The mystery of the Arabian miniatures

The Arabian Peninsula is bigger than western Europe, and ecologically similar to North Africa. However, an unexplained fact is that most of its terrestrial animals, of body mass 1 kilogram or more, are diminutive.

Not only are the wild ungulates and the major Carnivora of the Arabian peninsula all smaller-bodied than conspecifics (or their closest ecological counterparts) in Africa or mainland Asia, but five of the six wild species of ungulates, and both of the top carnivores, are smaller-bodied than the key deer ( - an extremely diminutive form restricted to small islands off Florida.

The body mass attained by the genus Lepus ( on the Arabian Peninsula is only a quarter of that attained by the same genus in semi-arid, subtropical North America.

The pattern applies also to the extinct Arabian form of the ostrich (Struthio camelus syriacus,, which was smaller-bodied than the various subspecies in Africa.

This means that

  • there has been no wild species of herbivore on the Arabian Peninsula - apart from the mysterious ancestor of a camel - with adult female body mass averaging more than 65 kg, and
  • both of the top predators of the Arabian Peninsula are only about the size of the Eurasian lynx (Lynx lynx).

The lion (Panthera leo) seems to have been naturally absent from the Arabian Peninsula ( Far from compensating for this absence by increased body size, the Arabian leopard (Panthera pardus nimr) is the smallest-bodied subspecies: adult females about 20 kg. The Arabian wolf (Canis lupus arabs) likewise weighs only about 20 kg. The lynx-like felid Caracal caracal, too, is relatively small-bodied on the Arabian Peninsula.

The following illustrates the various mammalian species.


Gazella saudiya is extinct, with no photos of the living animal ( It was the smallest of all gazelles, with adult female body mass probably less than 13 kg.

Gazella arabica has adult female body mass probably about 15 kg. See and and and and

Gazella marica has adult female body mass probably about 18 kg. See and and and

Arabitragus jayakari has adult female body mass about 20 kg ( See and and and and and

Capra nubiana has adult female body mass about 23.5 kg (, compared to 56 kg in Capra cylindricornis ( The Arabian species is also smaller-bodied than Capra caucasica ( and the Ethiopian species Capra walie ( See and and and However, Capra nubiana is weakened as a example of the trend towards miniaturisation because it occurs both west and east of the Red Sea.

Oryx leucoryx (which narrowly escaped extinction) has adult female body mass about 63 kg (,%3D%206%2C%20range%2048%E2%80%9390). See and and and


Camelus dromedarius, the ancestral dromedary, does not conform to the trend of miniaturisation on the Arabian Peninsula. According to the wild form last occurred in what is today Dubai. If so, it seems that this ungulate, which is large-bodied for an artiodactyl even by mainland standards, survived in Arabia partly because of the island-like freedom from predation by the lion and the spotted hyena (Crocuta crocuta). Even if the population of the wild dromedary in Arabia was merely relictual, it would depart from the pattern because C. dromedarius, with body mass at least 400 kg, is not small-bodied among the camelids of the Holocene, worldwide.


Equus hemionus hemippus (, a subspecies of the Asian wild ass, was associated mainly with the Levant and Mesopotamia. However, it is thought also to have extended to what is now Saudi Arabia. It was the smallest of all the wild forms of equid of the Holocene, worldwide. I estimate its body mass to have been less than 150 kg ( Although only marginally present on the Arabian Peninsula, it conformed somewhat to the general trend of miniaturisation.


Procavia capensis jayakari of the Arabian Peninsula is smaller-bodied (average adult total length 41 cm, skull length 8.1 cm) than P. c. syriaca (49 cm and 9.0 cm, respectively) of Jordan, Lebanon, and Israel (Harrison and Bates 1991), and P. c. capensis (52 cm and 9.1 cm, respectively) of southern Africa (Skinner and Chimimba 2005). Given that the average adult body mass in southern Africa is 3.5 kg, the Arabian subspecies probably weighs only about 2 kg. See


Lepus capensis weighs about 2 kg in Syria and Lebanon, in the north of its distribution, as well as in South Africa, in the south of its distribution.

It therefore comes as a surprise that subspecies omanensis ( and and and of mainland Oman weighs only 1 kg (Chapman and Flux 1990). Its total length averages only 38.5 cm, compared to 59.5 cm in South Africa (Harrison and Bates 1991, Skinner and Chimimba 2005). Lepus capensis cheesmani ( and and and, widespread in the sandy deserts covering much of Arabia, is also diminutive. (Also see


Panthera pardus nimr: see and

Caracal caracal schmitzi ( of the Arabian Peninsula is smaller-bodied than Caracal caracal caracal of southern Africa. The averages for total length are 91 cm and 112 cm respectively (Harrison and Bates 1991, Skinner and Chimimba 2005).

