Transitional Vertebrate Fossils FAQ
Part 2A

Part 1B

Contents

Part 2B

PART 2

Overview of the Cenozoic

The Cenozoic fossil record is much better than the older Mesozoic record, and much better than the very much older Paleozoic record. The most extensive Cenozoic gaps are early on, in the Paleocene and in the Oligocene. From the Miocene on it gets better and better, though it's still never perfect. Not surprisingly, the very recent Pleistocene has the best record of all, with the most precisely known lineages and most of the known species-to-species transitions. For instance, of the 111 modern mammal species that appeared in Europe during the Pleistocene, at least 25 can be linked to earlier European ancestors by species-to-species transitional morphologies (see Kurten, 1968, and Barnosky, 1987, for discussion).

Timescale

For the rest of this FAQ, I'll walk through the known fossil records for the major orders of modern placental mammals. For each order, I'll describe the known lineages leading from early unspecialized placentals to the modern animals, point out some of the remaining gaps, and list several of the known species-to-species transitions. I left out some of the obscure orders (e.g. hyraxes, anteaters), groups that went completely extinct, and some of the families of particularly diverse orders.

Primates

I'll outline here the lineage that led to humans. Notice that there were many other large, successful branches (particularly the lemurs, New World monkeys, and Old World monkeys) that I will only mention in passing. Also see Jim Foley's fossil hominid FAQ for detailed information on hominid fossils.

GAP: "The modern assemblage can be traced with little question to the base of the Eocene" says Carroll (1988). But before that, the origins of the very earliest primates are fuzzy. There is a group of Paleocene primitive primate-like animals called "plesiadapids" that may be ancestral to primates, or may be "cousins" to primates. (see Beard, in Szalay et al., 1993.)

• Palaechthon, Purgatorius (middle Paleocene) -- Very primitive plesiadapids. To modern eyes they looks nothing like primates, being simply pointy-faced, small early mammals with mostly primitive teeth, and claws instead of nails. But they show the first signs of primate-like teeth; lost an incisor and a premolar, and had relatively blunt-cusped, squarish molars.
• Cantius (early Eocene) -- One of the first true primates (or "primates of modern aspect"), more advanced than the plesiadapids (more teeth lost, bar behind the eye, grasping hand & foot) and beginning to show some lemur-like arboreal adaptations.
• Pelycodus & related species (early Eocene) -- Primitive lemur-like primates.

The tarsiers, lemurs, and New World monkeys split off in the Eocene. The Old World lineage continued as follows:

• Amphipithecus, Pondaungia (late Eocene, Burma) -- Very early Old World primates known only from fragments. Larger brain, shorter nose, more forward-facing eyes (halfway between plesiadapid eyes and modern ape eyes).

GAP: Here's that Oligocene gap mentioned above in the timescale. Very few primate fossils are known between the late Eocene and early Oligocene, when there was a sharp change in global climate. Several other mammal groups have a similar gap.

• Parapithecus (early Oligocene) -- The O.W. monkeys split from the apes split around now. Parapithecus was probably at the start of the O.W. monkey line. From here the O.W. monkeys go through Oreopithecus (early Miocene, Kenya) to modern monkey groups of the Miocene & Pliocene.
• Propliopithecus, Aegyptopithecus (early Oligocene, Egypt) -- From the same time as Parapithecus, but probably at the beginning of the ape lineage. First ape characters (deep jaw, 2 premolars, 5- cusped teeth, etc.).
• Aegyptopithecus (early-mid Oligocene, Egypt) -- Slightly later anthropoid (ape/hominid) with more ape features. It was a fruit-eating runner/climber, larger, with a rounder brain and shorter face.
• Proconsul africanus (early Miocene, Kenya.) -- A sexually dimorphic, fruit-eating, arboreal quadruped probably ancestral to all the later apes and humans. Had a mosaic of ape-like and primitive features; Ape-like elbow, shoulder and feet; monkey- like wrist; gibbon-like lumbar vertebrae.
• Limnopithecus (early Miocene, Africa) -- A later ape probably ancestral to gibbons.
• Dryopithecus (mid-Miocene) -- A later ape probably ancestral to the great apes & humans. At this point Africa & Asia connected via Arabia, and the non-gibbon apes divided into two lines:
  1. Sivapithecus (including "Gigantopithecus" & "Ramapithecus", mid- Miocene) -- Moved to Asia & gave rise to the orangutan.
  2. Kenyapithecus (mid-Miocene, about 16 Ma) -- Stayed in Africa & gave rise to the African great apes & humans.

