Mechanics of dust bathing in birds

Significance

Despite the common occurrence of dust bathing in birds, the physical mechanism by which it removes ectoparasites has remained unclear. This study demonstrates that sand grains act through immediate collisions, dislodging mites at a wing-flapping frequency that scales with body mass. These findings suggest that birds have evolved to coordinate their wing motions with the mechanical effects of sand, revealing a cleaning behavior shaped by parasite pressure and contributing to our understanding of grooming evolution.

Abstract

Dust bathing is a widespread behavior among birds that helps remove ectoparasites, but its mechanical basis remains poorly understood. Here we show that wing flapping during dust bathing generates sand-particle collisions sufficient to dislodge feather mites. Using high-speed video analysis across 34 bird species, we found that wing-flapping frequency scales inversely with body mass ($f\sim M^{-0.20}$), ranging from 1 to 10 Hz. Since feather motion matches wing motion, we use mechanical shakers to simulate dust-bathing dynamics. Mite-removal experiments reveal that parasite dislodgment efficiency increases rapidly up to around 5 Hz and plateaued thereafter. The frequencies around 5 Hz are sufficient to achieve maximal cleaning efficacy. Collision modeling indicates that the impact force from sand particles exceeds mite adhesion forces, thus collision is the dominant removal mechanism. Mites are removed within seconds when sand is present, suggesting that sand plays an active mechanical role in parasite dislodgment. These findings not only elucidate a physical mechanism underlying dust bathing, but also suggest that birds may have evolutionarily tuned their flapping behavior to optimize parasite removal through substrate interaction.

Dust bathing is most commonly observed in rodents, such as chinchillas, where it serves to remove oils and dirt from the fur surface and prevent fur matting (1, 2). Dust bathing is also common in both wild and domestic birds (3, 4), believed to help dislodge ectoparasites (5). However, despite its widespread occurrence, the physical mechanism underlying parasite removal, especially the role of sand and wing-flapping frequency, remains poorly understood. Previous studies have focused on behavior duration or preference (6–10), but rarely quantified biomechanical parameters such as impact force or frequency tuning. This study addresses this gap by establishing a mechanistic framework that links bird body mass to optimal wing-flapping frequency and provides evidence for a sand cleaning effect. Beyond its broad occurrence, clarifying the physical mechanism of dust bathing is also practically relevant for systems where ectoparasite control is critical.

The poultry industry loses an estimated over $100 million USD annually due to ectoparasite infestations (11), which reduce egg production, compromise animal welfare, and increase dependence on chemical insecticides. Among these parasites, the northern fowl mite is especially detrimental, leading to skin inflammation, lower egg production rate, and fluctuation of the price of egg (12). Current strategies to control parasites involve pesticides that risk environmental contamination and chemical resistance (13). Conventional parasite control in poultry farms primarily relies on chemical acaricides, which are applied through feed additives (14–17), sprays, or dusting powders (18). While these methods can suppress mite populations temporarily, their long-term efficacy is undermined by the development of resistance and the accumulation of residues in the environment and animal products (19, 20). Moreover, the overuse of chemical treatments has raised significant concerns regarding environmental pollution, occupational health for farm workers, and risks to food safety. The poultry industry thus faces a growing need for alternative parasite control strategies that are both effective and environmentally benign. Instead, birds have evolved a natural, chemical-free solution: dust bathing. These considerations motivate a mechanistic analysis of how wing and feather motions, interacting with sand, effect parasite removal.

Understanding this mechanism not only advances our knowledge of how birds naturally remove ectoparasites but also offers practical insights for developing sustainable cleaning technologies. Mimicking the physics behind sand bathing, frequency-tuned vibration observed in dust bathing may inspire eco-friendly dry-cleaning methods for nonwashable things such as down-filled jackets. Additionally, the same principles could guide the design of vibration-based systems for removing dust from air-conditioner filters, appliance vents, and other hard-to-clean crevices where conventional washing is ineffective.

