Genetic / taxonomic issues for the Cat Specialist Group

[BorneanTiger]

Leopard (Panthera pardus)

The CSGhad grouped the Zanzibari (formerly Panthera pardus adersi, possibly extant) and mainland African leopards (including the Barbary leopard (formerly Panthera pardus panthera) of the Maghreb (Northwest Africa)) into one subspecies (Panthera pardus pardus), and said that the Sinai (Panthera pardus jarvisi) and Arabian leopards (Panthera pardus nimr) could be the same subspecies as African leopards, and all other leopards in Asia and European Russia (which has Caucasian leopards (Panthera pardus ciscaucasia / tulliana)), including the Sri Lankan leopard (Panthera parsus kotiya), could be one subspecies (Panthera pardus fusca), with the exception of the Javan leopard (Panthera pardus melas):

"Luo et al. (2014) published a further molecular study which included more samples from Indochina and the Malay Peninsula. The phylogeographical patterns are not clear for all putative subspecies. For example, P. p. kotiya is close to East Asian leopards based on mtDNA, but groups with P. p. fusca based on microsatellites (Uphyrkina et al. 2001). P. p. saxicolor also seems to group differently depending on the analysis used (Uphyrkina et al. 2001, Luo et al. 2014). Luo et al. 2014 show that P. p. fusca is diphyletic based on mtDNA, which was not found in previous studies. Khorozyan et al. (2006) analysed the skull morphometrics of southwest Asian leopards, and concluded that saxicolor and ciscaucasica were consubspecific, but retained tulliana and millardi as distinct. However, sample sizes were very small for some of these putative subspecies. Rozhnov et al. (2011) examined sequences of mtDNA (NADH5) and 11 microsatellites from southwest Asian leopards. They concluded that all were consubspecific from Afghanistan through Iran to the Caucasus, but no western Turkish specimens (tulliana) were analysed. Here japonensis is included in orientalis; there is no clear biogeographical barrier between these two forms which appear to form a cline in northeastern Asia. As the molecular differences between continental Asian leopards are very small compared to differences in Javan leopards (P. p. melas; Wilting et al. 2016), there could be a case for including all Asian subspecies, excluding melas, in a single Asian subspecies. These conflicting results from different studies suggest that more comprehensive sampling is required from throughout the range, taking advantage of museum specimens of known provenance. Until such a study is carried out, we propose the following conservative arrangement of subspecies:

Panthera pardus pardus (Linnaeus, 1758). Distribution: Africa. Comment: Although there are two principal mtDNA clades in Africa, they both occur in southern Africa and appear to be partly sympatric. Thus it would appear that no subspecies can be distinguished within Africa. However, more comprehensive sampling is needed.

Panthera pardus tulliana (Valenciennes, 1856; 1039), including ciscaucasica, saxicolor. Type locality: Ninfi, village situé à huit lieues est de Smyrne [near Izmir, Turkey]. Holotype: MNHN-ZM-MO-1849-20 mounted skin (skull inside). Distribution: Turkey, Caucasus, Turkmenistan, Uzbekistan, Iran, Iraq, Afghanistan and Pakistan. Comment: This is the earliest name for leopards from South West Asia, and hence includes saxicolor and ciscaucasica. If tulliana proves to be distinct from other southwest Asian leopards, ciscaucasica is the earliest available name.

Panthera pardus fusca (Meyer, 1794). Distribution: Indian subcontinent, Burma and China.

Panthera pardus kotiya (Deraniyagala, 1949). Distribution: Sri Lanka.

Panthera pardus delacouri (Pocock, 1930b). Distribution: SE Asia and probably southern China

Panthera pardus orientalis (Schlegel, 1857), including japonensis. Distribution: Eastern Asia from Russian Far East to China.

Panthera pardus melas (Cuvier, 1809; 152). Distribution: Java. Comment: Distinct ancient island form (Meijaard 2004, Gippoliti & Meijaard 2007, Uphyrkina et al. 2001, Wilting et al. 2016).

Panthera pardus nimr (Hemprich and Ehrenberg, 1832). Distribution: Arabian Peninsula. Comment: Distinctively small form, but may prove to be consubspecific with subspecies pardus, although should be retained as a separate management unit if so."

   
   


As for Central Chinese leopards, like at Wolong Reserve in Sichuan, where they may attack sub-adult pandas, their exact taxonomic status is unclear, being alternatively grouped under the P. p. japonensis, P. p. delacouri, or even P. p. fusca.

In 1993, Brakefield noted that just as there is a North Chinese leopard, there is also a South Chinese leopard, which was "much more golden yellow" in colour, had shorter fur, and which people thought might be of the Indian (Panthera pardus fusca) or Indochinese (Panthera pardus delacouri) subspecies, or a subspecies of its own (possibly Panthera pardus sinensis), and Uphyrikina et al. said "Teeth of ancient leopards found in southern China and dated from the Middle of Pleistocene were similar to the recent subspecies P. p. sinensis; this led to the hypothesis of local evolution in eastern and southeastern Asia (Hemmer 1976)."