GAP: There are no known fossil hominids or apes from Africa between 14 and 4 Ma. Frustratingly, molecular data shows that this is when the African great apes (chimps, gorillas) diverged from hominids, probably 5-7 Ma. The gap may be another case of poor fossilization of forest animals. At the end of the gap we start finding some very ape-like bipedal hominids:

• Australopithecus ramidus (mid-Pliocene, 4.4 Ma) -- A recently discovered very early hominid (or early chimp?), from just after the split with the apes. Not well known. Possibly bipedal (only the skull was found). Teeth both apelike and humanlike; one baby tooth is very chimp-like. (White et al., 1994; Wood 1994)
• Australopithecus afarensis (late Pliocene, 3.9 Ma) -- Some excellent fossils ("Lucy", etc.) make clear that this was fully bipedal and definitely a hominid. But it was an extremely ape-like hominid; only four feet tall, still had an ape-sized brain of just 375-500 cc (finally answering the question of which came first, large brain or bipedality) and ape-like teeth. This lineage gradually split into a husky large-toothed lineage and a more slender, smaller- toothed lineage. The husky lineage (A. robustus, A. boisei) eventually went extinct.
• Australopithecus africanus (later Pliocene, 3.0 Ma) -- The more slender lineage. Up to five feet tall, with slightly larger brain (430-550 cc) and smaller incisors. Teeth gradually became more and more like Homo teeth. These hominds are almost perfect ape- human intermediates, and it's now pretty clear that the slender australopithecines led to the first Homo species.
• Homo habilis (latest Pliocene/earliest Pleistocene, 2.5 Ma) -- Straddles the boundary between australopithecines and humans, such that it's sometimes lumped with the australopithecines. About five feet tall, face still primitive but projects less, molars smaller. Brain 500-800 cc, overlapping australopithecines at the low end and and early Homo erectus at the high end. Capable of rudimentary speech? First clumsy stone tools.
• Homo erectus (incl. "Java Man", "Peking Man", "Heidelberg Man"; Pleist., 1.8 Ma) -- Looking much more human now with a brain of 775-1225 cc, but still has thick brow ridges & no chin. Spread out of Africa & across Europe and Asia. Good tools, first fire.
• Archaic Homo sapiens (Pleistocene, 500,000 yrs ago) -- These first primitive humans were perfectly intermediate between H. erectus and modern humans, with a brain of 1200 cc and less robust skeleton & teeth. Over the next 300,000 years, brain gradually increased, molars got still smaller, skeleton less muscular. Clearly arose from H erectus, but there are continuing arguments about where this happened.
• One famous offshoot group, the Neandertals, developed in Europe 125,000 years ago. They are considered to be the same species as us, but a different subspecies, H. sapiens neandertalensis. They were more muscular, with a slightly larger brain of 1450 cc, a distinctive brow ridge, and differently shaped throat (possibly limiting their language?). They are known to have buried their dead.
• H. sapiens sapiens (incl. "Cro-magnons"; late Pleist., 40,000 yrs ago) -- All modern humans. Average brain size 1350 cc. In Europe, gradually supplanted the Neanderthals.

