ÐÓ°ÉÔ­´´

Reptiles have long been regarded as morally corrupt, rather low sorts
of creatures – witness Victorian natural history books with their tales
of cold-blooded crocodiles and strange and terrible dinosaurs. And of all
their alleged corruptions, it is cannibalism that holds the greatest fascination.
But just how prevalent was cannibalism among the dinosaurs and their primeval
colleagues? Palaeontolo-gists have been debating this question for years,
scouring the fossil record for clues. The focus of their attention has been
the ichthyosaurs, dolphin-like animals that lived during the Mesozoic, between
245 and 65 million years ago.

Over the past 200 years, many examples have emerged of adult ichthyosaurs
with juveniles trapped in their rib-cages; the best come from Holzmaden,
near Stuttgart, in Germany. The key question has been whether these juveniles
are prey or embryos. Now, a study by Roland Bottcher, of Stuttgart’s natural
history museum, may have finally settled the matter. Having pored over the
evidence, Bottcher finds the icthy-osaurs not guilty: the juveniles are
unquestionably embryos, he says. Additional support for this verdict comes
from an ichthyosaur fossil found recently in Britain, now on show in Bristol
City Museum.

Bottcher’s study is important because it deals a hefty blow to the whole
notion of ancient reptiles as cannibals. Aside from the ichthyosaur skeletons,
evidence that extinct reptiles ate members of their own species is extremely
scarce. Indeed, there is only one unequivocal specimen, a skeleton of an
adult Coelophysis discovered in New Mexico in 1947 and dated at about 220
million years ago (the late Triassic). Tellingly, the fossil harbours the
skeleton of a second Coelophysis, not only curled up in a suspicious manner
inside the adult’s ribcage, but also far too large to be an embryo.

Bottcher’s study also establishes beyond doubt that the ichthyosaurs
bore live young instead of laying eggs, as is more usual for reptiles. It
confirms what many palaeontologists have long suspected: that for ichthyosaurs,
live birth would have been essential. Being confined to water, the creatures
would have been unable to lay eggs in the sand as modern aquatic turtles
and crocodiles do. And, being air-breathing, the embryos in any eggs laid
under water would simply have suffocated. The acceptance of these ideas
means that researchers can now start to make interesting comparisons between
the ichthyosaurs and the many groups of modern lizards and snakes which
have evolved viviparity (the ability to retain embryos in the oviducts until
they have developed sufficiently to lead independent lives).

The first ichthyosaur carrying a juvenile came to light at Holzmaden
as long ago as 1749. Yet the fossil was not described in print until 1824,
and even then the author, Georg Friedrich von Jaeger, failed to spot the
juvenile. It was only some 20 years later, following other finds, that the
controversy began in earnest. Some naturalists were happy to accept the
idea of ichthyosaurs giving birth to live young, while others immediately
interpreted the ‘included’ juveniles as evidence of cannibalism.

As more and more ichthyosaurs carrying young emerged, the debate gathered
momentum. Since the 1840s Holzmaden has yielded several hundred skeletons
of adult ichthyosaurs, of which some fifty appear to carry young; and researchers
are still finding new specimens. Supporters of the embryo theory have always
explained the high proportion of suspect fossils by arguing that pregnant
females, especially those close to, or at, term were particularly vulnerable.
But can such an argument account for all the specimens?

A German naturalist called Wilhelm von Branca believed not. In 1908
he published what became the authoritative account of why some of the skeletons
within skeletons had to be prey. His argument rested on five observations:
some adults contain juvenile skeletons of various sizes; some juveniles
lie close to remnants of chewed food; most skeletons face the same way as
the adult (which implies tail-first birth, something Branca considered unlikely);
most skeletons lie away from the region of the uterus; and ribcages commonly
contain up to 10 or 11 skeletons (Branca thought this significant because
whales and dolphins generally give birth to a single, large calf).

Initially, the influence of Branca’s study extended well beyond the
ichthyosaurs themselves. It encouraged the belief that many, if not most,
extinct reptiles had had cannibalistic tendencies, and even fuelled suspicions
about modern reptiles. As naturalists learnt more about the habits and reproductive
biology of modern reptiles, the study’s influence waned, but the notion
of the ichthyosaurs as cannibals persisted. Nobody bothered to challenge
Branca’s arguments in detail – that is, until Bottcher decided to reopen
the case.

