Mammal or Reptile?
Mammals are part of an evolutionary lineage called synapsids – a clade of amniotesnote recognizable by features like differentiated teeth and the presence of a single large opening behind the eye socket of the skull – that began to appear in the Late Carboniferous, about 310 million years ago. While true mammals would only appear in the Early Jurassic, alongside the dinosaurs in the Mesozoic, the non-mammalian synapsids ruled the Permian as the first and largest dominant land vertebrates, being highly numerous and diverse and occupying several niches in the ecossystem, from apex predators to massive herbivores. Most of them were wiped out during the Permian-Triassic mass extinction event, but a few continued to thrive in the Early Triassic even after that, with the dicynodonts in special becoming highly successful – those survivors are traditionally believed to have gone extinct up until towards the end of the period, leaving mammals and their closest relatives as the last synapsids for the rest of the age of the dinosaurs.
Considered to be good examples of evolution and missing links between mammals and reptile ancestors, non-mammalian synapsids were often called "mammal-like reptiles", due to the belief at the time that synapsids evolved from reptiles that gained increasingly more mammalian features. However, the term has fallen in disfavor in the scientific community nowadays, as these animals aren't considered reptiles anymore in the cladistic sensenote . Thus, the prefered word in technical literature for our non-mammalian ancestors and their relatives is "proto-mammals" or "stem-mammals" instead.
In media and popular consciousness, the most famous stem-mammal is without a doubt Dimetrodon (the animal portrayed in the image), no in small part thanks to its theatrical and distinguishable sail on its back and being one of the best-known stem-mammals, though it is commonly placed together with dinosaurs in toylines and media due to still being more reptile-like, despite being distant from them and living some tens of millions years before, in the Early Permian. Other non-mammalian synapsids are much more unheard of in fiction, often making small unnamed appearances and only getting more attention in natural documentaries focused on the Paleozoic and the Triassic. From these, the numerous dicynodonts Lystrosaurus, the apex predators gorgonopsians, like Gorgonops and Inostrancevia, the similarly sail-backed Edaphosaurus, the bulky Moschops, the much more mammalian Cynognathus and Thrinaxodon and others are the most likely to appear. This page focus on the non-stock stem-mammals and their evolution.
See Prehistoric Life - Mammals for the prehistoric animals that descended from the stem-mammals.
open/close all folders
Early Synapsids
When Mammals were still hairless "lizards"
- We traditionally call "pelycosaurs
" the most basal synapsids (the correct name for the mammal-like "reptiles"), or in other words, non-therapsid synapsids. The term is paraphyletic, i.e., it encompasses a common ancestor and some of its descendant sub-groups, but not all of them. They were the dominant group of land animal in the Early Permian (for the record, the Permian period was just before the Mesozoic era), until they were replaced in the Middle Permian by their descendants, the therapsids. Pelycosaurs were still lizard-like in general body shape but already showed mammalian traits: their head was laterally flattened and high-settled above the ground, and their teeth started to show some resemblance with ours. By far the most popular are Dimetrodon and, to a lesser degree, Edaphosaurus, because both shared a similar crest (the so-called "sail") on their back substained by elongated neural spines, still for uncertain purpose: a solar panel/radiator as traditionally said? A courtship device? Or both things? Most other "pelycosaurs" didn't have such a sail: these seem not to receive any attention, even in books. However, from little animals like the Carboniferous Archaeothyris, several big, cool-looking animals originated, not only the sailbacks. The plant-eating Cotylorhynchus, for example, was not only the biggest pelycosaur known so far, but also one of the oddest-looking, with its disproportionately small head compared to the bulky body; the more normal-looking Ophiacodon ("snake tooth") was one of the very first amniotes to reach large size (it was already living at the end of the Carboniferous); Varanosaurus (literally "monitor lizard") and the similarly-named Varanops ("monitor face") were even earlier than the latter. While Sphenacodon was a sort of sail-less Dimetrodon, and gave its name to the Dimetrodon's family, Sphenacodontids, which includes also other "sailed" members such as the croc-headed Secodontosaurus. The latter, because of its long slender jaws, had a weaker bite and arguably hunted smaller preys than the more robustly-skulled Dimetrodon.
Scales or Hairs?
