(SCOTT SAMPSON): Like many kids, I was into dinosaurs at a young age. Unlike most kids, I never grew out of it. It’s been very interesting watching the evolution of dinosaur paleontology. When I was a kid, dinosaurs were generally regarded as sprawled, sluggish, dim-witted, and lizard-like—in short, not very interesting. That view held sway for most of the last century. Then, in the late 1960s, Yale paleontologist John Ostrom proposed that dinosaurs were very bird-like and that birds may be the direct descendants of dinosaurs. Although the bones on museum shelves remained unchanged, paleontologists suddenly began to look at dinosaurs differently. Virtually overnight, these prehistoric beasts became supercharged— smarter, faster, warm-blooded, and more complex. Now more like birds than lizards, evidence was found for such behaviors as parental care and cooperative hunting. Ultimately, we started to question whether this new view was real. Nevertheless, all of this dynamism within the science sparked a whole new generation of children (of all ages) with a passion for dinosaurs.
People often forget that paleontologists are extremely limited in the data they have access to. For the most part, we only have bones and teeth to work with, so it’s really the ideas that drive the science. And the ideas, of course, are driven by the biases of that particular moment. Dinosaur paleontology went from a lizard bias to a bird bias, and now the pendulum is swinging back toward the middle. For example, the top speeds of many dinosaurs have been slowed somewhat, and the assumption of warm-bloodedness has shifted toward talk of a range of metabolic strategies concentrated between those of reptiles on the one hand and birds and mammals on the other.
Like their object of study, dinosaur paleontologists have undergone a major transformation in recent decades. In the 1800s, paleontologists were trained largely as biologists, with a strong grounding in anatomy. During most of the 20th century, there was a shift in emphasis and most practitioners were trained as geologists. This subject duality reflects the schizophrenic nature of the field, sitting on the cusp of two major disciplines: biology and geology. With a geologic focus, many paleontologists rarely considered the biology of their study organisms. That situation is changing, moving back toward the biological end of the spectrum.
These days, there’s a variety of new, often high-tech tools being applied to paleontology. Let’s say you find a virtually complete, intact dinosaur skull. Of course, there is much to be seen from the outside, yet a tremendous amount of anatomical data is locked within, obscured by the rocky matrix. Typically, this sediment can’t be removed, because the process would damage the specimen. However, we can run such specimens through a high resolution CT scanner that enables us to look inside the skull (or whatever) and reconstruct important features like brain size and shape. We can estimate which portions of the brain were most developed, which in turn permits hypotheses about sensory abilities, standard head position in life, and even aspects of behavior.
Another very different tool, yet one also dependent upon technology, is a widespread method of reconstructing the historical relationships of organisms. Known as phylogenetic systematics, or cladistics, this technique enables biologists to assess the distribution of shared, specialized features within a group of study organisms. Large datasets, sometimes involving dozens of different groups and hundreds of characters, are processed using high-speed computers, which can sift through hundreds of thousands of branching alternatives in search of the simplest one(s), requiring the fewest number of evolutionary steps. Generally depicted as branching diagrams, or trees, these hypotheses are effectively estimates of evolutionary patterns of descent. Cladistics has become standard operating procedure throughout paleontology, and biology generally, revolutionizing our ability to determine organismal relationships.
An understanding of these relationships turns out to be prerequisite to most other kinds of studies within evolutionary biology. For example, there are well over 100 shared, specialized morphological characters now that link dinosaurs to birds. This is strong evidence supporting the notion of a close evolutionary bond between these groups. Within dinosaurs, the evidence indicates that birds show the greatest affinity with small, carnivorous dinosaurs, informally known as “raptors.” Until recently, feathers were the quintessential feature of avians, associated only with flight. Now, thanks to an amazing fossil locality near Liaoning, China, we have specimens of feathered raptor dinosaurs that clearly did not fly. This tells us that feathers evolved prior to flight, meaning that they must have first evolved for some other reason—perhaps for controlling body temperature, or for use in display. The point is that, until we understood the evolutionary relationships, we couldn’t make an argument as to the progression, or evolution, of feathers. These are the types of hypotheses currently being tested that would have been impossible 15 or 20 years ago, because we simply didn’t have the computing power to assess alternative hypotheses.
