Genetic diversity in global populations of the critically endangered addax (Addax nasomaculatus) and its implications for conservation

Abstract

Threatened species are frequently patchily distributed across small wild populations, ex situ populations managed with varying levels of intensity and reintroduced populations. Best practice advocates for integrated management across in situ and ex situ populations. Wild addax (Addax nasomaculatus) now number fewer than 100 individuals, yet 1000 of addax remain in ex situ populations, which can provide addax for reintroductions, as has been the case in Tunisia since the mid-1980s. However, integrated management requires genetic data to ascertain the relationships between wild and ex situ populations that have incomplete knowledge of founder origins, management histories, and pedigrees. We undertook a global assessment of genetic diversity across wild, ex situ and reintroduced populations in Tunisia to assist conservation planning for this Critically Endangered species. We show that the remnant wild populations retain more mitochondrial haplotypes that are more diverse than the entirety of the ex situ populations across Europe, North America and the United Arab Emirates, and the reintroduced Tunisian population. Additionally, 1704 SNPs revealed that whilst population structure within the ex situ population is minimal, each population carries unique diversity. Finally, we show that careful selection of founders and subsequent genetic management is vital to ensure genetic diversity is provided to, and minimize drift and inbreeding within reintroductions. Our results highlight a vital need to conserve the last remaining wild addax population, and we provide a genetic foundation for determining integrated conservation strategies to prevent extinction and optimize future reintroductions.

Keywords: Addax nasomaculatus; Sahelo‐Saharan antelope; captive populations; conservation genetics; reintroduction; ungulate conservation.

FIGURE 1

Historical range of addax (a) showing the historic addax range with locations of the remaining population in the Tin Toumma region, Niger and free‐ranging and semiwild reintroduced populations. Inset (b) illustrates the reintroduction history of addax in Tunisia, indicating numbers of males (M) and females (F) translocated from each source population.

FIGURE 2

Network of addax mtDNA control region haplotypes. Circles represent haplotypes, with most likely evolutionary relationships indicated by solid black lines and alternative likely relationships shown by dashed grey lines. Mutational steps are shown by hashes. The 76‐bp indel was recoded as a single base pair 5th state and is indicated by *; all other indels were excluded. Haplotype circles are coloured according to populations as shown, with circle size indicating sample size (log scale) as indicated. Haplotype nomenclature follows Table S1. Wild (pre‐1930) haplotypes are from Hempel et al. (2021).

FIGURE 3

Pairwise F _ST estimates for the five primary ex situ populations (a) and for the three Tunisian populations (b). All pairwise F _ST estimates were significant (as indicated by *).

FIGURE 4

Visualization of population structure in the ex situ and reintroduced addax populations using two methods. (a) PCA using 1704 SNPs in the managed addax populations, showing PC1 and PC2 (percentage of variation explained shown in brackets). The Tunisian metapopulations are represented by different shapes, as shown in the legend. Inset shows the first 20 eigenvalues. (b) Admixture results using 1073 SNPs filtered to minimize linkage disequilibrium, showing, for each individual, the proportion of genetic membership to each ancestry cluster for the most informative number of clusters (K = 2 to K = 6). K = 5 was the best supported model.

FIGURE 5

Genetic diversity measures in the ex situ and reintroduced populations. (a) Mean observed heterozygosity (black points) and expected heterozygosity (grey points) for each population, with 95% confidence limits (black and grey bars, respectively). (b) Mean F _IS (coloured bars) within each population and 95% confidence limits (black error bars). (c) Mean allelic richness (Ar) (points) standardized to N = 14, with 95% confidence limits (black error bars), with the global average indicated by the horizontal dashed line. (d) Individual standardized multilocus heterozygosity (sMLH). (e) Private allelic richness (pAr) standardized to N = 14. In all cases, estimates for Tunisia summarize all three national parks, which are also shown independently (shaded in grey).