Over the past 30 years there has been a virtual explosion in technology associated with data collection, analytical techniques and computational science. Scientists interested in plant genetics, growth, physiology and management have unprecedented opportunities for using computer-based technologies to study and develop an integrated understanding of how plants function and can be optimally managed to meet the goals of growers in rapidly changing economic and climate contexts. Field and remote sensing and data transfer technology has made it possible to gather real-time data more quickly than ever before. But there is a lack of creative ideas about how these data can be optimally used. Similarly, genotype-specific genetic data can be obtained for a fraction of the time and cost of a decade ago but the application of this genetic information to solve practical production issues is still largely illusive. Determining optimal genotypes requires identifying optimal phenotypes, and optimizing phenotypes for specific environments requires dynamic and integrated understanding of how plants grow and respond to changing environments and management practices. The key to developing this understanding is computer simulation modelling. From my perspective, modelling is best used to develop an integrated understanding of specific processes or phenomena and then applications of the derived understanding can be applied to address practical problems; rather than starting with a specific applied goal and trying to build a model primarily based on empirically-derived relationships without a fundamental, mechanistic understanding of the system. To achieve the goal of developing integrated, mechanistic, functional structural plant models requires a paradigm shift that embraces rather than attempts to avoid complexity. Traditional crop modelling approaches attempt to simplify models to the fewest driving variables possible that can explain/predict plant behavior for the situation being addressed. Truly integrative modelling for increasing understanding recognizes and embraces the complexity of plants and tries to build that complexity into functional models. Ultimately the goals of these types of models will be to simulate phenotypic appearance and behavior, from cells to whole plants based on genotype. My attempts to begin building such a model will be illustrated by progress that has been made on the L-PEACH model,


Theodore Dejong

From Plant Cells to Plant Fields e-session

Photos by : Tyssul Patel