When travelers think of experiencing their first safari, its usually South Africa that tops their wish list. But its time to consider Kenya, a country known for its picture-perfect savannas loaded with wildlife. In fact, the animal population is so dense in this country that even its capital, the very urban Nairobi, has its own game reserve. Herewith, a primer on two of the countrys most beautiful districts, where communing with elephants, lions, giraffes, zebras, and buffalo will be the most normal part of your day.
The MaraIts hard to find a more perfect destination for game-viewing than the Masai Mara and its unparalleled inventory of wildlife. If youre after the Big Five, the density of animals in the Mara makes it so that its almost impossible not to see them. The most dynamic time to visit is during the great migration when millions of wildebeests, zebras, gazelles, and many other animals travel to and from Kenya. July is when they make their trek from the Serengeti to the Mara, and they return to Tanzania starting in October. The sheer volume of wildlife on view is staggering, but the migrations can also be really exciting, because these animals natural predators (the lions, the hyenas) follow. So prepare to witness some life-or-death drama on the plains of Africa.
Tented Camps: If solid four walls arent your thing, &Beyonds Kichwa Tembo Tented Camp is perfectly positioned to make reality the best version of your glamping fantasies. The camp is on the path of the migration and offers killer views of river crossings. But soaking in jaw-dropping views is one of the cornerstones of a stay here with many tents offering stunning vistas of your surroundings, especially if you splurge on one of the eight superior-view tents that face out to the sweeping plains of the Mara. In terms of activities, all the classicsnight game drives, sundowner cocktails, and bush dinnersare available.
From the Sky: A hot-air balloon ride over a game reserve is pretty special, but what about taking it to the next level with a private plane? Elewana Collection, a portfolio of luxury camps and lodges in East Africa, just launched SkySafari Kenya Connoisseur, an eight-day itinerary that takes you to three diverse regions of Kenya with private flight transfers between them. Theres a lot of once-in-a-lifetime experiences included in this package, but the best perk could be saved for the end of the trip when you arrive to the Sand River Masai Mara. Elevate your game drive by taking it to the air with a Cessna Grand Caravan. Imagine aerial views of the migration or a lioness chasing after her preyall youll need to complete the moment is David Attenborough narrating the action.
Tucked Away: Because of the Maras wildlife-packed landscape, its only natural for it to be popular with hospitality companies, too, who want to give you the safari of your dreams. Masai Mara National Reserve is a solid choice when trying to locate a lodge to book, and there are plenty of choices there, but for something a little more secluded, secure one of the 30 glass-walled luxury tents from Angama Mara. Located high on the Great Rift Valleys Oloololo Escarpment at the border of the less congested Mara Triangle, the property then splits up the 30 tents into two camps of 15 to ensure privacy of the highest order. But despite its off-the-beaten-path location, Angama Mara doesnt sacrifice access: Its still only a 15-minute drive to the Masai Mara National Reserve.
When travelers head for Kenya, they often focus entirely on safari experiences at the Mara, which is a shame because Mount Kenya, located in the center of the country, delivers rugged beauty with its fair share of wildlife. With a peak of over 17,000 feet, Mount Kenya is Africas second highest mountain (and the highest in the country), and with its craggy shape, one of its most picturesque. If youre already in Kenya, it would be a shame to miss it.
Where to Stay: From the outdoor pool of the Fairmont Mount Kenya Safari Club, which sits on 100 acres of manicured lawns, you can enjoy the breathtaking serenity of this pocket of Africa. And if you have the right camera, unforgettable photos of the pool with the mountains peaks in the background are sure to deliver the kind of Instagram envy were all aiming for. Spacious rooms toe the line between Colonial and country-club chic to great effect. Interiors are generally subdued, favoring four-poster beds, floral upholstery, and lots of antique-looking furniture. But its really the outdoor offerings that make this resort a standout, especially for animal lovers. The hotel has a sanctuary, where numerous animals are nursed back to health, from ostriches and tortoises to leopards and even neighborhood monkeys, that arent actually supposed to be there. You can visit them all and learn about the propertys wildlife initiatives. Another special activity on offer is a guided horseback ride through the slopes of the mountain for a lovely outdoor breakfast. The staff will set up everything; all you need to do is get there. If youre lucky, you may spot elephants on your way.
What to Do: Because youre in the mountains, theres plenty to do if youre game for a little adventure. Enlist the help of luxury tour operator Abercrombie & Kent, which can organize a fantastic mix of outdoorsy activities with more cultural excursions. You may not be in the Mara, but there are still plenty of game reserves to add to the itinerary including Solio Game Reserve, a private ranch and home to hundreds of rhinoceroses. If youre a skilled climber, Mount Kenya offers beautiful but challenging treks that not only push your physical limits but also bring you opportunity to see zebras, eland, and, if youre hashtag blessed, maybe even wild African cats. If youre really hard core, theres also the opportunity to climb the granite cliffs of the mountain and then abseil down. But life is all about balance. Abercrombie & Kent can also set you up to meet with local women from the Nanyuki Spinners and Weavers, who specialize in Kenyan highland wool. They make some scarves and shawls, but they mostly focus on handloomed rugs, which come in a variety of designs, from more novel elephant pictures to contemporary prints in chevron stripes and leopard spots.
After 40 years of searching, researchers can finally put a new face on a mysterious human ancestor whose skull was discovered 40 years ago in Kenya. The find is giving scientists a better look at an enigmatic species that was alive soon after the dawn of our genus Homo about 2 million years ago. It also shows that there were several species of Homo present 1.78 million to 2.03 million years ago in the Rift Valley of Africa, and that they probably had to adapt in different ways to coexist.
Ever since paleoanthropologist Meave Leakey got her first look at the skull of a strange, new kind of human ancestor in 1972 at Koobi Fora, famed fossil beds on the east side of Africas Lake Turkana where several different species of human ancestors have been found since the 1960s, she and others have searched in vain for more members of this species. The 2-million-year-old skull had a big brain that made it a member of our own genus Homo. But its long, flat face and other features distinguish it from the other two members of early Homo known at the time, so many researchers thought of it as a new species, Homo rudolfensis. Some questioned whether it was a new species, however, or just an unusual member of Homo habilis, which lived 2.3 million to 1.4 million years ago in East Africa. "It was always an anomaly," says Leakey, of the Turkana Basin Institute in Kenya and Stony Brook University in New York. "We always knew we had to find more of it."
When fossil hunters found the well-preserved fossilized bones of the mid-face and teeth of a juvenile protruding out of rock in 2008, it "was really exciting," Leakey says. The face looked like a small "pocket version" of the original H. rudolfensis skull, known as KNM-ER 1470 -- with an unusually flat visage, as opposed to the more jutting upper jaw found in H. habilis, says co-author Fred Spoor of the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany. Its small size also ruled out the older view that H. rudolfensis skulls were invariably larger than those of H. habilis or that the larger specimens were males and the smaller were females.
With the discovery of a remarkably complete lower jaw in 2009, the team got an even better look at this elusive species (1470 did not have a lower jaw). The jaw and the new face revealed that H. rudolfensis had an unusual, U-shaped palate, with canines facing the front of the jaw rather than aligned on the sides in a V-shaped palate, as in H. habilis. This suggests a significant developmental difference between two species, rather than variation within one species, Spoor says.
The new fossils, described online today in Nature, were all found on the Karari Ridge of Koobi Fora, within 10 kilometers of the fossil beds where the 1470 skull was found -- and within the same region where fossils of H. habilis and H. erectus have been discovered. Some researchers still think that 1470 and the new fossils could be members of the same taxa (or biological group), H. habilis, because so few fossils of H. habilis have been found that "we still don't understand H. habilis," says paleoanthropologist Timothy White of the University of California, Berkeley.
But others say the new material "is really good evidence that there has to be *H. erectus *plus two or three other taxa," paleoanthropologist Bernard Wood of George Washington University in Washington, D.C., says, who was not involved in the new work. And if three species did coexist at roughly the same time and place, how did they compete with each other for food and sleeping sites? Did they eat different foods, inhabit different terrain, or use stone tools in various ways? Now its time to "think about hypotheses to explain how they might have divided up their world," says paleoanthropologist William Kimbel of Arizona State University, Tempe.
WIRED is where tomorrow is realized. It is the essential source of information and ideas that make sense of a world in constant transformation. The WIRED conversation illuminates how technology is changing every aspect of our livesfrom culture to business, science to design. The breakthroughs and innovations that we uncover lead to new ways of thinking, new connections, and new industries.
Since its discovery in 1972 (ref. 1), the cranium KNM-ER 1470 has been at the centre of the debate over the number of species of early Homo present in the early Pleistocene epoch2 of eastern Africa. KNM-ER 1470 stands out among other specimens attributed to early Homo because of its larger size, and its flat and subnasally orthognathic face with anteriorly placed maxillary zygomatic roots3. This singular morphology and the incomplete preservation of the fossil have led to different views as to whether KNM-ER 1470 can be accommodated within a single species of early Homo that is highly variable because of sexual, geographical and temporal factors4,5,6,7,8,9, or whether it provides evidence of species diversity marked by differences in cranial size and facial or masticatory adaptation3,10,11,12,13,14,15,16,17,18,19,20. Here we report on three newly discovered fossils, aged between 1.78 and 1.95 million years (Myr) old, that clarify the anatomy and taxonomic status of KNM-ER 1470. KNM-ER 62000, a well-preserved face of a late juvenile hominin, closely resembles KNM-ER 1470 but is notably smaller. It preserves previously unknown morphology, including moderately sized, mesiodistally long postcanine teeth. The nearly complete mandible KNM-ER 60000 and mandibular fragment KNM-ER 62003 have a dental arcade that is short anteroposteriorly and flat across the front, with small incisors; these features are consistent with the arcade morphology of KNM-ER 1470 and KNM-ER 62000. The new fossils confirm the presence of two contemporary species of early Homo, in addition to Homo erectus, in the early Pleistocene of eastern Africa.
