GAIA Nº 15, LISBOA/LISBON, DEZEMBRO/DECEMBER 1998, pp. 399-403 (ISSN: 0871-5424)
THE EVOLUTION OF FEATHERS FROM
DINOSAUR HAIR
Peter J. GRIFFITHS
University of Wolverhampton, School of Health Sciences. Lichfield Street, Wolverhampton WV1 1DJ. UNITED KINGDOM
ABSTRACT: The significance of finding feathered theropod dinosaurs is discussed in terms
of the theories of the evolution of feathers, birds and endothermy. The plumulous proximal
portion of the isolated Archaeopteryx feather suggests that endothermy had already
evolved in the dinosaur - bird lineage by the Late Jurassic. It is suggested that adipose tis-
sue would have played a major role in providing insulation in endothermic ancestral thero-
pod dinosaurs. Small endothermic dinosaurs may have found an additional form of
insulation an advantage. This may initially have evolved as"hair" in the dinosaurs, rather
than the more morphologically complex branched feather. As avian epidermal appendages
are composed of the unique f keratin family, it is unlikely that feathers were derived directly
from archosaurian scales, but must have involved intermediate structures. It is suggested
that feathers could consequently have been derived from dinosaur "hair".
RÉSUMÉ: La découverte significative de dinosaures théropodes à plumes a entraîné une dis-
cussion en fonction des théories sur l'évolution des plumes, des oiseaux et de l'endother-
mie. La partie proximale duveteuse de la plume isolée de l' Archaeopteryx suggère une
évolution de l'endothermie dans la lignée dinosaure-oiseau à la fin du Jurassique. On pense
que le tissu adipeux aurait joué un rôle majeur en fournissant de la chaleur chez les dinosau-
res théropodes endothermes. Ainsi, les petits auraient pu en trouver un avantage supplé-
mentaire. Chez les dinosaures, il y aurait d'abord pu avoir une évolution vers des "poils"
plutôt que des plumes ramifiées, morphologiquement très compliquées. Etant donné que
les tissus épidermaux des animaux à plumes sont composés de l'unique famille kératine f, il
est improbable que les plumes aient directement dérivé des écailles archosauriennes, mais
ont dû faire intervenir des structures intermédiaires. Par conséquent, les plumes auraient
pu provenir des "poils" de dinosaures.
INTRODUCTION supporting evidence for the hypothesis that feathers
evolved initially for the purpose of insulation or
The recent discovery of several species of
display, becoming secondarily adapted for flight. In
feathered theropod dinosaur from the Liaoning
addition, the discovery of feathered non-avian
Province of China, locality of the early birds Sinornis
dinosaurs gives support to the hypothesis that birds
(SERENO & RAO, 1992) and Confuciusornis (HOU et
evolved from the theropods, and that dinosaurs
al. 1995), has fuelled the debate regarding the origin
ancestral to birds were endothermic.
and initial function of feathers, and also speculation
concerning the possibility that dinosaurs were
THE THEROPOD ANCESTRY OF BIRDS
endothermic. Sinosauropteryx prima (JI & JI, 1996),
a small compsognathid dinosaur is described as It is generally accepted by most palaeontologists
having epidermal appendages resembling the that birds evolved from the theropod dinosaurs
plumules of modern birds (CHEN, DONG & ZHEN, (OSTROM, 1973), and cladistic analyses suggests
1998). Protoarchaeopteryx robusta (JI & JI, 1997) that among the theropod dinosaurs, the dro-
and Caudipteryx zoui (JI et al., 1998) are both maeosaurs share the most characters with the earli-
described as having pennaceous feathers attached est known bird Archaeopteryx, and also with modern
to the fore limbs and tail, while Caudipteryx is birds (GAUTHIER, 1986; HOLTZ, 1994).
reported as also having plumulaceous feathers
The hypothesis that birds evolved from the thero-
around the body (JI et al., 1998). The discovery of
pod dinosaurs has recently received further support
feathered non-avian theropod dinosaurs provides
399 artigos/papers
P. J. GRIFFITHS
with the discovery of a new species of fossil bird morphologically very similar to those of modern
from the Upper Cretaceous of Madagascar, Rahona birds.
