Revealing the true identity of Zenkerella insignis, one of Africa’s most mysterious mammals

PictureThe enigmatic Cameroon flightless
anomalure, Zenkerella insignis.
My fascination with Africa’s endemic anomaluroid rodents — also known as "anomalures", or by the misnomer “scaly-tailed squirrels” (they do have scaly tails, but are not closely related to squirrels) — began about 15 years ago, when my colleagues and I first started to discover fossil bones and teeth of 37 million-year-old anomaluroids at what was then a newly-discovered site, “Birket Qarun Locality 2” (BQ-2), in the Fayum area of northern Egypt. At the time I couldn’t claim to have any expertise on anomalures, or even rodents generally, but I knew just enough about the fossil record of rodent evolution to be intrigued by the new specimens. For some reason the BQ-2 rodent fauna was radically different from those documented at younger fossil sites in the Fayum area — many of which were exceptionally rich, but none of which had ever yielded fossil anomaluroids. Instead, those younger sites are dominated by early “hystricognaths”, a large rodent group that ultimately gave rise to familiar species such as guinea pigs, porcupines, and mole-rats (1,2). Clearly, something had happened in northern Africa toward the end of the Eocene epoch (around 34 million years ago), that led to the successful diversification of hystricognaths and the local extinction of anomalures, the latter of which are still alive today in the equatorial forests of West and Central Africa, including those on Bioko, an island off the coast of Cameroon that is part of Equatorial Guinea. 

PictureHesham Sallam at BQ-2.
Fortunately it wasn’t long after we had found the new anomalure fossils that an Egyptian student, Hesham Sallam (image to the right, now Director of the Mansoura University Vertebrate Paleontology Center in Egypt, and a postdoctoral researcher at Duke University), applied to work with me on fossil mammals from the Fayum area. Hesham enthusiastically agreed to analyze the rodents from BQ-2 (and other sites) for his DPhil thesis at University of Oxford, and over the next several years, we worked together to describe the teeth and cranial parts of the two BQ-2 anomalures — a large species that we named Kabirmys (Arabic kabir = large, Greek mys = mouse, roughly translating to “big mouse” or “mighty mouse”) and a smaller species, Shazurus, whose teeth more closely resembled those of fossil anomaluroids known from early Miocene (~20 million-year-old) sites in Kenya and Uganda (3, 4). The teeth of Kabirmys and Shazurus were quite different, suggesting that they were probably derived from a very ancient common ancestor.

PictureAnomalurus pusillus
While the teeth of Kabirmys and Shazurus provided some good clues that helped us to figure out how these ancient species were related to other living and extinct anomaluroids, in some ways the most fascinating anomaluroid fossils that had been found at BQ-2 were the bones from these species' arms, hands, legs, and feet. This is because two of the living anomalure genera (the relatively large-bodied Anomalurus, see image to the left, and the tiny Idiurus) have anatomical adaptations that allow them to glide between trees, including a “patagium” (or gliding membrane), and we hoped that the fossil bones from BQ-2 might allow us to determine whether this remarkable mode of locomotion had already appeared by 37 million years ago. Among other things, if we could infer that the BQ-2 anomalures were gliders, they would be the only other mammals (aside from primates and possibly bats) that could be definitively tied to an arboreal habitat in this ancient community; indeed, they could have been hanging out in the trees right alongside our ancient monkey-like ancestors.

As research proceeded it became obvious, however, that the limb bones of Kabirmys were quite different from those of Anomalurus and Idiurus, suggesting that it probably was not a glider. But when we went to make comparisons with the third living anomalure, Zenkerella insignis, we ran into a major problem: despite having been discovered over a century ago, the species was still largely a mystery to scientists. Zenkerella was known to lack a patagium, and so was clearly not capable of gliding, but it had never been seen alive by trained mammalogists, and there was no information in the scientific literature about how it moved, what it ate, or when and where it spent its waking hours. Only eleven specimens were known in museum collections worldwide, and none of those specimens had been cleaned to reveal what the bones of the arms, legs, pelvis, and spinal column looked like. Awkwardly, we suddenly knew more about the postcranial anatomy of 37 million-year-old Kabirmys than we knew about a living species! To make matters worse, Zenkerella's DNA had never been sequenced, and so nobody knew for sure how it was related to the two gliding anomalurids. Some researchers had argued on the basis of skull and tooth anatomy that Zenkerella was most closely related to Idiurus, but this implied that either Zenkerella had lost its patagium (and its ability to glide), or that Anomalurus and Idiurus had evolved patagia and gliding behavior independently (see image below). Both of these evolutionary scenarios frankly seemed unlikely to us, but without DNA or postcranial skeletons of Zenkerella, we were stuck -- we could describe the postcranial bones from BQ-2, but there would be no solid evolutionary context in which to place them.
Possible relationships among anomalures, and scenarios for the gain or loss of gliding adaptations. Modified from Coster et al. (5).
PictureDavid Fernández (left) of The
University of the West of England
Photo: G. McCabe
Our first major breakthrough came when I was attending the doctoral dissertation defense of Spanish primatologist and conservationist David Fernández, at Stony Brook University in New York in 2013 (David is now at the University of the West of England in Bristol). I was surprised to hear that David would soon be moving to Bioko (one of the areas where Zenkerella occurs) to become the co-Director of the Moka Wildlife Center, as part of the Bioko Biodiversity Protection Program (BBPP) that is run by Drexel University and the National University of Equatorial Guinea. We talked briefly about Zenkerella and the need to obtain new specimens and DNA sequences, and fortunately David was very excited to search for the species on Bioko, and to interview locals to find out more about the Zenkerella's distribution and habits. Not long after arriving on Bioko, David was able to confirm, after talking with village elders and soldiers, that Zenkerella had recently been trapped in ground snares near a village called Ureca, but that the bodies had been discarded because they are not valued for their meat. He was even told that in the local Bubi language some call Zenkerella “musuló”, which means “inferior to all squirrels”. The few people who were familiar with the species indicated that Zenkerella is nocturnal. David asked the local trappers in Ureca to hold on to any specimens that they found in their ground snares. And we waited.

PictureFernández with the first whole
-body specimen of Zenkerella
Photo: G. McCabe
For eight months, there was no news. But finally, only a few days before he was scheduled to leave Bioko for good, David went back to Ureca and was told by the chief of the village that a specimen had been found in one of their traps. He sent me images of the specimen and I could barely contain my excitement — indeed, there it was (image to the left), the first whole-body specimen known for the entire species whose preservation would finally allow us to easily study its DNA, postcranial bones, muscles, brain, gastrointestinal system, etc. Several months later, David wrote again with even better news — two more specimens had been found (including the specimen figured below), giving us a total of three whole-body specimens! Finally, as of just a few weeks ago, David has confirmed that we have yet another two individuals, male and female. All have been found in ground snares during the rainy season. With five specimens, including male and female individuals, we will be able to study intraspecific variation and sex differences as well. The individual figured below is a male.

