Did a huge asteroid or comet thundering to Earth spell environmental gloom and doom for the dinosaurs? This is just one of more than 80 theories that seek to explain the demise of these giants. Some are more plausible than others, but any theory that presents a picture of dinosaurs and the other organisms disappearing across the world in the wink of a geological eye has a significant flaw: there is no global record of dinosaur extinction. No one can say whether dinosaurs died out overnight around the world, or whether they lingered for several million years in some places after disappearing elsewhere. Palaeontologists must base their understanding of such mass extinctions on regional data; a close look at one region, the Western Interior of North America, reveals what was happening to the dinosaurs and other creatures 65 million years ago.
The Western Interior, stretching from Alberta, Canada, in the north to Texas in the south, is at this time the only area in the world where the record of the extinction of the dinosaurs is known in any detail. Eastern Montana is especially fruitful for the dinosaur fossils that my colleague Laurie Bryant and I have used for a detailed study of species changes at the Cretaceous-Tertiary (K/T) boundary. We found major extinctions, but also many survivals, and there are even major evolutionary breakthroughs, which form a pattern of extinction, survival, and revolution termed 鈥榖iotic turnover鈥�.
IN THE STARS OR ON THE CARDS?
The pattern of biotic turnover in the rocks of eastern Montana is like a card game 鈥� more specifically, a game like blackjack, which depends on both skill and luck. At the K/T boundary, the pattern of biotic turnover depended on evolutionary 鈥榣uck鈥� and ecological 鈥榮kill鈥�. Some species were lucky enough to evolve under conditions that favoured them in the variety of environmental perturbations around the time of the K/T boundary. Ecological skill for these vertebrates meant surviving wide fluctuations in the availability of food and shelter. Both luck and skill determine whether a species becomes extinct, merely survives, or evolves further.
Advertisement
Blackjack and the pattern of biotic turnover at the K/T boundary have other parallels. In both cases, the eventual outcome depends on conditions as well as how events unfold. In a game of blackjack, the starting conditions include the number of players, the number of decks of cards used and, in many cases, what the players have been drinking. For Cretaceous creatures, the starting conditions were such factors as whether an animal lived on land or in the water, if aquatic whether it preferred salt or fresh water, was warm blooded or cold blooded, laid eggs, or was small, medium, or large.
In blackjack, the playing conditions are whether a player 鈥榞oes bust鈥� or can play on, the order in which cards are dealt, and whether a drinking player continues to imbibe. For biotic turnover at the K/T boundary, the playing conditions included the loss of large areas of shallow seaways that had existed worldwide, leading to fragmentation of habitats; increased competition; climate change; asteroid impacts; and major volcanic eruptions. As a model for biotic turnover, blackjack does a better job of explaining the changes at the K/T boundary than theories based on a single cause. We are only just beginning to sort through the factors of evolutionary luck, ecological skill, and starting/playing conditions that controlled the biotic turnover at this time in Earth history.
The approach Laurie Bryant and I have used in eastern Montana is relatively simple although extremely time-consuming. We started with a basic question 鈥� of those vertebrates alive in the last few tens or hundreds of thousands of years before the dinosaurs died out in eastern Montana, how many survived the extinction event or events? We based our work on an enormous collection of fossils from this interval in eastern Montana, made over more than 20 years by crews under the direction of William Clemens, of the University of California at Berkeley. This huge effort has produced almost 150 000 specimens of 112 species of vertebrates 鈥� fish, frogs, lizards, mammals and, of course, dinosaurs.
THE QUICK AND THE DEAD
Instead of the 75 per cent extinction that is often cited, we found almost the opposite 鈥� our survey showed that perhaps as many as 65 per cent of the vertebrate species survived. Although this is only a regional record, we believe that it serves as a litmus test for assessing global theories of extinction.
Our study is unusual because it focused on extinctions at species level. A major concern about such studies is the inherent imprecision of the fossil record. The vagaries of preservation and classification that seem to plague species-level studies are supposedly masked by working at higher taxonomic groupings because such groups are arguably more taxonomically stable, and the chances of at least some representation in the fossil record is supposedly increased. In reality, such higher taxonomic groupings are not more stable and, more importantly, such studies do not represent evolutionary change at the level where it actually happens and so may give a highly distorted view of extinctions.
