Economic modeling of mechanized and semi-mechanized rainfed wheat production systems using multiple linear regression model

Abstract

Mathematical modeling of economic indices is a challenging topic in crop production systems. The present study aimed to model the economic indices of mechanized and semi-mechanized rainfed wheat production systems using various multiple linear regression models. The study area was Behshahr County located in the east of Mazandaran Province, Northern Iran. The statistical population included all wheat producers in Behshahr County in 2016/17 crop year. Five input variables were human labor, machinery, diesel fuel, chemical (chemical fertilizers and chemical pesticides) costs, and the income was considered to be the output. The results showed that the cost of wheat production in the semi-mechanized system was higher than that of the mechanized system. In both systems, the highest cost was related to agricultural machinery input. Moreover, seed cost was lower in the mechanized system than that of the semi-mechanized system. The net return indicator was 993.68 $ ha−1 and 626.71 $ ha−1 for the mechanized and semi-mechanized systems, respectively. The average benefit to cost ratio was 3.46 and 2.40 for the mechanized and semi-mechanized systems, respectively, demonstrating the greater profitability of the mechanized system. The results of the evaluation of five types of regression models including the Cobb-Douglas, linear, 2FI, quadratic and pure-quadratic for the mechanized and semi-mechanized production systems indicated that in the developed Cobb-Douglas model, the R2-value was higher than that of the quadratic model while RMSE and MAPE of the quadratic model were determined to be smaller than that of the Cobb-Douglas model. Therefore, the best model to investigate the relationship between input costs and the income of wheat production in both mechanized and semi-mechanized systems was the quadratic model.