Bottcher overturns each of Branca’s arguments and marshals more evidence
to support the opposite view. First, the way the young are preserved points
to their being in the uterus rather than in the gut, he argues. There are
signs that in certain specimens the gases released by the decomposition
of the adult’s flesh have forced one or more of the embryos out of the cloaca
tail first. In some cases the embryo’s head remains lodged between the pelvic
bones, while its body lies outside. Had the juveniles been swallowed, gases
within the adult would have pushed them in the opposite direction, up into
the ribcage. Only partly digested specimens in the intestine would have
been within reach of the pelvic area – and no such fossils have been discovered.

Some specimens appear to contradict this interpretation, as they contain
juveniles at the front of the adults’ ribcages. Bottcher’s explanation is
that this position is close to that of the oviducts of living reptiles.
By analogy with modern viviparous lizards and snakes, he argues, ichthyosaur
oviducts probably expanded during gestation to occupy both sides of the
ribcage as far forward as the shoulder area. The right oviduct extended
further than the left, and the embryos probably moved from side to side
to make best use of the available space. According to specimens that preserve
traces of internal organs, the stomach, heart and lungs lay below the oviducts.
Any young contained in the stomach would end up lying nearer the underside
of the ichthyosaur skeletons.

A third piece of evidence comes from the way the skeletons lie. Some
of the very small skeletons are curled up as though confined by an egg membrane,
but most of the larger skeletons are stretched out, parallel to the backbone
of the adult. Had they been swallowed whole, they would be more disordered.
It seems that the ichthyosaurs had evolved beyond the first stage of bearing
live young, in which the mother retains her eggs inside her body to provide
a kind of protective nest, with the yolk sac providing all nutrition. At
least at larger sizes, it would appear, ichthyosaur young were nourished
by maternal nutrients that passed through a placenta. Further support comes
from the fact that the eggs of many modern groups of lizard have components
which function as a kind of placenta and which have evolved quite independently
of the better-known mammalian placenta.

The fact that the heads of the young skeletons nearly always point forward
is in itself an important clue. Bottcher notes that modern animals which
devour their prey whole, such as snakes and fishes, tend to swallow them
head first. Branca’s interpretation was quite different. Based on the behaviour
of modern animals, he argued that ichthyosaurs must have given birth to
their young head first, and that the small skeletons were therefore pointing
the wrong way to be embryos. This reasoning turns out to be flawed: mammals
that usually give birth to a single, large juvenile follow the ‘head-first’
rule, but mammals that bear large numbers of less-developed young show no
preferred orientation.

The behaviour of other aquatic animals also casts doubt on Branca’s
interpretation. Viviparous bony fishes give birth both ways, while young
sharks are usually born tail first. The coelacanth’s preference has changed
during the course of evolution. The young of the living fish Latimeria are
born tail first, but the young of the extinct coelacanth Undina were born
head first, according to fossil evidence. Limited observations suggest that
the young of whales and hippopotamuses are also born tail first.

Tail-first birth is an obvious adaptation for air-breathing animals
confined to water. The supply of oxygen from the mother is relinquished
only at the last minute, giving the juvenile ample time to swim to the surface
to fill its lungs for the first time. Also, studies of modern animals show
that the heaviest part of the body often lies lowest in the uterus and so
is born last. For whales and ichthyosaurs, unlike land-dwelling mammals,
this is the front part of the body because the hind limbs and hip, no longer
needed for walking, are comparatively small.

If the young are embryos, there ought to be a relationship between their
sizes and those of the adults. This is exactly what Bottcher finds: in general,
the bigger the adult icthyosaur, the bigger its juveniles. Although a similar
trend could result from cannibalism – bigger predators can tackle bigger
prey – the relationship would be expected to blur for the largest adults,
who would tend to eat both small and large prey. Branca made much of the
fact that a few adult ichthyosaurs do contain young of wildly different
sizes. But these could in fact be embryos at different stages of development.

