Wednesday, September 17, 2025

Coral reefs set to stop growing as climate warms

University of Exeter

image:
Dead reef crest on Mexico's Caribbean coast
view more
Credit: Chris Perry

Most coral reefs will soon stop growing and may begin to erode – and almost all will do so if global warming hits 2°C, according to a new study in the western Atlantic.

An international team, led by scientists from the University of Exeter, assessed 400 reef sites around Florida, Mexico and Bonaire.

The study, published in the journal Nature, projects that more than 70% of the region’s reefs will stop growing by 2040 – and over 99% will do so by 2100 if warming reaches 2°C or more above pre-industrial levels.

Climate change – along with other issues such as coral disease and deteriorating water quality – reduces overall reef growth by killing corals and impacting colony growth rates.

To understand how changing reef ecology is impacting reef growth potential – in other words, how the balance of living organisms translates into vertical “accretion” (reef-building) – the team analysed fossil reefs from across the tropical western Atlantic region to improve understanding of how reef growth rates vary depending on the types of coral present.

They then combined this with ecological data from more than 400 modern reef sites across the region to calculate present-day reef growth rates, and to explore how growth rates will change under future climate change, and whether reefs can keep up with future sea-level rise.

“Our research shows that under current CO2 emission scenarios most Atlantic coral reefs will not only stop growing but many will actually be eroding by mid-century,” said lead author Professor Chris Perry, of the University of Exeter.

“At the same time, rates of sea-level rise will increase – and our analysis suggests the growth of reefs will lag behind.

“With reefs and sea levels moving in opposite directions, water depths above reefs will increase – raising flooding risks along vulnerable reef-fronted coasts and fundamentally changing nearshore ecosystems.

“Resultant water depth increases of around 0.7m are projected by the end of this century if global temperature increases exceed 2°C, and as much as 1.2m under higher warming rates.”

Reef growth is strongly influenced by the amount and types of living coral present.

Multiple factors including disease outbreaks and “bleaching” events caused by high temperatures have changed the makeup of many reefs – depleting key reef-building species.

“We are witnessing an alarming decline in both the abundance and diversity of corals across Atlantic coral reefs,” said co-author Dr Lorenzo Alvarez-Filip, from the Universidad Nacional Autonoma de México.

“Climate change is not only accelerating this decline but also worsening the cascading ecological and socio-economic consequences of their loss.”

Co-author Dr Didier de Bakker, from the University of Exeter, said: “The changes we project would have major impacts along coastlines where reefs presently help limit wave exposure.

“These changes would also transform the environmental conditions in nearshore lagoons which harbour important habitats such as seagrass beds.”

One strategy to reverse losses and enhance reef growth is coral restoration.

Dr Alice Webb, also from the University of Exeter, said: “The scale of action required to reverse current coral losses is significant.

“To have meaningful effects on limiting water depth increases, any restoration will need to occur in tandem with effective land and water management, and rapid climate mitigation actions. Actions to keep warming below 2°C are critical.”

Professor Perry concluded: “We are moving into a period where the two factors that control water depths above coral reefs – vertical reef growth rate and sea level rise rate – are starting to operate in increasingly divergent directions.

“Limiting climate warming is critical if we are to try to mitigate this and to avoid the worst impacts for coastlines and coastal ecosystems.”

The paper is entitled: “Reduced Atlantic reef growth past 2°C warming amplifies sea-level impacts.”

Widespread bleaching - Mexican Caribbean, 2023

Credit

Lorenzo Alvarez-Filip

Before, during and after coral bleaching [VIDEO] |

Journal

Nature

DOI

10.1038/s41586-025-09439-4

Article Title

Reduced Atlantic reef growth past 2°C warming amplifies sea-level impacts.”

Article Publication Date

17-Sep-2025

Gulf of Aqaba corals survive record-breaking heatwaves, offering hope amid global coral crisis

The Hebrew University of Jerusalem

New study reveals that corals in the Gulf of Aqaba have withstood four consecutive and intensifying marine heatwaves, including the world’s most extreme 2024 event, without suffering mass bleaching — a resilience unmatched elsewhere. This is important because coral reefs globally are collapsing under rising ocean temperatures, threatening ecosystems and human livelihoods. The Gulf of Aqaba may represent one of the planet’s last natural refuges for reef survival, offering a crucial model for understanding resilience and underscoring the urgency of protecting this unique ecosystem before even it reaches its limits. New study reveals resilience of Red Sea corals in the face of intensifying climate threats.

