Une base de données phénotypiques : un prérequis pour la mise en place de programmes de sélection des races locales

51es Journées de la Recherche Porcine, 5 et 6 février 2019, Paris, p. 243-244, par Marie-José Mercat et al., poster


Peu de races locales porcines ont des programmes de sélection. En plus des généalogies déjà disponibles, mettre en place de tels programmes exige des phénotypes pour définir des objectifs de sélection puis estimer les valeurs génétiques des animaux. Dans le cadre du programme européen TREASURE, une base de données et un site web ont été développés afin de promouvoir la collecte et l’enregistrement de phénotypes et de mieux caractériser les races, viandes et produits.

An online phenotype database: a prerequisite for breeding programs in local pig breeds

Currently, breeding programs exist only for few local pig breeds in Europe. For some of them, no or very few phenotypes are recorded. To promote phenotyping, a dedicated database and a website were developed as part of the TREASURE H2020 project. This is the first step towards implementing breeding programs in European local pig breeds, based on standardised recording of growth, carcass and meat quality traits. First, the required variables were collected for six local breeds: Basque (FR), Bísara (PT), Crna slavonska (HR), Gascon (FR), Krškopolje (SI) and Schwäbisch-Hällisches (DE). In total, 74 variables were identified representing animal herdbook information (10), rearing and growth (22), carcass (22) and meat quality (20) attributes. The database is compatible with the various identifiers (IDs) used in the different countries: animal IDs, breed, farm, etc. codifications. Great attention was paid to describing methods for measuring traits. Thus, each carcass and meat quality phenotype is associated with a method description representing 35 additional variables. The website can be easily translated into several languages: English, Croatian, French, German and Slovenian languages are already available, and other languages can be easily included. Stakeholders of all the breeds studied in TREASURE are free to use these tools. Data can be hosted in a common database or in duplicate partner-specific databases. Once enough data have been recorded, genetic parameters (heritabilities, genetic correlations between traits) can be estimated and selection objectives defined for each local breed.