•Relatively high levels of polymorphism in Tunisian breeds, similar to Mediterranean sheep populations.•Uncontrolled crossbreeding is the main cause of genetic admixture between Barbarine and Queue Fine de l'Ouest.•A marked genetic differentiation of Noire de Thibar breed, confirming the European gene flow.

The aim of this study was to assess the genetic diversity and differentiation patterns of four Tunisian sheep breeds. A total of 186 animals belonging to Barbarin (n=59), Western Thin Tail (n = 46), Black Thibar (n = 40), and Sicilo Sarde (n=41) breeds were genotyped for a panel of 30 microsatellite markers. In addition, a sample of 29 Spanish Merino sheep breed was used as an outgroup for tree topology. All markers were highly informative with PIC values ranging from 0.588 to 0.885 and a mean number of private alleles of 0.016. In all breeds, inbreeding was indicated by heterozygosity deficiency with estimated F-IS values varying from 0.134 (Barbarin) to 0.165 (Western Thin Tail), and a highly significant departure (p < 0.01) from Hardy-Weinberg proportions was observed. Wright's F-ST index indicated low genetic differentiation between Tunisian breeds (0.017). The lack of clear genetic differentiation was also evidenced by cluster analysis using STRUCTURE as well as by the significant gene flow observed between breeds, especially Barbarin and Western Thin Tail, on the basis of the estimated Reynolds' genetic distances (D-R) and the number of migrants (N-m). The highest DR values were observed between Sicilo Sarde and the three other breeds, reflecting different phylogenetic origins of this breed. The study represents a first step towards further genetic characterization of Tunisian sheep breeds and results presented here are useful for better management and conservation of sheep resources. (C) 2015 Elsevier B.V. All rights reserved.

In this study, genetic variation of ovine prion protein in Tunisian sheep breeds was analysed. Sequencing of the entire coding sequence of prion protein gene (PRNP) was performed in a total of 201 samples belonging to four breeds (Barbarin, Western Thin Tail, Sicilo Sarde and Black Thibar). Five haplotypes (ARQ, ARR, ARH, AHQ and VRQ) and 10 genotypes were observed based on codons 136, 154 and 171, with different frequencies among the investigated breeds. The ARQ, ARR and ARH haplotypes were present in all breeds, the VRQ haplotype was observed at low frequencies in Barbarin and Western Thin Tail breeds. The ARQ and ARR haplotypes were the most common with frequencies ranging from 33.4% to 47.8% and from 26.5% to 46.5% respectively, in the different breeds. Moreover, 12 additional non-synonymous (Q101R, M112T, G127S/V, M137T, L141F, H143R, N146S, R151G, Y172D, N176K and H180Y) as well as 2 synonymous polymorphisms at codons 231 and 237 were found in the PRNP gene. Among them, the R151G polymorphism has never been described in sheep. Moreover an insertion of an octarepeat in the ARQ haplotype has been observed. These results represent the first survey on PRNP variability in Tunisian sheep and may serve as basis for the development of breeding programme to increase scrapie resistance in national sheep breeds.

INTRODUCTION The D’man breed originates from the oasis in the south of Morocco and is well known for its high prolificacy and its accelerated lambing rhythm (Lahlou-Kassi et al., 1989). Introduced in Tunisia in 1994, the D’man is being diffused in the south as pure breed and is a potential candidate for cross breeding with local breeds in other favorable areas (Lassoued and Rekik, 2001). In either case, the genetic characterization of the breed is a prerequisite step and this work addresses the main environmental factors affecting growth performances and the estimation of the genetic parameters of growth traits.

SUMMARY - "Estimation of heritability of growth characters in Barbarine lambs using three methods: MIVQUE(O), ML and REML". Three estimators of the sire variance (02J, representing the fourth of the additive variance (02J, and the error variance (0'3 were compared: MIVQUE(0) which is a minimum variance quadratic unbiased estimator with a priori values oza = O and oze = I; maximum likelihood (ML) and restricted maximum likelihood (REML). 104 Barbarine rams with at least 30 offspring each were used in this study. Our results indicate that ML estimators were lower than MIVQUE(0) and REML parameters. Estimators from the latter two methods were more or less comparable, exception made for birth weight, with slightly inferior values from MlVQUE(0). REML heritability estimates were (0.43+0.05), (O. 15+0.03), (0.20+0.04), (0.30+0.04), (0.36+0.05), (O. 15-t-0.03), (0.2920.04) and (0.4 I-t-0.05) respectively for lamb weights at birth, IO, 30, 70 and 90 days of age, average daily gain (ADG) from 10-30 days, ADG 30-70 and ADG 30-90.