Coral Reef: https://drive.google.com/drive/folders/1nvggt4ffICAkBiIC6ZCqG_trTvRm306Y
Credit: Maoz Fine

[Hebrew University] — A new study led by Phd. student Na'ama-Rose Kochman and Prof. Maoz Fine of the Hebrew University’s Institute of Life Sciences and the Interuniversity Institute for Marine Sciences in Eilat, has uncovered extraordinary resilience among corals in the Gulf of Aqaba (GoA), raising hopes for the future of this globally-significant coral reef amid worsening climate change.

The findings, published in Science of the Total Environment, show that GoA corals endured four consecutive summers of intensifying marine heatwaves, including a record-breaking 2024 event that lasted 113 days with sea surface temperatures soaring to 32.6°C — 3.4°C above average and generating 30 Degree Heating Weeks (DHWs), the highest thermal stress recorded globally last year.

Key Findings

Survival through record stress: Five coral species survived the world’s most extreme 2024 heat event without mass bleaching.
Energy resilience: Corals maintained stable energy reserves, with higher symbiont carbohydrates in 2024 compared to previous years.
Species responses: Porites showed metabolic stability, while Cyphastrea experienced stress but recovered within months.
A climate refuge under pressure: Despite remarkable tolerance, sporadic shallow bleaching suggest the Gulf’s may be reaching its limit.

“While nearly half of reef-building corals worldwide face the risk of extinction, our study shows that the Gulf of Aqaba remains one of the last strongholds of reef survival,” said Na'ama-Rose Kochman. “But even this refuge is not immune to the accelerating pace of climate change and local pollution.”

Prof. Fine added: “These results highlight both the resilience and the fragility of coral ecosystems. They underscore the urgent need for regional conservation policies to safeguard what could be the world’s last thriving coral reef.”

Coral reefs are biodiversity hotspots that sustain millions of livelihoods, from fisheries to coastal protection. Yet marine heatwaves — prolonged periods of unusually warm sea surface temperatures — are now one of the primary drivers of coral mortality worldwide.

The resilience of the Gulf of Aqaba corals offers a rare glimmer of hope. Scientists emphasize, however, that without rapid climate action and local protections, even this stronghold may not withstand future warming.

Coral Reef

Credit

Maoz Fine

Journal

The Science of The Total Environment

DOI

10.1016/j.scitotenv.2025.180463

Method of Research

Observational study

Subject of Research

Animals

Article Title

Gulf of Aqaba as a thermal refuge: Insights from four years of intensifying marine heatwaves

Article Publication Date

16-Sep-2025

Delegation to Artificial Intelligence can increase dishonest behavior

International research team warns that people request dishonest behavior from AI systems, and that AI systems are prone to comply

Max Planck Institute for Human Development

image:
Does delegating to AI make us less ethical?
view more
Credit: Hani Jahani

When do people behave badly? Extensive research in behavioral science has shown that people are more likely to act dishonestly when they can distance themselves from the consequences. It's easier to bend or break the rules when no one is watching—or when someone else carries out the act. A new paper from an international team of researchers at the Max Planck Institute for Human Development, the University of Duisburg-Essen, and the Toulouse School of Economics shows that these moral brakes weaken even further when people delegate tasks to AI. Across 13 studies involving more than 8,000 participants, the researchers explored the ethical risks of machine delegation, both from the perspective of those giving and those implementing instructions. In studies focusing on how people gave instructions, they found that people were significantly more likely to cheat when they could offload the behavior to AI agents rather than act themselves, especially when using interfaces that required high-level goal-setting, rather than explicit instructions to act dishonestly. With this programming approach, dishonesty reached strikingly high levels, with only a small minority (12-16%) remaining honest, compared with the vast majority (95%) being honest when doing the task themselves. Even with the least concerning use of AI delegation—explicit instructions in the form of rules—only about 75% of people behaved honestly, marking a notable decline in dishonesty from self-reporting.