Stuffed leopard for testing pandas at Wolong Nature Reserve, Central China, credit: Alamy 

*This image is copyright of its original author

Back to African leopards, though the CSG grouped them into 1 subspecies, they admitted that 2 principal mtDNA clades in Africa, particularly in southern Africa, and thus that they appeared to be partly sympatric, so though they couldn't distinguish between African leopards as different subspecies, "more comprehensive sampling" was needed. That's not the only issue facing African leopards. In another thread, @chui_ posted excerpts from the study by Anco et al. (2017):

"Historical mitochondrial diversity in African leopards (Panthera pardus) revealed by archival museum specimens

Abstract 

Once found throughout Africa and Eurasia, the leopard (Panthera pardus) was recently uplisted from Near Threatened to Vulnerable by the International Union for the Conservation of Nature (IUCN). Historically, more than 50% of the leopard's global range occurred in continental Africa, yet sampling from this part of the species' distribution is only sparsely represented in prior studies examining patterns of genetic variation at the continental or global level. Broad sampling to determine baseline patterns of genetic variation throughout the leopard's historical distribution is important, as these measures are currently used by the IUCN to direct conservation priorities and management plans. By including data from 182 historical museum specimens, faecal samples from ongoing field surveys, and published sequences representing sub-Saharan Africa, we identify previously unrecognized genetic diversity in African leopards. Our mtDNA data indicates high levels of divergence among regional populations and strongly differentiated lineages in West Africa on par with recent studies of other large vertebrates. We provide a reference benchmark of genetic diversity in African leopards against which future monitoring can be compared. These findings emphasize the utility of historical museum collections in understanding the processes that shape present biodiversity. Additionally, we suggest future research to clarify African leopard taxonomy and to differentiate between delineated units requiring monitoring or conservation action.

...

Leopards exhibited population structuring at large geographic scales (West, Central-East/Central-Southern, and Southern Africa), suggesting strong evidence against panmixia in this species. AMOVA and pairwise FST analyses support differentiation in the ND-5 locus spanning five major haplogroups: West Africa, Coastal West-Central Africa, Central- East-Africa, Central-Southern Africa, and Southern Africa. Distinction between CEA and CSA as two independent regional populations is supported by pairwise FST analyses (Figure 4). Although still high, FST[CEA-CSA]¼0.40, was the lowest among all African leopard population comparisons. CSA showed higher levels of differentiation from WA and CWCA leopard populations, than the latter two did to CEA, indicating that CSA leopards are reproducing in isolation from neighbouring populations (Figure 4). Furthermore, CSA exhibited the highest levels of differentiation when compared with the two selected Asiatic subspecies: FST[CSA-nimr]¼0.98 and FST[CSA-saxicolor]¼0.97 (Figure 4).

The African leopard harbours a greater degree of genetic diversity than previously indicated and is partitioned in a pattern providing strong support for significant genetic subdivision. Our pairwise FST analyses using mtDNA revealed leopard populations throughout sub-Saharan Africa retain highly divergent copies of the ND-5 locus on levels approaching, and in some instances exceeding, FST values observed between Asiatic populations (Arabian and Persian leopards) presently recognized by the IUCN as separate subspecies (Figure 4). AMOVA revealed population structuring indicating a lack of gene flow between larger geographic regions (West Africa, Central-East/Central-Southern Africa, and Southern Africa) and among all the populations within regions. Two populations, CEA and CSA showed decreased pairwise differences relative to other populations, which could be an artifact of decreased sampling. Lastly, the star-like phylogeny, widespread distribution, and connectedness of the H10 haplotype points to a likely origin of diversity for the ancestral haplotype of this locus in Central and East Africa. We caution this work may not fully express the degree of genetic diversity present in African leopards, especially given sampling deficiencies in North Africa, parts of West Africa, and in Northeastern Africa.

This study has raised important questions regarding the taxonomic status of leopards in Africa. First, these findings support a distinction between African populations and Arabian and Persian leopard populations. We found additional strong support for an East-West split in African leopards, which may correspond to previously hypothesized taxonomic groupings (Figure 1, Table 1) and is congruent with numerous recent phylogeographic analyses of widespread African taxa (Moodley & Bruford 2007; Lorenzen et al. 2012; Dobigny et al. 2013; Smitz et al. 2013; Bertola et al. 2016; Fennessy et al. 2016). More sampling is needed to accurately delineate geographic features acting as potential barriers to gene flow (e.g. Sanaga River in Central Cameroon), while a suture zone has been identified between CWCA and CEA populations (Figures 2 and 3). In addition, we have identified previously unrecognized levels of genetic diversity in historical collections of African leopards not represented in contemporary leopard populations. While only based on mtDNA, the reconstruction of a haplotype network using novel samples of African leopards has reopened a >15-year-old conversation regarding African leopard diversity and taxonomy. We acknowledge that our results are limited by the use of mtDNA, and consequently single locus data. We therefore, strongly recommend multilocus sampling to investigate whether African leopards exhibit evidence of discordance between mitochondrial and nuclear markers (Toews & Brelsford 2012). These findings will provide the foundation for our ongoing analysis of temporal changes in phylogeographic patterns using sequence capture from historical collections, which will contribute to management and planning strategies to conserve remaining genetic diversity in the African leopard."

Image of the Atlas lion, brown bear and leopard from the Maghreb (Northwest Africa), by Joseph J. Ortega:

*This image is copyright of its original author

Video:

Reply