Selection is the major force affecting local levels of genetic variation in species. The availability of dense marker maps offers new opportunities for a detailed understanding of genetic diversity distribution across the animal genome. Over the last 50 years, cattle breeds have been subjected to intense artificial selection. Consequently, regions controlling traits of economic importance are expected to exhibit selection signatures. The fixation index (Fst ) is an estimate of population differentiation, based on genetic polymorphism data, and it is calculated using the relationship between inbreeding and heterozygosity. In the present study, locally weighted scatterplot smoothing (LOWESS) regression and a control chart approach were used to investigate selection signatures in two cattle breeds with different production aptitudes (dairy and beef). Fst was calculated for 42 514 SNP marker loci distributed across the genome in 749 Italian Brown and 364 Piedmontese bulls. The statistical significance of Fst values was assessed using a control chart. The LOWESS technique was efficient in removing noise from the raw data and was able to highlight selection signatures in chromosomes known to harbour genes affecting dairy and beef traits. Examples include the peaks detected for BTA2 in the region where the myostatin gene is located and for BTA6 in the region harbouring the ABCG2 locus. Moreover, several loci not previously reported in cattle studies were detected.

The Melanocortin 1 receptor (MC1R) is one of the main genes implicated in the determination of the coat colour in mammals. This locus showed a relevant genetic variation between breeds that can be exploited for breed traceability of the animal productions. Modicana, Cinisara and Sardo-Modicana are three Italian endangered cattle breeds. Genetic characterisation by molecular markers is a fundamental prerequisite for managing genetic resources and for developing potential genetic traceability protocols. In order to improve the knowledge on Modicana, Cinisara and Sardo-Modicana breeds and to evaluate the possibility to develop DNA-based protocols for their mono-breeds products traceability, the genetic structure of MC1R gene was analysed. Four main alleles were observed in a representative sample of 162 animals. In the black coated Cinisara breed (n = 42), the ED and E+ alleles segregated with a frequency of 0.93 for ED allele. In the red coated Modicana (n = 60) and Sardo-Modicana (n = 60) breeds the E+ and E1 alleles segregated with frequencies of 0.42, 0.57 and 0.52, 0.47, respectively. The recessive allele e showed a low frequency (0.01) in both breeds. Sequencing a subsample of 34 animals the rare E2 allele was found only in Modicana and Sardo-Modicana at a good frequency (0.50). A new PCR-RFLP test, based on BstOI restriction endonuclease, was devised to assay for this allele. Results of the work indicate that red coat in Modicana and Sardo-Modicana cattle is genetically determined by the E+ and E1 alleles instead of the e allele at homozygote status, as occurs in other red European breeds. In these three Italian breeds of local importance, MC1R polymorphisms can be used to discriminate Cinisara from Modicana and Sardo-Modicana, but it was not able to distinguish between the two red coat populations.