ostromi (FORSTER et al., 1998). Rahona has bony
The earliest known fossil feather is that of Ar-
protrusions on the forearms which in modern birds
chaeopteryx lithographica (MEYER, 1861). A recent
serve as the attachment points for flight feathers,
analysis shows the feather to be morphologically
suggesting not only that it was feathered but also
identical to the tenth secondary flight feather of the
that it was capable of flight. Most interestingly, Ra-
magpie which has a wing shape (and presumably
hona has a large, retractable sickle-shaped claw on
aerodynamic characteristics) similar to that of Ar-
the second toe of the hind foot, similar to the slashing
chaeopteryx (GRIFFITHS, 1996). This is the only de-
claws of Velociraptor (OSBORN, 1924) and Deinony-
scribed individual fossil feather attributable to
chus (OSTROM, 1969).
Archaeopteryx, and so it is not known if the species
It has been suggested that the second toe of possessed feathers specialised specifically for insu-
Archaeopteryx may have been hyperextensible lation.
(PAUL, 1988; SERENO, 1997). In addition, although
However, an examination of the morphology of
the pedal claws are considerably smaller than those
the isolated Archaeopteryx feather reveals some in-
of the wing (G RIFFITHS , 1994), the claw of the
teresting features. On distal portions of the feather,
second toe is significantly larger than those of the
individual barbs and even barbules can clearly be
other toes, and has a different morphology. For
observed forming the usual pennaceous structure
example, in the Eichstätt specimen WELLNHOFFER
which allows the barbs to zip together creating the
(1974) reports that the claw of the second toe has a
typical flight-feather vane capable of generating
curvature with a greater angle from the vertical
aerodynamic lift. In comparison, the morphology of
chord across the articular facet to claw tip (160o)
the feather is quite different proximally where indi-
compared to the other toes (132 o -146 o ), which
vidual barbs cannot be distinguished. In this region,
supports the suggestion that the second toe of
the feather has a tufted appearance suggesting that
Archaeopteryx may have been hyperextensible.
the hooklets attached to the barbules are fewer in
While a slashing claw is found in Rahona, such a
number or are absent, and that the barbs are much
claw is not present in Iberomesornis (SANZ et al.,
thinner than in the distal parts of the feather. Conse-
1988) or Concornis (SANZ & BUSCALIONI, 1992) from
quently the isolated Archaeopteryx feather appears
the Lower Cretaceous of Spain, or Confuciusornis
to have a plumulaceous region at the base which
(HOU et al., 1995) and Sinornis (SERENO & RAO,
would efficiently trap air against the skin. This fea-
1992) from China. This would suggest that the
ture is identical to that found in the flight feathers of
slashing claw of Rahona may be a retained
modern birds which have few if any down feathers
plesiomorphic character.
on the wing. The plumulaceous base of the Archae-
opteryx feather indicates that it was capable of pro-
EVOLUTION OF ENDOTHERMY AND DOWN
viding insulation and so supports the hypothesis that
FEATHERS
Archaeopteryx was endothermic (GRIFFITHS, 1996).
The discovery of apparently flightless, feathered
As the Archaeopteryx feather is already morpho-
non-avian theropod dinosaurs provides strong
logically specialised and exhibits characteristics
supporting evidence for the hypothesis that feathers
enabling the dual functions both of flight and insula-
evolved initially for the purpose of display or insula-
tion, Archaeopteryx can give no indication of the ini-
tion, becoming secondarily adapted for flight during
tial role of feathers, but does make clear that
the subsequent evolution of birds (REGAL, 1975). An
feathers capable of providing insulation as well as
alternative hypothesis for the origin of feathers (FE-
flight were already present during the Late Jurassic.
DUCCIA, 1974) is that they evolved initially for flight,
It is logical to assume that endothermy and feathers
and subsequently become adapted for insulation
evolved in taxa ancestral to Archaeopteryx.
when the birds became endothermic, although this
now seems less likely in view of the recent finds.
