Abstract: The use of SNPs in combination with Bayesian statistics for the geographic traceability of cattle was evaluated
using a dataset comprising 24 breeds from Italy, France, Spain, Denmark, the Netherlands, Switzerland
and UK genotyped with 90 polymorphic markers.
The percentage of correct assignment of the individuals to their Country of origin was 90%, with an average
assignment probability of 93% and an average specificity of 92%. The higher value was observed for
UK breeds (97% of correct assignment) while Swiss animals were the most difficult to allocate (77% of correct
assignment).
Tracing of Protected Geographic Indication (PGI) products, the approach correctly assigned 100% of Guaranteed
Pure Highland Beef; 97% of ââVitellone dellâAppennino Centraleâ breeds; 84% of Ternera de Navarra,
and 80% of Boeuf de Chalosse.
Methods to verify Products of Designated Origin (PDO) and Protected Geographic Indication (PGI) products
will help to protect regional foods and promote the economic growth of marginal rural areas by
encouraging the production of high quality niche market foods.
Abstract: The effectiveness of single nucleotide polymorphisms (SNPs) for the assignment of cattle to
their source breeds was investigated by analysing a panel of 90 SNPs assayed on 24
European breeds. Breed assignment was performed by comparing the Bayesian and frequentist
methods implemented in the STRUCTURE 2.2 and GENECLASS 2 software programs. The use
of SNPs for the reallocation of known individuals to their breeds of origin and the assignment
of unknown individuals was tested. In the reallocation tests, the methods implemented
in STRUCTURE 2.2 performed better than those in GENECLASS 2, with 96% vs. 85% correct
assignments respectively. In contrast, the methods implemented in GENECLASS 2 showed a
greater correct assignment rate in allocating animals treated as unknowns to a reference
dataset (62% vs. 51% and 80% vs. 65% in field tests 1 and 2 respectively). These results
demonstrate that SNPs are suitable for the assignment of individuals to reference breeds.
The results also indicate that STRUCTURE 2.2 and GENECLASS 2 can be complementary tools to
assess breed integrity and assignment. Our findings also stress the importance of a highquality
reference dataset in allocation studies.
