How Did the Dunkleosteus Go Extinct?

Dunkleosteus disappeared roughly 359 million years ago during the end-Devonian mass extinction, a catastrophe that wiped out all placoderms and reshaped vertebrate life in the oceans. No single cause explains this extinction. Instead, a cascade of environmental disasters unfolded over the final stages of the Devonian period: the seas lost oxygen on a massive scale, volcanic eruptions destabilized the climate, expanding land plants choked waterways with nutrients, and the planet lurched from greenhouse warmth into glacial cold. For a heavily armored apex predator built to dominate shallow seas, those changes hit from every direction at once.

The Predator in Question

Dunkleosteus terrelli was among the most powerful predators of the Devonian seas. It belonged to the placoderms, a class of armored jawed fish that dominated marine and freshwater environments for tens of millions of years. Placoderm fossils span from the early Silurian, about 438 million years ago, through the end of the Devonian.¹ Dunkleosteus stood out even among placoderms for its sheer feeding power. Biomechanical modeling of its skull reveals a four-bar linkage mechanism that allowed rapid mouth opening for suction-style prey capture, paired with jaw muscles capable of generating more than 4,400 newtons of bite force at the jaw tip and over 5,300 newtons at the rear dental plates. That makes its bite among the most powerful of any fish, living or extinct.²

How big was it? For decades, popular sources pegged Dunkleosteus at five to ten meters long, roughly the size of a great white shark or larger. More recent statistical work using a large comparative dataset of arthrodires and living fishes paints a different picture. Typical adults likely reached about 3.4 meters, with the largest known individuals topping out around 4.1 meters.³ That is still a formidable animal, but it changes the ecological picture. Dunkleosteus was closer to the size of a bull shark than a school bus. This matters because it reframes how much food the animal needed, how far it ranged, and how sensitive it was to changes in the food web around it.

The Hangenberg Crisis

The extinction event that killed Dunkleosteus is known as the Hangenberg Crisis, named after a black shale horizon in Germany where the boundary is especially well preserved. It occurred at the very end of the Famennian stage, the last stage of the Devonian period, and it was not a gentle decline. High-precision dating places the event’s duration at a geologically brief interval, tied to a global shift from greenhouse to icehouse climate conditions, a worldwide marine transgression, and the widespread deposition of oxygen-starved black shale on the seafloor.⁴

The Hangenberg Crisis was actually the second major extinction pulse of the Late Devonian. An earlier event, the Frasnian-Famennian (or Kellwasser) extinction around 372 million years ago, had already culled reef ecosystems and thinned out many marine groups. But placoderms survived that first hit. It was the Hangenberg event that finished them. After it, the recovery of previously diverse groups including placoderms, lobe-finned fish, and acanthodians was minimal.⁵ For Dunkleosteus specifically, there is no fossil record beyond this boundary. Every placoderm lineage ends here.

Oceans That Could Not Breathe

The most consistently supported kill mechanism for the Hangenberg Crisis is ocean deoxygenation on a massive scale. Multiple independent lines of geochemical evidence, from uranium isotopes in marine limestones to iodine-to-calcium ratios in carbonate rocks, all point to the same conclusion: oxygen levels in shallow ocean waters dropped sharply right when marine life was dying.

Uranium isotope data from Late Devonian marine limestones shows that extinction events coincided with widespread deposits of anoxic sediment accumulating in subtropical shallow seas and some open ocean settings.⁶ Modeling based on these isotopic signals suggests that marine anoxia expanded to cover more than five percent of the continental shelf seafloor area, a dramatic expansion that coincided with the onset of the Hangenberg Crisis itself.⁷ Five percent might not sound like much in the abstract, but continental shelves are where most marine life concentrates. Losing oxygen across that much productive habitat would have devastated food webs from the bottom up.

Separate iodine-to-calcium measurements from carbonate rocks at multiple locations tell a consistent story. Values decline sharply in the critical interval just before the Devonian-Carboniferous boundary, coinciding with shifts in carbon and nitrogen isotope records that together indicate intensifying oxygen depletion in the waters where marine organisms actually lived.⁸ For a large, active predator like Dunkleosteus, which would have had high oxygen demands to power those massive jaw muscles and cruise through the water column, this kind of oxygen loss would have been catastrophic even before it collapsed the prey base.

Volcanic Eruptions and the Spread of Land Plants

The oxygen crash did not happen in a vacuum. Two major environmental forces were driving it, and both were operating on a planetary scale during the Late Devonian.

The first was volcanic. Three large igneous provinces erupted during this period: the Viluy Traps in eastern Siberia, the Kola large igneous province on the Kola Peninsula, and the Pripyat-Dnieper-Donets province in eastern Europe.⁹ Of these, the Viluy Traps are the most closely linked to the extinctions. Radiometric dating shows that the first phase of Viluy volcanism lines up with the Frasnian-Famennian boundary, the earlier extinction pulse, to within analytical uncertainty.¹⁰ A second volcanic pulse at roughly 363 million years ago would have overlapped with or preceded the Hangenberg Crisis. Massive volcanism pumps carbon dioxide, sulfur dioxide, and toxic metals into the atmosphere and ocean, destabilizing climate and ocean chemistry in ways that ripple through ecosystems for millions of years.

