Geoscience Reference
In-Depth Information
whereas a few authors have argued that there is sub-
stantial evidence for widespread saltwater (and
brackish water) tolerance in early stegocephalians
(Schultze & Maples 1992; Schultze et al. 1994;
Schultze 1995, 1999; Laurin & Soler-Gij
ยด
n 2001,
2006). The latter point of view was eloquently sum-
marized by Schultze (1985, p. 2):
Vertebrate remains are commonly used as terrestrial or
freshwater indicators, even all the fish... This tra-
ditional interpretation is mainly based on the fact that
complete vertebrates most commonly occur alone,
rather than together with invertebrates. In many
cases, this isolation results from preservational biases
(calcium phosphate v. calcium carbonate), and not
palaeoecological differences. Sometimes, the associ-
ation of isolated elements of vertebrates with marine
invertebrates has been explained as allochthonous,
with the vertebrate remains having been washed in.
No
dipnomorphs)
to
an
application
of
Witmer's
(1995) extant phylogenetic bracket.
Extant amphibians (almost all of which are
freshwater or terrestrial) have been thought to be
polyphyletic (Fig. 1a) for much of the 20th
century (Moodie 1916, pp. 46 - 49). More recently,
this point of view was developed by Carroll &
Holmes (1980) and Carroll & Currie (1975), who
argued for independent origins of urodeles and gym-
nophionans from 'lepospondyls' whereas anurans
were thought to be derived from 'temnospondyls.'
More recently, Schoch & Carroll (2003) suggested,
based on developmental data, that anurans and uro-
deles are temnospondyls whereas gymnophionans
are 'lepospondyls'. Schoch (2006) subsequently
reached different conclusions, based on a much
more rigorous analysis of developmental data. Fur-
thermore, early sarcopterygians were thought to
have inhabited only freshwater, as the extant lung-
fishes (Romer 1966).
Only reptiliomorphs were thought to include a
large number of saltwater-tolerant forms, as shown
by the large number of marine amniotes. For
instance, Neill (1958) listed 273 species or subspe-
cies of 'reptiles' (in his usage, contrary to ours,
this excluded birds) which lived at least occasion-
ally in salt- or brackish water. Given such premises,
the parsimony criterion (although it may not have
been explicitly used) suggested that most early ste-
gocephalians were restricted to freshwater and that
saltwater tolerance appeared within reptiliomorphs
(Fig. 1a). This does not imply that all reptiliomorphs
tolerated saltwater; these data and phylogeny make
no prediction about stem-reptiliomorphs which, at
the time (Romer 1966), were thought to include
embolomeres, seymouriamorphs, gephyrostegids,
Solenodonsaurus and diadectomorphs.
Later, the environments in which many early
sarcopterygians were found (including lungfishes)
were reinterpreted as coastal, deltaic or marine
(Carroll 1988; Janvier 1996; Schultze 1999).
Schultze (1997) even suggested that stegocepha-
lians originated in an intertidal environment.
Thomson (1980) also argued, on the basis of palaeo-
geographic arguments, that most groups of early sar-
copterygians were either marine or euryhaline; in
most cases, extinct stegocephalians of these two
ecological categories cannot be distinguished. The
term euryhaline is therefore used in this study, but
it should be understood that this only means that
the taxon could live in saltwater; this does not
exclude the possibility that it could also live in
freshwater and on land. Under such conditions, the
intolerance to the marine environment could be
seen as a specialization of the clade that includes
at least the last common ancestor of gymnophiones,
urodeles and anurans and all its descendants. Under
the phylogenies advocated by Carroll & Holmes
recent
example
of
such
association
has
been
recorded.
We provide a historical review of ideas and
recent evidence of the habitat of extant amphibians
and lungfishes and of Palaeozoic finned sarcoptery-
gians in a phylogenetic context. We demonstrate
that in the late 19th and early 20th century, palaeon-
tologists had objective reasons to expect early stego-
cephalians to be stenohaline, freshwater forms. The
review also shows that these objective reasons have
been refuted, and that there is no reason to expect
early stegocephalians to have been confined to
freshwater. We show how all data converge to
suggest a marginal marine habitat for the earliest
stegocephalians, and a long and widespread reten-
tion of salt- and brackish water tolerance in Palaeo-
zoic stegocephalians.
Parsimony and habitat of extant and
extinct sarcopterygians: a historical
perspective
In this section, we will consider how ideas about ste-
gocephalian phylogeny, observations of the habitat
of extant sarcopterygians and more recent infor-
mation about the habitat of early sarcopterygians
(especially dipnomorphs) may have influenced our
expectations about the habitat of early stegocepha-
lians, using parsimony as a criterion. Parsimony
may not have been explicitly invoked in early
works on this problem but (at least implicitly) it
has probably been used as a general scientific prin-
ciple. In this section, we will disregard direct evi-
dence about the habitat of early stegocephalians,
which will be presented separately (below). This
section can be seen as an attempt to use parsimony
to infer habitat of early stegocephalians (globally)
in the context of various phylogenies. It is analogous
(except in the inclusion of data on habitat of early