Another growing trend within paleontology (actually a re-emergence of earlier methods) is to use detailed studies of living animals in order to investigate the physiology, anatomy, and behavior of dinosaurs and other extinct organisms. Much of this work is directed toward reconstructing soft tissues, since bones are in many ways simply the framework used by vertebrates to anchor their other varied tissue types. So let’s say you’re interested in assessing the maximum running speed of Tyrannosaurus rex, the largest carnivore ever to walk the Earth. It’s been argued that this predatory behemoth, weighing in at about 6,000 kg, was capable of running at jeep-chasing speeds in excess of 40 mph. Others have claimed such excessive speeds to be nonsense.
How can we test this idea? Given that you can’t observe the behavior of T. rex directly, you might create a biomechanical model based on engineering principles. Yet any such models would be limited by the accuracy of its parameters. So there is a need for relatively high-resolution biological data. Since bones are the only tissue type for which we have good information, you might look at the cross-sectional properties of the hind limb elements. Yet even this is not sufficient, since we need some idea of the muscles involved. For muscles, we can turn to closely-related living animals, such as birds and crocodiles. If a particular soft tissue is found in one or both groups, its presence can be inferred with some confidence in the common ancestor, and thus in dinosaurs as well. This “phylogenetic bracket” method, pioneered by Larry Witmer, allows us to reconstruct not only muscles, but, at least in some cases, blood vessels, nerves, and other structures. These soft tissues often leave bony marks, or “osteological correlates”, in the form of scars, holes, grooves, and the like.
An exceptional young scientist named John Hutchinson, then a graduate student at UC Berkeley, recently employed just this combination of methods—an engineering model and muscle reconstructions—in order to test the locomotory abilities of Tyrannosaurus rex. John was able to model a bipedal, six tonne dinosaur predator and determine how much muscle mass would be necessary in the hind limbs to propel the animal at a speed of 40 mph. He concluded that such high running speeds would have required that T. rex devote something on the order of 80% of its body mass to hind limb muscles! Clearly this was not the case, so we’ve slowed those animals down somewhat. And lest you jump to the conclusion that the dinosaurian tyrant king must therefore have been a lowly scavenger (as has been argued), it is important to add that its likely prey were likely still significantly slower than this giant meat-eater.
Reconstructing soft tissues works best if we can use living close living relatives to reconstruct the anatomy of extinct forms. But what if the structure in question isn’t present among extant relatives? For example, let’s say you’ve got an eight-foot long skull of a Triceratops-like dinosaur adorned with horns over the nose and eyes, and an elongate bony frill sticking out behind. Traditionally, it was thought that such bizarre structures functioned first and foremost in defense against predators. Alternatively, others have suggested that these bony bells and whistles functioned in control of body temperature, in part by increasing the animal’s overall surface area. More recently, the “in vogue” hypothesis has been mate competition, with the horns and related features used either to attract members of the opposite sex or to intimidate same sex rivals. How could you assess these alternatives? In this case, you might turn not to close living relatives, but to extant analogues—that is, animals with similar kinds of structures. This is exactly what I did in a previous research project.
Many living animals have horns or hornlike organs; the list includes antelope, deer, chameleons, birds, and even ants. In virtually all these instances, these features function primarily in the competition for mates, either in display or in actual physical encounters. Importantly, species within these groups tend to be distinguished mostly or solely on the basis of these same characteristics. In other words, we are distinguishing species based on the same features that they themselves are using. This pattern holds for the dinosaurs as well. Whether it’s horned dinosaurs, crested duck-billed dinosaurs, plated stegosaurs, or dome-headed dinosaurs, paleontologists identify different species largely by these bizarre structures. Moreover, in both the living and the fossil examples, these structures tend(ed) to develop fully only as the animal’s approach(ed) sexual maturity and adult size. Together, this and other evidence strongly supports the mate competition hypothesis. It also underlines the importance of using living animals—for which we can examine more anatomy and observe actual behaviors—to assess the biology of extinct animals. Certainly much of the best paleontology done today synthesizes data from the modern and fossil realms.
