Fins to Feet

Pelycosaurs

This post is a continuation of “Coming of the Amniotes”

As the carboniferous period drew to a close, the tropics began drying up. The swampy equator-spanning rainforests of clubmosses, ferns and horsetails were gradually fragmented and replaced by communities of seed ferns and arid-adapted conifers. Amphibian diversity was on the decline and the amniotes – better equipped for dealing with the trials of a dry world – were well on their way to becoming the dominant vertebrate group on land. In this post, we shall consider the early evolution of an important clade of Amniotes that includes modern mammals as its only surviving members - the Synapsids.

What are pelycosaurs?

Primitive or “basal” synapsids from the late carboniferous and Permian periods are often referred to informally as “pelycosaurs” – although the term has fallen into disfavor of late among paleontologists for various reasons, it rolls off the tongue easily and is convenient for our purposes, and so we shall use it. Crudely put, the Pelycosaurs represent the very first step on the road to mammalhood in the reptile-to-mammal transition.

Pelycosaurs are distinguished from non-synapsid amniotes by a number of features, the most important of which is the presence of one hole on each side of the skull behind the eye-orbit, bounded above by the post-orbital and squamosal bones. These holes are called temporal fenestrae. A diagram will be instructive.

By contrast, other early amniote skulls bear either no holes (anapsids), two holes (diapsid) or a single highly placed hole (Euryapsid, similar to the synapsid condition, but differing in the location of the hole) behind each eye-orbit. The temporal fenestrae provide secure anchorage points for muscles associated with jaw function.

Pelycosaurs share a number of other general skull features. For example, the back of the skull slopes forward and the post-parietal bone is small and single rather than paired. These minor anatomical details need not concern us unduly here – and we shall take a cursory tour through the anatomy of the most famous Pelycosaur, Dimetrodon, in the closing section of this essay. The skeletons of pelycosaurs are structurally quite similar.

What did they look like? What did they eat?

Although Pelycosaurs share a closer phylogenetic affinity with mammals than with crocodiles or lizards, there is little externally mammalian about them. They were sprawling, scaly, ectothermic creatures and, if you discovered one ambling through your backyard, you’d probably class it, quite understandably, among the reptillia.

Archeothyris is among the earliest known pelycosaurs. It was a lizard-like creature, about 20 inches long with short limbs, a long snout and a number of slightly enlarged canine-like teeth towards the front of the mouth. Another early pelycosaur, Eothyris, had two pairs of very large teeth protruding from the upper jaw. These dental features suggest carnivory.  Here we see the beginnings of differentiation in the structure and function of the teeth among synapsids (more ‘primitive’ tetrapods have teeth that are essentially identical in terms of morphology) – an evolutionary process that would ultimately lead to the organization of the teeth, in modern mammals, into distinct morphological types: incisors, canines, premolars and molars.

The Caesids were herbivorous pelycosaurs. They had large nostrils and a shortened lower jaw. The body is almost comically large in comparison to the skull. This is a mark of herbivory – the enlarged, barrel-shaped ribcage housed a gut that was used as a chamber for the bacterial fermentation and digestion of large amounts of plant matter. The teeth are spatulate rather than pointed. A second set of teeth project from the palate and probably worked in concert with a muscular tongue to grind down food. And the low position of the joint between the lower jaw and the skull (below the tooth row) has been interpreted as a marker of increased bite force. These adaptations are all geared towards the processing of vegetation.

The most famous pelycosaurs, however, are the sailbacks – animals with elongated “neural spines” arising from the vertebrae of the neck and back. These spines supported a large ‘sail’ – a thin enveloping sheath of skin, ligaments and blood vessels. We shall discuss the probable function of these sails in the next section. Sails are observed in two pelycosaur groups: the sphenacodontia and the edaphosauridae. It is thought that sails evolved independently in the two groups. Edaphosaurus (a member of the Edaphosauridae) was a herbivore – armed with a battery of teeth designed to crop and process vegetation. Edaphosaurus had a suite of adaptations for a herbivorous diet that are similar to those found in Caesids: low jaw-joint, large body, grinding teeth on the roof of the mouth etc

Tell me about Dimetrodon!

Dimetrodon, the most celebrated and best studied of all the pelycosaurs, is a member of the Sphenacodontia. Dimetrodon was a large, apex predator that dined on everything from sharks to large land-going amphibians.

Dimetrodon made an appearance on Henry Levin’s Journey to the Center of the Earth (1959). Instead of the standard 50s claymation, the viewer is treated to some hilariously awful footage of iguanas tromping around with sails taped to their backs. Set to the right music, however, these absurd scenes acquire new and profound meaning. Check 1:08 of this video. Woah, trippy.

