Ground Reaction Forces in Monitor Lizards (Cieri et al., 2021)

Quick Facts

Paper Overview

How does a 26-kilogram reptile that walks with a bent-legged, sprawling gait avoid breaking its own bones? That question sits at the heart of a 2021 study published in Biology Letters by Robert L. Cieri, Taylor J. M. Dick, Robert Irwin, Daniel Rumsey, and Christofer J. Clemente at the University of the Sunshine Coast, Australia. The paper's full citation is: Cieri, R. L., Dick, T. J. M., Irwin, R., Rumsey, D., & Clemente, C. J. (2021). "The scaling of ground reaction forces and duty factor in monitor lizards: implications for locomotion in sprawling tetrapods." Biology Letters, 17(2), 20200612. DOI: 10.1098/rsbl.2020.0612.

The significance of this study lies in its exceptional size range. By including specimens spanning nearly 3.5 orders of magnitude in body mass — from the pygmy monitor Varanus brevicauda at just 7 grams to a full-grown Varanus komodoensis at 26.55 kilograms — the research team was able to test, across real animals, competing mathematical predictions about how ground reaction forces should scale as a sprawling tetrapod grows larger. The Komodo dragon is not merely a backdrop here: it is the largest data point and therefore most diagnostic for distinguishing between rival scaling models.

The Scaling Problem in Sprawling Tetrapods

To appreciate why Cieri et al.'s study is significant, it helps to understand the fundamental mechanical problem that sprawling locomotion poses as body size increases. In upright mammals such as horses and elephants, the limbs act as near-vertical columns beneath the body, transmitting weight efficiently to the ground. As mammals grow larger they tend to straighten their limbs further — a strategy called columnar posture — which reduces the bending moments experienced by bones and muscles. This is one reason elephants can reach several tonnes without structural failure.

Monitor lizards, and sprawling reptiles generally, keep their femora and humeri held laterally outward from the body, closer to horizontal than vertical. This creates large bending moments at each limb joint with every step, and those bending moments scale with body mass. Naive geometry predicts that peak ground reaction forces should scale in proportion to body weight (M^1.0). Because bone strength scales approximately as M^0.67 — governed by cross-sectional area — a purely geometric scaling would eventually produce forces that bones cannot bear. Something must give way; but what?

Two competing hypotheses existed before Cieri et al.'s study. The first, the geometric force-production hypothesis, predicted that muscle force output could be maintained relative to body weight by increasing muscle cross-sectional area isometrically, keeping peak forces near M^1.0. The second, the body-support hypothesis, predicted that sprawling animals would need to generate forces above M^1.0 simply to prevent their bellies from dragging on the ground — a steeper and potentially unsustainable scaling. Distinguishing between these two predictions required data from animals with a very large size range. The Varanus genus, with its extraordinary intrageneric size diversity from pygmy species to the Komodo dragon, offered precisely that range.

Methods: Force Plates, High-Speed Video, and Machine Learning

The experimental design combined three complementary measurement techniques that together captured both the forces acting on the ground and the kinematics of the limb producing those forces. Understanding the methods is essential for interpreting the findings correctly.

Force-Plate Instrumentation

The research team used three separate force-plate systems sampling at between 1,000 and 10,000 Hz. As each lizard walked or trotted across the platform, the plate recorded the three-dimensional ground reaction force vector at millisecond resolution: a vertical component supporting the animal's weight, a fore-aft braking/propulsive component, and a mediolateral component related to the sprawl angle. Impulse — the integral of force over time for a complete stance phase — was computed for each limb separately, allowing the team to ask not just how hard the animals pushed on the ground, but for how long.

High-Speed Kinematics

Synchronized high-speed cameras operating at 250 frames per second captured limb movement simultaneously with force-plate data. This allowed the investigators to determine duty factor: the fraction of a complete stride cycle during which a given limb remains in contact with the ground. Duty factor is a dimensionless gait descriptor — a value of 0.5 means a limb spends exactly half the stride in stance, half in swing. Higher duty factor generally corresponds to slower, more cautious locomotion; lower duty factor (approaching 0.5 and below) is associated with higher speeds and, eventually, aerial phases that define true running gaits.

