Imagine modern-day field biologists being sent back in time 66 million years ago to North Dakota on the brink of the end-Cretaceous extinction. Wandering the pine rainforests and palmetto-choked riverbanks, they set about recording the number of dinosaur species they observe. They have all sorts of traits to consider when sorting their subjects into species: color patterns, soft tissue, behavior and even genetic material. How many species would these time-traveling biologists identify? And how many of those species have left bones recognizable as such in the fossil record?
Parsing the taxonomy of extinct animals is a notoriously tricky business. For much of the history of paleontology, researchers have debated whether certain species should be lumped together or split apart, according to their own readings of incomplete and contestable fossils. Such taxonomic debates tend to hit the headlines most often where dinosaurs are involved. In the past two years, multiple studies have sought to clarify whether the animal known as Tyrannosaurus rex in fact represented a single species. Other research is weighing whether to split well-known taxa such as Velociraptor into multiple ones.
This isn’t just some esoteric academic debate. Some paleontologists argue that our estimates of prehistoric biodiversity are so far off base that they risk skewing our understanding of how life on Earth is faring today.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
“The actual numbers of species lost in previous environmental catastrophes are probably always worse than we currently record because we’re probably lumping more than one species under one name,” says Tom Holtz, a paleontologist at the University of Maryland.
Species are the “operational unit” of biology, says James Napoli, a paleontologist at the North Carolina Museum of Natural Sciences. They’re also famously difficult to delineate, with borders that blur and shift according to various definitions. Species may be defined based on physical differences, geographic range, evolutionary lineage and whether populations can—or do—interbreed, among other things. Perhaps not surprisingly, then, the question of how many species are alive today is contentious: estimates vary from two million to nearly nine million to a trillion, depending on the methods and assumptions used. Some researchers argue that many groups of organisms that we think of as species actually host complexes of multiple species that are only visible through genetics.
Over the past decades, population genetics has allowed scientists to study how genes flow—or don’t—across modern-day populations, leading to a number of previously unrecognized “cryptic species” in giraffes, orangutans, ostriches and crocodiles. There is even more potential diversity among invertebrates, with some researchers estimating that every insect species described on the basis of anatomy might contain anywhere from three to six cryptic species. “Cryptic species pose practical and theoretical challenges to the study of both the modern and the fossil records,” which is why it’s important to better understand them, says Erin Saupe, a paleobiologist at the University of Oxford.
That’s easier said than done where fossil organisms are concerned. Relatively few organisms are fossilized. Fewer still make it through the eons in particularly good shape. And while some fossil-rich areas are well studied, unequal histories of fossil-collecting by colonial powers and deep inequalities in scientific funding mean that others languish, says Nussaïbah Rajah, a paleontologist at the Friedrich Alexander University of Erlangen-Nuremberg in Germany.
Even in well-sampled, relatively well-preserved fossil ecosystems—such as the latest Cretaceous rocks of the Montana and Dakota badlands—it can be tricky to untangle how animals are related to one another. Genetic data generally aren’t available for fossil organisms. Instead paleontologists must typically rely on anatomical differences in preserved bones, shells and teeth. Those distinctions are what researchers use to assign animals to different species. But hard anatomy can be confusing: sometimes it’s not clear whether two specimens belong to different species or represent different growth stages of the same type of creature, Napoli says. Whichever interpretation one adopts will have serious ripple effects when one tries to track species richness over time.
Fossil species, therefore, can’t operate the same way modern species do as a unit of measurement in assessments of biodiversity, Holtz says. Vertebrate paleontologists in particular typically don’t have enough fossils of any given animal to understand the full range of variation in that species in the way that they can for a modern species.
Most studies of species richness in the fossil record therefore rely on “the modest but common instead of the rare but charismatic,” he observes. These include hard-shelled invertebrates such as trilobites, ammonites and miniature ostracods—organisms that leave dense, common remains. Vertebrate paleontologists, meanwhile, “barely have enough species to work with to allow a lot of our analyses to even run,” making it difficult to connect estimates of past species diversity to that of the modern day, says Emma Dunne, a paleobiologist at the Friedrich Alexander University of Erlangen-Nuremberg.
The question of how to measure the rise and fall of biodiversity numbers in the past has therefore been plagued by unknowns. Dunne notes that some researchers have tended to get around this problem by focusing less on the question of species numbers and looking instead at higher-level distinctions: the number of genera or families in a given unit of fossil-bearing rock and the way their distribution changes over time. (This approach is common with people who study insects or plants, fossils of which can be ambiguous even when they are well preserved.) Other scholars seek to compensate for the gaps in the record through statistical means. “When we compute biodiversity trends, we don’t look at absolute numbers because these are not representative of the ‘real’ biodiversity at the time,” Rajah says. Common species are more likely to be fossilized than rare ones. Statistical approaches can be used to account for preservation biases in the fossil record and allow for higher-level analyses of relative biodiversity trends, even if finer-grained analyses are difficult.
Investigators can also account for sampling biases. Which sites people choose to focus their time and attention on matters, and significant study can make an area appear more species-rich in comparison to others than it actually is. “There’s been a major push in the last five or 10 years to start standardizing between different fossil occurrences: standardizing for the amount of collection that’s occurred somewhere, for the amount of rock exposed for collections,” Napoli says, “so that our comparisons of diversity over time are as unbiased as can be.”
Some areas of the fossil record are detailed enough to track how environments changed over time and how species changed with them, Holtz notes. In Late Cretaceous North America, for example, one species of Triceratops seems to have evolved into another amid millennia-long shifts in climate and the landscape. In such cases, the fossil record is useful for determining whether “species migrated because the local environment was not viable anymore or there were changes in body sizes or ‘Has there been a change in abundance?’” Rajah says. Such research, she argues, can be used to directly target questions around climate change.
But Rajah maintains that the fossil record cannot be used as a yardstick by which to measure shifts in modern biodiversity. “We cannot and should not be comparing the fossil record and the modern world 1:1,” she says. “The fossil record provided invaluable information for understanding what is currently happening in the modern world. But the events that we recognize from the fossil record span over thousands and millions of years, so if we were to use those same values for addressing the current biodiversity loss, we would end up with artificially reduced rates.”
Even just trying to measure biodiversity changes in the past is thorny. The idea that statistical methods can compensate for gaps in the fossil record has been met with doubts from some experts. Hans-Dieter Sues, curator of vertebrate paleontology at the Smithsonian National Museum of Natural History, points to the oft-cited statistic that around 81 percent of marine species and 70 percent of land species vanished 252 million years ago amid the ravages of the Permian extinction. “We know a lot of things went extinct,” he says, but those numbers are all extrapolations and statistics that are not grounded in actual species data except in a few places. Elsewhere in the world, there are simply no fossils known from this time period.
In attempting to grapple with the question of how the diversity of life on Earth has shifted over the eons, Napoli says, the key thing is to acknowledge what we can and can’t know. There may be species unknown to us that are hidden within the bones of Tyrannosaurus and Triceratops. Teasing them out will offer a finer-grained understanding of how megafauna changed during the Cretaceous. For anything beyond that, people had better get working on that time machine.