Tetrapods Answer

Ecological Setting

Opportunity knocked. By the Late Devonian, a variety of new and promising aquatic habitats had developed. Extensive terrestrial vegetation and accelerated soil formation moderated flow regimes and stabilized banks and bottom sediments. This in turn generated a variety of microhabitats ranging from deeper channels to shallow zones to periodically inundated wetlands. Detrital inputs from terrestrial and wetlands vegetation provided a trophic base for increasingly complex food webs. Aquatic plants, which are less readily fossilized than their terrestrial counterparts, probably flourished in some of the shallow zones and would have also added substantially to the food supply of Late Devonian ecosystems.

The exploitation of these stable and productive shallows and wetlands probably provided the driving force behind the evolution of the tetrapods. Refinements and variations of the adaptations and features that enabled tetrapods to use these new shallow habitats would later enable them to expand onto the land.

Exploiting weedy shallows and wetlands would provide new food resources and possibly provide a relatively protected nursery for the young as well as an adult refuge from giant lobe-fin predators (e.g., Eusthenodon, Hyneria and rhizodonts). But how did the early tetrapods evolve the ability to exploit this new world? A variety of anatomical evidence indicates that early tetrapods evolved from osteolopiform lobe-fins. These fishes exhibited a number of preadaptations that appear to have been crucial to the early tetrapods’ success. Divergences from the osteolepiform body plan illustrate how tetrapods answered.

Fins to Limbs

One of the most conspicuous of these preadaptations —and a defining feature of lobe-fins— is their fleshy fins. These fins have long, ray-like dermal bones (lepidotrichia) that characterize the limbs as fins, but they also contain a well developed endochondral or internal skeleton. All lobe-fins fins have single appendicular bones (humerus and femur) that articulate with the pectoral and pelvic girdles, respectively. The pattern for most lobe-fins is for the remaining endochondral bones to form a dominant central axis with numerous side branches. This pattern, considered basal (or primitive) is evident in the living lobe-fins, lungfishes (e.g., Neoceratodus) and the coelacanth (Latimeria) as well as in a variety of fossil forms (e.g., the porolepiform Glyptolepis).

A loss in dominance of the central endochondral axis occurs in two sister groups of lobe-fins, the osteolepiforms and the rhizodonts. In the pectoral fins (the front paired fin comparable to our arms), for instance, the humerus distally (away from the body) articulates with a pair of bone, the radius and ulna. The ulna, in turn, distally articulates with another pair, the ulnare and intermedium. These bones are homologous with the lower limb bones of early tetrapods and with our own.

lobefin pectoral finsPectoral fins of two living (Latimeria and Neoceratodus) and three Devonian (Glyptolepis, Sauruipteris and Eustenopteron) lobe-fins. The basal condition of a central endochondral axis is evident for Latimeria, Neoceratodus and Glyptolepis whereas the derived condition is shown in the other two. ©

However, the bones that distally articulate with the radius, ulnare and intermedium vary considerably within these two sister groups and are not as clearly comparable with those found in tetrapods. For example, the rhizodont Sauripterus has a total of eight sets of radials (sometimes dubbed "fingers"), two of which articulate with the radius, two with the intermedium and four with the ulnare. In contrast, the tristichopterid osteolepiform Eustenopteron —long regarded as a transitional form between the basal lobe-fins and early tetrapods— has two sets of radials articulating with the ulnare, but none with the other bones.

