Analysis of Spray Aid's residue curve for timing fungicide applications for grey leaf spot on maize

Abstract

Maize grey leaf spot is still a problem in commercial maize production, although the effect has been reduced with the widespread use of resistant varieties. It is estimated that, with a yield reduction of 1 %, annual economic losses of R5.6 million for maize producers in KwaZulu-Natal (KZN) are possible (based on the current average price of R1 332.871⁻¹ and KZN producing approximately 422 0001 per annum). SprayAid, a model for the timing of fungicide applications to control maize grey leaf spot, was developed between 1992 and 1994 in Turbo Pascal. It was re-written in Delphi as a 32-bit application (SprayAid32) and changes were made to allow multiple scenarios to be simulated. However, differences between predicted and observed severity of disease in the 1998/99 and 1999/00 seasons were noted. Through a sensitivity analysis, the residue coefficient curve was evaluated to determine the cause of the apparent model bias. It was found that the original equation to describe the curve (power fit model) performed better than the equation found with CurveExpert 1.3. A feedback mechanism is proposed to eliminate the bias in the model.