Multivariate methods application in the classification of dual purpose cattle raising systems in north Carabobo state, Venezuela

Upon request of the Northern Carabobo state Association of Livestock Producers and the State Unit of the Ministry for Production and Commerce, together with the National Institute of Agriculture Research (INIA) and Romulo Gallegos University (URG), the dual purpose (DP) cattle production units of northern Carabobo state, Venezuela, were classified as part of a technical diagnosis. Using surveys information was obtained from 48 farms initially identified as DP. An integral data base was created using 20 structural and functional variables, processed with multivariate methods and the statistical software CSTAT and BMDP. A principal components factorial analysis explained 53.66% of the accumulated total variance of the third axis; the Ascendent Hierarchical Classification Analysis divided farms into two DP milk - meat classes, and one fattening milk - meat class, the last was eliminated due to its meat production bias. The new classification of 38 DP milk - meat farms resulted in: 50.55% explained Total Variance; Class 1 (13 farms), Class 2 (18), Class 3 (7), with respectively greater, lower, and intermediate technological levels. Discriminant Analysis corroborated the classification with variables of highest discriminatory power: 78.60, 77.89 and 64.20% of total females in the herd; 9.38, 1.88 and 3.42% of machinery; 9.5, 26.74 and 49.04% of family labor; 30.63, 23.60 and 49.38% of area with native grasses; 2.0; 1.55 and 4.28 times/year of vitamin application; and income due to milk sales of 10,400,815; 2,671,799 and 3,125,442 Bs/year respectively.