The pattern of distal bones also varies within the elpistostegelians, a paraphyletic group of lobe-fins that are transitional between tristichopterid osteolepiforms such as Eustenopteron and the early tetrapods. Panderichthys has a unique pattern of two distal elements that apparently correspond to the ulnare and intermedium, but these elements may have fused with their associated radials; independent radials are lacking as are their corresponding articular surfaces. In contrast, the pattern in Tiktaalik is more comparable with that found in tristocopterids and tetrapods. Four sets of radials directly articulate with the ulnare, one with the intermedium and none with the radius. The humerus in both Tiktaalik and Panderichthys resemble those of the early tetrapods in that they're both dorsoventrally flattened. On the other hand, the radius in both is substantially longer than the ulna, a condition shared with both osteolepiforms and rhizodonts.

evolutionary trends in pectoral limbs in lobefins and early tetrapods Pectoral fins from two Late Devonian elpistostegelian lobefins (Panderichthys and Tiktaalik), and pectoral limbs from two Late Devonian tetrapod (Acanthostega and Tulerpeton) and two Carboniferous tetrapods (Proterogyrinus and Limnoscelis); radial torsion is not depicted in Tulerpeton and the Carboniferous tetrapods. ©.

The limbs of tetrapods can be distinguished from the fins of lobe-fins by the complete loss of lepidotrichia and by the aquisition of digits. These distinctive elements differ from the radials of lobe-fins in that they're spool-shaped rather than cylindrical or flattened, and they don't bifurcate (split distally). All Late Devonian tetrapods for which there are preserved digits (Acanthostega, Ichthyostega, and Tulerpeton) exhibit polydactyly, a condition in which there are more than five digits on each manus (hand) or pes (foot). Sometime during the Late Devonian or Tournaisian (earliest Carboniferous) at least one group emerged that contained only the five digits that represent the ancestral condition for all living tetrapods.

The appearance of digits is one of the more perplexing aspects of early tetrapod evolution. Until recently, there was little in the way of transitional structures to indicate how these elements emerged; the disparity in the form and —more importantly— the pattern of radials in tristocopterids (e.g., Eusthenopteron) and those of digits in Late Devonian tetrapods (e.g., Acanthostega). In particular, there wasn't a central axis of endochondral bones from which the digits would arise. This has prompted some scientists to propose that the manus or pes were acquired de novo (i.e., they did not arise from pre-existing structures such as radials).

One widely considered proposal is that the digits resulted from a develomentally active zone that extended from the posterior margin of the fin (limb) and swung forward at the tip where it effectively bypassed any existing radials and give rise to the digits. However, the recent discovery of Tiktaalik suggest that digits were not aquired de novo. It exhibits a fin axis that extends beyond the ulna (3rd radial), a feature not found in any previously known tristochopterid or elpistostegelian lobe-fin.

In either case, development along the posterior half of the fin appears to be central to the emergence of the tetrapod limb. The ulna is conspiciously shorter than the radius in tristocopterid, rhizodont and elpistostegelian lobe-fins. This condition is also found in Acanthostega, but the two bones are nearly equal in length for two other Late Devonian tetrapods for which theses bones are known, Ichthyostega and Turlerpeton, as they are in all known tetrapods from the Carboniferous (i.e., crown tetrapods).

Associated with this extension of the ulna is the incoporation of two more distal elements (the ulnae and intermedium) into the wrist. This "joint" is actually a traverse plane involving several smaller bones that flex as a unit relative to the radius and ulna. The wrist is an important adaptation for terrestrial locomotion because it alows the manus (hand) to operate as a unit to both establish stable contact with the substrate (ground or bottom) and to more widely distribute weight from the supported body.

The wrist of crown tetrapods exhibits a number of derived features, including distal carpals (distal wrist bones that articulate with the digits) and the cross-articulation of individual wrist bones (they articulate laterally with other wrist bones as well as with proximal and distal elements). The wrists is usually poorly preserved in fossils because the small bones are often scattered and hard to identify, or are lost altogether. However, what is known of the wrist from two Late Devonian tetrapods suggest transitional structures. Except for an undifferentiated intermedium, the wrist elements in Acanthostega were either unossified or absent altogether. In contrast, Tulerpeton had simple, but apparently functional, wrist that lacked carpals and cross-articulation.

