Field Experiments, Behavioral Analyses, and Digestive Physiology Reveal the Selective Oyster-Feeding Strategy of Thais luteostoma

Simple Summary

Biofouling organisms, especially oysters, commonly attach to cultured pearl oysters and can seriously reduce growth performance and farming efficiency. Conventional fouling removal relies on frequent manual cleaning, which is labor-intensive, costly, and may cause physical damage to cultured animals. In this study, we found that the predatory gastropod Thais luteostoma naturally and selectively feeds on fouling oysters without harming the pearl oyster Pinctada fucata martensii. Field co-culture experiments showed that the presence of T. luteostoma effectively reduced biofouling organisms and indirectly promoted pearl oyster growth. Behavioral observations confirmed that T. luteostoma actively targets oysters, and digestive enzyme analyses demonstrated clear physiological responses following feeding. These findings suggest that T. luteostoma could serve as a potential biological control option for managing biofouling in pearl oyster aquaculture and potentially in other bivalve farming systems.

Abstract

Pearl oyster aquaculture is severely constrained by biofouling organisms, particularly fouling oysters, which substantially impair pearl oyster growth and farming efficiency. This study investigated the selective oyster-feeding behavior of the predatory gastropod Thais luteostoma and evaluated its potential as an ecological biofouling control agent in pearl oyster culture. Field co-culture experiments showed that T. luteostoma did not adversely affect the survival of Pinctada fucata martensii, while effectively reducing biofouling loads and significantly improving pearl oyster growth performance. Laboratory behavioral assays and quantitative analyses revealed a pronounced feeding preference for oysters in T. luteostoma, as evidenced by a higher number of feeding individuals, longer total feeding duration, and greater spatial overlap between feeding hotspots and oyster locations. In addition, digestive enzyme assays indicated marked post-feeding physiological responses in T. luteostoma, with a stronger induction of digestive activity in the digestive gland than in the stomach. Collectively, these findings suggest that T. luteostoma represents a promising and sustainable biological option for managing biofouling in pearl oyster aquaculture, with potential applicability to other high-value bivalve farming systems.

1. Introduction

Pinctada fucata martensii is a marine bivalve of high economic value and is widely cultivated for the production of high-quality seawater pearls. Approximately 1.938 tons of seawater pearls were produced from this pearl oyster species in China in 2024 [1]. It is primarily distributed in tropical and subtropical regions of the Pacific and Indian Oceans, including China, Japan, India, and Vietnam [2]. As a sessile species with limited mobility, P. f. martensii is highly susceptible to colonization by biofouling organisms on the shell surface during cage culture. The accumulation of biofouling organisms can impede water exchange between the inside and outside of culture cages [3,4]. Such accumulation can also reduce food availability and oxygen supply to cultured pearl oysters [5,6], hinder the removal of nitrogenous wastes from cages [7], and ultimately retard the growth of farmed individuals [8]. At present, biofouling in P. f. martensii aquaculture is mainly managed through periodic replacement of culture cages and mechanical removal from shell surfaces using knives or scrapers. The frequency of cage replacement and manual cleaning depends on latitude, season, and mesh size, and may be required as frequently as every five days under severe fouling conditions [9]. These high-intensity cleaning practices are labor-intensive and costly, and often cause unintended damage to the cultured pearl oysters. Typical injuries include shell fractures and compression or tearing of soft tissues (mantle and visceral mass), which reduce the pearl oysters’ resistance and make them susceptible to disease and predation, leading to reduced survival and growth rates [10,11,12].