The largest species of Dimetrodon hit a length of about 4 meters from nose to tail – making it the biggest land predator of its time. We shall begin our brief overview of the animal’s skeletal anatomy with the skull. Dimetrodon was an animal with a long, high snout. We notice the large opening behind the orbit, common to all early synapsids – the temporal fenestra.

The human jaw consists of a single bone, the dentary, which connects with the temporal bone of the skull. In Dimetrodon, however, the jaw is composed of multiple bones, of which the dentary is only one. Instead of the dentary-temporal bone connection seen in humans, we have a joint between the articular bone of the mandible and the quadrate bone of the skull. In an astonishing evolutionary transition, well-supported by the fossil record, the articular and quadrate bones would eventually be incorporated into the anatomy of the ear as ear ossicles (the malleus and incus; bones that help in the amplification of vibrations received by the ear-drum) in mammals. The dentary would ultimately come to be the sole bone in the mammalian mandible.

The position of the jaw-joint is noticeably lower than the tooth-row. The low placement of the joint lengthened the ‘moment arm’ of the force delivered by a set of muscles called adductors, which pull the jaw up and backwards, closing it. As a consequence, the torque experienced by the jaw as it snaps shut is greater. The back of the jaw is expanded into a “coronid eminence” to provide additional surface area for the attachment of muscles. These structural features of the jaw and its associated musculature, combined with the sturdy construction of the skull, enabled Dimetrodon to effectively snag and hold onto large, struggling prey with its mouth.

Dimetrodon walked with a sprawling gait, similar to that of many modern reptiles. The humerus and femur projected almost horizontally from the shoulder girdle and pelvis respectively. Rotation along the long axes of these bones (which moved the lower limbs through a broad arc) was important for locomotion.  Units called “intercentra” were present between the vertebrae – these intervening skeletal elements were lost over the course of amniote evolution in both synapsid and non-synapsid lines.

We now turn to the possible function of the animal’s spectacular dorsal sail:

Thermoregulation? Modern reptiles have evolved an interesting complement of biological adaptations to maximize the rate at which their bodies absorb heat and minimize the rate at which they lose it. This involves things like adjusting the distribution of blood flow in the body and controlling body color. Some reptiles also have bodies that are large enough to effectively store thermal energy due to their low surface-area-to-body-volume ratios  (think dinosaurs or certain modern crocodilians), a trait called “Gigantothermy”. The sail of Dimetrodon was, in all likelihood, well-supplied with blood vessels and may have acted as a powerful heat-exchanger. Dimetrodon could have basked with its sail positioned laterally to the sun’s rays, soaking up heat energy like a solar panel. The speed with which the animal could heat up, thanks to its sail, may have given it certain advantages over its poikilothermic prey. Dimetrodon might have achieved the minimization of unwanted heat-loss from the sail by methods similar to those employed by reptiles today – by varying the blood supply to the sail or by adjusting the coloration of the sail. Later Dimetrodon species were certainly large enough to maintain a constant body temperature over extended periods of time by mere virtue of their body size.

Here we see a possible early trend towards the regulation of body temperature in the remote ancestors of modern mammals.

Sexual display? The sail may have been a colorful sexual display used to attract potential mates. Examples of such secondary sexual characteristics that determine mating success are common throughout the animal kingdom, from the sail-like dorsal fins of certain mollies to the tail feathers of peacocks. It is unfortunate that we will never know what colors painted the strange sails of these pelycosaurs.

Other explanations involve the sail’s utility in swimming (where it really would function as a “sail” of the naval variety) or as a means of camouflage among the reeds, though these seem somewhat unlikely in my opinion.

The Sphenocodontia (the group to which Dimetrodon belongs) were ancestral to modern mammals (though the Sphenocodontid ancestor of mammals almost certainly did not bear a sail). The story of mammal origins is long and complex – but with our coverage of the pelycosaurs, we have set the stage for the next major leap – the origin of the Therapsids!

Coming of the Amniotes

Steamy, water-logged swamps, primitive amphibians trawling the undergrowth for food and gigantic insects buzzing overhead – this was the image of the Carboniferous period (which lasted from about 360 to 300 million years ago) that every illustrated natural history book I ever read as a child drilled into my mind’s eye. Great rainforests of clubmosses, horse-tails and ferns ranged throughout the tropics. Oxygen levels were much higher than they are today – allowing invertebrates to grow to extraordinary sizes. Amphibians were the dominant land vertebrates during this time and ran the gamut from small newt-like creatures to large predators similar in appearance to modern crocodiles. It was in this lush, wet world that the very first Amniotes, the rather diminutive ancestors of modern reptiles, mammals and birds, made their debut. This section of Fins to Feet, will deal with the appearance of this animal group – a critical event in evolutionary history that completed the vertebrate conquest of land.

