Movement is a fundamental characteristic of life, yet some taxa, such as barnacles, lose their capacity for locomotion as swimming larvae to become permanently affixed to a substratum as adults. Barnacles adopted this type of sessile life-stage at least 500 million years ago; however, we unequivocally confirm a prior report that the epizoic sea-turtle barnacle, Chelonibia testudinaria, has the capacity for self-directedlocomotion in the adult stage. We used time-series field and laboratory photographs to document its movement paired with transplant experiments employing various flow conditions and inter-individual configurations to test whether foraging or reproduction are ultimate causes of the activity. We also examined the barnacle cement microscopically for insight on its role in translocation. On loggerhead and green sea turtles, C. testudinaria moved distances up to 78.6 mm/year across its host substratum and in traverses on laboratory panels, occasionally altered course abruptly by 90^o. Our findings indicate these movements are behaviourally directed and not passively driven by external forces. Barnacles tended to move directly against water flow and independent of nearby conspecifics, suggesting that movement functions primarily to facilitate foraging, not reproduction. While the mechanism enabling this movement remains elusive, we observed that trails of cement bore signs of multi-layered, episodic secretion. We speculate that proximal causes of movement involve one or a combination of rapid shell growth, cement secretion coordinated with basal membrane lifting, and directed contraction of basal perimeter muscles.

Barnacle movements on captive and wild turtles

Loggerhead turtles were caught incidentally by local fisheries within the Communidad de Valencia, Spain and brought to the Área de recuperación y conservación de fauna marina (ARCA), which is managed by the Fundación Oceanogràfic. The facility has a permit from the Valencian Regional Government for sea turtle rehabilitation. We selected five of the turtles admitted to the ARCA between December 2019 and May 2020 as they each hosted at least 10 C. testudinaria. Upon admission to ARCA, each of these animals was suffering from decompression sickness and was thus treated and monitored for up to 18 weeks before being released back into the wild (for details on veterinary care, see ⁴²). Sea turtles were housed in circular tanks, ranging from 2 to 6 m in diameter with a water depth of 0.95 m and maintained at a water temperature of ~24°C. All animals were fed twice daily using a mix of vegetable and fish material.

While sea turtles were housed at the facility, we monitored barnacle movements on each by attaching a 1 cm length fabric measuring tape using clear epoxy to within 30 mm of any clusters of C. testudinaria or solitary large individuals on the carapaces of the selected turtles. We photographed these barnacles every two weeks, ensuring that the camera was placed perpendicular to the carapace and that the 1 cm tape measure sections were visible in every photograph. We measured the growth and movements of 10 haphazardly selected barnacles from each of the turtles in ImageJ (v. 1.52p) by using the tape measure sections as fixed reference points of known size. The barnacle shell diameters (maximum rostro-carinal length) ranged from 14 to 41 mm (mean 24 mm) at the start of the study and increased to 14 to 45 mm (mean 26 mm) by the end of the study (Fig. S1). To mitigate any effect of growth on the measurements of barnacle movements, we only considered that movement had occurred if the total distance travelled within 14 weeks was 5 mm as this exceeded the maximum observed growth in barnacle diameter over this same period. To assess directional movement of barnacles that translocated, we determined whether the barnacles moved in either an anterior, posterior, lateral, or medial direction. We additionally noted whether each individual C. testudinaria bore attached complemental males.

To make observations of barnacle movements in the wild, we queried professional underwater divers at Siao Liu Qiu Island, Taiwan for photos of turtles with barnacles. We received photos of three green turtles with barnacles that were repeatedly photographed over a 16-week period. We used the shape of the post-ocular scutes to confirm the identity of the turtles and then inspected the photos visually to determine the movements of the barnacles⁴³.

Barnacle cementation and translocation on synthetic substrata

We collected Charybdis crabs from the shores of Taiwan using crab traps and selected those individuals that had C. testudinaria barnacles attached to their carapaces (Fig. S3A). To remove the barnacles, the crabs were euthanized and then each crab’s carapace was carefully trimmed to the edge of each barnacle’s base. Within 3-4 days, the remaining carapace adhering to the barnacles’ bases had degraded (~3-4 days) without damaging the membranous base of the barnacles. The “cleared” barnacles were laced onto 15x15 cm acyclic plates until they reattached (Fig. S3B). Successful reattachment of the barnacles was confirmed by the appearance of white cement¹⁰, which was visible around the periphery of the base (Fig. S3C). In total, 15 specimens successfully reattached.

We monitored the movements of these 15 C. testudinaria using time-series photographs for up to one year in an aerated tank. Each barnacle was photographed once per week at the apical, lateral and basal view using a digital camera (Panasonic, Lumix DMC-G1). On each acrylic plate, four yellow marker dots were placed on the plate as reference points and empty shells of C. testudinaria were attached to the plate as calibration markers. During this experiment, the barnacles were cultured in a Polyethylene tray (70 x 20 x 10 cm) with closed circulation and continuous aeration. The specimens were fed live adult Artemia once per day and the shell surfaces were cleaned by brushing every three days to avoid algal overgrowth. Sea water was changed daily.

This same experiment was repeated a second time but this time, we used three C. testudinaria that were obtained from the carapaces of dead stranded green sea turtles. These barnacles were also photographed in apical, lateral and basal view but instead of recording photos weekly they were recorded daily. Each of these barnacles were monitored for 3, 5, and 8 months, respectively.