The use of biological control strategies (e.g., predators) to mitigate biofouling has therefore emerged as a promising alternative. A range of biological control agents, including fish [2], crabs [13], shrimps [14], sea cucumbers [15], and sea urchins [16], have been evaluated and shown to be effective in fouling control. Among various biofouling organisms, oysters often serve as a significant fouling group in marine bivalve aquaculture systems due to their strong attachment ability, rapid growth, and high population resilience [17]. For example, surveys of fouling organisms in major pearl oyster and scallop farming areas (Beibu Gulf, Guangxi; Leizhou Bay, Guangdong) indicate that fouling oysters frequently serve as the dominant fouling species and are present almost year-round [18,19]. Their firm adhesion to the shells of cultured bivalves makes manual removal particularly difficult and increases the risk of mechanical injury during cleaning. Consequently, it is both necessary and practical to target fouling oysters when developing biological control strategies within complex fouling communities. Currently, the use of carnivorous gastropods to control fouling oysters in aquaculture has been reported. Research indicates that introducing three carnivorous gastropods (Thais clavigera, Ceratostoma rorifluum, and Cantharus cecillei) into abalone (Haliotis discus hannai) aquaculture significantly reduces the survival number of oysters attached to abalone shells [20], revealing the potential application value of carnivorous gastropods in controlling fouling organisms, particularly oysters. Predatory behavior is typically accompanied by distinct digestive physiological responses. In marine gastropods, previous studies have reported that digestive enzyme activities increase markedly shortly after feeding, as observed in species such as Laevistrombus canarium and Neptunea cumingii [21,22]. Accordingly, changes in digestive enzyme activity can serve as a reliable physiological indicator of effective feeding following predation by T. luteostoma on oysters.

Based on the research status, our research group previously screened four nearshore gastropod species—Thais luteostoma, Trochus maculatus, Cuma lacera, and T. clavigera—for their potential to control biofouling organisms associated with P. f. martensii. The results demonstrate that T. luteostoma could selectively prey on oysters attached to the shells of P. f. martensii without damaging the host, indicating pronounced selective feeding behavior. To elucidate the selective oyster-feeding characteristics of T. luteostoma and to assess its potential for biofouling control in aquaculture systems, this study combines multiple approaches. Specifically, field co-culture experiments were conducted to evaluate fouling removal efficiency and ecological safety; laboratory behavioral observations and quantitative analyses were used to characterize selective feeding toward oysters; and digestive enzyme activities in the stomach and digestive gland were measured to assess feeding-associated physiological responses. Together, these results demonstrate the selective feeding capacity of T. luteostoma on oysters and its potential application value, providing critical evidence for ecological control of biofouling in pearl oyster aquaculture and other high-value bivalve farming systems.

2. Material and Methods

2.1. Animals Collection and Maintenance

Specimens of T. luteostoma (shell length: 43.66 ± 2.87 mm; shell width: 27.67 ± 1.61 mm; wet weight: 13.03 ± 2.40 g) were collected on 24 March 2025 from the intertidal zone along the coast of Naozhou Island, Zhanjiang, Guangdong Province, China (110°33′47″ E, 20°52′34″ N). All individuals were transported to the aquaculture facility of Guangdong Ocean University and acclimated for 10 days before experimentation. A total of 150 T. luteostoma underwent acclimation farming, achieving a 100% survival rate throughout the cultivation period. Following the acclimation phase, all rhinophore and tentacle extension activities, along with attachment behaviors, returned to normal in T. luteostoma, indicating that the specimens had recovered from the stress levels incurred during collection to their natural state (Figure 1a). During acclimation, seawater conditions were maintained at a temperature of 25.0–26.0 °C, dissolved oxygen of 5.5–6.5 mg·L⁻¹, salinity of 27–29 ppt, and pH of 7.9–8.1, with 50% of the water renewed daily. In the study area, T. luteostoma typically preys on barnacles and oysters as its primary food sources. Based on existing research reports on the prey selection of carnivorous gastropods [23,24] and observations from preliminary feeding trials, and to minimize experimental error and bias, frozen fish paste was selected as the feed for T. luteostoma during the acclimation period. T. luteostoma were fed frozen minced fish (Trachinotus ovatus) twice daily to satiation, and the photoperiod followed the natural light–dark cycle.

For field experiments, healthy 1.5-year-old adult P. f. martensii with intact shells were used (shell length: 62.46 ± 5.31 mm; wet weight: 35.81 ± 4.65 g). All biofouling organisms attached to the shell surface were carefully removed before the experiment.

For laboratory behavioral assays, P. f. martensii individuals had a shell length of 39.52 ± 3.27 mm and a wet weight of 7.03 ± 0.88 g, while oysters (Crassostrea rivularis) had a shell length of 36.71 ± 2.65 mm and a wet weight of 6.23 ± 1.09 g. Both oysters and P. f. martensii used in laboratory experiments were purchased from Daqiuzhuang Aquaculture Co., Ltd. (Leizhou, Guangdong Province, China).

