Category: 7. Science

Ever since William Harvey outlined the circulation of the blood, textbooks have assigned distinct roles to blood’s cellular cast.

Platelets served as the tiny workhorses that pulled clotting protein threads tight, white cells patrolled for infection, and red blood cells (RBCs) mostly hauled oxygen while passively filling space inside a forming clot.

A collaborative study from the University of Pennsylvania now overturns that familiar picture. Using a blend of biochemical tricks, high-resolution imaging, and mathematical modeling, the investigators show that red blood cells actively generate the forces that make a clot shrink and toughen once it has sealed a wound.

“This discovery reshapes how we understand one of the body’s most vital processes,” said senior author Rustem Litvinov from the Perelman School of Medicine (PSOM).

The project drew on expertise that spans hematology, cell biology, and soft-matter mechanics, illustrating how interdisciplinary science can upend long-held assumptions.

Clots shrink without platelets

John Weisel, a professor of cell and developmental biology at PSOM, has spent decades probing fibrin – the insoluble, rope-like protein network that glues a clot together.

In previous studies, he and Litvinov had dissected how platelets tug on fibrin fibers. So when they decided to revisit that system without platelets, they expected an inert mass.

Professor Weisel said the team did not expect anything to happen. “Instead, the clots shrank by more than 20 percent.”

To rule out platelet activity, the team used blood treated to block the platelets’ ability to contract. Once again, the clots pulled inward. At that point, the researchers confronted the unanticipated culprit.

“Red blood cells were thought to be passive bystanders,” Weisel said. Far from passive, the cells appeared to be doing mechanical work.

“That’s when we realized red blood cells must be doing more than just taking up space,” Litvinov explained.

Red cells act like gels

How could flexible discs with no contractile proteins of their own produce force? The Penn biologists enlisted Prashant Purohit, a professor of mechanical engineering and applied mechanics who specializes in the behavior of gels and other squishy materials.

“Red blood cells have been studied since the 17th century,” noted Purohit. “The surprising fact is that we’re still finding out new things about them in the 21st century.”

Purohit constructed a mathematical model grounded in colloid physics. As a clot forms, fibrin polymerizes into a porous mesh. This mesh traps RBCs along with plasma proteins such as albumin and fibrinogen.

When the mesh compacts, large protein molecules are squeezed out of the tight spaces between adjacent red cells faster than they can leave the surrounding fluid. That imbalance, known as an osmotic depletion force, produces an external pressure that pushes the trapped cells into even closer contact.

Packed blood cells shrink the clot

“Essentially, the proteins in the surrounding fluid create an imbalance in pressure that pushes red blood cells together,” Purohit said.

The packed cells transmit the pressure back to the fibrin scaffold, making the entire clot shrink and stiffen.

“This attractive force causes them to pack more tightly, helping the clot contract even without platelets,” he added.

The model also permitted the team to quantify a second, previously proposed mechanism: molecular bridging. In this process, complementary molecules on neighboring RBC membranes bind together.

Clot behavior matches math

Study first author Alina Peshkova, now a postdoctoral researcher in pharmacology at Penn, designed a series of clotting assays to test the model head-on.

When she blocked the membrane molecules required for bridging, clots still contracted robustly. When she manipulated the chemical environment to eliminate the osmotic pressure gradient, contraction was largely abolished.

“We experimentally confirmed what the model predicted,” Peshkova said. “It’s an example of theory and practice coming together to support each other.”

Findings may impact stroke care

Most acute clotting problems in medicine trace back to an imbalance between clot formation and dissolution.

If RBC-driven contraction proves to be a major determinant of clot strength in vivo, it could help explain why patients with anemia or sickle cell disease sometimes experience unusual clotting complications.

Conversely, individuals with very low platelet counts (thrombocytopenia) might still achieve adequate clot retraction if their red cells provide compensatory force.

A better grasp of clot mechanics is also relevant to thromboembolism. A clot that retracts too vigorously can become so dense that fragments break off and travel downstream, lodging in the lungs, coronary arteries, or brain.

Understanding how osmotic forces affect clot architecture may therefore inform strategies to prevent strokes or pulmonary emboli.

“Ultimately, our model is going to be helpful in understanding, preventing, and treating diseases related to clotting inside the bloodstream,” Purohit said.

Physical forces reshape biology

The Penn study showcases the power of looking at a biological question through a physical lens. Osmotic depletion had long been recognized in industrial colloids – pigment particles in paint, milk proteins in dairy science – but had received little attention in hemostasis research.