- If you see old paleo art, expect to see scaly-skinned mammal ancestors. This is due to a long-standing Taxonomic Term Confusion: since they are classically called reptiles in Linnaean systematics, most old-fashioned artists use to draw them using actual modern reptiles as model. But horny scales are a exclusive thing of the diapsids aka true reptiles (and possibly anapsids aka near-reptiles). It's very unlikely that Dimetrodon and its kin developed horny scales only to lose them altogether after becoming mammals. On the other hand, modern birds still retain the old reptilian scales on their hind limbs... indeed, cladistically speaking, birds are true reptiles. In terms of fur, it's unlikely they had any - it is difficult to determine exactly when hair started to appear in mammal evolution as it is more difficult to fossilize, but the synapsid Estemmenosuchus was found to have smooth hairless skin, and since it was a therapsid, even if an early one, it was more advanced than the pelycosaurs, suggesting that fur only came up later on in the lineage.
Horned Mammal-Ancestors?
- Scientific names are often misleading. Tetraceratops, for example, means "four-horned face" but is not a middle-way between a Triceratops ("three-horned face") and a Pentaceratops ("five-horned face"): it was a synapsid, thought by some a "missing link" between "pelycosaurs" and true therapsids, but with four small horns above of its head. Tetraceratops could actually belong to the most basal group of therapsids, the obscure Biarmosuchians, whose Biarmosuchus is the prototype. Only a bit more advanced was Eotitanosuchus: despite its name ("dawn titan-croc"), it was not related with the dinocephalian Titanosuchus (see just below).
Early Therapsids
Carnivores, Omnivores, and Herbivores: The Dinocephalians
- "Therapsids" is the classic name for the most advanced mammal-like "reptiles"; cladistically speaking, however, it contains mammals as well, so we humans are therapsids in this sense. We'll use this term with the traditional meaning. It's worth noting that most therapsids have been discovered in two precise places in the world: South Africa and Russia. Dinocephalians ("terrible head") included the largest therapsids of the Paleozoic. They were bulky-bodied and large-headed, and lived in the Middle/Late Permian, but were wiped out by the Permian-Triassic extinction event. Some of these were herbivorous, like Moschops and Tapinocephalus ("humble head"); and others were more likely carnivorous: Anteosaurus note and Titanophoneus ("titanic murderer") or omnivorous (Titanosuchus, meaning "titanic croc", and Jonkeria). The meat-eating ones were the apex predators of the Middle Permian, but were later substituted by another therapsid group, the gorgonopsians. The most awesome-looking among dinocephalians is probably the Russian herbivore Estemmenosuchus ("crowned croc"), with its bony, almost moose-like protrusions on its head whose purpose is uncertain. Despite their diversity, all dinocephalians shared common traits in their dentition. Curiously, Anteosaurus has often been depicted in drawings with a literal lion mane around its neck!
The Last Hunters before the Catastrophe: The Gorgonopsians
- Recently popularized by Walking with Monsters in the 2000s, gorgonopsians ("monstrous face") were the top predators of the Late Permian, but they too were deleted by the aforementioned mass extinction. More slender and usually smaller than dinocephalians, they are nicknamed "sabretooth" just like their mammalian namesakes; however their upper canines, though longer than most therapsids, were far less developed than those of a sabretooth-cat. The prototypical Gorgonops, the wolf-sized Lycaenops ("wolf face", often shown hunting Dicynodon) and the similar-sized Sauroctonus (portrayed in an old painting with the much bigger armored herbivorous near-reptile Scutosaurus), and the cow-sized Inostrancevia (named after a Russian geologist, and perhaps the biggest of the group) are among the most portrayed. Gorgonopsians, Cynognathus and other carnivorous therapsids are often described as dog-looking; indeed, in modern depictions, this resemblance is even more evident than in the older, more reptilian-like portraits.
From Mole-like to Elephant-like: The Dicynodonts
- Dicynodon has given his name to the Dicynodonts, the most successful group of herbivorous therapsids. They appeared in Late Permian as small diggers such as the tiny Diictodon ("two weasel teeth", portrayed in Walking with Monsters), Robertia broomiana (which honors Robert Broom, one of the greatest therapsid experts, which described in early XX century many south-africans mammal-ancestors), and the toothless Cistecephalus; flourished in the Early Triassic with Lystrosaurus; and then become bulky and vaguely elephant-like at the end of the Triassic: Russian/South African Kannemeyeria (named after Kannemeyer), Chinese Sinokannameyeria (meaning "Kannemeyeria from China" indeed), the Walking with Dinosaurs-famed North American Placerias ("broad body"), and the gigantic Polish Lisowicia (the biggest and last dicynodont known, the bulk of a small elephant, and the biggest non-mammal synapsid) are four main examples. Traditionally thought to have disappeared at the Triassic-Jurassic border, a recent discover seems indicating some Australian dicynodonts managed to make their way even in the Cretaceous. One thing does unify dicynodonts: their jaws. They had only two teeth at all, the upper canines vaguely Dracula-like, while the tip of their mouth was a sort of tortoise-like beak for cutting vegetation. Some species however lacked even these teeth, but still they made their way very well.