Known species-species transitions in primates:

Phillip Gingerich has done a lot of work on early primate transitions. Here are some of his major findings in plesiadapids, early lemurs, and early monkeys:

• Plesiadapids: Gingerich (summarized in 1976, 1977) found smooth transitions in plesiadapid primates linking four genera together: Pronothodectes, Nannodectes, two lineages of Plesiadapis, and Platychoerops. In summary: Pronothodectes matthewi changed to become Pro. jepi, which split into Nannodectes intermedius and Plesiadapis praecursor. N. intermedius was the first member of a gradually changing lineage that passed through three different species stages (N. gazini, N. simpsoni, and N. gidleyi). Ples. praecursor was the first member of a separate, larger lineage that slowly grew larger (passing through three more species stages), with every studied character showing continuous gradual change. Gingerich (1976) noted "Loss of a tooth, a discrete jump from one state to another, in several instances proceeded continuously by continuous changes in the frequencies of dimorphism -- the percentage of specimens retaining the tooth gradually being reduced until it was lost entirely from the population." The Plesiadapis lineage then split into two more lineages, each with several species. One of these lineages shows a gradual transition from Plesiadapis to Platychoerops,"where the incisors were considerably reorganized morphologically and functionally in the space of only 2-3 million years."
• Early lemur-like primates: Gingerich (summarized in 1977) traced two distinct species of lemur-like primates, Pelycodus frugivorus and P. jarrovii, back in time, and found that they converged on the earlier Pelycodus abditus "in size, mesostyle development, and every other character available for study, and there can be little doubt that each was derived from that species." Further work (Gingerich, 1980) in the same rich Wyoming fossil sites found species-to-species transitions for every step in the following lineage: Pelycodus ralstoni (54 Ma) to P. mckennai to P. trigonodus to P. abditus, which then forked into three branches. One became a new genus, Copelemur feretutus, and further changed into C. consortutus. The second branch became P. frugivorus. The third led to P. jarrovi, which changed into another new genus, Notharctus robinsoni, which itself split into at least two branches, N. tenebrosus, and N. pugnax (which then changed to N. robustior, 48 Ma), and possibly a third, Smilodectes mcgrewi (which then changed to S. gracilis). Note that this sequence covers at least three and possibly four genera, with a timespan of 6 million years.
• Early monkey-like primates: Gingerich (1982, also discussed in Gingerich, 1983) also describes gradual species-species transitions in a lineage of early Eocene primate: Cantius ralstoni to C. mckennai to C. trigonodus.

And here are some transitions found by other researchers:

• Rose & Bown (1984) analyzed over 600 specimens of primates collected from a 700-meter-thick sequence representing approximately 4 million years of the Eocene. They found smooth transitions between Teilhardina americana and Tetonoides tenuiculus, and also beween Tetonius homunculus and Pseudotetonius ambiguus. "In both lines transitions occurred not only continuously (rather than by abrupt appearance of new morphologies followed by stasis), but also in mosaic fashion, with greater variation in certain characters preceding a shift to another character state." The T. homunculus - P. ambiguus transition shows a dramatic change in dentition (loss of P2, dramatic shrinkage of P3 with loss of roots, shrinkage of C and I2, much enlarged I1) that occurs gradually and smoothly during the 4 million years. The authors conclude "...our data suggest that phyletic gradualism is not only more common than some would admit but also capable of producing significant adaptive modifications."
• Delson (discussed in Gingerich, 1985) has studied transitions in primates from the Miocene to the present. For instance, in a 1983 paper (see Chaline, 1983), he discussed a possible smooth transition from Theropithecus darti to T. oswaldi, and discusses transitions in hominids, concluding that Homo sapiens clearly shows gradual changes over the last 800,000 years.
• Kurten (1968) reports a smooth transition linking Macaca florentina to M. sylvana

Bats

GAP: One of the least understood groups of modern mammals -- there are no known bat fossils from the entire Paleocene. The first known fossil bat, Icaronycteris, is from the (later) Eocene, and it was already a fully flying animal very similar to modern bats. It did still have a few "primitive" features, though (unfused & unkeeled sternum, several teeth that modern bats have lost, etc.)

• Fruit bats and horseshoe bats first appear in the Oligocene. Modern little vespertiliontids (like the little brown bat) first appear in the Miocene.