2.2. Field Trials

Field trials were conducted in a P. f. martensii farming area located in Liushawan Bay, Zhanjiang, Guangdong Province, China (109°57′12″ E, 20°24′54″ N), approximately 1 km offshore. Four treatments were established, each containing 25 P. f. martensii per cage. The control group (C) contained no T. luteostoma. Based on a pilot study, the experimental groups were stocked with 3 (PT3), 6 (PT6), or 9 (PT9) T. luteostoma individuals per cage. Each treatment consisted of five replicates.

Based on local farming practices, experiments were conducted using plastic cages (42.5 cm × 32 cm × 14 cm) with a mesh size of 13.5 × 18.5 mm. The trial commenced on 3 April 2025 and lasted for 90 days. Survival rates of P. f. martensii and T. luteostoma were recorded every 30 days, and shell length measurements were taken to calculate growth rates. The number of oysters attached to the shell surface of P. f. martensii was also recorded at each sampling interval. At the end of the experiment (day 90), the total mass of inorganic sludge and the total mass of biofouling organisms attached to the shells were measured. Specifically, after retrieving the aquaculture cages from the sea, let them drain for 20 min. Then remove the pearl oysters from the cages and weigh each individually to obtain the “total weight”. Next, use a water flow to quickly rinse away the sludge from the surface of the pearl oysters. Similarly, let them drain for another 20 min before weighing each individually to obtain the “total weight after sludge removal”. Finally, manually remove all biofouling from the pearl oyster surfaces using knives, taking care not to damage the oysters, and weigh each individually to obtain the “net weight of pearl oysters” (Figure 1b–d). The total weight of shell surface sludge (WS), biofouling organisms (WBF), survival rate (SR), and relative growth rate (RGR) were calculated using the following equations:

WS = total weight − total weight after sludge removal

WBF = total weight after sludge removal − net weight of pearl oysters

SR = 100 × survival numbers/total quantity

RGR = 100 × (final shell length − initial shell length)/initial shell length

2.3. Behavioral Assays

Before behavioral experiments, T. luteostoma were starved for 3 days to eliminate potential satiation effects. Ten individuals were randomly selected and labeled with colored tags for individual identification. Experiments were conducted in 72 L aquaria, in which oysters (O group) and P. f. martensii (Pm group) were placed diagonally opposite each other, with T. luteostoma released at the center of the tank (Figure 2).

Animal movements were recorded using a time-lapse camera (Brinno TLC 200, Brinno, Taipei, Taiwan, China). Each trial lasted 4 h, and the recorded video duration corresponded to 1/450 of the actual observation time. Six independent replicates were conducted, with the positions of oysters and P. f. martensii alternated between trials to minimize spatial bias. Video files were analyzed frame by frame using Tracker software (Version 4.95, USA) to extract positional coordinates, time stamps, and velocities of T. luteostoma. These data were subsequently used to analyze feeding duration (based on the start and end of observed sucking), movement trajectories, and spatial feeding hotspots.

All behavioral trajectory data were further processed and visualized using a custom Python-based workflow. Exported tracking coordinates (x, y, time t, and velocity v) were mapped to image pixel positions according to the predefined coordinate system used during tracking. Time-lapse tracking points were directly connected in their original order, without any path smoothing or interpolation, and rendered as polyline segments on the background image using Matplotlib’s (V3.10.7) LineCollection, with segment-wise color coding by time or velocity to visualize movement dynamics.

Spatial hotspots were generated by pooling all trajectory points and calculating a velocity-inversed weighted density (1/v) within spatial grid cells to emphasize low-speed or lingering behaviors. The resulting density matrix was normalized across the arena, and spatial feeding hotspots were defined as contiguous regions exceeding the 75th percentile of the global density distribution. This percentile-based threshold objectively identifies sustained low-velocity residence without imposing an arbitrary minimum residence-time cutoff. The density matrix was smoothed using a Gaussian filter (σ = 3 grid units) to reduce discretization artifacts while preserving spatial resolution. Hotspot patterns were robust across σ values ranging from 2 to 4.

All analyses and visualizations were performed in Python 3.10 using NumPy (V2.3.5), Pandas (V2.3.3), Matplotlib (V3.10.7), SciPy (V1.16.3), and Pillow (V12.0.0).