Nevertheless, despite all the new perspectives, innovative technological applications, and revealing comparisons with living forms, it’s my concerted opinion that dinosaur paleontology (and indeed evolutionary biology generally) is currently sitting on the cusp of an entirely new era of discovery, one focused on connections. The great majority of current work within paleontology, and vertebrate paleontology in particular, is devoted to investigating patterns—for example, determining which animals lived where and when, as well as their interrelationships. While such work is obviously critical, it falls within the realm of alpha level science. Moreover, the underlying paradigm guiding this work emphasizes unique, historical events rather than common processes, let alone laws.
As many readers of Edge are aware, there is a strong trend within the physical, natural, and social sciences away from the traditional reductionist paradigm that has reigned over science for centuries. The new paradigm looks instead at the bigger picture of interrelationships among systems. Places like the Santa Fe Institute in New Mexico encourage scientists of various ilk to come together, learn to speak a common language, and concoct new ways of thinking about the world. This trend is just beginning to trickle down evolutionary biology, with increasing movement toward cross-disciplinary research programs. Consequently, the field is becoming much more interesting and dynamic, with collaborations bringing together, for example, paleontologists, ecologists, paleoclimatologists, and geologists.
The resulting questions, and thus the answers, tend to differ under this complexity-based paradigm. How did the world of dinosaurs differ from our own? Since we live in a miniscule snapshot in time, most people can’t relate to a thousand years, let alone millions, or billions of years. So how do we get our minds wrapped around Mesozoic timescapes? And once we’re there, how do we then recreate the world of dinosaurs? What role did dinosaurs play in their ecosystems? How did they relate to their environments, and what were these environments like? With often gigantic sizes, dinosaurs pushed the envelope of what it is to be a land-living animal; how were they able to do that? Perhaps most importantly, how did evolution and ecology converge to drive the various dinosaur radiations, and why were these oversized reptiles so successful for so long? In short, how did evolutionary and ecological processes combine to drive changes in dinosaurs? Paleontologists are only beginning to take this eco-evolutionary perspective, with important new insights.
It turns out that the Mesozoic Earth was both very different from and extremely similar to the world we know today. This was a time lacking in polar caps, tropical rain forests, and grasslands. Yet habitats functioned in exactly the same way as those we are familiar with. Nutrients and chemicals cycled through ecosystems. There were primary producers in the form of plants and bacteria. There was a diverse array of consumers, both herbivores and carnivores, and this of course is where the dinosaurs came in. Completing the cycle were numerous decomposing organisms. For example, paleobiologist Karen Chin has described evidence from the fossilized feces of dinosaurs demonstrating that dung beetles existed during the Mesozoic. As with their living descendants, these dung beetles metabolized fecal material, recycling components so they could be reused by other organisms. Another recent discovery has been the influence of bacteria on fossilization. It appears that many remarkable kinds of fossils, including the rare examples with preserved soft tissues such as skin and feathers, are due in large part to bacterial activity. Indeed it may be that bacteria are pivotal to the formation of fossils in general, something we hadn’t really thought much about previously.
Once we better understand the ecological role of a given morphological structure, we can then contemplate its evolutionary implications. For example, I noted above that closely related species of various dinosaur groups are often distinguished solely on the basis of structures interpreted as mate signals. It is remarkable how conservative these animals are in other aspects of their anatomy, including the teeth, limbs, and vertebral column. It’s as if these groups settled on a successful design and evolution then tinkered with the window-dressing.
We see the same thing today in birds, fishes, and other groups. There’s the oft-cited example of cichlid fishes in the East African great lakes, one of the greatest vertebrate radiations of all time. These animals identify members of their own species largely based on color patterns. These designs enable them to determine, “you’re one of mine and you’re not. I can mate with you but not you.” Recently, deforestation and subsequent erosion have been rampant along the margins of these lakes. Erosion transports abundant soil into the lakes. The water becomes murkier, and the fish can no longer see each other as well as they could before; certainly they can’t discern colors nearly as well. All of a sudden, they start to mate with individuals from different species because they can’t recognize members of their own kind any more. The species boundaries turn out to be quite fragile, with the cross-species unions actually generating viable offspring. This pattern underlines the importance of mating signals, which are often the very first things to change when new species form.