Many studies have simply asked how many families of dinosaurs became extinct at the K/T boundary, but documentation at family level is blind to possible changes in the number of species. For example, Peter Sheehan of the Milwaukee Museum and his colleagues found that of the eight families of dinosaurs they tracked through the last 10 million years of the Cretaceous Period, none became extinct until the K/T boundary. But these eight families included more than one species; in fact, more than 40 per cent of the species could have become extinct without producing a decline in the number of families.
But if palaeontologists are to use the information contained in the record of individual species, we must overcome several problems involved in reading the fossil record in such detail. Bryant and I realised that when a species disappears from the fossil record, four possibilities must be investigated. First, the species may have become extinct, and its entire gene pool eliminated, as happened to the dodo. Secondly, a species could have just disappeared locally. We found that some species from eastern Montana vanished at the K/T boundary, only to appear in early Tertiary rocks elsewhere in the Western Interior. Thirdly, a species may be very rare, and so appears to die out. Species that are very rare, fewer than 50 specimens out of the 150 000 in our sample, disappeared from the fossil record five times more frequently than more common species. Perhaps rare species were more susceptible to extinction if there were fewer individuals per species; other rare species are simply invalid taxa such as the dinosaur Ugrosaurus, which turned out to be the nose region of Triceratops. Fourthly, the disappearances of some species are pseudoextinctions. These false extinctions are examples of evolution in which the gene pool splits or becomes so greatly modified a new species forms. The clearest example at the K/T boundary, involves the mammals. Excluding pseudoextinction, it appears that most lines of mammals became extinct. In fact, the opposite is true; mammals were beginning a massive radiation at the K/T boundary. As many as 41 per cent of Late Cretaceous mammal disappearances in eastern Montana are pseudoextinctions.
These different sorts of disappearances are important in attempting to assess the pattern of extinction, survival, and evolution at the end of the Cretaceous. In the newest data set, the analysis begins with 103 species, excluding only the most problematic ones such as Ugrosaurus. If we allow for pseudoextinctions, species that survive outside eastern Montana, and exclude the rare species that disappear, we find that 72 per cent (74 out of 103 species) of the vertebrates survive the K/T boundary. If we include rare species, we find that 52 per cent (54 out of 103 species) of the vertebrates survive. Between one-half and three-quarters of all vertebrate species living in eastern Montana at the close of the Cretaceous Period survived. In a longer geological perspective, this range is comparable to percentages of biotic turnover 鈥� extinction, survival and evolution 鈥� measured for Late Cretaceous and early Tertiary times. Across the boundary, vertebrate species were continuing to arise and die out at much their usual rate.
So why does the K/T boundary have such a reputation for mass extinctions? For some groups, such as land plants in the mid-latitudes of North America and near-shore invertebrates, the level of extinction may well have been severe. But for others such as deep sea creatures and vertebrates, this old idea of mass extinction is incorrect or at least exaggerated. The extinction of all or nearly all truly large land vertebrates, however, continues to fuel the idea of massive K/T extinctions.
It is incorrect to say that all dinosaurs became extinct, because birds are descendants of dinosaurs, related most closely to some of the smaller carnivorous saurischians (reptile-hipped dinosaurs). Nevertheless, with the exception of a turtle and possibly a lizard, all the large vertebrates that became extinct at the K/T boundary were dinosaurs. But the loss of the large dinosaurs was not dramatic in terms of species number; at the very most they represented only 20 out of 111 vertebrate species.
Part of our overemphasis on dinosaur extinction is because in Western culture at least, we equate size with power and success. Nature has different ideas. Peter Dodson of the University of Pennsylvania has recently estimated that the dinosaurs had about 100 genera, worldwide, for each of the last two stages of the Cretaceous Period, which together lasted approximately 10 million years. In comparison, there were almost 200 genera of mammals in North America alone during the the first 10 million years of the Tertiary Period.
Although they never evolved very many species, the ultimate demise of the large dinosaurs did affect the terrestrial ecosystem. If nothing else, the explosive radiation of mammals that followed was a result of the death of the dinosaurs. But the extinction of the dinosaurs was not the only time most or all the largest land vertebrates were lost. The same pattern may characterise the several other mass extinctions in the history of our planet that included land vertebrates.