The second male specimen of Zenkerella insignis, from the near the village of Ureca on Bioko. From Heritage et al (6).
PictureSteven Heritage, Stony Brook
University Ph.D. student.
Enter Steven Heritage (image to the right), my Ph.D. student at Stony Brook University who is studying the phylogenetic relationships of another fascinating and obscure group of small African mammals — macroscelideans (also known as elephant-shrews or “sengis”) — for his dissertation, using both DNA and morphology. Fortunately Steven was happy to lend his expertise and take on the task of sequencing the new specimens' DNA and comparing it to that of other rodents, so that Zenkerella could finally be placed securely in the anomaluroid family tree. After several months of hard work, Steven successfully sequenced five genes (three from the mitochondrial genome, and two from the nuclear genome) from two of the Zenkerella individuals, pulled hundreds of other homologous rodent DNA sequences off of GenBank (a public repository for DNA sequences), aligned all of those sequences, and ran multiple analyses to determine the placement of Zenkerella among a large sample of 66 other rodent species. 

PicturePhylogenetic analysis of mitochondrial and
nuclear gene sequences showing the
placement of Zenkerella in anomaluroid phylogeny.
The results were as surprising as they were decisive (image to the left). As detailed in the study of Heritage et al. (6), which was published today in the open access journal PeerJ, Steven’s analyses showed that Zenkerella was unequivocally not closely related to Idiurus as had previously been suggested, but it also wasn’t closely related to the other anomalurid glider (Anomalurus). Instead, Anomalurus and Idiurus were found to be “sister taxa” — that is, they were more closely related to each other than to Zenkerella. This grouping of Anomalurus and Idiurus had perfect statistical support, and sat at the end of a long common branch, suggesting an extended period of common ancestry to the exclusion of Zenkerella. This result has profound implications for understanding anomaluroid evolution and the origin of gliding adaptations: given these relationships, it is highly probable that the last common ancestor of Anomalurus and Idiurus was a glider, and since Zenkerella diverged from the Anomalurus-Idiurus lineage long before gliding evolved along their “stem lineage”, there is no reason to believe that there were any reversals back to the primitive condition as some had envisioned for Zenkerella — that is, gliding probably had an ancient single origin within anomalures, and (so far as we know based on the fossil record) it was never lost once it appeared. Interestingly, these results aligned perfectly with the relationships that had been proposed for anomaluroids well over a century ago by a Swedish zoologist named Tycho Tullberg, in his seminal study of rodents published in 1899 entitled “Ueber das System der Nagethiere: eine Phylogenetische Studie” (7). Tullberg’s hypothesis was, of course, based solely on anatomy.

PictureIsolated teeth of Prozenkerella saharaensis, a
30 million-year-old species from Libya.
From Coster et al. (5).
We still didn’t know what the time-scale of these evolutionary changes might have been, but available evidence from the fossil record gave us good reason to suspect that Zenkerella was an exceedingly ancient branch in the anomalure family tree. It had been known since 1973 that fossils attributed to the genus Zenkerella were present at ~20 million-year-old sites in East Africa (8), and, remarkably, a new piece of evidence emerged mid-way through our project — late in 2015, a team of paleontologists led by the French scientist Pauline Coster announced the discovery of an ancient extinct relative of Zenkerella in ~31 million-year-old rocks in central Libya (5). The fossil species was only known from isolated teeth (see image to the right), but they were so similar to those of Zenkerella that Coster and colleagues named the new species Prozenkerella (“before Zenkerella”) saharaensis. When we analyzed the anatomical data from these and other living and extinct anomalurid species alongside the DNA evidence, using a relatively new Bayesian phylogenetic approach known as “tip-dating”, which takes into account rates of evolution to provide estimates for divergences between living and extinct species, we found that Zenkerella was estimated to have diverged from other anomalures about 49 million years ago. Our results further suggested that Anomalurus and Idiurus last shared a common ancestor near the close of the Paleogene, about 26 million years ago.

To put such ancient divergences into context — we last shared a common ancestor with chimpanzees about 6-to-8 million years ago, and all living anthropoids (including New World monkeys, Old World monkeys, apes, and humans) last shared a common ancestor between 35 and 45 million years ago (image below). If the Zenkerella lineage diverged 49 million years ago, then it was likely distinct and evolving independently in Africa long before that anthropoid common ancestor even existed, and maybe even before anthropoids had appeared in Africa! If we look to other primate groups, the ancient divergence and persistence of Zenkerella is somewhat analogous to the case of the strange aye-aye (Daubentonia) of Madagascar that probably also diverged from its lemur relatives around the same time, early in the Eocene, and is now represented by only one species. And, if the teeth of fossil relatives of Zenkerella are a reliable guide, by ~31 million years ago Prozenkerella had already acquired the feeding strategies of modern-day Zenkerella; the teeth of the living form and the ~31 million-year-old form only differ in slight details. Available genetic and fossil evidence seem to be converging on the conclusion that Zenkerella is a “living fossil” — that is, a species that differs little in its anatomy from its ancient fossil relatives, and that, in this particular case, may well have persisted in its current niche for well over 31 million years.

A comparison of major divergences within the family trees of primates (left) and anomaluroid rodents (right).
The estimated 26 million-year-old divergence between Anomalurus and Idiurus is also surprisingly ancient (image above), roughly equal in age to the time period when Old World monkeys diverged from apes, in the late Oligocene. Our tip-dating analyses suggest that the extinct early Miocene glider Paranomalurus might have diverged from the lineage leading to Anomalurus and Idiurus about 30 million years ago; if this is the case, then gliding must have appeared at some point between 49 and 30 million years ago in Africa. We speculated that such adaptations might have evolved in response to the dramatic environmental changes that occurred world-wide around 34 million years ago, at the beginning of the early Oligocene, when Antarctic glaciation accelerated and environments in the northern latitudes became much more seasonal. This was also when anomaluroids stopped showing up in the fossil-bearing rocks of northern Egypt, suggesting that their latitudinal distribution had been constricted. This hypothesis can be tested through future paleontological work in the Eocene and Oligocene of Africa. 
Such ancient divergences were difficult to reconcile with the prevailing taxonomy, which placed all anomalures in the single family Anomaluridae. We proposed that the gliding forms (Anomalurus and Idiurus) be retained in the family Anomaluridae, and that Zenkerella be placed in its own family, Zenkerellidae. Together, the living anomalures were placed in a higher-level group, Anomaluroidea, that would also include at least one extinct family known from the Miocene, Nonanomaluridae, if not also the older and phylogenetically basal Nementchamyidae — the family to which Kabirmys belongs.

PictureLandscape around Ureca on Bioko (photo by David Fernández).
Despite the important progress that we have made in figuring out Zenkerella's position in anomaluroid phylogeny, we still have much to learn about this genus. Research is currently underway on the anatomy of this fascinating rodent, and is revealing some interesting primate-like features that presumably reflect a largely arboreal lifestyle. Yet we know that Zenkerella comes to the ground, at least occasionally, because it has only been found in ground traps. David Fernández recently returned to Bioko, and, in collaboration with our co-authors Drew Cronin (Drexel University and BBPP), and José Manuel Esara Echube (National University of Equatorial Guinea and BBPP), will soon be making important headway on new studies of Zenkerella’s distribution, with the goal of ultimately studying the species’ behavior, locomotion, and feeding ecology. With any luck, in the coming years there will be lots of interesting new information to report about this remarkable "living fossil" that has eluded scientists for well over a century.