The amniote kinship of mammals and reptiles

In contrast to Amphibians, both Mammals and Reptiles are fully terrestrially adapted organisms. The exceptions to this rule include a relatively small number of reptile and mammal lineages that became secondarily adapted to living in the water – like, for example, whales and ichthyosaurs – though they still retain a number of features, common to all ‘amniotes’, that betray their terrestrial ancestry.

Both Mammals and Reptiles trace their origins to a common amniote ancestor – in other words, to creatures that laid waterproof eggs on dry land.

The first Amniotes were not the first backboned animals to walk on land – though they were more practiced land-lubbers than any vertebrate group that preceded them. Tetrapods, or four limbed vertebrates, made the first tentative forays onto land in the swamps of the late Devonian (374 – 359 million years ago). The epic vertebrate transition from water to land (the event referenced in the very title of this blog) will be the subject of a future post – and it can be regarded as having ended with the coming of the Amniotes. Early Amniotes appeared some 350 million years ago (in the first half of the Carboniferous period) in terrestrial environments already overrun with plant-life, insects and land-walking amphibians.

So what, then, are these “amniotes”? What distinguishes them from other tetrapods?

In younger years, I would have seen the anamniote-amniote transition as simply the transition from “amphibian” to “reptile” – and while this is not an altogether terrible way to think about it, some caveats are necessary. Modern amphibians, or Lissamphibians – that is, frogs, salamanders, newts, toads and the less familiar caecilians – are our only existing examples of non-amniote tetrapods. But we must not make the mistake of drawing a neat equivalence between modern Lissamphibia and the ‘amphibious’ tetrapod ancestors of amniotes. Lissamphibia lack ribs, have smooth, wet skins highly-adapted for cutaneous respiration and, for the most part, lead arboreal, aquatic or fossorial lifestyles. The ancient “amphibians” from which amniotes are derived were, by contrast, stout ribbed ground-dwelling creatures with bodies covered in thick dermal plates.

The earliest uncovered examples of Amniotes resemble  insectivorous terrestrial lizards,  at least superficially. These include fossil species like Hylonomus and Paleothyris. They have slender bodies that run a length of about 20 centimeters from nose to tail-tip. They shared their world with a wide range of small to giant primitive amphibians that together made up the bulk of vertebrate terrestrial life on earth, including – for example – temnospondylls and anthracosaurs. These amphibians lived under various degrees of dependence on bodies of water for reproduction and egg-development. The amniotes achieved emancipation from the aquatic realm through a number of important biological innovations. Foremost among these was the Amniote egg.

The Amniote egg is characterized by a hard semi-permeable shell membrane and various “extra-embryonic” tissues that surround the developing embryo  – these include the amnion, the allantois and the chorion. These eggs could be laid on relatively dry land, away from the hydrating embrace of a pond or river. Terrestrial egg-laying probably reduced the risk of predation.

Later in evolutionary history, various amniote groups would make massive modifications to this basic design – mammals, for example, did away with the fibrous shell and allowed embryonic development to proceed entirely within the uterus – but for now, let us consider what makes the primitive amniote egg a “terrestrial” egg.

Most modern amphibians deposit their eggs in water. Once the eggs hatch, an aquatic larval period commences – a “tadpole” phase during which the organism is equipped with gills and fins and lacks limbs. Eventually the tadpole metamorphoses into the adult form, growing limbs and discarding its gills. Amniote eggs are laid on land and give rise to developed hatchlings – which are, essentially, miniature versions of the adult form- without an intervening larval phase. This lack of a free-swimming tadpole stage is also seen in certain amphibian species– largely occupants of wet and humid tropical rainforests – which lay their eggs away from free-standing water (instead, the embryo undergoes its tadpole stage within the confines of the egg). Such a “direct developing” amphibian egg might have served as the precursor to the amniote egg.

The typical amphibian egg is bound by a vitelline membrane and one or more layers of egg jelly (together constituting the “egg capsule”). These structures protect the egg and provide support, but they do not help prevent water-loss – and thus, the egg must be laid in water or in a wet environment to avoid desiccation. Amniote eggs are more resistant to water-loss and can be laid in drier environments.

