Jvk. Afun et al., Weeds and natural enemy regulation of insect pests in upland rice; a case study from West Africa, B ENT RES, 89(5), 1999, pp. 391-402
Effects of five weed management regimes on abundance of weeds, insect pests
, generalist predators and on pest damage and rice yield were investigated
in upland rice in Cote d'Ivoire over two years. In both years there was a h
ighly significant negative correlation between weed biomass and grain yield
across all treatments. Only two pest insect groups, Nephotettix spp. (Cica
dellidae) and seed sucking Heteroptera, were consistently more abundant in
unweeded plots and had a consistent significant positive correlation betwee
n abundance and weed biomass across all seven treatments. These polyphagous
groups may have been more dependent on resources provided by weeds than th
e other pests studied. However, pest damage was not affected by presence or
absence of weeds, suggesting that populations were below a damage threshol
d. Ants were the most abundant predators in the rice canopy and were most a
bundant in unweeded treatments. Abundance of both ants and spiders was sign
ificantly positively correlated with weed biomass across treatments. Abunda
nce of reduviid bugs was positively correlated with weed biomass only in 19
95. Any benefits due to presence of weeds in the crop were completely outwe
ighed by loss of crop yield due to weed competition. However, if crop losse
s due to weeds were sufficiently reduced, it is possible that significant l
osses due to insect pests might emerge.