Functional, structural and agrohydrological sugarcane crop modelling: towards a simulation platform for Brazilian farming systems

Author(s)

Murilo dos Santos Vianna

• Document type: Doctoral Thesis
• Press: Piracicaba.
• Institution: Universidade de São Paulo (USP). Escola Superior de Agricultura Luiz de Queiroz (ESALA/BC)
• Defense date: 2018-04-06

Examining board members

Fábio Ricardo Marin; Marcos Vinicius Folegatti; Klaas Metselaar; Felipe Gustavo Pilau

• Advisor: Fábio Ricardo Marin

Abstract

Sugarcane crop is the main source of sugar and the second largest source of biofuel in the world. Since the 1980s, Brazil has been the largest sugarcane producing nation, producing half of the global amount. Ethanol and biomass from sugarcane account for more than 15% of the country´s energy source. Nevertheless, commercial Brazilian sugarcane yield has plateaued at 75 t ha-1, and to meet the increasing demand for sugar and ethanol, the crop has strongly expanded towards central-western regions, where irrigation is mandatory to offset water stress risks. To support decision making and scientific guidance towards where and how the crop should expand and/or to increase yields, a heuristic view of the crop system is needed, which can mathematically be translated into a crop model. In turn, the effects of crop management, land use change, climate variability and agro-economic change factors on crop production and associated quantities can and have been assessed by using crop process-based models (PBM). In contrast to other crops, however, sugarcane has only two PBMs available for end users (DSSAT-CANEGRO and APSIM-Sugar), and further modifications of these models are required to better assess and support sustainable sugarcane production in Brazil. Therefore, this study aimed to develop, calibrate and evaluate different crop modelling approaches for Brazilian sugarcane farming systems, water management strategies, climate change impacts and canopy structures to support improved decisions for private and public stakeholders in the sugarcane sector, provide scientific guidance and establish a Brazilian platform of crop simulations. A new version of the sugarcane process-based model (SAMUCA) was developed to operate at phytomer level, focusing on soil mulch effects on crop growth and development, tillering process under competition for light and sucrose accumulation based on source-sink relations. The model was embedded into a modular platform dedicated to simulating the soil-plant-atmosphere and the management of the sugarcane farm system. The previous version of SAMUCA was also re-structured and coupled to the SWAP (Soil, Water, Atmosphere and Plant) agrohydrological model platform, focusing on soil water relations to crop growth. Moreover, a Functional-Structural Plant Model (FSPM) for sugarcane was developed by integrating the main crop components at the organ level (phytomer), based on a relative source-sink approach and a robust light model embedded into a three-dimensional modelling platform (GroIMP). All approaches were evaluated, and the performance under experimental conditions for different Brazilian conditions was determined. The performance of the new version of SAMUCA in a long-term experiment and under different Brazilian conditions was satisfactory, with agreement indices close to those of other widely used sugarcane crop models (CANEGRO and APSIM-Sugar). In addition, the modulated crop simulation platform can be used to host more crop models and integrate new features of Brazilian farming systems. The coupling of the SWAP-SAMUCA model was accomplished, and although non-expressive improvements in model performance regarding crop yield were noticed (with an overall 6% lower RMSE), the ability of SWAP-SAMUCA to simulate soil water content was higher than that of the original \"tipping bucket\" approach (32% lower RMSE). The Functional-Structural Plant Model for sugarcane was able to satisfactorily simulate canopy development, tillering and sucrose accumulation at the organ level and its integration at the whole-plant level. Besides its ability to simulate competition for light, helping to understand intra-specific competition among tillers, the sugarcane FSPM framework can be used to support sucrose accumulation and translocation mechanism studies as well as intercropping studies for sugarcane, which has already successfully been done for other crops. (AU)

• FAPESP's process: 14/05887-6 - Development, enhancement, calibration and evaluation of a process-based model for Brazilian sugarcane
• Grantee: Murilo dos Santos Vianna
• Support Opportunities: Scholarships in Brazil - Doctorate