2.4. Digestive Enzyme Activity Assays

After a 3-day starvation period, six T. luteostoma individuals were randomly selected, and samples of the digestive gland (BD) and stomach (BS) were collected before feeding. Subsequently, individuals were fed oyster soft tissue ad libitum, and after 30 min, six feeding T. luteostoma were sampled to obtain digestive gland (AD) and stomach (AS) tissues.

Samples were homogenized in ice-cold conditions using 10 volumes of 0.86% physiological saline with a Tissue-prp-02 homogenizer (Jingxin, Shanghai, China), followed by centrifugation at 10,000× g for 10 min at 4 °C. The supernatants were collected for enzymatic analyses. Activities of trypsin, pepsin, and lipase were determined using commercial assay kits (Abbkine, Wuhan, China) according to the manufacturer’s instructions.

2.5. Statistical Analysis

All data were analyzed using SPSS 26.0 (IBM, New York, NY, USA). Normality and homogeneity of variance tests were conducted before parametric statistical analysis to ensure data compliance. For data involving three or more groups (relative growth rate, fouling biomass), differences were assessed using one-way analysis of variance (ANOVA), followed by Duncan’s multiple range test for post hoc comparisons. For behavioral data involving only two groups (feeding time), independent samples t-tests were used to assess differences. For percentage data such as survival rates, or data failing normality and homogeneity of variance tests, nonparametric tests (Kruskal–Wallis H test and Mann–Whitney U test) were employed, with post hoc comparisons conducted using the Nemenyi test. Statistical significance was set at p < 0.05.

3. Results

3.1. Effects of Field Co-Culture Experiments

3.1.1. Survival and Growth of P. f. martensii and T. luteostoma

Survival rate analyses indicated that no significant differences were observed among treatments for either P. f. martensii or T. luteostoma throughout the experimental period (p > 0.05; Figure 3a,b). Before the experiment, it was confirmed that the initial shell lengths of the P. f. martensii and T. luteostoma were statistically similar across, with no statistically significant differences (p > 0.05). Analysis of relative growth rates revealed that at 60 and 90 days of co-culture, the shell length–based relative growth rates of P. f. martensii in the PT6 and PT9 groups were significantly higher than those in the control group (p < 0.05). Although the PT3 group exhibited higher relative growth rates than the control, the difference was not statistically significant (p > 0.05; Figure 3c). No significant differences in the relative growth rate of T. luteostoma were detected among treatments (p > 0.05; Figure 3d).

3.1.2. Biofouling Removal Efficiency of T. luteostoma

The number of fouling oysters attached to the shell surfaces of P. f. martensii showed that after 30 days of co-culture, all three experimental groups (PT3, PT6, and PT9) had significantly fewer fouling oysters compared with the control group (p < 0.05; Figure 4a). At the end of the 90-day experiment, no significant differences were detected among treatments in the weight of inorganic sludge on the shell surfaces of P. f. martensii (p > 0.05). However, the total weight of biofouling organisms in the PT6 and PT9 groups was significantly lower than that in the control group (p < 0.05; Figure 4b).

3.2. Selective Feeding Behavior of T. luteostoma

Gantt charts of feeding duration indicated that across six independent trials, the majority of T. luteostoma individuals (5.67 ± 2.16) actively moved toward the oyster side and maintained feeding activity, whereas only a small number of individuals (1.50 ± 1.52) approached the P. f. martensii side (Figure 5a–f).

Total feeding duration analyses further showed that in five repeated trials, the cumulative feeding duration of T. luteostoma on oysters was consistently and significantly longer than that on P. f. martensii (p < 0.05; Figure 5g–l). Trajectory and spatial hotspot analyses revealed that in three of the six trials, feeding hotspots overlapped with oyster locations (Figure 6a,c,f). In two trials, no distinct hotspots were detected (Figure 6b,d), while in one trial, the hotspot occurred near P. f. martensii but did not overlap, as T. luteostoma primarily remained along the tank wall (Figure 6e). Overall, these results demonstrate a pronounced tendency of T. luteostoma to selectively target oysters for feeding.

