Forecasting olive ( Olea europaea) crop yield based on pollen emission

Abstract

Reliable crop volume estimations are an increasing necessity to allow optimised and effective agronomic management. In this paper, a methodology to evaluate future olive crop yield several months in advance is presented. Olive tree phenology, airborne pollen concentrations, meteorological data and fruit production data were analysed in the province of Córdoba (Andalusia, Spain) over a period of 20 years (1982–2002). Data were integrated to obtain models for predicting fruit production. In this study, annual Olea pollen emission is shown to be a reliable bio-indicator to forecast olive fruit production up to 8 months in advance. Hirst volumetric pollen traps were found to be an accurate tool in olive crop yield forecasting. May rainfall was the most important meteorological parameter affecting final fruit production. Three statistical models, with different elapsed time between crop estimation and harvest, were developed: 8, 4 and 2 months. All showed high determination coefficients (73–98%) with a significance of 99%. The models revealed no significant differences between expected and actual data in the 2001/2002 olive harvest (not included in the model). Comparisons with other estimations from the regional government indicated that the meteo-aerobiological methodology showed higher anticipation and also a higher level of coincidence between expected and actual data. In addition to the positive results from this study, the scientific basis of the statistical performance provides accuracy and objectivity to this forecasting methodology.