References (with links to the original papers, if available):

(1) Wood, A. E. 1968. Early Cenozoic mammalian faunas, Fayum Province, Egypt, Part II: the African Oligocene Rodentia. Peabody Museum Bulletin 28: 23-205.

(2) Sallam, H.M., Seiffert, E.R. 2016. New phiomorph rodents from the latest Eocene of Egypt, and the impact of Bayesian “clock”-based phylogenetic methods on estimates of basal hystricognath relationships and biochronology. PeerJ 4:e1717.

(3) Sallam, H.M., Seiffert, E.R., Simons, E.L. 2010. A highly derived anomalurid rodent (Mammalia) from the earliest late Eocene of Egypt. Palaeontology 53:803-813.

(4) Sallam, H.M., Seiffert, E.R., Simons, E.L., Brindley, C. 2010. A large-bodied anomaluroid rodent from the earliest late Eocene of Egypt: Phylogenetic and biogeographic implications. Journal of Vertebrate Paleontology 30:1579-1593.

(5) Coster, P.M.C., Beard, K.C., Salem, M.J., Chaimanee, Y., Jaeger, J-J. 2015. New fossils from the Paleogene of central Libya illuminate the evolutionary history of endemic African anomaluroid rodents. Frontiers in Earth Science 3:56.

(6) Heritage, S., Fernández, D., Sallam, H.M., Cronin, D.T., Esara Echube, J.M., Seiffert, E.R. (2016) Ancient phylogenetic divergence of the enigmatic African rodent Zenkerella and the origin of anomalurid gliding. PeerJ 4:e2320.

(7) Tullberg, T. (1899) Ueber das System der Nagethiere: eine Phylogenetische Studie. Uppsala: Akademische Buchdruckerei.

(8) Lavocat, R. 1973. Les Rongeurs du Miocène d’Afrique Orientale, Miocène inférieur. Mémoires et Travaux de l’Institut de Montpellier de l’Ecole Pratique des Hautes Etudes 1:1-284.
PictureA mouse lemur (Microcebus) in
Madagascar. Image by Blanchard
Lemurs, including species such as the mouse lemur in the image to the right, are a diverse group of primates that is now restricted to Madagascar and a few nearby islands. But the common ancestor that lemurs shared with the other extant "strepsirrhine" primates — the lorises and bushbabies — probably lived in the early Eocene, ~50-56 million years ago (and possibly even earlier) on what was then "Afro-Arabia", an isolated continent made up of the conjoined African and Arabian tectonic plates (1). An ancient ancestor of Asia's modern lorises presumably migrated from Arabia much later, perhaps during the early Miocene (~16-~23 million years ago), as that plate's eastern margin slowly made contact with the conjoined Asian and Indian plates (2). At least this seems to be the most likely scenario given the evidence that we have available; the early evolution of strepsirrhines is still shrouded in mystery, presumably because Afro-Arabia's Eocene fossil record has been so poorly sampled. 

PictureJaw of Djebelemur, an early relative of strepsirrhines
from Tunisia. Modified from Marivaux et al. (3).
Regardless, most of the potential close fossil relatives of the living strepsirrhines (a group that is also known as the "toothcombed" primates, for the comb-like shape of their lower incisor and canine teeth) are from Africa. Included among these early African species is a tiny primate named Djebelemur, from an early Eocene site in Tunisia [image to the left (ref. 3)]. The only known remains of Djebelemur — jaws, teeth, ankle bones, and ear bones — are all consistent with it being a strepsirrhine, and yet it lacks the toothcomb, suggesting that it is not the common ancestor of the living species, but rather an extinct side branch. Also present at sites of a similar age in Algeria are remains of other tiny strepsirrhines known as azibiids (Azibius and Algeripithecus). The ankle bones of Azibius closely resemble those of Djebelemur, but the arrangement of cusps and crests on its teeth is remarkably specialized for a species of such great antiquity (3).

But where did these early African strepsirrhine primates come from? Are they the descendants of a much more ancient Afro-Arabian lineage, or are they immigrants from some nearby landmass? The closest relatives of strepsirrhines are the tarsiers and anthropoids (monkeys, apes, humans), followed more distantly by two orders of non-primate mammals — flying lemurs and treeshrews. All of these groups have early fossil records in Asia, and on this basis it has been argued that the "stem lineage" of strepsirrhines (that is, the long line of successive ancestral populations that ultimately gave rise to the most recent common ancestor of lemurs, lorises, and bushbabies) probably also traces back to Asia (1). To be fair, though, not everybody agrees on this scenario. And there are, in fact, no known Djebelemur-like primates on either the Asian plate or on the Indian plate in the early Eocene (when those two tectonic plates were first coming into contact).

PictureA lower jaw of Anchomomys from Egerkingen,
Switzerland, held in the collections of
the Naturhistorisches Museum Basel.
Instead, the non-African fossil species whose teeth most closely resemble those of Djebelemur are anchomomyins — a group of tiny primates that lived in the middle and late Eocene of Europe, and that also lacked toothcombs. Anchomomys was first described in 1916, by the Swiss paleontologist Hans Stehlin (4), based on fragmentary jaws from middle Eocene sites in the Egerkingen area of Switzerland (which I discussed in an earlier post), but since that time teeth of several close relatives have been discovered, most recently in northeast Spain (5-7). Traditionally, anchomomyins have been linked to other fossil primates known from the Eocene of Europe, such as middle Eocene Europolemur and Protoadapis (8), but they are odd in being considerably smaller than their alleged relatives, and in having no older relatives that conclusively link them to other European species.

PictureA calcaneus bone of Anchomomys
compared to that of a mouse lemur
Mirza and an older strepsirrhine relative,
Asiadapis. Note the elongation of the
calcaneus in Anchomomys and Mirza.
One reason why anchomomyins have remained so mysterious over the course of the last century is that they have long been known only from partial jaws and isolated teeth. However early in the 1990s, work led by the Spanish paleontologist Salvador Moyà-Solà, at a ~42 million-year-old site called "Sant Jaume de Frontanyà-3C" in northeastern Spain, resulted in the recovery of numerous postcranial bones of a species named Anchomomys frontanyensis (9). Surprisingly, these fossils showed that, unlike its purported European relatives, Anchomomys was characterized by elongation of the end of the calcaneus bone (see image to the left) — a feature that is today seen only in small leaping prosimians, including bushbabies and some lemurs. In a recently published paper in Journal of Human Evolution (10) that I co-authored with lead author Judit Marigó, as well as Imma Roig, Moyà-Solà, and Doug Boyer, the ankle bones of Anchomomys are described in detail for the first time, and the information from these bones is incorporated into an analysis of relationships among early primates.