DeepLabCut Pose Estimation

To track body landmarks across dozens of individuals of very different sizes and scale accurately enough for quantitative analysis, the team employed DeepLabCut — a machine-learning framework originally developed for neuroscience that trains a neural network to identify specific body points (hip, knee, ankle, toe, etc.) in video frames without manual digitisation of every image. This approach substantially increased throughput and consistency compared to traditional hand-digitisation, and it represents a methodological advance that is increasingly standard in comparative biomechanics.

Statistical Analysis

All scaling relationships were estimated using mixed-effects linear models in log-log space, with body mass as the primary predictor and individual identity included as a random effect to account for repeated measurements from the same animal. Relative speed — speed normalised by hip height to allow fair comparison across animals of different sizes — was included as a covariate so that the team could separate the effect of being large from the effect of simply moving slowly.

Key Findings: Forces, Duty Factor, and Posture

Ground Reaction Forces Scale Between the Two Competing Predictions

The most consequential result of the study is that scaling of both vertical impulse and peak vertical force fell between the geometric force-production prediction and the body-support prediction — closer to the former but not perfectly matching either. Vertical impulse scaled at M^0.99–1.34 and peak vertical force at M^0.73–1.00, depending on limb and direction. These exponents confirm that larger varanids do experience proportionally greater ground reaction forces than smaller ones, but the increase is not catastrophically steep.

Crucially, the hindlimb and forelimb scale differently. Hindlimb vertical impulses showed a steeper exponent (M^1.32) than forelimb impulses (M^1.19), meaning the hindquarters bear a progressively greater share of body weight as varanids grow larger. In other words, the centre of mass shifts toward the tail with increasing body size — a geometric consequence of the trunk elongating relative to the limbs, and a pattern that has important implications for the heavy-bodied Komodo dragon specifically.

Duty Factor as a Load-Management Strategy

Duty factor increased significantly with body mass across the species studied (scaling M^0.04). While this exponent sounds small, it translates meaningfully across a body-mass range spanning 3.5 orders of magnitude: a Komodo dragon at the extreme upper end spends markedly more of each stride with its hindlimb on the ground than does a small Varanus. This is the mechanically elegant solution. By extending stance duration — keeping the foot planted longer — the animal distributes the same impulse over more time, keeping instantaneous peak force below critical fracture thresholds. The Komodo dragon does not need an elephant's erect posture to survive the forces of walking; it simply walks with a prolonged ground contact.

No Shift to an Upright Posture

A key negative finding — equally important — is that larger varanids do not adopt a more erect limb posture to compensate for higher forces. Unlike the mammalian pattern, where limb angle becomes more columnar as body mass increases, the data from Cieri et al. show varanids maintaining a sprawling posture across the full size range studied. The limb geometry does not change in the way that would reduce bending moments. Instead, the positive allometry of locomotor musculature — larger animals have disproportionately large muscles for their size — provides the additional force-generation capacity needed to sustain sprawling locomotion at high body mass. This finding aligns with a companion muscle-anatomy study from the same group (Cieri, Clemente et al., 2020, Journal of Anatomy, DOI: 10.1111/joa.13273), which documented widespread positive allometry in pectoral-girdle muscles of varanids.

The Respiratory Constraint: Gular Pumping & the Axial Paradox

A separate but deeply related body of research reveals that the biomechanical challenges of varanid locomotion extend beyond the skeleton and muscles to the respiratory system itself. The foundational paper here is: Owerkowicz, T., Farmer, C. G., Hicks, J. W., & Brainerd, E. L. (1999). "Contribution of gular pumping to lung ventilation in monitor lizards." Science, 284(5420), 1661–1663. DOI: 10.1126/science.284.5420.1661.

