Multivariate characterization of Wollo sheep populations in the cool highlands of Ethiopia

Abstract

The objective of this study was to assess variation among populations of Wollo sheep managed in different locations. Three locations were selected purposely based on their sheep population size, sheep production potential and experiences of farmers in sheep breeding. Body weight, linear body measurements and qualitative data were recorded from a total 405 sheep (322 females and 83 males) in 68 flocks. Body weight and most of the linear body measurements of the indigenous sheep populations were significantly affected by location (p<0.05). The stepwise discriminant analysis procedure identified eight most significant discriminating traits: rump height (RH), wither height (WH), body weight (BW), tail circumference (TC), body condition score (BCS), ear length (EL), chest girth (CG) and body length (BL) from females; and five: RH, BW, WH, BL and horn length (HL) from males. The overall average error count estimate of sheep was 30%, accounting for 44%, 33% and 14% for Wogide, Borena and Legambo sheep populations, respectively. This means that 70% of sheep for all populations were correctly classified in their source population. In conclusion, the lowest misclassification of Legambo sheep population could be an indication of more uniformity within the population, whereas the highest misclassification of Wogide and Borena might be due to low geographical distance between their habitats. The largest distance was found between the Wogide and Legambo sheep populations and between the Borena and Legambo sheep populations, whilst the sheep populations from the Wogide and Borena were poorly differentiated. More genetic analysis research is recommended to assess variations and similarities of sheep and to identify the potential of genetic resources.