Devil rays can move through a surprisingly large part of the water column. Some Mobula species are mainly associated with shallow water, while others have been recorded at depths approaching 2,000 m. That raises an interesting mechanical question: Does the cartilage making up their skeleton change between species that occupy such different environments?
A 2026 Journal of the Royal Society Interface study looked at neurocranial cartilage from four Mobula species. Ray skeletons are largely cartilaginous, but this tissue is not simply a uniform soft matrix. Parts of it are covered by small mineralized tiles called tesserae, which add another layer to how the tissue carries load.
The researchers approached the problem from a few different scales. Their cartilage mechanical testing included compression and stress relaxation of larger biopsies, followed by micro-indentation and nano-indentation on smaller regions of the tissue.
For the micro-indentation measurements, they used a MicroTester G2. This gave them a closer look at local stiffness, including measurements of the cartilage with its layered structure intact and again after the tessellated layer had been removed.
There was no single compositional measurement that neatly explained everything they saw. Mineral density was fairly similar between the species, and glycosaminoglycan content did not track the mechanical results particularly well. Water content was more closely related. The cartilage of the deeper-ranging species tended to be more compliant, while samples containing more water generally had lower stiffness, strength, and toughness.
Why compare cartilage mechanical testing results across different diving depths?
The four ray species occupied very different reported depth ranges:
- Mobula eregoodoo: 0 to 50 m
- Mobula thurstoni: 0 to 100 m
- Mobula mobular: 0 to 1,112 m
- Mobula tarapacana: 0 to 1,896 m
Those ranges matter because pressure changes quickly once an animal starts moving deeper. For roughly every 10 m of seawater, hydrostatic pressure rises by about 0.1 MPa. The authors wanted to see whether cartilage from the deeper-ranging species also behaved differently when it was mechanically tested.
Of course, depth is only one part of the picture. These are different species, so differences in development, tissue structure, hydration and habitat could all contribute to what was measured. The depth range still gave the researchers a useful way to compare the mechanical behaviour of animals living under quite different conditions.
The material they were working with is unusual as well. Shark and ray skeletons are mostly cartilage, but parts of that cartilage are reinforced by small mineralized tiles called tesserae. A fibrous outer layer, the perichondrium, sits over the tessellated surface. So even within one sample, the indenter may be interacting with several distinct structural layers.
For this study, the team sampled the neurocranium. They chose broad, relatively flat areas where muscular loading was less likely to dominate the tissue response. Those areas were then divided into smaller, repeatable regions so that the mechanical tests and compositional measurements could be made from comparable parts of each specimen.
A multiscale approach to cartilage mechanical testing
One thing that stands out in this study is that the researchers did not rely on a single stiffness measurement.
At the bulk scale, 5 mm cylindrical biopsies were subjected to unconfined compression testing. Stepwise compression was used to examine time-dependent relaxation, while separate compression-to-failure tests provided measures including Young’s modulus, ultimate strength and toughness. All of these tests were conducted with the samples submerged in PBS.
The stepwise tests also provided a direct look at stress relaxation. Each increase in strain produced an immediate rise in compressive stress followed by a gradual decline while deformation was held. That response is expected in hydrated cartilage, where both the solid matrix and movement of fluid contribute to the time-dependent mechanical behaviour.
The researchers then went to smaller length scales. Local micro-mechanical testing was performed using the MicroTester G2, followed by nano-indentation with a separate nanoscale system.
Using several scales matters here because bulk cartilage mechanical testing and local indentation are not necessarily answering exactly the same question. A compression-to-failure test samples the behaviour of the specimen as a whole. An indentation measurement probes a much smaller region and can be used to compare different layers or tissue components.
Cartilage mechanical testing showed a shift toward more compliant tissue
The bulk compression results showed a fairly clear pattern, although it was not perfectly monotonic across every species and measurement.
At high strains, Young’s modulus was highest in the shallow-ranging M. eregoodoo, at 53.5 MPa. The deepest-ranging species, M. tarapacana, had a mean high-strain modulus of 10.2 MPa. Ultimate strength followed a similar pattern, decreasing from 15.6 MPa in M. eregoodoo to 3.3 MPa in M. tarapacana.
