¿ìè¶ÌÊÓƵ

AS THEY glide through the air, they look like nothing so much as furry,
living kites. At rest, they look like the victims of over-enthusiastic mothers
who have bought them a skin three sizes too large, so that it bunches up
in ruffles between their front and rear legs. They are colugos, the so-called
flying lemurs that neither fly nor are lemurs, and they are more intimately
involved in our ancestry than anyone had ever supposed.

This is the import of two surprising papers in the 24 May issue of Nature
that have overturned the basic framework of the origin and adaptive radiation
of the order Primates. ‘In fact,’ says Richard Kay of Duke University in
Durham, North Carolina, the senior author of one of the papers, ‘You might
say that early primate evolution as we know it has just glided out the window.’

Few analyses of fossil species can have been so disruptive to the general
consensus on primate evolution. For the past 15 or 20 years, virtually all
the textbooks and scientific papers on early primate evolution have agreed
on the overall picture. In the Palaeocene, small-bodied, primitive primates
arose from an insectivore stock and then diversified into several families.
Not everyone agrees on exactly which families are primates and which are
simply their close relatives, but most would include three families among
the primates. These families, the Plesiadapidae, the Paromomyidae and the
Carpolestidae, known mostly from dental and cranial remains, are small,
shrewlike, snouty creatures, often with large incisors. Their molar teeth
resemble those of later, undoubted primates that appear in the Eocene, a
group generally known as the ‘first primates of modern aspect’.

Unlike Palaeocene primates, the Eocene forms show virtually all of the
traits usually cited as typifying the order Primates: the complete, bony
ring around the orbit; the relatively enlarged brain; the dominance of stereoscopic
vision over olfaction among the senses; the development of grasping feet
and hands with opposable thumbs and big toes; and the replacement of claws
by flat nails. Less functionally significant but equally important taxonomically
were the details of the blood supply to the brain – which in primates is
carried by the internal carotid artery through a bony canal in the base
of the skull – and the details of the bones that comprise the auditory bulla,
the bony balloon that holds the middle ear.

Although the received wisdom of the 19th and early 20th century was
that the order Primates was shaped by its arboreal habits, a perceptive
review by Matt Cartmill of Duke University, in 1973, argued compellingly
that this explanation was inadequate. After all, many tree squirrels are
adeptly acrobatic in the trees and yet they lack both a postorbital bar
and an opposable big toe. Worse yet, they retain their claws, poorly stereoscopic
vision, and elaborate olfactory apparatus and do not have brains significantly
larger than those of ground squirrels. Thus, the fundamental primates adaptation
could not be arboreality per se.

Cartmill’s alternative, known as the visual predation hypothesis, related
the diagnostic traits of the primates to the demands of nocturnal, predatory
behaviour in a terminal-branch, arboreal habitat. The wide-set eyes with
overlapping visual fields and a well-developed visual cortex parallels developments
seen in other predators, such as the cats, who locate their prey by sight.
The grasping digits and opposable thumbs and toes, equipped with flat nails,
are adaptive in small-branch environments, whereas squirrel-like claws are
more suitable for running up and down supports with a larger radius of curvature.
The development of good hand-eye coordination is clearly important in the
manual capture of small prey. In short, Cartmill proposed – and his fellow
palaeontologists largely accepted – that the basic primate adaptation was
a way of life such as that found among lorises, which sneak cautiously through
the trees to spot and seize insects and other small prey with their hands.

What was not so widely accepted was a corollary of Cartmill’s hypothesis.
Philosophically, the boundary between orders ought to reflect major adaptive
shifts, Cartmill argued; therefore, only species that had undergone this
shift to visual predation ought to be placed in the Primates. And that premise
removed virtually all of the Palaeocene primate species from the order,
as their fossil remains indicated that they still possessed clawed digits,
no postorbital bar, little overlap of visual fields and a long snouty face
with an apparently well-developed olfactory apparatus.

While the logic of Cartmill’s boundary was elegant, it made many palaeontologists
uncomfortable. The crux of the problem was that Palaeocene primates, though
lacking many or all of the ‘typical’ primate traits, had teeth and jaws
that were quite similar to the first primates of modern aspect in several
details. Relegating the Palaeocene primates to the Insectivora downplayed
the importance of those resemblances and cast a group of animals that was
clearly closely related to later primates adrift on the ill-defined and
shifting seas of the insectivores.

There matters rested until better fossil material became available recently.
And, as ‘better material’ is wont to do, it provided an alternative that
no one had anticipated. The first inkling that things were up in the air
again came from the research of K. Christopher Beard, now at the Carnegie
Museum of Natural History in Pittsburgh, Pennsylvania. For his doctoral
dissertation, Beard analysed two new, partial skeletons of Phenacolemur,
the first Palaeocene primate fossil specimens that had both cranial and
postcranial remains. As these Palaeocene genera and species are defined
on the basis of cranial remains, Beard’s advantage was that, for the first
time, he could be certain which heads went with which bodies. He could also
then go through collections of unassociated remains and pick out postcranial
remains of other individuals of Phenacalemur and another paromomyid, Ignacius.

