DEVELOPMENT OF APREDICTIVE MODEL TO ESTIMATE CORN YIELD WITH THE USE OF SPAD 502 CHLOROPHYLL METER. AZUERO, PANAMA. 1995-2001.

Abstract

A predictive model was developed to estimate grain yield of corn. The study was characterized by different rainfall regimes with normal (1996, 1998 and 2001), scarce (1995, 1997 and 2000) and abundant precipitation years (1999). A multiple correlation was found between the grain yield and the others evaluated variables. The highest correlation index was obtained with the chlorophyll meter (SPAD 502). A regression analysis was made by year and by year group, considering both the chlorophyll meter lecture and the density of plants per square meter. There were no differences among equation's coefficients for years 1998, 1999, 2000 and 2001. Neither among equation's coefficients for years 1995, 1996 and 1997. Three prediction equations were calculated according to the year group: The first included all years (Y= 0.07706 (Clor) + 0.31186 (Plants/m²)); the second incorporated data of the 1995-1997 period (Y= 0.06157 (Clor) + 0.35837 (Plants/m²)) and the last equation included years 1998 to 2001 (Y= 0.13535 (Clor) - 0.13518 (Plants/m²)). Determination coefficients of all equations were greater than 0.96. Finally, the annual equations and the three year group equations were validated with grain yield of the 2001 trial. Chi Square Test was used for this purpose. Results indicate that model representated by the all years equation is the best for predicting yield.