[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Re :_undiscovered_dinosaurs

In a message dated 98-05-22 03:02:21 EDT, forelf@internet19.fr writes:

<< > I remember seeing somewhere
 > (perhaps in the archives) that someone had estimated that there were
 > thousands of species of dinosaurs during the course of the Mesozoic.  Does
 > anyone have any references estimating the number of dinosaur species that
 > remain undiscovered?
 You could try the following refs: >>

I think those references considerably understimate the number of dinosaur
species that may have been extant throughout the Mesozoic. Here is my
estimate, as excerpted from my working manuscript of Mesozoic Meanderings #2
third printing:

Dinosaur Diversity and Extinctions

HOW MANY dinosaurs were there? This question was recently addressed by Dodson
(1990a), who applied statistical methods and a few seemingly reasonable
assumptions about collection and preservation biases to the dinosaur fossil
record. The total number of dinosaur genera that ever lived, according to his
calculations, lies somewhere between 645 and 3285, with a most probable value
between 900 and 1200.
I was asked the same question in an interview for a television show on
dinosaurs, and I estimated the answer differently: Dinosaurs dominated the
world's terrestrial biota for approximately 150 million years. Extant
vertebrates that likewise dominate their domains are bony fish, birds, and
mammals, and there are several thousand species of each: over 20,000 species
of fish, 9672 species of birds, over 3000 species of mammals. The present
epoch is, in my opinion, one of average vertebrate diversity, neither
particularly depauperate nor particularly bounteous, and "several thousand
species" may be taken as characteristic of the diversity of any of the high-
level vertebrate taxa. If anything, the ice ages and the advent of man have
made the present epoch more depauperate than usual. In any case, exactitude is
not necessary in this kind of order-of-magnitude calculation (as if we could
ever know the exact number of vertebrate species that existed during a
particular epoch!), so I chose a figure of 6000 as a likely number of dinosaur
species extant at any one time. If, like Dodson, one imagines dinosaurs to
have occupied only megafaunal niches, then the number could well be
substantially smaller. But I believe that the fossil record is hyperbiased
against the preservation of the smaller, Compsognathus-size dinosaurs.
Dinosaurs of this size class, known from scattered and fragmentary remains
worldwide, must have been much more numerous and diverse than might be
imagined solely from the minuscule number of specimens presently known in the
fossil record.
Only two known specimens, one German and one French, have been referred to the
genus Compsognathus. Very few places in the world have the fine-grained
lithographic limestones wherein the fossils of small, delicate-boned dinosaurs
like Compsognathus might be preserved, so it is extremely unlikely that the
only compsognathids that ever lived just happened to live there. I contend
that scores of as-yet-unknown compsognathid species existed worldwide?enough
species that two individuals were bound on death to drift into those rare
European limey lagoons to be preserved.
Dinosaur species appear to originate, flourish, and become extinct?or evolve
into new species?during a span of time considerably less than 4.5 million
years. This figure is obtained simply by dividing the number (four) of well-
known successional dinosaurian faunas in the Campanian to Maastrichtian of
North America (Fox, 1978) into the span of time (18 million years) during
which those faunas endured. On this time scale, even the generic turnover
between the faunas was for all practical purposes complete, and I have not
considered the non-depositional, trangressive periods between the faunas, when
species of intermediate morphology (Horner, Varricchio & Goodwin, 1992) must
have existed that largely escaped preservation.
One hundred fifty million years thus yields some 33 or 34 complete species
turnovers among the dinosaurs. When multiplied by an average of 6000 species
per epoch, this gives a grand total of about 200,000 dinosaur species that
ever lived. Assuming a conservative average of 10 species per genus?some
extant vertebrate genera have more than 30 species per genus, others only one
or two?this amounts to 20,000 genera: nearly an order of magnitude more than
the maximum figure calculated by Dodson.
What is the source of this rather huge discrepancy? I see nothing unreasonable
