Bat guano is a major food source for adult Astyanax in passages accessible to bats, where particles may float on the surface or sink to the floor. Many cavefish populations, however, inhabit underwater passages beyond sumps that bats cannot reach. Explorations in January 2020 at Tinaja Cave, in February 2023 at Sabinos Cave and in November 2023 at the Calera system examined deeper or less accessible areas.

Divers using underwater lamps observed the first two sites, while remotely operated vehicles explored more than 120 metres of underwater gallery in the Calera system. These surveys identified the mysid shrimp Spelaeomysis quinterensis as a potential food source in galleries without bat droppings. The shrimp were abundant in some of the explored areas and were observed mainly crawling on walls and ceilings, with fewer on the floor.

Laboratory work cited by the researchers showed that Astyanax eat commercially supplied mysid shrimp, and earlier field observations had recorded cavefish consuming a mysid placed in a tank. The new field evidence therefore linked the distribution of a known prey type with the physical surfaces along which cavefish swim, although the study did not quantify how much of their diet these shrimp contribute or directly observe the shrimp being captured in the surveyed passages. To measure body posture, the researchers divided a 5-gallon tank with a vertical glass panel and placed a camera perpendicular to that panel.

Fish swimming parallel to the glass were therefore recorded moving toward or away from the camera. The enucleated fish had therefore navigated without eyes throughout their lives, allowing the researchers to compare experimentally blind surface fish with fish whose ancestors had evolved in darkness. After at least 30 minutes of acclimation, the researchers recorded each group for five minutes in each tank compartment.

From the recordings, the team extracted clips showing fish swimming parallel to the wall and measured the dorsoventral axis against the vertical. They also noted whether the left or right side of each fish faced the glass. The analysis comprised 170 clips from Pachón cavefish, 177 from Tinaja cavefish, 90 from surface fish and 91 from enucleated surface fish.

Mann–Whitney U tests compared the cave populations with surface fish and compared right-facing and left-facing swims within each population. Because the study sampled swimming posture rather than attacks or feeding outcomes, the resulting values describe the geometry of wall-associated swimming and do not by themselves demonstrate that the posture is a hunting strategy. Surface fish in darkness held their bodies at an average angle of 3.2 degrees from vertical, with a standard deviation of 11.6 degrees.

The enucleated surface fish had a similar mean of 3.1 degrees, with a standard deviation of 6.5 degrees, and their posture did not differ significantly from that of the eyed surface fish in darkness (P = 0.560). By contrast, Pachón cavefish tilted away from the wall by 14.8 ± 9.9 degrees and Tinaja cavefish by 10.4 ± 11.1 degrees. Each cave population differed from the surface-fish group with P < 0.00001.

These comparisons indicate that loss of vision alone did not reproduce the mean posture of cave-adapted fish under the laboratory conditions used in the experiment. Side-dependent differences appeared in both cave populations. Pachón cavefish averaged 16.6 degrees when their right side faced the wall, compared with 12.9 degrees when their left side faced it; the difference had P = 0.012.

Tinaja cavefish averaged 12.6 degrees with the right side toward the wall and 7.8 degrees with the left side toward it, again with P = 0.012. Surface fish did not show a statistically significant side difference: their means were 6.1 degrees with the right side facing the wall and 1.0 degree with the left, with P = 0.214. The study thus measured a consistent directional asymmetry in cavefish body inclination, while finding no comparable significant asymmetry in the sampled surface fish.

The authors proposed, but did not prove, a sensory explanation for the larger cavefish tilt. Astyanax can construct hydrodynamic images of their surroundings and detect changes in water flow around the body through the lateral-line system. In a previously described mathematical model of larval prey capture, the detection range on each side of the body was represented as a semicircle or hemisphere.

The researchers suggest that a fish swimming upright beside a wall can scan the nearby surface primarily with the field on the side facing it. Tilting the body away from the wall may bring part of the opposite side’s receptive field into the detectable area, potentially enlarging the wall region that can be scanned. Direct measurements of sensory coverage, flow detection or prey detection would be required to test that proposal.

A second hypothesis addresses why the measured angles differed between left- and right-facing swims. Earlier work found that the lateral-line receptive fields on the two sides of cavefish can have different radii and that strike performance differs between sides. Related observations include a leftward-biased bend in the dorso-cranial skeleton, a positive correlation between vibration-attraction behavior and the number of left-side superficial neuromasts in Pachón cavefish, and population-level “handedness” in approaches to vibrating sources.

