Geoscience Reference
In-Depth Information
were still in the framework of a transcendental view
of homology (e.g. Geoffroy Saint-Hilaire 1807;
Owen 1861). Chambers' (1844) Vestiges suggested
a water-to-land transition that was inspired by the
Lamarckian views but was closer to modern scen-
arios since it considered adaptation to environ-
mental changes through time in a more elaborate
geodynamic context. Miller (1849), although a
creationist and catastrophist in the same vein as
his mentor Louis Agassiz and many other naturalists
of his time, also envisaged the transition of life
from water to land in a detailed chronology-based
sequence that was more accurate than Chambers'.
However, Chambers viewed this transition as
a continuum between for example, 'sauroid'
fishes 'calculated to breathe the atmosphere' and
'reptiles' (then including amphibians), whereas
Miller considered the earliest tetrapods ('reptiles')
as the result of a 'different creation'
Notably, the order of appearance of the six synapo-
morphies of the earliest land plants (multicellular
sporophyte, spore with sporopollenin, cuticle, sto-
matas, sporangia and gametangia) remains undocu-
mented by fossils.
In the case of vertebrate terrestrialization, lung-
fishes (and possibly the coelacanth since 1938)
were for a long time the only available hints. The
hesitation of the 19th century naturalists suggests
that these hints were unconvincing (see review in
Rosen et al. 1981). Vertebrates (and to some
extent arthropods) have a relatively good and anato-
mically informative fossil record that fills the mor-
phological gaps by providing examples of extinct
character combinations, however. Before the dis-
covery of such key fossil taxa, the rise of land ver-
tebrates had been a highly controversial matter.
Since Geoffroy St Hilaire's publication (1807),
living bichirs (polypteriforms) were regarded as
the most likely closest living piscine relatives of tet-
rapods until the discovery of the living lungfishes
and the controversies this raised. When finally
recognized as tetrapod-like fishes (and not fish-like
tetrapods), living lungfishes appeared as the sister
group of tetrapods in the first evolutionary trees of
vertebrates proposed by Haeckel (1866). This was
soon followed by Darwin's supporters, notably
Huxley (1876). Huxley, however, failed to recog-
nize the early fossil lungfishes and instead gathered
them with his crossopterygians (lobe-finned fishes;
now part of the sarcopterygians).
A first important clue came with Baur's (1896)
discovery that the earliest tetrapods known at that
time, the Carboniferous temnospondyls, were
similar to some of the fossil crossopterygians,
notably by their folded tooth structure. Watson
(1912) then provided evidence for choanae (internal
nostrils) in some crossopterygian fishes (Mega-
lichthys and Eusthenopteron), which also happened
to have somewhat limb-like paired fin skeletons.
The assembly of the tetrapod-like characters of the
late Devonian tetrapodomorph fish Eusthenopteron
culminated with Jarvik's (1942, 1980) detailed
description of this now iconic fossil.
Around much the same time, the Danish Green-
land expeditions also yielded the fish-like tetrapod
Ichthyostega (or 'four-legged fish') which com-
bined limbs with digits and a caudal fin covered
with fish-like fin rays that are composed of lepido-
trichs (Jarvik 1952). The junction between fish and
tetrapods was further completed by a number of
new early tetrapod discoveries in the 1980-1990s
and the identification of elpistostegalians as the
closest piscine relatives of tetrapods, thereby deth-
roning Eusthenopteron from this position (Clack
2002; Daeschler et al. 2006). Although the scenarios
of the fish-tetrapod transition are diverse, their
polarity is now firmly established: from fishes to
in the
Carboniferous.
Vertebrates and embryophytes: two
radically different case studies
Since the early 20th century, vertebrate history has
been regarded as a good case study for the transition
of animals from water to land. This is despite a
discontinuous extant taxonomic record; that is, the
phylogenetic pattern of the living taxa shows gaps
which frustratingly correspond to major adap-
tive events (i.e. the agnathan-gnathostome, fish-
tetrapod, amphibian-amniote or reptile-bird gaps).
This paucity of living 'intermediate forms' or, at
any rate, forms that display 'intermediate charac-
ters', is the result of deep divergences between
large clades and extensive extinctions (Donoghue
& Purnell 2005).
Such gaps seem to be frequent in metazoans,
notably in protostomes and deuterostomes, but vir-
tually absent in embryophytes (land plants) whose
major classical living groups (e.g. bryophytes, pter-
idophytes and gymnosperms) are almost all para-
phyletic (except for angiosperms) (Donoghue
2005). However, the monilophytes (ferns and horse-
tails), formerly unrecognized as a group, are now
regarded as a clade, sister to the lignophytes
(Pryer et al. 2001). Therefore, the progressive
assembly of the bauplan of the most terrestrialized
embryophytes, except for the homoplastic large
leaves (megaphylls) of pteridophytes and gymnos-
perms, can be more readily reconstructed on the
basis of the living taxa than that of land vertebrates
(living amphibians and amniotes being clades, and
only reptiles being paraphyletic).
Conversely, 'intermediate' structures are perhaps
more difficult, for material reasons, to infer from
fossil embryophytes than from fossil vertebrates.