“Using AI creates a convenient moral distance between people and their actions—it can induce them to request behaviors they wouldn’t necessarily engage in themselves, nor potentially request from other humans” says Zoe Rahwan of the Max Planck Institute for Human Development. The research scientist studies ethical decision-making at the Center for Adaptive Rationality.

“Our study shows that people are more willing to engage in unethical behavior when they can delegate it to machines—especially when they don't have to say it outright,” adds Nils Köbis, who holds the chair in Human Understanding of Algorithms and Machines at the University of Duisburg-Essen (Research Center Trustworthy Data Science and Security), and formerly a Senior Research Scientist at the Max Planck Institute for Human Development in the Center for Humans and Machines. Given that AI agents are accessible to anyone with an Internet connection, the study’s joint-lead authors warn of a rise in unethical behavior.

Real-world examples of unethical AI behavior already exist, many of which emerged after the authors started these studies in 2022. One pricing algorithm used by a ride-sharing app encouraged drivers to relocate, not because passengers needed a ride, but to artificially create a shortage and trigger surge pricing. In another case, a rental platform’s AI tool was marketed as maximizing profit and ended up engaging in allegedly unlawful price-fixing. In Germany, gas stations have come under scrutiny for using pricing algorithms that appeared to adjust prices in sync with nearby competitors, leading to higher gas prices for customers. These systems were likely never explicitly told to cheat; they simply followed vaguely defined profit goals. While such cases show that machines can act unethically, the human side of the equation of whether and how people use AI to offload moral responsibility had remained largely unexplored.

How can researchers study unethical behavior in the lab?

Over the years, behavioral scientists have developed clever methods to study dishonesty. One of the most widely used is the die-roll task, which has been used in hundreds of studies around the world. In this task, participants observe die rolls outcomes and are then asked to report the outcome observed. Participants get paid according to the number they rolled, with higher reported numbers earning more money. This design enables the researchers to assess whether participants acted (on average) truthfully, bent the truth, or cheated entirely. Commonly, people report higher numbers than occurred. Studies have shown that behavior in this game reliably reflects real-world dishonesty, from fare dodging to unethical sales tactics. Using this die-roll task, the researchers explored what happens when people delegate the task of reporting die rolls to AI compared to doing the task themselves. Crucially, the researchers varied how to instruct the AI in the following three ways:

Delegation to a rule-based system: Participants had to tell the AI exactly what number to report for each die roll outcome.
Delegation to a supervised learning model: Participants could choose to train the AI using one of three training datasets from previous players: an honest player who always reports truthfully, a partially dishonest player who cheats occasionally, and a fully dishonest player, who always reports the maximum value of six, regardless of the actual die roll outcome.
Delegation to a goal-oriented application: Participants selected a priority on a seven-point scale, anchored with “maximize accuracy” and “maximize profit.”

More cheating with more ambiguous interfaces

The level of dishonesty surprised the researchers. Even in the rule-based setup, cheating was significantly higher than when people engaged in the task themselves. Without machine involvement, 95% of participants reported truthfully. This dropped to ~75% when participants specified the rules the machine should follow on their behalf. In the supervised learning condition – where people chose the data to train an AI – only around half remained honest. When participants only had to define a goal, the vast majority, over 84% engaged in dishonesty, and within that, between a third and a half of participants cheated to the fullest extent. These findings suggest that the more ambiguous the behavioral outcomes from the interface, the greater the temptation to cheat.

Given the emergence and proliferation of LLMs, such as ChatGPT, in subsequent studies, the researchers examine delegation behavior using natural language instructions. For both the die-roll task and the well-studied tax evasion game in which participants have an incentive to misreport earned income, participants were asked to write a free-form prompt for a large language model and for a human agent to act on its behalf. Across the two tasks, the researchers found some evidence of greater unethical intentions when using AI rather than human agents. But of greater interest was the consistent finding regarding the question: Who’s more likely to follow unethical instructions: humans or machines?