Results

Relation Between Wing-Flapping Frequency and Bird Body Mass.

We analyzed 38 dust-bathing videos from YouTube and high-speed recordings conducted at a chicken farm, representing 38 individual birds from 34 species, with no repeated measurements on the same individual. All data are listed in SI Appendix. These recordings included a broad range of bird species with body masses ranging from 0.02 to 120.7 kg. A complete dust-bathing cycle begins with digging into the sand using the wings by conical motion, followed by wing-flapping movements that allow sand particles to interact thoroughly with the wing feathers (Fig. 1A). Conical motion coats the wings with sand, but the primary mechanism for removing mites from feathers is the wing-flapping motion. We simplify the wing flapping motion as rotational motion in this study.

Fig. 1.

Schematic and scaling of dust bathing in birds. (A) Diagram of a typical dust-bathing bout showing two phases: conical digging and rotational wing-flapping that drives sand into feathers. (B) Relationship between wing-flapping frequency $f$ and bird body mass $M$ across 34 species. The black dashed line is the best fit to experimental data: $f=4.09M^{-0.20}$, N = 38 individuals. The red dashed line shows predicted scaling from physical model: $f=1.11M^{-0.36}$. (C) Mites can be observed under a microscope to be randomly distributed across the surface of the feathers.

The wing-flapping frequency during dust bathing, f, was found to scale inversely with body mass M according to the relation:

$$f=4.09M^{-0.2},R^2=0.65,$$ [1]

This result (Fig. 1B) suggests that, despite body mass spanning four orders of magnitude, wing-flapping frequencies decrease slightly with increasing body mass but remain in a relatively narrow range of 1 to 10 Hz for dust bathing. For example, ostriches (94.3 ± 24 kg, N = 3 individuals) flap at about 1.1 ± 0.2 Hz, chickens (2.26 ± 1.4 kg, N = 12 individuals) flap at 4.2 ± 0.6 Hz, while bee-eaters (about 0.02 kg) reach 6.25 Hz. A full physical model based on a harmonic oscillator predicted a similar scaling follows in the next section, reinforcing that body mass and wing-flapping frequency impose biomechanical constraints on dust-bathing behavior (Fig. 1B, red dashed line).

Physical Modeling for Flapping Frequency Scaling With Body Mass.

To understand the biomechanical basis behind the observed scaling law of flapping frequency $f\sim M^{-0.2}$, we constructed a physical model based on the principles of simple harmonic motion. A bird wing was approximated as a rotational system (Fig. 2A), where the flapping behavior follows:

$$I\stackrel{\dot{}\dot{}}{\theta }+c\dot{\theta }+k_t\theta =0,$$ [2]

Fig. 2.

Physical modeling of avian wing-flapping dynamics. (A) Bird wing is modeled as a rotational system with mass moment of inertia $I$ and rotational stiffness $k_t$, leading to the analytical frequency equation. (B) Empirical data show wing mass scales with body mass as $m_w=0.077M^{1.10}$, N = 12 individuals. (C) Wing length scales as $L=0.48M^{0.35}$, N = 12 individuals. (D) Rotational stiffness scales as $k_t=0.32M^{1.08}$, N = 8 individuals.

Where $I$ is the moment of inertia of the wing, $k_t$ is the rotational stiffness, and the rotational damping coefficient $c$, representing the resistance offered by air and suspended sand grains. $\theta$ is the angular displacement. Assuming the wing behaves like a hinged beam, and the wing is fully extended, the moment of inertia scales as $I=\frac{1}{3}{m}_w{L}^2$, where $m_w$ is the wing mass and $L$ is the wing length. The joint angle as a function of time is denoted by $\theta (t)=A_w{e}^{-\zeta \omega t}cos(\omega t)$, where $A_w$ is the amplitude, $\omega$ is the angular frequency, and $\zeta$ is the damping ratio. $\zeta$ is defined as the rotational damping coefficient divided by the critical damping coefficient $c_c$, which is defined as $c_c=2\sqrt{k_{t}m}$. We rewrite the damping ratio, $\zeta =c/c_c=c/2\sqrt{k_{t}m}$. Combining this with the formula for angular frequency, $\omega =2\pi f$, and substituting the parameters above, the frequency $f$ of the system is given by:

$$f=\frac{1}{2\pi }\sqrt{\frac{3k_t}{m_w{L}^2}(1-{\zeta }^2)}.$$ [3]