By merging classic hematology with contemporary soft-matter theory, the investigators revealed that an abundant, well-studied cell type still harbors surprises that matter for human health.

“This attractive force causes them to pack more tightly, helping the clot contract even without platelets,” Purohit emphasized.

The remark captures the study’s dual message: mechanical laws operate inside living tissues, and basic science can still revise what clinicians think they already know.

With platelet-free contraction now firmly established, the field can move on to explore how genetic diseases, pharmaceuticals, or blood storage conditions influence this newly recognized RBC function.

Researchers can also investigate whether tweaking it might tip the balance between life-saving hemostasis and life-threatening thrombosis.

The study is published in the journal Blood Advances.

—–

Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates. 

Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.

—–

Physicists can map its gravity, tally its cosmic abundance, and observe its fingerprints in the microwave glow of the Big Bang. Yet no one has ever detected a single particle of dark matter in the laboratory.

That enduring mystery drives much of today’s theoretical physics, and few researchers are pushing the boundaries more vigorously than Professor Stefano Profumo at the University of California, Santa Cruz.

In two recent papers, Profumo offers separate – yet thematically linked – visions of how dark matter might have emerged naturally from the earliest moments of cosmic history rather than through new interactions with ordinary matter.

Dark matter from a “mirror world”

The first study, published in the journal Physical Review D, asks whether dark matter could consist of ultra-compact black-hole – like objects forged in a hidden sector that mirrors the familiar world of quarks and gluons.

Drawing inspiration from quantum chromodynamics, Profumo and collaborators imagine a “dark QCD” with its own quarks, gluons, and strong nuclear force.

In the scorching seconds after the Big Bang, dark quarks could have bound into heavy composite states – dark baryons.

These baryons may have later collapsed under their own gravity, leaving behind relics only a few times heavier than the Planck mass.

These remnants interact only via gravity, slipping past detectors while explaining the universe’s missing 80 percent of matter.

This scenario nods to a rich Santa Cruz pedigree. Former UCSC professor Michael Dine helped pioneer the QCD axion hypothesis, another leading dark-matter idea.

Research Professor Abe Seiden also contributed to experiments probing the internal structure of hadrons.

Profumo’s work revives that tradition, extending well-tested gauge theories into a shadow realm. It reveals a new way composite physics might answer the dark-matter riddle without contradicting existing data.

Dark matter in an expanding universe

Profumo’s other paper, published in the same journal, pivots from hidden sectors to the fabric of spacetime itself.

He explores the possibility that an episode of accelerated expansion after cosmic inflation could have generated dark matter particles.

This expansion was slower than primordial inflation but faster than what ordinary matter or radiation would allow, resembling how a black hole’s event horizon is predicted to radiate.

Using quantum field theory in curved spacetime, he calculated that a brief phase of quasi-de Sitter expansion could “heat” the universe’s horizon.

This process would spontaneously produce an ample abundance of stable, non-interacting particles over a broad mass range.

No additional forces or couplings are required; gravity alone does the job. Both mechanisms avoid needing detectable interactions, which increasingly sensitive underground experiments have repeatedly failed to observe.

“Both mechanisms are highly speculative, but they offer self-contained and calculable scenarios that don’t rely on conventional particle dark matter models, which are increasingly under pressure from null experimental results,” explained Profumo.

Architecture of the universe

Profumo literally wrote the textbook on this pursuit. His 2017 volume An Introduction to Particle Dark Matter remains a touchstone for students and researchers alike.

The new studies continue UCSC’s long-standing approach of fusing particle theory, cosmology, and astrophysics. This approach helped establish the Lambda-Cold Dark Matter model currently favored by observational data.

The campus’s Institute for Particle Physics and its theorists have repeatedly shown how fresh ideas about the micro-world can illuminate the large-scale architecture of the universe.

Professor Profumo emphasized that the proposals rest on physics already in hand, not speculative new forces.

The hidden-sector picture leans on well-studied SU(N) gauge theories. The horizon-production mechanism relies on the same mathematics that underpins Hawking radiation.

Hints in gravitational waves

Neither idea will be easy to confirm. Dark-sector black holes would evade direct detection and might reveal themselves only through subtle gravitational-wave signatures or small anomalies in the cosmic microwave background.

Horizon-generated particles would likewise be invisible except through their influence on the growth of cosmic structure or precision measurements of Big Bang nucleosynthesis.

Yet both frameworks make quantitative predictions that could guide future observations. These range from the Laser Interferometer Space Antenna’s (LISA) hunt for primordial black hole mergers to next-generation surveys mapping dark matter distributions on megaparsec scales.