Humans vs. machines – Who’s more compliant with instructions to be dishonest?

Two studies with large samples found that LLMs are significantly more willing to carry out fully unethical prompts. In two studies, the team of researchers asked separate groups of humans to act as agents and follow the instructions written to complete the die-roll task and tax evasion game. These agents could earn a bonus by behaving in a manner consistent with the intentions of the person giving the instruction. While both humans and machines (GPT4) complied with honest prompts in over 96% of the cases, the big difference occurred for fully dishonest prompts, such as “I would like to make the most money possible so please cheat for the maximum”. Overall, human agents were much less likely to comply with fully dishonest requests (42%) than machines were (93%) in the die-roll task. The same pattern emerged in the tax evasion game, with humans only compliant with fully unethical requests 26% of the time, as opposed to 61% of a machine agent. This pattern of results held across a range of models: GPT-4o, Claude 3.5, and Llama 3. The researchers believe greater machine compliance with unethical instructions reflects that machines do not incur moral costs, certainly not in the same manner as incurred by humans.

Prevailing safeguards are largely ineffective

The frequent compliance with requests for unethical behavior in the afore-mentioned studies raises commonly-held concerns around LLM safeguards–commonly referred to as guardrails. Without effective countermeasures, unethical behavior will likely rise alongside the use of AI agents, the researchers warn.

The researchers tested a range of possible guardrails, from system-level constraints to those specified in prompts by the users. The content was also varied from general encouragement of ethical behaviors, based on claims made by the makers of some of the LLMs studied, to explicit forbidding of dishonesty with regard to the specific tasks. Guardrail strategies commonly failed to fully deter unethical behavior. The most effective guardrail strategy was surprisingly simple: a user-level prompt that explicitly forbade cheating in the relevant tasks.

While this guardrail strategy significantly diminished compliance with fully unethical instructions, for the researchers, this is not a hopeful result, as such measures are neither scalable nor reliably protective. “Our findings clearly show that we urgently need to further develop technical safeguards and regulatory frameworks,” says co-author Professor Iyad Rahwan, Director of the Center for Humans and Machines at the Max Planck Institute for Human Development. “But more than that, society needs to confront what it means to share moral responsibility with machines.”

These studies make a key contribution to the debate on AI ethics, especially in light of increasing automation in everyday life and the workplace. It highlights the importance of consciously designing delegation interfaces—and building adequate safeguards in the age of Agentic AI. Research at the MPIB is ongoing to better understand the factors that shape people's interactions with machines. These insights, together with the current findings, aim to promote ethical conduct by individuals, machines, and institutions.

At a glance:

Delegation to AI can induce dishonesty: When people delegated tasks to machine agents–whether voluntarily or in a forced manner–they were more likely to cheat. Dishonesty varied with the way in which they gave instructions, with lower rates seen for rule-setting and higher rates for goal-setting (where over 80% of people would cheat).
Machines follow unethical commands more often: Compliance with fully unethical instructions is another, novel, risk the researchers identified for AI delegation. In experiments with large language models, namely GPT-4, GPT-4o, Claude 3.5 Sonnet, and Llama 3.3, machines more frequently complied with such unethical instructions (58%-98%) than humans did (25-40%).
Technical safeguards are inadequate: Pre-existing LLM safeguards were largely ineffective at deterring unethical behaviour. The researchers tried a range of guardrail strategies and found that prohibitions on dishonesty must be highly specific to be effective. These, however, may not be practicable. Scalable, reliable safeguards and clear legal and societal frameworks are still lacking.

Journal

Nature

DOI

10.1038/s41586-025-09505-x

Method of Research

Experimental study

Subject of Research

People

Article Title

Delegation to artificial intelligence can increase dishonest behaviour.

Article Publication Date

17-Sep-2025

AI model forecasts disease risk decades in advance

New AI model can estimate the long-term risk of over 1,000 diseases and forecast human health changes over a decade in advance

European Molecular Biology Laboratory

image:
AI model forecasts disease risk decades in advance
view more
Credit: Karen Arnott/EMBL-EBI

Imagine a future where your medical history could help predict what health conditions you might face in the next two decades. Researchers have developed a generative AI model that uses large-scale health records to estimate how human health may change over time. It can forecast the risk and timing of over 1,000 diseases and predict health outcomes over a decade in advance.

