Integrating emergence and phenology models to determine windows of action for weed control: A case study using Senna obtusifolia

The success of integrated weed management strategies is contingent on the accuracy of control actions in both time and space. While emphasis has been given to spatial accuracy, timing accuracy has been largely neglected. Weed control timing must consider not only the total duration of weed interference with the crop, as done by the traditional critical period of weed control (CPWC) based on yield protection only, but also weed growth, size, and susceptible phenological stages. In this study, we expand upon the idea of using weed emergence models for timing weed control by integrating them with phenology probability models for key weed growth stages to optimize timing of control actions, here referred as Critical Control Windows (CCW). Combining the CCW with thresholds for yield loss due to weed interference and thresholds for weed survival risk makes it possible determining the frequency with which control actions should be implemented to maintain crop yield and weed populations at desired levels. Using Senna obtusifolia as a study case, vegetative and reproductive phenological stages were modeled as a function of seedling emergence for different cohorts. Chronological and thermal-time models provided robust predictions of S. obtusifolia phenology. CCW did not always coincided with CPWC for several crops when considering 10-cm tall plants as the threshold for control. In general, for summer row crops, CCW required 2 postemergence control actions and sometimes 1 action outside the CPWC. The results of the present research illustrate how predictive models can be used to develop CCW that will complement the traditional CPWC. These two concepts when used complementary can increase not only timing accuracy, but also efficiency of weed control.