Application of GOSSYM to analysis of the effects of weather on cotton yields

Abstract Cotton ( Gossypium hirsutum L.) yields have declined since 1964 despite improvements in technology and introductions of higher-yielding cultivars. Cotton scientists have been unable to identify exact causes for the yield decline. As part of the effort to examine possible causes for the cotton yield decline, the cotton crop simulation model, GOSSYM, was used to analyze long term effects of weather on cotton yield trends. Weather, soil, and cultural input data recorded at five locations for over 20 years were acquired for this study. Simulations were completed for crops grown at the cotton breeders' variety trial locations at Florence, SC; Stoneville, MS; College Station, TX; Phoenix, AZ; and Fresno, CA. The weather factors were analyzed for each of the five locations to determine if there was a significant trend which could account for the yield decline. The results of the simulation analysis showed that, although weather varied greatly among years and caused large yearly fluctuation in yield, the weather effect on lint yield trends was neutral across the entire US cotton belt. Observed and simulated yields using weather variables agreed reasonably for each of the locations studied, and any discrepancies between observed and simulated yields appeared to be caused by other factors.