On a cool September morning in 2010 my crew and I began our daily descent from camp back into deep time, walking single file down a steep, knife-edge ridge of sandstone and mudstone in southern Utah's Grand Staircase-Escalante National Monument. Each of us carried water, a field notebook, lunch, a rock hammer and other hand tools. Heavier tools and materials—rock saws, picks, shovels, bags of plaster and swaths of burlap—awaited us half a mile away at the dig site. Even from the hilltop we could easily see the plaster jackets down in the quarry—alabaster beacons in a wilderness of arid, gray-striped badlands. Some of the irregular lumps were not much bigger than a loaf of bread. Others spanned 10 feet and tipped the scales at more than a ton. All contained the bony remains of animals that coexisted here 76 million years ago.

Over the course of two field seasons this single quarry—one of many in the fossil-rich rocks of the Kaiparowits Formation—had yielded a striking array of creatures, including several dinosaurs. Most impressive was a largely complete skeleton of Gryposaurus, a massive, duck-billed plant eater approaching the size of Tyrannosaurus. The crew was now under pressure to finish excavating the remaining fossils before the helicopter came in a few days to airlift the priceless cargo to a nearby road. From there the fossils would travel by truck to the Natural History Museum of Utah in Salt Lake City, where trained volunteers would painstakingly open the jackets, remove the rock and glue the bones back together over a period of months.

Pausing on a sandstone ledge to soak in the sprawling vista below, I imagined for the umpteenth time how this place might have appeared when these dinosaurs roamed. Back then, much of the territory was an immense, waterlogged floodplain. Sluggish rivers from mountains to the west meandered across a verdant landscape interspersed with ponds and lakes. Cypress trees thrived in the swampy lowlands; better-drained settings supported forests of conifers and flowering trees. Vines draped the tree branches, and the buzzing of insects filled the humid air. The scene would have called to mind the swamplands of northern Louisiana today—but with the addition of more than a dozen dinosaur species, from herbivorous duck-billed hadrosaurs and horned dinosaurs (called ceratopsids) to carnivorous, sickle-clawed dromaeosaurs and a type of giant tyrannosaur.

Our excavations in this remote region over the past 14 years—a joint project of the Utah museum, the U.S. Bureau of Land Management, and the Denver Museum of Nature & Science—have opened a fascinating window on the mix of dinosaur species that lived during the so-called Campanian Stage of the Late Cretaceous period, between 83.5 million and 70.6 million years ago, a time when dinosaurs here were undergoing perhaps their greatest florescence.

In a sense, the Kaiparowits fossil assemblage is unremarkable, preserving the same broad dinosaur groups unearthed from sediments of similar antiquity farther north in Montana and in Alberta, Canada. Yet the particular species in the Kaiparowits are unique, with many large-bodied forms—findings that are forcing us to reconsider much of what we thought we knew about dinosaur evolution and ecology. Let me explain.

South vs. North
During the late cretaceous the earth was a hothouse world. The planet's polar regions were free of ice caps, and global sea levels tended to be exceptionally high. A warm, saltwater sea, the Western Interior Seaway, inundated the central region of North America, connecting the Arctic Ocean with the Gulf of Mexico and dividing the continent into eastern and western portions: Appalachia and Laramidia, respectively. The dinosaurs, plants and other organisms we have been recovering from the Kaiparowits lived on Laramidia, a landmass less than one fifth of the size of its parent continent.

Beginning in the 1960s, fossil hunters working in the Western Interior began to notice that Late Cretaceous dinosaurs found in Montana and Alberta belonged to distinct species from those recovered from similarly aged rocks farther south in such places as New Mexico and Texas. In the 1980s Thomas Lehman of Texas Tech University tabulated the geographical occurrences of dinosaurs and other vertebrate creatures on Laramidia and found evidence of distinct northern and southern assemblages during the final 15 million years of the Cretaceous period, including the Campanian Stage. Lacking any indication of a physical barrier to north-south dispersal, Lehman hypothesized that a latitudinal climate gradient had produced distinct communities of plants and animals, including a bevy of big-bodied dinosaurs. His was a bold theory. Other researchers questioned the likelihood of multiple dinosaur communities coexisting on the diminutive landmass of Laramidia. It simply did not seem possible that so many kinds of giants could have shared such a small chunk of real estate.

Critics of Lehman's hypothesis pointed out that any apparent provincialism might be illusory, the result of uneven sampling of fossils through time. Given that the Late Cretaceous spans many millions of years, they noted, perhaps paleontologists working at different latitudes in the Western Interior have effectively been “time traveling” to different intervals. If so, this imbalanced sampling through time could generate the perception of distinct, coeval faunal provinces even if only a single cosmopolitan dinosaur fauna existed on Laramidia at any given geologic moment. Alternatively, skeptics observed, the seemingly distinct dinosaur communities could be the result of poor geographical sampling. Until recently, the vast majority of Laramidian dinosaurs were known from the north, particularly Alberta and Montana. Perhaps a more thorough sampling of dinosaurs from southern Laramidia would ultimately reveal a single, widespread community. These issues remained unresolved when my colleagues and I began our work in southern Utah in 2000.