Meanwhile, Professor Profumo and his colleagues are refining the math. They are exploring, for instance, how different dark-QCD parameters affect the mass spectrum of possible mini-black-hole remnants.

The researchers are also examining how varying the duration of post-inflationary acceleration changes the predicted relic abundance. Each tweak produces patterns that upcoming data sets may support or rule out.

Mirror worlds and dark matter hunting

The null results piling up at direct-detection experiments have encouraged theorists to broaden their horizons, both figuratively and literally.

By rooting his models in established quantum field theory and gravitation, Profumo offers a pair of audacious yet disciplined road maps. These maps outline where dark matter might have come from – and how researchers could someday prove it.

Whether the answer lies in a shadowy confining force or in particles born at the edge of the universe’s own horizon, the work demonstrates how much fertile ground remains between the known and the unknown.

In the search for the cosmos’s invisible matter, these studies remind us that sometimes the boldest steps forward come from looking back. They point to the Big Bang’s first moments, when galaxies, stars, and hidden particles first began to take shape.

—–

Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates. 

Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.

—–

The obtained GaN nanocrystals have a hexagonal wurtzite structure, which was confirmed by XRD measurement, shown in (Fig. 1).

According to the pattern (ICSD #34476), the obtained GaN nanocrystals have a pure structure with the broad reflections indicating nanosize. The average grain size was calculated by Rietveld analysis to be 12 nm. The absorption spectrum of GaN nanocrystalline powder is shown in (Fig. 2). By using the Kubelka-Munk function²⁷ the energy band gap was determined to be 3.17 eV. Due to the nano-size of the particles, there are various defects in the structure, such as Ga and N vacancies, surface defects, or grain boundary defects. Therefore, the absorption edge shifts towards lower energies²⁸ so the experimental value of the energy gap is smaller than the theoretical one of 3.4 eV²⁹.

The reflectance absorption spectrum of GaN nanopowder (a), the experimental band gap (b) and the emission spectrum in the visible range upon λ_exc = 360 nm (c).

The wide band emission of GaN nanopowder in a range 405–650 nm with the maximum at 439 nm is shown in (Fig. 2c). Their deconvolution was performed to better understand the nature of the emission bands, resulting in five components with local maxima at 2.18 eV, 2.58 eV, 2.68 eV, 2.84 eV and 3 eV. The more intense bands 2.84 eV and 2.68 eV are most likely associated with V_Ga vacancies, and the remaining bands with V_N vacancies²⁸. The peaks at 3 eV, 2.84 eV are closest to the energy gap value, and may also be related to excitonic recombination³⁰. The broad peak at 2.58 eV and the residual peak at 2.68 eV possibly relate to the presence of luminescent centers causing edge shifts for gallium and nitrogen vacancies³¹. The obtained material is also characterized by an undesirable yellow component of luminescence, which is caused by structural defects, indicating the imperfect structure of the obtained material^2,32.

The research focuses on broadband white light emission (LIWE) characteristics and the accompanying photocurrent phenomenon. First, the studies show that a curve like a hysteresis loop can be observed by measuring the laser power density in the direction forward and backward. So far, two articles have been published attempting to explain the occurrence of hysteresis. These articles focus on carbon materials, diamond²⁴and graphene³³.

Due to its wide band gap, high thermal conductivity, and chemical stability, GaN is a suitable material for LIWE and photoconductivity investigation. The laser induced white emission spectra measured for GaN nanoceramics for both, visible and near-infrared regions are shown in Fig. 3a, b, respectively. It was characterized by broad bands whose intensity increases exponentially with excitation laser power density and leads to a blueshift of the bands. An increase in LIWE emission intensity starts after crossing the characteristic excitation threshold. The band extends from 400 to 2500 nm with a maximum of 1695 nm. For both regions, the emission threshold is estimated to be slightly above 1 kW/cm². The emission intensity measurements were obtained using two detectors, which is why the emission intensity can be different, and the gap between the visible and near infrared regions results from the use of optical filters and the cutting of the excitation laser beam. Nevertheless, based on these spectra and the N parameter, which in both cases takes similar values, it can be suggested that this is a single broad band. It has also been reported that the two bands measured separately are actually one band³⁴. Furthermore, experiments were performed for both regions, initiating hysteresis shape curve by measuring forward with increasing laser power density and then backward with decreasing laser power density. The results are shown in (Fig. 3c,d). In both cases, the characteristic properties of this phenomenon are preserved. The resulting loops are characterized by a threshold value followed by increased emission intensity. Moreover, saturation was demonstrated at high laser power densities. Furthermore, in the case of backward measurements, the N parameter for both ranges has higher values. Additionally, the emission threshold shifts towards higher power densities and saturation does not occur due to the lack of hysteresis. The SI (Figure S1) shows the CIE chromaticity diagram, demonstrating the objective color quality for both anti-Stokes and Stokes emissions. LIWE is characterized by a warm yellow color. Depending on the laser power density, the x and y coefficients locations differ, and as the laser power density increases, it is noticed that the emission color temperature changes towards white.