The predicted rotational stiffness is defined as $k_t=E_t{A}_t{r}^2/l_0\sim M$, where r is the bone radius, $E_t$, $A_t$, and $l_0$ are the Young’s modulus, the cross-sectional area, and the original length of soft tissues surrounding the shoulder joint, respectively (21). $\zeta$ is a dimensionless quantity ranging from 0 to 1. Assuming that birds are isometric, $m_w\sim M$ and $L\sim M^{\frac{1}{3}}$. Substituting these parameters into Eq. 3 yields a theoretical scaling of wing-flapping frequency of $f\sim M^{-0.33}$.

In SI Appendix, we collected data from the literature across various bird species. Based on previous studies, the wings were excised at the shoulder joint from the carcasses, and their masses were measured (22). The resulting scaling relation was $m_w=0.077M^{1.10}$ (N = 12 individuals) in Fig. 2B. Wing length was measured from the shoulder joint to the feather tip using digital image analysis or calipers (22). Across species, wing length is scaled as $L=0.48M^{0.35}$ (N = 12 individuals) in Fig. 2C. Rotational stiffness of chicken was measured by experiment: wings were mounted on a fixed base, and incremental torques were applied using weights, as shown in SI Appendix, Fig. S1. Angular displacement was tracked via high-resolution video, from which stiffness was calculated by linear regression of torque vs. angular deflection. Across 8 samples from the experiment, rotational stiffness scaled as $k_t=0.32M^{1.08}$ (N = 8 individuals) in Fig. 2D. Because of insufficient sample availability, rotational damping experiments for the shoulder joint could not be performed. However, a previous study reported that the damping ratios of birds’ lower-limb joints are approximately 0.32 ± 0.24, based on measurements of 21 joints from 8 individuals across scales (21). While differences between joint types may exist, their functional similarities suggest that these values provide a reasonable first-order approximation; thus, this parameter was adopted for the shoulder joint in our model.

Substituting the scaling laws into Eq. 3 yields:

$$f=1.11M^{-0.36}.$$ [4]

The exponent (−0.36) obtained from the analytical model closely matches the theoretical prediction (−0.33), based on the assumptions of geometric similarity and resonance-based oscillation. The empirically measured scaling exponent from video analysis (−0.20) is slightly smaller but exhibits the same inverse trend. This suggests that bird’s flap substantially faster than the resonance-based expectation during dust bathing, on average 3.65 times higher than the model prediction across species. In particular, dust-bathing likely imposes different functional demands than steady-state locomotion, prioritizing short-term mechanical agitation over long-term energetic efficiency. As a result, birds may behaviorally modulate their wing motion away from the energetically optimal point to better meet the mechanical requirements of parasite removal. The functional significance of this behavior is addressed in the Discussion, where we consider its potential role in enhancing ectoparasite removal. Taken together, these observations suggest that dust-bathing represents a specialized behavioral context in which kinematic scaling laws are subtly adjusted to meet ecological demands.

Sand and Wing-Flapping Frequency Jointly Determine Mite Removal.

