Dynamics of pre-emerging herbicides on corn and brachiary straws and effectiveness of control on buva and capim-amargoso in different conditions of precipitation

Abstract

No-tillage crop system is widely used in several regions of Brazil. In this system, pre-emergent herbicides are applied on straw residues, which influences herbicide dynamics and can affect herbicide movement into the soil, thus impacting on weed control effectiveness. In addition, the dynamics of herbicide movement depends directly on the amount of rainfall and the time of application after a rainfall. If the amount of rainfall is not sufficient or if the straw residue is not sufficiently moist, herbicide movement may not be adequate to ensure an effective weed control. Studies on the dynamics and efficacy of herbicides applied on straw residues are still scarce, and many of them have been done on sugarcane straw. The objective of this research is to evaluate the dynamics of diclosulam, flumioxazin and diuron+sulfentrazone in straw residues (corn and brachiaria, considering a no-tillage scenery) and soil, and to observe how this behavior affects the efficiency of controlling Digitaria insularis and Conyza ssp., important weeds in various agricultural crops. The experiments will be carried out in the field, laboratory and greenhouse. The straws will be obtained and characterized in the field, and later used in the other experiments. The herbicide dynamics in straw residues and soil, testing different rainfall (0, 10, 20 and 30 mm) after application and times of rainfall occurrence (0, 2, 5, 7 and 10 days) after application, will be studied in pots, and the herbicide quantification will be performed in the laboratory (HPLC). The efficacy of herbicide control will be tested in greenhouse, using the same rainfall and times mentioned above. The data will be previously submitted to the tests of homogeneity of variances (Levene) and normality of residues (Shapiro-Wilk). Depending on the result, data can be analyzed by parametric statistics (F test and Tukey test) or non-parametric (Kruskall-Wallis test). The level of probability of error adopted will be 5%. (AU)