Seedling emergence model for tropic ageratum (Ageratum conyzoides)

Abstract

The timing of weed seedling emergence relative to the crop is important in planning and optimizing the time of weed control, but very little work has been done to predict seedling emergence of tropical weed species, especially in low-input and small-scale farms. We developed a simple model based on hydrothermal time to predict seedling emergence of tropic ageratum. Hydrothermal time at 2-cm soil depth was calculated from soil moisture and soil temperature simulated from several micrometeorological and soil physical variables. The model was developed using 5 yr of field emergence data from a continuous corn–cassava production system in southwestern Nigeria. Percentage of cumulative seedling emergence from the 5-yr data set was fitted to cumulative soil hydrothermal time using a Weibull function. The predicted cumulative emergence curve significantly matched observed field emergence (r2 = 0.83). Model predictions were evaluated with root mean square error (RMSE) using four field emergence data sets from southeastern Nigeria (RMSE ≤ 10.1) and Los Banos, Philippines (RMSE = 8.9). RMSE values ≤ 10 indicated that predictions represented observations well. With such models, extension personnel working on tropical soils, especially in West Africa, may be able to provide additional advice to farmers on the appropriate time for the management of tropic ageratum.