In any case, Bottcher resolves any doubts by showing that the adult
ichthyosaurs and their respective young obey an established equation governing
the sizes of reptiles and their hatchlings. The equation, first proposed
by Phil Currie of the Royal Tyrrell Museum of Palaeontology in Alberta and
Bob Carroll of McGill University in Montreal, is purely empirical but holds
true for 120 species of living lizards and crocodilians. It would be remarkable
if adult ichthyosaurs had chosen to feed only on juveniles whose sizes match
those predicted for their offspring at term.

The parameters for the equation vary according to reproductive strategies.
Robin Andrews of the Virginia Polytechnic Institute and State University,
Virginia, has found that the parameters are slightly different for each
group of living reptiles. The hatchling sizes of crocodilians and turtles
vary little with adult size, probably because these creatures lay extremely
large clutches of eggs (up to 200). Fewer eggs at a sitting means that larger
mothers can invest more per egg and produce larger hatchlings. Lizards lay
5 eggs on average, and most snakes lay between 1 and 20 eggs. The parameters
for ichthyosaurs are close to those of lizards and snakes.

The final piece of evidence is based on the anatomy of the gut. The
ichthyosaur’s stomach appears simply too small and complex to have accommodated
stretched-out juveniles. Bottcher doubts that an ichthyosaur could have
digested such large, bony prey, and the behaviour of modern crocodilians
bears this out. Large prey animals get stuck in the throats of crocodiles
until the parts in the stomach have been digested. If ichthyosaurs really
did eat large juveniles, one would expect to find adults with juvenile skeletons
poking out of their mouths.

None of the fossils contains evidence of partially digested juveniles.
Studies show that ichthyosaurs ate small cephalopods and fishes, which were
rapidly broken down into a sludgy mess inside the stomach. The hard bits
of a prey animal, such as the chitinous hooklets of the cephalopods, were
retained in the stomach; small fish bones were probably dissolved by the
digestive juices. It would have been quite impossible for complete skeletons
of young ichthyosaurs to pass into the deepest regions of the stomach or
into the intestine.

Only two observations are difficult to square with the embryo theory.
The first is that some adults contain juveniles of various sizes. Bottcher’s
reply is that most of these juveniles lie above the gut. Instead of reflecting
cannibalism, he argues, the size variation suggests ichthyosaurs carried
young at different stages of development as do some modern viviparous reptiles.
Alternatively, the smaller skeletons may be of young that died during development
and became trapped. Some modern animals abort dead embryos only after the
birth of their surviving siblings.

Secondly, why do the majority of specimens carry much fewer than the
maximum number of young – an average of about 3 instead of 10 or 11? One
possibility is that at the time of death most ichthyosaurs had already given
birth to some young. Another is that a high proportion of embryos vanished
during death, or afterwards during preservation and fossilisation.

The best evidence for an ingested juvenile comes not from Holzmaden
but from southern England. An ichthyosaur fossil found in Somerset in 1846
contains a tiny skull covered by the contents of the creature’s stomach.
Whether the skull lies in the stomach itself or the uterus, however, is
unclear: the fossil may thus be an accident of preservation. A similar fossil
was apparently found in Holzmaden in the 1840s, but it has since been lost.

A more recent finding from Somerset, dated at about 200 million years
ago, offers little comfort to the cannibalism camp. The fossil shows a tiny,
curled fetus just behind the cloacal opening; it was probably expelled by
decomposition gases. No other young are present, suggesting that the rest
of the litter vanished during decay.

Given the lack of hard evidence, why has the cannibalism theory proved
so resilient? Naturalists have always been rather zealous in looking for
signs of cannibalism among ‘lower organisms’, whether they be worms, sharks,
reptiles or cavemen. This tendency has exaggerated the significance of what
little evidence there is. Studies show that modern reptiles do indulge in
cannibalism, but not nearly to the extent implied by some naturalists’ tales.

One alleged offender is the world’s largest lizard, the Komodo monitor
(Varanus komodoensis). There have been many reports of the Komodo monitor
digging up nests of eggs and eating hatchlings and juveniles of its own
species. Yet, on closer inspection, the true extent of these activities
appears quite limited. Some research suggests that 8 per cent of Komodo
monitor droppings carry remnants of cannibalistic meals, but most put the
figure at less than 2 per cent.