Suwa, G., White, T. D. & Howell, F. C. Mandibular post-canine dentition from the Shungura Formation, Ethiopia: c rown morphology, taxonomic allocation, and Plio-Pleistocene hominid evolution. Am. J. Phys. Anthropol. 101, 247282 (1996)
Lieberman, D. E., Pilbeam, D. R. & Wood, B. A. A probabilistic approach to the problem of sexual dimorphism in Homo habilis: A comparison of KNM-ER 1470 and KNM-ER 1813. J. Hum. Evol. 17, 503511 (1988)
Kramer, A., Donnelly, S. M., Kidder, J. H., Ousley, S. D. & Olah, S. M. Craniometric variation in large-bodied hominids: testing the single-species hypothesis for Homo habilis . J. Hum. Evol. 29, 443462 (1995)
We thank the Governments of Kenya and Tanzania for permission to carry out this research, the Kenya Wildlife Service for permission to work in the Sibiloi National Park, the National Museums of Kenya and the National Museum of Tanzania for access to specimens in their care, and the Turkana Basin Institute for support. The National Geographic Society, the Leakey Foundation and the Max Planck Society funded fieldwork or laboratory studies. Many people helped us with this research, including N. Adamali, R. Blumenschine, C. Boesch, F. Brown, P. Gunz, J. J. Hublin, W. Kimbel, K. Kupczik, R. Leakey, C. Lepre, D. Lieberman, P. Msemwa, R. Odoyo, R. Quinn, P. Rightmire, L. Schroeder, U. Schwarz, M. Skinner, H. Temming, A. Winzer and B. Wood. Curatorial assistance was given by A. Kweka, F. Manthi, E. Mbua, M. Muungu and J. Thiringi. KNM-ER 60000 was discovered by C. Nyete, KNM-ER 62000 by D. Elgite and KNM-ER 62003 by R. Moru. We particularly thank the Koobi Fora Research Project field crew: A. Aike, S. Aila, D. Elgite, M. Kirinya, D. Gidole, O. Kyalo, A. Longaye, A. Lawri, E. Linga, J. Lonyericho, S. Lomeiku, D. Muema, A. Moru, R. Moru, S. Muge, C. Nyete, L. Nzuve, H. Sale and A. Sharamo whose fieldwork led to the discovery of these specimens, and camp managers J. Mutuku and T. Ngundo. H. Churcher, J. Coreth, A. Hammond, J. LaCarrubba, F. Kirera, C. Lepre, M. Noback, R. Quinn, M. Skinner, I. Wallace and S. Wright participated in one or more of the 2007, 2008 or 2009 field expeditions when these specimens were discovered. We are grateful to F. and J. Pinto, W. Philips, M. Hettwer, P. Sylvester, H. Buchi, N. Seligman, E. von Simpson, J. Doerr and B. and J. Chelberg for their financial support of this fieldwork.
Author Contributions M.G.L. and L.N.L. directed the field research, in which C.S.F. and F.S. participated. C.K. and F.S. prepared the hominin fossils, F.S. and M.C.D. made the virtual reconstructions, and C.S.F. studied the geological context. M.G.L., F.S., M.C.D., S.C.A. and L.N.L. made comparative observations and carried out analyses. F.S. took the lead in writing the paper, and S.C.A., M.C.D. and C.S.F. contributed.
Leakey, M., Spoor, F., Dean, M. et al. New fossils from Koobi Fora in northern Kenya confirm taxonomic diversity in early Homo. Nature 488, 201204 (2012). https://doi.org/10.1038/nature11322
Three hominin fossils newly discovered at Koobi Fora, east of Lake Turkana in Kenya, will greatly improve our understanding of the early radiation of the genus Homo, clarifying the iconic but enigmatic hominin cranium KNM-ER 1470, first described by Richard Leakey in Nature in 1973. The three are an exceptionally well-preserved lower jaw (KNM-ER 60000), a fragmentary lower jaw and, importantly, a well-preserved face. At between 1.78 million and 1.95 million years old, they broadly support the idea that there were at least two contemporary Homo species, in addition to Homo erectus, in the early Pleistocene of eastern Africa.
In 1967 G.G. Simpson described three partial mandibles from early Miocene deposits in Kenya that he interpreted as belonging to a new strepsirrhine primate, Propotto. This interpretation was quickly challenged, with the assertion that Propotto was not a primate, but rather a pteropodid fruit bat. The latter interpretation has not been questioned for almost half a century. Here we re-evaluate the affinities of Propotto, drawing upon diverse lines of evidence to establish that this strange mammal is a strepsirrhine primate as originally suggested by Simpson. Moreover, our phylogenetic analyses support the recognition of Propotto, together with late Eocene Plesiopithecus from Egypt, as African stem chiromyiform lemurs that are exclusively related to the extant aye-aye (Daubentonia) from Madagascar. Our results challenge the long-held view that all lemurs are descended from a single ancient colonization of Madagascar, and present an intriguing alternative scenario in which two lemur lineages dispersed from Africa to Madagascar independently, possibly during the later Cenozoic.
Strepsirrhine or toothcombed primates include three ancient clades that diverged early in the PaleogeneChiromyiformes (represented by one living and one subfossil species of aye-aye, both placed in the genus Daubentonia), Lemuriformes (containing extant lemurids, indriids, cheirogaleids, and lepilemurids, as well as the recently extinct archaeolemurids, palaeopropithecids, and megaladapids1), and Lorisiformes (lorisids and galagids). Recent molecular divergence estimates suggest that Madagascars lemurs (the clade containing both chiromyiforms and lemuriforms) split from lorisiforms in the Paleocene or early Eocene1,2,3,4,5, with lemurs colonizing Madagascar and then rapidly splitting into chiromyiform and lemuriform lineages1,2,3. The absence of a terrestrial Paleogene or Neogene fossil record on Madagascar6 has prevented paleontologists from testing this hypothesis. Given its geographic proximity to Madagascar, the adjacent African landmass is currently viewed as the most likely source for ancestral lemurs (and other endemic terrestrial mammals of Madagascar)6,7,8. Indeed, Paleogene ocean currents have been reconstructed as favoring west-to-east dispersal across the Mozambique Channel9, and together with the presence of both stem strepsirrhines and lorisiforms in Africas Eocene fossil record10,11,12 support this interpretation.
The primary challenge to the hypothesis of a single colonization of Madagascar by lemurs is the morphological evidence provided by the fossil primate Plesiopithecus teras, represented by a partial cranium and multiple mandibles from a single ~34Ma (terminal Eocene) site in Egypt13,14. Godinot15 suggested a relationship between Plesiopithecus and extant Daubentonia, noting that [Plesiopithecus] lower jaw has exactly the morphology that would be predicted for a daubentoniid ancestor, having already markedly enlarged its anterior tooth and reduced the teeth posterior to it (p. 457). Although marked enlargement of a single highly procumbent anterior lower tooth occurred more than once in primate evolution (e.g., in Eocene omomyiform haplorhines16), it is nevertheless an exceptionally rare pattern, and among extant primates is now seen only in Daubentonia, a taxon with highly specialized rodent-like anterior teeth. Godinots phylogenetic hypothesis linking Daubentonia and Plesiopithecus depends heavily upon their enlarged anterior lower teeth being homologous, as the cheek teeth of Daubentonia are highly modified and bear little resemblance to those of any living or extinct primate; the anterior teeth of Daubentonia are likely incisors17, whereas those of Plesiopithecus could be either canines or incisors18. Although Godinot did not test his hypothesis with an algorithm-driven phylogenetic analysis, it recently gained some support from a Bayesian tip-dating analysis of morphological and molecular data19 that, for some treatments of the morphological data, recovered evidence for a Daubentonia-Plesiopithecus clade within lemurs, despite the fact that, following previous interpretations13,20, the enlarged anterior lower tooth of Plesiopithecus was scored as a canine rather than as an incisor.
Here we present several new lines of evidence suggesting that the purported pteropodid fruit bat Propotto from the early Miocene of western Kenya is not only a strepsirrhine primate as was originally suggested by Simpson21, but represents a close relative of both Plesiopithecus and Daubentonia. Plesiopithecus was first described in 199214, a quarter century after the debate surrounding Propottos affinities appeared to have been resolved, hence its significance for interpreting Propotto was not appreciated. Our comparisons indicate that Propotto shares a number of specialized morphological features with Plesiopithecus, including all of the features that originally led Walker22 to doubt Simpsons proposed lorisid affinities for Propotto. Furthermore, digital reconstruction of the badly damaged upper molars of Plesiopithecus and comparisons with other specimens led us to the identification of two upper molars of Propotto, previously identified as Lorisidae indet.23, from the early Miocene site of Chamtwara, Kenya. These specimens offer the first glimpse of Propottos upper dentition and further buttress the case for its strepsirrhine primate affinities. Finally, a mandible of Plesiopithecus is here interpreted as retaining a small toothcomb-like canine distal to its enlarged anterior tooth, establishing that the enlarged anterior tooth is probably an incisor, thereby increasing the likelihood that the procumbent anterior lower teeth of Daubentonia, Propotto, and Plesiopithecus are homologous incisors.
