Digital Zookeepers Take a Census
|Professor Arnold Miller and his students are contributors to the massive database.
Credit: Dottie Stover
If the name of each species on Earth were put on a single recipe card, the box containing them all would stretch for over 6 miles. There are approximately 1.4 million species that have been named by researchers, but the true number of species on earth may be anywhere between 5 and 30 million species.
If just keeping those animal names straight wasn’t challenging enough, the shuffling of cards over time has captured the combined interest of a 25 member scientific team, first organized by Dr. Charles Marshall (now at Harvard). In reporting some of their analysis of the timelines for fossil records, the team has created a kind of digital zoo. Cataloguing the taxonomy of an entire planet’s history, their database holds great promise for resolving century-old debates about how the Earth got to be such a rich spawning ground for life’s diversity.
That history of life is buried in the rocks: "Our goal is quite literally to provide a reasonable sampling of the entire fossil record", says Professor Arnold Miller, a team member of the database project and geologist at the University of Cincinnati. Miller, along with 4 of his graduate students, are all catalog contributors. The National Center for Ecological Analysis and Synthesis (NCEAS) houses the mammoth database in Santa Barbara, California.
So far, the team’s fossil catalog comprises 8,500 collections and over 110,000 taxonomic occurrences, with an initial emphasis on the rich collections of marine invertebrates. Including 90% of all the animals on the planet, invertebrates encompass worms, sponges, mollusks and other species without a backbone. The analysis of the catalog spans 300 million years, an era that geologists classify as Middle Ordovician-Carboniferous (approximately 450 Million years ago) and Late Jurassic-Paleogene (approximately 150 Million years ago).
Evolution: Smooth or Bumpy?
As recently published in the Proceedings of the National Academy of Sciences (PNAS), one key question the research team has pondered statistically dates back to the earliest fossil collectors. Has biodiversity in the oceans increased dramatically throughout the history of animal life, or moved erratically to an average plateau? Does a curve or flat-line best describe the type and frequency as new species arise or die off?
No doubt, during past periods life has simply sprouted wildly, in what some paleobiologists enthusiastically refer to as the ‘explosive radiation of animals’. Buried in the sedimentary layers on both land and sea, previous fossil records show distinctly these innovative bursts. But such unbounded biological optimism contrasts markedly with sharp disturbances and interruptions, including seven mass extinction peaks in the last 250 million years, each spaced 20 to 60 million years apart.
Untangling these competing forces may help the team resolve the limitations of the simplest model of slow expansion and diversification, called logistic growth (a gradual upward exponential curve). Primarily attributed to two factors–the favorable geographic niches (e.g. islands, desert, mountains) and cooperative interactions (e.g. predator-prey relations)– life’s diversity has been commonly assumed to be an expansion, or radiation of animals. Self-evidently drawing a straight line from the first appearance of life on Earth to as many as 30 million species shows expansion. But the journey itself seems less smooth, particularly when peering into the sometimes sporadic fossil records.
|400 million-year-old trilobite fossil. Trilobites dominated the as a marine fauna during the Cambrian Period.
Proving exactly how steady this diversification has gone has required a detective’s insight into the fossil record. The researcher’s particular attention to the bin sizes is new, and as past progress has burst forward in fits and starts, their recent analyses account for exactly when’ and how much’. In general, the newer the fossil, the higher the chance it has preserved whatever evidence might point to a more diverse taxonomy.
Miller says: "The discovery of more taxa during a given interval may imply that the world genuinely contained more taxa at that time (i.e., the increase is biologically meaningful). However, it is also possible that the fossil record is simply better–more ‘complete’ or more extensive–for that interval, leaving open the possibility that the apparent increase in biodiversity during the interval may not be biologically meaningful, but might, instead, be tied simply to the increased availability of fossils– and taxa– from the interval." Therefore, according to the researchers, diversity may indeed increase or may alternatively— as an artifact — just appear better because of a more extensive database of rocks and recognized species’ taxonomy.
Clearly, before a digital zoo reveals deeper patterns of change, more data needed cataloguing. Miller notes: "This issue has been debated back and forth over the years, but has not been resolved, partly because the data have not been available to resolve it."
Data sampling: Does size matter?
One particularly promising statistical method used so far has been to adjust for what may appear as more diversity, but which in truth may just arise from better fossils. "If it works, this overcomes the problem because we can ‘adjust down’ the intervals with larger samples," says Miller. An immediate conclusion from the researchers’ compendium was evident: "Interval size matters! All else being equal, a longer interval should contain a larger sample," says Miller.
In addition to picking which time interval to sample, or bin size for the number of species to classify as a genuine growth spurt, a second and formidable challenge remains — to go to more remote places to get fossils. For sheer numbers and variety however, some of the richest biological areas on the globe, like the Amazon, haven’t been charted equally in depth with those found in North America and Europe. This regional selection may introduce an added bias: "In assembling the data, it is much easier to acquire information from some parts of the world than it is from others," says Miller. "One of the real challenges for us is to ensure that our data are truly global in scope." A large chunk of the database now represents mainly northern industrialized areas, and also tends to capture more marine fossils than land samples.
In pulling together such a large and accomplished group of scientists, a common purpose and unity was found between them about the importance of looking at global diversity.
|Professor Miller used the database of marine fossil genera to examine longevity trends throughout the Phanerozoic.
Credit: Dottie Stover
"One of the reasons we were able to get off the ground and gain momentum on what has now turned into a large project," says Miller, "is that several of us were like-minded in the first place with respect to the analysis of global diversity." Getting older fossils from more remote venues remains one of their top research goals.
The research group has already highlighted a number of intriguing initial conclusions. As anticipated, the older the rocks, the less they reveal about biological diversification. This bias indeed does make for more of a plateau than previously thought, and flattens some expectations for radiative explosion of animal species. For example, in the most recent geological time periods, instead of a sharp increase in biodiversity, there was very little change in overall diversity over time. Younger rocks mean more fossils, but, in reality, not necessarily more biodiversity.
If that conclusion continues to gain support as more remote geography and land species get catalogued, then the team will have moved closer to resolving one of the grand debates of paleobiology: The pace and scope of species leaping forward on evolutionary branches.
However that debate settles itself, one summary seems clear: tapping the historical record has predictable pitfalls in comparing together young and old rocks. "We are reliant on the efforts of paleontologists over the past centuries to catalogue the known fossil record," concludes Miller.
As the current collection expands beyond the 8,500 fossil catalogs so far, more biological diversity will be sought out to include in the database. For the number of species studied statistically in the database so far, "there is nothing magical about the 110,000 figure," says Miller. "It simply represented where we were at the time we did our first set of analysis last winter. The database has grown since then."
But more revealing than just the quality of their initial data records, richer comparisons across different venues and epochs can approach a standard kind of encyclopedic compendium of life: "It certainly is our hope that these data will be useful for other purposes, " says Miller, "in addition to sampling standardization, given that we are collecting a variety of ancillary information on each occurrence (e.g., rock type and paleoenvironment, when available) that should enable all sorts of investigations in which diversification is compared and contrasted among different environmental or geographic regimes."