Published: 6 December 2022| Version 3 | DOI: 10.17632/d5sfx6y4rg.3
Ferenc Istvan Bankuti,
Pedro Gustavo Lima


Data from 155 dairy farms were collected through on-site, semi-structured interviews. Partial least squares structural equation modeling (PLS-SEM) was used for data analysis. Four theoretical constructs (forage, genetics, management, and market) were analyzed to explore dairy farmers’ perceptions. We hypothesized that the four constructs have a positive and significant influence on concentrate feeding in dairy farms. Our results revealed that the genetic construct, followed by the market construct, has the greatest influence on concentrate use. The influence of the management construct was moderate, and that of the forage construct was low. PLS-SEM was effective in reducing the set of study variables to factors with a significant influence on concentrate feeding in Brazilian dairy farms.


Steps to reproduce

The sampled dairy farms were located in Paraná State, southern Brazil. Data collection was performed according to previous studies on animal production systems. The first step was to consult with experts in milk production, including state researchers and technical assistance providers from dairy industries, government agencies, and rural extension programs. A list of dairy farms containing information on geographical coordinates, telephone contacts, and milk transportation routes was prepared in collaboration with the experts. Farmers were contacted and invited to participate in the study. After consent was obtained, a face-to-face and semi-structured interview was conducted on-site (farm) by the first author. At each interview, farmers were asked to indicate other dairy farm operators to participate in the study. A total of 155 dairy farmers were interviewed. Responses were recorded using a tablet computer. The interview questionnaire was divided into two parts. The first section contained questions on sociodemographic, structural, and production variables. These data were used to descriptively characterize the sample. The second part contained questions on variables that were used to generate the constructs, such as herd breed composition, breeding techniques, and perceptions about management practices, forage production, herd genetics, and relationships with the dairy industry. Data on the daily amount of concentrate fed to lactating cows and the total daily volume of concentrate used on the farm were also collected. Responses to questions related to herd breed composition, breeding technique, concentrate level, and total volume of concentrate were presented in ascending order. Responses to questions evaluating farmers’ perceptions were rated on a 5-point Likert scale anchored at 1 (the most negative response) and 5 (the most positive response).


Universidade Estadual de Maringa


Agricultural Economics, Farming Systems Analysis, Farming System Modeling, Farming Systems Research