Carnivores

• Creodonts -- early placental mammals with minor but interestingly carnivore-like changes in the molars and premolars. Had a carnivore- like shearing zone in the teeth, though the zone moved throughout life instead of staying in particular teeth. Also had a carnivore- like bony sheet in the brain dividing cerebrum & cerebellum, details of ankle. Closely related to & possibly ancestral to carnivores. The origin of the creodonts is unclear. They probably were derived from condylarths.
• Cimolestes (late Cretaceous) -- This creodont (?) lost the last molar & then later enlarged the last upper premolar and first lower molar. (In modern carnivores, these two teeth are very enlarged to be the wickedly shearing carnassial teeth, the hallmark of carnivores.) Still unfused feet & unossified bulla. This genus is probably ancestral to two later lines of Eocene carnivores called "miacoids". Miacoids were relatively unspecialized meat-eaters that seem to have split into a "viverravid" line (with cat/civet/hyena traits) and a "miacid" line (with dog/bear/weasel traits). These two lines may possibly have arisen from these slightly different species of Cimolestes:
• Cimolestes incisus & Cimolestes cerberoides (Cretaceous) -- These are two species that lost their third molar, and may have given rise to the viverravid line of miacoids (see Hunt & Tedford, in Szalay et al., 1993).
• Cimolestes sp. (Paleocene) -- A later, as yet unnamed species that has very miacid-like teeth.
• Simpsonictis tenuis (mid-Paleocene) -- A very early viverravid. The upper carnassial was large; the lower carnassial was of variable size in different individuals.
• Paroodectes, Vulpavus (early Eocene) -- Early miacids. Enlarged carnassials now specialized for shearing. Still had unfused foot bones, short limbs, plantigrade feet, unossified bulla.

GAP: few miacoid skulls are known from the rest of the Eocene -- a real pity because for early carnivore relationships, skulls (particularly the skull floor and ear capsule) are more useful than teeth. There are some later skulls from the early Oligocene, which are already distinguishable as canids, viverrids, mustelids, & felids (a dog-like face, a cat-like face, and so on). Luckily some new well-preserved miacoid fossils have just been found in the last few years (mentioned in Szalay et al., 1993). They are still being studied and will probably clarify exactly which miacoids gave rise to which carnivores. Meanwhile, analysis of teeth has revealed at least one ancestor:

• Viverravus sicarius (mid-Eocene) -- Hunt & Tedford (in Szalay et al., 1993) think this viverravid may be the ancestral aeluroid. It has teeth & skeletal traits similar to the first known Oligocene aeluroids (undifferentiated cat/civet/hyenas).

From the Oligocene onward, the main carnivore lineages continued to diverge. First, the dog/bear/weasel line.

Dogs:

• Cynodictis (late Eocene) -- First known arctoid (undifferentiated dog/bear).
• Hesperocyon (early Oligocene) -- A later arctoid. Compared to miacids like Paroodectes, limbs have elongated, carnassials are more specialized, braincase is larger. From here, the main line of canid evolution can be traced in North America, with bears branching out into a Holarctic distribution.
• Cynodesmus (Miocene) -- First true dog. The dog lineage continued through Tomarctus (Pliocene) to the modern dogs, wolves, & foxes, Canis (Pleistocene).

Bears:

• Cynodictis (see above)
• Hesperocyon (see above)
• Ursavus elmensis (mid-Oligocene) -- A small, heavy doglike animal, intermediate between arctoids and bears. Still had slicing carnassials & all its premolars, but molars were becoming squarer. Later specimens of Ursavus became larger, with squarer, more bear-like, molars.
• Protursus simpsoni (Pliocene; also "Indarctos") -- Sheepdog-sized. Carnassial teeth have no shearing action, molars are square, shorter tail, heavy limbs. Transitional to the modern genus Ursus.
• Ursus minimus (Pliocene) -- First little bear, with very bearlike molars, but still had the first premolars and slender canines. Shows gradual tooth changes and increase in body size as the ice age approached. Gave rise to the modern black bears (U. americanus & U. thibetanus), which haven't changed much since the Pliocene, and also smoothly evolved to the next species, U. etruscus:
• Ursus etruscus (late Pliocene) -- A larger bear, similar to our brown bear but with more primitive dentition. Molars big & square. First premolars small, and got smaller over time. Canines stouter. In Europe, gradually evolved into:
• Ursus savini (late Pleistocene, 1 Ma) -- Very similar to the brown bear. Some individuals didn't have the first premolars at all, while others had little vestigial premolars. Tendency toward domed forehead. Slowly split into a European population and an Asian population.
• U. spelaeus (late Pleistocene) -- The recently extinct giant cave bear, with a highly domed forehead. Clearly derived from the European population of U. savini, in a smooth transition. The species boundary is arbitrarily set at about 300,000 years ago.
• U. arctos (late Pleistocene) -- The brown ("grizzly") bear, clearly derived from the Asian population of U. savini about 800,000 years ago.. Spread into the Europe, & to the New World.
• U. maritimus (late Pleistocene) -- The polar bear. Very similar to a local population of brown bear, U. arctos beringianus that lived in Kamchatka about 500,000 years ago (Kurten 1964).

