Geoscience Reference
In-Depth Information
so that the presumed presence of this genus in
Permo-Carboniferous localities has been considered
an indication of marine influence (e.g. Cassle et al.
2006). The persistence of a taxon traditionally
ranked as a genus from the Carboniferous to the
present should in any case have been suspect,
since even Lingula, often considered a living
fossil, does not occur in the Palaeozoic; fossils pre-
viously attributed to this taxon have been reassigned
to other genera of the Lingulidae (Emig 2003).
Nevertheless, this change in classification of lingu-
lids illustrates, to an extent, the subjective nature
of absolute (Linean) ranks (Laurin 2008).
Interestingly, Palaeozoic Spirorbis (a micro-
conchid) is a euryhaline form which occurs in the
microtidal Joggins (Falcon-Lang et al. 2006), Saar-
Nahe, Saale and Bohemia (Schultze & Soler-Gij ´ n
2004). Extant (genuine) Spirorbis, and more gener-
ally serpulids (which include Spirorbis) and at least
some other marine annelids, tolerate a wide range of
salinity (Ushakova 2003).
It might be useful if the parsimony criterion were
used in an explicit phylogenetic context (Fig. 1) to
reassess the significance of palaeoenvironmental
markers whenever possible. This might be feasible
for at least some mollusks, brachiopods and arthro-
pods. Also, all metazoans which do not belong to the
crown-groups which appeared before the Devonian
(and in some cases, well after) may plausibly have
been marine, given that the oceans and seas appear
to be the cradle of early metazoan diversification
(Barnes 1987; Clarkson 1998). In the absence of
the contrary, a marine habitat is a more reasonable
null hypothesis than a freshwater habitat, although
it should always be tested.
Detailed phylogenies are now available for many
relevant clades, and some of these include extinct
taxa (Wheeler et al. 1993; Waggoner 1996) or
include enough morphological characters to enable
inference of the position of extinct taxa (Jenner &
Schram 1999; Giribet et al. 2001; Collins 2002;
Jacobs et al. 2005; Collins et al. 2006); use of parsi-
mony and of an explicit phylogeny might therefore
yield additional data. For instance, the tenuous
interpretations by vertebrate palaeontologists of
several Palaeozoic localities as freshwater environ-
ments may have influenced other palaeontologists.
The suggestion that in the Devonian, 'spirorbiform
microconchids began to inhabit brackish and fresh-
water environments in addition to marine settings'
(Taylor & Vinn 2006, p. 227) may rest partly on
the interpretation of palaeoenvironmental prefer-
ences of stegocephalians.
It would be useful to try to resolve the phyloge-
netic position of microconchids and use the parsi-
mony criterion to assess their environmental
preferences. Given the number of problems (noted
above) which have marked the freshwater/marine
controversy, such an approach might yield useful
new insights.
Note that much of the discussion above has
treated saltwater tolerance as a discrete character
because, when little detailed information is avail-
able, this is the only applicable technique. When
more quantitative data are available (as for salinity
tolerance in some species of lissamphibians and
Nereis) however, squared-change parsimony and
independent contrasts could conceivably be used
to estimate environmental tolerance with confidence
intervals. Such techniques were recently used to
study body size evolution (Laurin 2004). Another
possible approach to determine habitat preference
and breadth of ancient organisms might be inference
models built upon observable characters which can
be shown to be correlated to the relevant environ-
mental variable. However, such models should be
applied only within the clades in which such corre-
lations have been tested, as suggested by an exten-
sion of the extant phylogenetic bracket principle to
continuous characters (Laurin et al. 2004, p. 607).
Habitat and the fossil record
The frequent, relative intolerance of lissamphibians
to saltwater could explain (at least partly) why their
fossil record is so much poorer than that of most
other groups of stegocephalians. Stem-amphibians
closer to lissamphibians than to lysorophians (or
dissorophoids) may have lived away from the
coast, possibly in upland environments, from
which the fossil record is generally poor. Sediments
deposited inland, above the sea level, are much more
likely to be eroded quickly than sediments deposited
in coastal areas, just below the sea level. Even rocky
shores have an extremely poor fossil record because,
despite their low altitude, they are areas of erosion
(Schultze 1999, p. 373). This could explain the
large stratigraphic gap between presumed sister-
groups of lissamphibians (which appear in the
middle Upper Carboniferous, such as lysorophians
and dissorophoids) and the oldest known lissamphi-
bians, such as Triadobatrachus (Rage & Rocek
1989) and Czatkobatrachus (Evans & Borsuk-
Bialynicka 1998), from the Early Triassic. That
poor record hampers direct comparisons between
molecular and palaeontological estimates of the
age of Lissamphibia (Zhang et al. 2005), although
indirect comparisons show no incompatibility and
suggest that the record is sufficiently good to
assess the age of origin of several lissamphibian
taxa (Marjanovi´ & Laurin 2007).
Habitat of early stegocephalians
This review attempts to shed new light on the long-
debated
problem
of
the
original
environment
Search WWH ::




Custom Search