Most lizards that locomote using lateral body undulation — the sinusoidal side-to-side bending of the trunk — face what physiologists call the axial constraint. During each stride, the lateral bending of the ribcage alternately compresses and stretches the lungs, interfering with the costal aspiration (rib-driven breathing) that lizards normally use at rest. The result is that running lizards cannot breathe efficiently simultaneously with locomotion: oxygen debt accumulates rapidly, and the animal is forced to stop and recover. This constraint is one of the principal reasons that most lizard species are burst sprinters rather than endurance runners.

Owerkowicz and colleagues demonstrated that monitor lizards have evolved a physiological workaround: a positive-pressure gular pump. The gular region — the loose skin and musculature beneath the lower jaw and throat — acts as a bellows. During locomotion, rhythmic contraction of the gular muscles forces air from the throat cavity into the lungs at positive pressure, supplementing or temporarily replacing costal aspiration and maintaining adequate oxygen delivery even as the ribcage is mechanically constrained by running movements.

The team tested this by experimentally immobilising the gular region in Varanus exanthematicus (the savannah monitor) and measuring lung ventilation during treadmill locomotion. With gular pumping blocked, the animals immediately showed evidence of the axial constraint: ventilation fell and oxygen saturation dropped. With the gular pump intact, the constraint was effectively masked. This single finding distinguishes varanid stamina from that of other lizards: the gular pump is the physiological key that allows monitors — including the Komodo dragon — to sustain moderate-intensity locomotion far longer than a typical squamate of comparable size could manage.

For the Komodo dragon, this has direct ecological relevance. The dragon's documented ability to track prey over kilometres and to engage in extended feeding competitions at carcasses depends on an aerobic capacity that would be impossible without respiratory compensation for the axial constraint. The gular pump is not merely an anatomical curiosity; it is a functional prerequisite for the species' predatory and scavenging ecology.

Sprint Speed, Body Mass, and the Optimal-Size Hypothesis

A third dimension of varanid locomotion research addresses the question of how maximum sprint speed varies with body size. The key reference here is: Clemente, C. J., Thompson, G. G., & Withers, P. C. (2009). "Evolutionary relationships of sprint speed in Australian varanid lizards." Journal of Zoology, 278(4), 270–280. DOI: 10.1111/j.1469-7998.2009.00559.x.

Clemente and colleagues measured sprint speed for 18 Varanus species across a wide range of body masses and fitted both linear and curvilinear scaling models. Their key finding was that sprint speed does not simply increase with body mass indefinitely. Instead, the relationship follows a hump-shaped curve with an optimum around 2.83 kg — the mass at which a varanid can achieve the highest maximal speed. Below that optimum, smaller animals are slower than expected because their absolute stride length is short. Above it, the mechanical costs of moving a heavier sprawling body overcome the advantage of longer strides, and speed declines.

When Komodo dragon data are added to the dataset, the optimum shifts slightly to approximately 2.23 kg — the inclusion of such a large, relatively slower outlier pulls the peak leftward. The model then allows a prediction for the extinct giant monitor Varanus (Megalania) prisca, which may have exceeded 500 kg: estimated maximum sprint speed of 2.6–3.0 m s^-1, considerably slower than a modern Komodo dragon's peak of roughly 5–6 m s^-1 over short bursts.

For the living Komodo dragon, the implication is clear: at roughly 50–70 kg in large males, the species lies well above the optimal sprint mass. Short bursts of speed remain possible — important for ambush predation at close range — but sustained high-speed pursuit is not the dragon's biomechanical forte. The combination of the Clemente et al. (2009) sprint-speed data with the Cieri et al. (2021) force-scaling data and the Owerkowicz et al. (1999) respiratory data thus builds a coherent picture: the Komodo dragon is an ambush predator with above-average reptilian endurance (thanks to gular pumping) but with a mass-imposed ceiling on sprint performance.