about the assumptions and back-of-the-envelope calculations outlined above.
But as I just explained, Dodson may have underestimated the effects of
preservation biases. There is not enough space here to question his
assumptions in detail; that is the stuff of a separate paper. Instead, I will
offer my own thoughts about such biases. Perhaps these will explain the
One factor apparently not considered in calculating preservational biases is
that all known dinosaur fossils necessarily occur in depositional
environments: coastal, fluviatile, lacustrine, or eolian. Such environments
(e.g., seashores, river deltas, lake bottoms, and deserts) presently comprise
about 10% of the earth's total land area (a crude estimate based on visual
examination of world maps). At any given time, it seems, some 90% of the
world's land area is subject to erosion and subsequent deposition on the other
10%. Epicontinental seas, such as the one that covered North America during
the Cretaceous Period, foster the development of coastal, fluviatile, and
lacustrine depositional regions, and the fraction of earth's land area that
was depositional during periods of major epicontinental transgression may have
attained 25% (Haubold, 1990). Even so, this means that between 75?90% of the
dinosaur species extant in any one epoch may never be found as fossils, simply
because they did not frequent depositional environments. Dinosaur fossils are
often found in assemblages that include the remains of freshwater vertebrates
such as fish and crocodylians. We are quite certain that the fish and
crocodylians lived exclusively in or near depositional environments. Why,
then, should we imagine that most of the known dinosaur species?particularly
the smaller species, for which extensive migrations were unlikely?did
Weishampel (1990a) listed 416 dinosaur-bearing localities worldwide. This
seems a large number, but it is as nothing?a few data points?when compared
with the great expanse of the earth's surface where dinosaurs could and did
once flourish. I believe we have already collected a reasonable fraction of
all the dinosaurs that ever lived in depositional environments. But we have
only meager evidence?a few isolated, fragmentary remains of animals that
happened to fall into rivers when they died?of the other kinds of dinosaurs
that once inhabited the earth. Dinosaurs undoubtedly lived in the Mesozoic
equivalents of grasslands, prairies, rain forests, and highlands, but because
these are not places where sediment accumulates, dinosaurs that lived there
would have been entirely lost to the fossil record.
Another factor to consider is that typical depositional environments were
probably not very speciose. Today's tropical rain forests contain the most
diverse faunas by far, but such forests, in which the carcasses of even large
animals are quickly dismembered and decompose, are taphonomically inimical to
the preservation of fossils. Triassic "rain forests," not swamps, river
deltas, and lakeshores, would have been the most favorable environments for
the evolution of arboreal animals. This might well explain why the fossil
record of basitheropods remains highly impoverished.
If dinosaurs diversified into an average of 6000 species per 4.5-million-year
depositional epoch, we would expect by my previous area-ratio calculation that
only 600?1500 inhabited environments where fossilization might occur. This,
then, becomes a loose upper bound on the number of species per epoch we could
in principle discover. Quite a few such deposits have already eroded away,
taking their loads of species with them; as a rule, the farther back we go in
the Mesozoic, the fewer such deposits have remained to be prospected. It is
well known, for example, that the Upper Cretaceous provides the largest number
of dinosaur-bearing formations (126 out of 416; Weishampel, 1990a) throughout
the world and, correspondingly, exhibits the greatest dinosaur diversity. This
I ascribe in large part to the formation-age fossilization bias.
Further supporting my assertion that dinosaur diversity was considerably
higher than Dodson's estimates is the observation that formations laid down at
different times in different places seldom yield the same dinosaur species.
The dinosaurs of China, Mongolia, Argentina, Africa, North America, Europe,
and Australia all belong to distinct species and almost always to distinct
genera. Whenever a fresh locality with good dinosaur material is prospected
(and there are never enough such sites), new dinosaur species are soon added
to my table. It does not seem to be the case that dinosaurs wandered between
