The Mysterious Origins of the Dromedary

Introduction

The dromedary is amongst the most important domesticated animals, playing a key role in the history of North Africa and South-West Asia as a means of transporting people and goods across vast expanses of desert efficiently (1). The dromedary has been dispersed by man as far as Australia (4) and is incredibly common both in domesticated and feral form across its range.

Extant camelids have a wide pattern of domestication, the llama (Lama glama) and the alpaca (Vicugna pacos) arose from the guanaco (Lama guanicoe) and vicuña (Vicugna vicugna) respectively (7). The bactrian camel (Camelus bactrianus) appears to have descended from a subspecies of the wild bactrian camel (Camelus ferus), or perhaps a closely related species (3). The dromedary is then left as a somewhat enigmatic species, its origins being highly contentious. No close extant relative exists, and numerous possible ancestors have been suggested, including the Bactrian camel, and various fossil species.

Fig 1. The dromedary (Camelus dromedarius) is the smallest of the three extant old world camels and one of the most important domesticated animals in human history

Terms of use: This image is licensed under a creative commons attribution-ShareAlike 3.0 Unported. It is attributed to Jjron. The image is unedited and the original can be found here.

The Bactrian Camel

One prominent theory has been that the dromedary is a specialised lineage of Bactrian camel, which was bred for the hot deserts of South-West Asia, or alternatively that the dromedary is derived from a different population of the same ancestor somewhere in Central Asia (13). Despite this theory being somewhat intuitive, it has fallen from favour in the light of recent studies.

The dromedary and bactrian camel appear rather distinct at a glance, however this is not unusual amongst breeds of other domesticated species such as cattle or horses. Indeed, some of the most obvious differences such as size and body covering could be adaptations to different climates. One major feature difficult to explain is that the dromedary possesses only a single hump, whilst the bactrian camel has two. This discrepancy has for a long time been deemed less significant because dromedary foetuses reportedly possessed a vistigial second hump, suggesting they are derived from a two humped ancestor (13). However, the embryological observations describing this are from the late 19th century (13) and could not be replicated in recent experiments by Kinne et al 2010. Additionally, differences in the vertebrae structure are observable in the two species, likely as a result of skeletal adaptations to the humps (9).

Other morphological differences between the Dromedary and Bactrian camel in skeletal structure are relatively small and vary heavily within each species, thus it can be hard to distinguish them based on bones alone. Even so, A recent quantitative study of the dimensions of the skeleton of the dromedary and bactrian camels showed that there were numerous diagnostic characteristics. Most significantly the shape of the skull is quite different in the dromedary and bactrian camel, and somewhat larger in the latter. Differences in skull morphology could indicate fundamental differences in feed regime, the characteristics of the dromedary appear to be more in line with that of a browser and the Bactrian camel more like grazer. This suggests different wild ancestors as no key differences in diet are observed in the domesticated speices (9).

In addition to the morphological argument, proponents of the Bactrian camel origin theory also point to the biological species concept. Simply put, this states that two animals capable of producing fertile offspring belong to the same species. The bactrian camel and dromedaries comply with this criterion, in fact their hybrids, commonly called the ‘Tülu’ have often been bred historically as they attain larger sizes and make stronger beasts of burden (14). Hybrids are capable of producing at least four generations of offspring (9), though these are reportedly more aggressive, smaller and have a reduced fertility (9, 14). In addition to this it appears that despite historic hybridisation genetic material have not flowed between the two species (3). This presents a somewhat borderline case on the biological species concept, wherein hybrids are fertile, but potentially unviable in the long-term, which could indicate that even if they could be deemed conspecific there may be a deep evolutionary divide.

Such a divide has been supported by the advent of genetic analyses, which have shed light on the relationship between the bactrian and dromedary camels. A study by Cui et al 2007 suggested that the two species diverged about 8 million years ago, based on differences in the mitochondrial genomes and another more recent study found divergence dates of 4.4 million years (19). In either case the molecular evidence suggests a divergence dating far beyond the origin of modern humans, let alone the domestication of camels. Altogether this indicates that the dromedary and bactrian camels originated from two separate species, and despite similarities are sufficiently different to consitute different species from a morphological and phylogenetic standpoint, and almost under the biological species concept.

