We live in a world characterised by biodiversity loss and global environmental change. The extinction of large carnivores can have ramifying effects on ecosystems like an uncontrolled increase in wild herbivores, which in turn can have knock-on impacts on vegetation regeneration and communities. Cheetahs (Acinonyx jubatus) serve important ecosystem functions as apex predators; yet, they are quickly heading towards an uncertain future. Threatened by habitat loss, human-wildlife conflict, and illegal trafficking, there are only approximately 7,100 individuals remaining in nature. We present the most comprehensive genome-wide analysis of cheetah phylogeography and conservation genomics to date, assembling samples from nearly the entire current and past species' range. We show that their phylogeography is more complex than previously thought, and that East African cheetahs (A. j. raineyi) are genetically distinct from Southern African individuals (A. j. jubatus), warranting their recognition as a distinct subspecies. We found strong genetic differentiation between all classically recognised subspecies, thus refuting earlier findings that cheetahs show only little differentiation. The strongest differentiation was observed between the Asiatic and all the African subspecies. We detected high inbreeding in the Critically Endangered Iranian (A. j. venaticus) and North-western (A. j. hecki) subspecies, and show that overall cheetahs, along with snow leopards, have the lowest genome-wide heterozygosity of all the big cats. This further emphasizes the cheetah’s perilous conservation status. Our results provide novel and important information on cheetah phylogeography that can support evidence-based conservation policy decisions to help protect this species. This is especially relevant in light of ongoing and proposed translocations across subspecies boundaries, and the increasing threats of illegal trafficking.

Dataset 2 – Mitochondrial DNA: We amplified mtDNA regions of 929bp for 56 individuals and 681bp for 57 individuals. We targeted two mitochondrial genes that included the 14 previously described diagnostic SNPs from Charruau et al. (2011), the NADH-dehydrogenase subunit 5 (MT-ND5) and the control region (MT-CR). The amplicons were Sanger sequenced. We further included the 78 individuals from Charruau et al. (2011) in the 681bp dataset. All individuals of the 929bp data set are included in the 681bp set, as the 929bp fragment includes the shorter 681bp fragment. The mtDNA data were used to infer population structure and included samples from their present and past distributions. We followed the protocol of Rohland, Siedel & Hofreiter (2010) for the DNA extraction from museum samples. To avoid DNA contamination, we carried out all extractions in a dedicated laboratory for museum samples at the Vetmeduni in Vienna, Austria.

Dataset 3 – mini-barcodes: We designed a mini-barcode approach to investigate whether all A. j. soemmeringii carry the 3bp deletion in the MT-ND5 gene as described in Charruau et al. (2011), and as a quick subspecies assignment test. We designed three mini-amplicons that amplify a total of 190bp, based on diagnostic sites inferred from our mitochondrial haplotype data (dataset 2). Primer sequences can be found in Supplementary Table S5. We used the extracts of all A. j. soemmeringii specimens from dataset 1 and sequenced the amplicons on an Illumina iSeq 100 (2x150bp) following the two-step sequencing protocol of Lange et al. (2014).

Dataset 4 – immune response genes: We sequenced the MHC class II DRB exon 2 of 46 individuals (belonging to four of the five classical subspecies; excluding A. j. hecki) to investigate their immunogenetic diversity. We used the Qiagen DNeasy Blood and Tissue kit for DNA extraction from hair and tissue samples, and the VWR PeqGold™ Tissue DNA Mini Kit Plus for blood samples. We carried out PCR amplifications of the target region as described in Castro-Prieto, Wachter & Sommer (2011). Indexing, multiplexing, and sequencing were carried out following the Illumina Nextera XT DNA Library Prep kit reference guide. Sequencing was performed on an Illumina MiSeq (2x250bp).