Propotto leakeyi was originally described by Simpson as a lorisiform strepsirrhine that might be related to the extant lorisid Perodicticus (commonly known as the potto)21. The hypodigm available to Simpson included the holotype (KNM-SO 508, his specimen R; KNM=National Museums of Kenya), a right mandible with P3-M2 and alveoli for P2 and M3 as well as a small portion of the root of an enlarged anterior tooth (Figs.1g and 2b); KNM-RU 1879 (specimen S), a left mandible with a very shallow P2 alveolus, an erupting P3, fully erupted dP4 and M1, alveoli for M2 and an erupting M3 (see M1 and M3 in Fig.1c, d, respectively; this specimen also exhibits a laterally compressed and forward-facing alveolus for an anterior tooth); and KNM-RU 2084 (specimen T), a right mandible with M23 that is most likely from Songhor but labeled as being from Rusinga (Fig.1i). KNM-RU 1879 has Songhor written on the specimen despite the fact that the label suggests it might be from Rusinga; we consider it probable that the specimen is, in fact, from Songhor. If all of these specimens are indeed from Songhor, they would originate in the Chamtwara and the Kapurtay Conglomerates of Butler24, which Pickford25 put in his Set I fauna, and dated at 18.520Ma. This estimate is mainly based on K-Ar dates of Bishop et al.26 published in 1969, so additional work is needed to provide more precise age constraints for these localities using contemporary methodologies. Digital models of all the fossil specimens figured here are available on MorphoSource (Table1).
Comparison of lower molar morphology of latest Eocene Plesiopithecus teras and early Miocene Propotto leakeyi and mandibular morphology and lower dentition of Plesiopithecus teras. a M13 of DPC 11636, left mandible of Plesiopithecus teras (reversed for comparison, latest Eocene, Quarry L-41, Fayum Depression, Egypt); b M13 of CGM 42291, holotype right mandible of Plesiopithecus teras; c Left M1 of KNM-RU 1879, mandible of Propotto leakeyi (reversed for comparison; Simpsons specimen S; note that this specimen is probably from Songhor despite the Rusinga accession number); d Left M3 of KNM-RU 1879, mandible of Propotto leakeyi, reversed for comparison; e KNM-CA 1832, isolated right M1 of Propotto leakeyi (early Miocene, Chamtwara, Kenya); f KNM-CA 2195, isolated right M2 of Propotto leakeyi (early Miocene, Chamtwara, Kenya); g M1 of KNM-SO 508, holotype right mandible of Propotto leakeyi (early Miocene, Songhor, Kenya; Simpsons specimen R); h M2 of KNM-SO 508, holotype right mandible of Propotto leakeyi; i M23 of KNM-RU 2084, right mandible of Propotto leakeyi (possibly from Songhor despite the Rusinga accession number; Simpsons specimen T); j DPC 13607, left mandible of Plesiopithecus teras, with an alveolus that we interpret as being for a small canine, and tooth crowns that we interpret as I1 or I2 and P2-M2. Digital models were created using CT scans made available by the Duke Lemur Center Division of Fossil Primates and the National Museums of Kenya, which were downloaded from www.morphosource.org and made available for reuse under a CC BY-NC license. Map of Africa is adapted from Google Earth
Comparison of lower molar morphology of Daubentonia, Plesiopithecus, and Propotto, and volume rendering of the enlarged anterior teeth of Plesiopithecus and Propotto. a Left M13 of AMNH M-41334, extant Daubentonia madagascariensis, with individual surfaces reoriented slightly to facilitate comparison (teeth are from right side in AMNH M-41334 but are reversed for comparison); b left P3-M2 of Propotto leakeyi (holotype mandible KNM-SO 508, teeth are from right side but are reversed and reoriented slightly to facilitate comparison); c left mandible with I1 or I2 and canine-M3 of Plesiopithecus (DPC 11636). Scale in left panel is for ac (2mm). df Volume renderings of the enlarged anterior tooth (probable I1 or I2, rendered orange-yellow) in d KNM-KO 101, left mandible with partial root of I1 or I2 and crowns of P3-M2, Propotto leakeyi (early Miocene, Koru, Kenya); e KNM-RU 3690, right mandibular fragment with root and partial crown of I1 or I2 and crowns of P34, cf. Propotto leakeyi (early Miocene, Rusinga Island, Kenya); note that in this specimen the root of I1 or I2 extends under the roots of M1; f DPC 11636, left mandibular corpus with complete crowns of I1 or I2 and canine-M3, Plesiopithecus teras. Digital models were created using CT scans made available by the Duke Lemur Center Division of Fossil Primates and the National Museums of Kenya, which were downloaded from www.morphosource.org and made available for reuse under a CC BY-NC license
Simpson21 noted the highly peculiar cheek teeth (p. 51) of Propotto and the fact that its mandible deepened anteriorly, but nevertheless considered this taxon to be similar enough to the extant lorises Perodicticus and Nycticebus to recognize Propotto as an aberrant lorisid. Walker22 re-examined the hypodigm of Propotto and pointed out that the single-rooted P2 was probably small (though no crown is preserved), and not enlarged and caniniform as in lorisiforms. Further, he interpreted the alveolus of Propottos enlarged anterior lower tooth as being for a caniniform canine, and contrasted that with the canine morphology that would be expected in lorisiforms, which incorporate the canine into a toothcomb. Finally, he noted that the mandibular corpus was also unlike those of lorisiforms in deepening anteriorly and having a deep masseteric fossa. Walker concluded that Propotto was a pteropodid fruit bat and not a primate, a conclusion that was accepted by Simpson in correspondence exchanged before the 1969 publication of Walkers work (Supplementary Fig.1).
In 1984, Butler24 described a few additional Propotto specimens from the early Miocene sites of Koru, Chamtwara, and Rusinga Island in western Kenya25. The specimens from Rusinga localities located in the Hiwegi Formation would be considerably younger, dated to ~17.9 Ma27. Butler discussed the resemblance of Propottos cheek teeth to those of primates such as Cheirogaleus, Perodicticus, and Pithecia, and also with the Neotropical phyllostomid bat Artibeus. He further noted that the enlarged anterior lower tooth of Propotto (which he also interpreted as a canine) is relatively larger than the lower canines of extant pteropodids, having a root that extends posteriorly to at least P3. Despite these observations, Butler ultimately supported the idea that Propotto represented a side-branch of the chiropteran family Pteropodidae, erecting a new subfamily, Propottininae, for the genus.
For the last half-century, discussion of Propottos significance as a possible primate has been deterred by the authoritative consensus reached by Simpson, Walker, and Butler that Propotto is a bat. However, it is now clear that the features that Walker cited in his criticism of Simpsons identification of Propotto as a lorisid are all characteristic of the undoubted strepsirrhine primate Plesiopithecus and so do not necessarily exclude Propotto from Strepsirrhini (Fig.1j). The laterally compressed and presumably highly procumbent lower anterior tooth of Propotto (Fig.2d, e; unknown to both Simpson and Walker because this feature is only preserved in specimens described by Butler in 1984) does not occur in fruit bats, or for that matter any known living or extinct chiropteran. This feature is, however, present in Plesiopithecus (Figs.1j and 2df) and Daubentonia.
Despite being very low-crowned, the lower molars of Propotto are fundamentally strepsirrhine in structure, and are very similar to those of Plesiopithecus (Fig.1ai). Differences from Plesiopithecus include extension of the oblique cristids to meet the protoconid apices, reduction or elimination of hypoflexids, reduction of metaconids, and presence of a cingulid around the lingual margin of the metaconids. The P34 of Propotto and Plesiopithecus are very similar in having mesially shifted protoconids from which two dominant crests run distobuccally and distolingually to enclose well-developed talonids (Fig.2b, c). An automated geometric morphometric analysis of lower molar morphology in strepsirrhines, pteropodids, Propotto, and various living and extinct euarchontans demonstrates that the shape of Propottos M2 is most similar to that of strepsirrhines (and particularly those of cheirogaleids, Daubentonia and Plesiopithecus; Fig.3). The molar morphology of this set of taxa occupies a morphospace that is distinct from sampled pteropodids. Strepsirrhine lower molars also have low principal component (PC) 1 values that separate them from all non-euarchontans, non-primates, tarsiers, and almost all sampled Paleogene primates. The only fossils that group with modern strepsirrhines on PC1 are Adapis, Propotto, and Plesiopithecus. On PC2, strepsirrhines are divided into a cluster of lemurids, indriids, lorisiforms, and Adapis with high values, and another including cheirogaleids (Microcebus, Cheirogaleus, Mirza, Phaner), Daubentonia, Plesiopithecus, and Propotto with low values. The pteropodid fruit bats Pteropus and Rousettus are well-separated from Propotto in having much higher PC1-2 scores, although they do overlap with the second group of strepsirrhines. Although we only plot PC1 (20% of variance) and PC2 (14% of variance) in Fig.3, the most important clustering patterns are maintained on PC3 and PC4, as well (Supplementary Data13).
First two principal components (PC) resulting from principal component analysis of 1100 pseudolandmarks on a broad taxonomic sample of second lower molar teeth. Each point represents the tooth of one individual. Convex hulls and different colors indicate distinct extant taxonomic groups. Gray hulls with different marker symbols represent different fossil taxa. This sample includes teeth of extant non-primate treeshrews (Ptl: Ptilocercus, Tp: Tupaia sp.), cynocephalid dermopterans (Cn), and pteropodid fruit bats (Ptp). It also includes fossil non-primates and possible stem primates including Leptacodon sp. (Lp), various plesiadapiforms (Pls), and Altanius (Alt). It includes a number of extant primates including strepsirrhine lemurids (Lm), indriids (Id), cheirogaleids (ch), lorisids (Ld), and galagids (Gg), as well as Daubentonia madagascariensis. The extant haplorhine Tarsius is also included (Ts). Fossil haplorhines include Eosimias (Es) and Phenacopithecus (Ph). Early fossil prosimians include Donrussellia sp. (Dr), Cantius torresi (Ct), and Teilhardina sp. (Th). See Supplementary Data1 for a list of all specimens included. See Supplementary Data2 and 3 for principal component scores of additional components (e.g., 34), which also support the groupings of PC12
In addition, digital reconstruction of the damaged upper molars of Plesiopithecus (Fig.4a) reveals similarities with two upper molars from Chamtwara that were previously identified as Lorisidae indet. by Harrison23 (Fig.4b). Manipulation of digital surfaces of these upper molars allowed us to confirm that they occlude perfectly with Propotto lower molars from Chamtwara (Fig.1e, f), and they are accordingly identified here as the first known upper teeth of Propotto. The morphology of these upper molars also resembles that of the possible stem lorisiform Karanisia from the earliest late Eocene of Egypt (Fig.4d)10, the stem strepsirrhine Djebelemur from the early or middle Eocene of Tunisia28, and, intriguingly, the enigmatic late Eocene primate Nosmips from Egypt, which has been placed with Plesiopithecus in some phylogenetic analyses20. Among extant primates, Propottos upper molars (particularly M1) are most similar to those of the dwarf lemur Cheirogaleus. Similarities to Djebelemur, Karanisia, Nosmips, and Plesiopithecus include the broad lingual and more restricted buccal cingula, absence of a metaconule, and a concave distal margin of M2. Although some of these features may be primitive within Strepsirrhini, this character suite is nevertheless clearly characteristic of early strepsirrhines, and is not found in any living or extinct chiropteran. The upper molars of Propotto differ from those of known Paleogene strepsirrhines in being very low-crowned (matching the pattern seen in the lower molars), and in exhibiting a massive lingual cingulum, flattened lingual surfaces of the buccal cusps, and a reduced protocone.