The egg capsule of amphibians is more of a barrier to the diffusion of respiratory gases (namely carbon dioxide and oxygen) than is the fibrous egg shell membrane of modern reptiles. This imposes stricter limits on the size and metabolic activity of the growing embryo in amphibians. To illustrate: As embryo size increases, the radius of the egg increases. As the radius of a spherical egg increases, the volume (4/3 pi r^3) increases much faster than does the surface area available for gas exchange (4 pi r^2 ). Given that the egg capsule is a poor agent of respiratory exchange to begin with, this relationship means that oxygen uptake through the capsule is insufficient to support embryos beyond a certain size.  Furthermore, an increase in size also means an exponential increase in the weight of the supporting tissues around it. The egg capsule would have to get thicker to support the mass, thereby reducing effective diffusion.

Thus, the range of egg sizes that can be realized under the amphibian egg-plan is limited.

So how might early Amniotes have overcome this size/metabolic constraint? And how did they ultimately make the transition to egg-laying on dry land? 

1)      By enveloping the egg in a fibrous air-permeable shell membrane that (a) allowed for easier exchange of gases but kept fluids in and (b) provided enough mechanical support to resist the force of gravity on land. This proteinacious shell membrane is secreted onto the egg within the oviduct. This may have antecedents in certain amphibian species, where a layer of glycoproteins is secreted onto the egg-capsule in the oviduct before deposition.

2)      By transferring water from the egg-jelly to the yolk over evolutionary time, leaving the outer portions of the egg dry and effectively eliminating the egg-jelly as a barrier to gas diffusion. A process like this is observed during egg development in reptiles, when water is withdrawn from the albumen and transported to the yolk.

3)      By the evolution of an extra-embryonic membrane specialized for respiratory exchange called the chorioallantois. We shall discuss this briefly in my section on extra-embryonic membranes.

All reptile eggs measured to date are larger than those observed in amphibians and generally appear to display a higher rate of metabolic activity and oxygen consumption. Amniotes on the whole display a vast range of egg sizes. Freedom from the mentioned size constraints allowed for bigger offspring and, by extension, bigger adults. This newly available range of body-sizes was crucial to the evolutionary success of the amniotes in the long run.

Some researchers suggest that elevated oxygen levels in the late Carboniferous may have eased the various size constraints (related to gas diffusion) discussed earlier and helped along with the evolution of large eggs.

Why are Amniote eggs “yolkier” than the typical amphibian egg?

Since development within the egg among Amniotes is geared towards the production of a fully-formed miniature adult rather than a free-living, immature aquatic larva (as in most amphibians), a large store of nutrients (in the form of “yolk”) is needed to carry the embryo through the whole extent of development. In other words, amniote eggs are stocked up for the long haul – and have large yolk sacs richly provided with blood vessels to convey nutrients to the growing embryo. The increased nutritional requirement imposed by “direct development” is the reason Amniotes needed bigger eggs and larger amounts of yolk. Because more has to be invested in each egg, reptiles also tend to lay fewer eggs than amphibians or fish.

What about those extra-embryonic membranes?

The position and function of the various extra-embryonic membranes has been a source of confusion for me throughout the course of my schooling, but hopefully a simple diagram will help us suss out the complicated attendant tissues and boundaries that compartmentalize the egg and regulate material exchange both between compartments and between the egg and the external world. Although there are amphibians that lay eggs on land, none of them show the complex arrangement of extra-embryonic tissues depicted below.

Let us use the chicken egg, a familiar household food-item, as our model for understanding the essential structure of the amniote egg. The embryo and supporting tissues arise from a small disc of cells, called the blastoderm, settled on surface of the yolk of the egg. The movements of the various embryonic tissue layers need not concern us here, but sufficeth to say that the growth and rearrangement of the tissues of this disc gives rise to three extra-embryonic layers: [1] the chorion is the outer-most extra-embryonic layer and, apart from providing overall enclosure for the embryo, plays a role in respiration along with the allantois [2] the yolk sac encloses the yolk and is supplied with blood vessels that convey food from the yolk to the embryo [3] during development, the chorion folds over the embryo (as diagrammed below) to enclose it in the amniotic cavity. The fluid in this cavity buffers shocks and acts as a protective mechanical barrier. Finally, a fourth extra-embryonic membrane, called the allantois, develops as an outpocketing of the hindgut. It functions to store nitrogenous wastes (i.e. uric acid) produced by the metabolic activity of the chick. It expands to make contact with the undersurface of the chorion (forming the chorioallantois mentioned earlier) and serves as the principal respiratory organ for the embryo. It is richly supplied with blood vessels for gaseous exchange. As development proceeds, the yolk sac diminishes in size while the allantois grows.

The Albumen surrounds the yolk and provides additional support and nutrition.

Although mammals do not lay external eggs (with the exception of Platypi and echidnae), similar extra-embryonic membranes are seen around the developing fetus in the maternal womb.

