Marine Biodiversity Dynamics over Deep Time

See allHide authors and affiliations

Science  03 Sep 2010:
Vol. 329, Issue 5996, pp. 1156-1157
DOI: 10.1126/science.1194924

The fossil record provides our only direct window into how Earth's biodiversity has changed over the past 530 million years. Efforts to dissect diversity dynamics, however, have been impeded by gaps in the fossil record and uneven sampling by paleontologists. As a result, two of the most fundamental questions in paleobiology remain contentious. How, in detail, has diversity changed over geologic time? And what is the nature of the rules that governed these diversity changes? More specifically, was the diversification of most groups of organisms unconstrained, or was it subject to some sort of limit? On page 1191 of this issue, Alroy (1) presents a new analysis of the marine fossil record that goes to the heart of these questions. The results are unlikely to be universally accepted, but they help to pinpoint the critical issues.

Diverse views of diversity.

(A) Sepkoski's marine animal diversity curve has long been the standard, but lacks correction for uneven temporal or spatial sampling. (B) A more recent diversity curve based on the Paleobiology Database (PBDB) attempted to correct for uneven sampling and offers a different view of diversity dynamics (6). (C) The latest diversity curve from the PBDB uses a new method for correcting for uneven sampling (1) and, for the Cenozoic (yellow), is strikingly similar to Sepkoski's. The challenge now is to reach agreement on the methods for correcting for uneven sampling and for rigorously demonstrating that the PBDB provides a balanced reflection of the fossil record.

Alroy is not the first to attempt to graph the historic ebb and flow of marine biodiversity. One early effort came in 1860 (2), and in the 1980s Sepkoski combed the paleontology literature to develop a massive compendium that listed the first and last occurrences of known marine genera (3) and drew what has become the standard diversity curve (see the figure, panel A). This approach left many unsatisfied, however, because it did not account for biases in sampling. One problem, for instance, is that older fossils may be underrepresented because they are less likely than younger fossils to have survived erosional processes and to be found by paleontologists. Another problem is that investigators have collected more heavily in some geographic areas and rock strata than others. In an attempt to overcome such biases, Alroy and colleagues (including Sepkoski) created the Paleobiology Database (PBDB) (4), a large and growing compilation of data from nearly 100,000 fossil collections. Over the past decade, researchers have used statistical techniques to “subsample” the PBDB and draw “sample-standardized” diversity curves (5), including one published by Alroy and colleagues in 2008 (6) (see the figure, panel B).

Alroy has now used the PBDB to produce a diversity curve (panel C) with some features that were seen in Sepkoski's original uncorrected curve (3), but were absent from earlier PBDB curves (5, 6). Using this curve, Alroy reaches the unsurprising conclusion that the rules governing diversification of major groups were not fixed over time. The end-Permian mass extinction, for example, “reset” the rules for crinoids and “articulate” brachiopods, which led to lower diversity in these groups after the mass extinction. A second conclusion, more controversial and more dependent on the details of the new curve, is that most groups have caps on their diversity, although those caps varied at different times. That is, Alroy's analysis appears to refute some previous work that argues for unconstrained, or exponential, diversification of the marine biota (7, 8).

To place Alroy's results in context, it helps to note that the first results from the PBDB in 2001 (5) were contentious in part because they showed an almost complete absence of a diversity increase in the Cenozoic (∼65 million years ago to the present). This absence of a diversity rise, however, was challenged by other data; for example, researchers had reported that the average number of species found per locality increased dramatically toward the present (9). Furthermore, doubts arose over the veracity of the subsampling methods used to create the first PBDB curve (10). The 2008 PBDB curve rested on essentially the same methods but reflected data from about eight times as many collections (6). Nonetheless, the curve remained flatter over the Cenozoic than many expected.

One important reason for the differences in the PBDB curves, at least in the Cenozoic, seems to be the use of a new method of sample standardization. Alroy calls it shareholder sampling and asserts that, unlike previous methods, it ensures that uncommon taxa are fairly represented in the results. This method suggests that previous sample-standardized curves were indeed severely “damped” or showed less diversity than actually was present. Now, other similarly damped curves will need to be reexamined. For example, a sample-standardized study of phytoplankton has produced a remarkably flat trajectory toward the present that may understate diversity trends (11).

This is not likely to be the last word on marine diversity patterns over the past 530 million years. First, shareholder sampling needs to be further vetted by the scientific community. A second, deeper question is whether the PBDB is currently an accurate reflection of the fossil record. Some feel that it is not (1214). For example, most groups of living organisms show a pronounced latitudinal diversity gradient, with diversity decreasing from the equator to the poles. The PBDB's fossil record of bivalves over the past 11 million years, however, does not show a similar pattern. In part, this is because some key literature is missing from the PBDB, and also we have poor knowledge of the fossil record of the tropical realm, particularly of the diverse Indo-West Pacific, relative to temperate regions (12). These and similar deficiencies might be biasing the PBDB toward curves that support the view that there were constraints on diversification for groups such as bivalves. To further reduce such biases, projects such as the recent €.2.7 million (U.S. $3.4 million) effort by the European Union to describe Indonesia's tropical marine fossil record will be of critical importance.

Although Alroy's curve represents the latest contribution to our understanding of diversity dynamics on geologic time scales, there will be no immediate consensus on the details of the pattern of diversity. Still, the pieces are falling into place for rigorously understanding the dynamics that have shaped, and continue to shape, the biosphere.


Navigate This Article