Acinonyx jubatus (locally extinct) was smaller-bodied on the Arabian Peninsula than in Africa (Harrison and Bates 1991).


Canis lupus arabs: see and

Vulpes vulpes arabica, with adult body mass about 2.7 kg (, is far smaller-bodied than the same species in Europe, which has adult body mass about 6 kg (


Genetta genetta grantii of the Arabian Peninsula is smaller-bodied than G. g. genetta of the Iberian Peninsula, based on the length of the hind foot (7.5cm vs 8.5 cm). My sources are Harrison and Bates (1991) and


Hyaena hyaena sultana of the Arabian Peninsula has average adult body mass about 31 kg ( "All body measurements averaged smaller in the two southern subspecies, H. h. dubbah and sultana, than in the three northern subspecies, barbara, syriaca, and hyaena (Rieger, 1979a)" ( and The length of the skull averages 22.5 cm in H. h. sultana, compared with 24 cm in H. h. syriaca of Israel, Jordan, and Lebanon (Harrison and Bates 1991). See and and,March%202020%2C%20April%202020.


Mellivora capensis pumilio of the Arabian Peninsula is smaller-bodied than both M. c. wilsoni of Israel and M. c. capensis of southern Africa (Harrison and Bates 1991).


Papio hamadryas arabicus "appears to average smaller than P. h. hamadryas from Africa, with the teeth especially small" (Harrison and Bates 1991). See and

Based on Harrison and Bates (1991), this leaves only the following species as exceptions:

One of the reasons why this remarkably consistent diminution has been overlooked is that some of the species listed above lack the body proportions usually associated with reduction in body size, such as large-looking heads and eyes, and short-looking legs and horns.

For example:

Reduction in body size is a familiar pattern on islands ( and and, usually explained by limitation on resources in confined areas.

This can theoretically apply to peninsulas too, inasmuch as peninsulas produce isolation from mainlands. And a further parallel with islands is the tendency of the fauna to be driven rapidly to extinction by human hunters.

However, isolation is an unsatisfactory explanation for the pattern of diminution in the Arabian fauna, because this peninsula

Arabia, for most of the last two million years, has therefore been a subcontinent as much as a peninsula.

The mystery of the Arabian miniatures remains an intriguing puzzle for naturalists to ponder. Even allowing that aridity limits the productivity and reliability of resources, how can the ecological and evolutionary effects in Arabia have been so different from those in the Sahara?

Posted by milewski milewski, 02 October, 2020 12:53


Please see my paper in Oryx (Harding, L. E., O. Abu Eid, N. Hamidan, and A. Shah'lan. 2007. Re-introduction of the Arabian oryx (Oryx leucoryx) in Jordan: war and redemption, Oryx 41:478-487).

Posted by hardinglee over 1 year ago (Flag)

@hardinglee Dear Lee, how nice to meet you here. Thank you for your work on caribou and various other interesting species. You and I share a research interest in trace elements, including selenium. I look forward to discussing several topics here on iNaturalist with you in the future, if you would be willing...

Posted by milewski over 1 year ago (Flag)

I had started a major paper about selenium pharmacokinetics about 5 years ago, going so far as to get permission from my clients (I was a consulting biologist) for use of the data I had collected. But I lost interest--or rather, began to focus my (waning, it must be admitted) energy on trying to save the few mountain caribou we have left.
I enjoyed reading your blogs.

Posted by hardinglee over 1 year ago (Flag)

Could a possible explanation be hunting rather than isolation? Humans tend to hunt megafauna, and the largest elements are eliminated first.
Perhaps early enough exposure to developing human skills allowed miniaturisation as an escape from hunting?
Europeans landed at the Cape in 1652 and by 1700 had eliminated most mammals over 50kg from 200km from Cape Town. Admittedly this was in the era of gunpower and pans, and unrifled guns, but even today animals below 50kg are relatively common around Cape Town, whereas all species over 50kg have been (re-)introduced. .
I would envisage hunting as having two effects: smaller individuals are less desirable as food items, and smaller size also allows larger populations to persist in refugia around settlements.
Curiously: how many of the Arabian species are nocturnal?

The Camel of course was domesticated: not being hunted it supports this hypothesis!

Posted by tonyrebelo about 1 year ago (Flag)

@tonyrebelo Many thanks for these comments, which I'll think more about...

Posted by milewski about 1 year ago (Flag)

The levant experienced high levels of defaunation 5 millennium ago, early peoples exploited the region of modern syria immensely (, pressure like this, over thousands of years most certainly has made impacts on the general mass and size of the megafaunal taxa present in the arabian peninsula.

