Managing genetic diversity in breeding programs of small populations: the case of French local chicken breeds

Abstract

Background: On-going climate change will drastically modify agriculture in the future, with a need for more sustainable systems, in particular regarding animal production. In this context, genetic diversity is a key factor for adaptation to new conditions: local breeds likely harbor unique adaptive features and represent a key component of diversity to reach resilience. However, local breeds often suffer from small population sizes, which puts these valuable resources at risk of extinction. In chickens, population management programs were initiated a few decades ago in France, relying on a particular niche market that aims at promoting and protecting local breeds. We conducted a unique comprehensive study of 22 French local breeds, along with four commercial lines, to evaluate their genetic conservation status and the efficiency of the population management programs.

Results: Using a 57K single nucleotide polymorphism (SNP) chip, we demonstrated that both the between- and within-breed genetic diversity levels are high in the French local chicken populations. Diversity is mainly structured according to the breeds' selection and history. Nevertheless, we observed a prominent sub-structuring of breeds according to farmers' practices in terms of exchange, leading to more or less isolated flocks. By analysing demographic parameters and molecular information, we showed that consistent management programs are efficient in conserving genetic diversity, since breeds that integrated such programs earlier had older inbreeding.

Conclusions: Management programs of French local chicken breeds have maintained their genetic diversity at a good level. We recommend that future programs sample as many individuals as possible, with emphasis on both males and females from the start, and focus on a quick and strong increase of population size while conserving as many families as possible. We also stress the usefulness of molecular tools to monitor small populations for which pedigrees are not always available. Finally, the breed appears to be an appropriate operational unit for the conservation of genetic diversity, even for local breeds, for which varieties, if present, could also be taken into account.

Fig. 1

Partial correlation matrix between current population diversity indices and management program features for breeds of Group 1. Values correspond to partial Pearson’s correlation coefficients. Each correlation was corrected by management program features that are not involved in the correlation, except for correlations involving ${\mathit{Ne}}_{\mathit{founders}}$, which were only corrected for the number of generations. Colored cells stand for significant correlations (p < 0.05), either positive (blue) or negative (red); color intensity reflects the strength of the correlation; white cells represent non-significant correlations. “Nb.Generation” stands for the number of generations since the start of the management program and “Fixed” is the proportion of fixed alleles

Fig. 2

Genetic diversity summary statistics per population. From the top to the bottom, Fit, Fis, observed heterozygosity (Ho), expected heterozygosity (He), minor allele frequency (MAF), and proportion of fixed alleles (Fixed). Populations were sorted by group (Group 1 in orange, Group 2 in yellow, and Group 3 in green) and by Fit within groups. Error bars correspond to standard errors

Fig. 3

Runs of homozygosity (ROH) summary statistics per population. From top to bottom, inbreeding coefficient (F-ROH), mean number of ROH per individual, and mean length of each ROH. Populations were sorted by group (Group 1 in orange, Group 2 in yellow, and Group 3 in green) and by F-ROH within groups. Error bars correspond to standard errors

Fig. 4

Unrooted neighbor joining tree of individuals with respect to populations. Dotted lines represent grouping features: brown egg laying breeds (blue), broiler breeds (red), and Bresse breeds (green)

Fig. 5

Mean center of origin of populations and cluster assignment of the regions. The color of each French region (“department”) corresponds to the mean assignment of their chicken populations to each of the two clusters of the DAPC, red for Asian and blue for European

Fig. 6

Neighbornet tree of the chicken populations. The color of each diamond corresponds to the population cluster affiliation, red for Asian and blue for European

Fig. 7

Unrooted neighbor-joining tree of the Marans breed. Colors of the surrounding circles corresponds to the breeders (inner circle) and the phenotypes (outer circle)