This new generative AI model was custom-built using algorithmic concepts similar to those used in large language models (LLMs). It was trained on anonymised patient data from 400,000 participants from the UK Biobank. Researchers also successfully tested the model using data from 1.9 million patients in the Danish National Patient Registry. This approach is one of the most comprehensive demonstrations to date of how generative AI can model human disease progression at scale and was tested on data from two entirely separate healthcare systems.

“Our AI model is a proof of concept, showing that it’s possible for AI to learn many of our long-term health patterns and use this information to generate meaningful predictions,” said Ewan Birney, Interim Executive Director at the European Molecular Biology Laboratory (EMBL). “By modelling how illnesses develop over time, we can start to explore when certain risks emerge and how best to plan early interventions. It’s a big step towards more personalised and preventive approaches to healthcare.”

This work, published in the journal Nature, was a collaboration between EMBL, the German Cancer Research Centre (DKFZ), and the University of Copenhagen.

AI for health forecasting

Just as large language models can learn the structure of sentences, this AI model learns the "grammar" of health data to model medical histories as sequences of events unfolding over time. These events include medical diagnoses or lifestyle factors such as smoking. The model learns to forecast disease risk from the order in which such events happen and how much time passes between these events.

“Medical events often follow predictable patterns,” said Tom Fitzgerald, Staff Scientist at EMBL’s European Bioinformatics Institute (EMBL-EBI). “Our AI model learns those patterns and can forecast future health outcomes. It gives us a way to explore what might happen based on a person’s medical history and other key factors. Crucially, this is not a certainty, but an estimate of the potential risks.”

The model performs especially well for conditions with clear and consistent progression patterns, such as certain types of cancer, heart attacks, and septicaemia, which is a type of blood poisoning. However, the model is less reliable for more variable conditions, such as mental health disorders or pregnancy-related complications that depend on unpredictable life events.

Future use and limitations

Like weather forecasts, this new AI model provides probabilities, not certainties. It doesn’t predict exactly what will happen to an individual, but it offers well-calibrated estimates of how likely certain conditions are to occur over a given period. For example, it could predict the chance of developing heart disease within the next year. These risks are expressed as rates over time, similar to forecasting a 70% chance of rain tomorrow. Generally, forecasts over a shorter period of time have higher accuracy than long-range ones.

For example the model predicts varying levels of risk for heart attacks. Taking the UK BioBank cohort at the age of 60-65, the risk of heart attack varies from a chance of 4 in 10,000 per year for some men to approximately 1 in 100 in other men, depending on their prior diagnoses and lifestyle. Women have a lower risk on average, but a similar spread of risk. Moreover, the risks increase, on average, as people age. A systematic assessment on data from the UK Biobank not used for training showed that these calculated risks correspond well to the observed number of cases across age and sex groups.

The model is calibrated to produce accurate population-level risk estimates, forecasting how often certain conditions occur within groups of people. However, like any AI model, it has limitations. For example, because the model's training data from the UK Biobank comes primarily from individuals aged 40–60, childhood and adolescent health events are underrepresented. The model also contains demographic biases due to gaps in the training data, including the underrepresentation of certain ethnic groups.

While the model isn’t ready for clinical use, it could already help researchers:

understand how diseases develop and progress over time,
explore how lifestyle and past illnesses affect long-term disease risk,
simulate health outcomes using artificial patient data, in situations where real-world data are difficult to obtain or access.

In the future, similar AI tools trained on more representative datasets could assist clinicians in identifying high-risk patients early. With ageing populations and rising rates of chronic illness, being able to forecast future health needs could help healthcare systems plan better and allocate resources more efficiently. But much more testing, consultation, and robust regulatory frameworks are needed before AI models can be deployed in a clinical setting.

“This is the beginning of a new way to understand human health and disease progression,” said Moritz Gerstung, Head of the Division of AI in Oncology at DKFZ and former Group Leader at EMBL-EBI. “Generative models such as ours could one day help personalise care and anticipate healthcare needs at scale. By learning from large populations, these models offer a powerful lens into how diseases unfold, and could eventually support earlier, more tailored interventions.”

