Rice yield was highest under conventional tillage (2.73 t ha
−1
)
and differed significantly from the rice yield after maize intercropped with Brachiaria under the no-till system (1.79 t ha
−1
), but
not from the rice yield after maize intercropped with soybean with no tillage (2.20 t ha
−1
). Among yield components, only the number of spikelets per panicle differed significantly between cropping
systems, despite an interaction between the season and cropping system (for the component ‘number of panicles per plant’ the interaction between the season and the cropping system was
higher (P < 0.0006) than the cropping system effect (P < 0.0093), see
Table 4).
Significantly higher rice yields were obtained with organic
inputs and mineral fertilizer (MF) than with no fertilization
(NF), although the difference in yield between fertilization
treatments varied with the year (source of the interaction fertilization × season). Among yield components, the number of panicles
per plant and the number of spikelets per panicle differed significantly between cropping systems. Conditions that affect these
components occurred mainly between early tillering and heading.
The percentage of filled grain was not affected by the management system or fertilization. No difference was observed in
1000-grain weight between seasons, management systems, and
fertilization treatments.