Increasingly, biologists point to two distinct factors necessary for the origin of species in macro-sized vertebrates like dinosaurs. First, there must be persistent isolation of sub-populations of a given species, so that interbreeding cannot occur, or at least is severely limited. Second, the genetic make-up of those sub-populations must differentiate to the point that individuals of one group can no longer reproduce successfully with those of the other. Recognizing these minimal requirements, we can explore the process of evolutionary radiations rather differently. Let’s take an example from dinosaurs.
I’m most fascinated by the Late Cretaceous, in particular the last 15 million years of the Mesozoic (80-65 million years ago), just before a giant asteroid (or whatever it was) slammed into the planet. We know more about dinosaurs from this time than from any other. Similarly, the place I’m most interested in is western North America, because we know more about the dinosaurs from this region than from any other.
It turns out that there was a great deal of environmental change going on in North America during the Late Cretaceous. Increased plate tectonic activity translated into rampant volcanism, which in turn pumped abundant CO2 into the atmosphere. Global climates responded with increased warming and higher sea levels, which in turn resulted in flooding of most major continents. The climate even at high latitudes during much of this period was warm and equable year-round, described by one investigator as “wall-to-wall Jamaica.” One of these continental seaways extended from the today’s Arctic Ocean to the Gulf of Mexico, splitting North America in two. Exquisite beachfront property could have been had at this time in Colorado, Montana, or Utah. The adjacent seaway wasn’t static but rather expanded and contracted over time. During times of expansion, or transgression, the animals living on the western North American landmass were sandwiched between the seaway to east and a rising mountain chain, the Cordilleran thrust belt, to the west. The flowering plants, or angiosperms, literally blossomed during this interval, forming dense, closed canopy forests.
Amidst this dynamic environmental backdrop, various groups of dinosaurs underwent dramatic radiations, with apparently rapid rates of both speciation and extinction. Now, it may be coincidence, or it may be that the environmental changes were key factors driving this evolutionary change. Certainly the transgressing seaway would have reduced available habitat on land, likely fragmenting populations. Another factor in this regard may have been the increased abundance of angiosperm forests. Once populations became fragmented and isolated, evolution apparently targeted mating signals, likely driven, at least in part, by sexual selection.
Ecologically speaking, once two closely related species differing only in reproductive structures (e.g., horns, frills, crests, etc.) come back into contact, it’s unlikely that both will persist for very long, since they will be doing the same thing to make a living. The geologic record of North American dinosaurs appears to support this pattern—that is, one of the daughter lineages lives on while the other goes extinct. So here we have the makings of an evolutionary scenario that combines a dense fossil record with physical and biological environmental changes to postulate an integrated hypothesis of change over time.
A similar eco-evolutionary problem I have been pursuing, one that also involves Late Cretaceous dinosaurs from North America, is the evolution of gigantism in large carnivores. How does evolution generate a 6,000 kg carnivore like T. rex? Few other predatory dinosaurs approached such incredible masses, and no other group of terrestrial carnivores has come close before or since. In general, when confronted by such problems, the tendency within evolutionary biology has been to focus on single-cause explanations. But Nature is rarely so simple; typically it’s necessary to consider multiple causal factors. So I got together with some paleontological colleagues of varying expertise— Mark Loewen, Jim Farlow, and Matt Carrano—to tackle the question of giant dinosaur carnivores. We set out to consider not only those forces that might drive animals to gigantic sizes, but, equally important, forces that would limit the attainment of such sizes.
Ultimately, we outlined several factors that, depending on their timing and combination, could limit or promote gigantism in terrestrial carnivores. First, we argued that the largest carnivorous dinosaurs likely had intermediate metabolic rates—that is, metabolic requirements higher than those of ectothermic (cold-blooded) lizards but significantly below those typical of endothermic (warm-blooded) mammals and birds. A low maintenance metabolism appears necessary, since, in order to keep its hot-blooded furnaces stoked, a lion-sized endotherm must consume several times more food than an ectothermic lizard of the same body mass. Calculations of estimated daily caloric intake suggest that a T. rex-sized endothermic carnivore is highly improbable, since it is very unlikely that it could have consumed enough food to maintain its six tonne body mass.