This limited evidence from Late Devonian tetrapods suggest that the wrist developed after the arm and digits. Strikingly, a functional analog of the wrist may have evolved in Tiktaalik, a "fishapod" elpistegelian lobe-fin that predated these tetrapods by several million years. It had was appears to be a series of transverse joint planes among the radials of its front fin. The bones of its "wrist" are not homologous with those in tetrapods, but the functional similiarity suggest some of the evolutionary pressures that led to the appearance of tetrapods. adaptations.

The newly evolved limbs would have been able to prop-up the body in shallow water or on land, but they would have less efficient in open water than fins. However, they were probably more effective in navigating shallow water filled with aquatic plants and other debris.

The universal tetrapod condition of front limbs that bend backward at the elbow and hind limbs that bend forward at the knee can plausibly be traced to early tetrapods living in shallow water. In addition to paddling, the front limbs could have been used to prop up the front half of the body. Body mass, which is normally supported by the buoyancy of water, can become crushing weight. Propping up the upper trunk would prevent the constriction of lungs and damage to other internal organs. Propping up the head would facilitate breathing in somewhat deeper water. Propping isn’t as important in hind quarters so the hind limbs could be reserved for providing propulsion (i.e., pushing back).

Gills and Lungs

One of the most conspicuous adaptations of early tetrapods is air breathing. Breathing is an obvious necessity for terrestrial life, but it would also be favorable for any large animal trying to exploit shallow stagnant waters.

Unfortunately, the origin and evolution of lungs is unclear. A classical theory proposed by Alfred Romer stated that osteichthyans (bony fishes) first evolved lungs and then developed gills, but many authorities believe that it was the other way around. What does seem clear, however, is that the diversion of the esophagus that formed either swim bladders or lungs had evolved in early osteichthyans. The gas-filled swim bladder of ray-fins enabled them to achieve neutral buoyancy and was probably crucial to their extraordinary evolutionary success. The fat-filled swim bladder in Latimeria (the coelacanth), and quite possibly some of the extinct lobe-fins, apparently also provides buoyancy.

Lungs may have evolved independently within two lobe-fin lineages: lungfishes and the clade that includes osteolepiforms and tetrapods. Most earlier, marine lungfishes probably didn’t have functioning lungs, but some Late Devonian and Carboniferous species apparently evolved an air-gulping mechanism that allowed at least some air breathing. Gill respiration is substantially reduced in later lungfishes and some modern species will drown unless they can gulp air.

The first appearance of functional lungs in the osteolepiform/tetrapod lineage is unknown. It’s clear that functional internal gills were present in virtually all osteolepiforms. They were also present in at least one Late Devonian tetrapod, Acanthostega. On the other hand, it has been suggested that at least one other Late Devonian tetrapod, Hynerpeton, did not have internal gills. Anatomical evidence of internal gills in other Late Devonian tetrapods is either absent or inconclusive.

The Choana

A tantalizing line of evidence associated with air breathing involves the choana. This structure connects the external nostril to the buccal cavity (the inside of the mouth). It also has an opening that leads to the orbits (eye openings) via a small duct. (The choana is apparently homologous with the primitive paired external nostrils that are still found in lampreys, sharks, Latimeria and ray-fins. These external nostrils are used for smelling in water and not for respiration.) The choana is found only in osteolepiform lobe-fins and tetrapods; their presence or absence in rhizodont lobe-fins is unresolved. Interestingly, a similar structure is also found in the more derived lungfishes. However, it lacks the duct that leads to the obits.

The choana is the primary external passage through which modern tetrapods breathe. Its function in early tetrapods and osteolopid lobe-fins is less clear. It may have been used for air breathing, even in fully aquatic forms such as Eusthenopteron and Osteolepis. On the other hand, the choana may have served another function in gill-breathing fish. The duct leading to the orbits is comparable to the lacrimal duct that is found in modern tetrapods; secretions from the lacrimal gland bathe the eyes and nasal passages to prevent damage via exposure to the air. Similar secretions may have enabled certain lobe-fin fishes to operate near or at the water’s surface. Breathing via the choana may have evolved later.