During dust bathing, the birds’ wings and feathers undergo steady-state oscillations, meaning they vibrate at a single, consistent frequency. Following the approach of Wang et al. (23), we modeled the feather shaft as a cantilever beam anchored to the wing, and applied the Euler–Bernoulli beam theory for free vibration (24). The first natural frequency of the feather was calculated using the following equation:

$$f_1=\frac{{1.875}^2}{2\pi L^2}\sqrt{\frac{E_s{I}_s}{\rho A_s}}(Hz),$$ [5]

where $E_s$ is the Young’s modulus of keratin feather shaft, $I_s$ is the second moment of area, $\rho$ is the density, and $A_s$ is the cross-sectional area of the shaft. Because the feather shaft tapers distally, we adopted position-dependent flexural stiffness values from (25), who measured $I_s$ along the length of pigeon and barn owl primaries. Combined with average values of $E_s$ from (26), we estimated that the first-mode natural frequencies within the proximal 88% of the shaft reach about 45 Hz for pigeons and about 16 Hz for barn owls. Dust-bathing wing-flapping frequencies range from 1 to 10 Hz, is lower than the feather’s natural frequency, thus the feathers in dust bathing do not undergo resonance.

Building upon this, we used a shaker to simulate the wing-flapping motion. We designed a series of mite removal experiments using a shaker capable of generating harmonic vibration up and down (Fig. 3A) at specified frequencies to mimic the rotational motion of a bird’s wing-flapping. This orientation was chosen because it represents the dominant normal acceleration observed in dust bathing, but the apparatus could not reproduce the lateral components of natural wing motion. Chicken feathers infested with mites were secured on the shaker platform and subjected to vibrations in the presence or absence of sand. Quartz sand with a mean grain diameter of about 0.4 mm was used to replicate field conditions. Based on our observations, each dust-bathing event comprises five bouts, each lasting approximately 5 s. We therefore applied vibration for a total duration of 25 s in each experimental trial to replicate natural dust-bathing conditions. After vibration at 6 Hz, the remaining mite count on each feather was recorded under a microscope.

Fig. 3.

Mite removal via sand collisions driven by wing-flapping. (A) Schematic of mite-removal experiment using mechanical shaker. Mite-infested chicken feathers are vibrated vertically with or without sand. (B) Mite removal over time with (orange triangles) and without sand (blue circles) at 6 Hz. Sand increases removal efficiency significantly, supporting the collision hypothesis. $N\left(t\right)$ is the number of mites at time $t$. (C) Mite number after 25 s at different vibration frequencies. Exponential decay fitting (red dashed line) shows optimal removal at 5 Hz. (D) Mite length vs. bird body mass across 19 species.

Results demonstrated a clear and significant effect of sand on mite removal efficacy, as shown in Fig. 3B. After 25 s of vibration, the feathers treated with sand retained only 25.6 ± 12.6% of the initial mites, whereas those vibrated without sand retained 76.6 ± 7.7% (independent t test, P = 0.007). The difference in residual mite proportion became significant after 5 s and persisted throughout the trial, indicating that sand substantially enhances the mechanical removal of mites. This suggests that sand provides a critical mechanism, through momentum transfer during collisions, that can not be achieved by vibration alone. Photographic evidence indicated that, in no-sand trials, mites predominantly remained attached to feather barbs, whereas in sand-treated trials, the feather surfaces were largely free of mites. Moreover, increasing vibration duration resulted in a progressive reduction in residual mite presence.

Furthermore, we examined the effect of vibration frequency on cleaning efficacy. Frequencies ranging from 1 to 50 Hz were tested using identical feathers and shaking durations. The result is shown in Fig. 3C. A sharp reduction in mite count was observed between 1 and 5 Hz, after which the benefit plateaued. At frequencies above 5 Hz, no further significant decrease was observed. When the vibration frequency was below the first natural frequency of the feather, the amplitude increased with frequency, consistent with vibration theory (24). If vibration amplitude were the dominant factor governing cleaning efficacy, a continued reduction in mite count would be expected at higher frequencies. However, no such trend was observed above 5 Hz. This observation indicates that cleaning efficacy is not strongly correlated with vibration amplitude, but instead depends on vibration frequency. The experimental data were best fit by an exponential decay model:

$$N\left(f\right)=9.2e^{-0.47f}+1.4,R^2=0.92.$$ [6]

Mite removal efficiency reached a local maximum near 5 Hz, with 95% CI from replicate variability indicating a range of 4 to 6 Hz. This frequency range also overlaps with the wing-flapping frequency of chickens during dust bathing, demonstrating a frequency-dependent effect that facilitates ectoparasite removal. Collectively, the experiments demonstrate that effective mite removal requires both the presence of sand and vibration higher than the natural frequency of the wing but lower than that of the feathers. The most effective flapping frequency occurs at about 5 Hz for chicken-scale feathers. The observed frequency in birds may balance biomechanical constraints and the effectiveness of cleaning.