unconnected depositional areas, for then we would find many more instances of
the same genera and species in widely separated localities. Some depositional
areas, however, seem to have extended for thousands of miles without
significant geographical barriers?along the shores of epicontinental seas, or
when all the continents were joined into a single Pangaean land mass?which did
permit certain genera and species considerable ranges. But I feel on
reasonably firm ground when I maintain that the dinosaurs in between the
regions of deposition?the dinosaurs unavailable in the fossil record?belonged
to genera and species different from those already known.
The diversity of the dinosaurs we already know practically demands the
existence of those numerous "transitional" genera and species so notoriously
absent from the fossil record. For example, Horner, Varricchio & Goodwin
(1992) reported no less than three pachyrhinosaurine taxa from the Two
Medicine Formation of Montana transitional between the closely related genera
Styracosaurus and Pachyrhinosaurus (as well as one transitional
pachycephalosaurid, lambeosaurid, and tyrannosaurid species), and there is
good reason to expect several more besides. (Two of the three transitional
pachyrhinosaurines represent new genera: Einiosaurus and Achelousaurus:
Sampson, 1994.) If three or more? in most cases many more?unknown taxa stand
behind every known dinosaur taxon, it is easy to see how the total number of
dinosaur species might approach or even exceed 200,000.
Although I have not analyzed the situation with any rigor whatsoever, it seems
that the number of new dinosaurs described at any time correlates best with
the number of paleontologists at work and with the strength of establishment
support for paleontology during that time (see, e.g., Dodson & Dawson, 1991).
This certainly shows that we have not yet approached a true upper limit to
dinosaur diversity, when excavating new sites would just add more specimens of
dinosaur taxa already known, no matter how many persons are at work.
Finally, it should be noted once more that the bulk of species diversity in
extant vertebrate groups generally lies among the smaller forms. If there were
only large dinosaurs, then Dodson's estimates could well be near the mark, for
as he asserts, his diversity estimates agree in order of magnitude with those
of extant large mammals. But I contend that there were many more small
dinosaurs?including but by no means limited to pigeon-size and chicken-size
arboreal, climbing, gliding, and flying forms?than anyone has previously
guessed. Their existence follows from the pattern of archosaurian evolution as
outlined in this work. Their physical traces include the unidentifiable teeth
and other skeletal fragments of small theropodomorphs and ornithischians found
in many museum collections, and rich footprint faunas that document
significantly higher local dinosaur diversities than do skeletal remains
(Lull, 1953; Ellenberger, 1970, 1972, 1974; Olsen & Galton, 1984; and others).
Dodson (pers. comm.) has suggested that juvenile and subadult large dinosaurs
occupied many niches that would otherwise have been taken by adult small
dinosaurs. This effect would have depressed the number of small-dinosaur
genera and species and kept the total number of species well below my
estimates. Unfortunately, juvenile large dinosaurs are often difficult to
distinguish from adult small dinosaurs (Callison & Quimby, 1984), especially
in view of the poor fossil record of small dinosaurs, so a definitive test of
this thesis is probably not yet possible. It bears pointing out, however, that
many diagnostic small-dinosaur remains do exhibit adult features (fusion of
the skull elements, fusion of neural arches to centra, and so forth), so small
adult dinosaurs probably did exist. On the other hand, Horner (in Horner &
Gorman, 1988) found patterns of rapid growth in Maiasaura hatchlings and
juveniles, suggesting that dinosaurs in general and large dinosaurs in
particular spent most of their lives at or near fully adult size. Taken
together, these factors suggest a relatively minor ecological role for the
small juveniles of large-adult dinosaurs.
After taking these kinds of biases into account, we are left with the handful
of species per epoch that are actually present in our museum collections. The
point to ponder here is that enough slack still exists in the statistics to
allow the fossil record to support either Dodson's "lowball" diversity
estimates or my "highball" estimates. Consequently, we may ultimately have to
look elsewhere than the fossil record to estimate dinosaur diversity more