Multilevel modelling of the draft force required by seeder-fertilizers

Abstract

The high production costs in agriculture have guiding the adoption of farming systems and new management techniques as well as the sizing of agricultural machinery. In this regard, planning entails knowledge on the efficiency according to the energy requirements parameters, so the farmer shall consider the characteristics of the soil on which the implement operates. The performance assessment of sowing-fertilize machines shows the effect of some variables on the draft requirement, so the experimental conditions, which might lead to different outcomes of implement operation, must be regarded. Therefore, it is necessary obtain metanalytical estimates to integrate the results available in the literature. Grounded on systematic review, this study aimed to model the draft imposed by sowing-fertilizing machines considering fixed effects, mainly soil characteristics, and random effects associated to the selected experiments. It was found that the best models according to information criteria may not always meet the assumptions as normality of the distribution of residuals and homoscedasticity. The variables such as bulk density of soil, stubble conditions, depth of fertilizer placement, and speed could accurately explain the draft requirement with mean squared deviation of 2.93 whereas the referred evaluator for the ASABE standard was 63.51. Forthcoming works may analyze the repeatability of the models considering different seeders under diverse configurations and operation conditions