Upper molars of early African strepsirrhines and extant Daubentonia from Madagascar. a Left M1 (on right) and M2 (on left) of DPC 12393, partial cranium of Plesiopithecus teras (latest Eocene, Quarry L-41, Fayum Depression, Egypt); reversed for comparison, the M1 is badly damaged and has been digitally reconstructed by segmenting out multiple fragments and repositioning them, while the M2 is lacking much of the buccal margin; b isolated right M1 [KNM-CA 1796, on right] and M2 [KNM-CA 1797, on left] of Propotto leakeyi (early Miocene, Chamtwara, western Kenya); c right M1 (on right) and M2 (on left) of AMNH M-41334, adult Daubentonia madagascariensis individual from Madagascar, locality unknown; d right M1 [on right, DPC 21639C] and M2 [on left, DPC 21636E] of Karanisia clarki (earliest late Eocene, Quarry BQ-2, Fayum Depression, Egypt); e oblique mesial view of DPC 21639C, right M1 of earliest late Eocene Karanisia clarki, showing the tall primary cusps, low parastyle, low lingual cingulum, and paraconule typical of early strepsirrhines; f oblique mesial view of KNM-CA 1796, right M1 of Propotto leakeyi, showing the low primary cusps, relatively tall parastyle, tall lingual cingulum, and absence of paraconule that is characteristic of this species. Scale is equal to 1mm. Digital models were created using CT scans made available by the Duke Lemur Center Division of Fossil Primates and the National Museums of Kenya, which were downloaded from www.morphosource.org and made available for reuse under a CC BY-NC license
The holotype of Plesiopithecus teras (CGM 42291; CGM=Egyptian Geological Museum) preserves a single enlarged and procumbent tooth mesial to P2-M3. A different mandibular specimen, DPC 11636 (DPC=Duke Lemur Center Division of Fossil Primates), was figured and discussed by Simons and Rasmussen13 (their Fig.3) and preserves a small tooth (which the authors interpreted as a P1) between the enlarged anterior tooth and the P2. They did note, however, that the tooth might also be the lateral canine derived from a toothcomb (p. 9949). At some point after the description of this specimen in 1994, the crown of the tooth was broken and glued back onto the root, although rotated into an incorrect orientation. We have reconstructed the tooth using digital models and provide additional views of the specimen for the first time (Figs.2c and 5). The tooth differs markedly in morphology and orientation from the adjacent P2, and has several features that are more consistent with it being a vestigial canine. The evolution of Daubentonias rodent-like incisor morphology from a toothcombed ancestor would likely involve topographically shifting the canine out of the toothcomb to accommodate an enlarged incisor. Indeed, in Plesiopithecus, this canine is strongly procumbent relative to its root, has a flattened surface (corresponding to the mesial face of a typical toothcomb canine, but more appropriately described as topographically lingual in DPC 11636) demarcated by a distinct ridge from the surface best exposed in occlusal view (lingual in a typical toothcomb canine, better described as topographically distal in Plesiopithecus), which has a gently curving and apically convex buccal margin. Morphological evidence supporting identification of this tooth as a lower canine rather than as a first premolar is supplemented by the dental formulae of all known living and extinct crown strepsirrhines, which unequivocally indicates that the loss of the upper and lower first premolar occurred along that clades stem lineage, and therefore before the appearance of both the strepsirrhine crown group and the split between the chiromyiform and lemuriform lineages. Indeed, no African strepsirrhine, living or extinct, is known to retain a P1. In light of this, it is more parsimonious to interpret this tooth of Plesiopithecus as a reduced lower canine, requiring the enlarged anterior tooth of Plesiopithecus to be an incisor, and therefore more likely homologous with the anterior tooth of Daubentonia. This enlarged anterior tooth also shows some thinning of the lingual enamel (relative to that on the buccal surface), though not to the extent seen in Daubentonia. The Plesiopithecus mandible DPC 13607 has also been digitally reconstructed, revealing a tiny canine alvelous anterior to the P2 (Fig.1j); therefore the holotype is unlike the two other known specimens in lacking a canine.
Comparative morphology of the lower dentition in crown strepsirrhines in phylogenetic context. From top to bottom, Galago senegalensis (MCZ 34381), Eulemur fulvus rufus (MCZ 16356), Microcebus (MCZ 45125), composite mandible of Plesiopithecus teras (DPC 11636; right side mirror-imaged), and Daubentonia madagascariensis (composite mandible using the corpus and incisor of AMNH M-185643 and the M13 of AMNH M-41334). Digital models were created using CT scans made available by the Museum of Comparative Zoology and Harvard University, the American Museum of Natural History, and the Duke Lemur Center Division of Fossil Primates, which were downloaded from www.morphosource.org and made available for reuse under a CC BY-NC license
Phylogenetic analysis (see Methods) of our combined molecular and morphological data matrix using the Bayesian tip-dating method with fossilized-birth-death parameterization recovered Propotto as exclusively related to Daubentonia, and Plesiopithecus as the sister taxon to this Daubentonia-Propotto clade (Fig.6). Standard Bayesian (non-clock) analysis recovered an exclusive Propotto-Plesiopithecus clade that is sister to Daubentonia. In both analyses, Propotto and Plesiopithecus are strongly supported as crown lemurs (posterior probability=0.9) and are situated as stem chiromyiforms. Importantly, this result emerged despite controlling for two scoring biases that could have provided additional support for a Daubentonia-Plesiopithecus-Propotto clade. First, though there are sound reasons to consider Propottos enlarged anterior lower tooth to be homologous with that of Plesiopithecus, Propotto was conservatively not scored for either canine or incisor characters (i.e., only premolar and molar characters were scored). Second, Plesiopithecus enlarged anterior upper teeth were scored as canines and not incisors, although they could conceivably be enlarged incisors homologous with those of Daubentonia. To further avoid bias, we did not create any new characters or character states to capture novel observations of derived dental features shared by Daubentonia and Propotto to the exclusion of Plesiopithecus (see discussion below).
Phylogenetic relationships and biogeography of living and extinct strepsirrhines. Time-scaled tree derived from Bayesian tip-dating analysis of the combined molecular and morphological dataset. Terminal branches are color coded according to continental biogeography, and internal branches are color coded according to Bayesian ancestral biogeographic analysis. Numerical values to the right of nodes represent clade support (posterior probabilities) and circled numbers at each strepsirrhine node represent the posterior probability of each biogeographic reconstruction. Complete time-scaled phylogenetic trees and biogeographic reconstructions are available at the Dryad Digital Repository associated with this study(https://doi.org/10.5061/dryad.gb182)
Bayesian stepping-stone estimation of marginal likelihoods for alternative placements of Plesiopithecus and Propotto, using the morphology matrix and constraining these two taxa to fall in different positions within the optimal time-scaled tree derived from the tip-dating analysis of molecular and morphological data, reveals that there is strong evidence (based on a Bayes factor of 15.64) for favoring a (Plesiopithecus (Daubentonia, Propotto)) topology over the (Daubentonia (Plesiopithecus, Propotto)) topology derived from the non-clock analysis. Other alternative constraints, such as situating Plesiopithecus and Propotto as advanced stem strepsirrhines or as stem lemurs, were decisively rejected by stepping-stone analyses (based on Bayes factors of 572.49 and 651.97, respectively). Bayesian reconstruction of ancestral morphological character states on the optimal clock topology identified 15 character state changes along the chiromyiform stem leading to the (Plesiopithecus (Daubentonia, Propotto)) clade, and 18 character state changes along the lineage leading to the Daubentonia-Propotto clade. The monophyly of Eocene-Recent chiromyiforms is supported by character changes relating to the modification of the anterior dentition to include only a single enlarged and procumbent incisor, as well as numerous details of premolar and molar crest and cusp development/placement, and increased depth of the mandibular corpus (see supporting data files held in the Dryad Digital Repository associated with this study(https://doi.org/10.5061/dryad.gb182)).
Bayesian reconstruction of strepsirrhine biogeographic history strongly supports (posterior probability=1) African origins for both Chiromyiformes and Lemuriformes, implying independent dispersals across the Mozambique Channel. Our analyses place the last common ancestor of Daubentonia and Propotto on the African continent at 27.9Ma (near the early-late Oligocene boundary), suggesting that the dispersal to Madagascar that ultimately gave rise to Daubentonia likely occurred some time after the early Oligocene. The continental divergence of the chiromyiform and lemuriform lineages is estimated at 41.1Ma (late middle Eocene) and the island origin of crown lemuriforms is estimated at 19.9Ma (early Miocene).