Fig 2. The Bactrian camel (Camelus bactrianus) is the closest living relative of the dromedary and played a similar role to the dromedary in Central and East Asia

Terms of use: This image is licensed under a Creative Commons Attribution-Share Alike 4.0 International. It is attributed to Edmart80. The image is unedited and can be found here.

The Wild Dromedary

Instead we must look for an ancestral species in the fossil record, to best evaluate this we must determine the location and time of domestication so we can get an idea of the fossil species in the viscinity at the time. This was approximately determined by Almathen et al 2016. Genetic diversity for species which have recently spread is usually greatest at their point of origin, this is because only a subset of the population colonises new land, only bringing part of the total genetic diversity with them, this is known as the founder effect. This is not the case for modern populations, the genetic structure of domesticated and feral dromedaries is fairly homogenous, indicating a large degree of interbreeding between populations, as might be expected for an animal used for long-distance transportation (1). Instead genetic diversity from historic populations has been used, in this case it appeared that genetic diversity was highest in South-Eastern Arabia, making it the most likely origin of the domesticated camel (1). The exact date for domestication is a little less clear, it is suggested that it dates to about the late second millennium BCE based on an increase in Camel bone frequency found at archaeological sites in the Arabien peninsula, a general decrease in bone size and the presence of figures depicting domesticated dromedaries (1). From there dromedaries appear to have spread to Mesopotamia and North Africa in the 1st  millennium BCE (1).

This supports a survival of the wild ancestor of the dromedary to at least the second millennium BCE. This date may be a little conservative, modern dromedary camels are divided into two haplogroups, which each subdivide into multiple haplotypes. This genetic diversity probably reflects several domestication or restocking events, which suggests a window of co-existence between domesticates and wild dromedaries. In any case a late Holocene survival of a wild camel population in the Arabian Peninsula is established. This is corroborated by the fossil record, nearly 18,000 bones attributed to wild dromedaries have been discovered at the Sufouh 2 site in Dubai dating to between the 2nd and 3rd millennia BCE (18), with a single specimen dated to the 1st millennium BCE. Unfortunately, remains are degraded and fragmented, but the most intact fossils clearly belong to the dromedary. Considering the age of the site most specimens were probably wild (1, 18). Fossil remains predating the estimated domestication time tend to be larger, indicating a size reduction which is commonly associated with domestication. (1, 18) Further wild dromedary remains have been dated as far back as the 5th millennium BC at other South-Eastern Arabien sites (2).

A few things are known about this population of wild dromedaries. The wild dromedary much like its domesticated counterpart appears to have inhabited deserts, possibly feeding on salt-tolerant plant species along the Arabian coastline (18). Interestingly, they appear to be a remnant population. Genetic studies suggest that the effective population of dromedaries was very low prior to domestication (1, (3), perhaps with only about 3,000 females producing offspring. This appears to have been the culmination of a long-term trend dating back nearly 100,000 years where the effective population size was around 100,000 (3). The reasons behind this decline are unknown but could plausibly be tied in with the drying of the Arabian peninsula during the last glacial period (15, 17), the arrival of Homo sapiens (6) in the Arabian Peninsula or some combination of the two. Human hunting of wild dromedaries appears to have been common, at least in the Holocene. Al Sufouh 2 with its abundance of camel remains appears to have been a butchery site, with many bones showing signs of cutting by tools. Artwork from Central Saudi Arabia dating to 3000 BCE, also clearly depicts the hunting of a dromedary (16). The absence of clear dromedary fossils from the Late Pleistocene and earlier is puzzling, it probably is partly due to the sparse fossil record in the region, but with a much greater population during this time it is peculiar no fossil material is known.

Fig 3. A section of the Arabian desert, showing scant pieces of vegetation amidst the dunes which dromedary feeds on

Terms of use: This image is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 2.0 Generic. It is attributed to Cliff Hellis. The image is unedited and can be found here.

Pleistocene Origins

The wild dromedary may therefore be derived from a distinct Middle or Late Pleistocene species. This is most often suggested as Camelus thomasi, a large camelid originally described from the late Early Pleistocene of Algeria (10). Much like the wild dromedaries of the Holocene, C. thomasi was somewhat larger than the extant dromedaries. An analysis of the post-cranial skeleton indicates a closer relationship to the Dromedary than Bactrian camel (13). Further fossils from middle-Pleistocene Morocco (10), Late Pleistocene Northern Sudan (13) and Late Pleistocene Israel (10) have all been assigned to this taxon. This would indicate C. thomasi as a prime candidate as the ancestor of the wild dromedary, being both close both geographically and temporally to the wild dromedary as well as showing affinities to it.

