Compensatory phenotypic plasticity in irrigated rice: Sequential formation of yield components and simulation with SAMARA model

Abstract

High-yielding rice varieties (HYV) show strong compensation among sequentially developed yield components (YC). This phenotypic plasticity has adaptive value but for crop improvement, more information is needed on its effects on yield. SAMARA, a deterministic crop model predicting trait-trait and trait-environment interactions by simulating morphogenetic processes and competition among sinks for assimilates, was developed to study crop phenotypic plasticity. Dynamics of YC and morphology were observed on the HYV IR72 planted at standard and 4-fold greater density in 4 environments in the Philippines in 2012/13. Data for other years/seasons were obtained for model validation. Sequential path analysis was used to determine the phenotypic plasticity of traits consecutively contributing to yield. Tiller number at flowering (R2=0.94) and maturity (R2=0.84) and grain yield (R2=0.77) were predicted accurately for independent datasets. The model also predicted accurately density effects on aboveground dry weight (agdw), plant height, leaf size, spikelet number per panicle and filling percentage. Tiller and leaf mortality were over-estimated under high density. Overall, the model predicted satisfactorily the sequential compensation processes among YCs. They led to stable grain yield despite large morphological differences among density treatments and environments. Sensitivity analysis of simulation outcomes vs. variation in crop parameters indicated that modified genotypic tillering ability, phyllochron or leaf size had little effect on final grain yield because of compensations by other traits, although IR72 appeared to have an optimal combination of parameter values. Larger effects on grain yield were predicted for variation of parameters affecting the sensitivity of leaf and tiller mortality to assimilate resources and the ability to mobilize stem non-structural carbohydrates during grain filling. The model will be used next to perform physiological trait dissection and plasticity analyses for diverse genotypes.