The use of crop models to classify environments and rank cultivar traits based on importance for target environments

Description of the topic:
Due to the complexity of crop production systems, mechanistic or process-based crop models are often used to understand genotype x environment x management (GxExM) interactions in order to optimize them to increase grain yield (Teixeira et al. 2017). Different crop trait impact scenarios could be simulated, such as increased radiation use efficiency, enhancing photosynthetic pathways, specific changes in plant architecture to increase biomass, optimized harvest index, optimizing phenology, deeper and faster root growth to enhance water and nutrient capture, etc. These are based on traits that have previously indicated promise for example in boosting yield potential (Reynolds et al., 2012), adaptation to heat and drought stress (Cossani and Reynolds, 2012), etc. Crop modelling can help confirm the potential value of traits and their combinations, information that can help to design crosses targeting different environments (Reynolds and Langridge, 2016). 

Work expectations: 
At the end of the period here in CIMMYT, it is expected to have a scientific article ready to submit to a peer-reviewed journal. 

Objectives: 

  • Implement SIMPLE and DSSAT wheat, maize, and rice models to simulate crop growth, development, and yield inside the EBS software;
  • Classify environments using SIMPLE and DSSAT crop models based on drought and heat crop modeling simulations.
  • Identify and rank cultivars traits based on importance for target environments and agronomic practices

Required skills:

  • Knowledge of agronomy, plant science, and crop modeling to be able to run simulations and interpret crop model outputs. 
  • Knowledge of computer programming language such as R, Python, and Fortran;
  • Use High Performance Computing (HPC) Linux clusters for crop model simulations and data processing.
  • Experience on writing scientific articles.

Activities:

  • Apr/May:
    • Write wrapper code using R (SIMPLE model) and Python (DSSAT models) to read pre-set input data, execute crop model simulation, and display output variables (proof of concept using SIMPLE and DSSAT models)
  • Jun/Jul
    • Implement wrapper code to allow for user to enter their own experimental data and run simulations
  • Aug/Sep
    • Run pre-set experiments using EBS software user interface
    • Run user’s experimental data using EBS software user interface