Use of a plastic temperature response function reduces simulation error of crop maturity date by half
- Publisher:
- ELSEVIER
- Publication Type:
- Journal Article
- Citation:
- Agricultural and Forest Meteorology, 2020, 280
- Issue Date:
- 2020-01-15
Closed Access
| Filename | Description | Size | |||
|---|---|---|---|---|---|
| 1-s2.0-S0168192319303867-main.pdf | Published version | 2.2 MB |
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
© 2019 Elsevier B.V. Understanding how crop development rate responds to the environment provides the basis for evaluating the impact of climate change on crop yield. In most crop simulation models, temperature response functions of development rate during the reproductive growth period (RGP) are assumed to only vary with temperature and not with other environmental factors. However, studies have indicated that the response functions may be plastic with other factors. Until now, little attention has been paid to this type of response. Here, using extensively collected field observations and data from intentionally designed interval planting experiments with winter wheat (Triticum aestivum L.), rice (Oryza sativa L.), and spring maize (Zea mays L.), we show that temperature response functions during RGP are plastic with day of year of flowering/heading (DOYR). Coefficients of determination between DOYR and development rate were significant for 69% sites. Partial correlation coefficients between development rate, temperature, and DOYR suggest that DOYR explains almost the same variability in maturity date as temperature. The plastic model was developed by coupling DOYR with a linear temperature response function. The model can improve the fitting efficiency by 112%, while dependency between DOYR and temperature explains less than 25% of this improvement. The average RMSEs of simulated maturity date estimated by the plastic model in the three crops were 2.1, 2.5, and 3.7 d, respectively, while the corresponding values given by widely applied traditional models were 3.1, 6.5, and 7.4 d, respectively. Therefore, the plastic model can reduce simulation error by half. Moreover, simulation errors resulting from the plastic model have less systematic bias than traditional models. The plastic model simply and effectively provides accurate estimates of crop maturity and reduces the system deviation of the estimates. Coupling the plastic model of crop development with crop simulation models will likely decrease uncertainties in simulated yield under warming conditions. Additionally, results of this study will encourage future studies of other phenotype plasticity considered in current crop simulation models.
Please use this identifier to cite or link to this item: