Leslie A. Hunt

Multimodel ensembles of wheat growth: many models are better than one.

Crop models of crop growth are increasingly used to quantify the impact of global changes due to climate or crop management. Therefore, accuracy of simulation results is a major concern. Studies with ensembles of crop models can give valuable information about model accuracy and uncertainty, but such studies are difficult to organize and have only recently begun. We report on the largest ensemble study to date, of 27 wheat models tested in four contrasting locations for their accuracy in simulating multiple crop growth and yield variables. The relative error averaged over models was 24-38% for the different end-of-season variables including grain yield (GY) and grain protein concentration (GPC). There was little relation between error of a model for GY or GPC and error for in-season variables. Thus, most models did not arrive at accurate simulations of GY and GPC by accurately simulating preceding growth dynamics. Ensemble simulations, taking either the mean (e-mean) or median (e-median) of simulated values, gave better estimates than any individual model when all variables were considered. Compared to individual models, e-median ranked first in simulating measured GY and third in GPC. The error of e-mean and e-median declined with an increasing number of ensemble members, with little decrease beyond 10 models. We conclude that multimodel ensembles can be used to create new estimators with improved accuracy and consistency in simulating growth dynamics. We argue that these results are applicable to other crop species, and hypothesize that they apply more generally to ecological system models.

GUMCAS: a model describing the growth of cassava (Manihot esculenta L. Crantz)

Abstract A process-oriented model, GUMCAS, describing the growth of cassava ( Manihot esculenta L. Crantz) was developed for inclusion in the IBSNAT decision support system. Potential dry matter production is calculated from existing leaf area, and is modified by effects of light, temperature, water stress, and vapour pressure deficit. Leaf and stem growth are assumed to be the dominant sinks for assimilate, with fibrous roots receiving a fraction of that allocated to the shoot, decreasing as the crop ages. The storage roots receive any remaining assimilate. The Ritchie water balanced model is used to estimate water status. Leaf size is calculated empirically as a function of time. However, there was a strong correlation between leaf size and the assimilate supply/demand ratio represented by δ W/W (daily change in weight/total plant weight). This approach is included in the model as an option for simulating environmental influences on leaf size. This relationship functioned well, particularly on release of drought stress. Phenology is described by assuming two independent “clocks” controlling vegetative and reproductive development respectively. Both clocks are influenced by temperature and water status, while reproductive development is also influenced by photoperiod (φ). Fitting the standard model for photoperiod response to observed reproductive branching data gave a minimum optimum photoperiod ( φ 0 ) of 15.5 h and a sensitivity ( S φ ) of 0.25 h −1 , assuming the crop was sensitive to photoperiod from emergence. However, an improved fit was obtained by also assuming branching is inhibited when the rate of change of photoperiod (d φ /d t ) was above 0.01 h day −1 . Sensitivity analysis confirmed previous reports that leaf longevity is an important character determining storage root yield; other characters are the age at which first branching occurs, and specific leaf area. However, the importance of some characters changes under drought, such as the date at which the maximum leaf size occurs. The model was validated with the limited number of datasets available; good agreement between simulated and measured values was obtained for a range of cultivars in a range of environments. Further validation is necessary, however, before widespread application. The factors controlling leaf size and times of branching were identified as areas in which there is a scarcity of knowledge, and to which future cassava physiology research should be directed.

Simulating the influence of vernalization, photoperiod and optimum temperature on wheat developmental rates

BACKGROUND AND AIMS Accurately representing development is essential for applying crop simulations to investigate the effects of climate, genotypes or crop management. Development in wheat (Triticum aestivum, T. durum) is primarily driven by temperature, but affected by vernalization and photoperiod, and is often simulated by reducing thermal-time accumulation using vernalization or photoperiod factors or limiting accumulation when a lower optimum temperature (T(optl)) is exceeded. In this study T(optl) and methods for representing effects of vernalization and photoperiod on anthesis were examined using a range of planting dates and genotypes. METHODS An examination was made of T(optl) values of 15, 20, 25 and 50 degrees C, and either the most limiting or the multiplicative value of the vernalization and photoperiod development rate factors for simulating anthesis. Field data were from replicated trials at Ludhiana, Punjab, India with July through to December planting dates and seven cultivars varying in vernalization response. KEY RESULTS Simulations of anthesis were similar for T(optl) values of 20, 25 and 50 degrees C, but a T(optl) of 15 degrees C resulted in a consistent bias towards predicting anthesis late for early planting dates. Results for T(optl) above 15 degrees C may have occurred because mean temperatures rarely exceeded 20 degrees C before anthesis for many planting dates. For cultivars having a strong vernalization response, anthesis was more accurately simulated when vernalization and photoperiod factors were multiplied rather than using the most limiting of the two factors. CONCLUSIONS Setting T(optl) to a high value (30 degrees C) and multiplying the vernalization and photoperiod factors resulted in accurately simulating anthesis for a wide range of planting dates and genotypes. However, for environments where average temperatures exceed 20 degrees C for much of the pre-anthesis period, a lower T(optl) (23 degrees C) might be appropriate. These results highlight the value of testing a model over a wide range of environments.

The uncertainty of crop yield projections is reduced by improved temperature response functions

Increasing the accuracy of crop productivity estimates is a key element in planning adaptation strategies to ensure global food security under climate change. Process-based crop models are effective means to project climate impact on crop yield, but have large uncertainty in yield simulations. Here, we show that variations in the mathematical functions currently used to simulate temperature responses of physiological processes in 29 wheat models account for >50% of uncertainty in simulated grain yields for mean growing season temperatures from 14 °C to 33 °C. We derived a set of new temperature response functions that when substituted in four wheat models reduced the error in grain yield simulations across seven global sites with different temperature regimes by 19% to 50% (42% average). We anticipate the improved temperature responses to be a key step to improve modelling of crops under rising temperature and climate change, leading to higher skill of crop yield projections.

The effect of daylength on shoot growth and formation of root tubers in young plants of cassava (Manihot esculenta grantz)

Abstract Plants of cassava ( Manihot esculenta Crantz) “Llanera” were grown from stem cuttings for 16 weeks under three daylength treatments of 8 h (short days), 14 h, and 20 h using an 8 h basic period of high intensity light in all treatments, extended with weak incandescent light to give long day treatments. Results of a harvest at 8 weeks indicated that the initiation of root tubers was earlier in the 8 h daylength than the 14 h or 20 h, but there was no difference in the number of root tubers in the 16 week harvest. However, the weight of root tubers was 75 g in the 8 h daylength, compared with 30 and 35 g in the 14 h and 20 h daylengths, and this was accompanied by a reduction in stem dry weight to 18 g, compared with 47 and 41 g in the long day treatments. It is concluded that a long photoperiod promotes shoot growth and reduces root tuber development in young cassava plants, without influencing total dry weight.

Author Correction: The uncertainty of crop yield projections is reduced by improved temperature response functions

Nature Plants3, 17102 (2017); published online 17 July 2017; corrected online 27 September 2017.