Environmental variables related to wheat yields in the semiarid pampa region of Argentina

Abstract

The current paper analyses various environmental parameters in relation to wheat yields in Bordenave, Province of Buenos Aires, Argentina. The variables used are: precipitation (ppt), maximum (Tx) and minimum temperature (Tmi) as well as those obtained by applying the Palmer model. Decadic and phenological scales are used for data corresponding to the period 1977-2000. The stepwise method is used to obtain a multi-variate equation to calculate yield taking into account environmental variables only. For a five variates model the coefficient of determination, R2 equals 95.79% and the standard error of estimation is 129.0 kg ha-1. In the sample yields, the incidence of total variability for thermal variables is 42.7% and for hydrological variables, 53%. The value and sign of the correlation coefficients were analysed throughout the cultivation cycle. The α coefficient is mainly responsible for yield variance during tillering and stem elongation. There is good correlation with the values of Palmer's Drought Severity Index (PDSI) for the flowering and grain filling stages.