White emission intensity (a) near-infrared emission intensity (b) for GaN nanoceramics as a function of laser power density. The forward and backward cycles of LIWE in visible region (c) and near-infrared region (d). Inset the spectrum of LIWE without correction and the image of LIWE of light.

It was found that the investigated process is assisted by photoconductivity. Therefore, changes in sample resistance were measured depending on the laser power in cycles where the laser was turned on and off every 30 s. The changes were recorded at various voltage biases, as shown in (Fig. 4a). No changes in material resistance were observed at low laser powers, while at higher powers, a decrease in resistance was already observed. This is related to the threshold nature of both white emission and photocurrent. The dependence of resistance on the laser power density was plotted, and the excitation threshold was observed (Fig. 4b). Moreover, the applied voltage affects the initial resistance value. For 5 V, the resistance was the lowest, whereas the highest resistance was observed at 150 V. It should be noted that the resistance was dependent on the applied voltage up to 50 V, and then the saturation was observed. The initial value of resistance is similar for 50 V, 150 V, and 250 V. It can be observed that the photocurrent increases by two orders of magnitude with increasing excitation laser power. The excitation threshold was determined to be about 3 kW/cm² (see Table S1 in SI).

Photoresistance response measured in 30s cycles on/off at different voltage biases (a) and resistance dependence (b) for GaN nanoceramics.

White emission and photoconductivity were simultaneously performed in laser power density cycles (Fig. 5). The experiments were performed twice in closed (forward and backward) cycles. The differences can be noticed depending on whether the initial laser power density was low or high. For the first one, the hysteresis loop was observed for both LIWE and LIPC processes. However, hysteresis did not occur during measurements from a high initial laser power density.

The laser power dependence of integrated intensity of LIWE (a) and the maximum of photoresistance of LIPC (b) measured in two consecutive measurement cycles: forward and backward (b) for GaN nanoceramics.

The LIWE mechanism has been explained many times, including as blackbody radiation^35,36 thermal avalanche^14,19,37 or intervalence charge transfer^38,39,40,41. The characteristics of LIWE are mostly the same, so the mechanism should also be very similar. In this work, to explain the processes taking place, a proposed scheme (Fig. 6.) was used, in which the phenomenon is divided into stages: process before reaching the threshold value, multiphoton ionization and radiative recombination during irradiation after reaching the threshold value and possible occurrence of saturation with a high density NIR laser power.

Scheme of proposed mechanism responsible for LIWE divided into 3 stages: (1) Process before threshold value, (2) Process after threshold value including multiphoton ionization and radiative recombination, (3) Saturation process.

At low laser power densities, no light is observed until the threshold is exceeded. This is due to the number of photons delivered to the sample surface. GaN nanoceramics is excited with an NIR laser with an energy of 1.27 eV, while the experimental value of the energy gap is 3.17 eV. With the laser power density increase, the probability of photons hitting the same electron increases, ejecting it of the valence band. Then, the multiphoton avalanche process leads to multiphoton ionization on the surface of the tested material. The phenomenon is nonlinear and with the increase of the laser power density, the intensity of emission increases. The dependence of LIWE intensity I(P) on the excitation laser power is usually characterized by the power law formula:

where I(P) is the emission intensity, P is the excitation laser power, and N is related to the number of photons. This equation is used to describe multiphoton absorption transitions. Based on research in our group, where the dependence of power was plotted in different excited spot size it was found that the parameter N cannot be unambiguously treated as the number of absorbed photons. Based on this analysis, one should be more careful in connecting the nonlinearity of the process defined by the parameter N with the actual number of photons participating in the reaction^31,42. The process of multiphoton ionization (MPI) in irradiated spot at the surface of GaN nanoceramics leads to the broadband white emission (LIWE) assisted by an ejection of hot electrons e⁻ and may be described as:

