Wednesday, February 10, 2010

Rotting the tree of life

There are good papers, great papers, and those clever little papers that make you say "I wish I'd thought of that!". Before I get to that, a little preamble:

Taphonomy is the branch of research that is interested in describing what happens to an organism between dying and ending up as a fossil (or even why it won't end up as a fossil). A lot can happen to an organism in that period of time, as the earth is a dynamic spheroid. The older a fossil, the more possible disturbances it can experience. Taphonomy can tell us a lot about the environment an organism was deposited in and it can provide important controls on the inferences we make about the environment we think a fossil organism once lived in. But taphonomy is also an important consideration in considering what an organism is. That is, the 'life' of a fossil after death, might have a profound impact on how we place that fossil in the tree of life.

Enter the experiments of Rob Sansom and colleague's experiments on lamprey larvae and the title organism of this blog, reported in this week's issue of Nature. Sansom et al. wanted to examine what happens to 'primitive' vertebrates that lack hard, mineralized tissues, the type of tissues that normally fossilize. I say "normally", because there are some 'abnormal' cases in which soft-bodied creatures with no bones, teeth, or hard cuticles actually form as fossils. Some such fossils have played an important role in understanding the timing and early origin of vertebrate animals.

For instance, this species known as Yunnanozoon (Chen et al. 1999) from the Cambrian of China. It represents one of the earliest known vertebrates or vertebrate-like forms.

Yunnanozoon is remarkably well preserved, but other Cambrian chordates can be even more incomplete. The problem with such fossils is that they're difficult to interpret because they're squished, and they're made of soft parts. We have no idea how much they might have decayed, apart from the fact that they seem to be an exception to the rule that soft parts don't fossilize. This usually implies some sort of exceptional conditions favouring preservation, but doesn't necessarily rule out decay or other types of disruption.

Sansom et al. let larval lamprey and lancelets rot in buckets of sea water and recorded the progress of the decay over the period of several months.

The impressive and startling results of watching fish decay are below the fold:

As the animals rotted away, Sansom et al. recorded details of their anatomy. Not just general features, but the types of characters that would be used to score an organism for a phylogenetic analysis. These include classically important features, like the gill filaments, cartilages of the gill arches, the type of heart, the shape of the body muscles, the dorsal rod known as a notochord, and so on. These are characters that have normally played a significant role in establishing the relationships of vertebrates and their nearest non-vertebrate relatives, such as the lancelet.

What this figure shows is the length of time each character survived as the animal rotted. What's striking is that the characters that lasted longer all tend to be characters that we consider phylogenetically more primitive. Characters such as a notochord and segmented axial musculature are all considered to be primitive features shared by the last common ancestor of lancelets and lamprey. On the other hand, features such as eyes, or a chambered heart are more derived features found in modern vertebrates.

This figure shows nicely how the decay features plot out in phylogenetic history. If you go back to figure above, there is a graph showing the relationship between phylogenetic rank and decay stage.

What we see is that the level of decay would lead one to think that the taxon was signicantly more distantly related to the vertebrates, much like the early chordates we find in the Cambrian.

Not only do these results provide a caution against how we interpret soft-bodied Cambrian chordates, but it illustrates a framework for studying the phylogenetic effects of decay. As decay is studied across a wider phylogenetic scope, the more we can determine about the generality of these types of patterns. That will have a profound effect on how we study and interpret the exceptional cases of soft-tissue preservation in fossils.

Chen, Y.-J., Huang, D.-Y., and Li, C.-W. 1999. An early Cambrian craniate-like chordate. Nature 402:518-522 link

Sansom, R.S., Gabbott, S.E., and Purnell. M.A.2010. Non-random decay of chordate characters causes bias in fossil interpretation. Nature 463:797-800 link

Briggs, D.E.G. 2010. Palaeontology: Decay distorts ancestry. Nature 463:741-743 link

Read full post

Thursday, January 07, 2010

Early tetrapod footprints from Poland: what changes and what doesn't.

A few years ago, I wrote a post about prediction in the historical sciences. This post came in anticipation of the publication of Tiktaalik roseae, a remarkably well-preserved fossil of an aquatic animal very closely related to the first tetrapods. I offered this figure to illustrate the position of a gap in the fossil record as a predicate for how/where we might look for fossils that would fit within those 'gaps' in phylogeny.

Yesterday, this picture got a lot more complicated. With the publication of tetrapod footprints some 18 million years in advance of this gap by Nied┼║wiedzki et al., the nice congruence between the node order of the tree and stratigraphy no longer appears so nice.

Here's a picture of the trackways (click for the full view):

[More below the fold]

This news has already broken, and I'm usually the last person to blog about it these days. Naturally, Ed Young has a good take on it. As noted by Jenny Clack in Ed's piece, the individual footprints are more convincing than the trackways in terms of their identity. Here's one from the supplementary information file (which is free to access).

[Picture deleted: I made a mistake, reading too fast and posted an image of Triassic temnospondyl footprints! A better image is this digital surface rendering from the main paper posted below]

But how "bad" is the record now that we know this? Well, bear in mind that lungfishes and tetrapods are considered (at least by the phylogenies in question) to be sister groups. That is, they split from their last common ancestor at exactly the same time. The earliest 'true' lungfishes (i.e. lungfishes with toothplates, and upper jaws fused to the braincase, etc.) are Emsian in age, roughly the same age as these new footprints. Recognizable close relatives of lungfishes, such as Diabolepis, Youngolepis and Powichthys, however, are considerably older. Do, we can set a minimum age for the origin of the tetrapod stem lineage down in very earliest part of the Devonian even if we do not have the fossils.

So, not surprisingly, the record is pretty bad. It's certainly 'gappy'. But we should not be too surprised that tetrapods emerged long before the time when we find their first actual fossils.

Read full post

Wednesday, January 06, 2010

MAJOR news about early tetrapods today.

Okay, I've exaggerated by putting MAJOR in all-caps, but here it is:

Tetrapod trackways from the early Middle Devonian period of Poland

The fossil record of the earliest tetrapods (vertebrates with limbs rather than paired fins) consists of body fossils and trackways. The earliest body fossils of tetrapods date to the Late Devonian period (late Frasnian stage) and are preceded by transitional elpistostegids such as Panderichthys and Tiktaalik that still have paired fins. Claims of tetrapod trackways predating these body fossils have remained controversial with regard to both age and the identity of the track makers. Here we present well-preserved and securely dated tetrapod tracks from Polish marine tidal flat sediments of early Middle Devonian (Eifelian stage) age that are approximately 18 million years older than the earliest tetrapod body fossils and 10 million years earlier than the oldest elpistostegids. They force a radical reassessment of the timing, ecology and environmental setting of the fish–tetrapod transition, as well as the completeness of the body fossil record.

I'll update with a detailed explanation soon.
Read full post