Speaker
Description
To develop new cropping strategies (e.g. early sowing, double cropping) and explore suitable cultivation areas for grain legumes under current and future climates, an accurate prediction of crop phenology under different photo-thermal conditions is a pre-requisite. This means the simulation of the onset and length of the main phenophases between emergence and physiological maturity as a function of temperature and photoperiod.
For that purpose, a simple phenology algorithm (SPA) was developed and calibrated for soybean (Glycine max) in controlled conditions then satisfactorily evaluated at field level in SW France (Schoving et al., 2020) but with less accuracy at national level (Bourgeois et al., 2023).
Using the same approach, SPA was applied to three other spring-sown grain legumes: chickpea (Cicer arietinum), faba bean (Vicia faba), and field pea (Pisum sativum) with as an attempt to propose a generic method for phenology prediction in grain legumes. In 2022, SPA was calibrated for 2 cultivars of chickpea (Elixir, Twist), faba bean (Espresso, Victus) and pea (Karpate, Kayanne), and for 6 cultivars of soybean (from MG 000 to II).
Before being applied, SPA requires the calibration of 7 genotypic parameters determined from the literature, dedicated experiments or optimization. Two experiments were set up for this calibration: a) a pot experiment on the Heliaphen platform at INRAE Auzeville (lat. 43°31’ N, long. 1°28’ E) under natural conditions with 4 to 5 planting dates (from March to July) and a phenological monitoring thrice a week in order to unravel the effects of photoperiod from temperature accumulation; b) an experiment on germination response to temperature (from 3 to 43°C) under controlled conditions (cold room and incubator) in order to determine the three cardinal temperatures (Tmin, Topt, Tmax) for germination. A photothermal time formalism was used to evaluate by optimization the sensitivity of each cultivar to photoperiod, and the optimal Physiological Development Days of the cultivar to complete a given phenological phase (PDD). The optimization of the PDD plant parameter from hourly temperature values instead of daily values improved the prediction of phenology by SPA.
Data from the national post-registration network of Terres Inovia used for varietal recommendations were collected from to 2017 to 2021 for the corresponding varieties of grain legumes in order to evaluate the predictive quality of SPA.
Historical series of daily temperature were retrieved from the SAFRAN historical reanalysis which covers France at 8 × 8 km resolution. The SPA model was applied on each of the 8602 grid cells (height < 600 m) on the 2001-2021 period to produce maps of suitability at national level (fig).
Cardinal temperatures were determined for the 4 grain legumes species: Topt ranged from 24.4 °C (faba bean) to 28.0 °C (field pea) and Tmax from 33.6 °C (chickpea) to 36.8°C (field pea). Faba bean and chickpea were confirmed as long day plants and soybean as a short day plant especially for the late-maturing varieties. Surprisingly, pea behave as a short day plant until flowering. With the assumption that the simulated date of physiological maturity was close to the actual date of harvest, SPA resulted in a reasonable prediction of the length of the growing season (with average RRMSE of 13.9 % to 16.7 % as a function of the crops). Using SPA, maps of crop feasibility at France level could be proposed for each variety as illustrated (fig).
After some improvement (effect of more extreme daylengths, refined simulation of emergence and harvest dates…), this methodology combining cultivar phenotyping and simulation could be expanded to winter-sown grain legumes, later sowing dates (double crops) and future climates in order to promote the diversification of cropping systems with grain legumes.
Acknowledgements : This work was supported by the Ecodiv project (Carnot Plant2Pro)
References
Bourgeois D, Kang L, Duchalais A, Schoving C, Constantin J, Champolivier L, Maury P, Debaeke P,. 2023. Proc. 11th World Soybean Research Conference, 18‐23 June 2023, Vienna (Austria), p 307
Schoving C, Stöckle C, Colombet C, Champolivier L, Debaeke P, Maury P, 2020. Frontiers in Plant Science 10, 1755
| Keywords | Agroecology ; Grain legumes ; Phenotyping ; Photoperiod sensitivity ; Crop feasability |
|---|