Changes in the Skull

Most lobe-fins exhibit a cranial kinesis (skull mobility) that enhances bite strength. This facility is provided via articulations within a jointed braincase and its associated dermal skeletal elements. On the other hand, the skulls of Elpistostegelians (e.g., Elpistostege, Panderichthys, and Tiktaalik) and early tetrapods have a progressive loss of intracranial mobility. The dermal skull roof becomes solidified in both groups, while the braincase in tetrapods becomes solidified and more closely integrated with the skull roof.

The overall proportions of the skull also changed. Those of the elpistostegelians and tetrapods became flatter and broader. In tetrapods, the part of the skull anterior to the orbits lengthened while the area behind the orbits shortened. In addition, the dermal bones that cover the gills in lobe-fins (e.g., the opercular) are apparently lost in at least one elpistostegelian (Tiktaalik) and in the early tetrapods. Consequently, the pectoral girdle and forelimbs are freed from the skull and the animal has a neck.

skullls of Eustenopteron, Panderichthys and AcanthostegaTop and side views of the skulls of Eusthenopteron (Frasnian), Panderichthys (Givetian-Frasnian) and Acanthostega (Famennian) ©.

Some explanations have been offered for the changes that occurred in these skulls. For example, a flattened skull would be useful in shallow water, and a solidified skull would been required to support a head not otherwise supported by buoyant water. However, the changes may be better attributed to enhanced air breathing. A broadened and flattened skull, combined with the elongation of the anterior skull, would generate a more efficient buccal pump. This in turn would greatly improve the ability of the tetrapod to ventilate its lungs. Moreover, a wider and flatter skull would have substantially reduced the utility of cranial kinesis and thereby reduce selective pressure for this kind of skull mobility.

Whatever the reasons for changes in skull shape and structure, the increase in air breathing efficiency would have led to the reduction or modification of the internal gill apparatus. One of these bones, the hyomandibular, became the stapes, a bone that ultimately became central to terrestrial hearing. The freeing of the pectoral girdle from the rear of the skull may have also been a consequence of the shift from gill breathing. However, it may be plausibly attributed to the benefit of greater axial flexibility in weed-filled shallows.

Hips and Shoulders

The freeing of the shoulder or pectoral girdle from the head in early tetrapods is associated with a number of changes in the structure of the pectoral girdle itself. There’s a general trend in the reduction of dermal bones and the enlargement of endoskeletal elements; a similar trend in dermal and endoskeletal elements is also evident in the limbs. In particular, the endoskeletal scapulocoracoid makes up the bulk of the tetrapod pectoral girdle. The glenoid fossa (the surface of the scapulacoracoid that articulates with the humerus) shifts from the posterior orientation characteristic of lobe-fins to a more lateral orientation and generally become larger.

Significant changes also occurred in the pelvic girdle. The pelvis of lobe-fin fishes consist of two isolated sets of bones that don’t come into contact with the either the spine or each other. In contrast, the two halves of tetrapod pelvic girdles join ventrally to form a pubic symphysis and connect to the spine via sacral ribs. As in the pectoral girdle, the endoskeleton expands to provide additional support and surface area for muscle attachment. The acetabulum (the pelvic counterpart to the glenoid fossa) is also enlarged and more laterally oriented.

The pectoral and pelvic fins of lobe-fins such as Eusthenopteron were used primarily for braking and steering. The transition from these fins to paddles required increased structural support, increased surface area for muscle attachments and a more lateral orientation of the limb articulations (glenoid and acetabulum). These modifications provided preadaptations for propping-up the head and ultimately to walking.