Barnacle cement trail analysis

After observing that moving barnacles often left behind a trail of cement, we visually inspected the cement to see if it would provide insights into the mechanism enabling barnacle locomotion. This cement could be readily detached after air-drying for more than one week. We attempted further air-drying (up to a month) but this resulted in cracking of the material (see Fig. 3A). We also attempted cutting the cement with a razor blade, but this destroyed the microstructure along the cut edge (data not shown). Thus, the edges of natural cracks that ran almost perpendicular to the longitudinal axis of the cement strip were examined under SEM (FEI Quanta 200) to obtain cross-section views of the internal microstructure of the cement trail. Pieces of cement trail were mounted on SEM stubs and gold coated prior to observation.

Barnacle detachment force

To compare the attachment strength among barnacles that were either experimentally reattached to acrylic plates or naturally attached to crab carapaces, a shear force was applied to each test individual parallel to the substratum following methods specified in ASTM (2005)⁴⁴ and the force requires to detach them was measured with a force gauge (FG-20G, Taiwan). Barnacles on acrylic plates were divided into four groups for analysis based on their reattachment time and nested by size (1-10 (n = 10), 11-20 (n = 5), 21-30 (n =10) and >365 days (n = 5)). Individuals from the latter cohort were similar in size to those naturally attached to crabs to which they were compared. Before detachment with the force gauge, the underside of each transparent acrylic plate was photographed and the area of cement secreted was measured using the image analysis software Sigma Scan Pro. Variations in the force needed to detach the barnacles were analysed using one-way analysis of variance (ANOVA) after the data passed a homogeneity of variance test⁴⁵.

Barnacle locomotion relative to current flow

To study the effect of unidirectional water flow on barnacles, six individuals were reattached to the centre of replicate 60 x 60 mm acrylic plates. Each plate was placed in a 60 x 50 x 28 cm tank and positioned in the flow of an underwater pump (flow rate 10 l min^-1) directed out of a cylindrical PVC funnel, made by cutting the bottom out of a 250 ml commercial PVC bottle. The internal diameter of the outflow tube of the underwater pump was 9.24 mm and thus the flow velocity was 2.48 m s^-1 (Velocity = flow rate x tube cross section area). The experimental design included six replicate plates for three different treatments: (1) barnacles with the rostrum (cirral net) facing towards the flow (i.e., facing the narrow opening of the funnel), (2) barnacles with the rostrum (cirral net) facing away from the flow (i.e., facing the wide opening of the funnel), and (3) barnacles on control plates inserted into the funnel but with the water pump shut off. Water temperatures were maintained at 25 °C for all experiments (Fig. 5A, B). The distances traversed and the angles of movement of barnacles were measured after three months (Fig. 5D) using the intersection point of the paired scutum and tergum opercular plates of the barnacle as the center point. All barnacles were fed with Artemia nauplii, dispersed and recirculated for a duration of six hours per day, after which the tank water was exchanged.

Barnacle locomotion relative to conspecifics

To investigate whether barnacles move to reduce mating distance, we attached six pairs of barnacles at inter-individual distances of 5 cm, 10 cm, and 15 cm. Each pair of barnacles was laced onto acrylic plates 16 x 16 cm (5 cm and 10 cm pair treatment) and 16 x 24 cm (15 cm pair treatment). The plates were maintained in 25^oC aerated seawater. Barnacles were fed six hours per day and the experiment was monitored for 10 months. We analyzed variation in inter-individual distances in each treatment using paired t-tests. This experiment was repeated again but this time using five barnacles each laced on to 16 x 16 cm acrylic plates, at an inter-individual distance of 5 cm (Fig. S6E). We ran five replicates for this experiment and each experiment lasted for 3 months. To assess whether barnacles were clustering, separating, or moving randomly we calculated the area of the minimum convex polygon needed to encompass the area occupied by the barnacles both at the start and end of the experiment. The differences in areas were analysed using paired t-tests (Fig. S4E).

To further assess whether the nearby presence of conspecifics affects the movement patterns of C. testudinaria, we reattached 29 and 31 individuals onto separate acrylic plates 36 cm x 36 cm. The barnacles were placed at inter-neighbour distances ranging between 30-50 mm. Barnacles were allowed to reattach for three weeks prior to taking any measurements. Photographs of barnacles were collected weekly over 12 weeks. In each photograph, the distance to the nearest neighbour for each barnacle was measured using image analysis software (Sigma Scan Pro 5) and a nearest neighbour index (NNI) was used to assess the distribution⁴⁶. The value was calculated using the following formula:

R_n=D(Obs)0.5 an  

where ‘R_n’ is the NNI value, ‘D(Obs)’ is the average observed nearest neighbour distance in cm between the barnacles, ‘a’ is the spatial area in cm² occupied by the barnacles, and ‘n’ is the total number of individual barnacles. Values of the R_n index range in a continuum from 0 to 2.15. Rn values close to 0 indicate a clustered distribution, a value of 1 indicates a random distribution and a value near 2.15 indicates a regular distribution.

During all laboratory movement experiments, barnacles were kept at 25 ^oC water temperature, which is typical of seawater temperatures during summer in Taiwan when most barnacles produce mature gonads for reproduction. Chelonibia testudinaria had mature ovary and testis upon dissection of dead individuals in the experiments, suggesting that individuals were reproductively active during all movement experiments.