One important conclusion of our study is that the astragalus (or talus) bones of Anchomomys differ very little from those of Azibius and Djebelemur. Another important conclusion is that our analyses of primate relationships place anchomomyins closer to azibiids, Djebelemur, and living strepsirrhines than to any European species. This result raises the intriguing possibility that African strepsirrhines might be derived from a much older anchomomyin-like ancestor that lived on the Iberian Peninsula in the earliest Eocene. This part of the strepsirrhine family tree is, however, still poorly resolved, and instability in the placement of anchomomyins relative to African species leaves open the possibility that movement was not from Europe to Africa, but rather from Africa to Europe -- that is, perhaps anchomomyins might be derived from the early African radiation of non-toothcombed strepsirrhines that also gave rise to azibiids and Djebelemur.

Another possibility that must be entertained is that our phylogenetic results are entirely due to convergent evolution in the teeth and ankle bones of these early primates, which might have occurred if similar habitats were available to small proto-strepsirrhines in northwestern Africa and the Iberian Peninsula during the Eocene. Interestingly, a distantly related group of small primates known as microchoerines — remains of which have also been found in Spain alongside those of Anchomomys -- also evolved particularly long calcaneus bones, which (by analogy with living primates) presumably facilitated acrobatic leaping between trees. Selection pressures clearly favored the evolution of leaping adaptations in multiple primate lineages that were present in Europe during the Eocene. 

Digital models of the Anchomomys tarsal bones can be viewed (with registration) on MorphoSource, where they can be compared with models of astragali and calcanei belonging to hundreds of other living and extinct primate species.

References (with links to the original papers, if available):

(1) Seiffert E.R. 2012. Early primate evolution in Afro-Arabia. Evolutionary Anthropology 21: 239-253.

(2) Seiffert E.R. 2007. Early evolution and biogeography of lorisiform strepsirrhines. American Journal of Primatology 69: 27-35.

(3) Marivaux L., Ramdarshan A., Essid E.M., Marzougui W., Ammar H.K., Lebrun R., Marandat B., Merzeraud G., Tabuce R., Vianey-Liaud M. 2013. Djebelemur, a tiny pre-tooth-combed primate from the Eocene of Tunisia: A glimpse into the origin of crown strepsirhines. PLoS ONE 8: e80778.

(4) Stehlin H.G. 1916. Die Säugetiere des schweizerischen Eocaens. Critischer Catalog der Materialen. Siebenter Teil, zweite Hälfte: Caenopithecus--Necrolemur--Microchoerus--Nannopithex--Anchomomys--Periconodon--Amphichiromys--Heterochiromys —Nachträge zu Adapis —Schlussbetrachtungen zu den PrimatenAbhandlung der Schweizerischen Paläontologischen Gesellschaft 41 :1299–1552.

(5) Marigó J., Minwer-Barakat R., and Moyà-Solà S. 2010. New Anchomomyini (Adapoidea, Primates) from the Mazaterón Middle Eocene locality (Almazán Basin, Soria, Spain). Journal of Human Evolution 58:353-361. 

(6) Marigó J., Minwer-Barakat R., and Moyà-Solà S. 2011. New Anchomomys (Adapoidea, Primates) from the Robiacian (Middle Eocene) of northeastern Spain. Taxonomic and evolutionary implications. Journal of Human Evolution 60:665-672.

(7) Marigó J., Minwer-Barakat R., and Moyà-Solà S. 2013. Nievesia sossisensis, a new anchomomyin (Adapiformes, Primates) from the early late Eocene of the southern Pyrenees (Catalonia, Spain). Journal of Human Evolution 64:473–485.

(8) Godinot M. 1998. A summary of adapiform systematics and phylogeny. Folia Primatologica 69 (Suppl. 1): 218-249.

(9) Moyà-Solà S, and Köhler M. 1993. Middle Bartonian locality with Anchomomys (Adapidae, Primates) in the Spanish Pyrenees - preliminary report. Folia Primatologica 60:158-163. 

(10) Marigó J., Roig I., Seiffert E.R., Moyà-Solà S., and Boyer D.M. 2016. Astragalar and calcaneal morphology of the middle Eocene primate Anchomomys frontanyensis: Implications for early primate evolution. Journal of Human Evolution 91:122-143.

PictureA boomslang, Dispholidus typus (Colubridae)
in Tanzania. Photo by William Warby.
Presumably humans have always had a complex relationship with snakes; some cultures revere and worship them, others see them as evil, and the estimated one-third of the population that suffers from ophiophobia (fear of snakes) simply want nothing to do with them. Love them or hate them, there is just something about their shape and movement that freaks us out, and we are remarkably good at detecting snakes (or just snake-like objects) in our peripheral vision and amongst camouflage (1, 2). These and other observations underpin the controversial hypothesis that past predation by snakes has, over the course of tens of millions of years of evolution, played a key selective role in shaping parts of the primate brain that relate to visual threat detection (3). It has even been argued that perhaps “venomous snakes were such an important selective pressure favoring greater visual specialization in primates that they were ultimately responsible for the emergence of anthropoids” (Isbell, ref. 3, p. 12). It is true that snakes eat non-human primates (at least occasionally), and we all know stories about humans who have been bit by snakes, but what do we really know about the snakes that lived alongside the earliest members of the primate group that now includes monkeys, apes, and humans? Until very recently, almost nothing.

Today Africa is home to some of the deadliest venomous snakes on Earth, all of which belong to a large group known as Colubroidea, a superfamily that includes most living snake species. The majority of colubroids are non-venomous, but counted among its members are undeniably scary venomous African species such as puff adders (genus Bitis, a member of the family Viperidae), boomslangs (Dispholidus, a member of the family Colubridae, seen in the picture above), and elapids such as cobras (genus Naja) and mambas (genus Dendroaspis). The ancestral lineage that gave rise to all living colubroids probably originated in Asia very early in the Cenozoic, but colubroids have since undergone a remarkable diversification and are now found on all continents aside from Antarctica.

There is growing evidence that anthropoids probably also arose in Asia early in the Cenozoic, but, at some point in the Eocene, an ancestral population of anthropoids made it across the ancient Tethys Sea onto the then-isolated Afro-Arabian landmass (see for instance Seiffert (ref. 4), among others). The group subsequently diversified into parapithecoids (an extinct side branch), platyrrhines (the New World monkeys), and catarrhines (the Old World monkeys, apes, and humans). The fossil evidence for this phase in our shared evolutionary history with other anthropoids is best-documented in an area of northern Egypt known as the Fayum Depression, where there are multiple fossil sites that range from 37 to 29 million years in age. However, snake fossils are rare at the Fayum sites, and in fact the only species that have been described, way back in 1901 -- the giant madtsoiid Gigantophis garstini and the palaeophiid Pterosphenus schweinfurthi (both non-colubroid) -- were found in near-shore marine deposits that do not preserve anthropoid fossils (5). In light of this, for the last 115 years we had no evidence that the earliest African anthropoids lived alongside colubroids at all.