Beard’s anatomical analysis revealed a most surprising feature in the
hands and feet of these paromomyids. Like many other primates, these paromomyids
had three bones (phalanges) in each digit except the thumb and big toe.
Unlike other primates, the proportions of these bones were peculiar. Almost
invariably among mammals, the bone nearest to the palm or sole, the proximal
phalanx, is longer than the middle phalanx, the next bone in the digit.
In the fossils, the middle phalanx, was much elongated, a condition found
only among two groups of living mammals.

The Edentata, the group including the anteaters and armadillos, have
very short, stubby proximal phalanges, which makes the middle phalanges
relatively long. Their third or distal phalanges form curved claws. This
pattern is clearly an adaptation for powerful digging, either in burrowing
or in search of prey. But Phenacolemur and Ignacius, while clawed, do not
have shortened proximal phalanges. In fact, the living mammals that have
normal-sized proximal phalanges and even longer middle phalanges are the
flying lemurs.

Also known as colugus, the flying lemurs are the sole living representatives
of an entire order, the Dermoptera or ‘skin wings’. There is but a single
genus, Cynocephalus, containing two species that live in Southeast Asia.
Like all gliding mammals, colugos possess a patagium, a furry web of skin
stretching between fore and hindlimbs. The patagium enables them to float
aloft over long distances – more than 100 metres per glide has been measured
– rather than dropping like a stone when they jump out of a tree. They soar
rather than flapping and flying, as do bats.

Where colugos are extraordinary, even among gliders, is in the extent
of their patagium, which not only connects front to hind legs but also webs
between the neck and shoulders, in between all the fingers, between all
the toes and from the hind leg to the tail. (Other gliding mammals, such
as ‘flying’ squirrels, apparently evolved their less-developed patagia independently
from the colugos.) To control the direction of the glide and the rate of
drop, colugos adjust their patagia with their digits, calibrating the degree
of parachuting or billowing as well as the angle of the ‘wing’. Thus, fine
control of the hands and exquisite hand-eye coordination, to ensure safe
landings, is extremely important in colugos.

Like the fossil paromomyids, colugos have very strong ridges for the
attachment of the patagium on the proximal phalanges as well as somewhat
short and flattened claws at the terminal ends of their digits. Additional
specialisations among colugos involve a reorganisation of the wrist bones
and some changes in the ankle – both of which Beard also found among the
paromomyid fossils he analysed. Beard’s claims that paromomyids were gliders
might have been swept aside as too unexpected to be reasonable – despite
his careful analysis – were they not being reinforced by studies of another
paromomyid by the team of Richard Kay, of Duke University, Richard Thorington,
Jr., of the Smithsonian Institution in Washington DC, and Peter Houde, of
Princeton University.

Simultaneously with Beard’s work, this team was studying a new skull
of Ignacius, one of the species whose postcrania Beard was working on. Collected
by Houde, the Ignacius skull preserved regions of the base of the cranium
that were crushed or missing in all other known specimens. It offered the
opportunity to settle debates about the composition of the middle ear and
the blood supply to the brain, two crucial regions for settling the taxonomic
debates about whether these Palaeocene species ought to be classed as Primates.

They, too, found some surprises. The auditory bulla of Ignacius is not
made up of the petrosal bone, as in primates, but is formed from the entotympanic.
While entotympanic bullae have evolved in many different groups, the anatomical
details of the arrangement of bones in this ear region show that Ignacius’
anatomy resembles only that in colugos.

Tracing the blood supply to the brain also showed that Ignacius was
unlike primates. There is no evidence in the specimen for either an internal
carotid artery or an ascending pharyngeal artery by which the brain could
be nourished. By process of elimination, the team concluded that the blood
supply to the brain was solely through the vertebral artery, a vessel that
passes through the foramen magnum in the base of the skull. The same pattern
is found in another known Palaeocene primate, Plesiadapis, and in the extant
dermopterans. So Palaeocene primates are not Primates, and they almost certainly
are dermopterans.

What of the origin of the order Primates, then? Three points can be
made. One is that the fossil record of primates has just shrunk noticeably,
leaving the Eocene forms as the earliest primates. There is a certain satisfactory
clarity about this turn of events, however unsettling it may be to have
all of the Palaeocene fossil record of our order disappear in one fell swoop.

Secondly, the anatomical features that typify the primates as visual
predators are disconcertingly similar to those needed for controlled gliding
between trees. Reliance on stereoscopic vision, the evolution of grasping
hands and feet, the development of excellent hand-eye coordination to control
gliding and to ensure safe landings – all these make sense for gliders as
well as visual predators. No one has suggested that a lineage specialised
for gliding reversed itself to evolve towards a visual predator in Cartmill’s
sense. Still, the fundamental direction of the two adaptations is remarkably
similar and still bespeaks a close relationship. The long-noted dental similarities
between the Palaeocene dermopterans and the Eocene primates continue to
attest to a common ancestry.

Finally, if there was ever a reason to argue for more funding for field
work and fossil collection hand-in-hand with analyses of functional anatomy,
these studies have strengthened the claim. That a very small number of new
specimens have so revised our views of what was thought to be a well-understood
time period is heartening. There are still surprises in store for us; there
are still new worlds to be discovered in the past.

Pat Shipman is a freelance science writer based in Maryland.