A 62-million-year cycle in biodiversity emerges from scrutiny of a marine-fossil database, but its causes remain mysterious. Thus, this discovery is likely to provoke a flurry of theoretical speculation.
For decades, palaeobiologists have been finding large-scale cycles and patterns in fossil records, and one might assume that there was nothing substantial left to be discovered. Not by a country mile! On page 208 of this issue1, Rohde and Muller demonstrate a 62-million-year cycle in fossil biodiversity during the Phanerozoic, the past 500-plus million years of life on Earth. This 62-million-year wave is surprisingly strong and — so far — there's no good explanation for it.
In some fields of science, particular data sets achieve iconic status, focusing the attention of whole communities of researchers as they try to understand the patterns they contain. For the past two decades, many palaeobiologists have been exploring the marine-fossil databases created by the late Jack Sepkoski2,3. Sepkoski spent years combing through the palaeontological literature, compiling the first and last known occurrences of every animal genus in the marine-fossil record, more than 30,000 in all. He used genera because, being the taxonomic level above species, they are revised less often by palaeontologists and their numbers are more manageable. And he dealt exclusively with marine animals, because their fossil record is longer and better preserved than that of their terrestrial counterparts. Sepkoski's labours produced a detailed history of biodiversity through time, which in turn has generated a kaleidoscope of analyses4,5,6,7,8,9 aimed at both detecting its patterns and explaining them.
Over time, these analyses have explored increasingly subtle features of the fossil record (Box 1), but Rohde and Muller's discovery shows that there are still broad patterns left to be found. They have taken Sepkoski's fossil diversity curve — a graph of the total number of marine-fossil genera at each point in the past 542 million years — and subtracted a smooth mathematical function that accounts for the broadest trends of diversity through time. What is left is a diversity curve that rises and falls somewhat irregularly, but roughly in step with a 62-million-year cycle. Rohde and Muller use spectral analysis to show that these ups and downs oscillate much more rhythmically than one would expect by chance.
The 62-million-year wave is too big to ignore, with biodiversity growing and shrinking by several hundred genera between the peaks and the troughs. The cycle is visually obvious, and some palaeobiologists will wonder why they haven't seen it before. In fact, a similar long wave in diversity was suggested almost 30 years ago in the pages of this journal10, but only with the most recent refinements to the geological timescale can the cycle in fossil diversity be shown to be too regular to have arisen at random.
Rohde and Muller show that the 62-million-year cycle is particularly strong when one considers only short-lived genera, which they define as those that survive less than 45 million years. This seems to make perfect sense; longer-lived genera are thought to survive longer because they are more diverse and more widespread, and thus more resistant to disturbance. But it is also almost true by definition, because it would be hard to construct a wave from components that endure for more than half the wavelength, and that would therefore bridge the peaks and troughs.
Two other aspects of Rohde and Muller's discovery may also attract sceptical attention. First, the 62-million-year cycle is largely absent from the past 150 million years, which is precisely the part of the fossil record that is most accurately known. Second, many will have nagging suspicions that the cycle is somehow a statistical fluke, because some apparent patterns in the fossil record have evaporated when subjected to statistical scrutiny. To their credit, Rohde and Muller have tested the 62-million-year cycle in several different ways, and in each case it seems to be statistically significant.
If the 62-million-year cycle is real, it demands an explanation, and the search for one will be interesting. Rohde and Muller are physicists, and they examined 14 possible geophysical and astronomical drivers for the cycle. Several of these are at best highly speculative, including a companion star to the Sun and a hypothesized large planet for which no evidence exists. But in any case, all of these candidate driving mechanisms seem to be insufficient to cause the 62-million-year cycle.
As other disciplines enter the fray, the range of possible explanations will grow. Earth scientists and palaeontologists will point out that marine-fossil diversity depends on the diversity of marine habitats, and thus on the size and configuration of the continental shelves. They will therefore ask whether the 62-million-year cycle could potentially reflect changes in the continental margins through time, as sea level fluctuates and the continents rearrange themselves. Others will observe that this long wave in biodiversity is broadly consistent with the reported phase shift between fluctuations in the rate of extinction of existing organisms and the diversification of new ones7, and will search for a theory that unites these observations. Theoretical biologists will also note that global biodiversity is a tapestry that weaves itself, so the 62-million-year cycle in fossil diversity need not be generated by similar cycles in external driving factors. Instead, biodiversity could swing like a pendulum, with a rhythmic cycle that is governed by its own internal dynamics rather than by rhythmic external forcing.
But if the 62-million-year cycle is caused by a biological pendulum, it swings so slowly that it will be challenging to discern the underlying mechanisms. By any biological yardstick, 62 million years is a very long time; 62 million years ago last Tuesday, we mammals had only recently embarked on our striking morphological diversification following the mass extinction at the end of the Cretaceous. Clever modellers should have little difficulty creating biological models that exhibit very long oscillations in biodiversity. But the hard work will lie in showing that the premises behind these models are themselves accurate, or at least plausible.
It is often said that the best discoveries in science are those that raise more questions than they answer, and that is certainly the case here. Let the theorizing begin.
Rohde, R. A. & Muller, R. A. Nature 434, 208–210 (2005).
Sepkoski, J. J. Jr Paleobiology 19, 43–51 (1993).
Sepkoski, J. J. Jr in A Compendium of Fossil Marine Animal Genera (eds Jablonski, D. & Foote, M.) Bull. Am. Paleontol. No. 363 (Paleontological Research Institution, Ithaca, 2002).
Raup, D. M. & Sepkoski, J. J. Jr Science 231, 833–836 (1986).
Miller, A. I. Science 281, 1157–1160 (1998).
Jablonski, D. Science 284, 2114–2116 (1999).
Kirchner, J. W. & Weil, A. Nature 404, 177–180 (2000).
Peters, S. E. & Foote, M. Nature 416, 420–424 (2002).
Miller, A. I. & Foote, M. Science 302, 1030–1032 (2003).
Thomson, K. S. Nature 261, 578–580 (1976).
Miller, A. I. in Deep Time: Paleobiology's Perspective (eds Erwin, D. H. & Wing, S. L.) 53–73 (Univ. Chicago Press, 2000).
Sepkoski, J. J. Jr Milwaukee Public Mus. Contrib. Biol. Geol. 51, 1–125 (1982).
About this article
Paleontological Journal (2012)