3.3. Changes in Digestive Enzyme Activities Before and After Oyster Feeding in T. luteostoma

Digestive enzyme assays showed that, compared with pre-feeding levels, the activities of trypsin, pepsin, and lipase in the digestive gland were all significantly elevated after feeding (p < 0.05; Figure 7a–c). In stomach tissues, trypsin and pepsin activities were also significantly higher after feeding than before feeding (p < 0.05), whereas lipase activity did not differ significantly between pre- and post-feeding conditions (p > 0.05; Figure 7d–f). The quantitative fold changes in trypsin, pepsin, and lipase in the digestive gland before and after feeding were 1.93, 1.23, and 1.39, respectively. Moreover, the stomach tissues were 1.65, 2.02, and 0.83, respectively. Overall, feeding on oysters resulted in significant alterations in digestive enzyme activities in T. luteostoma.

4. Discussion

4.1. Potential Application of T. luteostoma in Biofouling Control

The excessive proliferation of fouling and sessile epibionts represents a major ecological constraint on bivalve aquaculture productivity. Building on our prior laboratory observations, we found that T. luteostoma displays a strong and selective feeding preference for oysters, a representative fouling taxon, highlighting its potential utility as a biological antifouling agent. To evaluate ecological safety and operational feasibility under realistic farming conditions, we conducted a field co-culture experiment. Importantly, co-culturing T. luteostoma did not significantly affect the survival of P. f. martensii. This outcome is ecologically plausible given their clear niche segregation in both habitat use and trophic strategy: T. luteostoma primarily inhabits nearshore intertidal zones and feeds on oysters and a range of benthic epifauna (e.g., small gastropods, barnacles, and bryozoans) [25,26,27,28], whereas P. f. martensii is typically associated with coral reefs, rocky substrates, and sandy–muddy bottoms and relies mainly on filter-feeding phytoplankton and organic particulates [29,30]. Accordingly, neither an explicit predator–prey relationship nor direct resource competition is expected, providing an ecological basis for the biosafety of introducing T. luteostoma into pearl oyster cage systems.

After establishing safety, the field trial further demonstrated the antifouling efficacy of T. luteostoma and its potential to enhance culture performance. As few as three individuals produced a significant removal effect on oysters, and six individuals achieved significant clearance of common fouling assemblages. Notably, co-culture not only maintained survival, but also significantly increased the growth of P. f. martensii. This growth enhancement is likely an indirect consequence of reduced fouling pressure and subsequent improvement of the cage microenvironment. Biofouling has been shown to hinder water exchange across cage boundaries, depress dissolved oxygen availability, and limit the dispersal of metabolic wastes [3,7]. In our system, sustained grazing by T. luteostoma on fouling biota from both shell surfaces and cage structures likely increased hydrodynamic permeability, thereby facilitating oxygen delivery and metabolite diffusion. In parallel, because fouling organisms can compete with cultured bivalves for particulate food and oxygen [31,32], their removal may reduce non-target consumption of nutrients and oxygen, allowing P. f. martensii to allocate more effectively to feeding and growth. Moreover, dense epibiont coverage can physically restrict shell opening, compromise filtration and respiration efficiency, and ultimately depress growth or elevate mortality [6,33,34,35]. By clearing shell surfaces and freeing culture space, T. luteostoma may alleviate mechanical interference, shading, and chronic physiological stress associated with heavy fouling loads, thereby providing a more favorable environment for P. f. martensii.

Although trophic segregation suggests ecological compatibility between T. luteostoma and P. f. martensii, cage culture is a confined, artificial system that may increase encounter frequency and potentially compress microhabitat space. Under such conditions, interference or competition for attachment surfaces could occur despite niche differentiation. However, within the stocking densities tested here, we detected no evidence of negative interactions: survival was unaffected, no shell damage or persistent attachment behavior was observed, and pearl oyster growth was significantly enhanced. These results indicate that spatial confinement did not translate into measurable interference under experimental conditions, although density-dependent effects warrant further evaluation in long-term or large-scale applications.

In summary, T. luteostoma in pearl oyster aquaculture exhibits a combined ecological profile characterized by negligible impacts on the cultured host, effective suppression of fouling organisms, and concomitant growth benefits, supporting its potential application as a functional biological antifouling component in cage-based culture systems.