Myths vs Facts

Key Takeaways

• The Cieri et al. (2021) study is the most comprehensive force-plate analysis of varanid locomotion to date. Spanning 12 species from 7 g to 26.55 kg, it directly measured ground reaction forces in Varanus komodoensis and revealed how the group manages biomechanical scaling without adopting mammal-like erect postures.
• Duty factor is the primary load-management mechanism. Larger monitors extend the proportion of the stride during which each limb contacts the ground, distributing impulse over time and capping instantaneous peak forces below fracture thresholds.
• Positive muscle allometry is the structural companion to duty-factor adjustment. Locomotor muscles grow disproportionately large relative to body mass in bigger varanids, compensating for the mechanical disadvantages of a sprawling posture at high mass.
• The gular pump (Owerkowicz et al., 1999) is the respiratory key to varanid endurance. By decoupling breathing from rib-cage undulation, monitor lizards achieve aerobic stamina that other sprawling lizards cannot — a physiological prerequisite for the Komodo dragon's active predatory and scavenging lifestyle.
• Sprint speed peaks at ~2–3 kg in varanids (Clemente et al., 2009). Adult Komodo dragons at 50–70 kg operate well above this optimum, making prolonged high-speed pursuit biomechanically expensive. Ambush hunting at close range is the ecologically coherent consequence.
• These findings are specific and quantitative. They provide a foundation for modelling energy expenditure, home-range requirements, and the ecological carrying capacity of Komodo National Park that general natural-history descriptions cannot support.

Frequently Asked Questions

What is a ground reaction force and why does it matter for lizard biology?

A ground reaction force is the force the ground exerts back on a foot during stance — equal and opposite to the force the animal pushes down with. In biomechanics it is the fundamental quantity linking locomotor behaviour to bone and muscle mechanics: bones fail when ground reaction forces exceed the strength of the skeletal material. Measuring how these forces scale with body mass in lizards tells us the mechanical limits on body size in sprawling tetrapods and how living animals have evolved to approach without exceeding those limits.

How were Komodo dragons safely used in the force-plate study?

The Komodo dragon data in Cieri et al. (2021) were collected from captive specimens at zoological institutions under institutional animal care protocols. Force plates were embedded in walkways and the animals crossed them voluntarily as part of natural movement, without restraint. High-speed cameras recorded limb kinematics simultaneously. No surgical procedures were involved. The ethical protocols are described in the supplementary methods of the original paper.

What is the "axial constraint" and how does gular pumping solve it?

During lateral-undulatory locomotion, the sideways bending of a lizard's trunk periodically compresses one lung while stretching the other, disrupting the rhythmic costal aspiration that drives breathing at rest. This is the axial constraint. Gular pumping bypasses it: the gular musculature forces air directly into the lung at positive pressure, independent of rib-cage movement, maintaining oxygen delivery during continuous locomotion (Owerkowicz et al., 1999).

Why do varanids not simply evolve erect posture to handle increasing body mass?

Erect posture requires fundamental reorganisation of limb musculature, joint geometry, and neuromotor control — an evolutionary pathway that took tens of millions of years to evolve in the mammalian and dinosaurian lineages. Varanids appear to have found an alternative solution — duty-factor adjustment and muscle allometry — that is evolutionarily accessible within the sprawling body plan. Whether the sprawling solution has an absolute upper size limit is a remaining open question in comparative biomechanics.

Does duty factor change when a Komodo dragon runs versus walks?

Yes. Across all varanids studied by Cieri et al. (2021), duty factor decreases with increasing relative speed — animals keep their feet on the ground for a smaller fraction of each stride as they move faster. The mass-related increase in duty factor (M^0.04) describes the baseline value at a given relative speed; at higher speeds, even large dragons reduce duty factor and approach the walk-to-run transition.

How does the gular pump compare to mammalian diaphragm breathing during exercise?