We don’t know whether any of the Late Devonian tetrapods spent a considerable amount of time out of the water. Ichthyostega has several features that suggest at least some terrestriality. Its vertebrae have zygapophyses (bony projections that brace against adjoining vertebrae) that could help support the spine in the absence of buoyant water. It also had massive, overlapping ribs that could have helped to prevent the crushing of lungs and other internal organs. In addition, the structure of the forelimbs suggests that they could have borne weight. However, Ichthyostega could not have traversed the land in a fashion even vaguely resembling the typical quadrapedal tetrapod gait because its broad, overlapping ribs would have inhibited side-to-side movement. Instead, it may have dragged itself along with its powerful forelimbs, or, as recently suggested by Per Ahlberg and Jennifer Clack, it may have flexed its spine dorsal-ventrally in a manner comparable to a modern elephant seal. In any case, Ichthyostega —once regarded as the ideal link between fish and later tetrapods— probably represents an evolutionary blind alley, an early experiment in tetrapod terrestriality.

Two other lines of evidence suggest some degree of terrestriality in other Late Devonian tetrapods. The shoulder of Hynerpeton contains anatomical features (i.e., muscle attachment scars and its articulation with the humerus) that indicate considerable load-bearing capability and a range of appendicular movement. Unfortunately, little else is known about this animal. Tetrapod trackways also suggest terrestriality. However, uncertainties regarding their age and depositional environment (created in shallow water or on land) make the interpretation of these trackways problematic.

While it’s unclear how far Late Devonian tetrapods developed the terrestrial habit, their diversity and geographic range indicate that they were very successful in shallow aquatic habitats. Indeed, a diverse assortment of aquatic tetrapods (e.g., Greerepeton, Crassigyrinus, and Diploceraspis) characterize vertebrate faunas throughout of the Late Paleozoic. The earliest confirmed terrestrial form, Pederpes, is known from the late Tournaisian (Early Carboniferous), some 15 million years after the Late Devonian tetrapods occurred, while a more diverse terrestrial fauna appears some five million years later in the Viséan. These forms exhibited a number of distinctive adaptations for life on land. One of these was the development of a bony spine characterized by interlocking vertebrae with well-developed centra rather than a largely unrestricted notochord. Another adaptation is the ability to flex wrists and ankles. The Carboniferous tetrapods are also characterized by pentadactly (five-digit hands and feet). Whether this reduction in digits is a direct result of the transition from a paddle operating in water (e.g., in Acanthostega) to a foot walking on land is unknown.

Top of Page .