PictureSearching for fossil snakes and monkeys,
among other species, at BQ-2.
Photo by Hesham Sallam.
Thanks to more recent paleontological work in the Paleogene of Egypt and Tanzania, we finally know more about the early African record of Colubroidea. And, of particular significance for Isbell's hypothesis, we know more about snakes whose fossil remains occur right alongside those of the oldest undoubted African anthropoids. In a recent paper (6) in Journal of Vertebrate Paleontology that I co-authored with snake expert Jacob McCartney of SUNY Geneseo, a diverse snake fauna from the 37 million-year-old Fayum Birket Qarun Locality 2 ("BQ-2", image to the left) is described. BQ-2 has also yielded remains of several primates: the early anthropoid Biretia, the adapiform Afradapis, the lorisiforms Karanisia and Saharagalago, and the enigmatic possible anthropoid Nosmips (ref. 4; in addition to a few additional species, including anthropoids, that have not yet been described). This is the only example, from the entire Eocene epoch (34-56 million years ago) of Africa, of snakes and primates co-occurring at the same locality.

So what can be said of relevance to the "Snake Detection Theory" in light of what has been found at BQ-2? A key piece of Isbell's hypothesis is that early anthropoids lived alongside venomous snakes. The BQ-2 snake fauna provides no positive support for that hypothesis. Only one colubroid snake is present, the new genus and species Renenutet enmerwer, and it accounts for about 15% of the recovered snake remains, but it cannot be assigned to a particular lineage of colubroids -- instead it is simply a generalized species that can only be excluded from certain colubroid groups with venomous members (elapids, viperids) based on the absence of the specialized features that are seen in those groups. This shouldn't be taken to mean that there weren't potential primate predators in the BQ-2 snake fauna. Also present was the enormous Gigantophis, adults of which might have been 9-11 meters long, and the fauna is dominated by booids (which probably would have been similar to living medium-sized boas or pythons). Also present, strangely enough, is a tropidophiid -- a group that is now represented solely by the "dwarf boas" of the New World (West Indies, Central and South America; see image below), but that is also known from older fossil members in Europe. Perhaps tropidophiids can be added to the small list of land animals (including platyrrhine anthropoids and caviomorph rodents) that somehow dispersed across the South Atlantic, from Africa to South America, during the Eocene.

The dwarf boa Tropidophis melanurus, in Cuba. Picture by Jerry Oldenettel.
The presence of a generalized colubroid in the late Eocene of Africa raises a number of questions that are difficult to answer given available material, most obvious of which is -- where does Renenutet fall in colubroid phylogeny? The molecular phylogeny of snakes published by Pyron et al. (7) is consistent with an Asian origin of Colubroidea, and multiple independent dispersals of colubroids into Africa, but when did these different dispersals occur? Some might have been relatively recent, but one diverse group that could be of ancient African origin is Lamprophiidae -- a family that Renenutet cannot be excluded from, or included in, based on vertebral morphology. The recent "tip-dating" analysis of Hsiang et al. (8) suggested an Eocene divergence of lamprophiids from elapids, perhaps close in time to the deposition of BQ-2, but the Pyron et al. phylogeny suggests that the common ancestor of that clade was Asian, not African. This result is again not necessarily consistent with the presumed coexistence of anthropoids and venomous snakes in the Eocene of Africa.

But even if there were no venomous snakes in Africa during the Eocene or the early Oligocene, that period of isolation might have been short-lived (in geological terms). Thanks to the work of McCartneyNancy Stevens, and Patrick O'Connor, we now know that 12 million years after the deposition of BQ-2, at about 25 million years ago in the late Oligocene, there was a more diverse colubroid fauna present in Tanzania (9). The fossil snakes that they described were found in the Rukwa Rift Basin, which has also yielded the oldest fossil remains of apes (Hominoidea) and Old World monkeys (Cercopithecoidea). Notably, included among the Rukwa colubroids is a member of the family Elapidae, living members of which have hollow fangs that are used to inject venom (though note that the fossils from Tanzania are, like most fossil snakes, only known from vertebrae, so we don't know if these species had hollow fangs). In contrast to the BQ-2 fauna, these discoveries provide positive evidence that the earliest hominoids and cercopithecoids likely coexisted with venomous snakes in the late Oligocene of Tanzania, though they have not yet been documented to co-occur at the same sites.

As always, we need more fossils (and ideally much more complete fossils), from different parts of Africa and different time periods, to more thoroughly test the hypothesis that early African anthropoids co-existed with venomous snakes; it is, after all, a huge continentAnd clearly it is time for a taxonomically broad analysis of divergence dates within Colubroidea (ideally including fossils and using the same "tip-dating" methods employed by Hsiang et al.), with attention paid to the time and place of origin of venom delivery in the various colubroid lineages that evolved that adaptation. It may well be that there is little reason to expect that venom delivery had evolved in any colubroid family by the Eocene. Another key question would be whether diurnal arboreal anthropoids would have regularly encountered these venomous snakes at all -- Hsiang et al. (8) found that elapids and colubrids were likely to have been ancestrally diurnal, but were they ancestrally arboreal or terrestrial?

References (with links to the original papers):

(1) Lean Q.V., Isbell L.A., Matsumoto J., Nguyena M., Horia E., Maiorc R.S., Tomazc C., Trana A.H., Ono T., and Nishijoa H. 2013. Pulvinar neurons reveal neurobiological evidence of past selection for rapid detection of snakes. Proceedings of the National Academy of Sciences, U.S.A. 110: 19000–19005.

(2) Soares S.C., Lindström B., Esteves F., and Öhman A. 2014. The hidden snake in the grass: Superior detection of snakes in challenging attentional conditions. PLoS ONE 9: e114724.

(3) Isbell L.A. 2006. Snakes as agents of evolutionary change in primate brains. Journal of Human Evolution 51: 1-35.
(4) Seiffert E.R. 2012. Early primate evolution in Afro-ArabiaEvolutionary Anthropology 21: 239-253.

(5) Andrews C.W. 1901. Preliminary note on some recently discovered extinct vertebrates from Egypt (Part II). Geological Magazine 8: 436-444.

(6) McCartney J.M. and Seiffert E.R. 2016. A late Eocene snake fauna from the Fayum Depression, Egypt. Journal of Vertebrate Paleontology e1029580.

(7) Pyron R.A., Burbrink F.T., and Wiens J.J. 2013. A phylogeny and revised classification of Squamata, including 4161 species of lizards and snakes. BMC Evolutionary Biology 13: 1-53.

(8) Hsiang A.Y., Field D.J., Webster T.H., Behlke A.D.B., Davis M.B., Racicot R.A., and Gauthier J.A. 2015. The origin of snakes: revealing the ecology, behavior, and evolutionary history of early snakes using genomics, phenomics, and the fossil record. BMC Evolutionary Biology 15: 1-22.