Mite Detachment Via Sand-Feather Collisions.

We developed a collision-based theoretical model grounded in experimental observation to examine whether sand particles can exert sufficient force to detach mites from feathers. We estimated the magnitude of the impact force during a typical sand–feather interaction under dust-bathing conditions. We analyzed high-speed video recordings at 1,000 frames per s to determine the duration of interaction between the sand grain and the feather during their collision. (Supplementary Video 1). The effective contact time is the interval from the onset of deceleration of the sand particle by the feather to its detachment, which was consistently measured as $\mathrm{\Delta }t=51.7\pm 2.2$ ms (N = 6 trials). This interval captures the main phase of force transmission in the collision and excludes postimpact passive motion. During the contact time, the sand and the feather oscillate synchronously. The oscillation is approximately simple harmonic motion, exhibits velocity symmetry, such that the peak upward and downward velocities are identical in magnitude. The maximum velocity of sand landing is $v_0=L_{\mathit{feather}}\omega =-2\pi L_{\mathit{feather}}f$, where $f$ is the frequency of the feather, and the end velocity is $v_1=2\pi L_{\mathit{feather}}f$. The average mass of one sand grain $m_{\mathit{sand}}$ was measured to be 0.4 mg. We then applied the impulse-momentum theorem to estimate the impact force as:

$$F=\frac{m_{\mathit{sand}}\mathrm{\Delta }v}{\mathrm{\Delta }t}$$ [7]

where sand velocity change is $\mathrm{\Delta }v=4\pi f{L}_{\mathit{feather}}$. With $f=4.1M^{-0.2} \mathrm{Hz}$ and $L_{\mathit{feather}}=0.17M^{0.24}$ (27–34) in SI Appendix, $\mathrm{\Delta }v\cong 8.7M^{0.04}(m/s).$ The calculated impact force $F=0.12M^{0.04} (\mathrm{mN})$. The mass of birds that have dust-bathing behavior ranging from 0.02 kg to 120.6 kg kg, thus, the collision forces of a sand particle are between $0.1 \mathrm{mN}$ and $0.15 \mathrm{mN}$. This derived force estimate represents the average force transmitted to a mite during dust bathing.

We compared the estimated impact force to previously reported ectoparasite adhesion forces to evaluate whether it was sufficient for detachment. From (35), the adhesive force of feather mites ranges between $0.012$ and $0.05 \mathrm{mN}$. Our measured impact force thus exceeds the upper bound of mite adhesion, indicating that a single collision event at typical feather-flapping speed is capable of dislodging mites. This conclusion is further supported by the observation that mite size remains largely consistent across bird species. As shown in Fig. 3D, across 19 avian species, mite body length showed no significant correlation with host body mass, following the relation $L_{\text{mite}}=0.42M^{-0.04}$. Although host body mass ranged from 0.01 kg to over 100 kg, mite length remained relatively consistent, averaging 0.41 ± 0.19 mm (N = 19 individuals) (36–44). We assumed that mite mass is the same across different bird species with similar mite sizes. This uniformity implies that the required detachment force is relatively constant across bird species, even as bird mass and flapping frequency vary. Beyond previously proposed mechanisms such as feather lipid reduction (6) and pore plugging the breathing pores of ectoparasites (8), we hypothesize that discrete collisions between sand grains and mites may also contribute. Such collisions deliver impulsive forces leading to detachment.

Discussion

Scaling of Wing-Flapping Frequency.