G.R. Morton's Transitional Forms: Fish to Amphibian:
Clack, J.A. 2002. Gaining Ground: The Origin and Early Evolution of Tetrapods. Bloomington: Indiana Univ. Press.
Janvier, P. 1996. Early Vertebrates. Oxford: Claredon Press.
Long, J.A. 1995. The Rise of Fishes: 500 Million Years of Evolution. Baltimore and London: John Hopkins Univ. Press.
Maisey, J.G. 1996. Discovering Fossil Fishes. New York: Henry Holt & Co.
Zimmer, C.. 1998. At the Waters' Edge. New York: Touchstone Press.
Discover Magazine. June 1995: "Coming Onto the Land."
National Geographic. May 1999. "From Fins to Feet."
New Scientist. 19 August 2000. "One Small Step for Fish, One Giant Leap for Us."
Scientific Papers:
Ahlberg, P.E. 1989. "Paired fin skeletons and relationships of the fossil group Porolepidormes (Osteichthyes: Sarcopterygii)." Zoological Journal of the Linnean Society. 96: 119-166.
Ahlberg, P.E., J.A. Clack, and H. Blom. 2005. "The axial skeleton of the Devonian trtrapod Ichthyostega." Nature 437(1): 137-140.
Carroll, R.L., J. Irwin, and D.M.Green. 2005. "Thermal physiology and the origin of terrestriality in vertebrates." Zoological Journal of the Linnean Society 148: 345-358.
Clack, J.A. 1994. "Acanthostega gunnari, a Devonian tetrapod from Greenland; the snout, palate and ventral parts of the braincase, with a discussion of their significance." Meddelelser om Grønland: Geoscience, 31: 1-24.
Clack, J.A. 1997. "Devonian tetrapod trackways and trackmakers: a review of the fossils and footprints". Paleogeography, Paleoclimatology, Paleoecology 130: 227-250.
Coates, M.I., 1994. "The origin of vertebrate limbs." Development 1994. Supplement, 169-180, p. 174.
Coates, M.I. 1996. "The Devonian tetrapod Acanthostega gunnari Jarvik: postcranial anatomy, basal tetrapod interrelationships and patterns of skeletal evolution." Trans. Royal Soc. Edinburgh; Earth Sciences 87:363-421.
Coates, M.I., J. E. Jeffery and M.Ruta. 2002. "Fins to limbs: what the fossils say." Evolution and Development 4:5, 390-401
Daeschler, E.B. and N. Shubin. 1995. "Tetrapod origins." Paleobiology 21(4) 404-409.
Daeschler, E.B. and N. Shubin. 1998. "Fish with fingers?" Nature 391:133.
Holmes, R. "Proterogyrinus scheelei and the early evolution of the labryrinthodont pectoral limb." In: A.L. Panchen (ed.). The Terrestrial Environment and the Origin of Land Vertebrates. London: Academic Press.
Jarvik, E., 1980. Basic structure and evolution of vertebrates. Vol 1. London: Academic Press.
Lebedev, O.A. and M.I. Coates. 1995. "The postranial skeleton of the Devonian tetrapod Tulerpeton curtum Lebedev." Zoological Journal of the Linnean Society 114: 307-348.
Long, J.A. and M.S. Gordon. 2004. "The greatest step in vertebrate history: a palaeobiological revies of the fish-tetrapod transistion" Physiological and Biochemical Zoology 77(5): 700-719.
Moy-Thomas, J. A. and R. S. Miles, 1971. Paleozoic Fishes. London: Chapman and Hall.
Rackoff, J.S. 1980. "The origin of the tetrapod limb and the ancestry of the vertebrates." In: A.L. Panchen (ed.). The Terrestrial Environment and the Origin of Land Vertebrates. London: Academic Press.
Shubin, N.H., E.B. Daeschler, and M.I. Coates. 2004. "The Early Evolution of the Tetrapod Humerus." Science 304: 90-93.
Shubin, N.H., E.B. Draeschler, and F.A. Jenkins. Jr. 2006. "The pectoral fin of Tiktaalik roseae and the origin of the tetrapod limb. Nature 440(6): 764-771.
Thomson, K.S. 1980. "The ecology of Devonian lobe-finned fishes." In: A.L. Panchen (ed.). The Terrestrial Environment and the Origin of Land Vertebrates. London: Academic Press.
Thomson, K.S. 1995. "The origins of the tetrapods." Amer. J. Science 293-A: 33-62.
Vorobyeva, E.I. 1977. "Morphology and the features of the evolution of crossopterygian fish." Trudy Paeont. Inst. 163: 1-240. (in Russian).
Williston, S.W. 1925. In: W.K. Gregory (ed.). The Osteology of the Reptiles. Cambrige: Harvard Univ. Press.
Image Credits:
The two images above are copyrighted © 2006, Dennis C. Murphy, (see Terms of Use). The fin and limb reconstructions are based on Ahlberg (1989), Coates (1994), Coates, Jeffery and Ruta (2002), Daeschler & Shubin (1998), Holmes (1980), Jarvik (1980), Lebedev and Coates (1995), Shubin & Daeschler (2006), and Williston (1925). The skull reconstuctions are based on Clack (1994), Coates (1996), Jarvik (1980), Moy-Thomas & Miles (1971), Shubin et. al. (2006), and Vorobyeva (1977).

Top of Page .