Yield Prediction of some Orchard Trees Based on Soil Characteristics under Rainfed Conditions

Studying the relationship between soil properties and crop yield is one of the most important tasks that should be considered under rainfed conditions. The cropping system in the north western coastal zone is mainly includes of olive and fig grown under rainfed conditions. The intercorrelation among soil properties and their influences on olive and fig yields was investigated. Therefore, multiple regression analysis was employed to generate coefficients for relative contributions of selected soil properties such as soil depth, gravel, soil texture (expressed by sand, silt, and clay), salinity, pH, and calcium carbonate on crop yield. Data of olive and fig yield were collected during 3 consecutive years of 2013, 2014, and 2015 from two different locations, namely wadi Hashem basin (7 sites) and wadi El Heriga basin (9 sites). The collected soil data were interpolated and mapped across the study areas. Statistically, the Pairwise comparisons of crop production demonstrated that there was a significant difference among some of the studied sites in regard to their yield potentiality. To predict the crop yield of the studied plants based on the selected soil properties, Partial Least Square Regression Model (PLSRM) was used. It depicted a profound predication model of olive and fig yield, where it produced R^2 of 0.892 with RMSE of 0.093 and R^2 of 0.995 with RMSE of 0.042 for olive and fig respectively. Eventually and for the current study, it could be concluded that most of the studied soil properties have a great influence on olive and fig yield under the rainfed condition.