Our study provides a mechanistic framework for understanding how most ground-dwelling birds remove ectoparasites through dust bathing, with a focus on the scaling of wing-flapping frequency with body mass. We observed that the wing-flapping frequency during dust-bathing decreases with increasing body mass, following an empirical scaling of $f\sim M^{-0.2}$. This general trend aligns well, yet the scaling exponent is slightly smaller than theoretical predictions (approximately −0.33) or our analytical predictions (−0.36) would suggest. This discrepancy indicates that birds do not flap at the natural frequency based on their morphology. Instead, they may increase their flapping rate beyond resonance to improve the effectiveness of parasite removal. As a methodological limitation, our mechanical shaker imposed only vertical oscillations. This simplification captures the dominant normal component of wing-driven sand impacts but does not reproduce the full multiaxis kinematics of natural dustbathing.

Such intentional deviations from purely mechanical scaling have been documented across various animal locomotor behaviors, where the functional demands of a behavior take precedence over minimizing energy expenditure or structural stresses. For instance, Taylor et al. (45) demonstrated that small rodents expend disproportionately more energy generating muscular force during running than large mammals, yet both adjust stride frequency and ground contact times minimally in response to increased load, highlighting that locomotor energetics are modulated by immediate biomechanical constraints rather than solely by theoretical energetic optimization. Similarly, aerodynamic studies conducted on pigeons have demonstrated that birds actively modulate wing kinematics even in nonflight contexts, such as controlled hovering or stationary wing rotation. These adjustments optimize localized force generation according to the immediate functional requirements of the behavior, independent of flight performance (46). Together, these studies support the interpretation that dust-bathing behavior involves active modulation of wing motions to specifically enhance sand acceleration and impact forces crucial for ectoparasite removal.

The distinct scaling pattern observed for wing-flapping frequency during dust bathing compared to typical flight behavior further implies that dust bathing is not merely a passive reuse of flight kinematics. Specifically, wing-flapping frequencies measured during dust bathing are consistently lower than those observed during sustained flight in birds of comparable body mass. The scaling relationship between wing-flapping frequency and body mass in flight exhibits a more pronounced power-law pattern, with exponents typically ranging from approximately −0.17 to −0.33 (47, 48). This indicates that as body mass increases, wing-flapping frequency decreases more steeply in flight compared to dust bathing, reflecting distinct biomechanical and energetic constraints underlying these different motor behaviors. Moreover, observations from hummingbirds clearly illustrate that birds are capable of deploying identical anatomical structures—the wings—for vastly different functional objectives, ranging from routine forward flight to specialized, high-speed courtship dives (49). This behavioral versatility underscores the modularity of avian motor control systems. In the context of dust bathing, wing-flapping alone is insufficient to immediately remove ectoparasites. Our results show that significant mite detachment—up to 80% reduction within 25 s—occurs only when wing motion is combined with sand collisions. From a neurophysiological perspective, this behavioral specialization likely arises from the recruitment and activation of independent neural control modules—central pattern generators—that can orchestrate similar muscular movements to achieve distinctly different functional outcomes depending on the ecological or behavioral context (50).

Collisions Drive Mite Detachment.

Our experiments provide quantitative support for the hypothesis that collisions between sand particles and ectoparasites play a primary role in mite removal during dust bathing. Using identical feather samples mounted on a mechanical shaker, we compared mite removal under two conditions: with and without the presence of sand. We found that the addition of sand particles significantly and immediately enhanced mite removal efficiency after a short period of vibration, as shown in our results. In contrast, when no sand was present, mite removal was minimal under the same mechanical stimulation. This rapid effect supports the interpretation that collisions between sand particles and ectoparasites are the dominant removal mechanism.

The instantaneous removal of mites confers a direct fitness benefit to the host, consistent with the widespread physiological and reproductive costs of ectoparasitism documented in birds (51). If the reduction in mite numbers were instead primarily driven by processes such as depletion of feather surface lipids, a delayed response would be expected, as mites would require time to suffer from food scarcity. The observed rapid removal effect is therefore consistent with momentum transfer through collisions.