THE ROLE OF ADIPOSE TISSUE FOR
Modern birds are unquestionably endothermic INSULATION
and have down feathers which are specialised for in-
In contrast to birds, mammals use two different
sulation as well as flight feathers. At some point
mechanisms for insulation, hair and also a layer of
along the dinosaur - bird lineage, both endothermy
adipose tissue under the dermis. Hair plays a similar
and feathers for insulation evolved, although not
role in insulation as down feathers, trapping a layer
necessarily at the same time. The first discovered
of air against the skin thus reducing heat loss by con-
fossil down feather has been classified as Ilerdop-
vection and radiation. In adipose tissue, lipid drop-
teryx viai (LACASA, 1985) from the Lower Creta-
lets are stored in large specialised cells which are
ceous of El Montsec, Spain. This location has
embedded in connective tissue, particularly in the
yielded a variety of specialised feathers which are
400
THE EVOLUTION OF FEATHERS FROM DINOSAUR HAIR
dermis. The dermal layer of adipose tissue prevents Cretaceous birds Iberomesornis, Concornis and
the loss of heat from deep within the body, helping to also Sinornis were much smaller than Archaeop-
maintain core temperature. teryx. Bird evolution therefore appears to have in-
volved an initial reduction in size, probably because
The potential role of adipose tissue for insulation
it is easier for small birds to generate aerodynamic
must be taken into account when considering the
lift (SANZ & BONAPARTE, 1992), although Confuciu-
evolution of endothermy in birds, particularly with re-
sornis appears to be an exception as they were of a
gard to the ancestral dinosaurs and the debate as to
similar size to Archaeopteryx.
whether they were endothermic. Once there is a ne-
cessity for regulation of the internal temperature to
THE EVOLUTION OF FEATHERS
within small tolerances, then insulation becomes an
immediate advantage due to the high energetic ex- MADERSON (1972) has suggested that proto-
pense of the metabolic generation of heat. This feathers may have evolved from the tips of archo-
would be particularly important for smaller species saurian scales with the scale eventually regressing
with a high surface area to volume ratio where heat to leave the feather, on the basis that both feathers
loss would be more significant. There would there- and the epidermis of scales in birds were thought to
fore be a very high selection pressure, particularly contain b keratin, while the interfollicular skin adja-
for small species of dinosaur living in a colder cli- cent to feathers and the inner surface of scales con-
mate, to evolve some form of insulation. This could tained a keratin. This hypothesis received support
most simply be achieved with a layer of adipose tis- from the work of DHOUAILLY , HARDY & SENGEL
sue under the skin. Adipose tissue is widely distrib- (1980), who showed that it was possible to convert
uted among modern sauropsids, and in modern the prospective scales on the feet of chick embryos
birds is used not only for energy storage but also in- to feathers with the use of Retinioc Acid, known to
sulation, particularly in species which come into fre- play an important role in positional signalling and
quent contact with water. pattern formation during cell differentiation in devel-
A thick layer of adipose tissue, while it may pro- oping embryos.
vide excellent insulation, is however relatively However, BRUSH (1993) has shown that the
heavy. In terms of insulation efficiency to weight, the avian keratins in feathers, down, scutes and beaks
ratio is low (GRIFFITHS, 1996). This is not normally a are quite different from the keratin of reptiles, and is
problem, but becomes a serious disadvantage composed of a family of f keratins. These are more
where weight is of consideration such as when at- closely related to a keratins than b keratins. There is
tempting to fly.
little evidence that f keratin and a keratin diverged
Conversely, down feathers and hair have a high from a common ancestral gene, or that f keratin was
insulation efficiency to weight ratio and therefore derived from existing a keratin genes, although f
have the advantage over adipose tissue where keratin could perhaps be derived from a cytokeratin
weight is of consideration. The early mammals are protein. f keratin forms b -pleated sheets which
thought to have been small shrew-like creatures spontaneously assemble into filaments by self asso-
with a large surface area to volume ratio, thus heat ciation, rather than the a helixes which a keratin can
loss from the surface would have been a major prob- form. The f keratins exist as two main groups: a
lem. Hair would have been a big advantage in pro- larger molecule of 13,500 Da found in scales, claws
viding extra insulation and reducing heat loss, and beaks, and a smaller molecule of 10,500 Da
particularly if the early mammals were mainly noc- found in feathers and down. The amino acid se-
turnal and active when ambient temperatures were
quences of the f keratin of feathers and down is
lower. Mammals seem to have increased in size
more similar to the sequence of amino acids in
later in their evolutionary history, and large mam-
scutes than in reticulate scales, and closer still to
mals frequently have reduced hair cover, particu-
beak and then claw keratins (BRUSH, 1980).