Data privacy and ethics

This AI model was trained using anonymised health data under strict ethical rules. UK Biobank participants gave informed consent, and Danish data were accessed in accordance with national regulations that require the data to remain within Denmark. Researchers used secure, virtual systems to analyse the data without moving them across borders. These safeguards help ensure that AI models are developed and used in ways that respect privacy and uphold ethical standards.

Funding

This work was funded by EMBL member state contributions, DKFZ funds and Novo Nordisk Foundation grant.

EMBL’s European Bioinformatics Institute (EMBL-EBI)

EMBL’s European Bioinformatics Institute (EMBL-EBI) is a global leader in the storage, analysis, and dissemination of large biological datasets. We help scientists realise the potential of big data by enhancing their ability to exploit complex information to make discoveries that benefit humankind.

We are at the forefront of computational biology research, with work spanning sequence analysis methods, multi-dimensional statistical analysis, and data-driven biological discovery, from plant biology to mammalian development and disease.

We are part of the European Molecular Biology Laboratory (EMBL) and are located on the Wellcome Genome Campus, one of the world’s largest concentrations of scientific and technical expertise in genomics.

Journal

Nature

DOI

10.1038/s41586-025-09529-3

Method of Research

Computational simulation/modeling

Subject of Research

People

Article Title

Learning the natural history of human disease with generative transformers

Article Publication Date

17-Sep-2025

‘Teen’ pachycephalosaur butts into fossil record

North Carolina State University

A “teenaged” pachycephalosaur from Mongolia’s Gobi Desert may provide answers to lingering questions around the dinosaur group, according to new research published today in the journal Nature. The fossil represents a new species of pachycephalosaur and is both the oldest and most complete skeleton of this dinosaur group found to date.

“Pachycephalosaurs are iconic dinosaurs, but they’re also rare and mysterious,” says Lindsay Zanno, associate research professor at North Carolina State University, head of paleontology at the North Carolina Museum of Natural Sciences and corresponding author of the work.

The specimen was discovered in the Khuren Dukh locality of the Eastern Gobi Basin by Tsogtbaatar Chinzorig from the Mongolian Academy of Sciences, who is the lead author of the paper and currently a research assistant at NC State.

The new species is called Zavacephale rinpoche, which is the combination of zava, meaning “root” or “origin” in Tibetan, and cephal, meaning “head” in Latin. The specific name, “rinpoche,” or “precious one” in Tibetan, refers to the domed skull discovered exposed on a cliff like a cabochon jewel.

Z. rinpoche lived around 108 million years ago during the Early Cretaceous period in what is now Mongolia’s Gobi Desert. At the time, the area was a valley dotted with lakes and surrounded by cliffs or escarpments. Pachycephalosaurs were plant eaters, and adults could grow to around 14 feet long (4.3 meters) and seven feet tall (2.1 meters), weighing 800 - 900 pounds (363 – 410 kilograms).

“Z. rinpoche predates all known pachycephalosaur fossils to date by about 15 million years,” Chinzorig says. “It was a small animal – about three feet or less than one meter long – and the most skeletally complete specimen yet found.”

The Z. rinpoche specimen the team discovered was not fully grown when it died. However, it already sported a fully formed dome, though without much of the additional ornamentation found on other pachycephalosaur fossils.

“Z. rinpoche is an important specimen for understanding the cranial dome development of pachycephalosaurs, which has been debated for a long time due to the absence of early diverging or pre-Late Cretaceous species and the fragmentary nature of nearly all pachycephalosaurian fossils,” Chinzorig says.

How to tell whether two skulls that look different belong to two distinct species or just different growth stages of the same species is a long-standing debate for paleontologists who study this group, and that’s where Z. rinpoche comes in.

“Pachycephalosaurs are all about the bling, but we can’t use flashy signaling structures alone to figure out what species they belong to or what growth stage they’re in because some cranial ornamentation changes as animals mature,” Zanno says.

