Are fossils essential to understanding evolution?

In a very interesting post about agamids and chameleons at Tetrapod Zoology, my fellow ScienceBlogger Darren states the following;

One of the greatest fallacies held about evolutionary theory is that fossils are essential in demonstrating the existence of change (don't believe me? Look at 'creation science' books like Duane Gish's Evolution: the Challenge of the Fossil Record and Evolution: the Fossils Say No!). Of course fossils do indeed show how characters were accrued and modified over time, and it's that 'time' aspect of the data that they shed crucial information on. But we most certainly do not need fossils to demonstrate the fact of evolution, as we are surrounded by evolutionary intermediates right here in the modern world. In fact, if we didn't have any fossils at all we would still conclude - from the living organisms that surround us - that evolution happens... - [emphasis mine]

What if there was no fossil record? Would our understanding of evolution be the same? It is impossible to know for sure, but such questions bring up some important points about the way we have come to understand the details of the evolution of life on earth.

When Darwin wrote On the Origin of Species by Means of Natural Selection paleontology presented more problems than support. There was a general feeling that the record of ancient life was fairly complete, or at least becoming so, and outside of a few controversial transitional forms (like Archaeopteryx) the fossil record did not produce the evolutionary gradations Darwin's theory predicted. Different naturalists dealt with this problem in different ways. Huxley, until he changed his views in the 1870's, maintained that much of evolution occurred during "non-geologic time," or at a time so distant that the strata that might have preserved the fossils was missing. He later changed his views so that evolution could be observed in the fossil record, although Huxley was often cautious to say that certain fossils represented a "type" in a line of descent rather than a direct ancestor or descendant.

Darwin, on the other hand, argued that the fossil record was so imperfect that the transitions might not be preserved at all, although he realized that the rejection of this notion would mean the rejection of his entire theory. If his mechanism of natural selection was correct (if the evolutionary forces seen acting today were acting in the past, meaning that all organisms shared a common ancestry) then paleontology should have produced evidence to confirm his hypothesis. The fact that it did not when Darwin wrote his book was therefore a significant obstacle.

The best positive case that Darwin could make using the fossil record was that what his theory predicted was consistent with the succession of different forms of life over time. The succession of different "ages" had been recognized for quite some time (but not generally used as evidence of evolution), and the "big picture" of the history of life fit nicely with common descent. Eventually the history that evolution by natural selection would be confirmed by fossil discoveries made after Darwin published, particularly of fossil horses. Gaudry and Kowalevsky, for example, began to put together a series of fossil horses that showed the evolution of smaller, three-toed forms to the modern one-toed horse. Eventually O.C. Marsh's phylogeny of North American fossil horses would bury the Old World model, but generally speaking the evidence for the gradual evolution of the horse was among the best fossil evidence for Darwin's theory of evolution.

This did not mean that everyone accepted natural selection as the main evolutionary mechanism, however. Although Darwin's work was admired for the detail and force of argument not everyone could agree that natural selection was the primary mechanism for evolution, particularly since it was seen as being so violent. Neo-Lamarckism was a competing school of thought during this time, and it made some people more comfortable in that it was seen as more "positive" and made evolution easier to reconcile with religion. Even into the early 20th century the place of natural selection as the primary evolutionary mechanism was unsure, at least until it was shown to be the central concept during the formation of the Modern Synthesis.

Although G.G. Simpson made an important contribution to the Modern Synthesis in the form of his book Tempo and Mode in Evolution, the work of geneticists is more often cited and remembered for giving natural selection pride of place in evolutionary studies. Unfortunately paleontology was seen as recording the history of life and doing little else, the discovery of the way in which evolution actually worked being up to geneticists and population biologists. While it certainly is true that understanding evolution requires the "present being a key to the past" in terms of evolutionary mechanisms, the history of life is extremely difficult to understand without fossils.

The search for the origins of whales provides a good example. For about 80 years the Eocene fossil whales Basilosaurus, Dorudon, and Protocetus were the oldest representatives of whales. Even beyond the fragmentary nature of these creatures, all of them were already aquatically adapted and were difficult to connect to terrestrial ancestors or the two great groups of living whales, the toothed (Odontoceti) and baleen whales (Mysticeti). The teeth of the earliest whales suggested that they had evolved from terrestrial carnivorous mammals, but from what kind was difficult to discern (the favored groups went from the Carnivora to creodonts to mesonychids). Indeed, nearly 100 years after Huxley proposed that whales had evolved from terrestrial carnivores in 1870, some researchers were considering the Odontoceti and Mysticeti to by diphyletic due to their differences and the general lack of fossil forms connecting them. That whales had evolved was not in question, but the details of their evolution were frustratingly absent.

Fortunately there has been a revolution in the study of whale origins since the description of Pakicetus in 1981, the dry times of earlier decades giving way to a flood of new fossil whales. There are still many questions that need to be resolved, but as the gap in early whale evolution has been filled in more and more hypotheses have been able to be tested and revised. Although it is a common claim of critics that paleontology is purely observational and in this way distinct from "hard" sciences like physics, the history of discovery involving whales (and the relationship of paleontology to evolution in general) shows that predictions can be made and hypotheses tested by looking at the fossil record. If we could only look at living organisms it would be difficult to avoid the conclusion that evolution occurs, but even if we were able to figure out how it occurred the big question of the origins of the diversity we see today would remain obscure.

Evolution is such a powerful notion because it combines the discoveries of many disciplines, lines of evidence intertwining with each other as new ideas are formulated and tested. There is controversy, of course, but the recent integrations of studies of genetics, development, and paleontology (i.e. to discover the date of divergence of our ancestors from chimpanzee ancestors and the relationship of whales to artiodactyls) reveal that understanding evolution requires an interdisciplinary approach. Each discipline has different strengths and weaknesses, but together they shed light on the minute details of the history of life. So yes, if there was no fossil record natural selection may have still been discovered and evolution observed, but I am certainly glad that we are not bounded by such a narrowed view of natural history!