larly those living in warmer climates, as seen in
modern elephants, although large mammals such As Archaeopteryx has feathers which are essen-
as mammoths retained a thick hair covering when tially modern in structure, it could be assumed that
faced with adverse climatic conditions. they were also composed of f keratin. In addition to
The situation appears to be the reverse for birds feathers, the keratin sheaths of the claws have been
which probably originated from larger theropod an- preserved in Archaeopteryx. There is no sign of a
cestors and initially became progressively smaller beak in any of the specimens although it is unlikely
as flight evolved. Archaeopteryx is smaller than that Archaeopteryx possessed a beak as teeth were
most known theropods, with the exception of Comp- still present. Also, there are no indications of scutes
sognathus (WAGNER,1861) which was an unusually or reticulate scales in any of the specimens.
small theropod. Similarly, the thrush sized Lower
401
P. J. GRIFFITHS
This introduces additional complexity into any structures are described as consisting of a primary
hypothesis concerning the origin of feathers as it is system of subparallel fibres arranged perpendicular
necessary to postulate the evolution of a new family to the bone surface, with a less conspicuous secon-
of genes coding for these new proteins, which must dary system orientated in parallel. Again, it is not
have been present in epidermal appendages prior to clear if these are branched structures rather than
the evolution of feathers. The molecular structure of single filaments.
the f keratins would suggest that it is unlikely that Branching filamentous structures which can only
feathers evolved directly from archosaurian scales. be described as symmetrical feathers have been
As feather keratins are closer to beak and claw found associated with the two new species from the
keratins than scute or scale keratins, it again sug- Liaoning province of China, Protoarchaeopteryx
gests that a number of intermediate stages were in- and Caudipteryx (J I et al., 1998). While the
volved in feather evolution (BRUSH, 1996). phylogenetic analysis places Caudipteryx as a
sister group to the Avialae, the systematic position of
DINOSAUR HAIR Protoarchaeopteryx is less clear and appears to be
Av i a n f e a t h e r s c a n b e m o r p h o l o g i c a l l y unresolved from the Velociraptorinae (GAUTHIER,
distinguished from mammalian hair on the basis that 1986; HOLTZ, 1994) root, and may also be a sister
they have a complex branching structure, while group to the Avialae (JI et al., 1998). Although
mammalian hair consists of single filaments. From a Protoarchaeopteryx has relatively long arms
developmental perspective, hair is more simple to compared to non-avian coelurosaurs they are
form than the structurally complex branching shorter than Archaeopteryx. The arms of
feather, and consequently mammalian hair follicles Caudipteryx are shorter than non-avian
are morphologically simpler than avian feather coelurosaurs. In either species it appears to be
follicles. If the evolution of epidermal appendages unlikely that they were able to fly. However, there is a
were driven by the necessity of providing insulation distinct possibility that they were the flightless
rather than providing an aerodynamically functional descendants of birds, and inherited feathers from
vane, it is parsimonious to consider the initial flying ancestors. This hypothesis is supported by the
evolution of an unbranched structure to be more observation that the some of the feathers of
probable than a complex branching one. Protoarchaeopteryx and perhaps Caudipteryx were
Unbranched single filaments (hair) are therefore pennaceous. PAUL (1988) has even suggested that
more likely than branched feathers to have initially all the dromaeosaurs are descended from flying
evolved as a structure required for the purposes of protobirds. This is not impossible as they appear to
insulation. have evolved after Archaeopteryx, and even
Protoarchaeopteryx, Caudipteryx and
If the dinosaurs were to be considered endother- Sinosauropteryx are found in deposits that probably
mic, hair would have the advantage of conferring ad- date from the early Cretaceous (SMITH, EVENSEN &
ditional insulation to that provided by a layer of YORK, 1996).
adipose tissue. This may have been important to
smaller species of dinosaur with a high surface area FEATHERS FROM DINOSAUR HAIR
to volume ratio which were challenged with the prob-
lem of heat loss that larger dinosaur species would Feathers could conceivably have evolved from
not experience. dinosaur "hair", perhaps similar in morphology to the
structures reported associated with Pelicanimimus.
