Résumé
High-yielding rice varieties (HYV) show strong phenotypic plasticity, notably in compensatory tiller production and panicle size. In a precursor study on the high-tillering cv. IR72 we showed that the plasticity is adaptive by enabling yield stability across different stand densities, and we presented and validated the functional-structural crop model SAMARA, developed for the study of phenotypic plasticity. The present study analyses the phenotypic plasticity of high and low tillering plant types. Twelve contrasting, high-yielding irrigated rice genotypes were studied for the plasticity of traits contributing to grain yield under different season/site environments and stand densities in the Philippines. For the density factor, sequentially developing yield components along phenology, beginning with tillering, showed decreasing plasticity, indicating progressive adjustment to the initial stand density. For the site/season factor (climate), tillering was less plastic and panicle size and grain filling ratio provided for the strongest adjustments. Large plasticity was observed also for LAI and little plasticity for SLA and leaf morphology, particularly blade width. Genotypes differed strongly in organ size vs. number (leaves, panicles), representing organ "number" or "size" plant types. Up to 10 Mg ha(-1) grain yield were observed for both extreme types. The plasticity of panicle size was caused by variation in spikelet number and not kernel weight. All genotypes were able to fully adjust yield components to differential plant stand density, resulting in unchanged yield. However, between environments, genotypes having greatest maximal yields and those having stable (but on average smaller) yields showed different patterns of phenotypic plasticity for yield component and plant morphology traits. The SAMARA model was calibrated for five genotypes showing contrasting morphology and plasticity. It was able to predict most of the genotypes' variation in morphological and yield component traits. Future opportunities to predict improved rice ideotypes for changing environments and cultural practices through combined field experimentation and modelling are discussed. (C) 2016 Elsevier B.V. All rights reserved.