See more from this Division: A03 Agroclimatology & Agronomic Modeling
See more from this Session: Symposium--Incorporating Genomic Knowledge Into Crop Simulation Models
Tuesday, November 2, 2010: 3:15 PM
Long Beach Convention Center, Room 103A, First Floor
Simulating plant phenotypic plasticity depending on environmental condition is a drawback in either process based or functional structural plant models. Although usually the biomass production per m² can be well predicted based on leaf area index (LAI) and water use efficiency, the underlying plant architecture that depends on the biomass partitioning can vary a lot upon environmental conditions such as plant density. On the other hand, plant breeding is based on measurements on plant architecture (plant height, leaf surface, etc.) that are often too variable to be used in plant growth model. These measurements, which mix up both environmental effect and the plant functioning, are not stable parameters. Instead, hidden parameters that control both architecture development and growth (biomass production and partitioning) are more stable and must be computed by inverse method once the model is built from the measurement on the plant architecture. These parameters should be better candidates to predict the yield in quantity and quality (for instance the organ number and size), and to be linked with QTL in genetic issues.
We will present a mathematical model GreenLab which simulates the interaction between plant architecture and functioning. In GreenLab, organs in the plant architecture play roles as source and/or sink, and the plant source-sink ratio has effect on plant development, such as branch appearance, or fruit set. By inverse method, parameters of source and sink functions can be identified from measured plant data. The possible link between GreenLab and genomic knowledge is discussed. Examples are shown on applying GreenLab on analysis of several crops with different genetic background.