(9) McCartney J.M., Stevens N.J., and O'Connor P.M. 2014. The earliest colubroid-dominated snake fauna from Africa: Perspectives from the late Oligocene Nsungwe Formation of southwestern Tanzania. PLoS ONE 9: e90415.
PictureLudwig Rütimeyer (1825-1895)
Almost all primates — a group of mammals that includes humans as well as apes, monkeys, tarsiers, lemurs, lorises, and bushbabies — live in warm and forested regions of either Asia, Africa, Madagascar, or South America. Primate species that have managed to survive in cold climates are exceedingly rare exceptions to a general pattern of tropical living that has persisted within the group since its members first appeared in the fossil record about 56 million years ago, at the beginning of the Eocene epoch (a division of the geological timescale that lasted about 22 million years, ending around 34 million years ago). This hasn't always been obvious; paleontologists have slowly pieced together the evidence for this conclusion over the last century and a half. So imagine the conundrum that the Swiss naturalist Ludwig Rütimeyer (1825-1895) faced in the mid-1800s when a fragmentary fossil of a primate — an upper jaw — was extracted from Eocene rocks in northern Switzerland, in an area known as Egerkingen. As far as he could tell, a fossil like this had never been found before, anywhere in the world. When Rütimeyer compared the upper molar teeth of the extinct species with those of living primates and other mammals, he discovered that they looked similar to those of howler monkeys (genus Alouatta), which live today in Central and South America. To a lesser extent, its teeth looked like those of some lemurs that live in Madagascar. Could it possibly be that he was looking at the remains of a monkey or lemur that had managed to carve out an existence north of the Swiss Alps?

PictureDarwinius masillae, from the middle Eocene of Germany
Rütimeyer was no doubt puzzled by the find, but was emboldened to identify the species as a primate by the earlier work of a famous English anatomist, Sir Richard Owen (1804-1892), who had inferred that an even more fragmentary fossil (a single tooth implanted in a jaw fragment) from Eocene rocks in southern England was that of a monkey (1,2). Owen named that specimen Macacus eocænus, implying that it was related to living macaque monkeys — some of which, such as the Japanese “snow monkey”, do live in areas with seasonally cold climates. But Owen was wrong; upon closer inspection the alleged fossil macaque turned out to be Hyracotherium, a tiny ancient relative of modern horses. Owen didn’t rectify his error until after Rütimeyer had prepared his description of the Swiss primate (3), which he placed in "die Klasse der Affen" (implying that it was a "monkey" of some sort) and named Caenopithecus lemuroides, hedging his bets about the species’ relationships by mixing references to both monkeys (-pithecus, from the Greek word for ape) and lemurs (lemuroides, Greek for “lemur-like”) in the name. Ironically, the remains of a fossil primate from the Eocene of France — a genus named Adapis — were known to science even before Rütimeyer was born, but both Rütimeyer and the famous paleontologist Georges Cuvier (1769-1832, who named Adapis in 1822) had failed to recognize it as such. Instead, they thought that it was related in some way to hoofed mammals — hence the name Adapis, which refers to the ancient Egyptian bull diety Apis. Rütimeyer’s task would have been so much simpler if Adapis had been correctly identified by Cuvier, or if he himself had made what seems, in hindsight, to be an obvious connection! Regardless, by identifying Caenopithecus as a primate, Rütimeyer became the first paleontologist to correctly identify a primate of Eocene age.

Thanks to paleontologists, we now know that primates lived all over the northern Hemisphere during the Eocene epoch — not only in Europe, but also in North America and Asia. From time to time, the ranges of lemur-like and tarsier-like primates must have even extended up into the Arctic (4). This isn’t because these species were able to tolerate cold climates, but rather because the world was considerably warmer at that time, due to higher concentrations of the “greenhouse” gas carbon dioxide in the atmosphere. As a result, there wasn’t such a big difference in temperature between the equator and the poles, so during particularly warm phases primates were able to move between Asia and North America in areas that are now often covered in snow.

We also now know that Europe was home to quite a diverse group of primates during the Eocene. Some of these primates, known as “adapiforms” (including species of Adapis, Caenopithecus, and the relatively well-known, and relatively complete, Darwinius, to the right, known by many as “Ida”) are, in my opinion, probably distant relatives of lemurs and lorises, while another group of small and bug-eyed leaping primates, known as “omomyiforms”, might have been related to the bizarre tarsiers that are found today on various islands in southeast Asia. But adapiforms have long been a problem group for primate paleontologists, because, as
Rütimeyer first discovered, they combine monkey-like and lemur-like features (omomyiforms are just as controversial due to their combination of lemur-like and tarsier-like features, but that is a story for another day). Those who accepted Rütimeyer's identification of Caenopithecus as a primate were similarly perplexed by its genealogical position; for instance, in his review of the paltry primate fossil record that existed by 1872, Charles Forsyth-Major (1843-1923) echoed Rütimeyer's confusion about Caenopithecus, noting that "...Cænopithecus is to a certain extent intermediate between the Lemuridae and the Simiadæ..." (5, p. 160); similar observations have repeatedly been made about Caenopithecus and other adapiforms up to the present day (most recently, and notably, about Darwinius).

The work that I have undertaken with my collaborators on the relationships of early primates, taking into account most of the evidence that is now available from the Eocene record of primate evolution, nevertheless suggests that there were no monkeys in Europe during the Eocene, and that, based on its teeth and jaws, Caenopithecus was (despite having some monkey-like features) probably more closely related to lemurs and lorises than to monkeys, apes, and humans (though not all agree, see. ref. 6). If our results are correct, then Rütimeyer was right about Caenopithecus being a primate, but probably for the wrong reasons — the strong similarity of its teeth to those of howler monkeys was entirely the result of convergent evolution, an evolutionary phenomenon in which distantly related species acquire similar adaptations due to their similar lifestyles (a famous example are the wings of bats and birds, which were not present in their common ancestor).

PictureA howler monkey chomping on a leaf.

Photo by Steve Jurvetson.
In the case of Caenopithecus, it appears that, like howler monkeys, this species regularly used its molar teeth to chew up leaves. Leaves can be tough and difficult to slice through, so primate lineages whose distant ancestors depended on leaves for their survival typically have evolved (through the process of natural selection) long and tall crests on their molars that act somewhat like small scissors when food is being compressed between them. Caenopithecus’ molar tooth crests coincidentally came to look a lot like those of the howler monkeys via natural selection, but their common ancestor didn’t have such specialized tooth crests; in fact, the ancestor of living howler monkeys probably didn’t appear on Earth until about 30 million years after Caenopithecus individuals were living and dying in northern Switzerland around 44 million years ago — and furthermore it seems very likely that howler monkeys have only ever lived in Central and South America. Whatever the basis for his conclusions, Rütimeyer’s identification of Caenopithecus as a primate — notably only three years after Darwin published The Origin of Species — set in motion important debates about the existence of primates in the ancient rocks of Europe, and the evolutionary and/or climatic changes that might have allowed such primates to have existed so far north of their modern distributions.