The alligator, though not completely innocent (dissected adults occasionally
contain the remains of young alligators) has suffered similarly. Female
alligators, for instance, assist their young out of the nest and into the
water, crushing and swallowing rotten or infertile eggs while doing so.
Predictably, casual observers have tended to interpret such behaviour as
evidence of cannibalism. Even more unjust, female crocodiles often carry
hatchlings, in their mouths, to the water for a first swim: a caring act,
but one often misconstrued as a mother eating her own offspring.

The moralistic view of natural history prevalent in Victorian times
no doubt encouraged such misjudgments. But at least for the ichthyosaurs,
the spectre of cannibalism has now been laid to rest. With the acceptance
of Bottcher’s ideas about the trapped skeletons, researchers can at last
begin to use the many excellent specimens to build up a fuller picture of
the developmental biology of ichthyosaurs. The embryos may even shed light
on the vexed question of the origin of the ichthyosaurs, and their acquisition
of such advanced aquatic adaptations from a fully terrestrial ancestry.

Michael Benton is at the Department of Geology at the University of
Bristol. His latest book is Vertebrate Palaeontology (Chapman & Hall).

* * *

Caring, sharing dinosaurs bury old prejudices

The myth of the wicked reptile is dealt a further blow by palaeontologists
working in the US and in Mongolia who have found evidence of caring group
behaviour among dinosaurs.

John Horner, of Montana State University, has been excavating dinosaur
nesting sites in Montana for nearly 15 years. He discovered early on that
these dinosaurs, herbivorous hadrosaurs and hypsilophodontids, laid their
eggs in batches of 20 to 30 in large circular nests which they scraped out
in the sand. The female, it seems, scooped a hollow about 1 metre wide and
laid her eggs in concentric circles. She then covered them with sand and
sometimes with leaves and twigs: as with some modern crocodilians, the rotting
vegetation would have helped to warm up the nest and incubate the eggs.

Horner found that the Montana dinosaurs nested communally – like many
modern birds – spacing their nests out so that no mother dinosaur could
quite reach the neighbouring ones. By digging down beneath the nests, he
found successive layers of buried rock with nests, suggesting that the dinosaurs
returned to the same nesting site year after year.

Many of the nests still contain eggs, some of which are unhatched. One
hypsilophodontid egg contains a tiny embryo which would have been just 20
centimetres long, compared to an adult length of 2.5 metres. Horner argues
that these tiny hatchlings were too small and poorly developed to move far
from the nest, and that their parents must therefore have fed and protected
them.

Further evidence that dinosaurs reared their young comes from skeletons
of adult and juvenile hadrosaurs, of the genus Maiasaura (‘good mother reptile’),
which Horner found near the nests. The juveniles were 1 metre long and had
worn teeth – evidence, says Horner, that they were probably one-year-old
siblings who had come back to the nesting ground to help their parents with
the next generation. These findings point to complex family group behaviour
in dinosaurs.

Recent work on Mongolian dinosaur nests has both confirmed some of Horner’s
findings and added new details. The first known dinosaur nests were discovered
in Mongolia in the 1920s by an expedition from the American Museum of Natural
History. Yet very little detailed work had been done on these finds until
Konstantin Mikhailov, of the Palaeontological Institute in Moscow, and Karov
Sabath, of the Palaeontological Institute in Warsaw, began their studies.

The two researchers have now identi-fied eggs and nests produced by
all the major dinosaur groups, including the large herbivorous sauropods,
the herbivorous horned ceratopians, and the carnivorous theropods. At one
of their Mongolian sites, they found nests in the sand made by two or three
different kinds of dinosaur, as well as one kind of bird, all on an ancient
beach beside a lake. Mikhailov’s findings support the idea that dinosaurs
nested communally and returned to the same nesting grounds each year.

Mikhailov made an even more startling discovery in one theropod nest
in Mongolia. Among the shell fragments were numerous bones of herbivorous
ceratopians and other small reptiles. The conditions at the site showed
that these small bones had not been washed in, nor were the ceratopians,
being herbivores, feeding on the eggs. Mikhailov’s interpretation is that
the theropods had stocked the nest with small prey as food for their ravenous
hatchlings. But this remains controversial because of the methods used to
identify the eggs (This Week, 7 September).