Understanding effects of genotype × environment × sowing window interactions for durum wheat in the Mediterranean basin
Introduction
Durum wheat (Triticum turgidum L. subsp. durum) is one of the most important crops in the Mediterranean basin, which contributes to more than 38 % to the global durum wheat production (IGCC, 2017). In general, grain yield and quality are strongly influenced by the variability of weather conditions during the season (Porter and Semenov, 2005). Moreover, wheat is mostly sensitive to late frosts as well as to high temperatures and water deficit during the reproductive and grain filling phases (Porter and Gawith, 1999; Farooq et al., 2011; Alghabari et al., 2014). Terminal droughts could render the crop more sensitive to heat stress, as well as heat stress could be exacerbated by water stress, resulting in even more negative effects on the grain yield (Asseng et al., 2011). Such environmental conditions, typical of the Mediterranean climate, are the main climatic constraints for durum wheat yield in the Mediterranean basin (Ferrise et al., 2015).
Appropriate management practices are essential to ensure high wheat production. The sowing time and the cultivars choice are among the key factors (Connor et al., 1992; Bassu et al., 2009), since they act on the duration and timing of developmental phases and have the potential to avoid adverse environmental conditions in wheat sensitive stages (Ferrise et al., 2010; Bassu et al., 2009; Tapley et al., 2013). It is proved that the combination of optimal sowing date with early or late cultivars allows maximizing the yield and avoid stresses due to frost, heat and drought stresses around the anthesis and during grain filling (Ababaei and Chenu, 2020; Zheng et al., 2012), while favouring precipitation, temperature and radiation accumulation in the growing season (Tapley et al., 2013). Long-season cultivars may benefit from early sowing, thanks to an increased intercepted radiation during winter, thereby resulting in a higher accumulation of dry matter by the crop, while escaping terminal droughts (Zheng et al., 2012). On the other hand, this practice is not appropriate with early cultivars, when frost risk is generally high during winter or early spring (Andrarzian et al., 2015). Late sowing is usually recommended in locations with high frost risk (Connor et al., 1992), because it helps the crop to escape from frost conditions during the sensitive phases of crop growth. However, late sowing may induce a shortening of crop cycle (Sharma et al., 2008) and expose the crop to warmer and drier conditions during anthesis and grain filling (Panozzo and Eagles, 1999; Subedi et al., 2007). In the Mediterranean basin, the sowing window of durum wheat starts with the first significant rainfall after the summer season and ends when a sowing date is too late to allow the completion of the crop cycle. In the Mediterranean basin, the optimal combination of sowing window and cultivar should result in avoiding (or minimizing the impact of) frost damage during early springs as well as heat and drought conditions during anthesis and grain filling (Bassu et al., 2009; Chen et al., 2020).
It is widely recognised that wheat yield and grain quality are affected by genotype (G), environment (E), sowing window (SW) and their interactions (Sharma et al., 2008; Tapley et al., 2013; Haq et al., 2017). The complexity to test G × E × SW interactions in field experiments could be reduced by testing alternative scenarios with crop simulation models (Stapper and Harris, 1989; Chenu et al., 2017), which are able to reproduce crop growth and development, considering the interactions between soil, weather and crop management. Crop models have been largely used to extrapolate agronomic research findings over time and space (Chenu et al., 2017), to assess crops performance in response to climatic conditions (Salado-Navarro and Sinclair, 2009; Soltani et al., 2013; Dettori et al., 2017) and to identify the best management practices in a given environment (Soltani and Hoogenboom, 2007; Rozbicki et al., 2015). Several crop modelling studies have been carried out to investigate the effect of shifting the sowing date in future climatic scenarios (Moriondo et al., 2010; Dettori et al., 2017; Nouri et al., 2017) or to analyze the combination of adaptation strategies, including the shifting of the sowing date (Ruiz-Ramos et al., 2018; Giuliani et al., 2019) or to define the optimal flowering period in Australian environments (Flohr et al., 2017; Chen et al., 2020) but only few studies focused on the optimization of the sowing window in the Mediterranean basin under present climatic conditions (Bassu et al., 2009).
The objective of this study was to investigate the effects of genotype × environment × sowing window in four areas of the Mediterranean basin for current climate, by using the wheat simulation model SiriusQuality. First, the model was calibrated and evaluated in the targeted environments; then, a scenario analysis was carried out to identify the optimal sowing windows for early and late cultivars in order to quantify the impact of G × E × SW interactions on wheat growth and development.