The exponential decay model in the Result section indicates that mite removal increases nonlinearly with frequency. However, due to the diminishing slope of the exponential decay curve, the removal efficiency gains become progressively smaller at higher frequencies. This saturation effect suggests that flapping beyond a certain frequency yields limited additional benefit. This trend closely resembles the shaking behavior observed in wet mammals during self-drying. Notably, the critical control parameter differs between the two systems: in mammals, the critical point corresponds to a nondimensionalized centrifugal acceleration of approximately 10, whereas in birds dust bathing, saturation occurs at a vibration frequency of approximately 5 Hz (52). The typical dust-bathing frequency observed across bird species ranges from 1 to 10 Hz, with observed scaling exponent of −0.2. Domestic chickens flap at approximately 4 to 6 Hz. This frequency falls near the steep region of the exponential response curve, where small increases in frequency yield large improvements in removal efficiency, which is consistent with our experimental results, as shown in Fig. 3C. Compared to time-consuming behaviors like preening, dust bathing delivers immediate removal benefits, aligning with previous experiments comparing grooming efficiencies (53).

Moreover, our results show that the estimated impact force of sand particles during dust bathing exceeds the reported adhesion force of feather mites by a substantial margin, supporting the mechanical feasibility of collision-based removal. This estimate aligns with known adhesion thresholds for small arthropods, indicating that sand-grain-scale collisions are sufficient to overcome mite attachment forces (51). In addition, mite size was found to be largely independent of host bird body mass, exhibiting minimal scaling across species, as demonstrated in our results. This suggests that a relatively uniform force threshold is required to detach mites, enabling birds of various sizes to achieve effective mite removal through appropriately tuned flapping frequencies.

Other animals also remove attached materials by modulating their flapping frequency. For example, mosquitoes remove drops by increasing their wingbeat frequency to nearly twice that used during flight, thereby generating sufficient centrifugal acceleration to shed droplets (54). In contrast, birds do not rely on such frequency amplification alone. The first natural frequency of feathers is higher than that of the wing, increasing the flapping frequency below the feathers’ first natural frequency leads to an amplification of amplitude. This amplification produces sufficient centrifugal forces and, when combined with sand grains, increases the probability of collisions with mites, ultimately enabling effective parasite removal.

Such functional consistency parallels scaling trends seen in other biological systems, such as the invariant duration of urination across mammals of different sizes (55). While the present model captures the dominant mechanical principles of dust bathing, biological variability—such as feather structure and species-specific mite behavior—may further modulate cleaning performance in natural contexts, as suggested by studies on tree swallows and nest feather composition (56).

Evolutionary Implications of Dust-Bathing Behavior.

Dust bathing is most commonly observed in terrestrial taxa such as sparrows (57), quails (58), chickens, and ostriches (59). Although numerous interspecific studies have examined dust-bathing behavior, the potential influence of body size has received little attention. Furthermore, the occurrence of this behavior is not consistent across all bird species (60). Many aquatic birds rely primarily on water bathing or preening for feather maintenance (61). This ecological pattern suggests that dust bathing likely evolved as a behavioral adaptation to terrestrial habitats where access to water is limited but ectoparasite control remains essential.

From an evolutionary perspective, the ability to harness ambient particulate matter to dislodge ectoparasites through oscillatory wing motion mechanically represents a low-cost, substrate-mediated solution to a widespread biological challenge. The wing-flapping frequencies used during dust bathing converge with the frequency range that maximizes parasite removal via collisions. This alignment implies that the behavior has been shaped by selective pressures favoring effective cleaning within biomechanical constraints. Supporting this notion, morphological traits such as the small maxillary overhang—a tiny forward projection of the upper beak that enhances the bird’s ability to shear and remove ectoparasites during preening—have been shown to evolve under parasite-driven selection, reinforcing the connection between ectoparasite pressure and host phenotype optimization (62). Together, these findings suggest that dust bathing and related morphological adaptations represent an integrated evolutionary strategy to facilitate parasite control within biomechanical and ecological constraints.