Posted by gingko_biloboa1 11 months ago (Flag)

@gingko_biloboa1 Hi Michael, Many thanks for your comment, with regards from Antoni.

Posted by milewski 11 months ago (Flag)

I would think aridity had a lot to do with it, but it also sounds like there must be something else at work here. Hunting by ancient inhabitants?

What a great puzzle to ponder!

Posted by beartracker 7 months ago (Flag)

@beartracker I'm glad you enjoy this topic. It's hard to explain via aridity because the nearby Saharan region is equally arid but has a fauna showing no similar pattern. With pressure of predation by humans, likewise, it's hard to see how this would have been greater in Arabia than in North Africa, not so?

Posted by milewski 7 months ago (Flag)

Those are both good points. What a puzzle!

Posted by beartracker 7 months ago (Flag)

Indeed a great and amazing puzzle! Fascinating. Many thanks and much appreciation! Ruth

Posted by grinnin 7 months ago (Flag)

@grinnin You're most welcome and it's good to see you here again, with regards from Antoni.

Posted by milewski 7 months ago (Flag)

In north Africa afrotropical mammals are i think more likely to benefit from connectivity with the sahel region and drought refugia north of the Atlas moutains during range contraction ( whereas their arabian counterparts are more likely to evolve isolated in drier condition (leading to the distinctives desert ecomorphotypes we know today)

A good example of this would be north African genets (Genetta genetta ssp) who show a strong morphological homogeineity, dispalying the mediteranean phenotype even in desert/semi-desert region (Gaubert et al., 2004)(Gaubert et al., 2009), where they in addition carry west African mt haplotypes. (Gaubert et al., 2009) Likely sign of recent introgression of WA into NA (Gaubert et al., 2009). In addition to this coastal mediteranean genets (Northern Algeria) have been found for uniquely carrying north African autochtonous mt haplotypes and displaying high allelic richness (Gaubert et al., 2015). It's reasonable to assume that semi-desert genets of NA may be a subset of genet from mesic habitat. Whereas on the orther hand arabian genets, likely share no mt haplotypes with Africa and according to Fernandes CA, 2004 they would be not consepcific at all with G. genetta, being highly cytochhrome b divergent. A likely sign of long term isolation and in situ adaptation in south Arabia ( author highlight the need of nuclear markers to fully resolve the case of arabian genets).

Otherwise i think that some species like North African Hystrix cristata would be pretty conform to the pattern followed by arabian species in respect of their evolution intensely impacted by the sahara dessication (Trucchi et al. 2016) and their average smaller size compared to their sub-saharan counterparts (Angelici et al. 2021).

Posted by redwan21 6 months ago (Flag)

@redwan21 Many thanks for your interesting comment.

Posted by milewski 6 months ago (Flag)

The Arabian miniatures are mysterious partly because Arabia is not particularly isolated, i.e. it is a stretch to call it a peninsula, let alone an island. The Arabian 'peninsula' was even less isolated for most of the last two million years, and until about 10, 000 years ago. This is because the lowered sea-levels of the 'Ice Ages' (which collectively occupied most of the Pleistocene) meant that most of the Persian Gulf was dry land, and the narrowest strait of the Red Sea/Gulf of Aden ( was even narrower than it is today. Please see Figure 2 in and Figure 2 and Figure 5 in

Posted by milewski 6 months ago (Flag)

@milewski Thank you for the share and your reply
There is no doubt that past sea level oscillation promoted some biotic interchange, as sahara climate oscillation do in North Africa. However the core of the question remain, how viable and frquent were these dispersal routes to ensure the break of Arabia biotic isolation ? Today no Artiodactyles species are directly shared between the two sides of the Bab al-Mandab, and this despite Artiodactyle overall high vagility... Isolation of Arabian lineages from East-Africa seems thus not unlikely

Posted by redwan21 6 months ago (Flag)

@redwan21 You are right, one of the remarkable aspects of the biogeography of Arabia is the degree to which the Red Sea has split the ungulate faunas. There are two exceptions at the species-level, namely Capra nubiana ( and Tragelaphus imberbis (see Harrison and Bates 1991, page 191). Both of these occur (or formerly occurred in the case of the lesser kudu) both west and east of the Red Sea. Harrison and Bates (1991) suggest that the lesser kudu was introduced to Arabia by humans, which seems plausible; and most authors (including me in this Post) have ignored the lesser kudu as a member of the Arabian fauna.

Posted by milewski 6 months ago (Flag)

Good photo showing body size of Panthera pardus in South Africa: The body masses of these individuals are at least double those of conspecifics in Arabia.

Posted by milewski 5 months ago (Flag)

Add a Comment

Sign In or Sign Up to add comments