Advanced Therapsids
Very Mouse-Like: The last Cynodonts
- Not only the classic carnivorous Cynognathus and Thrinaxodon: the cynodonts were very diverse in habits. The closest-to-mammals were often not predatory at all: traversodonts (Traversodon, Massetognathus, and others), for example, were omnivores or even herbivores; tritylodonts (so-named from their prototype Tritylodon) even developed rodent-like teeth for gnawing. However, both achieved their traits independently from modern herbivores, and were not direct ancestors of mammals. However, the last common mammal ancestor has surely to be searched among cynodonts. Tritylodonts were particularly close to mammals, while the Traversodonts were closer to Cynognathus than to mammals in the evolutive tree. In the Late Triassic, cynodonts were the only successful therapsid group: dicynodonts were still surviving but were rare at that point. Non-mammalian cynodonts survived until the Early Jurassic with two groups, the Tritylodonts like Oligokyphus, and the Ictidosaurs ("weasel-lizards") like Tritheledon. Tritylodonts reached even the Early Cretaceous, with the enigmatic Xenocretosuchus ("strange cretaceous croc"). But cynodonts sensu stricto made a minor part of the synapsid fauna after Triassic: their true mammalian descendants were dominant at this point.
A Great Unknown Epic
- Non-mammalian synapsids were extraordinarily numerous in prehistory. More than 500 genuses have been described so far, expecially among the therapsids. The most abundant and diversified subgroups of therapsids were the Dinocephalians (Moschops and kin), the Dicynodonts (Lystrosaurus and kin), and the Theriodonts (meaning "mammal-tooth"). The last term is now usually in disuse (being probably paraphyletic) and indicates the more mammal-looking therapsids, like the relatively few gorgonopsians (all similar to each other) and the much more numerous & diversified cynodonts. Other subgroups of therapsids were very generic-looking and are usually ignored by paleo-artists: for example the basal anomodonts, closely related with the dicynodonts (aka the advanced anomodonts) but more primitive. One example of non-dicynodontian anomodont was the multi-toothed, beakless and lizard-shaped Galepus ("weasel-foot"). There is still another numerous & diversified group of "theriodont" therapsids to be mentioned, considered the sister-group of the cynodonts: see below.
Obscure Relatives: The Therocephalians
- Among the least portrayed among the main therapsid groups, therocephalians ("mammal-head") were similar in size and shape to the cynodonts, but less mammal-like, and only distantly related to mammals. Therocephalians began in the Permian and survived until the Early Triassic, but then were replaced by their more evolved relatives, cynodonts indeed. Just like the latter, they were initially carnivores (ex. Lycosuchus "wolf-croc", and Pristerognathus, "saw-jaw"), but then some of the last forms became vegetarian (Bauria, once considered in a distinct group, the Bauriamorphs), while others became more similar to lizards (Ericiolacerta: "lacerta" just means lizard in Latin). One unidentified predatory therocephalian appears in Walking with Monsters, portrayed with a totally speculative venomous bite; it could be Euchambersia.
Thank You Dinosaur! MESOZOIC MAMMALS
Milk, Hair, Limbs, Jaws & Ears
- The boundary between mammals and non-mammals has always been a hard issue for paleontologists. Since typical mammalian features such as hair, milk glands, etc... do not fossilize most the time, the key to separate the two ensembles lays in their skull. True mammals have a mandible made by a single couple of bones, and three ossicles in the mid-ear. Non-mammalian synapsids have several pairs of bones in the lower jaw and a single ossicle in the ear. It's also worth noting that mammalian features probably didn't appear all in the same instant: perhaps some therapsids already produced milk and had hair, though they didn't have erect limbs yet, unlike modern mammals (except platypus and echidna, aka the Monotremes, that still have splayed limbs and primitive milk glands). Some quasi-mammals (more correctly called Mammaliaforms) began in Late Triassic and were tiny, very shrew-like, and insectivorous: Morganucodon and Megazostrodon are the two most portrayed. Both were once classified as "triconodonts", but today this term only indicates some more evolved true mammals from the Jurassic and Cretaceous, like the cat-sized Triconodon indeed. Another group of mammaliaforms were the omnivorous docodonts, which managed to reach the Late Jurassic with species such as the namesake Docodon. The first true mammals (Mammalia) appeared in the early Jurassic, and were shrew-like just like their Triassic mammaliaform ancestors. They remained so for all the Mesozoic... at least this is what scientists used to think. Traditionally, fossils of Mesozoic mammals are extremely rare and fragmentary due to their smallness; but some very interesting new mammal fossils have been discovered since the Turn of the Millennium, and we now know mammals and even mammaliaforms were already very diversified at the Age of Dinosaurs. Some were mole-like diggers (Fruitafossor), some were beaver-like swimmers (Castorocauda), and some were even gliders (Volaticotherium).