The transitions between each of these bear species are very well documented. For most of the transitions there are superb series of transitional specimens leading right across the species "boundaries". See Kurten (1976) for basic info on bear evolution.

Raccoons (procyonids):

• Phlaocyon (Miocene) -- A climbing carnivore with non-shearing carnassials and handlike forepaws, transitional from the arctoids to the procyonids (raccoons et al.). Typical raccoons first appeared in the Pliocene.

Weasels (mustelids):

• Plesictis (early Oligocene) -- Transitional between miacids (see above) and mustelids (weasels etc.)
• Potamotherium (late Oligocene) -- Another early mustelid, but has some rather puzzling traits that may mean it is not a direct ancestor of later mustelids. Mustelids were diversifying with "bewildering variety" by the early Miocene.

Pinniped relationships have been the subject of extensive discussion and analysis. They now appear to be a monophyletic group, probably derived from early bears (or possibly early weasels?).

Seals, sea lions & walruses:

• Pachycynodon (early Oligocene) -- A bearlike terrestrial carnivore with several sea-lion traits.
• Enaliarctos (late Oligocene, California) -- Still had many features of bear-like terrestrial carnivores: bear- like tympanic bulla, carnassials, etc. But, had flippers instead of toes (though could still walk and run on the flippers) and somewhat simplified dentition. Gave rise to several more advanced families, including:
• Odobenidae: the walrus family. Started with Neotherium 14 my, then Imagotaria, which is probably ancestral to modern species.
• Otariidae: the sea lion family. First was Pithanotaria (mid- Miocene, 11 Ma) -- small and primitive in many respects, then Thalassoleon (late Miocene) and finally modern sea lions (Pleistocene, about 2 Ma).
• Phocidae: the seal family. First known are the primitive and somewhat weasel-like mid-Miocene seals Leptophoca and Montherium. Modern seals first appear in the Pliocene, about 4 Ma.

Now, on to the second major group of carnivores, the cat/civet/hyena line. Civets (viverrids):

• Stenoplesictis (early Oligocene) -- An early civet-like animal related to the miacids. Might not be directly ancestral (has some puzzling non-civet-like traits).
• Palaeoprionodon (late Oligocene, 30-24 Ma) -- An aeluroid (undifferentiated cat/civet/hyena) with a civet-like skull floor. Probably had split off from the cat line and was on the way to modern viverrids.
• Herpestides (early Miocene, 22 Ma, France) -- Had a distinctly civet-like skull floor, more advanced than Palaeoprionodon.
• More advanced modern civets appeared in the Miocene.

Cats:

• Haplogale (late Oligocene, 30 Ma) -- A slightly cat-like aeluroid (cat/civet/hyena).
• "Proailurus" julieni, (early Miocene) -- An aeluroid with a viverrid-ish skull floor that also showed the first cat-like traits. The genus name is in quotes because, though it was first thought to be in Proailurus, it's now clear that it was a slightly different genus, probably ancestral to Proailurus.
• Proailurus lemanensis (early Miocene, 24 Ma) -- Considered the first true cat; had the first really cat-like skull floor, with an ossified bulla.
• Pseudaelurus (early-mid Miocene, 20 Ma) -- A slightly later, more advanced cat.
• Dinictis (early Oligocene) -- Transitional from early cats such as Proailurus to modern "feline" cats
• Hoplophoneus (early Oligocene) -- Transitional from early cats to "saber-tooth" cats

