Aim:

The aim of this research was to characterize agricultural systems using cattle in the Guayas River basin, Ecuador.

Materials and Methods:

The samples from 50 farms, and 19 variables (16 input and 3 output variables) were considered study cases. All instances were classified by factorial and conglomerate analysis during the first stage, based on the principal component and non-hierarchical clusters (k-mean) methods. Four conglomerates were defined and codified according to the mean values of the related variables, from I to IV. Finally, the productive response was evaluated by one-way analysis of variance, which considered each particular codification as study factors, based on the conglomerates; the output variables were deemed as the productive response.

Results:

The results showed the priority of the components, which expressed more variability in the systems studied, and depended on the use of inputs, residues, and food produced. The classification made according to the variables included in the input component comprised half the sample in conglomerates III and IV, with the highest values.

Conclusion:

The agricultural productive response was determined by the greater use of inputs; however, cattle response was higher in the categories with the lowest utilization levels.