A recent analysis by Martini & Geraads 2018 however casts doubt upon this claim. The morphology of a wealth of specimens from the type location in Algeria (material which was not available to Peters 1997) was compared to the modern camel species. The species appears to have had a number of distinct features distinguishing it from either the dromedary or bactrian camel, but overall had stronger morphological affinities to the bactrian camel than the dromedary (10), which is suspect for a supposed direct ancestor. Additionally, it appears that the Late Pleistocene remains assigned to C. thomasi were assumed to belong to this taxon, because it was the only camelid known from the region during the Pleistocene so their validity is questionable (10). This suggests Camelus thomasi as an Early-Middle Pleistocene species endemic to North Africa, and an unlikely ancestor of the dromedary.

Until recently C. thomasi was the only species described from the region dating to the middle-Pleistocene or later, however three additional species have been discovered from this time in the El Kowm Basin of Syria (11, 12). Camelus roris is the earliest speices at the site, dating to between 550-150kya (11). The species was intermediate in size between C. thomasi and the dromedary, placing it around the size of the Bactrian camel. The skull of Camelus roris was larger and more massive than the dromedary, and it possessed a variety of cranial features which differs from both the extant dromedary and Bactrian Camel, furthermore it can be quite readily distinguished from any other species of Camelid based on its dental morphology. Overall, it did not exhibit similarities to the dromedary beyond what is observed in other members of the genus (11). Whilst it is possible that it survived beyond the 150kya there is nothing to indicate that it did, and its diplacement by the other two species make a Late Pleistocene survival unlikely. Thus, whilst presenting a more plausible candidate than C. thomasi, it is not a particularly persuasive case. A second middle-Pleistocene camel may have been present as well but remains are too rare and fragmentary to make an assessment (11).

Fig 4. The Middle Pleistocene-Holocene fossil sites of various camel species from the Middle East and North Africa.

More relevant are the two Late Pleistocene camelids from El Kowm, Camelus moreli and Camelus concordiae. Both disappear from El Kowm around 50kya, placing them close in time to the wild dromedary remains, but this time period may extend further as both species have tentatively been assigned to other Late Pleistocene remains in the region. C. moreli has been suggested as the true taxonomic identity of the Israeli and Sudanese fossils previously assigned to C. thomasi. Camelus moreli is however an unlikely candidate for the ancestor of the wild dromedary as it is morphologically disimilar. To start off with, it is much larger than any other Late Pleistocene camel, except for Camelus knobblochi from Northern Asia, weighing perhaps a full metric tonne. Additionally, large differences exist in the dental and skull morphology, for instance it had a much shorter skull than the dromedary. The limb and vertebral anatomy does not indicate that it was capable of supporting a lot of weight on top of it, leading to the conclusion that it may have lacked a hump. These large differences would preclude it as a likely ancestor of the dromedary (12).  

That leaves Camelus concordiae. Despite, only being conclusively known from El Kowm, a number of contemporary remains from Jordan, Israel and other parts of Syria also probably belonged to this species. Morphologically C. concordiae is much closer to the dromedary than any other known camelid, though there are differences. Dromedaries have relatively large skulls in relation to their bodies when compared to C. concordiae, and there are slight differences in the lower jaw proportions. The study describing these morphological differences concluded that two were sufficiently different to consitute separate species, but most likely they were closely related. Additionally, C. concordiae is actually smaller than the extant dromedary, let alone the wild dromedary of the Holocene. There are key lifestyle differences between the two species as well, where the dromedary was a desert specialist, C. concordiae is known only from the mediterrenean and arid steppes of the Levant (12). So how to reconcile the relationship between these two taxa? One possibility is that they are indeed two separate species, sharing a common dromedary-like direct ancestor, possibly the undescribed middle Pleistocene taxon from El Kowm. This would be similar to the relationship between the modern bactrian camel and wild bactrial camel (3). Alternatively, dromedaries may be a highly derived sub-species of C. concordiae. In this scenario a southern population of C. concordiae could have colonised the Arabian peninsula and adapted to desert lifestyle. This might explain the size reduction, by either placing the population outside the range of Camelus moreli allowing it to occupy part of the niche of the larger camelid. Or perhaps a size increase was an adaptation to storing water for longer distances. The changes to skull proportions may also reflect a difference in diet. This is all conjecture. Until more Pleistocene fossil remains of camels from the Arabian Peninsula is uncovered, the exact nature of the relationship between C. concordiae and the dromedary will likely remain unclear along with the origins of the dromedary.