Our analyses suggest that Propotto and Plesiopithecus are stem chiromyiform lemurs that are closely related to the extant aye-aye Daubentonia from Madagascar, and that stem chiromyiforms were present in Africa from at least the late Eocene through the early Miocene. Our time-scaled tree, when combined with Bayesian biogeographic analyses, strongly supports an African origin for the common ancestor of lemuriforms and chiromyiforms, and independent dispersals of these groups across the Mozambique Channel to Madagascar. Although independent African origins for two closely related Madagascan lineages might appear overly coincidental, recent molecular phylogenies of chameleons indicate that there were two independent colonizations of Madagascar by African lineages in the Cenozoic29, establishing an independent precedent for the feasibility of such a pattern. Our results suggest that the chiromyiform dispersal to Madagascar occurred no earlier than the Oligocene (based on the divergence date of Propotto and Daubentonia), while the lemuriform dispersal could have occurred no later than the early Miocene (based on the time of origin of the crown lemuriform clade). Our results do not allow us to address the question of when the lemuriform dispersal to Madagascar occurred within the ~41 to ~20Ma dispersal window provided by our analyses, but it is noteworthy that the combined phylogenetic and biogeographic evidence can now accommodate a scenario in which lemuriforms dispersed to Madagascar quite late in the Cenozoic (i.e., as late as the earliest Miocene), where they then underwent an adaptive radiation. Because our results do not require lemuriforms to have been present on Madagascar until the early Miocene, they render the recently proposed hypothesis of a mass extinction of lemuriforms on Madagascar near the EoceneOligocene boundary1 less likely, but not impossible.
Previously, lemurs have been regarded as the first placental mammals to colonize Madagascar (with the possible exception of the enigmatic subfossil mammal Plesiorycteropus, whose affinities are uncertain). However, our results suggest much later dispersal windows for lemuriforms and chiromyiforms which overlap with those that have been recently estimated for Madagascars other endemic terrestrial mammalsi.e., euplerid carnivorans30,31, nesomyine rodents31, and tenrecids31,32. The Oligocene to early Miocene interval also overlaps with recently proposed dispersal windows for hyperoliid frogs8,33, lamprophiid snakes8,34, zonosaurine lizards8,35, as well as multiple scincids and gekkonids8. Importantly, this interval was also characterized by the lowest sea levels in the Cenozoic36, prior to the onset of middle Miocene cooling.
Several additional derived dental features shared by Propotto and Daubentonia (to the exclusion of Plesiopithecus) would be consistent with a close relationship of the former two genera, but were not captured by the morphological matrix used here. In the upper dentition, both taxa have particularly well-developed parastyle and metastyle cusps that are close in height to the paracone and metacone, respectively; flat (as opposed to convex) lingual surfaces of the buccal cusps; and low molar protocones. In the lower dentition, Daubentonia and Propotto share very shallow talonid basins; tall and wall-like oblique cristids; and highly reduced (Propotto) or absent (Daubentonia) metaconid and entoconid cusps, protocristid crests, and hypoflexids on M12. The Propotto specimen KNM-RU 3690 shows that the mesial aspect of P3 was much closer to the alveolus for the enlarged anterior lower tooth than in Plesiopithecus, suggesting that the P2 of Propotto was smaller than that of Plesiopithecus and that there was no room for a canine as in Plesiopithecus; this pattern is derived toward the total loss of the lower premolars seen in Daubentonia.
From a functional perspective, the distally oriented postprotocrista and mesiodistally aligned lingual cingulum on the M1 of Propotto, combined with the low M12 protocones, flattening of the talonid basins and reduction of the hypoflexids, protocristids, and metaconids on the lower molars (features not present in Plesiopithecus) reflects increased emphasis on propalinal mastication and associated development of predominantly mesiodistally aligned upper and lower molar wear facets, as in Daubentonia. If Propotto is indeed a close relative of Daubentonia, the formers massive lingual cingula and reduced protocone and metaconid cusps might even help to explain the strange lingual region of M12 in Daubentonia. These teeth superficially appear to lack a lingual cingulum, but nevertheless bear hypocone cusps, structures that are always derived from the distolingual cingulum in crown strepsirrhines. The fact that the M12 hypocones in Daubentonia are connected buccally to the postmetacristae by what appear to be tall and thick postcingula, and are continuous mesiolingually with similarly tall and thick elongate ridges, hints at the possibility that the latter features might be derived from the lingual cingulum and not the protocone, and that the protocones (which are highly reduced in Propotto) are effectively absent in Daubentonia. A possible mechanism for this transformation is provided by the occlusal morphology of the Propotto molars from Chamtwara, the M1 of which bears a cingulid lingual to the very reduced metaconid that occludes with the lingual cingulum of M1. Digital manipulation of these surfaces indicates that the gutter between the mesial aspect of the M1 lingual cingulum and the protocone occludes on top of the reduced M1 metaconid, which is a remarkably odd arrangement, but one that, taken to an extreme, would yield a morphology like that seen in Daubentonia (Fig.2a). The morphological similarity of Daubentonia and Propotto M2 surfaces, as supported by our automated geometry analyses, lends further credibility to this hypothesis. Again, these novel interpretations of the cusp/crest homologies of Daubentonia were not scored as characters in our phylogenetic analyses to avoid circularity in our assessment of phylogenetic relationships of the taxa in question.
Marked restructuring of interpretations of strepsirrhine biogeographic history suggested by our analyses presently depends almost exclusively on dental morphology, hence more rigorous tests of these hypotheses will only be possible as new and more complete fossils are discovered. An obvious challenge to the hypothesis that chiromyiforms and lemuriforms independently dispersed to Madagascar is the current lack of diagnostic stem lemuriforms in the African fossil record. Notably, the Paleogene fossil record of Afro-Arabia is notoriously poor and geographically biased toward northern Africa37. Nonetheless, a handful of fragmentary fossils provide tantalizing evidence of possible lemuriform-like strepsirrhines from other parts of Afro-Arabia, such as Notnamaia from Namibia38, and Omanodon and Shizarodon from Oman39. Furthermore, an early Miocene origin for the crown lemuriform clade does not necessarily require that the Africa-to-Madagascar dispersal event be coincident with, or even close in age to, the origin of that cladethat dispersal could have occurred at any point along the long lemuriform stem lineage. Therefore, absence of lemuriform fossils from the Paleogene of Afro-Arabia might be explained by an early dispersal to Madagascar closer to the chiromyiformlemuriform split, followed by the extinction of basal stem lemuriforms in Africa. An analogous pattern is seen in the fossil record of platyrrhine anthropoid primates that likely also originated in Afro-Arabia and then dispersed to South America, from an as-yet unsampled stem lineage in Afro-Arabia.
Regardless of the phylogenetic and biogeographic history of Daubentonia, it is of great paleoecological significance that Cenozoic African primate communities gave rise to a somewhat Daubentonia-like (and presumably tree gouging) primate lineage, as occurred in various non-primate mammalian radiations on other continents, such as Apatemyidae and various plesiadapiform euarchontans on northern continents, and the marsupial groups Petauridae and Yalkaparidontia in Australasia40,41. It is also significant that the strepsirrhine lineage represented by Plesiopithecus and Propotto persisted in Africa well into the Miocene, long after major perturbations in Earth climate history such as global cooling at the EoceneOligocene boundary and biotic events in Africa such as the immigration of multiple mammalian lineages in the late Oligocene or early Miocene. These patterns add to the evidence that equatorial Africa likely had a key role as a relatively temperate refugium for primate communities at a time of marked mid-Cenozoic ecological restructuring42,43.
We tested Simpsons original hypothesis of Propotto-lorisid affinities, and Godinots hypothesis of a Daubentonia-Plesiopithecus clade within Strepsirrhini, by adding Daubentonia and Propotto to an augmented version of Seiffert et al.s44 morphological character matrix, rescoring the anterior lower dentition of Plesiopithecus (taking into account the observations detailed above), and combining those morphological data with the molecular dataset of Springer et al.4. The dermopteran Galeopterus was also added to the matrix and scored for morphological characters, so that two extant euarchontan outgroup taxa were scored for both molecular and morphological data; the augmented morphological dataset now includes 102 fossil and 23 living euarchontans and 395 characters. To maximize phylogenetic signal, 264 ordered characters were constructed using intermediate states to code for polymorphic observations45. However, MrBayes currently imposes a six-state limit for ordered transformation series, necessitating that 39 of these characters be recoded using standard polymorphic scoring.
The DNA supermatrix of Springer et al.4 consists of 69 nuclear and 10 mitochondrial gene segments totaling 61,119 positions. To optimize data completeness, molecular sequences for Lepilemur ruficaudatus and Propithecus verreauxi were selected to accompany morphological character scores for their respective genera. All other taxa were scored for both molecular and morphological data at the species level, and species present in the DNA supermatrix but absent from the morphological dataset were removed. PartitionFinder v2.1.146 was used to select a DNA subset scheme and nucleotide substitution models. Each gene segment was assigned as an input block and testing included all models available in MrBayes. The best parameterization was assessed using the Bayesian Information Criterion (BIC), which recommended a scheme with 13 subsets and a combination of General Time Reversible (GTR+G and GTR+I+G) models.
The molecular and morphological datasets were concatenated with Mesquite v3.2047. We used the parallel (MPI) version of MrBayes v3.2.648 to conduct two total-evidence Bayesian phylogenetic analyses. The first was a standard non-clock analysis. The second was a time-scaled clock analysis which implemented the tip-dating method with fossilized-birth-death (FBD) parameterization. For the molecular portion of the matrix, settings for partitions and models were assigned based on the PartitionFinder results. The morphological portion of the matrix was set as a single partition using a gamma-distributed Markov k model and variable coding to accommodate ascertainment bias. Parameters for models across all 14 partitions were set as unlinked. Two hard topological constraints were appliedone enforced monophyly for primates and the other constrained scandentians as an outgroup. Metropolis coupled Markov chain Monte Carlo (MCMCMC) parameters were set for 2 runs with 4 chains each and to sample in 1000 generation increments. To promote chain swapping, the heating temperature was set to 0.02.
