SELECTION INDICES FOR EARLY SELECTION IN DUAL-PURPOSE MERINO PRECOZ SHEEP UNDER LOW-INPUT CONDITIONS

Resumen

In central Chile’s low?input drylands, sheep breeding is challenged by early selection at weaning, often without pedigree records, and limited resources. These constraints reduce genetic progress, as farmers rely on visual selection. Nevertheless, selection indices provide an alternative for more efficient and accurate genetic improvement of dual-purpose breeds such as Merino Precoz (MP) sheep. This study aimed to estimate regression coefficients and relative contributions (RC) of key traits within several selection indices to optimize breeding objectives and estimate genetic progress in MP sheep. Five indices were simulated under three economic scenarios based on the relative price of fibre diameter (FD) to greasy fleece weight (GFW), integrating pre-weaning and maternal data. The breeding objective included GFW, FD, number of lambs weaned (NLW), weaning weight (WW) and adult weight (AW). Results showed that WW consistently had the greatest RC (8.74–8.82%), while NLW contributed less (4.18–4.22%), and excluding NLW notably increased index accuracy (0.47–0.49 vs. 0.35). Wool traits gained importance as the economic value of FD increased, although WW remained the most influential trait across indices. Expected annual genetic gains ranged from 0.014 to 0.025 kg for GFW, -0.044 to 0.022 µm for FD, 0.279 to 0.300 kg for WW, and 0.525 to 0.621 kg for AW. These findings suggest that excluding NLW from the breeding objective improves accuracy and increases the predicted response for several production traits under low-input systems, while dam live-weight records provide little improvement in accuracy. The coefficients presented in this study are valuable for farmers practicing early selection at weaning, enabling more efficient genetic improvement of wool and meat traits under resource-limited conditions.