ONLY THE SMALL SURVIVE?
Details of the fate of the vertebrates at the K/T boundary bring further insight into the outcome of our game of palaeontological blackjack. Although between 52 and 72 per cent survived, four major groups of species fared much worse 鈥� the sharks, dinosaurs, lizards, and marsupials. Smaller animals, and those that thrive in water tended to fare better in this game of biotic turnover, but these four groups do not share a common ecological or evolutionary thread.
We should look to the environment of the Western Interior some 65 million years ago for reasons behind this pattern of extinctions. At the time of the last dinosaurs, eastern Montana was part of the low coastal plain along the western margin of a large seaway that split North America from north to south. Sea level was at one of its all time highs, here and around the world. All the continents were divided or partially swamped by these very shallow seas. During the Cretaceous, these seas flooded the Western Interior four times, finally receding from the continents at the end of the Cretaceous.
It is difficult to calculate how quickly the seas retreated for the last time, although it must have taken between tens of thousands and a few hundred thousand years. These marine regressions provide reasons for the selectivity and severity of some vertebrate extinctions such as the sharks, dinosaurs, lizards, and marsupials.
Sharks and their relatives can travel many kilometres upstream into fresh water, although most are marine creatures. When the sea levels fell, the oceans lost their links with the shallow seaways and sharks could no longer live in these fresh waters. They disappeared at least regionally, returning for a geologically short stay in the middle of the Paleocene, the first epoch of the Tertiary, when the seas rose and again reached the Western Interior. As the seas left eastern Montana, more poorly drained, swampy conditions prevailed. With the wetter conditions, the numbers of different lizard fauna plummeted.
Although not directly tied to the marine regression, the precipitous drop in the number of marsupial species does appear related. From their first appearance 85 million years ago until their decline 65 million years ago, marsupials were the taxonomically most diverse group of therian mammals (the group that includes both marsupials and placentals) in North America. The other major group of therians, the placentals, to which we belong, appears to have originated in central Asia. Although some species of placental mammals do show up in North America soon after marsupials arise, the greater number do not arrive and radiate until the end of the Cretaceous with the regression of the seaways. Competition with one group in particular, the archaic ungulates or condylarths, appears to have been the undoing of marsupials in North America.
DECLINE AND FALL
Although some species of sharks, lizards and marsupials became extinct, the groups flourished elsewhere. The stars of the show, the dinosaurs, were not so lucky. Birds thrive today as the direct descendants of saurischian dinosaurs and represent perhaps 20 times the number of species of dinosaurs that lived at any given time. Nonetheless, the really large dinosaurs disappeared at the end of the Cretaceous. Without a global record of this process, we do not know its geographic extent, except for the detailed information from North America. There, the loss of these creatures set the stage for the great radiation of the mammals.
The diverse extinction pattern alone argues strongly against a single cause of their extinction, such as an asteroid impact. Marine regression seems to be the major, albeit indirect, cause of the disappearance of dinosaurs in North America. For many years no one knew how such a process could be an agent of extinction. The answer has come in the past 10 years as ecologists have recognised the phenomenon of habitat fragmentation. We have known for some time that destruction of habitats can and does lead to massive extinction, but habitat fragmentation is more subtle and insidious. In the rainforests of the Amazon Basin (or in the canyons of my own city of San Diego), as habitats are chopped into smaller pieces, the diversity of plants and animals decreases rapidly, sometimes leading to local or even total extinction. As the seaways bisecting North America receded some 65 million years ago, it reduced and fragmented the low coastal plains where the dinosaurs lived. Size alone probably made these animals more susceptible to extinction. At the same time, freshwater systems lengthened giving a reason for higher survival among freshwater vertebrates.
We need to expand our database for the patterns of K/T biotic turnover to other continents. For now all that we can say with confidence is that if what we see in North America is any indication, we may have finally begun to understand the old conundrum of the decline and fall of the big beasts that we call dinosaurs.
J. David Archibald is professor of biology in the Department of Biology at San Diego State University.