The fossils that we have recovered from Grand Staircase-Escalante go a long way toward filling the southern dinosaur gap on Laramidia and bolstering Lehman's theory. Dating of the fossils has been essential to answering the question of whether distinct, coeval dinosaur provinces existed in the north and south. Team geologist Eric M. Roberts of James Cook University in Australia found layers of volcanic ash scattered throughout the Kaiparowits strata that he was able to date using radiometric techniques. The results indicated that the key, fossil-rich hot zone formed over a period of one million years between 76.5 million and 75.5 million years ago. Comparing these ash dates with those from other Laramidian formations revealed that the Kaiparowits Formation closely overlapped in time with the Dinosaur Park Formation in Alberta. We now had strong evidence that at least one pair of northern and southern dinosaur assemblages lived concurrently.

The next step was to assess whether the dinosaurs themselves differed from north to south. Of the 17 dinosaur varieties that the team has recovered so far from the one-million-year interval, 13 are sufficiently complete to allow species-level identifications. Only one—a duck-billed hadrosaur in the Gryposaurus genus—was possibly present farther north. This species looks very similar to Gryposaurus notabilis from Alberta, but the identification is currently uncertain, and investigations are under way to assess whether ours is a distinct species.

Outside of this single question mark, the emergent picture is clear. Every other Kaiparowits dinosaur species thus far identified differs from those found farther north. When small to midsize carnivores such as the oviraptorosaur Hagryphus and the troodont Talos prowled ancient Utah, the same groups were represented instead by Chirostenotes and Troodon, respectively, up in Alberta. Similarly, whereas a large-bodied, short-faced tyrannosaur named Teratophoneus was the top land carnivore in the Utah region, other tyrannosaurs such as Gorgosaurus filled this role in the north. Plant-eating dinosaurs in the Kaiparowits Formation are similarly distinct from the northern forms. One of these is Parasaurolophus, a bizarre hadrosaur with a long, tubular crest on the top of its head. Three species of Parasaurolophus have been found previously, one in Alberta and two in New Mexico; the Utah species appears to be new to science.

The pattern repeats among the ceratopsids. One species that we dubbed Utahceratops has a long-frilled skull approaching seven feet in length. The skull of a second, shorter-frilled species, Kosmoceratops, is ornate in the extreme, bearing 15 horns on its head, the most of any dinosaur. While Utahceratops, Kosmoceratops and a third ceratopsid dinosaur, Nasutoceratops, foraged in Utah, different species of horned dinosaurs munched on plants up north.

The newly discovered Kaiparowits dinosaur assemblage in Grand Staircase-Escalante provides by far the strongest evidence of isolated dinosaur provinces on Laramidia. Although the same major groupings of dinosaurs occurred in both the north and south, northerners and southerners were distinct species. None of the more than 50 Campanian dinosaur species from numerous formations can yet be confidently placed in both the north and south. These findings effectively refute the possibility that distinct northern and southern assemblages are merely an artifact of incomplete temporal or geographical sampling. Instead we must come to grips with the fact that at least two dinosaur communities coexisted on this landmass for about a million years of late Campanian time.

Land of Giants
That many of the dinosaurs in these two communities were giants deepens the Laramidia mystery. Studies of modern-day terrestrial mammals show a tight connection between maximal body size and land area. Large-bodied forms tend to range farther, both as individuals and as species, because bigger animals require more area to obtain enough food. For the same reason, species with more extensive home ranges tend to have lower population densities. Maximum body size in giant terrestrial mammal species, then, reflects a balance between maintaining population densities low enough to avoid overexploitation of food resources yet high enough to avoid extinction. Ultimately the upper limits of both body size and species diversity among megavertebrates are constrained by a combination of physiology (higher metabolic rates demand greater food intake), food availability and landmass area, with more extensive landmass areas typically supporting more kinds of large-bodied species. This relation places the severest demands on big-bodied carnivores, which must maintain relatively larger home ranges than herbivores because only a small fraction of an ecosystem's total energy budget reaches the top of the food chain.

In theory, giant dinosaurs should have followed a pattern similar to that of today's large terrestrial mammals, with few species on the relatively small Laramidian landmass. Yet taken together, the animal communities represented in the Kaiparowits Formation and the Dinosaur Park Formation contain at least 17 to 20 coeval species of giant dinosaurs—that is, forms exceeding one ton in adult body mass—with most weighing in at more than two tons. By modern standards, that scenario seems downright bizarre. Today the only place on earth where you can find an abundance of giants is Africa, which harbors six mammals with a mean body mass in excess of one ton, all of them herbivores: the giraffe, the hippopotamus (which spends most of its time in freshwater rather than on land), and two species each of elephants and rhinoceroses.

