Dry matter partitioning of sugarcane in Australia and South Africa

Partitioning of dry matter (DM) in sugarcane is of interest for two fundamental reasons. Firstly, sugar production depends directly on partitioning of crop biomass to the stalk and then to sucrose stored largely in stalk parenchyma. Secondly, various DM components of the stalk and particularly sucrose concentration are used to calculate the value of cane consignments delivered to the mill. In this study we review data from serial harvesting experiments in South Africa and Australia to discover similarities and differences in DM partitioning of sugarcane grown in a wide range of conditions and to gain a better understanding of the factors influencing DM partitioning as a basis for improving functional responses in sugarcane simulation models. Firstly, climatic and genetic variations in DM partitioning to various above ground plant components of sugarcane and to sucrose within the stalk component, are examined at a broad level. We then assess the robustness of sucrose partitioning rules used in the two international sugarcane models (APSIM and CANEGRO) and provide enhancements on predictions of sucrose content (SC) on a DM basis in sugarcane stalks.Well-adapted cultivars in widely different climatic conditions were similar with respect to partitioning of biomass to various above ground organs. The trash component (dead leaves and dead stalks) was one that varied most between cultivars and growing conditions. The stalk fraction of green biomass (biomass less trash) reached a maximum of about 0.85 when green biomass yield exceeded 60 t ha⁻¹ regardless of cultivar or extremes of water regime. Opportunities for further improvements of this trait through breeding were not obvious but harvesting could be delayed to ensure that stalk fraction is as high as possible.Variations in SC with respect to age, biomass and seasonal variation in climate, were explained by a simple conceptual model of cane stalk segments which are rapidly filled with sucrose when young and may remobilise sucrose when carbohydrate is required for rapid expansive growth. Season and age effects on whole stalk SC are due to varying proportions of young segments with low SC and older segments with high SC. A mechanistic model capturing these concepts is required but as an interim measure an empirical model based on stalk mass, leaf number per stalk, leaf area index and temperature accounted for a large proportion of the variation in SC across Australian and South African cultivars and conditions. Uses for this model together with CANEGRO or APSIM include improved SC through manipulating irrigation and the harvest schedule and more realistic targets for cane quality.