Paleontologists（古生物学者） can't always get what they want, to paraphrase¹ the Rolling Stones, but sometimes they can get what they need, according to a study that will appear in the May issue of the American Naturalist². The fossil record captures both the broad sweep of evolutionary³ changes in life on earth as well as ecological⁴（生态的） responses to shorter-term local and regional environmental shifts. And yet the amount of variability（可变性，差异性） seen among successive fossil assemblages（集合体，与会者） tends to be low compared to that ecologists see over shorter time periods. This suggests that communities are extremely resilient（弹回的，有弹力的） or resistant⁵ to（有抵抗力的） change over decades to centuries.

In the American Naturalist article, University of Chicago paleontologists Adam Tomašových and Susan Kidwell used data on living and fossil communities to explore where the low variability in fossil assemblages comes from. Does it come from the natural "time averaging" of skeletal（骨骼的） remains⁶ that occurs during postmortem（死后的，尸体解剖） accumulation（积累，累积） , or from biological processes actively⁷ maintaining a particular community composition?

Processes in the latter category include strong species preferences for particular niches⁸（生态位） that may enable them to avoid extinction⁹ owing to high growth rates at small population sizes, or the buffering¹⁰（缓冲作用，减震） of population growth during unfavorable times.

On most seafloors, lake bottoms and land surfaces, however, sediment¹¹（沉淀物，沉积） accumulates quite slowly compared to the rates at which local animal populations generate skeletal remains. Multiple generations of skeletal remains may therefore become mixed within a single fossil assemblage. "Samples from a series of such time-averaged fossil assemblages are thus very different from the snapshots（快照） that come from repeatedly sampling a living community," said Tomašových, a postdoctoral scientist in Geophysical Sciences at UChicago.

Simulating the effects of time-averaging on living communities from a Texas lagoon¹²（泻湖，环礁湖） and an east African lake, Tomašových and Kidwell, the William Rainey Harper Professor in Geophysical Sciences, found that low temporal variation in species composition can be produced or reinforced by time averaging alone. Typical paleoecological data will thus seem to support models of local community stability, whether those communities have actually been stable.

"Fossil assemblages are, in most cases, the only source of information about past ecosystems¹³ unaffected by human disturbance," Tomašových said, "and so it is important that we evaluate all factors that might affect their ability to record rates of change. Post-mortem（事后剖析） processes that might bias¹⁴ assemblages toward under-estimating past variability are especially important."

On the positive side, Tomašových and Kidwell note that although many fossil assemblages will be too time-averaged to discriminate¹⁵（区别，歧视） variability generated by biological processes operating at small spatial¹⁶（空间的） scales (approximately a meter) —such as competition or predation-prey interactions—they excel at capturing the larger picture, such as the identities and relative abundances of species in the regional diversity pool that local communities draw upon.

The composition of that pool is determined¹⁷ by speciation（物种形成） , extinction and biogeographic（生物地理的） processes that are usually extremely challenging to measure using the limited spatial and temporal scope of conventional biological sampling. "It's a matter of scale," Kidwell said. "The fossil record can't give us everything we want at the fine scale, but for some questions, we can get what we need."