Under the proposed model, a cavefish facing a wall with a smaller receptive field may gain a proportionally greater scanning benefit by tilting, whereas one facing the wall with a larger field may gain less. The present data are consistent with that prediction but do not establish that anatomical or physiological asymmetry caused the posture. Together, the laboratory measurements and field observations provide a broader picture of wall-associated behavior without demonstrating a complete foraging mechanism.

Cavefish are already known to prefer remaining near walls, and earlier experiments found that this preference can persist across straight, concave and convex surfaces. The current study adds precise measurements of body inclination and shows that the mean posture differs between cave-adapted fish and both sighted and experimentally enucleated surface fish. Its field observations also show that mysid shrimp can occupy walls and ceilings in passages beyond bat access.

Whether cavefish use the measured tilt to improve prey detection, what the complete receptive fields look like in adult wall swimmers, and which physiological or morphological factors produce the side-dependent angles remain open questions requiring direct behavioral and sensory testing.

Subject of Research: Laboratory measurements and field observations of wall-associated body inclination in Pachón and Tinaja Astyanax cavefish.

Article Title: Cavefish dorsoventral axis angle during wall swimming: laterality asymmetry

Article References: Espinasa, J., & Espinasa, L. (2024). Cavefish dorsoventral axis angle during wall swimming: laterality asymmetry. Subterranean Biology, 49, 19-29. https://doi.org/10.3897/subtbiol.49.121747

Image Credits: AI Generated

DOI: 10.3897/subtbiol.49.121747

Keywords: Astyanax cavefish, wall swimming, dorsoventral axis, laterality, Pachón cavefish, Tinaja cavefish, lateral line, underground ecology, mysid shrimp, sensory ecology, Cavefish, dorsoventral

Motor differences in autism have long been recognized but poorly measured. Autistic children and adults tend to score lower on classic dexterity assessments such as the Purdue Pegboard test, inserting fewer pegs into small holes within a time limit with either hand. Maximum finger tapping speed is reduced, and grip strength tends to be smaller. Studies of walking have documented larger step-to-step variability and abnormal gait patterns in autistic children, while autistic adults show less stable postural control. Yet these observations have been gathered with a patchwork of different tests, expensive motion capture systems, and separate assessments for arms and legs, making it nearly impossible to ask whether the same underlying factor drives motor differences throughout the body.

The new study, led by Atsushi Takagi and Chihiro Itoi of NTT Communication Science Laboratories together with colleagues, set out to close that gap with a single, common framework. The team built on a motor assessment previously validated on more than 1,600 people aged three to eighty-eight, which measures how consistently a limb repeats a circular movement. In the test, participants hold a smartphone and trace a ten-centimeter circle on a guide sheet, completing one revolution per metronome beat at 2.5 hertz for the hands. For the feet, the phone is strapped in an armband above the ankle, and the participant swings the straight leg from the hip to trace a fifteen-centimeter circle on the floor at a slower 1.5 hertz pace. Each limb is tested for just fifteen seconds.

The technical trick lies in how the raw accelerometer signal is turned into a measure of skill. The three-axis acceleration data are first centered and normalized, then rotated using principal component analysis so that the circular motion lies mainly in a two-dimensional plane. Each revolution is identified by successive crossings of the movement phase, and the similarity between neighboring revolutions is quantified with the Hausdorff distance, a metric that is fast to compute and robust to differences in sampling. Averaging the Hausdorff distances across all adjacent revolution pairs yields a single variability score for each limb. When every circle looks alike, variability is low and control is judged robust; when each revolution wanders along a slightly different path, variability is high and control is less consistent.

Crucially, the researchers first confirmed that both groups kept tempo equally well. Seventeen clinically diagnosed autistic adults, with a mean age of twenty-eight, and twenty age- and IQ-matched comparison participants traced circles at statistically indistinguishable speeds for every limb. This mattered because movement variability is known to rise with speed, a relationship described since Fitts’s classic 1954 work, so any group difference in variability could only be interpreted fairly if speed was matched. Prior studies have also found that autistic individuals can keep tempo with a single-frequency external rhythm without difficulty, and the present results are consistent with that picture.