Hyaenids:

• Though there are only four species now, hyaenids were once very common and have an abundant fossil record. There is a main stem of generally small to medium-sized civet-like forms, showing a general trend toward an increase in size (Werdelin & Solounias, 1991):
• Herpestes antiquus (early Miocene) -- A viverrid thought to be the ancestor of the hyenid family.
• Protictitherium crassum (& 5 closely related species) (early Miocene, 17-18 Ma) -- Fox-sized, civet-like animals with hyena-like teeth. Transitional between the early civet-like viverrids and all the hyenids. Split into three lines, one of which led to the aardwolf. Another line eventually led to modern hyenas:
• Plioviverrops orbignyi (& 3 closely related species)
• Tungurictis spocki, a mid-Miocene fox-sized hyenid. Truly hyena-like ear capsule.
• Ictitherium viverrinum (& 6 closely related species)
• Thalassictis robusta (& 5 other spp.)
• Hyaenotherium wongii
• Miohyaenotherium bessarabicum
• Hyaenictitherium hyaenoides (& 3 other spp.)
• Palinhyaena reperta
• Ikelohyaena abronia
• Belbus beaumonti
• Leecyaena lycyaenoides (& 1 other) We're now in the Pliocene.
• Parahyaena brunnea
• Hyaena hyaena. Pliocrocuta (below) split off from Hyaena via cladogenesis. Hyaena itself continued on mostly unchanged as the modern striped hyena, with one more recent offshoot, the brown hyena,
• Hyaena brunnea.
• Pliocrocuta perrieri
• Pachycrocuta brevirostris (& 1 other)
• Adcrocuta eximia, which split into: Crocuta crocuta (the modern spotted hyena), C. sivalensis, and C. dietrichi.

Species-species transitions among carnivores:

• Ginsburg (in Chaline, 1983) describes gradual change in the early cats, from Haplogale media to Proailurus lemansis, to (in Europe) Pseudaelurus transitorius to Ps. lorteti to Ps. rmoieviensis to Ps. quadridentatus. These European lineages gave rise to the modern Lynx, Panthera, etc. Different lineages of Pseudaelurus evolved in North American, Africa, and Asia.
• Hecht (in Chaline, 1983) describes polar bear evolution; the first "polar bear" subspecies, Ursus maritimus tyrannus, was a essentially a brown bear subspecies, with brown bear dimensions and brown bear teeth. Over the next 20,000 years, body size reduced and the skull elongated. As late as 10,000 years ago, polar bears still had a high frequency of brown-bear-type molars. Only recently have they developed polar-bear-type teeth.
• Kurten (1976) describes bear transitions: "From the early Ursus minimus of 5 million years ago to the late Pleistocene cave bear, there is a perfectly complete evolutionary sequence without any real gaps. The transition is slow and gradual throughout, and it is quite difficult to say where one species ends and the next begins. Where should we draw the boundary between U. minimus and U. etruscus, or between U. savini and U. spelaeus? The history of the cave bear becomes a demonstration of evolution, not as a hypothesis or theory but as a simple fact of record." He adds, "In this respect the cave bear's history is far from unique."
• Kurten (1968) also described the following known species-species transitions:
  • Felis issiodorensis to Felis pardina (leopards)
  • Gulo schlosseri to Gulo gulo (wolverines)
  • Cuon majori to Cuon alpinus (dholes, a type of short-faced wolf)
• Lundelius et al. (1987) describe a study by Schultz in 1978 that showed an increase in canine length leading from the dirk-tooth cat Megantereon hesperus to Megantereon/Smilodon gracilis, then to Smilodon fatalis (a saber-toothed cat), and then to Smilodon californicus. Note the genus transition and the accompanying striking change in morphology.
• Werdelin & Solounias (1991) wrote an extensive monograph on hyenids. They discuss over one hundred (!) named species, with extensive discussion of the eighteen best-known species, and cladistic analysis of hundreds of specimens from the SIXTY-ONE "reasonably well known" hyaenid fossil species. They concluded:

  "We view the evolution of hyaenids as overwhelmingly gradual. The species, when studied with regard to their total variability, often grade insensibly into each other, as do the genera. Large specimens of Hyaenotherium wongii are, for example, difficult to distinguish from small specimens of Hyaenictitherium hyaenoides, a distinct genus. Viewed over the entire family, the evolution of hyaenids from small, fox-like forms to large, scavenging, "typical" hyenas can be followed step by step, and the assembly of features defining the most derived forms has taken place piecemeal since the Miocene. Nowhere is there any indication of major breaks identifying macroevolutionary steps."