An analysis of the epidermal appendages of Si- Such appendages may well have been composed of
nosauropteryx (CHEN, DONG & ZHEN, 1998), de-
filaments of f keratin, rather than a keratin as in
scribes the structures as being possibly hollow,
mammalian hair. Subsequently, the epidermal ap-
rather course structures that may resemble the plu-
pendages became split to form the branching struc-
mules of modern birds, with short quills and long fila-
ture of barbs. Such a branching structure could
mentous barbs, but with no signs of barbules or
possibly be represented by the appendages re-
hooklets. However, as it has not yet been possible to
ported associated with Sinosauropteryx, with short
isolate a single structure for analysis, it is not clear if
quills and long filamentous barbs. The branching
these structures really are branched rather than be-
then become more complex to form barbules. Hooks
ing composed of single filaments, or if they even
subsequently evolved allowing the barbules to zip
have any real relationship with modern feathers.
together to form the pennaceous structure of mod-
Preserved impressions of possible integumen- ern feathers, as exhibited by Archaeopteryx and
tary structures have also been described associated modern birds. This hypothesis therefore proposes a
with the non-avian theropod Pelecanimimus polyo- series of progressively more complex structures,
den (PÉREZ-MORENO et al., 1994) from the Lower each of which is functionally advantageous in terms
Cretaceous location of Las Hoyas, Spain. These of initially providing insulation, and then finally lead-
402
THE EVOLUTION OF FEATHERS FROM DINOSAUR HAIR
ing to a feather capable of generating aerodynamic PAUL, G.S (1988) - Predatory Dinosaurs of the World. Simon &
Schuster, New York, 403 pp.
lift. A change in the morphology of the epidermal ap-
PÉREZ-MORENO, B.P.; SANZ, J.L.; BUSCALIONI, A.D.; MORATALLA,
pendage from hair to feather would involve an in- J.J.; ORTEGA, F. & RASSKIN-GUTMAN, D. (1994) - A Unique
crease in complexity of the morphology of the Ornithomimosaur Dinosaur from the Lower Cretaceous of
appendage follicle, and so would be caused by epi- Spain. Nature, 370: 363-367.
genetic factors. REGAL, P. (1975) - The Evolutionary Origin of Feathers. Quart.
Rev. Biol., 50(1): 35-66.
BIBLIOGRAPHY SANZ, J.L.; BONAPARTE, J.F. & LACASA, A. (1988) - Unusual Early
Cretaceous Birds from Spain. Nature, 331(6155): 433-435.
BRUSH, A.H. (1980) - Patterns In The Amino Acid Composition Of
Avian Epidermal Proteins. Auk, 97: 742-753. SANZ, J.L. & BONAPARTE, J.F. (1992) - A New Order of Birds
(Class Aves) from the Lower Cretaceous of Spain, in
BRUSH, A.H. (1993) - The Origin of Feathers: A Novel Approach. CAMPBELL, K.E. (Ed.), Papers in Avian Palaentology Ho-
Avian Biol., 9: 121-162.
noring, Pierre Brodkorp, Los Angeles County Museum Sci.
BRUSH, A.H. (1996) - On The Origin of Feathers. J. Evolut. Biol., 9: Series, 36: 39-49.
131-142. SANZ, J.L. & BUSCALIONI, A. (1992) - A New Bird From the Lower
CHEN, P.-J.; DONG, Z.-M. & ZHEN, S.-N. (1998) - An Exceptionally Cretaceous of Las Hoyas, Spain and the Early Radiation of
Well-preserved Theropod Dinosaur from the Yixian Birds. Palaeontology, 35(4): 829-845.
