The bones tell a brutal story. At a quarry in Colorado, palaeontologists have pieced together food webs from 150 million years ago, and the message is clear: if you were a baby sauropod in the Late Jurassic, life was cheap. Very cheap indeed.
The Dry Mesa Dinosaur Quarry sits on Colorado’s Uncompahgre Plateau, where flash floods once swept through a drought-stricken landscape, entombing dinosaurs by the dozen. The site has yielded an extraordinary snapshot of Morrison Formation life. Six species of long-necked sauropods, multiple theropods including Allosaurus and the massive Torvosaurus, armoured Stegosaurus, and assorted smaller creatures. What makes Dry Mesa special isn’t just its diversity. It’s that the assemblage preserves animals across their entire life stages, from hatchlings to hulking adults.
Cassius Morrison at University College London and colleagues have now reconstructed the quarry’s food web in unprecedented detail, mapping who ate what across more than 12,000 possible food chains. The analysis reveals that immature sauropods—defenceless, abundant, and essentially abandoned by their parents—were the foundation supporting the entire predator guild.
“Life was cheap in this ecosystem and the lives of predators such as the Allosaurus were likely fuelled by the consumption of these baby sauropods,” says Morrison. The largest adult sauropods could exceed 30 tonnes, but they started life from eggs just a foot wide. That’s a staggering size difference, and it created what Morrison calls “ontogenetic niche shifts,” different life stages occupying completely different ecological roles.
The food web analysis shows this dramatically. Baby sauropods had 15 potential predators. By the time they reached intermediate size classes, that dropped to 11, then nine, then seven. The largest adults? Zero predators. They were immune, moving through the landscape as walking fortresses that nothing dared challenge.
This is where sauropod reproductive strategy becomes critical. Unlike mammals, which produce relatively few offspring and invest heavily in parental care, sauropods appear to have laid large clutches of eggs and provided minimal protection afterwards. Think of modern sea turtles—hatchlings emerge en masse and scatter, overwhelmed by predation, with only a tiny fraction surviving to adulthood. Evidence from sauropod nesting sites suggests something similar, though the sheer body mass involved makes direct parental care essentially impossible. How do you incubate eggs without crushing them when you weigh 20 tonnes?
The predators, meanwhile, showed remarkable dietary flexibility. Allosaurus jimmadseni, the most common large theropod at Dry Mesa, had an “in-degree” (number of potential prey items) of just 10 when small but 47 when fully grown. Torvosaurus, even more impressive, clocked 72 potential food sources at maximum size. These weren’t specialists. They were opportunists, and the most obvious opportunity was freshly hatched sauropods.
Morrison’s team used sophisticated regression models to calculate maximum prey sizes for different predator life stages. An adult Allosaurus could theoretically take down sauropods weighing several tonnes—but why bother when defenceless babies were everywhere? Modern predators on the African savannah follow similar logic. Lions can hunt adult elephants, but they predominantly target the young, the sick, or the isolated. It’s less dangerous, requires less energy, and the failure rate is lower.
The fossil record supports this preference. Bite marks on Morrison Formation bones show that while predators occasionally tackled large prey, immature individuals were overwhelmingly the targets. One study found that 70 million years later, in Late Cretaceous ecosystems dominated by tyrannosaurs, similar patterns held, though with a crucial difference. Those later ecosystems lacked the diversity of enormous sauropods that characterised the Morrison.
This absence may have driven the evolution of tyrannosaurs towards greater killing power. With fewer easy meals available, predators needed stronger jaws, better vision, and more sophisticated hunting strategies to take down well-defended hadrosaurs and ceratopsians. The Morrison predators had it easier, in a way. Why develop a bone-crushing bite when you can snap up baby diplodocids?
The analysis also reveals something about injured predators. Allosaurus specimens from the Morrison show horrific pathologies—stress fractures, bone infections, healed breaks. One individual appears to have been speared by a Stegosaurus tail spike, surviving long enough for some bone remodelling before eventual death. Yet many injured allosaurs lived with severe disabilities. How? Morrison suggests the abundance of vulnerable prey allowed even injured hunters to survive, scavenging or targeting the easiest marks whilst healing.
There’s a darker calculation here. The food web modelling suggests that in a stressed environment—during drought, for instance—cannibalism among theropods would increase. Conspecifics become just another food source when options narrow. Evidence from other Morrison sites, like Mygatt-Moore Quarry, shows possible cannibalistic bite marks on Allosaurus bones. Whether this represents true predation or scavenging of already-dead conspecifics remains debatable, but the option was clearly there.
The sauropods themselves showed niche partitioning by feeding height. Diplodocids like Diplodocus and Barosaurus were low-level browsers, hoovering up ground vegetation and plants below 1.5 metres. Macronarians like Brachiosaurus and Camarasaurus fed higher, reaching into tree canopies. This reduced competition and may have allowed six or seven sauropod species to coexist in the same region—though Morrison cautions that immature animals of different species likely had more dietary overlap than adults, potentially driving some of that coexistence pattern.
What emerges from the analysis is an ecosystem structured from the bottom up by sauropod biology. The sheer abundance of sauropods, combined with their extreme reproductive output and lack of parental care, created a conveyor belt of vulnerable prey. This in turn supported a diverse predator community with less need for extreme specialisation. Later Cretaceous ecosystems, lacking this foundation, developed differently—fewer, more dominant predators with enhanced capabilities for tackling dangerous prey.
The food web also raises questions about gregarious behaviour. Some researchers argue that multi-individual Allosaurus bonebeds indicate pack hunting. Morrison is sceptical. Modern crocodilians occasionally show non-communicative mob behaviour—multiple individuals converging on prey—but this doesn’t require the social coordination of wolf packs. Evidence for true cooperative hunting in Morrison theropods remains weak, and pathologies suggesting intraspecific violence argue against tight social bonds. More likely, these were solitary hunters that sometimes tolerated each other around abundant resources.
There’s a final twist. The Dry Mesa assemblage may represent a drought-induced concentration, where animals converged on shrinking water sources. Modern drought assemblages in Africa show similar patterns—elevated death rates, mixed species aggregations, extensive scavenging. If Dry Mesa captures such an event, the food web represents not normal times but an ecological crisis. Yet even in extremis, the pattern holds: baby sauropods at the bottom, apex predators at the top, and a complex web of interactions filling the space between.
The research provides a framework for comparing other Morrison Formation sites and, potentially, ecosystems from different periods. Each quarry will have its own characteristics—different species mixes, different preservational biases—but the methodology allows quantitative comparisons. As Morrison notes, understanding these ancient food webs helps us grasp the evolutionary pressures that shaped dinosaur biology and behaviour across millions of years.
For now, though, the message from Dry Mesa is stark: in the Late Jurassic, being born a sauropod was a dangerous lottery. You’d either die young or grow too large to kill—and the odds weren’t great.