So what really distinguishes the Amniote egg from the eggs of other tetrapods is the fibrous shell membrane and a host of extra-embryonic membranes. What else marked the first amniotes apart from their amphibious cousins?

The first Amniotes appeared to have smaller, narrower and deeper skulls than other tetrapods. A diagnostic feature that is often used to differentiate amphibians from amniotes is the absence in reptiles of an invagination called the “otic notch” which is present behind the eye-orbits in anamniote tetrapods. The anamniote-amniote transition also appears to have involved alterations in the musculature of the jaw, elaboration of the tongue, a reduction in the number of movable elements in the skull, strengthening of the ankles and the appearance of more slender bones. All in all, however, the skeletal differences between known basal amniotes and closely related tetrapods are rather minimal.

A large number of ancient amphibians, very probably including the carboniferous ancestors of amniotes, were armored in heavy dermal plates – the first amniotes appear to have traded these for horny keratinous epidermal scales that cover most of the body (with dermal gastralia on the underside). This loss of massy dermal bone lightened the body and made speedier locomotion on land possible.

What did the first amniotes eat?

Amniotes today exhibit a far wider range of diets than do amphibians – and this was key to their evolutionary success on land. Early fossil amniotes like Holonymus and Paleothyris have sharp teeth designed to pierce through the tough carapaces of invertebrates. It has been suggested that the radiation of amniotes may have to do with rising levels of insect diversity.

The development of terrestrial herbivory was a key event in the evolution of the amniotes and we shall spend a few moments trying to make sense of it.

Diadectes, a reptile-like amphibian closely related to basal amniotes, is one of the earliest terrestrial herbivores known to science (its diet is inferred from its dental and skeletal anatomy). The evolution of terrestrial herbivory (a trait unseen in modern amphibians, but common among amniotes) probably involved some sort of progression through the following stages: [a] an omnivorous stage where a diet of invertebrates was supplemented with low-fiber, high-nutrition, cellulose-poor plants or plant-parts (buds, shoots, young leaves etc.) [b] a stage in which the primary diet was high-quality, low-fiber plants and [c] finally, a stage of obligate herbivory where the diet consisted of abundant low-quality, high-fiber vegetation – [c] is an ecological role that has been continuously occupied by a succession of large land-walking animals, from Diadectes to Apatosaurus to modern grazing mammals, for the last 300 million years. High-fiber mature stems and leaves, though widely distributed, are not very digestible. They are rich in cellulose, a carbohydrate which cannot be broken down by the vertebrate digestive system without the aid of certain kinds of bacteria. It is likely that animals foraging in the leaf-litter picked up bacterial species that survived in the gut and, over evolutionary time, became members of the microbial gut flora of a species. What may have begun as incidental commensalism – where neither party benefited particularly from interaction – may have eventually evolved into symbiosis, where the bacteria broke down cellulose in ingested plant matter, making nutrients available to the host, while the host provided shelter and food to the bacteria. The disproportionately large, barrel shaped body in animals like Diadectes is designed to help bacterially ferment ingested plant material (bacterial fermentation is slow and a large space to store plant material for extended periods of time is supremely useful for an obligate herbivore). The evolution of terrestrial herbivory in Amniotes (and in closely related groups) allowed them to attain truly enormous sizes on land.

What divisions do we see among early amniotes?

We now turn, briefly, to consider some of the taxonomic divisions that can be established among early fossil amniotes by studying the structure of the skull. Temporal fenestrae are openings seen in the skull behind the eye-orbits. The establishment of these holes lightens the skull and provides additional edges for the attachment of muscles. The presence, number and position of these holes can be used to categorize amniotes. We observe four different conditions:

1)      Anapsid: No temporal fenestra is seen. Examples: Turtles and possibly Hylonomus and Paleothyris.

2)      Synapsid: A lower temporal fenestra is seen. Examples: Mammals and their extinct fossil relatives and ancestors

3)      Diapsid: Two temporal fenestrae are seen. Examples: Dinosaurs, birds, mosasaurs, lizards

4)      Euryapsid: A temporal fenestra is seen in the upper skull. This condition is probably derived from the Diapsid condition by the loss of the lower fenestra. Example: Icthyosaurs

The crucial bifurcation in the tree of Amniote life is between the Synapsida (mammals and their ancestors) and the Sauropsida (including animals with the Diapsid, Anapsid and Euryapsid conditions – that is, birds and reptiles).

The story of the Sauropsids and the Synapsids is matter for future posts. The sizes and shapes that amniotes would attain over the course of their evolution is truly astonishing. Once vertebrate life solidified its grip on land, sky really was the limit.