Yet a low maintenance metabolism was not enough to result in the evolution of gigantism. Over the entire 160-million-year duration of dinosaurs, it was only during the Late Jurassic, and particularly the Cretaceous, that Mesozoic ecosystems were inhabited by truly giant meat-eating dinosaurs. So what was going on earlier? We postulate that geography plays an important role. Gary Burness, Jared Diamond and co-authors have argued that geographic area dictates maximal body sizes in terrestrial vertebrates. They were able to establish that the larger the land area, the larger the maximal body mass. These authors found somewhat different regression lines for warm-blooded carnivores, cold-blooded carnivores, warm-blooded herbivores, and cold-blooded herbivores. The regression lines were highly predictive as well, such that you can accurately estimate the shrinkage in body size that a mammoth species, for example, would ultimately undergo if marooned on an island of a particular size. Moreover, this and other studies show that, in order to maintain populations sizes large enough to stave off extinction, large warm-blooded carnivores such as lions require vast, continent-sized species ranges. Humans, of course, have restricted the movements of virtually all large animals, but such extensive ranges were the norm traditionally.
Given the remarkable, virtually law-like consistency of this relationship among living and recently extinct vertebrates, we assumed that geography must also have been a major factor governing dinosaur body sizes. We further postulated that T. rex-sized carnivorous dinosaurs, whether warm- or cold-blooded, would require vast species ranges. An examination of the fossil record bears out this prediction; the largest carnivorous dinosaurs occur only on continent-scale landmasses.
Yet this pattern in the fossil record raises a fundamental question. When the dinosaurs originated in the Late Triassic, about 230 million years ago, all of the continents were united as the supercontinent Pangaea. Given this proposed relationship between maximal body size and landmass area, you would expect the largest carnivorous dinosaurs to occur when all the continental landmasses were connected. But that’s not what we find. It’s only during the Cretaceous, after most of the continental fragmentation was completed, that the largest forms, such as T. rex, existed. Clearly, then, intermediate-grade metabolic rates combined with vast species ranges were insufficient to provoke the evolution of meat-eating titans. We realized that at least one other ingredient was necessary.
Next we turned to competition, a dominant concept in evolutionary biology, yet one that has fallen out of favor somewhat in recent years. Over the past two decades or so, paleobiological hypotheses founded on competition have been brought into question, and there has been much emphasis—rightfully so I think—on the role of cooperation and symbiosis. Yet in this case we argue for interspecific competition as a limiting factor.
When all of the continents were united as Pangaea, and even during the initial phases of fragmentation, virtually every terrestrial ecosystem for which we have good data indicates the presence of multiple, perhaps two to four, kinds of large carnivorous dinosaurs, in the range of 750-2000 kg. Given the extensive continental connections, this was a time when terrestrial animals were able to move around much of the planet. It is also why we find remains of dinosaurs on every continent. They didn’t need to fly or swim across major marine barriers—they simply walked from landmass to landmass. With all of this faunal mixing, it is not surprising that we find multiple species of large carnivores in most ecosystems.
Unlike living carnivorous mammals, which often have highly specialized teeth and jaws for particular diets (meat, bone marrow, etc.), large carnivorous dinosaurs apparently lacked such ecological diversity. So, given that they were all doing pretty much the same thing to make a living, it seems reasonable to postulate that inter-species competition would have limited the maximal body size for any one species. It’s highly unlikely that a given lineage could have evolved to be a giant of five or six tonnes when several other species were in direct competition in the same ecosystem. As the continents split apart, dinosaurs and all other parts of the terrestrial biota went along for the ride on these giant rafts of continental crust, setting sail on independent different evolutionary courses. We postulate that it was only after all the continents broke apart that opportunities arose for a single species to dominate an ecosystem and grow to T. rex proportions.