PictureCaenopithecus maxilla described by Stehlin in 1916
In 1916, over half a century after Rütimeyer first described Caenopithecus, another Swiss paleontologist named Hans Stehlin (1870-1941) brought to light additional fossils of this mysterious species (7). By that time, paleontologists working in the western United States had documented several additional Eocene primate species, and four years earlier Stehlin had written a monograph on Adapis, clearly establishing, once and for all, that it was a primate. With the new and more complete Caenopithecus fossils, he was able to show that it differed from Adapis in some important ways — for instance, Caenopithecus only had six molar and premolar teeth on each side of the jaw, whereas Adapis had seven (humans typically have four or five, depending on whether our “wisdom” teeth have erupted, or have been removed by a dentist). In Caenopithecus, the first premolar behind its fang-like canine tooth was peculiar and unlike that of Adapis in being very small, almost vestigial. Stehlin argued that Adapis showed the greatest anatomical similarity to living lemurs, and should not be aligned with "higher primates", but he could not make sense of Caenopithecus' odd combination of features. Caenopithecus probably seemed even more distinctive to Stehlin than it should have, because he mistakenly identified a foot bone from Egerkingen — an astragalus (or talus) bone, which connects the shin bone (tibia) to the rest of the foot — as belonging to Caenopithecus, but it was actually that of a very distantly related carnivorous mammal, called a “creodont” (8), that lived in the same time and place as Caenopithecus (and possibly preyed upon it). And that was the end of the story for Caenopithecus — the fossils from Egerkingen showed what its jaws and teeth looked like, but for the last century we’ve known nothing else about paleontology’s “first" Eocene primate.

PictureView of the Egerkingen area in 2011
In 2011 I met with colleagues in Switzerland to see if there was any chance that the Egerkingen sites could be reopened for excavation, with the hope that some day we might learn more about Caenopithecus and the other adapiform and omomyiform primates that lived alongside it. Unfortunately, the prospects for future work at Egerkingen are not particularly good — the tall and steep rock faces that yielded the fossils are now used primarily by local rock climbers, and the paleontologically interesting fissures are either gone or largely exhausted. However when I visited the museum that both Rütimeyer and Stehlin worked at, the Naturhistorisches Museum Basel, and looked through the old Egerkingen collections, I was surprised to find four primate foot bones, about the size that one would expect for a Caenopithecus individual, that had never been formally described. They were particularly interesting to me because the primate astragalus bone from Egerkingen looked very similar to one that has been discovered in late Eocene rocks in Egypt several years earlier (9), and that probably belonged to a close relative of Caenopithecus that my colleagues and I had named Afradapis in 2009 (10). With the realization that the astragalus bone (and three other calcaneus, or “heel”, bones) from Egerkingen probably belonged to Caenopithecus, we appeared to have our first new information about this enigmatic primate in almost a century...and conveniently didn’t even have to excavate to find them!

Together with my co-authors Doug Boyer, of Duke University, and Loïc Costeur, of the Naturhistorisches Museum Basel, we have just described these fossils (11) in the journal PeerJ. The “new” remains of Caenopithecus provide some interesting clues about its lifestyle that we didn’t have before.

PictureA slow-moving African loris (genus Perodicticus)
Using three-dimensional digital models of the bones, which we constructed by CT-scanning the specimens, we used quantitative approaches to compare their shapes with those of other living and extinct primates (if you are interested, these models can be downloaded, viewed, and manipulated here, on MorphoSource, an extraordinarily useful digital library of CT-scanned bones that was created by my co-author Doug Boyer). We found that while the bones of Caenopithecus are quite unique when compared with those of other early primates, they do show some specializations that are similar to those of the slow lorises — prosimian primates that (as their name would suggest) typically move very slowly and cautiously, and do not leap at all. This raises some interesting questions, because slow lorises have diets made up largely of insects or fruits, but recall that Caenopithecus had the teeth of a leaf-eater. So Caenopithecus might have moved through the trees somewhat like a loris, but perhaps for very different reasons. For instance, lorises are active at night, and their stealthy movement through the darkness appears to be part of a strategy to avoid being detected by both potential insect prey, and by potential predators (such as mammalian carnivores and birds), but the leaves that Caenopithecus liked to eat obviously couldn’t have run away, so there would have been no need to sneak up on them.

Picture3D models of the Caenopithecus bones
Given that Caenopithecus was a leaf eater, perhaps the cautious movement that we envision for Caenopithecus would have been related more to a need to safely move through complex networks of small leaf-bearing twigs that exist at the ends of larger branches, without slipping and falling to the forest floor. The joint surfaces of Caenopithecus’ ankle bones suggest that its feet were very mobile, and, by analogy with the foot bone anatomy of some living lemurs, we inferred that a deep groove on the base of the astragalus bone might have helped to constrain the line of action of the tendon of an important toe-flexing muscle, and help to maintain the mechanical advantage of that muscle when the foot is pointed away from the shin, as would have happened if a Caenopithecus individual suspended itself under a branch, holding on exclusively by one or both feet (thereby freeing up the hands to grab leaves or fruit). In fact, these kinds of movements are commonly observed in a primate group mentioned before — the howler monkeys — which also eat lots of leaves; this would be yet another example of convergent evolution between Caenopithecus and howler monkeys. Howler monkeys also use their prehensile tails to hang from branches, but we don't know if Caenopithecus had such capabilities.

PictureThe recently extinct lemur Babakotia
The foot bones of Caenopithecus are also somewhat similar to (but less specialized than) those of a recently extinct large “sloth” lemur from Madagascar, named Babakotia, that also ate leaves and probably spent much of its time hanging under branches by its arms and legs (12). As in the case of the howler monkeys, the last common ancestor of Caenopithecus and Babakotia wouldn’t have had this lifestyle, and so this would be yet another example of convergent evolution. It is fascinating to think that a much more ancient, and much smaller, ancestor of Babakotia might have closely resembled a species like Caenopithecus lemuroides due solely to similar selection pressures that were operating entirely independently in Europe and Madagascar. These recurring patterns in the primate fossil record serve as a reminder of just how strongly our evolutionary paths have been shaped by the combination of constraints imposed externally (by life in the trees and the foods that exists in those trees) and internally (by the genetic and developmental constraints on anatomy that we inherited from the ancient last common ancestor of all primates); we should not be surprised to find that similar evolutionary “solutions” to survival have occurred time and time again over the course of the last 56 million years.

The ankle bones of primates are also of interest for determining relationships, because some features of the astragalus and calcaneus consistently sort living species onto different sides of the primate family tree. For instance, the orientation of the joint between the fibula and astragalus
seems to be one of only a few reliable skeletal indicators of whether an early fossil primate is a member of the strepsirrhine group (lemurs, lorises, bushbabies, and their early relatives) or the haplorhine group (humans, apes, monkeys, tarsiers, and their early relatives) (13, 14). In Caenopithecus this joint shows a very pronounced flare, as in strepsirrhines (which appear to be specialized in this feature). When we re-evaluated the relationships of Caenopithecus taking the new foot bones into account, we found that it is a distant relative of strepsirrhines, and probably a close relative of Adapis (see below)
— which, you might recall, was actually the first fossil primate ever discovered (of any age), but wasn't recognized as such until late in the 19th century. However even as early as 1873, the French paleontologist Paul Gervais (1816-1879) had argued that perhaps Caenopithecus and Adapis were very close relatives (if not members of the same genus), so it isn't a new idea...but it's still remarkable, in hindsight, that Cuvier and Rütimeyer didn't recognize the anatomical evidence for this connection.