When variability itself was compared, a clear asymmetry emerged. For the hands, statistical analysis revealed significant effects of dominance, group, and their interaction. Follow-up tests showed that the autistic group’s non-dominant hand was significantly more variable than that of the comparison group, while the dominant hand showed no reliable difference. For the feet, the pattern was broader: the autistic group showed greater variability across both the dominant and non-dominant lower limbs, with no specific dominance interaction. In other words, the upper limb difference was confined to the non-dominant side, whereas the lower limb difference appeared globally. The autistic group also displayed significantly greater hand laterality, meaning the gap between their dominant and non-dominant hands was larger than in comparison participants.

Perhaps the most intriguing result came from linking the limbs together. Using a linear mixed-effects model, the researchers asked whether a person’s foot variability predicted their hand variability after accounting for group, dominance, and their interaction. It did: the association was positive, with a slope of 0.28, meaning that individuals with wobblier feet also tended to have wobblier hands. Notably, this relationship held equally in both groups, since adding a group-by-foot-variability interaction did not improve the model. The authors are careful to stress that this does not prove a causal link or a single shared neural mechanism, because the group effects differed between the upper and lower limbs. Still, it suggests that common factors, such as motor timing variability, sensory variability, or habitual limb use, may influence movement consistency throughout the body, coexisting with limb-specific effects.

The increased hand laterality in the autistic group came as a surprise. Earlier questionnaire studies and meta-analyses have suggested that autistic individuals are more likely to be mixed-handed, which led the team to predict smaller, not larger, differences between the two sides. Yet on the Edinburgh Handedness Inventory both groups reported equally strong right-handedness, and a slight difference in footedness on the Coren quotient vanished after correction for multiple comparisons. The authors propose that what people say about their hand preferences may diverge from how skillfully they actually use each hand, and that questionnaire measures suffer from ceiling effects that blunt their sensitivity. Motor-based assessments, by contrast, may capture real differences in lateralized skill that self-report misses.

Why would the non-dominant hand and both feet be selectively affected? The researchers weigh several possibilities. One candidate is heightened sensory variability: autistic individuals show larger trial-to-trial variability in sensory percepts, and tasks like peg insertion that demand tactile processing are disproportionately affected. But it is unclear why such a difference would localize to the non-dominant side alone. Another candidate is motor timing. Autistic children show reduced cerebellar activity and connectivity during finger tapping, and the cerebellum is critical for precise timing; timing variability is itself known to be larger in the non-dominant hand. A third possibility involves practice and generalization. Autistic children have been reported to use their non-dominant hand less often, and previous work suggests that motor learning in autism may generalize poorly across limbs, with internal models formed in a highly limb-specific way and an increased reliance on proprioceptive feedback. Skills refined through years of dominant-hand use may simply fail to transfer, leaving the non-dominant hand and the comparatively unfamiliar feet less robustly controlled. The authors caution that habitual use was not measured, so this remains a hypothesis.

The practical implications may prove the most consequential part of the story. Because the test requires only a smartphone, fifteen seconds of movement, and no laboratory equipment, it could be deployed in clinics, schools, or homes as a rapid, quantitative probe of coordination. Related work by the same group has linked circular leg movement variability to postural stability, suggesting the measure carries functional meaning beyond the task itself. The study does have limits: nearly all participants were right-handed, so laterality findings need testing across a range of handedness, and only adults were studied, leaving open when in development these differences emerge. Even so, the image of autistic motor control that emerges is more nuanced than a simple deficit. Dominant-hand skill is intact, tempo-keeping is intact, and yet the less-practiced corners of the motor repertoire betray a measurable, quantifiable difference, one that a device in nearly every pocket can now detect.

Subject of Research: Motor variability during circular movements as a measure of limb control differences in autistic adults

Article Title: Motor Variability of Circular Movements Reveals Differences in Non-Dominant Hand and Foot Control in Autistic Adults

Article References: Takagi, A., Itoi, C., Honda, K., & Kashino, M. (2026). Motor Variability of Circular Movements Reveals Differences in Non-Dominant Hand and Foot Control in Autistic Adults. Journal of Autism and Developmental Disorders. https://doi.org/10.1007/s10803-026-07546-3

Image Credits: AI Generated

DOI: 10.1007/s10803-026-07546-3

Keywords: autism spectrum disorder, motor variability, sensorimotor control, smartphone accelerometer, non-dominant hand, lower limb control, laterality, Hausdorff distance, motor timing, cerebellum, clinical motor assessment, adults