November 6, 2001

Kidwell shows fossil record works

The history of life became a little clearer this week. A paper published in the current issue of Science by the University's Susan Kidwell has helped show that the marine fossil record, a vital source of data for scientists studying the early history of life on the planet, is reliable.

"The fidelity of fossil assemblages to their source communities has haunted paleontologists for decades," said Kidwell, professor in the department of geophysical sciences and the College and the committee on evolutionary biology, in the Science paper.

The reliability question is a serious one: the fossils preserved in marine sediments around the world are used to track species abundance over time, allowing ecologists to better understand the factors that lead to the success or extinction of a species. Species abundance also hints at previous climactic conditions, the positions of continents, and a great many other pressing geological and biological questions.

The problem, though, is that no one has been sure whether the species preserved in the fossil record provide an accurate measure of the species that were alive in the ecosystem. If some species fossilize more readily than others, the fossil record would give a distorted picture of the true abundances of each sort of organism in an ecosystem, making research on it less reliable.

Kidwell's paper has helped shore up research in this crucial area. While earlier studies had been done to compare living communities to the remains they left behind, they took place under so many different conditions and in so many different environments that it was difficult to know whether the fossil record was generally reliable.

"There has been so much variation in methods used by different workers and in the particular environments studied that it was basically impossible to see what the overall 'answer' in terms of quality really was," Kidwell said.

"That is, how much variation in results among studies was because of differences in methods, and how much due to actual environmental differences in post-mortem bias."

Kidwell's paper helps bridge that gap through a statistical technique known as meta-analysis, in which the raw data used in previous studies is analyzed together while controlling for potentially distracting local effects.

Looking for studies that compared the relative abundances of mollusks in a living ecosystem to their shells dredged from ocean sediment, Kidwell gathered data from the past decade to use in her analysis. When the data were brought together, the overall pattern was quite clear: species abundance in the living systems matched the abundance of shells in the sediment quite closely. The marine fossil record held.

There were some holes, however. Some studies showed a much poorer correlation between living and dead groups than others. After analyzing the data in more detail, Kidwell tied this mismatch to the size of the mesh the experimenters used to dredge the ocean sediments. "When I tested for live-dead agreement in abundance, I did indeed find a huge scatter, verifying original authors' impressions, but that's when I discovered the mesh-size effect," Kidwell said.

Meshes with smaller holes, which pick up smaller shells, tended to yield relative species abundances in the shells that were quite different from those in the living communities. Larger meshes, which let the smallest shells slip through, showed a much closer match between the shells and their live counterparts.

Kidwell suspects this effect is due to the small size of young mollusks. Juvenile populations are ecologically unstable because the young are so vulnerable to predators and changing environments. Fluctuating constantly in living systems, their living numbers at any given moment are unlikely to match the number of the dead.

When the mesh-size effect is accounted for, a strong correlation remains between the living ecosystem and the shells that remain. That correlation provides the essential support that researchers interested in past climates and ecosystems need to be sure that the marine fossil record will give them a clear view of the past worlds they are studying.

Kidwell's conclusions remain to be tested for other sorts of ecosystems and to be extended by the addition of more data. She continues to look for new studies to add to her meta-analysis and to consider other aspects of the marine fossil record.

"For a while [my work will be] a bit more biologically oriented than my usual field-based work on the rock record," Kidwell said. "That's so that I can make these analyses as relevant as possible to applying fossil/dead data to issues of modern environmental and biosphere change."