Interference of redroot pigweed (Amaranthus retroflexus L.), large crabgrass (Digitaria sanguinalis (L.) Scop.) and hairy galinsoga (Galinsoga ciliata (Raf.) Blake) in pepper (Capsicum annuum L.)

Field experiments were conducted to determine the influence of emergence time and density of redroot pigweed (Armaranthus retroflexus L.), large crabgrass (Digitaria sanguinalis L.) and hairy galinsoga (Galinsoga cillata L.) on yield or biomass loss of bell pepper (Capsicum annuum L.) in 1998 and 1999. Selected weed densities of 0, 1, 2, 4, 8, 16, 32 plants m–1 were established within a 15cm band on either side of crop row for each weed species and the mixture of the three weed species. The effects of weed emergence time relative to pepper were studied by investigating the different yield responses to weeds emerging right after and two weeks after transplanting of pepper. ^ Both weed density and emergence time affected the magnitude of yield loss. The relationship of pepper yield or biomass with weed density and emergence time in single and mixed species interference was adequately described by the single and expanded version of hyperbolic nonlinear equations for each year. Variation in environmental conditions caused variation in competitiveness of weeds, which in turn caused variation in parameter estimates between years. The empirical nature of these models determined their inability to explain the variation of parameter estimates between years. Large crabgrass was the most competitive species in 1998 while in 1999 redroot pigweed exhibited the strongest competitive ability. Hairy galinsoga was the least competitive species in both years. Economic threshold was calculated based on parameter estimates for each weed species, which was lower for weed species with stronger competitive ability. ^ One important contribution of this study is relating to model development. Expansion of the existing model was proposed to accommodate multiple weed species interference. Introduction of one parameter into the existing models to account for the change of maximum yield loss with relative emergence time of weeds could improve the fit of data in both single and multiple weed species study. The performance of two types of the currently most used models was also evaluated. The relative leaf area model did not show obvious superiority in this study. The results of this study are important to weed management of pepper. ^