Investigation of carbon flux partition in metabolism of cassava storage roots for developing a model for crop yield prediction under the influence of genetic and environment

Cassava is the important staple crop for people worldwide, and being a cash crop for Thai farmers. Under the current climate changes, sufficient cassava production to the global demand is on risk. Improvement of cassava yield per unit cultivated area, not only in terms of amounts but also stability, is thus a grand challenge to be achieved for securing food sufficiency, especially when agricultural areas become limiting by urbanization. Cassava breeding program has offered various elite lines by improving their genetics. However, the process always takes substantial time, especially at crop selection step where a number of bred lines are growing to test whether they contain the desired traits. Many candidate lines are dropped out because of their phenotypic variation under tested environments. This problematic circumstance underlines the need of rationale that enables cassava yield estimation by including both endogenous and environmental influences. Here, we proposed to use carbon flux pattern, which will be predicted from the constraint-based model to describe how carbon is utilized in metabolism towards production of root biomass under the exposed condition, to estimate cassava yield. The predicted carbon flux pattern will be calculated from growth of plant under cultivated condition, thus reflects the inclusive effects of endogenous and exogenous factors. In this study, the predicted carbon flux distribution in cassava storage under the difference of genetics (Kasetsart 50, Rayong 9, Rayong 11, CMR38-125-77) and environment (irrigation, early-drought, and late-drought) will be investigated as to infer the pattern of carbon utilization in metabolism of cassava storage roots. With these carbon flux patterns, the mathematical model based on data mining approach will be formulated for relative evaluation of yield under the influence of genetics and environments i.e., cassava varieties, plant ages, and growing conditions. We expected that the resulting mathematical model would facilitate plant breeders during crop screening process, i.e. by reducing time for crop testing, and further apply for crop yield prediction.