Toughness also differed significantly between species. In the compression-to-failure curves, M. eregoodoo, M. thurstoni and M. mobular generally reached higher stresses than M. tarapacana. At lower strain, however, differences in Young’s modulus were much less obvious.
That distinction between low- and high-strain behaviour is useful. The cartilage did not behave like a simple linear elastic material with one stiffness value that described everything it did.
Stress relaxation told a related but somewhat different story. Relaxation times generally became longer across species toward the deeper-ranging M. tarapacana, particularly as strain increased. The tissue was therefore not simply weaker or less stiff. Its time-dependent response was also changing.
For researchers working more broadly in cartilage biomechanics and tissue engineering, this is one reason cartilage mechanical testing often extends beyond a single modulus value. Stiffness, relaxation, strength and loading history can each describe a different part of how hydrated cartilage responds.
How was micro-indentation used to measure cartilage stiffness?
The MicroTester portion of the study provided a more localized measurement.
The researchers prepared 2 mm diameter cartilage biopsies and tested them in a PBS bath at 25°C. A 1 mm zirconium oxide spherical bead was attached to a 1 mm diameter CellScale cantilevered tungsten microbeam.
Each specimen underwent three load-unload cycles to an indentation depth equal to 10% of the sample height. Each loading phase lasted 30 seconds and was followed immediately by 30 seconds of recovery. Sample height was measured with the integrated optical system while force and displacement were recorded continuously. The third loading cycle was used to calculate Young’s modulus with a Hertz contact model.
This is a useful example of indentation testing because the researchers were able to compare two related structures from the same tissue.
First they tested the complete layered specimen, including the perichondrium, tesserae and unmineralized cartilage. They then removed the tessellated layer and repeated the measurements on the unmineralized cartilage.
The resulting modulus differed significantly among species for both conditions. The deeper-ranging M. tarapacana showed some of the lowest local modulus values, while the shallower species tended to be stiffer. Interestingly, the difference between the full layered sample and isolated unmineralized cartilage was relatively small. The authors suggest that this may reflect a buffering effect from the perichondrium or a limited contribution of the tesserae under this particular local loading condition.
Because the specimens remained submerged during testing, the experiment is also a good example of hydrated mechanical testing. For cartilage and other highly hydrated biological materials, keeping the sample in a controlled fluid environment can be important because water is part of the mechanical system being measured.
The same general challenge appears in other small biological specimens. For example, researchers have also used microscale force measurements for mechanical testing of tissue spheroids, where conventional bulk fixtures would be poorly matched to the size of the sample.
Water content tracked the cartilage mechanical testing results
After seeing differences in stiffness, strength and relaxation, the next question was what might be changing within the tissue.
The researchers looked at three possibilities: glycosaminoglycan content, tissue mineral density and water content. None of them followed depth in a perfectly clean way.
Mineral density, for example, was fairly similar across the four species. GAG content varied, but those differences did not line up well with the bulk mechanical measurements.
Water content was more interesting.
The samples with more water generally had lower high-strain stiffness, lower toughness and lower ultimate strength. That relationship showed up statistically across the species, whereas the same pattern was not seen for mineral density or GAG content.
There was also a difference between the ends of the depth range. M. tarapacana, the deepest-ranging species in the study, had a mean total water content of about 84%. For the shallow-ranging M. eregoodoo, it was closer to 80%.
That does not mean greater diving depth caused the cartilage to become more hydrated. The study cannot really answer that. What the authors could say was that the deeper-ranging species tended to have cartilage that was both more hydrated and more compliant. Changes in collagen organization, development or other parts of the matrix may also be involved.
For cartilage mechanical testing, the result is a useful reminder that water is not just something surrounding the sample during a test. In a tissue such as cartilage, it is part of the material being measured, and differences in hydration can show up directly in the mechanical response.
Does formalin fixation change cartilage mechanical properties?
There was another practical problem the researchers had to work around.
These ray samples were not easy to come by. Mobulid rays are protected, many live offshore, and collecting fresh tissue is rarely straightforward. In some cases, specimens are available only after they have been fixed for transport or storage.