The position of "adapiforms" in the primate family tree, based on the results of our analyses
And the relationships among the adapiforms, based on this most recent genealogical estimate that we provide (greatly simplified in the picture above), bring us to another interesting issue —  which is that the closest relative (or "sister taxon") of Caenopithecus is Afradapis, a younger form (about 37 million years old) from Egypt. This might not seem so odd given the current arrangement of the continents, but throughout the Eocene, Africa was separated from both Europe and Asia by a vast sea that paleontologists call "Tethys" (the ancient, and much larger, precursor of the modern-day Mediterranean). If Afradapis' close relatives, such as Adapis, Caenopithecus, and Darwinius, are all found in Europe and are descended from an ancestor that lived there as well, the adapiform family tree implies that somehow an ancestor of Afradapis managed to cross the Tethys Sea to get to Africa, most likely from Europe but perhaps from Asia (the Eocene record of which is not as well-sampled by paleontologists, and still holds many mysteries). It is an inescapable, but perplexing, conclusion that primates have made these improbable sea voyages multiple times, probably by floating on rafts of vegetation that were washed out to sea by strong storms or tsunamis — it's the only way to explain lemurs' presence in Madagascar, and somehow monkeys made an even more impressive journey across the Atlantic Ocean from Africa to South America (though those continents were much closer at the time; have a look at this recent post for a good discussion of the problem). Looking back into the fossil record, we see that members of several other extinct groups, including adapiforms and our own monkey-like ancestors, probably also crossed the Tethys, but perhaps taking advantage of island chains that have since been erased from the geological record as the conjoined African and Arabian continental plates slowly moved to the northeast and smushed into Asia (thereby closing off the Tethys, and forming the Zagros Mountains).

Finally, the four errors that have been mentioned so far, all made by very skilled anatomists and paleontologists, nicely demonstrate how knowledge in paleontology often advances by taking two steps forward and one step back. If people like Cuvier, Owen, Rütimeyer, and Stehlin (and so many others) all made mistakes when analyzing the early fossil primates of Europe, we have to wonder…what were we wrong about in our study? There are a few obvious possibilities. Perhaps we have attributed the foot bones to the wrong species, and they actually belong to another adapiform from Egerkingen named Leptadapis priscus. We think this is unlikely, though, because remains of Caenopithecus are far more abundant in the Egerkingen sites, and furthermore the ankle bones are very different from those of Leptadapis priscus’ close relatives. Maybe the evolutionary trees that we present have incorrect groupings; in fact this would not be surprising, because we have so little information from the fossil species that are included, and anatomy isn't always a reliable guide to relationships (again, due to convergent evolution). And fortunately we avoided at least one embarrassing oversight thanks to the process of peer review — unbeknownst to us, one of the world's experts on primate foot bones, Marian Dagosto (now of Northwestern University), actually briefly described and figured the Caenopithecus tarsals in her unpublished doctoral dissertation, which was completed in 1986. One of the reviewers of our paper caught the omission and we are glad that we were able to give her due credit in the final version of our manuscript; fortunately she had come to very similar conclusions about the functional and evolutionary implications of the fossils, which is reassuring. Only time will tell what conclusions of our study will turn out to be inaccurate once more evidence has been collected, but we can be pretty confident that we didn’t get everything right; it would be foolish to think otherwise.

References (with links to the original papers):

(1) Owen R. 1839. Description of the fossil mentioned in the preceding letter. The Magazine of Natural History 3: 446-448.

(2) Owen R. 1846. A History of British Fossil Mammals and Birds. London: John Van Voorst, 561 pp.

(3) Rütimeyer L. 1862. Eocäne Säugethiere aus dem Gebiet des schweizerischen Jura. Neue Denkschriften der Allgemeinen Schweizerischen Gesellschaft für die Gesammten Naturwissenschaften 19:1–98.

(4) Smith T., Rose K.D., and Gingerich P.D. 2006. Rapid Asia–Europe–North America geographic dispersal of earliest Eocene primate Teilhardina during the Paleocene–Eocene Thermal Maximum. Proceedings of the National Academy of Sciences, U.S.A. 103: 11223–11227.

Forsyth Major C.I. 1872. Note on some fossil monkeys found in Italy, preceded by a review of the fossil Quadrumana in general. The Annals and Magazine of Natural History 10: 153-166.

(6) Franzen J.L., Gingerich P.D., Habersetzer J., Hurum J. H., von Koenigswald W., and Smith B.H. 2009. Complete primate skeleton from the Middle Eocene of Messel in Germany: morphology and paleobiology. PLoS ONE 4: e5723.

(7) Stehlin H.G. 1916.
Die Säugetiere des schweizerischen Eocaens. Critischer Catalog der Materialen. Siebenter Teil, zweite Hälfte: Caenopithecus--Necrolemur--Microchoerus--Nannopithex--Anchomomys--Periconodon--Amphichiromys--Heterochiromys —Nachträge zu Adapis —Schlussbetrachtungen zu den Primaten. Abhandlung der Schweizerischen Paläontologischen Gesellschaft 41 :1299–1552.

(8) Decker R.L. and Szalay F.S. 1974. Origins and function of the pes in the Eocene Adapidae
(Lemuriformes, Primates). In: Jenkins, F.A., ed. Primate Locomotion. New York: Academic Press,

(9) Boyer D.M., Seiffert E.R., and Simons E.L. 2010. Astragalar morphology of Afradapis, a large adapiform primate from the earliest late Eocene of Egypt. American Journal of Physical Anthropology 143 :383-402.

(10) Seiffert E.R., Perry J.M.G., Simons E.L., and Boyer D.M. (2009) Convergent evolution of anthropoid-like adaptations in Eocene adapiform primates. Nature 461: 1118-1121.

(11) Seiffert E.R., Costeur L., and Boyer D.M. (2015) Primate tarsal bones from Egerkingen, Switzerland, attributable to the middle Eocene adapiform Caenopithecus lemuroides. PeerJ.

(12) Jungers W.L., Godfrey L.R., Simons E.L., Chatrath P.S., and Rakotosamimanana, B. 1991. Phylogenetic and functional affinities of Babakotia (Primates), a fossil lemur from northern Madagascar. Proceedings of the National Academy of Sciences 88: 9082–9086.

(13) Gebo DL. 1988. Foot morphology and locomotor adaptation in Eocene primates. Folia Primatologica 50:3–41.

(14) Boyer D.M. and Seiffert E.R. 2013.  Patterns of astragalar fibular facet orientation in extant and fossil primates and their evolutionary implications. American Journal of Physical Anthropology 151:420-447.