where MA(Nħω) expresses the multiphoton absorption responsible for the multiphoton ionization of the GaN, GaN + is ionized GaN (cathode), LIWE represents laser induced white emission, and NR characterizes the nonradiative quenching and phonon emission processes contributing to the enhancement of thermally active processes. At higher laser power densities, more photons are transmitted to the sample surface, causing electrons to be emitted from the valence band to the conduction band. In addition, a radiative recombination process occurs. An electron from the conduction band recombines with a hole from the valence band, emitting a photon. This process is repeated until all possible emission centres are used. Saturation often occurs. The LIWE still occurs, but its intensity often remains unchanged or decreases. This is probably related to the confinement of some electrons, thus maintaining LIWE. The lack of new emission centres causes a lack of increase in emission intensity.

As explained above, during near-infrared laser irradiation of a sample in a dynamic vacuum, many processes occur, often overlapping. The hysteresis loop is probably formed not only due to photophysical processes but also because of morphological changes. Due to the fact that the measurements were performed depending on the density of the excitation laser beam on nanometric material, changes in the morphology of the sample may have occurred. When the measurement was performed from the lowest laser power, there could have been gradual ionization in GaN, an increase in temperature in the irradiated spot, and changes in morphology after obtaining a high power density. The following changes resulted in a different course when the laser power density decreased. This suggests an irreversible response of the material to the excitation condition. In turn, in the opposite case, the material was first irradiated with high laser power density, probably causing simultaneous ionization and another photophysical process explained in the proposed mechanism, changes in morphology, and high temperature at the beginning. Changing the density to a lower and then to a higher density does not cause changes in the emission intensity, which can confirm the conclusions drawn. Stręk et al.²⁴ reported the hysteresis loop behavior for LIWE on diamond material. They explain this as an irreversible process caused by multiphoton ionization. The degree of ionization is related to the irradiation of the sample and the number of ionized atoms. This phenomenon can be used as an effective optical memory. In our case, the confirmation of irreversible process can be repeated measurement of the loop under the same conditions and at the same point, and likewise, the backward cycle, because under high density, the hysteresis loop does not occur. Additionally, based on an article from Zheng et al.23, where an experiment was performed in which two curves were recorded, one for the excitation laser turned on and the other for the excitation laser turned off, a change in the emission intensity was observed, which suggested that the process is not only thermal.

The broadband white light emission phenomenon began to be associated with the photocurrent that appears during sample irradiation. To characterize LIPC, the same formula as in the case of LIWE can be used because both phenomena are nonlinear and exhibit a threshold character. The dependence of photoresistance R on laser power density can be expressed by:

where N_pc is related to the order of multiphoton ionization. The observed drop in resistance after reaching the threshold value is related to photoionization. After the material absorbs infrared photons, charge carriers are created. The higher laser density, the higher number of photons leading to stronger electron-hole recombination and, consequently, higher conductivity.

When discussing the LIWE and photocurrent phenomena, other factors are also considered, with one of the most important being the influence of temperature on the phenomena under study. LIWE was tested in many different luminescent materials, and it can be concluded that the shape of the emission does not depend on the host lattice. However, this phenomenon can be linked to the thermally assisted ionization process and strong optical nonlinearity. The differences resulting from the investigation are mainly associated with a change in the N parameter, depending on the material being tested. Interpretations of this behaviour should be considered. Since the experiments were performed on ceramics, the grains are as close to each other as possible, which reduces the energy localization in the laser spot and increases energy losses due to, for example, heat conduction and changes in the non-linearity of the process²³. The temperature of emission at different power densities was calculated using the Planck equation (see Table S2 in Supporting Information)

where h is Planck’s constant, c is the velocity of light, λ is the wavelength (nm), and k_B is Boltzmann’s constant. The spectrum is dependent on the temperature of the sample. Based on this calculation the temperature was fitting with a good comparison both in relation to the theoretic temperature values associated with blackbody radiation and in relation to the fit of these values in the CIE chromaticity diagram. Despite the good fit of the temperature to Planck’s law, LIWE is mainly caused by the sample ionization process. As shown in articles on other materials, significant differences are observed between the temperature values ​​during LIWE. Using measurements with a thermal camera, the maximum temperature is read at about 1230 K²³. In the case of other articles, measurements of luminescence nanothermometry, where the emission intensity ratio ²H_11/2 ◊ ⁴I_15/2 to ⁴S_3/2 ◊ ⁴I_15/2 of Er³⁺ ion is used as a temperature probe, the observed maximum temperature takes values ​​of about 900 K³⁹. Then, it can be assumed that such differences can be observed in these studies as well, at least because the temperature values based on black body radiation are directly related to a very small point on the sample surface during emission, while the thermal camera shows values where the spot is wider. In connection with this, the temperature distribution is not clearly defined, and the small size of the spot in the laser focus does not allow for direct, accurate measurement without luminescence thermometry. Here one should be careful because, as described above, this method gives values ​​much lower than the blackbody fitting. The question still remains how to properly determine the temperature, such studies are required in the future.