4.2. Selective Feeding Behavior and Physiological Responses of T. luteostoma

Characterizing feeding behavior is a fundamental prerequisite for elucidating the mechanisms underlying selective feeding. Our indoor behavioral assays showed that T. luteostoma exhibits a pronounced feeding preference for oysters, which is consistent with its effective removal of fouling oysters under field co-culture conditions. Because marine gastropods generally have limited visual resolution, they typically rely on highly developed chemosensory organs (e.g., tentacles, rhinophore, and osphradium) to detect prey-derived feeding cues in aquatic environments [36]. Trajectory reconstruction and residence hot-spot analyses in this study indicated that T. luteostoma frequently performs non-directional exploratory movement before feeding; once it approached oysters, its movement shifted from exploratory to directed orientation, followed by the formation of a stable residence hot spot in the oyster area. This spatiotemporal transition from “exploration–orientation–feeding” suggests that the effect of attractant chemical cues on T. luteostoma may be distance-dependent. Supporting evidence has been reported in other marine gastropods; for example, during predation by Rapana venosa, the proportion of individuals achieving successful foraging reached 33.00–51.11% when the predator–prey distance was less than 1 m, but declined markedly to 8.89–15.57% when the distance increased to 2 m or more [37]. Based on this, to study the effectiveness of feed behavior, we selected an experimental tank with a length and width, and a diagonal edge, each less than 1 m. We observed that T. luteostoma can trigger a linear path when it is close to the prey (within the tank’s length or width). When T. luteostoma is located at the diagonal of the prey, it is rarely observed that its linear path tends to feed, this is consistent with the distance-dependent in the above study. Accordingly, we infer that T. luteostoma switches from exploration to sustained feeding only after approaching oysters, when local dissolved cues reach a perceivable intensity, whereas low-concentration cues at longer distances may be insufficient to elicit a clear oriented feeding response. However, the reasons for T. luteostoma’s specific feeding preferences for oysters in complex aquatic environments compared to P. f. martensii remain unclear. We speculate that oysters may stimulate specific digestive enzyme preferences in T. luteostoma, or that their nutrients are better aligned with T. luteostoma’s growth needs. Further research is urgently needed to validate these speculations.

Predation and effective ingestion typically induce pronounced digestive physiological responses. For instance, digestive enzyme activities increase rapidly after feeding in marine gastropods such as Laevistrombus canarium and Neptunea cumingii [21,22]. To verify that the selective feeding observed in T. luteostoma represents effective feeding, we quantified changes in key digestive enzyme activities before and after feeding. We found that following selective feeding, protease and lipase activities in the digestive gland of T. luteostoma increased significantly, indicating that selective feeding indeed triggers digestive physiological responses and thus constitutes effective feeding.

Notably, in contrast to the general upregulation of enzyme activities in the digestive gland, gastric lipase activity did not change significantly before versus after feeding. This discrepancy may reflect functional partitioning within the gastropod digestive system. The stomach primarily mediates mechanical grinding and initial digestion and may exhibit a more sensitive enzymatic response to components that are readily hydrolyzed (e.g., proteins), whereas lipid emulsification and subsequent hydrolysis depend more strongly on the enzyme system of the digestive gland and its associated intra-/extracellular digestive processes [38,39,40]. Consequently, lipid digestion may occur predominantly at the digestive-gland stage, resulting in a non-significant change in lipase activity in the stomach tissue.

5. Conclusions

Collectively, the results from controlled behavioral observations, field co-culture experiments, and digestive enzyme assays demonstrate that T. luteostoma functions as an effective biocontrol grazer in pearl oyster culture systems. Under co-culture conditions, T. luteostoma did not measurably compromise the survival of P. f. martensii, yet it significantly reduced oysters and other common sessile fouling taxa. This fouling suppression is likely to improve local hydrodynamic exchange and resource availability within culture units, thereby alleviating microenvironmental constraints and ultimately enhancing host growth. These findings support the feasibility of deploying T. luteostoma as a functional biological antifouling factor in pearl oyster aquaculture. Given a robust balance among biofouling removal efficiency, production costs, and potential space constraints in marine bivalve aquaculture, we recommend deploying 6 T. luteostoma per 25 cultured individuals.

Mechanistically, selective feeding in T. luteostoma was characterized by a stereotyped “exploration–orientation–feeding” behavioral shift, consistent with a distance-dependent, threshold-mediated activation by dissolved attractant cues. Moreover, the marked post-feeding elevation of protease and lipase activities in the digestive gland confirms that the observed selectivity reflects effective ingestion rather than transient probing, providing physiological support for its antifouling effect.