Section snippets
Experimental sites
Data were collected from field experiments carried out in seven Mediterranean locations where durum wheat (Triticum turgidum L. subsp. durum) is widely grown (Fig. 1). The experimental sites were located in Florence (43.76 °N, 11.21 °E, 42 m elevation) and Foggia (41.26 °N, 15.30 °E, 90 m elevation), in central and southern Italy, respectively, in Carmona (37.47 °N, 5.63 °W, 253 m elevation) and Santaella (37.57 °N, 4.85 °W, 238 m elevation), in southern Spain, in Marchouch (33.98 °N, 6.49 °W,
Model evaluation
For both the calibration and validation data sets, MAE for anthesis or heading date and maturity date were < 9.3 days (Table 2, Fig. 3C-E), and nRMSE for grain yield and total above ground biomass ranged from 6 to 18 % and from 10 to 32 %, respectively (Table 2, Fig. 3A-d-F-G-H). The overall nRMSE for grain yield was only 5 % higher in validation than in calibration data set. All d values for phenology and grain yield except heading date in Carmona (that is, the Spanish site) were > 0.70. On
SiriusQuality evaluation
The value of seven varietal parameters describing the four wheat cultivars used in this study were within the literature range for Sirius (Semenov et al., 2014) and SiriusQuality (Tao et al., 2017). Moreover, they were comparable to previous modelling studies with Karim, Creso and Simeto (Bassu et al., 2011; Dettori et al., 2017; Bregaglio et al., 2015). No published parametrization for Amilcar cultivar was found.
SiriusQuality provided a good estimation of phenology, grain yield, grain N yield
Conclusions
In this study, a crop model was used to analyze the effect of environment, genotype and sowing window on wheat yield in the Mediterranean basin. The results indicated that different strategies, depending on the location, could be used to increase wheat yield. In hotter and drier environments, where temperature and water are the main limiting factors, it is crucial that grain filling occurs earlier in the season to guarantee a longer duration of this phase, due to lower temperature and adequate
CRediT authorship contribution statement
Gloria Padovan: Conceptualization, Methodology, Writing - original draft, Writing - review & editing. Pierre Martre: Conceptualization, Methodology, Writing - original draft, Writing - review & editing. Mikhail A. Semenov: Conceptualization, Methodology, Resources, Writing - original draft, Writing - review & editing. Alberto Masoni: Formal analysis, Writing - original draft, Writing - review & editing. Simone Bregaglio: Resources, Writing - review & editing. Domenico Ventrella: Resources,
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The authors acknowledge financial support from the following sources: FACCE JPI MACSUR2 through the Italian Ministry for Agricultural, Food and Forestry Policies (D.M. n. 24064/7303/15); the European Union’s Horizon 2020 research and innovation program under the Grant Agreement No. [727247]; the Biotechnology and Biological Sciences Research Council, through the Designing Future Wheat (DFW) program (BB/P016855/1); the AgriDigit-Agromodelli project (DM n. 36502 of 20/12/2018), funded by the
References (81)
- et al.
Using pattern recognition for estimating cultivar coefficient of a crop simulation model
Field Crop Res.
(2009) - et al.
Optimising sowing date of durum wheat in a variable Mediterranean environment
Field Crop. Res.
(2009) - et al.
Consequences of breeding on biomass, radiation interception and radiation-use efficiency in wheat
Field Crop Res.
(1997) - et al.
Spatial patterns of estimated optimal flowering period of wheat across the southwest of Western Australia
Field Crop Res
(2020) - et al.
Contribution of crop models to adaptation in wheat
Trends Plant Sci.
(2017) - et al.
Simulation of climate change impacts on production and phenology of durum wheat in Mediterranean environments using CERES-Wheat
Field Crop Res.
(2017) - et al.
Sowing date and nitrogen fertilisation effects on dry matter and nitrogen dynamics for durum wheat: an experimental and simulation study
Field Crop Res.
(2010) - et al.
Simulation of environmental and genotypic variations of final leaf number and anthesis date for wheat
Eur. J. Agr.
(2012) - et al.
Effects of water deficits on evapotranspiration from barley
Agric. For. Meteorol.
(1995) - 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)