A process-based model to simulate changes in tiller density and light interception of sugarcane crops

Abstract

Tiller density on commercial sugarcane fields is a dynamic variable and part of the canopy that interacts with the environment. Three internal processes influence tiller density in well-watered crops. Those are (1) primary tiller germination, (2) underground branching (tillering) and (3) tiller senescence. Current sugarcane models depend on light interception, partly determined by tiller density, to simulate photosynthesis and evapotranspiration. In this paper a model is described that takes mechanistic canopy processes, like tillering and light competition into consideration. Five phases of canopy development are described. The model was calibrated on one cultivar and has mean errors close to half those of the CANEGRO model in tiller numbers per ha (RMSE=78,577 cf. RMSE=149,084) and%PAR light interception (RMSE=8.5 cf. RMSE=13.8), respectively. The model's mechanistic nature enables its application in more sensitive tiller and canopy dependent processes, such as high tiller densities, extensive ratooning and stool damage.