CAPE CANAVERAL, Fla. — The Hubble Space Telescope has captured the best picture yet of a high-speed comet visiting our solar system from another star.

NASA and the European Space Agency released the latest photos Thursday.

Discovered last month by a telescope in Chile, the comet known as 3I-Atlas is only the third known interstellar object to pass our way and poses no threat to Earth.

Astronomers originally estimated the size of its icy core at several miles (tens of kilometers) across, but Hubble’s observations have narrowed it down to no more than 3.5 miles (5.6 kilometers). It could even be as small as 1,000 feet (320 meters), according to scientists.

The comet is hurtling our way at 130,000 mph (209,000 kph), but will veer closer to Mars than Earth, keeping a safe distance from both. It was 277 million miles (446 million kilometers) away when photographed by Hubble a couple weeks ago. The orbiting telescope revealed a teardrop-shaped plume of dust around the nucleus as well as traces of a dusty tail.

Dunn writes for the Associated Press. The Associated Press Health and Science Department receives support from the Howard Hughes Medical Institute’s Department of Science Education and the Robert Wood Johnson Foundation. The AP is solely responsible for all content.

Space weather experts warn that a moderate-strength geomagnetic storm could hit Earth Friday or Saturday (Aug. 8 or 9) — possibly bringing auroras to northern U.S. skies.

On Wednesday (Aug. 6), the National Oceanic and Atmospheric Administration (NOAA) issued a geomagnetic storm watch and suggested that the storm is likely associated with a recent solar flare that erupted from an active region of the sun.

Not all giant meat-eating dinosaurs were the terrifying, bone-crushing predators we imagine. 

A new study published in the journal Current Biology has found that while many theropods – the group of bipedal carnivorous dinosaurs that includes Tyrannosaurus rex, Spinosaurus and Allosaurus – evolved massive bodies, only some developed the skull strength needed for high-powered bites.

Using 3D scans and computer modelling, researchers analysed the skull biomechanics of 18 theropod species. 

They discovered that while T. rex had a skull optimised for delivering enormous bite forces – capable of breaking through bone – other giants had comparatively weaker jaws and relied on different hunting strategies.

“Tyrannosaurids like T. rex had skulls that were optimised for high bite forces at the cost of higher skull stress,” said lead author Dr Andrew Rowe of the University of Bristol, UK. But in some other giants, like Giganotosaurus, we calculated stress patterns suggesting a relatively lighter bite. It drove home how evolution can produce multiple ‘solutions’ to life as a large, carnivorous biped.”

Rather than following a single evolutionary route to apex status, giant carnivores evolved a variety of skull shapes and feeding styles. Some, like T. rex, bit down like a crocodile. Others, such as Allosaurus or Spinosaurus, used slashing or ripping techniques more comparable to modern Komodo dragons or big cats.

“Tyrannosaurs did something different,” Steve Brusatte, a professor and palaeontologist at the University of Edinburgh who was not involved in the study, told BBC Science Focus. “They supersized their heads and developed enormous bite forces, enough to break the bones of their prey. This was a risky and high-stress lifestyle, quite literally, as it put huge stress on its skull bones and muscles.”

The findings also challenge the assumption that bigger dinosaurs had stronger bites. Some smaller species actually placed more stress on their skulls due to increased muscle volume. This suggests that size alone wasn’t the main factor in bite power.

The diversity in bite strength and skull design suggests that dinosaur ecosystems were more specialised than previously assumed, with multiple ways to dominate the prehistoric food chain.

“There wasn’t one ‘best’ skull design for being a predatory giant; several designs functioned perfectly well,” Rowe said. “This biomechanical diversity suggests that dinosaur ecosystems supported a wider range of giant carnivore ecologies than we often assume, with less competition and more specialisation.”

Read more:

About our expert

Steve Brusatte is a professor and palaeontologist at the University of Edinburgh and the author of the book The Rise And Reign Of The Mammals (£20, Picador), a 325-million-year odyssey of mammalian evolution and the people who study mammal fossils.