Insignificant Critters?
- If you'll read a paleo book you have good chance to see Mesozoic mammals/mammaliaforms described as insignificant little creatures ruled by the mighty dinosaurs. Actually, thanks to their possibly dense populations, these synapsids could have affected their ecosystem the same way dinosaurs did; and remember that small animals are often key species in their natural environments. Another unexpected discovery from the 2000s showed Mesozoic mammals not being necessarily preys for dinosaurs as well: the badger-sized Repenomamus was discovered with baby dinosaur remains in its stomach. Another commonplace to debunk is that Mesozoic mammals were all insectivores. Actually, a whole group, the Multituberculates were rodent-like and herbivorous: their name "multi-tubercled tooth" is due to an unique couple of protruding cheek-teeth. They were the most abundant early mammals at the end of the Cretaceous, and managed to survive after the mass extinction. At the beginning of the Cenozoic they became even more successful, until true rodents replaced them in the Oligocene (or not). Multituberculates were the longest living mammalian group ever before gone extinct. One of the largest, Taeniolabis from the Palaeocene, weighed 100 kg (the bulk of a giant panda). The direct common ancestor of modern mammalian groups (placentals, marsupials, and monotremes) is unknown, but all the three groups became widespread only in the Cretaceous, expecially in Late Cretaceous. We can mention: the platypus-like Steropodon, an early monotreme belonging to a distinct lineage than the ones of true platypuses & echidnas, portrayed in Walking with Dinosaurs as a scavenger; the early otter-like "marsupial" Didelphodon, also portrayed in WWD as an opportunist; and the oddly-named placental Purgatorius, which is often considered the first known ancestor of primates, or at least, a close relative. Together, eutherian mammals (the placentals) and metatherians (the marsupials) make their own group: the Therians (literally "the beasts" in Greek). Monotremes, on the other hand, are much more primitive than the former, and are traditionally called Prototherians ("the first beasts"). You could also read the names "allotherians", "symmetrodonts", and "pantotherians" especially in older texts. Allotherians ("the other beasts") included the multituberculates and their relatives; Symmetrodonts ("symmetrical teeth") had three-pointed cheek-teeth, and the word still indicates a natural group of mesozoic mammals though more strict than formerly; "Pantotherians" ("the totally-beasts") included the common ancestors of marsupials and placentals. As you'll see in the following mammal section, -therium is the common suffix for most extinct mammals, a bit like -saurus for extinct reptiles.
Dawn Parents
- In the 2000s two animals discovered from the famous Early Cretaceous deposits of China were object of some sensationalism: Sinodelphys the "first marsupial ever", and even more Eomaia "the first ever placental" and thus "the first Man's ancestor" ("Eomaia" meaning "dawn mother"). However, as mammal fossils from the Mesozoic are such a rarity, it's virtually impossible understanding which one was really the most basal placental / marsupial. Both are very precious, though, because they have preserved their fur — before that, the oldest fossilized furs were from the Early Cenozoic (the famous Messel tar pits of Germany). Finally, let's debunk another tenacious myth about mammal evolution: we must thank if non-avian dinosaurs went extinct, otherwise humans couldn't have appeared on Earth. Maybe we could have appeared just the same, maybe a bit later... It's more probable that dinos actually guided mammal evolution in an indirect way. Being competitors of and preying upon our ancestors, they selected actively the most adapted, most evolved traits us mammals are proud of: among them, intelligence and parental care. If you are here to read this now, it's in part thanks to the dinosaurs, in a way. Everything has always been better with dinosaurs!
