Are There Disagreements Between the Fossil Record and Molecular Data?
Molecular biologists have a tradition of reworking a lot of the evolutionary relationships and timescales that morphologists and paleontologists worked so hard to figure out. This can really piss off the non-molecular folks, but I prefer to think of it as a cooperative relationship. The molecular clock, for example, would not be possible without calibration from the fossil record. It is important to note that molecular and morphological data tell two different stories, which I outline below the fold . . .
When I wrote that the Cambrian Explosion never happened, I did not mean that the Cambrian Explosion never happened -- I meant, the Cambrian Explosion never happened. Confused? Well, it all depends on how we define the Cambrian Explosion. My preferred definition of the Cambrian Explosion is "The sudden appearance of many animal body plans in the fossil record." (Of, course, by sudden we mean over the span of millions of years.) If we use that definition, then the Cambrian Explosion is a reality -- there are some forms that did fossilize prior to the explosion. But this definition is unsatisfying because it only tells us about the fossil record, not the underlying evolutionary process.
The Cambrian Explosion is just one example of an apparent disagreement between molecular and morphological data (and I'll get back to it in a moment). Before we return to it, we must outline what the fossil record and what the molecular data tell us. The fossil record shows a preserved body plan -- in the case of earliest body plans, it shows the first discovered body plan. The molecular data, on the other hand, reveals the relationship of evolutionary lineages. It is possible to reconstruct evolutionary relationships with morphological data (both extant morphologies and fossilized ones), but the molecular data does this far more objectively. The molecular data does not contain direct information regarding the actual time two lineages diverged (the molecular clock must be calibrated with the fossil record), nor does it inform us about the morphologies of the organisms in a particular lineage.
Even a perfect fossil record will give us different information than molecular data. The fossil record documents the evolution of morphologies; the molecular data shows evolutionary relationships. Getting back to the Cambrian Explosion, it may be possible that both the fossil record and molecular data are correct. Many of those body plans may have first emerged during Cambrian. The molecular data tells us that many of the lineages that gave rise to those body plans were around prior to the Cambrian, but it says nothing about what the animals looked like prior to the Cambrian. Those lineages may have been present but in their ancestral body plans -- some worm-like animal. In that case, the Cambrian Explosion is a rapid diversification of body plans, but not a rapid evolutionary radiation.
Another example of disagreements between morphological and molecular data is the radiation of mammals. The fossil record reveals many new mammalian body plans following the extinction of the dinosaurs. The molecular data, however, suggests that the lineages that gave rise to the different mammalian body plans were present prior to the phenotypic diversification. Both the Cambrian Explosion and mammalian radiation are examples of animals exploiting new niches that appeared after a major change in the earth's climate. In the case of the Cambrian Explosion, a snowball earth may have predated the appearance of new body plans. When the earth thawed, new niches appeared, and the drastic morphological changes occurred. The lineages that gave rise to those new morphologies were just "chilling" in the ocean. In the case of the mammals, they were restricted to nocturnal critters while the dinosaurs filled all of the day-time niches. After the catastrophic extinction of the dinosaurs, all of those mammalian lineages that were little shrew-like critters during the reign of the dinosaurs, evolved so as to take advantage of the newly vacated niches.
This logic can be extended to any other disagreement between molecular and morphological data. It is important to realize that these are not conflicting sources of information, but complementary. The fossil record informs regarding morphologies and absolute dates. The molecular data reveals evolutionary relationships, but it is restricted to evolutionary lineages, not morphology or absolute time. If we combine the fossil record and molecular data with geological information regarding major environmental changes, we can gain a clearer picture of historical evolution.
I would add that molecular clock estimates have very large error bars molecules are better for determining the relative order of branching events in an evolutionary tree than they are for determining absolute dates, even with geology-based calibration. Some discrepancies between molecular data and fossil data could be due to this (one should never confuse molecular clocks with atomic clocks despite their similar sci-fi names, if you set your watch to a molecular clock, you might miss your appointment by a few million years).
In addition, I think your point could also be stated this way: sequence divergence and taxonomic divergence are two distinct events, so of course they might happen at different times. Perhaps 80 million years ago the ancestor of humans had already diverged from the ancestor of bats, but if some ceratopsian taxonomist were to study them at that time, unaware of what the future would bring, the two mammals would probably be assigned to the same order. Only later would the bat order diverge from the primate order, in the sense that a taxonomist of the time would recognize the distinction (illustrating that higher taxonomic categories are somewhat contrived, subjective things).
Yes, molecules are excellent for determining branching orders, and molecular clocks do have large error bars.
I also agree that taxonomy (above the species level) is contrived. I think the evolutionary lineages are most important (who cares if our ancestors would have been classified as primates prior to the extinction of dinosaurs?). Taxonomy only becomes more absurd when it includes extinct species (in this case, even species level taxonomy is contrived).
My argument stands, however, whether the molecular clock works or not. We obviously have lineages that lead to certain types of organisms prior to the morphology evolving. I think the focus should be shifted from discrepancies between the fossil record and molecular data, to the different stories the types of data tell.