The second force was biological, and it came from land. The Late Devonian saw the rapid spread of the first large rooted trees and forests. As these plants expanded, their root systems accelerated the weathering of rocks on land, which flushed huge amounts of phosphorus into rivers and ultimately into the ocean. Phosphorus is a nutrient that fuels algal growth, and a sudden surplus of it triggers eutrophication: algal blooms explode, die, sink, and decompose, consuming dissolved oxygen in the process. Modeling based on phosphorus flux data from Devonian sediments shows that this enhanced nutrient delivery promoted oceanic eutrophication and deoxygenation. The resulting burial of organic matter in the ocean drew down atmospheric carbon dioxide levels and contributed to climate cooling on the order of half a degree to a degree and a half Celsius.⁹

This is one of the more striking aspects of the end-Devonian extinction: a major marine die-off was partly driven by what was happening on land. Life on the continents was revolutionizing itself, building soils and forests for the first time, and one side effect was choking the oceans of oxygen.

A Climate That Flipped

While volcanism pumped greenhouse gases into the atmosphere and land plants simultaneously pulled carbon dioxide back out through weathering and organic burial, the net result over time was a dramatic climate reversal. The Late Devonian world shifted from a long-running greenhouse state into icehouse conditions, complete with glaciation at high latitudes. This is not a subtle seasonal shift. It represents a fundamental reorganization of the planet’s energy balance.

For marine ecosystems, the cooling brought additional stress. Temperature changes alter ocean circulation, which controls where nutrients and oxygen are distributed. Cooling at the poles can strengthen the density contrast between polar and tropical waters, potentially changing how well deep water gets renewed with oxygen. On top of that, cooling can trigger the buildup of ice sheets, which locks up water on land and causes sea levels to drop.

And sea level did drop. A major fall during the Late Devonian likely terminated the growth of carbonate platforms, the shallow-water reef-like habitats, over wide areas of western North America and elsewhere.¹¹ These shallow marine environments were exactly the habitats that supported Dunkleosteus and its prey. When sea levels dropped and those platforms dried out or became too shallow, the living space for large marine predators contracted sharply. Dunkleosteus was not built for the deep open ocean; its heavy armor and ambush-style feeding suited shallow, productive waters. Losing those habitats would have been like pulling the floor out from under the entire ecosystem.

Why a Top Predator Was Especially Exposed

Not everything in the Devonian seas died at the same time or for the same reasons. So why was Dunkleosteus, along with all other placoderms, unable to weather the storm?

Large predators sit at the top of the food chain, which makes them inherently vulnerable to disruptions further down. They need more food per individual, they reproduce more slowly, and their populations are smaller to begin with. When oxygen-starved water kills off the small fish and invertebrates that form the base and middle of the food web, the apex predators are the first to starve and the last to recover. This pattern repeats across mass extinctions throughout Earth’s history.

Dunkleosteus had an additional vulnerability: its body plan. Placoderms were built around heavy external armor plates made of bone. That armor provided excellent protection against other predators, but it came at a metabolic cost. Maintaining and carrying all that bone requires energy, and it limits maneuverability compared to lighter-bodied fish. In stable, oxygen-rich waters full of prey, that trade-off worked in Dunkleosteus’s favor. In a world where oxygen was plummeting and food was scarce, the arithmetic reversed. A heavily armored body demands more oxygen per unit of activity, exactly the resource that was disappearing.

The animals that survived and thrived after the extinction were built differently. Small, fast-breeding ray-finned fishes, sharks, and early tetrapods dominated the post-extinction world. The Hangenberg event acted as a filter that precipitated the demise of placoderms, acanthodians, and most lobe-finned fish lineages, while giving an opening to highly successful radiations by ray-finned fishes, sharks and rays, and tetrapods.¹² The survivors tended to be smaller, lighter, and faster at reproducing, traits that let them bounce back in degraded environments where a giant armored predator could not.

The Shrunken World That Followed

One of the most telling consequences of the end-Devonian extinction is what happened to body size afterward. Vertebrates experienced persistent reductions in body size for at least 36 million years following the extinction. This was not a brief dip; it was a fundamental restructuring of what it meant to be a successful vertebrate in the ocean. Small ray-finned fishes, sharks, and tetrapods, most under one meter from snout to tail, radiated to dominate post-extinction ecosystems and vertebrate biodiversity.¹³

Researchers have investigated whether this prolonged downsizing was driven by atmospheric oxygen levels or temperature, and the answer appears to be no. The shrinkage persisted through conditions that varied considerably in both respects, which suggests that ecological factors, like the collapse of the large-bodied food web and the competitive advantage of being small in a recovering ecosystem, were the primary drivers.¹³ Vertebrates did not reach sizes of five meters or greater again until the Carboniferous, tens of millions of years later.³

The extinction of Dunkleosteus was not just the loss of one species. It marked the end of an entire body plan that had dominated the oceans for roughly eighty million years. Placoderms as a whole vanished, and with them went the armored-predator model that had defined marine ecosystems since the Silurian. The world that emerged was one of lighter, faster, smaller vertebrates, the ancestors of the fish and sharks that fill today’s oceans. In a real sense, the extinction of Dunkleosteus cleared the stage for modern vertebrate life.

