Featured in the newest Dialogue Magazine »
Upgrading Dinosaur Diets

Upgrading Dinosaur Diets


Have you ever tried to imagine what a world without grasses would look like? The scenery might well be pretty bleak. Scientists have long declared that there were no grasses present in dinosaur communities. The plant-eating dinosaurs had to make do, we have been assured, with cycads, ferns, horsetails and trees of flowering plants such as palms and magnolias.

Today however, grasses (which are also flowering plants), actually contribute far more than beauty to the environment. They represent extremely important sources of nutrition. Indeed, our whole ecosystem could well crash if grasses disappeared.

This plant group includes barley, oats, fescue and timothy grasses, rice, millet, sorghum, sedges, bamboo and corn among others. What large grazing animal does not depend upon one or other of these sources of nutrition? Nevertheless, scientists have steadfastly maintained that large grazing dinosaurs survived without grasses. This view will certainly be reevaluated in the light of an article on dinosaur diets which appeared in the journal Science in November 2005. In retrospect, scientists may well decide that it makes sense that dinosaurs ate grass. Many grasses grow continuously when they are grazed. Thus they provide a reliable source of food. Ferns and cycads on the other hand, restrict growth to once per year. Cycads, which look like squashed down palms, have indigestible leaves much like those of conifers. Cycad seeds could well be eaten, but some contain compounds that are poisonous, at least to people. The stems might be good, but the plant might not survive to grow next year. Altogether, without grasses, the vegetation would be hard pressed to support gigantic grazing dinosaurs.

The amazing teeth of the duckbill dinosaurs suggest that their lifestyles were similar to ungulate grazers (deer, bison, gazelles) that we know today. There were also mammals present, now extinct, which had teeth shaped much like today’s grazers. It certainly makes sense that the teeth of those gondwanatherians were used for eating grass. In hindsight it all seems so obvious. Why did it take so long for scientists to recognize the presence of grasses in these communities? Plant spores and pollen are very resistant to decay. Why was grass pollen not identified in sediments with the dinosaurs? That is a very good question.

There are few issues in palaeontology as controversial as the study of fossil pollen grains. It is the case that non seed producing land plants reproduce by means of spores, while seed plants release pollen as part of the reproductive process. The outer coats of both spores and pollen grains are made of an exceedingly tough organic compound called sporopollenin.

It so happens also that these outer coverings are characteristically shaped and decorated in ways which are diagnostic of the plants which released them. The main criterion distinguishing the non-seed plants – spores from the pollen of seed plants, is size. Also the pollen grains of flowering plants show lots of variation.

Because the sporopollenin is so resistant, these small artifacts can be concentrated from sedimentary rock samples by treatment with strong acids or alkalis which dissolve away everything else. Obviously this technique provides a wonderful way to establish the presence of plant material in sediments which lack large plant fossils.

One of the early companies to exploit this technique was Royal Dutch Shell in the 1930s in Venezuela. They used the presence of specific pollen types to establish the sequence of rock layers and thus to locate oil bearing rocks. Apparently this technique was highly successful for Shell.

None of this, of course, is controversial. What is controversial is the suggestion that these pollen grains came from plants like our modern ones. Based on evolutionary assumptions, most scientists refuse to acknowledge that deep sedimentary layers might contain remains like those of modern plants.

Palynologists (spore/pollen grain experts) routinely recognize pollen grains of extant genera and species, if these are found in recent sediments. When it comes to the sediments which buried the dinosaurs, and also even lower sedimentary layers, it is a widely accepted convention that the pollen grains will be identified by appearance only. Any resemblance to modern plants is entirely ignored. Specialists who had previously identified such pollen collections in terms of modern plant names, later saw their conclusions discounted by others, and names based on form only were substituted for the modern names.

The problem of identifying the presence of grass plants in dinosaur communities now becomes clearer. The sediments might contain ample grass pollen, but nobody would identify it as such because of the custom of describing appearance only. At the higher levels where pollen grains are compared with living plants, grasses have variously been said to have first appeared in Oligocene rocks (about 35 million years ago by evolutionary estimates) or Paleocene rocks (about 55 million years ago by similar reckoning). According to such evolutionary scenarios, this would place the first appearance of grasses ten to twenty million years after the extinction of the dinosaurs.

It was the identification, in dinosaur dung, of tiny silica crystals produced only by grasses, that led to the dramatic announcement concerning grasses in dinosaur diets.

These crystals, called phytoliths, form in the thickened walls of certain plant cells. These artifacts are typically destroyed by the strong acids used to concentrate pollen grains. Thus phytoliths would never be found along with collections of pollen grains. The shapes of these crystals are so specialized, that scientists can identify the general characteristics of the plants which produced them.

Recently five very different types of grass phytolith were found in dinosaur dung. Some of the phytoliths resemble those produced by modern rice. Some of the plants may have looked like bamboo, only smaller. All we know to this point is that there was considerable variety in the characteristics of the grasses. Evidently these plants were an established part of the ecosystem.

Thus we return to the question as to why it took the scientists so long to document the presence of grasses with the dinosaurs. It is evident that the scientists were not looking for grasses at these deep levels in the rocks. The pollen grains which could have been assigned to the grasses, were instead described merely in terms of appearance. Phytoliths did not show up because nobody was looking for them. The preparation of pollen grains, in fact, eliminated phytoliths. The fact is, this recent discovery complicates theories about the speed of plant evolution. Indeed this is a classic example of the way in which evolutionary preconceptions can bias discovery. It is always good to be a little skeptical about conclusions which scientists draw from fossils.

April 2006

Subscribe to Dialogue