Indeed the evidence suggests that this is exactly what happened with the tyrant king himself. About 75 million years ago, when North America was divided into two landmasses by a seaway, several smaller-bodied tyrannosaurs such as Daspletosaurus and Gorgosaurus lived alongside one another. These animals were large, about 1,000 to 2,000 kg, and no doubt menacing, yet a fraction the size of their subsequent relative, Tyrannosaurus rex, which lived about 67-65 million years ago. In contrast to its predecessors, T. rex lacked the direct competition with other large carnivores. For whatever reason, all other tyrannosaur lineages died out. Almost simultaneously, the seaway receded for good, reconnecting east and west America for the first time in 25 million years and effectively doubling the geographic area for North American dinosaurs. The additional area allowed Tyrannosaurus to increase in body size while maintaining population densities high enough to avoid extinction, at least for awhile. Thus, according to this hypothesis, at least three factors—intermediate metabolism, reduction in interspecies competition, and dramatic increase in geographic area—were necessary to allow Tyrannosaurus rex to pump up to record-breaking body sizes. It is important to note that integrative hypotheses like these result in testable predictions that can be falsified or supported by future observations.
Yet why, you might be thinking, should we bother with fossils at all, given that the record of ancient life is sporadic and limited, whereas the modern record is so much denser? Part of the answer is deep time. Evolution unfolds not during human life spans, but over thousands, millions, and billions of years. Despite what geneticists may argue, any understanding of evolutionary mechanisms will be grossly incomplete without a consideration of processes operating over deep time. In other words, if your window is restricted to the present, you will by necessity have a myopic view of life. By analogy, how could you really understand a given person if you knew nothing of their past? Paleobiology, by making use of the fossil record, has the ability to gaze backward and watch evolution play out over vast time spans.
Although there have been outstanding exceptions, like George Gaylord Simpson and Elizabeth Vrba, most of the theory work in paleobiology has been conducted by invertebrate, rather than vertebrate, paleontologists. This is in part because invertebrate sample sizes are typically so much larger than those available to fossil vertebrate workers. Nonetheless, because of their sheer size and propensity for fossilization, the record of dinosaurs (and mammals) is reasonably dense. So I think vertebrate paleontology will have much to add to this discussion in the coming years.
There is much to learn from a connections-based perspective, both in terms of the contingency of unique events (i.e., bifurcation points) that history provides, and the general rules that guide long-term change in natural systems. From this vantage point, the unfolding of life can be viewed as a tapestry in which every new thread is contingent upon the nature, timing, and interweaving of virtually all previous threads. This is an extension of the idea that the origin of new life forms is fundamentally contingent upon interactions among previous biotas. As Stephen J. Gould described it, if one could rewind the tape of life and let events play out again, the results would almost certainly differ dramatically. The point of distinction here is a deeper incorporation of the connections inherent in the web of life. Specifically, the origin of new species is inextricably linked both to evolutionary history and to intricate ecological relationships with other species. Thus, speciation might be aptly termed “interdependent origination.”
For example, it is often said that the extinction of dinosaurs 65 million years ago cleared the way for the radiation of mammals and, ultimately, the origin of humans. Yet the degree of life’s interconnectedness far exceeds that implied in this statement. Dinosaurs persisted for 160 million years prior to this mass dying, co-evolving in intricate organic webs with plants, bacteria, fungi, and algae, as well as other animals, including mammals. Together these Mesozoic life forms influenced the origins and fates of one another and all species that followed. Had the major extinction of the dinosaurs occurred earlier or later, or had dinosaurs never evolved, subsequent biotas would have been wholly different, and we almost certainly wouldn’t be here to contemplate nature. An equivalent claim could be made for any major group at any point in the history of life.
While a few investigators, such as Stuart Kaufmann and Niles Eldredge, have begun to work around the problem from different angles, we still seem to be a long way off from a synthesis of ecology, which focuses on matter-energy transfer systems, and evolution, which emphasizes genetic (information) change over time within complex adaptive systems. At this point, we are beginning to discern the steps in this eco-evolutionary dance, yet the music eludes us. Ultimately, if a true synthesis is to come, it will be accomplished only by combining insights from the modern realm—for example, through genetics, biochemistry, microbiology, and ecology—with those from the deep past. To my mind, this search for law-like properties amidst the numerous patterns in Nature is one of the principle challenges facing evolutionary biologists and ecologists alike in this century. All indicators suggest that we are approaching an exciting time of discovery.