Rodents

Lagomorphs and rodents are two modern orders that look superficially similar but have long been thought to be unrelated. Until recently, the origins of both groups were a mystery. They popped into the late Paleocene fossil record fully formed -- in North America & Europe, that is. New discoveries of earlier fossils from previously unstudied deposits in Asia have finally revealed the probable ancestors of both rodents and lagomorphs -- surprise, they're related after all. (see Chuankuei-Li et al., 1987)

• Anagale, Barunlestes, or a similar anagalid (mid-late Paleocene) -- A recently discovered order of primitive rodent/lagomorph ancestors from Asia. Rabbit-like lower cheek teeth, with cusps in a pattern that finally explains where the rabbits' central cusp came from (it's the old anagalid protocone). Primitive skeleton not yet specialized for leaping, with unfused leg bones, but has a rabbit-like heel. No gap yet in the teeth. These fossils have just been found in the last decade, and are still being described and analyzed. Barunlestes in particular (known so far from just one specimen) has both rodent-like and rabbit-like features, and may be ancestral to both the rodents and the lagomorphs. This lineage then apparently split into two groups, a eurymyloid/rodent-like group and a mymotonid/rabbit-like group.
• Heomys (mid-late Paleocene, China) -- An early rodent-like eurymyloid. Similar overall to Barunlestes but with added rodent/lagomorph features (enamel only on front of incisors, loss of canines and some premolars, long tooth gap) plus various rodent-like facial features and rodent-like cheek teeth. Probably a "cousin" to the rodents, though Chuankuei-Li et al (1987, and in Szalay et al. 1993) think it is "very close to the ancestral stem of the order Rodentia."
• News flash Tribosphenomys minutus (late Paleocene, 55 Ma) -- A just-announced discovery; it's a small Asian anagalid known from a single jaw found in some fossilized dung (well, we all have to die somehow). It still had rabbit-like cheek teeth, but had fully rodent-like ever-growing first incisors. This probably is the "ancestral stem" of the rodents. (see Discover, Feb. 1995, p. 22).
• Acritoparamys (was "Paramys") atavus (late Paleocene) -- First known primitive rodent.
• Paramys & its ischyromyid friends (late Paleocene) -- Generalized early rodents; a mostly squirrel-like skeleton but without the arboreal adaptations. Had a primitive jaw musculature (which modern squirrels still retain). Rodent-like gnawing incisors, but cheek teeth still rooted (unlike modern rodents) and primitive rodent dental formula.

Squirrels:

• Paramys (see above)
• Protosciurus (early Oligocene) An early squirrel with very primitive dentition and jaw muscles, but with the unique ear structure of modern squirrels. Fully arboreal.
• Sciurus, the modern squirrel genus. Arose in the Miocene and has not changed since then. Among the rodents, squirrels may be considered "living fossils".

Beavers:

• Paramys (see above)
• Paleocastor (Oligocene) -- Early beaver. A burrower, not yet aquatic. From here the beaver lineage became increasingly aquatic. Modern beavers appear in the Pleistocene.

Rats/mice/voles:

• Paramys (see above)
• Eomyids -- later Eocene rodents with a few tooth and eyesocket features that show they had branched off from the squirrel line.
• Geomyoids -- primitive rodents that have those same tooth & eyesocket features, and still have squirrel-like jaws; Known to have given rise to the mouse family only because we have intermediate fossil forms.
• In the Oligocene these early mice started to split into modern families such as kangaroo rats and pocket gophers. The first really mouse- like rodent, Antemus, first appeared in the Miocene (16 Ma) in Asia. In the Plio-Pleistocene, modern mice, hamsters, and voles appeared and started speciating all over the place. Carroll (1988, p. 493) has a nightmarish diagram of vole speciation which I will not try to describe here! The fossil record is very good for these recent rodents, and many examples of species-species transitions are known, very often crossing genus lines (see below).

