They have the advantage of being often dual-purpose, i.e. they produce eggs and meat, thus making it possible to valorize both sexes, contrary to specialized "commercial" breeds (laying hens or broilers), and to consider production channels that better meet ethical expectations. However, these breeds are often represented by small numbers, which makes them particularly vulnerable to drift and inbreeding. Their management (reproduction, generation renewal, ...) is not always coordinated and sometimes relies on amateur breeders for whom a follow-up of the pedigree is complicated or even impossible.

In the framework of the BioDivA project conducted in collaboration with ITAVI and SYSAAF, we studied the genetic diversity and structure of 22 local French chicken breeds using both genotyping (57K SNPs) and pedigrees when available. We demonstrated that inter- and intra-breed genetic diversity levels are high in French populations with strong genetic differentiation between them (measure of differentiation, mean Fst=24%) and high effective genetic numbers (i.e. indicators of genetic diversity), Ne, considering the real numbers of breeds (Ne=[22; 285] individuals). The diversity is mainly structured according to the selection objective (meat, egg-laying, mixed) and to the history of the breeds (constitution, crossing...). Nevertheless, we have observed an important substructuring within the breeds according to the practices of the breeders in terms of exchanges and selection of animals, in particular among the amateur breeders (clubs, groups, constitution of varieties within the breeds), leading to more or less genetically isolated flocks. By analyzing demographic parameters and molecular information, we have shown that current management/conservation programs, in which pedigrees and matings are monitored, are effective in maintaining genetic diversity. Indeed, breeds that have initiated such programs earlier have older inbreeding and higher overall genetic diversity. For example, the effective population size, Ne, is greater than 100 individuals for breeds that initiated a management program in the 1990s while Ne is less than 100 individuals when initiated in the 2000s.

We recommend that future breed management (conservation/breeding) programs sample as many individuals, males and females, as possible (relative to breed numbers) at inception, and focus on increasing population size rapidly and significantly, while maintaining as many families as possible. We also emphasize the usefulness of molecular tools for monitoring and managing small populations for which pedigrees are not always available. Finally, the breed seems to be an appropriate operational unit for the conservation of genetic diversity, even for local breeds. Indeed, although one might expect blurrier boundaries between local breeds than between commercial breeds/lines, especially unmanaged ones, we have shown that they are all perfectly distinct genetically. However, also adding consideration of intra-breed varieties could improve conservation programs by better sampling the available diversity (see Figure illustrating the case of the Marans breed).