Format of a Weed Control Database

Abstract

Access to detailed descriptions of the effects of applying specific rates of herbicides to crops and weeds in various growth stages is hampered by the format in which the relevant information is stored. Compared to traditional formats of journal articles and herbicide registration labels, computer database systems could easily cross-reference data from large numbers of experiments and answer specific questions concerning herbicide performance under particular conditions. Availability of this type of information could have far-reaching consequences for herbicide users, consultants, researchers, and regulators. A preliminary format for storing weed control information in IBM-PC compatible computers was developed, including procedures to enter data and retrieve information. Weed control efficacy or crop injury data for all rates of a herbicide or tank-mixture applied at a specific growth stage in a single test are used to generate dose/response equations by means of regression analysis routines. The best fitting of these equations is then used to estimate herbicide rates that would provide ten categories of control, ranging from a “no observable effect level” (NOEL) up to complete control. Rates are estimated only for those categories either within or bordering the range of the observed data, the remaining categories are empty. The estimated rates are stored in the database, along with the original data and other qualifying information. Access to information is organized around searches for a single herbicide, plant species, or pair of species. Search output is presented in a tabular format listing species, growth stage, herbicide name, and herbicide rates for the ten categories: NOEL, 10, 30, 50, 70, 83, 90, 95, 98, and 100% control or injury.