Run length was assessed and chosen using built-in MrBayes diagnostics for convergence and sampling sufficiency. First, we targeted run lengths yielding an average standard deviation of split frequencies (ASDSF)0.01. The value of this diagnostic should approach zero as individual runs converge on topological distributions. Second, we targeted run lengths in which the minimum estimated sampling size (minESS) of all parameters was100. Burn-in settings were selected to optimize these two diagnostics. We explored absolute burn-in generation values in increments of 5M, ranging from 5M up to 50% of run lengths. At each increment, tree and parameter summarizations were run and diagnostic data were collected. The final burn-in value for each analysis was selected by identifying all values that yielded the ASDSF target (0.01), and then from this group, choosing the one which yielded the minESS set whose least sampled parameter most exceeded the minESS target (100). Under this run-length strategy, the non-clock analysis was run for 92M generations of which 5M were discarded as burn-in. The resulting ASDSF was 0.00995 and the smallest minESS in the parameter set was 532.5906. The tree distribution was summarized using the allcompat (majority-rule plus compatible groups) option.
Our clock analysis included several additional settings. (1) The node age prior was set as calibrated with living taxa fixed to zero (=present day) and fossil taxa constrained to an age range. Age ranges were estimated using (when possible) the currently recognized upper and lower bounds of magnetochrons, land mammal ages, and/or other radiometric constraints in which each fossil taxon may be reasonably placed44. Fossil-tip ages were set to sample from uniform distributions across these ranges. (2) For the clock variance prior, we used the independent gamma rates (IGR) model to estimate relaxed clock rates. For the associated IGR variance prior, we used the default MrBayes setting (exponential distribution, =10). (3) We applied the FBD model for the prior probability distribution of branch lengths and used the fossil-tip sampling strategy. The FBD extinction and fossilization priors were set as flat (beta distributions, =1, =1) and the FBD speciation prior was set to the MrBayes default (exponential distribution, =10). The FBD model also requires a sampling probability prior which is an estimate for the proportion of extant taxa included in the study. With about 450 living primate species4, 2 currently recognized living dermopteran species49 (though it should be noted that there is now strong evidence for cryptic diversity within Demoptera50), and 20 living scandentian species51, the 23 extant taxa in our matrix constitute ~0.0487 of living euarchontan diversity. (4) The clock rate prior is an initial estimate for a distribution describing the number of substitutions per site per million years. To derive this prior, we used novel R code that utilizes the non-clock tree, age estimates for each taxon and an age estimate for the tree root. First, the dist.nodes function from the R package APE v3.452 was used to extract path lengths from each tree tip to the tree root. Next, each path length was scaled by the difference between the root age estimate and a tip age estimate. For the root age, we used 65.2Ma that corresponds to the earliest bound for Purgatorius, the oldest fossil taxon in the dataset. For tip ages, we used zero for living taxa and age range midpoints for fossil taxa. Finally, the fitdist function from the R package fitdistrplus v1.0-653 was used to fit normal, lognormal and gamma distributions to the set of scaled path lengths. The best model was assessed using the BIC, in this case a lognormal distrbution with a mean of 3.983465721 and a standard deviation of 0.564231504. This model and its parameter values were used directly for the MrBayes clock rate prior. (5) Age calibrations were applied to both the Primates and Euarchonta nodes. Calibration settings specified truncated normal distributions with minimum and mean ages corresponding to the earliest bound of the oldest fossil taxon in the group (Primates: Teilhardina=55.8Ma, Euarchonta: Purgatorius=65.2Ma). Pre-Eocene ghost lineages were penalized by setting 1Ma standard deviations on these distributions. The clock analysis was run for 150M generations, of which 20M were discarded as burn-in. The resulting ASDSF was 0.009053 and the smallest minESS in the parameter set was 138.0835. The tree distribution was summarized using the allcompat option.
Both the clock and non-clock analyses recovered stem chiromyiform positions for Plesiopithecus and Propotto, but with different placements relative to Daubentonia. As time is a fundamental factor in estimating branch lengths of phylogenetic trees, and statistical methods such as Bayesian inference inseparably consider branch lengths and topology in likelihood calculations, we regard the result of the clock analysis as optimal. To compare the estimated marginal likelihood of the clock result to those of alternate topologies, we conducted a set of post hoc analyses using MrBayes. To do so, Plesiopithecus and Propotto were pruned from the time-scaled clock tree, and the remaining tree was used as a soft constraint for reanalysis of the morphological partition with stepping-stone sampling. (A) The first reanalysis allowed Plesiopithecus and Propotto to be placed anywhere in the tree. The result was identical to the clock analysisspecifically that Plesiopithecus is positioned as sister to a Propotto-Daubentonia clade [H1, optimal]. (B) The second reanalysis disallowed a Propotto-Daubentonia group. The resultant topology for Chiromyiformes was identical to the non-clock analysisspecifically that a Propotto-Plesiopithecus clade is positioned as sister to Daubentonia [H2]. Comparison of the estimated marginal likelihood of H1 to that of H2 yielded a Bayes factor of 15.6. (C) The third reanalysis disallowed placements of Plesiopithecus and Propotto as stem chiromyiforms. Given this constraint, a Propotto-Plesiopithecus clade was instead positioned as the immediate sister clade of crown strepsirrhines [H3]. Comparison of the estimated marginal likelihood of H1 to that of H3 yielded a Bayes factor of 572.5. (D) The fourth reanalysis also disallowed placements of Plesiopithecus and Propotto as stem chiromyiforms, but forced these taxa to be placed within crown Strepsirrhini. Given these constraints, a Propotto-Plesiopithecus clade is positioned as the immediate sister group of the chiromyiformlemuriform clade [H4]. Comparison of the estimated marginal likelihood of H1 to that of H4 yielded a Bayes factor of 652.0. We interpret Bayes factors >10 as strong evidence, and >100 as decisive evidence, in favor of the optimal hypothesis versus alternate hypotheses54,55.
MrBayes was used to conduct ancestral state reconstructions (ASR) for all morphological characters that were scored for either Plesiopithecus or Propotto (n=166). Given the topology and branch lengths of the allcompat tree derived from the clock analysis, ASR provides the probabilities of all states of all characters for each ancestral node. For each character, we considered the state with the highest probability to be the best estimate at an ancestral node, provided that the probability of that state exceeded all others by >10%. If the probability of a runner-up state was 10% of the highest, we considered the best ancestral estimate to be inclusively polymorphic.
Geographic distribution for each taxon in the dataset was coded as one of the following six land masses: North America, South America, Europe, Asia, Afro-Arabia, Madagascar. For both the optimal [H1] and first runner-up [H2] trees, MrBayes was used to conduct ASR of biogeography; both analyses support independent dispersals of the chiromyiform and lemuriform lineages from Afro-Arabia to Madagascar. A concise summary of ancestral biogeography on the optimal tree is presented in Fig.6.
To assess the phenetic affinities of the molar teeth of Propotto in an objective and quantitative way, we used an automated three dimensional (3D) geometric morphometric analysis. Our sample consists of lower second molars of 222 individuals representing 42 genera (Supplementary Data1). Digital models of these specimens were created by micro-CT scanning physical specimens (either casts or originals); Avizo versions 6-8.156 were used to fit and crop surfaces, followed in some cases by further processing (patching and smoothing) in Geomagic57. Euarchonta is comprehensively represented (with the exception of anthropoids), including a diversity of early fossil prosimians. We also include molars of two pteropodid genera, as a test of the hypothesis that Propotto is a primitive fruit bat. Automated analyses allow an objective and comprehensive representation of shape. The software evenly spreads a user-specified number of pseudolandmarks over the surface of a 3D object and then algorithmically determines the correspondence among landmarks on different bones58. We used the MATLAB version of auto3dgm, the most recent version of which can be accessed by contacting the authors or through the github address: https://github.com/trgao10/PuenteAlignment/. We used the following parameters in our analysis: 300 initial pseudolandmarks, 1100 final pseudolandmarks, 3000 iterations of iterative closest points, and reflections allowed. The analysis was run on the mathematics computing cluster at Duke University. The Procrustes transformed pseudolandmark coordinates are available (Supplementary Data1). The aligned coordinates of the pseudolandmarks from TableS1 were then analyzed using principal components analysis of tangent space in morphologika2.5 (Supplementary Data2 and 3). We decided to remove Prolemur from the analysis post hoc given the small sample (n=2), and problems with breakage on one tooth and heavy wear on the other. Additionally, we identified seven outlier specimens (in the sense that they plotted far from other individuals of their species) in an early version of the analysis. These seven outliers turned out to have mesh or alignment issues as noted in the footnote of Supplementary Data1.
Input data files, settings, and results from phylogenetic, stepping stone, ASR, and biogeographic analyses are available on the Dryad Digital Repository (https://doi.org/10.5061/dryad.gb182). Digital surface models for all of the figured fossil specimens are available on MorphoSource (www.morphosource.org, see Table1 for digital object identifiers).
Steiper, M. E. & Seiffert, E. R. Evidence for a convergent slowdown in primate molecular rates and its implications for the timing of early primate evolution. Proc. Natl Acad. Sci. USA 109, 60066011 (2012).
Crottini, A., Madsen, O., Strau, A., Vieites, D. R. & Vences, M. Vertebrate time-tree elucidates the biogeographic pattern of a major biotic change around the KT boundary in Madagascar. Proc. Natl Acad. Sci. USA 109, 53585363 (2012).
Seiffert, E. R., Simons, E. L., Ryan, T. M. & Attia, Y. Additional remains of Wadilemur elegans, a primitive stem galagid from the late Eocene of Egypt. Proc. Natl Acad. Sci. USA 102, 1139611401 (2005).
