There is a high degree of coat colour variation between geographic populations of leopards (Panthera pardus L., Carnivora: Felidae) (Kingdon et al. 2013). Individuals from arid regions are generally pale with dispersed and open-centred rosettes, in contrast to those residing in forests which are darker with clustered and small-centred rosettes. These patterns are thought to correspond with differing vegetation types and light levels in order to conceal the animal from prey and possibly other predators (Allen et al. 2010; Kingdon et al. 2013). This adaptive explanation is supported by the frequent occurrence of melanistic leopards in humid habitats such as the Malayan peninsula (Kawanishi et al. 2010; Schneider et al. 2012). The frequency of ‘black panthers’ dramatically decreases across more arid regions (Kawanishi et al. 2010). The release of eumelanin (black pigmentation) into mammalian pelage is known to be regulated by the extension gene and phaeomelanin (yellow-red pigmentation) by the agouti gene (Fontanesi et al. 2009). Mutations to either of these genes can produce melanism in felids; however, it is a mutation in the agouti gene which results in melanism in leopards which is inherited as a recessive trait (Schneider et al. 2012).

In contrast, extreme pale (albino) colour morphs, which lack any pigmentation, or erythrism, which contain red pigmentation instead of black, are rarely documented in wild leopards (Divyabhanusinh 1993; Hartwell 2015; Sunquist & Sunquist 2014). Although the cause of erythrism in large felines is unknown, Peterschmitt et al. (2009) found evidence for a recessive mutation in the extension gene which produces more phaeomelanin, resulting in an amber colour in the domestic Norwegian Forest Cat (Felis catus). Similar mutations may also be responsible for the red colouration seen in dogs, humans, and other mammals (Fontanesi et al. 2009; Majerus & Mundy 2003).

Reports of erythristic leopards (also informally known as strawberry or red leopards, or pink panthers; Anonymous 2013, 2014a, 2015; Dell’Amore 2012) are exceptionally rare. A detailed search of the literature found only one paper (Divyabhanusinh 1993) which reported that five pale leopards with light brown spots (one male, one female, and the rest undetermined) had been shot in India between 1905 and 1965. To our knowledge, no other records of wild erythristic leopards were documented until 2012 when a male was photographed by a guide at the Madikwe Game Reserve in the North West Province of South Africa (Figure 1). This was subsequently reported in the popular press (Dell’Amore 2012). Here, we report new sightings from Mpumalanga and the results of a survey of managers and section rangers of National Parks, wildlife reserves, and conservation organisations in South Africa, supplemented by press reports and social media, to understand the possible distribution and abundance of this leopard colour morph.

From the Internet search and local reports, five individual erythristic leopards, identified through pelage patterns have been captured on camera trap, killed, or caught, in the Lydenburg area, Mpumalanga; and two animals in Madikwe Game Reserve and the surrounding area, North West Province (Figure 1, Table 1). Of the 28 individual leopards recorded at TTWR during this 3-year study, two (7.14%) were erythristic (Figures 2, 3a and b); one of these was born to a normal coloured female (Figure 3a and b).

Of the 25 individuals approached we received replies from 19 managers, section rangers, and researchers from reserves and organisations from across South Africa. The skin of one other erythristic animal (Table 1: animal 1) was provided as evidence. No other responses were received from the social media call for information on strawberry leopard sightings.

To our knowledge, only one previous paper has reported the presence of erythristic leopards (in India; Divyabhanusinh 1993). Here, we provide the first formal report of the presence of wild erythristic leopards outside of India. In total, there are seven records of wild erythristic leopards in South Africa, five in the Lydenburg area.

South Africa’s first erythristic leopard report in 2012 was recorded in the North West Province, some 400 km from our Mpumalanga study site. Although such distances are likely to preclude dispersal of offspring of any given individual, it has been suggested that these widely separated leopard populations may be considered part of a single core population (Friedmann & Traylor-Holzer 2008).

