (a) to derive empirical relationships to predict cynara seed yield per head and per unit area, avoiding laborious extraction of seeds from its complex inflorescence structure; (b) to determine the headweight distribution per unit area; and (c) to estimate the range of Cynara cardunculus seed yield. We analyzed 16 field experiments, varying in crop age and environmental conditions in Greece. Seed yield per head (SYhead) can be accurately predicted as a linear function of dry head weight (Hw): SYhead = 0.429 · Hw– 2.9 (r 2=0.96; n=617). Based on this relationship, we developed a simple two-parameter equation to predict seed yield per unit area (SY): SY = HN · (0.429 · µ – 2.9), where HN is the total number of heads per unit area and µ is the mean head weight (g head-1), respectively. The models were tested against current and published data (n=180 for head-level; n=35 for unit arealevel model), and proved to be valid under diverse environmental and management conditions. Attainable cynara seed yields ranged from 190 to 480 g m–2 y–1, on dry soils and on aquic soils (shallow ground-water level). This variation in seed yield was sufficiently explained by the analyses of head-weight distribution per unit area and variability of seed/head weight ratio at head level. This work provides basal information of crop reproductive effort and reveals the great potential of cynara as a biomass and oil crop for bioenergy production.
[Show abstract][Hide abstract] ABSTRACT: Nearly three decades ago Farquhar, von Caemmerer and Berry published a biochemical model for C3 photosynthetic rates (the FvCB model). The model predicts net photosynthesis (A) as the minimum of the Rubisco-limited rate of CO2 assimilation (Ac) and the electron transport-limited rate of CO2 assimilation (Aj). Given its simplicity and the growing availability of the required enzyme kinetic constants, the FvCB model has been used for a wide range of studies, from analysing underlying C3 leaf biochemistry to predicting photosynthetic fluxes of ecosystems in response to global warming. However, surprisingly, this model has seen limited use in existing crop growth models. Here we highlight the elegance, simplicity, and robustness of this model. In the light of some uncertainties with photosynthetic electron transport pathways, a recently extended FvCB model to calculate Aj is summarized.
Preview · Article · Dec 2009 · NJAS: wageningen journal of life sciences
[Show abstract][Hide abstract] ABSTRACT: Genes contributing to the quantitative variation of a complex crop trait can be numerous. However, using existing approaches, the number of quantitative trait loci (QTL) detected for a trait is limited. Therefore, rather than looking for QTL for a complex trait itself, determining QTL for underlying component traits might give more information. In this study the potential of component analysis in QTL mapping of complex traits was examined using grain yield in spring barley as an example. Grain yield was divided into three components: number of spikes/m2, number of kernels/spike, and 1000-kernel weight. These traits were measured for individuals of a recombinant inbred-line population in field trials conducted over 2 years. By the use of an approximate multiple QTL model, one to eight QTL were detected for each trait in a year. Some QTL were mapped to similar positions in both years. Almost all QTL for yield were found at the position of or in close proximity to QTL for its component traits. A number of QTL for component traits were not detected when yield itself was subjected to QTL analysis. However, relative to the QTL for yield itself, all component-trait QTL did not explain the variation in yield better. The results in relation to the potential of using component analysis in studying complex crop traits are discussed.