That matters because fixation can change the tissue itself. If the cartilage has already been sitting in formalin, the number measured during a mechanical test may not match what would have been measured in fresh tissue.
The researchers checked this separately by comparing frozen and frozen-fixed neurocranial samples from three cartilaginous fish species.
The clearest change appeared in M. eregoodoo. At high strain, Young’s modulus was 89.6 MPa in the frozen samples and 33.8 MPa after fixation. Ultimate strength went from 32.1 to 11.8 MPa, and toughness from 7.9 to 2.6 MPa.
The size of the change was not the same for every species. Even so, the authors treated the fixed-tissue values cautiously, since they likely underestimate the properties of the native cartilage. The comparisons between species were still useful, though, because the overall ordering of the mechanical response was largely retained.
It is an easy experimental detail to overlook. How a tissue is stored before testing can affect the numbers that come out at the end. Fixation, freezing, hydration and the time spent in storage are therefore worth considering alongside the loading protocol itself.
For readers working with a wider range of tissues and biomaterials, our overview of mechanical testing of biomaterials discusses some of the broader considerations that go into choosing and interpreting mechanical tests.
What this study adds to cartilage mechanical testing
There is an appealing simplicity to the depth comparison, but the details are less tidy.
The deeper-ranging rays generally had more compliant cartilage. Water content tracked the changes better than mineral density or GAG content. Local MicroTester measurements broadly agreed with the whole-tissue mechanical trends.
At the same time, there were only two individuals from each Mobula species. The samples had been fixed, and the fixation study showed that preservation itself changes absolute mechanical values. The authors therefore treat the results as an initial view of a pattern rather than a definitive demonstration of pressure-driven skeletal adaptation.
That caution is worth keeping.
What the study does provide is an unusual example of cartilage mechanical testing across scales. Bulk compression describes how the complete specimen carries load. Stress relaxation adds the time-dependent response. Micro-indentation looks at local tissue stiffness. Compositional measurements then give the researchers something to compare those mechanical differences against.
A similar multiscale question comes up in engineered cartilage as well. In previous work highlighted on the CellScale blog, researchers used cell-only bioprinting to organize articular cartilage-like tissue. The biological system is very different, but the underlying challenge is familiar: the organization and composition of a tissue can change the way it responds mechanically.
Using the MicroTester for small-scale cartilage mechanical testing
The MicroTester is often used when the part of the sample you care about is quite small, or when a bulk test would average over too much of the tissue. In this study, the researchers used it to indent small cartilage biopsies while they were submerged in fluid. A spherical probe was brought into contact with the tissue, then force and displacement were recorded over repeated loading cycles.
Those measurements were used to calculate a local Young’s modulus and compare the different species and tissue layers.
That type of setup can be useful for cartilage mechanical testing, but the same approach can be applied to other small or delicate biological specimens where forces are low and sample geometry is limited. The instrument’s integrated imaging is used to position the sample and observe deformation, while interchangeable microbeams provide the force measurement appropriate to the test.
Researchers interested specifically in localized loading can find more detail on CellScale’s indentation testing and micro-mechanical testing pages. For experiments where the material needs to remain immersed throughout the protocol, the hydrated testing page covers the same general testing consideration used in this study.
The devil ray samples make a good example of why this kind of testing exists in the first place. These were only 2 mm biopsies from a tissue that is difficult to obtain, and the researchers were interested in local differences rather than only the response of the whole specimen. A conventional bulk test can answer part of that question. The micro-indentation work filled in another part.
Citation
Theda HinrichsJingxiao ZhongChun-Wei ChangGobiraj RamajeyamAnusha NeranjanShahrouz AminiMason DeanShaun CollinDietmar W. HutmacherTravis J. KleinVictoria Camilieri-Asch; Biomechanics and composition of fixed neurocranial tessellated cartilage from the neurocranium of four mobulid ray species with different depth ranges. J. R. Soc. Interface. 1 September 2026; 23 (242): 20251331. https://doi.org/10.1098/rsif.2025.1331