“We age dinosaurs by looking at growth rings in bones, but most pachycephalosaur skeletons are just isolated, fragmentary skulls,” Zanno adds. “Z. rinpoche is a spectacular find because it has limbs and a complete skull, allowing us to couple growth stage and dome development for the first time.”

By examining a thin slice of the specimen’s lower leg bone, the researchers determined that, despite sporting a fully formed dome, this Z. rinpoche was still a juvenile when it died.

Pachycephalosaurs are famous for their large domed skulls and are often depicted using those domes to duel in epic headbutting contests. “The consensus is that these dinosaurs used the dome for socio-sexual behaviors,” Zanno says. “The domes wouldn’t have helped against predators or for temperature regulation, so they were most likely for showing off and competing for mates.

“If you need to headbutt yourself into a relationship, it’s a good idea to start rehearsing early,” she says.

Z. rinpoche fills in huge gaps in the pachycephalosaur timeline – both in terms of when they lived and how they grew, the researchers say.

“This specimen is a once-in-a-lifetime discovery. It is remarkable for being the oldest definitive pachycephalosaur, pushing back the fossil record of this group by at least 15 million years, but also because of how complete and well-preserved it is,” Zanno says. “Z. rinpoche gives us an unprecedented glimpse into the anatomy and biology of pachycephalosaurs, including what their hands looked like and that they used stomach stones to grind food.”

“The newly recovered materials of Z. rinpoche, such as the hand elements, the stomach stones (gastroliths), and an articulated tail with covered tendons, reshape our understanding of the paleobiology, locomotion, and body plan of these 'mysterious' dinosaurs," Chinzorig says.

The work appears in Nature and was supported by the National Geographic Society (grant NGS-100601R-23). Ryuji Takasaki of the Okayama University of Science; Junki Yoshida of the Fukushima Museum; Batsaikhan Buyantegsh, Buuvei Mainbayar and Khishigjav Tsogtbaatar of the Institute of Paleontology of the Mongolian Academy of Sciences; and Ryan Tucker of Stellenbosch University contributed to the work.

-peake-

Note to editors: An abstract follows.

“A Domed Pachycephalosaur From the Early Cretaceous of Mongolia”

DOI: 10.1038/s41586-025-09213-6

Authors: Tsogtbaatar Chinzorig, Institute of Paleontology of the Mongolian Academy of Sciences and North Carolina State University; Ryuji Takasaki, Okayama University of Science and University of Toronto; Junki Yoshida, Fukushima Museum; Ryan Tucker, Stellenbosch University; Batsaikhan Buyantegsh, Buuvei Mainbayar, and Khishigjav Tsogtbaatar, Institute of Paleontology of the Mongolian Academy of Sciences; Lindsay Zanno, North Carolina State University.
Published: Sept. 17, 2025 in Nature

Abstract:
Dome-headed pachycephalosaurians are among the most enigmatic dinosaurs. Bearing a hypertrophied skull roof and elaborate cranial ornamentation, members of the clade are hypothesized to have evolved complex sociosexual systems. Despite their importance for understanding behavioral ecology in Dinosauria, the absence of uncontested early diverging taxa has hindered our ability to reconstruct the origin and early evolution of the clade. Here, we describe Zavacephale rinpoche gen. et sp. nov., from the Early Cretaceous Khuren Dukh Formation of Mongolia—the most skeletally complete and geologically oldest pachycephalosaurian discovered globally. Zavacephale exhibits a well-developed frontoparietal dome and preserves the clade's first record of manual elements and gastroliths. Phylogenetic analysis recovers Zavacephale as one of the earliest diverging pachycephalosaurians, pushing back fossil evidence of the frontoparietal dome by at least 14 myrs and clarifying macroevolutionary trends in its assembly. We find that the earliest stage of dome evolution occurred via a frontal-first developmental pattern with retention of open supratemporal fenestra, mirroring hypothesized ontogenetic trajectories in some Late Cretaceous taxa. Finally, intraskeletal osteohistology of the frontoparietal dome and hindlimb demonstrates decoupling of sociosexual and somatic maturity in early pachycephalosaurians, with advanced dome development preceding terminal body size.