The effect of scope of analyzed data on accuracy of selection indices for evaluation of sow breeding value

Abstract

Objective of this paper was to establish to what extent the accuracy of constructed selection indices will be changed depending on the scope of analyzed data relating to fertility based on which parameters necessary for construction of selection indices (SI) were established. Fertility results of Swedish Landrace sows obtained on three farms in Republic of Serbia (farms 1, 2 and 3) were analyzed. Parameters necessary for construction of SI were determined by application of different mixed models of the method of Least Squares. For each farm three SI were constructed for evaluation of sow breeding value based on realized fertility in the way that parameters necessary for SI construction were calculated based on fertility results in the first (SI1), first two (SI2) and first three farrowings (SI3). Accuracy of constructed Sis varied within following limits - from low rIH = 0.255 (SI3 for Farm 2) to rIH = 0.405 (SI1 for Farm 3), and only in the case of SI2 for Farm 2 it was in the category of very low (rIH = 0.231). Introduction into analysis of fertility results realized in the second and third farrowing resulted in decrease of accuracy of constructed selection indices.

U ovom radu rezultati istraživanja su pokazali da je tačnost selekcijskih indeksa konstruisanih na osnovu parametara ocenjenih analizom rezultata plodnosti (BŽP i BOP) ostvarenih na posmatranim farmama bila od rIH = 0.255 (SI3 za Farmu 2) do rIH = 0.405 (SI1 za Farmu 3), a samo je u slučaju SI2 za Farmu 2 bila u kategoriji jako slabe (rIH = 0.231). Povećanje obima analiziranih rezultata plodnosti dovelo je do smanjenja tačnosti SI.