While some environmental or sensory factors, such as sunlight exposure, do not significantly enhance the efficiency of other grooming behaviors like preening (63), the strong frequency dependence observed in dust bathing suggests a more direct mechanical mode of action. Finally, the persistence and refinement of dust-bathing behavior may reflect an evolutionary arms race between birds and their ectoparasites. Reciprocal natural selection between host cleaning strategies and parasite clinging adaptations has been demonstrated in avian systems, indicating that such behaviors are not simply innate habits but evolved outcomes of long-term antagonistic coevolution (64). Future comparative studies may help elucidate how this behavior coevolved with substrate type, parasite diversity, and feather morphology across avian lineages.

Materials and Methods

Recording and Data Sources.

Dust-bathing behaviors were recorded using high-speed cameras (Phantom VEO-E310L) and DSLR equipment (Sony α1 with 100 mm lens). A total of 38 videos from field recordings and YouTube videos were analyzed to extract wing-flapping frequencies across species.

Feather–Wing Frequency Relationship Via Mechanical Modeling.

The relationship between feather and wing motion was addressed through mechanical analysis. Feathers were modeled as cantilever beams anchored at the wing joint, and their first-mode natural frequencies were estimated using Euler–Bernoulli beam theory. Material and geometric properties of feather shafts were obtained from published sources to confirm that feather vibration would remain synchronized with wing motion in the 1 to 10 Hz range observed during dust bathing.

Experimental Setup for Mite Removal.

Feathers naturally infested with Cheyletidae mites were collected from silkie fowl raised on a local farm. With the assistance of Dr. Jhih-Rong Liao, we were only able to narrow down the classification of the mite to the family Cheyletidae, without being able to determine its genus further. Images of the Cheyletidae mites, with scale bar labeled, are provided in SI Appendix, Fig. S2.

Each feather was fixed onto a wooden stick and vertically mounted on a shaker (LABWORKS ET-126), driven by a function generator and amplifier. Vibrations were applied at controlled frequencies (1 to 50 Hz) for 25 s, with or without sand (ROMAY beach sand, 0.1 to 2 mm). Mite removal efficacy was assessed by counting the number of remaining mites under a light microscope. Each feather was examined under a stereomicroscope prior to testing, and the number of visible mites was recorded. Because infestations typically consisted of 5 to 20 mites per feather, direct counting was feasible with an estimated precision of ±1 to 2 individuals.

Estimation of Sand–Feather Impact Force.

Sand particle acceleration was tracked from high-speed video (1,000 fps) and used to estimate the collision force between sand and feather. The mass of a single grain was calculated by weighing 100 particles using an analytical balance. Contact time was measured as the overlap duration between the sand grain and feather in video frames. Impact force was then calculated under the assumption of collision using impulse–momentum theory.

Measurement of Wing Rotational Stiffness.

The rotational stiffness of bird wings was measured by fixing the dissected wing to a rigid base and applying torque via suspended weights. Angular deformation was recorded, and stiffness was derived from the slope of the moment–angle relationship. All animal materials were obtained with appropriate permissions from farms and wildlife rehabilitation sources. This estimate assumes uniform material properties along the rachis. In practice, the shaft narrows distally and the effective modulus changes but applying the methods of Bonser (26) and Bachmann et al. (25) showed that in chickens the proximal ~88% of the rachis can be regarded as mechanically rigid and vibrates with the wing as a whole. Accordingly, we focused on wing–feather frequency matching and neglected distal variation in material properties.

Supplementary Material

Appendix 01 (PDF)

Movie S1.

Dust bathing of Lohmann Brown Chicken, mass of 1.76 kg. Time slowed by 2X.

Data, Materials, and Software Availability

All study data are included in the article and/or supporting information.