It is true that Africa and some other landmasses housed considerably more large terrestrial species in the past. For example, during the Early Pleistocene, between about 2.5 million and two million years ago, Africa supported on the order of 16 megaherbivore mammals: multiple varieties of giraffes, elephants, hippos and rhinos, plus several types of big-bodied antelope that weighed nearly a ton. Nevertheless, a number of lines of evidence indicate that the Laramidian dinosaur example is exceptional.

First, Laramidia was less than one fifth of the size of Pleistocene Africa, so those 17 to 20 dinosaur behemoths were confined to a much smaller area than their mammalian counterparts. Furthermore, abundant evidence from mass death accumulations, or bone beds, indicates that many species of hadrosaurs and ceratopsids congregated for at least part of every year in large “herds,” numbering in the hundreds (and perhaps thousands) of animals. Second, mammal-dominated ecosystems from the Pleistocene onward have few terrestrial carnivores approaching a ton. Indeed, mammal evolution on land has yet to produce carnivores that even approach the magnitude of a tyrannosaur. Africa's largest predator, the lion, typically weighs less than 600 pounds; Laramidia, in contrast, was home to at least three giant tyrannosaurs, all of which apparently exceeded a ton. Third, whereas paleontologists have found Early Pleistocene fossils in multiple African countries, current sampling on Laramidia is limited to two coeval geologic formations. Given that Laramidian dinosaurs appear to have had considerably smaller species ranges than modern terrestrial mammals, with minimal overlap between concurrent communities, it seems highly probable that additional dinosaurs lived on this landmass during the Campanian. If so, the total number of contemporaneous giants on Laramidia may have far surpassed 20 species. In short, the new Kaiparowits evidence strongly suggests that dinosaurs exceeded known mammalian limits for species richness at large body sizes.

Conservative Dinos or Productive Plants?
This comparison between African mammal giants and Laramidian dinosaur giants brings us back to the burning question elicited by Lehman's hypothesis. How did dinosaurs manage to squeeze so many varieties of giants into such a small area? Either these dinosaurs got by on less food than modern giants do, or the environments they inhabited produced more food than we see in modern settings.

Scientists have long debated whether dinosaur metabolism was more like that of the cold-blooded ectotherms (such as amphibians and reptiles) or the warm-blooded endotherms (such as birds and mammals). If their metabolic rates were intermediate between these groups, causing them to have lower energy requirements than large mammals do, this difference could help explain how so many big-bodied species coexisted on the relatively tiny landmass of Laramidia. Recent research by Brian K. McNab of the University of Florida supports this notion of “Goldilocks” dinosaurs—not cold-blooded, not hot-blooded, but something in between. McNab has found a range of evidence indicating that lower energy expenditures of dinosaurs may have enabled their communities to support biomasses up to five times greater than those of mammalian herbivores in present-day Africa.

Alternatively, relative to present-day land ecosystems, Late Cretaceous plants may have offered megaherbivores foods that were more abundant or more nutritious, or both. Plant diversity and abundance are controlled by such factors as precipitation, temperature, length of growing season and availability of niches. Today the greatest diversity and biomass of plants tend to occur in the tropics, but during the hothouse of the Late Cretaceous high temperatures may have limited plant and animal diversity in the equatorial regions. At midlatitudes such as those occupied by much of Laramidia, in contrast, the climate was mild and the growing season long. To the west, Laramidian mountain ranges and rivers multiplied the number of available niches. To the east, the Western Interior Seaway ameliorated temperatures while providing a major source of precipitation. Paleobotanist Ian Miller of the Denver Museum of Nature & Science has thus far recovered more than 100 different plant varieties from the Kaiparowits strata. Although much more work needs to be done, all indicators suggest that the dinosaur communities in Laramidia were founded on a great bounty and diversity of plants.

Figuring out whether slower metabolic rates or augmented food supplies enabled dinosaurs to reach such gargantuan proportions and high richness of species will require further testing. My hunch is that both factors were involved. One thing we can say with confidence is that the hothouse world of the dinosaurs was very different from the world of today. Many of the major biomes of our present-day icehouse earth—for example, grasslands, tundra and rain forests—were absent during the dinosaurs' reign, and we are still trying to glean even a basic understanding of their hothouse predecessors. The good news is that paleontology is becoming increasingly interdisciplinary—involving collaborations with geologists, paleoecologists and paleoclimatologists, to name a few—increasing the odds that fruitful insights will emerge.

Meanwhile our work in the Kaiparowits Formation, like any scientific research worth its salt, is generating as many questions as answers. How many distinct dinosaur communities existed at any one time on the lost continent of Laramidia? What was the nature of the barrier separating the northern and southern communities? Was this boundary based solely on climatic variation between the north and south, as first thought? Or, as some geologists now suspect, was some kind of physical barrier present, perhaps a series of large rivers flowing from the mountains to the sea at the latitude of northern Utah and Colorado?

One final, intriguing implication is worth noting. If it turns out that dinosaurs tended toward much smaller species ranges than equivalent-size mammals, the richness of dinosaur species globally may have been far greater than previously anticipated. Which is to say that many, many more weird and wonderful dinosaurs are probably still buried out there, patiently awaiting discovery.