They are mechanistically different solutions to respiratory demands during locomotion. Mammals use a muscular diaphragm that is mechanically independent of trunk bending, allowing continuous costal and diaphragmatic breathing throughout a stride cycle without interference. Varanid gular pumping is a positive-pressure accessory mechanism that supplements rather than replaces costal aspiration. The mammalian solution is arguably more efficient at high aerobic intensities; the gular pump provides a meaningful but partial compensation that gives varanids a substantial advantage over other lizards.

What does this research imply for the Komodo dragon's hunting strategy?

The biomechanical picture is internally consistent: the Komodo dragon's mass places it above the optimal sprint speed for varanids, making sustained high-speed pursuit costly. However, gular pumping provides above-average aerobic endurance relative to other reptiles of similar size, supporting the documented behaviour of tracking wounded prey over distances of several kilometres. The animal is therefore best characterised as an ambush predator with moderate endurance — a profile that differs from both the pure burst-sprinter model and the pure coursing-predator model sometimes applied inaccurately to the species. For more on hunting behaviour, see our general overview at Komodo Dragon Locomotion & Speed.

Are there plans for further biomechanical research on Komodo dragon locomotion?

The Cieri et al. (2021) paper itself identifies several open questions, including the three-dimensional limb posture during stance, the contribution of the tail to balance and centre-of-mass stabilisation, and whether force-scaling relationships differ between juvenile and adult dragons (which differ substantially in body mass and dietary ecology). The Clemente laboratory at the University of the Sunshine Coast has an active programme in varanid biomechanics, and further studies on these questions are likely.

Sources & Further Reading

1. Cieri, R. L., Dick, T. J. M., Irwin, R., Rumsey, D., & Clemente, C. J. (2021). "The scaling of ground reaction forces and duty factor in monitor lizards: implications for locomotion in sprawling tetrapods." Biology Letters, 17(2), 20200612. https://doi.org/10.1098/rsbl.2020.0612 — Primary paper reviewed in this article; includes V. komodoensis force-plate data.
2. Owerkowicz, T., Farmer, C. G., Hicks, J. W., & Brainerd, E. L. (1999). "Contribution of gular pumping to lung ventilation in monitor lizards." Science, 284(5420), 1661–1663. https://doi.org/10.1126/science.284.5420.1661 — Foundational paper demonstrating the gular pump's role in overcoming the axial respiratory constraint during varanid locomotion.
3. Clemente, C. J., Thompson, G. G., & Withers, P. C. (2009). "Evolutionary relationships of sprint speed in Australian varanid lizards." Journal of Zoology, 278(4), 270–280. https://doi.org/10.1111/j.1469-7998.2009.00559.x — Sprint-speed versus body-mass scaling for 18 Varanus species; derives optimal sprint mass and predicts speed of extinct Megalania.
4. Cieri, R. L., Clemente, C. J., et al. (2020). "Monitoring muscle over three orders of magnitude: Widespread positive allometry among locomotor and body support musculature in the pectoral girdle of varanid lizards (Varanidae)." Journal of Anatomy, 237, 1114–1135. https://doi.org/10.1111/joa.13273 — Companion anatomy study demonstrating muscle-mass allometry that underpins force generation in large varanids.
5. Owerkowicz, T., Brainerd, E. L., & Carrier, D. R. (2001). "Electromyographic pattern of the gular pump in monitor lizards." Bulletin of the Museum of Comparative Zoology, 156(1), 237–248. — EMG follow-up to Owerkowicz et al. (1999) characterising the muscular activation sequence of the gular pump.
6. Clemente, C. J., Withers, P. C., & Thompson, G. G. (2013). "Lizard tricks: overcoming conflicting requirements of speed versus climbing ability by altering biomechanics of the lizard stride." Journal of Experimental Biology, 216(20), 3854–3862. https://doi.org/10.1242/jeb.087668 — Shows that arboreal varanids modulate stride frequency rather than stride length to maintain speed across substrates.
7. Auffenberg, W. (1981). The Behavioral Ecology of the Komodo Monitor. University Presses of Florida. — Foundational field study of Komodo dragon natural history; predates modern biomechanics methods but remains the primary behavioural reference.