Cavies:

GAP: No cavy fossils are known between Paramys and the late Oligocene, when cavies suddenly appear in modern form in both Africa and South America. However, there are possible cavy ancestors (franimorphs) in the early Oligocene of Texas, from which they could have rafted to South America and Africa. Known species-species transitions in rodents:

• Chaline & Laurin (1986) show gradual change in Plio-Pleistocene water voles, with gradual speciations documented in every step in the following lineage: Mimomys occitanus to M. stehlini to M. polonicus to M. pliocaenicus to M. ostramosensis. The most important change was the development of high-crowned teeth, which allows grass-eating. They say: "The evolution of the lineage appears to involve continuous morphological drift involving functional adaptation processes. It presumably results from changes in diet when Pretiglian steppes were replaced in Europe by a period with forest...In our opinion phyletic gradualism [in this lineage] seems well characterized. It lasts for 1.9 my and leads to very important morphological changes, and the transitional stages in the chronomorphocline are sufficiently easily recognizable that they have been described as morphospecies..."
• In a previous paper, Chaline (1983, p. 83) surveyed speciation in the known arvicolid rodents. About 25% of the species have fossil records complete enough to study the mode of appearance. Of those 25%, a wide variety of modes was seen, ranging sudden appearances (taken to mean punctuated equilibrium), to quick but smooth transitions, to very slow smooth transitions. Both cladogenesis and anagenesis occurred. Overall, smooth species-to-species transitions were seen for 53% of the studied species, but no single mode of evolution was dominant.
• Chevret et al. (1993) describe the transition from mouse teeth to vole teeth (6-4.5 Ma).
• Fahlbusch (1983) documents gradual change in various Miocene rodent transitions.
• Goodwin (in Martin, 1993) describes gradual transitions in prairie dogs, with Cinomys niobrarius increasing in size and splitting into two descendants, C. leucurus and C. parvidens.
• Jaeger (in Chaline, 1983) describes gradual shifts in tooth size and shape two genera of early mice, related to the development of grazing.
• Kurten (1968) describes a transition in voles, from Lagurus pannonicus to L. lagurus.
• Lundelius et al. (1987) summarizes and reviews species-species transitions in numerous voles, grasshopper mice, jumping mice, etc., from at least 11 different studies. Ex: Sigmodon medius to Sigmodon minor, and Zapus sandersi to Zapus hudsonius. The authors point out that some promising, well-fossilized groups have not even been studied yet for species-to-species transitions (e.g. the packrats, Neotoma).
• Martin (1993) summarizes and reviews the numerous known Pleistocene rodent species-to-species transitions in muskrats, water voles, grasshopper mice, prairie voles, pocket gophers, and cotton rats. Michaux (in Chaline, 1983) summarized speciations in mice. He found a wide variety of modes of speciation, ranging from sudden appearance to gradual change.
• Rensberger (1981) describes a likely lineage in the development of hypsodonty (high-crowned teeth for eating grass), among seven species of meniscomyine rodents in the genus Niglarodon.
• Stuart (1982, described by Barnosky, 1987) showed smooth transitions in water voles, including a genus transition. Mimomys savini gradually lost its distinctive tooth characters, including rooted cheek teeth, as it changed into a new genus, Arvicola cantiana, which in turn smoothly changed into the modern A. terrestris.
• Vianey-Liaud (1972) showed gradual change in two independent lineages of the mid-Oligocene rodent genus Theridomys. For example, the molars become gradually more hypsodont over time from species to species.
• Vianey-Liaud & Hartenberger (in Chaline, 1983) also describe gradual shifts in size and shape in Eocene rodents (mainly theridomyids), concluding that gradual evolution explains their data better than punctuated equilibrium.

Part 1B

Contents

Part 2B