General colour resemblance (where an animal resembles the general colour of their environment) may be the reason for the higher frequency of melanistic leopards in moist habitats sporting thick vegetation (Allen et al. 2010), but this is unlikely to provide an explanation for the presence of the erythristic forms recorded here, as this region does not exhibit a prolonged dry season and leopards in savannah habitats are thought to be predominantly nocturnal hunters (Bailey 1993), hence a pale pelage would not be beneficial.

It is worth considering other explanations for the recent sightings of erythristic leopards in Mpumalanga, and we posit three hypotheses. Firstly, this is simply a reflection of reporting bias. The area covered by the camera traps is limited and although our reports are unlikely to present a precise density of this colour morph, the numbers of observers and sightings consulted though social media from large National Parks to smallholdings across South Africa are a good indication of its rarity. Therefore we feel that reporting bias is unlikely (although because of the secretive nature of leopards, it is possible that unusual behaviours or forms are overlooked, e.g. Pirie et al. 2014). Secondly, and highly speculatively, this may reflect leopards released or escaping from captive breeding programmes, where animals are reared for trophy hunting. Some nine game ranches in South Africa breed leopard (Lindsey et al. 2011) and the captive breeding of colour morphs of other species for hunting, such as lion (Panthera leo), is known to occur (Crowley 2015). Indeed, there is a record of a captive bred male strawberry leopard born to parents which came from the same area as the wild individual seen in 2012 (Anonymous 2014a). This is possible, but unsubstantiated. Thirdly, population fragmentation and isolation and therefore a highly reduced effective population size, may have resulted in the expression of a de novo or previously rare allele at higher frequencies. Such inbreeding effects are thought to result in the presence of the very rare king cheetah (Acinonyx jubatus) and white lion colour morphs in southern Africa (Sunquist & Sunquist 2014); it has also been documented with leopards in the Malaysian peninsula, where the rapid near fixation of melanism occurred following population fragmentation (Hedges et al. 2015; Kawanishi et al. 2010). Similarly, Haag et al. (2010) reported genetic drift within small fragmented jaguar populations over a relatively short time frame and McManus et al. (2015) showed that leopard populations can become isolated within a few generations. Perhaps this is the most reasonable explanation for observing the erythristic morph in relatively high numbers in a single area; however we can only speculate, and future research is required to test this explanation.

The geographical range of the leopard has diminished by an estimated 37%, which underpinned the need to reclassify the leopard on the International Union for Conservation of Nature (IUCN) (Red data list from least concern to near-threatened in 2008 (Balme, Slotow & Hunter 2010). Within South Africa, the destruction of suitable leopard habitat has produced highly fragmented areas with depleted prey densities, (Chase-Grey 2011; Swanepoel et al. 2013) which combined with persecution (Lindsey et al. 2011) has substantially reduced leopard numbers and caused populations to become isolated (Friedmann & Traylor-Holzer 2008). Until January 2016, 150 CITES (Convention for the International Trade in Endangered Species) trophy animal permits were allocated annually to South Africa (Balme et al. 2010; Lindsey et al. 2011). Zero permits have been allocated for 2016; however, this is temporary and will be reassessed for 2017 (Anonymous 2016). Documenting this rare leopard colour morph could result in negative outcomes (e.g. encouraging illegal capture for breeding); however, the expression of erythrism could be a visible indicator which highlights the consequences of population fragmentation and decline.

We collated seven records of erythristic leopards in South Africa. Two of 28 individuals recorded at our study site (Thabo Tholo Wilderness Reserve; Mpumalanga) were of this form, and we found five other records from Mpumalanga and North West provinces. The majority of records were recent, and the oldest was from the late 1990s we were unable to find earlier records of erythristic leopards from South Africa. Although speculative, we suggest that this may reflect the consequences of increased population fragmentation and consequent inbreeding. Further research is needed in order to ascertain whether erythristic leopards are a visible sign of increasing threats to the viability of local leopard populations.