Fig 5. A Late Pleistocene faunal assemblage at the El Kowm Basin in Syria. The site was an arid steppe. Depicted are the camels Camelus moreli (Centre) and Camelus concordiae (Left), The Syrian Wild Ass (Equus hemionus hemippus), Leopard (Panthera pardus), Goitered Gazelle (Gazella subgutturosa), and Northern Bald Ibis (Geronticus eremita).

Terms of use: Artwork by Hodari Nundu and Commissioned by The Extinctions

Conclusion

The ancestry of the dromedary in the Bactrian camel can be safely dismissed on the basis of morphological and genetic evidence, which suggest a much deeper divergence. Analysis of historic dromedary remains from the middle east reveals an origin of domestication in the South-East of the Arabian Peninsula, from a heavily reduced population of wild dromedaries. Wild dromedaries are known exclusively from this region during the Holocene, and have no clear Pleistocene remains, therefore its possible that they are derived from a different Pleistocene ancestor in the region, this was generally ascribed to Camelus thomasi, but reanalysis of this species suggests past affinities to the dromedary were exaggerated. Instead, Camelus concordiae a recently described camel from the Late Pleistocene of Syria appears the closest known species to the dromedary and may present either a direct ancestor or a close relative of the dromedary.

References

1.       Almathen, F., Charruau, P., Mohandesan, E., Mwacharo, J. M., Orozco-terWengel, P., Pitt, D., Abdussamad, A. M., Uerpmann, M., Uerpmann, H-P., De Cupere, B., Magee, P., Alnaqeeb, M. A., Salim, B., Raziq, A., Dessie, T., Abdelhadi, O. M., Banabazi, M. H., Al-Eknah, M., Walzer, C., Faye, B., Hofreiter, M., Peters, J., Honatte, O., Burger, P. A.. (2016). Ancient and modern DNA reveal dynamics of domestication and cross-continental dispersal of the dromedary. PNAS 113(24), 6707-6712.

2.        Beech, M., Mashkour, M., Huels, M., Zazzo, A.. (2009). Prehistoric camels in south-eastern Arabia: the discovery of a new site in Abu Dhabi’s Western Region, United Arab Emirates. Proceedings of the Seminar for Arabian Studies 39. 17-30.

3.       Burger, P. A.. (2016). The history of Old World Camelids in the light of molecular genetics. Tropical Animal Health and Production 48, 905-913

4.       Cherifi, Y., Gaouar, S. B. S., Guastamacchia, R., El-Bahrawy, K. A., Abushady, A. M. A., Sharaf, A. A., Harek, D., Lacalandra, G. M., Saidi-Mehtar, N., Ciani, E.. (2017). Weak Genetic Structure in North African Dromedary Camels Reflects Their Unique Evolutionary History. Plos ONE 12(1): e0168672

5.       Cui, P., Ji, R., Ding, F., Qi, D., Gao, H., Meng, H., Yu, J., Hu, S., Zhang, H.. (2007). A complete mitochondrial genome sequence of the wild two-humped camel (Camelus bactrianus ferus): an evolutionary history of camelidae. Biomedical Central Genomics 8 (241)

6.       Groucutt, H. S., Grun, R., Zalmout, I. S. A., Drake, N. A., Armitage, S. J., Candi, I., Clark-Wilson, R., Stephens, N. B., Stock, J. T., Stewart, M., Price, G. J., Kinsley, L., Sung, W. W., Alsharekh, A., Al-Masari, A., Bahameem, A. A., Al Murayyi, K. S. M., Zahrani, B., Scerri, E. M. L., Petraglia, M. D.. (2018). Homo Sapiens in Arabia by 85,000 years ago. Nature Ecology & Evolution 2, 800-809.

