The cheetah (Acinonyx jubatus) is one of the most remarkable yet unusual members of the family Felidae. Not only does its mostly diurnal lifestyle deviate from most other wild felids, but it is also the fastest carnivore and reaches the highest speed of any land animal of up to 105kmh (Sharp 1997). Due to its uniqueness, the species inhabits a distinct niche within African and Asian ecosystems, with distinctive subspecies occurring in different habitats (Mills et al., 2004). Adapted to high-speed running, their main source of prey consists of small to medium-sized antelopes that are captured by short-distance sprints (Mills et al., 2004). Hunting smaller prey than lions, leopards, and hyenas, cheetahs avoid interactions with the larger carnivores, one of their major natural threats (Durant, 2000; Hunter et al., 2007). Historically, cheetahs inhabited most parts of non-tropical Africa, the Arabian Peninsula, and large parts of Southwest Asia (Durant et al., 2017). Presently, their distribution consists only of a small fraction of the species' former range, and most populations are isolated from each other (Fig. 2a; Durant et al., 2017). The International Union for Conservation of Nature (IUCN) Cat Specialist Group currently recognizes four cheetah subspecies: the Southern African cheetah (Acinonyx jubatus jubatus); the Asiatic cheetah (A. j. venaticus); the Northeast African cheetah (A. j. soemmeringii) and the Northwest African cheetah (A. j. hecki) (Kitchener et al., 2017). Traditionally the cheetah was divided into five subspecies (Krausman & Morales, 2005). However, the East African cheetah (A. j. raineyi) has been recently subsumed into the Southern African cheetah (A. j. jubatus) due to a non-monophyletic clustering on the mtDNA level (Charruau et al., 2011; Kitchener et al., 2017), a change since challenged (Prost et al., 2022). Here, we follow the classical five subspecies concept from Krausman and Morales (2005) while acknowledging that A. j. raineyi is currently not accepted as a fifth subspecies by the IUCN. The species as a whole is listed as "vulnerable" by the IUCN, with approximately 7,100 adult and adolescent individuals left in the wild (Durant et al., 2017), showing a drastic reduction in population size compared to the estimated 15,000 individuals 45 years ago (Myers & Resources, 1975). The subspecies A. j. venaticus and A. j. hecki are already listed as "critically endangered" (Durant et al., 2017), and, with fewer than 50 individuals, A. j. venaticus is likely to face extinction in the near future (Durant et al., 2017; Farhadinia et al., 2016, 2017; Khalatbari et al., 2018). Thus, while the species as a whole is likely to survive, current conservation efforts, e.g., reintroduction of cheetahs into former range countries, should take subspecies assignments into account as the loss of certain subspecies might irreversibly alter the corresponding ecosystems (Bertola et al., 2022; Liu et al., 2018).
Multi-factorial causes affect cheetah population declines and can be divided into manmade causes such as habitat loss, illegal wildlife trade (Klaassen & Broekhuis, 2018; Marker & Dickman, 2004), climate change (Khalatbari et al, 2018), and non-manmade causes, mainly related to inbreeding (Merola, 1994; O'Brien et al., 1985, 1986). In recent years, illegal trafficking especially for the exotic pet trade has become a major threat to cheetahs (Tricorache et al., 2018). This issue is further intensified by the political instability in many of the cheetah's range countries (Brito et al., 2018; Jacobs & Schloeder, 2001). As the species is listed in Appendix I of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES), the legal trade of captive-bred cheetahs is strongly regulated, and any trade with wild-caught individuals is banned with few exceptions (https://cites.org/eng/app/appendices.php). In most countries within the species' range, the cheetah is protected by law and different conservation plans act to ensure legislation (Balint et al., 1997; Nowell & Jackson, 1996). Cheetah trade is of high economic value, which creates a black market for illegal trading and organized poaching of live cheetah cubs (Magliolo et al., 2021). The main customers of exotic pets are primarily found within the Arabian Peninsula where a pet cheetah counts as a status symbol. In many Arabic cultures, hunting with captive cheetahs has a long tradition, and the pet trade might be a relic of those habits (Spee et al., 2019). Poaching of live cheetahs occurs in different intensities throughout their range and therefore affects the subspecies differently. With A. j. raineyi and A. j. soemmeringii close to one of the major illegal trading hubs, the Horn of Africa, poaching is an increasing threat, especially in countries such as Ethiopia, Somalia, and Kenya (Tricorache et al., 2021). To monitor illegal trade and to identify poaching hotspots, identifying cheetah subspecies and reliably detecting each individual's geographical origin is crucial. In the case of East Africa, delineating the subspecies could reveal whether an individual was poached in Ethiopia, South Sudan, Northern Somalia, or in Kenya, Tanzania, Uganda, or Southern Somalia, as these are the respective distributions of the two subspecies, A. j. soemmeringii and A. j. raineyi.
It is challenging to distinguish cheetah subspecies morphologically, and they are often assigned to subspecies-level solely based on their geographic origin. However, cheetahs have been recorded to move > 1000 km (Durant et al., 1988; Farhadinia et al., 2016). Genetics has shown promise to effectively identify cheetah subspecies (Charruau et al., 2011; Schmidt-Küntzel et al., 2018; Prost et al., 2022). In general, it has been shown that mitochondrial DNA (mtDNA) barcodes can reliably discriminate between species (Tavares & Baker, 2008) and even subspecies (Soares et al., 2019; Gaber et al., 2020). However, several conflicting signals regarding mitochondrial DNA-based subspecies identification in cheetahs have been identified. For example, A. j. raineyi and A. j. jubatus have recently been merged into a single subspecies, namely A. j. jubatus, based on mtDNA data (Kitchener et al., 2017). Yet, this finding does not fit recent evidence obtained from genome-wide SNP data (Prost et al., 2022), raising questions about the validity of mtDNA markers for subspecies assignment in cheetahs. In their study, Prost et al. (2022) suggest incomplete lineage sorting (ILS) or mitochondrial capture (MC) as possible reasons for the presence of two mitochondrial haplogroups in East African cheetahs, one of which is shared with the Southern African subspecies, A. j. jubatus. ILS is the retention of two or more alleles originating from an ancestral population in one or two populations after divergence/ speciation (Hahn, 2019). MC can be caused by mitochondrial introgression from one population into another, if the introgressed mitochondrial genome fixates in the receiving population (Allendorf et al., 2022).
To obtain better insights into the mitochondrial phylogeography of cheetahs and to investigate potential polymorphisms for subspecies identification, we used 15 mitochondrial SNPs and a 3-bp deletion in the NADH dehydrogenase subunit 5 (ND5; Charruau et al., 2011 and Prost et al., 2022), amplified in five short mtDNA mini-barcodes (cheetah subspecies specific amplicons; CSAs), to infer the presence of mitochondrial haplotypes/haplogroups throughout the cheetah's current and former range. The CSAs, not exceeding 200 bp in length, were chosen to enable amplification even in highly degraded samples e.g., museum or confiscated material. Three of these amplicons have previously been used in Prost et al. 2022. Using an extended CSA and sample set we show that differences exist across mtDNA haplogroups, but that especially in East Africa, subspecies assignment using only mtDNA is confounded by the presence of two mitochondrial haplogroups. Furthermore, as the Horn of Africa is a poaching hotspot, we designed an easy-to-use PCR approach (Amplification-Refractory Mutation System; ARMS) for the identification of the A. j. soemmeringii haplogroup, which does not require sequencing of the amplicons.