Managing Phenotypic Data For Use In Crop Simulation: Lessons From The ICASA Standards

Ecophysiological models characterize crop responses to genetics, environment and management. They find application in research ranging from precision crop management to phenotyping. Phenotypic data are required both for model parameterization and evaluation. Traits vary in scale from biochemical or cellular to crop community. Many models output similar data for crop growth (e.g., leaf area index and mass), phenology and environmental responses. However, no standards are in widespread use, with the result that interchange of data among model developers and end-users has suffered. Metadata for crop management, soil conditions and weather are similarly neglected. Standards developed by the International Consortium for Agrotechnology Transfer (ICASA) are being revisited for prediction of climate change impacts on agriculture via the global Agricultural Model Intercomparison and Improvement Project (AgMIP.org). The current standard contains over 600 crop descriptors (phenotypes) and 620 metadata fields. AgMIP is developing data translators that map the ICASA variables to nomenclature and units used by individual models in order to test models with common datasets. These efforts have confirmed the utility of the ICASA standards and fomented interest in their expansion to include data recorded on sub-daily time steps and to document measurement protocols. This progress is emerging from a confluence of forces that include relaxing formatting requirements associated with the standards and improvements in software and hardware. The modeling community also seems to have evolved toward a more collaborative research ethic that supports data sharing and model improvement, especially with respect to including genomic information in crop simulation models.