Everson, K. M., Soarimalala, V., Goodman, S. M. & Olson, L. E. Multiple loci and complete taxonomic sampling resolve the phylogeny and biogeographic history of tenrecs (Mammalia: Tenrecidae) and reveal higher speciation rates in Madagascars humid forests. Syst. Biol. 65, 890909 (2016).
Nagy, Z. T., Joger, U., Wink, M., Glaw, F. & Vences, M. Multiple colonizations of Madagascar and Socotra by colubrid snakes: evidence from nuclear and mitochondrial gene phylogenies. Proc. R. Soc. B 270, 26132621 (2003).
Blair, C. et al. Multilocus phylogenetic and geospatial analyses illuminate diversification patterns and the biogeographic history of Malagasy endemic plated lizards (Gerrhosauridae: Zonosaurinae). J. Evol. Biol. 28, 481492 (2015).
Gheerbrant, E., Thomas, H., Roger, J., Sen, S. & Al-Sulaimani, Z. Deux nouveaux primates dans lOligocene inferieur de Taqah (Sultanat dOman): premiers adapiformes (?Anchomomyini) de la peninsule arabique? Palaeovertebrata 22, 141196 (1993).
Wiens, J. J. in Phylogenetic Analysis of Morphological Data (ed. Wiens, J. J.) Coding morphological variation within species and higher taxa for phylogenetic analysis (Smithsonian Institution Press, Washington, DC, 2000).
Lanfear, R. F. P., Wright, A. M., Senfeld, T. & Calcott, B. PartitionFinder 2: new methods for selecting partitioned models of evolution for molecular and morphological phylogenetic analyses. Mol. Biol. Evol. 34, 772773 (2016).
Roberts, T. E., Lanier, H. C., Sargis, E. J. & Olson, L. E. Molecular phylogeny of treeshrews (Mammalia: Scandentia) and the timescale of diversification in Southeast Asia. Mol. Phylogenet. Evol. 60, 358372 (2011).
We thank the National Museums of Kenya for access to the Propotto specimens figured and discussed here, and the Egyptian Mineral Resources Authority, the Egyptian Geological Museum, and the Egyptian Environmental Affairs Agency for supporting paleontological fieldwork in the Fayum area of Egypt that led to the recovery of the Plesiopithecus specimens figured and described here. The late Alan Walker graciously provided copies of his correspondence with George G. Simpson. This research was funded by Leakey Foundation grants to N.J.S.; National Science Foundation grants BCS-0416164 and BCS-0819186 to Elwyn L. Simons and E.R.S.; BCS-1231288 to E.R.S., D.M.B., G.F.G., and John G. Fleagle; BCS-1440742 to D.M.B. and G.F.G.; DBI-1458192 to GFG; BCS-1552848 to D.M.B.; and BCS-1127164 and BCS-1638796 to N.J.S; and DDIG 0925793 to Lynn Lucas and Lynn Copes. This is Duke Lemur Center publication #1400.
N.J.S. proposed reanalysis of Propotto to G.F.G. G.F.G., D.M.B., S.H., F.K.M., E.R.M., N.J.S., and E.R.S. designed the study; G.F.G., E.R.M., and E.R.S. created the first draft, which was reviewed and edited by all authors; S.H. and E.R.S. compiled and concatenated data matrices for phylogenetic analysis, ran phylogenetic analyses, and interpreted results; E.R.S. scored morphological characters, created digital surface models, and created Figs. 1, 2, 4, and 5; D.M.B. created digital surface models for the automated geometry analyses, ran the analyses, and created Fig. 3; S.H. designed the stepping-stone analyses and reconstructions of ancestral states, created Fig. 6, and created digital surface models; S.H. and H.M.S. micro-CT scanned specimens.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the articles Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
Gunnell, G.F., Boyer, D.M., Friscia, A.R. et al. Fossil lemurs from Egypt and Kenya suggest an African origin for Madagascars aye-aye. Nat Commun 9, 3193 (2018). https://doi.org/10.1038/s41467-018-05648-w
Travelling through the Central Highlands, Kenyas political and economic heartland, offers some great rewards. Mount Kenya, Africas second-highest peak, gave the colonial nation its name and presents plenty of scope for hiking. Walks lower down and in the Aberdare range are easier but still dramatic, with better chances of seeing wildlife. Travel itself is never dull here, and the range of scenery is a spectacular draw in its own right: primary-coloured jungle and shambas, pale, windswept moors and dense conifer plantations, all with a mountain backdrop. People everywhere are friendly and quick to strike up a conversation, the towns are animated and the markets colourfully chaotic. Most roads are in good shape, and bus and matatu journeys are invariably packed with interest and amusement.
After the main game-viewing areas and the coast, the circuit provided by the Mount Kenya ring road is one of the most travelled in Kenya, and there are always a few tourist vehicles to be seen. Apart from the high forests, moors and peaks, little of this remains wild country, with shambas steadily encroaching upon the ridges. The Kikuyu, Meru and Embu peoples have created an extraordinary spectacle of cultivation on the steep slopes, gashed by the road to reveal brilliant red earth.
As you travel, the mountain is a constant, looming presence, even if you cant often see much of it. With a base 80km across, Mount Kenya is one of the largest free-standing volcanic cones in the world. The twin peaks are normally obscured by clouds, but early in the morning and just before sunset the shroud can vanish suddenly, leaving them magically exposed for a few minutes. To the east and south, the mountain drops steeply away to the broad expanse of Ukambani (Kamba-land) and the Tana River basin. Westwards, and to the north, it slopes away more gently to the rolling uplands of Laikipia.
The Aberdare range, which peaks at 4001m, is less well known than Mount Kenya. The lower, eastern slopes have long been farmed by the Kikuyu (and more recently by European tea and coffee planters), and the dense mountain forests covering the middle reaches are the habitat of leopard, buffalo, some six thousand elephants and a few small herds of critically endangered bongo antelope. Above about 3500m, lions and other open-country animals roam the cloudy moorlands. Melanistic forms, especially of leopard, but also of serval cat and even bushbuck, are also present.
The park stretches 60km along the length of the peaks, with the Salient on the lower slopes reaching out east. Like Mount Kenya National Park, it attracts the worst of the weather: rainfall up here is high, often closing the park to vehicles in the wet season, although in the Salient the tree-hotel game lodges The Ark and Treetops stay open all year. The towns of Naivasha and Nyeri are the usual bases. Nyahururu, the other important town in the region, has Thomsons Falls as a postcard attraction, and is also the setting-off point for a wild cross-country journey to Lake Bogoria, 1500m below, in the Rift Valley. Also from Nyahururu begins the main route to Maralal and Loiyangalani on the eastern shore of Lake Turkana.
The Central Highlands are utterly central to Kenyan history. The majority of British and European settlers carved their farms from the countryside around Mount Kenya. Later, and as a direct consequence, this was the region that saw the development of organized anti-colonial resistance culminating in Mau Mau.
Until independence, the fertile highland soils (A more charming region is not to be found in all Africa, thought Joseph Thomson, exploring in the 1880s) were reserved largely for Europeans and considered, in Governor Eliots breathtaking phrase, White Mans Country. The Kikuyu peoples, skilled farmers and herders, had held the land for several centuries before the Europeans arrived. They were at first mystified to find themselves squatters on land whose ownership, in the sense of exclusive right, had never been an issue in traditional society. They were certainly not alone in losing land, but, by supplying most of the Mau Mau fighters for the Land and Freedom Army, they were placed squarely in the political limelight. In return, they received a large proportion of what used to be known as the Fruits of Independence. Today, most of the land is in African hands again, and it supports the countrys largest rural population. Theres intensive farming on almost all the lower slopes and much of the higher ground as well, beneath the national parks of Mount Kenya and the Aberdare.
Mount Kenyas lush, green foothills are a beautiful place to go horseriding. A good option for a guided horseback tour is German family-run Sandai, 11km northeast of Mweiga (0733 734619), a charming and relaxing rural homestay, with comfortable and very attractive rooms. Besides horseriding, activities include day hikes to the Aberdares, overnight excursions, painting and yoga. The five self-catering cottages are also beautifully furnished, with fireplaces and kitchens. To get here from Mweiga, head north for 4.2km from the town centre note the white tyre in the earth on the east side of the road, marked Sandai 7km (beneath a sign announcing St Joseph Mahiga Secondary School). Follow this for 5.3km, turn left and after a further 600m right onto a road marked by a white tyre planted in the earth, marked Sandai 7km.
Kenyas main highland forests are on Mounts Kenya, Elgon and Marsabit, on the Aberdare range and on the Mau Escarpment. The characteristic natural landscape in the highlands is patches of evergreen trees separated by vast meadows of grasses often wire grass and Kikuyu grass. The true highland forest, typically found only above 1500m, contains different species of trees from lowland forest, and does not normally grow as tall or dense. Typical species include camphor, Juniperus procera (the East African cedar) and Podocarpus. The better-developed forests are found on the wetter, western slopes of the highlands. Above the forest line, at altitudes of 2500m and higher, you get stands of giant bamboo, while along the lower, drier edges of the highlands, the stands of trees tend to be interspersed with fields of tall grass, where you commonly also find various species of olive.
The ancestors of the Kikuyu migrated to the Central Highlands between the sixteenth and eighteenth centuries, from northeast of Mount Kenya. Stories describe how they found various hunter-gatherer peoples already in the region (the Gumba on the plains and the Athi in the forests), and a great deal of intermarriage, trade and adoption took place. The newcomers cleared the forests and planted crops, giving the hunters gifts of livestock, honey or wives in return for using the land.
