RICE CROP DURATION AND LEAF APPEARANCE RATE IN A VARIABLE THERMAL ENVIRONMENT - I - DEVELOPMENT OF AN EMPIRICALLY BASED MODEL

Variable crop duration is a major constraint to rice double cropping i n arid irrigated environments, such as the Sahel. Photoperiodism and l ow air and water temperatures during the cool season are the major cau ses of variability, and cultivars are needed whose photothermal respon se provides a more stable crop duration. A previous study analyzed cul tivar photothermal constants on the basis of progress to flowering. Th e present study sought to identify, on the basis of leaf appearance ra tes, the phenological stages that are most sensitive photothermally, a nd to explore technical options to screen germplasm for stable crop du ration. Three Oryza sativa, indica-type rice cultivars (Jaya, IKP, IR6 4) were sown in the field at 15-day intervals during the dry season of 1995 (11 sowing dates) and 1996 (5 sowing dates) in Ndiaye, Senegal, under full irrigation and wide spacing to reduce microclimate variabil ity. Mean daily water temperature (T-w) varied from 13 to 35 degrees C . After seed soaking, the rate at which the first leaf (L-1) appeared was linearly related with T-w, with a base temperature (T-base) of abo ut 10 degrees C. Appearance rates of the subsequent three leaves (L-2- L-4) had a similar T-base, and presented a distinct temperature optimu m (T-opt) at about 23 degrees C, beyond which development rates decrea sed. Errors were too large to determine differences among cultivars in thermal constants. No significant temperature response was observed f or the leaf appearances between L-5 to the flag leaf (L-12 to L-20). C rop duration to flowering varied by 45 (IR64) and 63 days (Jaya). Thes e variations were associated with highly variable leaf numbers in all cultivars, including photoperiod-insensitive IKP. One-third of the var iable duration was hypothesized to be due to a variable basic vegetati ve phase (BVP), caused by variable germination and leaf appearance rat es, and two-thirds to variable duration of panicle induction after BVP . Water temperature was the main determinant of both sources of variab ility. A simulation model, describing these temperature and photoperio d effects on leaf number, growth duration and leaf appearance rates, w as developed using the 1995 data, and satisfactorily validated with th e 1996 data. The model was used to identify phenological-stage and cul tivar-specific causes of variable crop duration. (C) 1998 Elsevier Sci ence B.V.