Formation of China. Nature, 391: 147-152. SERENO, P.C. & RAO, C. (1992) - Early Evolution of Avian Flight
DHOUAILLY, D.; HARDY, M.H. & SENGEL, P. (1980) - Formation of and Perching: New Evidence from the Lower Cretaceous of
Feathers on Chick Foot Scales: A State-Dependant Morpho- China. Science, 255: 845-848.
genetic Response to Retinoic Acid. J. Embryol. Experimental SERENO, P.C. (1997) - The origin and evolution of Dinosaurs. Ann.
Morphol., 58: 63-78.
Rev. Earth Planet. Sci. 1997, 25: 435-489.
FEDUCCIA, A. (1974) - Endothermy, dinosaurs and Archaeop- " !'
SMITH, P.E.; EVENSEN, N.M & YORK, D. (1996) - Ar- Ar Lazer
teryx. Evolution, 28(3): 503-504.
Probe Dating of the Yixian Formation, Northeast China and
FORSTER, C.A.; SAMPSON, S.D.; CHIAPPE, L.M. & KRAUSE, D.W. the Early Evolution of Birds. J. Vertebr. Paleontol., 16(3): 67A.
(1998) - The Theropod Ancestry of Birds: New Evidence from WAGNER, A. (1861) - Neue Beiträge zur Kenntnis der urweltlichen
the Late Cretaceous of Madagascar. Science, 279: 1915- Fauna lithographischen Schiefers; Compsognathus longipes
1919. Wag. Abh. bayer. Acad. Wiss., 9: 30-38.
GRIFFITHS, P.J. (1994) - The Claws and Digits of Archaeopteryx
WELLNHOFFER, P. (1974) - The Fifth Skeletal Specimen of Ar-
lithographica. Geobios, M. S., 16: 101-106.
chaeopteryx. Palaeontographica, A, 147: 169-216.
GRIFFITHS, P.J. (1996) - The Isolated Archaeopteryx Feather. Ar-
chaeopteryx, 14: 1-26.
GAUTHIER, J. (1986) - Saurischian Monophyly and the Origin of
Birds, in PADIAN, K. (Ed.), The Origin of Birds and the Evoluti-
on of Flight. Mem. Calif. Acad. Sci. 8:1-55.
HOLTZ, T.R., JR. (1994) - The phylogenetic position of the Tyran-
nosauridae: Implications for theropod systematics. J. Pa-
leontontol., 68(5): 1100-1117.
HOU, L.; ZHOU, Z.; MARTIN, L.D. & FEDUCCIA, A. (1995) - A Beaked
Bird from the Jurassic of China. Nature, 377: 616-618.
JI, Q., & JI, S.A. (1996) - On Discovery of the Earliest Bird Fossil in
China and the Origin of Birds. Chinese Geol., 233: 30-33.
JI, Q., & JI, S.A. (1997) - Protoarchaeopterygid Bird (Protoar-
chaeopteryx gen.nov.) - Fossil Remains of Archaeopterygids
from China. Chinese Geol., 238: 30-33.
JI, Q.; CURRIE, P.J.; NORELL, M.A. & JI, S.A. (1998) -Two Feathe-
red Dinosaurs from Northeastern China. Nature, 393: 753-
761.
LACASA, A. (1985) - Nota sombre Las Plumas Fosiles del Yacimi-
ento Eocretacio de "La Pedera - La Cabrua" en la Sierra del
Montsec (Prov. Lleida, España). Ilerda Inst. Estud. Ilerdenses
Diputatión Prov. Lleida, 46: 227-238.
MADERSON, P.F.A. (1972) - On How an Archosaurian Scale May
Have Given Rise to an Avian Feather. Am. Naturalist, 106
(949): 424-428.
MAYER, H.V. (1861) - Vogul-Federn und Palpipes priscus von Sol-
nhofen. N. Jb. Mineral. Geol. Palaeontol., 1861: 561.
OSBORN, H.F. (1924) - Three New Theropoda. Protoceroatops
Zone, Central Mongolia. Am. Museum Nat. Hist. Novit., 144:
3-12.
OSTROM, J.H. (1969) - Osteology of Deinonychus antirrhopus, an
unusual theropod from the Lower Cretaceous of Montana.
Bull. Yale Peabody Museum Nat. Hist., 30: 1-165.
OSTROM, J.H. (1973) - The ancestry of birds. Nature, 242 (5393):
136.
403