Analysing productivity in intercropping systems ANALYSIS OF THE PRODUCTIVITY OF UPLAND RICE AND COVER CROPS IN RELAY INTERCROPPING SYSTEMS USING A MECHANISTIC COMPETITION MODEL

Relay intercropping systems with rice cultivars (WAB450-24-3-2-P18-HB (V4) and WAB56-50), the competitive Cajanus cajan and the weakly competitive Stylosanthes hamata were evaluated in the tropical savanna zone in Côte d’Ivoire. The INTERCOM model for competition served for data analysis and understanding of the system based on phenological, physiological and morphological characteristics of the species. Relationships between accumulated legume biomass and yield loss (YL) in rice showed that rice yield was reduced up to 63 % and 100 % by S. hamata and C. cajan respectively, depending on the introduction time. At low levels of YL both legumes performed similarly. At higher levels of rice YL, C. cajan performed better as it produced higher amounts of biomass. Differences in legume biomass at a specified level of rice YL were mainly created in the growth period following rice maturity, which is restricted by water availability. The lesser performance of S. hamata in that period was attributed to the difference in phenology, combined with the short-duration character of the species. The combination of early introduction and short growth duration of this species prevented a rapid leaf canopy development after rice harvest, resulting in the relatively poor biomass production. Options to improve legume biomass production in relay systems should focus on the period following rice maturity. Key-words : Legumes, cropping systems, yield loss, Côte d’Ivoire.