In West Africa, weeds are the major constraint to rice production. Native African rice Oryza glaberrima, however, is very competitive against weeds and research is ongoing to introgress competitive traits of O. glaberrima into high-yielding O. sativa types (Jones et al 1997). To support this work, the crop-weed competition model INTERCOM (Kropff and van Laar 1993) is being adapted and parameterized to evaluate weed-competitive rice plant types in a prebreeding support project.
The accurate prediction of phenology is the first step in applying process-based mechanistic models as tools for rice research and decision support at the site or regional level. Phenology is important in situations in which weeds compete with rice because competitiveness is affected by the partitioning of dry matter to organs and partitioning is driven by phenology. Phenology in INTERCOM is calculated from a thermal time (°C day) approach that assumes a linear developmental rate from plant emergence to flowering and another from flowering to physiological maturity.
Rice development to flowering, however, has been shown to respond to both temperature and photoperiod at specific stages. Thus, subphases-a basic vegetative phase (BVP), a photoperiod sensitive phase (PSP), and a panicle formation phase (PFP)- have been identified (Vergara and Chang 1985). Varietal sensitivity to photoperiod modulates the duration of PSP, depending on the positive deviation of daylength from a critical daylength value. These were used to modify the phenology routine in INTERCOM in a way similar to that in ORYZA_1 but allowing for multispecies structure, which is characteristic of INTERCOM.