7.       Kadwell, M., Fernandez, M., Stanley, M. F., Baldi, R., Wheeler, J. C., Rosadio, R., Bruford, M. W.. (2001). Genetic analysis reveals the wild ancestors of the llama and the alpaca. Proceedings of the Royal Society of London 268, 2575-2584.

8.       Kinne, J., Wani, N., Wernery, U., Peters, J..(2010). Is There a Two-Humped Stage in the Embryonic Development of the Dromedary. Anatomia Histolgia Embryologia 39(5), 479-480.

9.       Martini, P., Schmid, P., Costeur, L.. (2017). ‘Chapter 1: Comparative morphometry of Bactrian camel and Dromedary’ in: Martini, P..(2019). Camel fossils from the El Kowm Basin, Syria. Diversity and evolution. PhD Thesis, University of Basel, Basel.

10.       Martini, P., Geraads, D..(2018). ‘Chapter 2: Camelus thomasi Pomel, 1893, from the Pleistocene type-locality Tighennif (Algeria): Comparisons with modern Camelus’ in: Martini, P..(2019). Camel fossils from the El Kowm Basin, Syria. Diversity and evolution. PhD Thesis, University of Basel, Basel.

11.       Martini, P., Costeur, L., Jagher, R., Le Tensorer, J-M..(2019). ‘Chapter 4 A new species from Nadaouiyeh Aïn Askar (Syria) contributes to the diversity of Pleistocene Camelidae’ in: Martini, P..(2019). Camel fossils from the El Kowm Basin, Syria. Diversity and evolution. PhD Thesis, University of Basel, Basel.

12.       Martini, P., Costeur, L., Le Tensorer, J-M., Schmid, P..(2019). ‘Chapter 5: A giant and a small camel lived side by side in the Late Pleistocene of Syria’ in: Martini, P..(2019). Camel fossils from the El Kowm Basin, Syria. Diversity and evolution. PhD Thesis, University of Basel, Basel.

13.       Peters, J..(1997). Camelus thomasi Pomel, 1893, a possible ancestor of the one-humped camel? Zeitschrift fur Saugetierkunde 63, 372-376.

14.       Potts, D. T..(2004). Camel Hybridization and the Role of Camelus bactrianus in the Ancient Near East. Journal of the Economic and Social History of the Orient 47(2), 143-165.

15.       Rosenberg, T. M., Preusser, F., Risberg, J., Plikk, A., Kadi, K. A., Matter, A., Fleimann, D..(2013). Middle and Late Pleistocene humid periods recorded in palaeolake deposits of the Nafud desert, Saudi Arabia. Quarternary Science Reviews 70, 109-123.

16.       Spassov, N., Stoytchev, T..(2004). The dromedary domestication problem: 3000 BC rock art evidence for the existence of wild One-humped camel in Central Arabia. Historia naturalis bulgarica 16, 151-158.

17.       Vaks, A., Bar-Matthews, M., Matthews, A., Ayalon, A., Frumkin, A..(2010). Middle-Late Quarternary Paleoclimate of northern margins of the Saharan-Arabian Desert: reconstruction from speleothems of Negev Desert, Israel. Quarternary Science Reviews 29, 2647-2662.

18.       Von der Driesch, A., Bruckner, H., Obermaier, H., Zander, A..(2007). The hunt for wild dromedaries at the United Arab Emirates coast during the 3rd and 2nd millenia BC. Camel bones from the excavations at Al Sufouh 2, Dubai, UAE. Archaeozoology of the Near East 8, 488-497.

19.       Wu, H., Guang, X., Al-Fageeh, M. B., Cao, J., Pan, S., Zhou, H., Zhang, L., Abutarboush, M. H., Xing, Y., Xie, Z., Alshanqeeti, A. S., Zhang, Y., Yao, Q., Al-Shomrani, B. M., Zhang, D., Li, J., Manee, M. M., Yang, Z., Yang, L., Liu, Y., Zhang, J., Altammami, M. A., Wang, S., Yu, L., Zhang, W., Liu, S., Ba, L., Liu, C., Yang, X., Meng, F., Wang, S., Li, L., Li, E., Li, X., Wu, K., Zhang, S., Wang, J., Yin, Y., Yang, H., Al-Swailem, A. M., Wang, J.. (2014). Camelid genomes reveal evolution and adaptation to desert environments. Nature Communications 5(5188)