Likewise, there was trade and intermarriage between the Kikuyu and the Maasai, both peoples placing high value on cattle ownership, with the Maasai depending entirely on livestock. During bad droughts, Maasai would raid Kikuyu herds, with retaliation at a later date being almost inevitable. But such intertribal warfare often had long-term benefits, as ancient debts were forever being renegotiated and paid off by both sides, thus sustaining the relationship. Married Kikuyu women enjoyed a special immunity that enabled them to organize trading expeditions deep into Maasai-land, often with the help of a hinga, a middleman, to oil the wheels.
Like the Maasai, the Kikuyu advanced in status as they grew older, through named age-sets and rituals still important today. For Kikuyu boys, circumcision marks the important transition into adulthood (female circumcision, or clitoridectomy, is illegal and rarely performed today). In the past, boys would grow their hair and dye it with ochre in the style of Maasai warriors (in fact, the Maasai got their ochre from the Kikuyu, so it may really have been the other way around). They also wore glass beads around their necks, metal rings on their legs and arms, and stretched their ear lobes with earplugs. Women wore a similar collection of ornaments and, between initiation and marriage, a headband of beads and discs, still worn today by most Maasai women.
Traditionally, the Kikuyu had no centralized authority. The elders of a district would meet as a council and disputes or important decisions would be dealt with in public, with a party to follow. After their deaths, elders now known as ancestors continued to be respected and consulted. Christianity has altered beliefs in the last few decades, though many churchgoers still believe strongly in an ancestor world where the dead have powers over their living descendants. The Kikuyu traditionally believed that the most likely abode of God (Ngai), or at least his frequent resting place, was Mount Kenya, which they called Kirinyaga (Place of Brightness). Accordingly, they used to build their houses with the door always looking out towards the mountain, hence the title of Jomo Kenyattas book, Facing Mount Kenya.
Today, the Kikuyu are at the forefront of Kenyan development and, despite entrenched nepotism, are accorded grudging respect as successful business people, skilled media operators and formidable politicians. There is considerable political rivalry between the Kiambu Kikuyu of the tea- and coffee-growing district north of Nairobi and the Nyeri Kikuyu, based in the fertile area of Othaya who rely on a more mixed economy.
The GEMA (Gikuyu, Embu and Meru Association), created in 1971 to further Kikuyu interests, at first concerned itself primarily with countering Daniel Arap Mois ascent to the presidency, and although it was banned in 1980, it is believed to continue to operate clandestinely throughout Kenya.
The emergence in Kikuyuland in the early 2000s of the secret and violent Mungiki cult, somewhat modelled after the colonial eras Mau Mau independence movement but based primarily around extortion and gangster operations rather than emancipation, brought terror to slum districts in parts of Central Kenya. In a twist of jaw-dropping chutzpah, its leader escaped justice, declaring himself a born-again Christian. He is now wooed by mainstream politicians, while Mungiki has become a deeply corrupting force within the political process.
Throughout Kenya, and especially in the Central Highlands and on the coast, youll often see people selling and chewing what looks like a bunch of twigs wrapped in a banana leaf. This is miraa, more commonly known abroad by its Somali name qat, a natural stimulant that is particularly popular among Somalis, Somali Kenyans and Yemenis. The shrub (Catha edulis) grows in the hills around Meru (the world centre for its production), and the red-green young bark from the shrubs new shoots is washed, stripped with the teeth and chewed, with the bitter result being something of an acquired taste (its sometimes taken with bubble gum to sweeten it). Miraa contains an alkaloid called cathinone, a distant relative of amphetamine, with similar effects, though you have to chew it for some time before youll feel them. When they do kick in, they include a feeling of alertness, ease of conversation and loss of appetite. Long-term daily use can lead to addiction. Its not always looked upon favourably, with signs prohibiting the chewing of it in many hotels and bars.
Miraa comes in bundles of a hundred sticks called kilos (not a reference to their weight) and various qualities, from long, twiggy kangeta, which is the ordinary, bog-standard version, to short, fat gisa kolombo, which is the strongest. As it loses its potency within 48 hours of picking, its wrapped in banana leaves and transported at speed. Street stalls selling it often display the banana leaves to show that they have it, and the best place to buy miraa in many towns is where the express matatus arrive from Meru. The use of miraa by bus, truck and matatu drivers goes a long way towards explaining why they have so many accidents. There are no legal restrictions on the use of miraa in Kenya, although imams have issued a fatwa (legal judgement) condemning it as an intoxicant, like alcohol, which means that it is forbidden to true believers. In fact, in most countries (but not the UK), miraa is a controlled narcotic, the possession of which is a criminal offence.
North of Naro Moru, the A2 runs across the yellow-and-grey downs, scattered with stands of tall blue gums, roamed by cattle and overflown by brilliant roller birds, before dropping to NANYUKI, the gateway to Laikipia and parts of northern Kenya. You might be forgiven for expecting something momentous to take place at the equator, just south of town. Theres a sprouting of curio shops and signs (This sign is on the Equator) and even an Equator Professor who claims to demonstrate the Coriolis effect of the earths rotation using a bucket of water and a matchstick (aided by sleight of hand). In the northern hemisphere a large body of still water in a perfectly formed vessel would gurgle through a plug hole anticlockwise, whereas in the southern hemisphere it would flow clockwise though in practice the direction of flow is controlled by the operator because the Coriolis effect is too tiny to have an impact, especially anywhere near the equator itself where the effect is zero. The demonstration is free; the certificate comes for a fee.
Nanyuki has the dual distinction of being Kenyas air-force town and playing host to the British Armys training and operations centre. And although in recent decades it has taken in thousands of refugees, escaping from rural poverty and ethnic violence, it remains very much a country town in atmosphere, and an oddly charming one.
A wide, tree-lined main street and the mild climate lent by its 2000m altitude bestow an unfamiliar, cool spaciousness that seems to reinforce its colonial character. Yet the town is becoming popular with foreign and Kenyan investors, and real-estate prices have doubled in recent years. The towns modern Nakumatt supermarket and assorted new coffee shops and restaurants are a sign of things to come.
The first party of settlers arrived in the district in 1907 to find several old Maasai manyattas, a great deal of game and nothing else. Nanyuki is still something of a settlers town and European locals are always around. The animals, sadly, are not. Although you may see a few grazers on the plains, the vast herds of zebra that once roamed the banks of the Ngare Nanyuki (Maasai for Red River) were decimated by hunters seeking hides, by others seeking meat (particularly during World War II, when eighty thousand Italian prisoners of war were fed a pound of meat each day), but most of all by ranchers protecting their pastures.
Like Nanyuki, NYAHURURU is almost on the equator, and it shares much of Nanyukis character. Its high up (at 2360m, Kenyas highest town), cool and set on open savanna lands with patches of indigenous forest and plenty of coniferous plantation. Since the B5 road to Nyeri was completed, Nyahururu has been less cut off, but its still something of a frontier town for routes north to Lake Turkana and the desert. A tarmac road goes out as far as Rumuruti and then the fun begins.
Joseph Thomson gave the town its original name when he named the nearby waterfall after his father in 1883. Many still call it T. Falls, and not just the old settlers as you might expect. Thomsons Falls was one of the last settler towns to be established. The first sign of urbanization was a hut built by the Narok Angling Club in the early 1920s to allow its members to fish for the newly introduced trout in the Ewaso Narok, Pesi and Equator rivers. In 1929, when the railway branch line arrived, the town began to take shape. The line has closed now, but the hotel built in 1931, Thomsons Falls Lodge, is still going strong, and Nyahururu remains an important market town, and not really a tourist centre. The market is well worth a browse, especially on Saturdays. It sprawls out over most of the district west of the stadium, an indication of the towns rapid growth over the last couple of decades.
On the northeast outskirts of town, Thomsons Falls are pretty rather than spectacular, though they can be dramatic when the Ewaso Narok is in flood after heavy rain. The falls are a popular stopoff for tourists travelling between Samburu and Maasai Mara game reserves, and the hotel lawns above the falls get crowded with picnickers from town at weekends. Uniformed council officials have taken to extracting an entrance fee of Ksh200 from unwary tourists: only pay if they can give you an official ticket or receipt, otherwise tell them you have business at the hotel, whose grounds overlook the falls. The path leading down to the bottom of the 75m falls is somewhat dangerous, especially when wet, and you should ensure there have been no recent incidents of robbery. Dont attempt to climb up again by any other route, because the cliffs are extremely unstable.
The self-styled capital of Kikuyu-land a title the Kikuyu of Kiambu might dispute NYERI is the administrative headquarters of Nyeri County and a lively, chaotic and friendly highland town, whose name derives from the Maa word nyiro, meaning reddish brown, after its earth. An attractive trading centre, it nestles in the green hills where the broad vale between Mount Kenya and the Aberdare range drops towards Nairobi. Tumultuous markets, scores of dukas, and even a few street entertainers, lend it an air of irrepressible commercialism.
Another former British military camp, Nyeri emerged as a market town for European coffee growers in the hills and for settlers on the ranching and wheat farms further north. Nyeri was also the last home of Robert Lord Baden-Powell, founder of the worldwide scouting movement, whose cryptically named Paxtu cottage, now a small museum, stands in the grounds of the Outspan Hotel and whose grave and memorial are to be found on the north side of town in the cemetery.
Test the class III, IV and V whitewater rapids of Kenyas longest river, the Sagana, at Savage Wilderness Camp (0737 835963), an adventure activities base specializing in rafting. The camp also offers a 60m bungee jump over the river, nature walks, zipline and an artificial rock-climbing wall. The shaded, grassy campsite is entirely powered by a home-made hydroelectric system. Theres plenty of space to pitch your own tent; other options include renting a small tent (with bedding and mattress) or a large one (with bedding and two camp beds), or getting a double cottage.
Copyright 2020 Apa Digital AG, all rights reserved. Rough Guides is a trademark owned by Apa Group with its headquarters at 7 Bell Yard, London WC2A 2JR, United Kingdom This site is owned by Apa Digital AG, Bahnhofplatz 6, 8854 Siebnen, Switzerland