Revised Size Estimates and What They Change

The recent downward revision of Dunkleosteus’s size is worth dwelling on because it shifts how paleontologists think about Devonian ecology more broadly. The old estimates of five to ten meters were not based on rigorous statistical analysis; they were extrapolations from skull proportions using rough comparisons to living fish.³ The newer method uses a measurement called orbit-opercular length and a dataset of nearly a thousand species and over three thousand observations to produce estimates that are statistically grounded.

If Dunkleosteus was around three to four meters rather than six to ten, the energy requirements of the animal change. A four-meter predator needs less food than a ten-meter one, but it is also more vulnerable to habitat loss because it can be outcompeted or starved more quickly when prey populations crash. The revised size also suggests that the Devonian ocean ecosystem was structured differently than we assumed. If the top predators were smaller, the food chains they sat atop were probably shorter and less energy-rich than previously modeled. That would make the whole system more fragile and more susceptible to the kind of cascading collapse triggered by anoxia and habitat loss.

The size question also affects how we think about the bite force numbers. The biomechanical study that produced those impressive figures modeled an individual estimated at six meters total length, which now appears to be above the likely maximum.² A more typically sized animal would still have had an extremely powerful bite, but the absolute numbers would scale down somewhat. The feeding mechanism itself, that four-bar linkage allowing fast opening and powerful closing, is not in dispute. Dunkleosteus was still a phenomenally effective predator for its time. It just was not as colossal as museum displays often suggest.

What the Fossil Record Can and Cannot Tell Us

One persistent challenge in reconstructing the extinction of Dunkleosteus is that the fossil record for placoderms is heavily biased toward their head and trunk shields, the bony armor that preserves well. The rest of the body, which was likely less heavily armored or even cartilaginous in some species, rarely fossilizes. This means we have good information about skull mechanics and armor construction but much less about locomotion, metabolism, soft tissue anatomy, and behavior. Most of what we infer about Dunkleosteus’s ecology comes from the morphology of its jaws and head plates, plus its position in the food web as interpreted from associated fossils and, occasionally, preserved gut contents.

The timing of the extinction is well constrained by conodont biostratigraphy, the use of tiny tooth-like fossils from eel-like animals to date marine sediments. But the resolution still limits how precisely we can separate cause from effect. Did anoxia kill Dunkleosteus directly, or did it collapse the food web first? Did sea-level drop remove habitat before the oxygen crash hit? These questions are hard to answer when the relevant events all happened within a geologically narrow window. What we can say is that multiple stressors converged and that no single cause explains the extinction. The Hangenberg Crisis was a compound catastrophe, and Dunkleosteus was built for a world that no longer existed by the time it was over.

1
Current Biology. Placoderms
2
PubMed Central. Feeding mechanics and bite force modelling of the skull of Dunkleosteus terrelli, an ancient apex predator
3
Diversity. A Devonian Fish Tale: A New Method of Body Length Estimation Suggests Much Smaller Sizes for Dunkleosteus terrelli (Placodermi: Arthrodira)
4
Terra Nova. High‐precision U–Pb age and duration of the latest Devonian (Famennian) Hangenberg event, and its implications
5
PubMed Central. End-Devonian extinction and a bottleneck in the early evolution of modern jawed vertebrates
6
Earth and Planetary Science Letters. Global seawater redox trends during the Late Devonian mass extinction detected using U isotopes of marine limestones
7
Earth and Planetary Science Letters. Extensive marine anoxia associated with the Late Devonian Hangenberg Crisis
8
Geochemistry, Geophysics, Geosystems. Intensified Ocean Deoxygenation During the end Devonian Mass Extinction
9
Communications Earth & Environment. The expansion of land plants during the Late Devonian contributed to the marine mass extinction
10
Palaeogeography, Palaeoclimatology, Palaeoecology. New 40Ar/39Ar and K–Ar ages of the Viluy traps (Eastern Siberia): Further evidence for a relationship with the Frasnian–Famennian mass extinction
11
Palaeogeography, Palaeoclimatology, Palaeoecology. The role of sea-level change and marine anoxia in the Frasnian–Famennian (Late Devonian) mass extinction
12
General and Comparative Endocrinology. Observations on the radiation of lobe-finned fishes, ray-finned fishes, and cartilaginous fishes
13
PubMed. Body-size reduction in vertebrates following the end-Devonian mass extinction