Movement Behavior of the Dusky Grouper Epinephelus marginatus (Lowe, 1834) in Early Life Stages

: The dusky grouper ( Epinephelus marginatus ) is a vulnerable predatory ﬁsh found in the Atlantic and Indian oceans, and in the Mediterranean Sea. This study investigates the movement patterns of grouper larvae (151 individuals) during their ﬁrst feeding period (three to ten days post hatching) th rough laboratory experiments oﬀering rotifers ( Brachionus sp.) and copepod nauplii as food. High-speed cameras and imaging techniques including bright ﬁeld microscopy, matched ﬁl-ters, and holography captured rapid displacements (up to 25 body lengths per second), peduncle caudal beats (up to 40 beats s −1 ), turns, and resting periods. Reynolds numbers ranged from <45 for routine swimming to 222 for burst swimming. Speciﬁc behaviors, including changes in swimming velocity and body movements, were observed from three days post hatching, when feeding begins, suggestive of an array of responses to environmental forcing, predator avoidance, food search, and feeding success. These ﬁndings provide insights into the critical movement performances of E. mar-ginatus larvae, which help to unravel their ecological interactions and survival strategies. Understanding grouper larval swimming behavior is pivotal for devising eﬀective cultivation strategies aimed at replenishing wild stocks and enhancing production for human consumption.


Introduction
Epinephelus marginatus (Lowe, 1834), commonly known as the dusky grouper, is a member of the Serranidae family within the order Perciformes.These fish are distributed across the Atlantic and Indian oceans, including the Mediterranean Sea, and are frequently found along the south, southeast, and northeast coasts of Brazil [1,2], typically inhabiting rocky shores and coral reefs.These groupers are listed as vulnerable species due to their status as predatory fishing targets and their high commercial value [3,4].They exhibit the potential for aquaculture due to their resilience in densely populated conditions and rapid growth at temperatures ranging between 23 and 25 °C [5].Both aquaculturists and researchers share the goal of achieving high production levels to meet market demands while alleviating pressure on wild populations [6].
During the early life stages of groupers, particularly the first feeding phase occurring around 3 days post hatching, a significant mortality rate is observed [7][8][9].This phase represents a critical period as larvae transition from endogenous to exogenous feeding, making behavioral studies on grouper larvae imperative yet limited.Larval fish with altricial development are highly vulnerable to predation [10] and must find suitable food (both in terms of quality and quantity) to ensure survival [9,11].The first feeding phase is crucial as it marks the larva's initiation into capturing food in the marine environment, accompanied by the development of sensory organs and fins [12,13].Larvae in this phase exhibit fragile bodies, typically characterized by a preflexion stage notochord, and display a continuous swimming behavior interspersed with rest periods [14,15].
Studies by Mahjoub et al. [16,17] on Epinephelus malabaricus larvae and Fui et al. [18] on Epinephelus fuscoguttatus larvae have explored the behavioral strategies and changes in sinking velocity and vertical distribution with age, respectively.Efforts have been made to breed E. marginatus in captivity [19], highlighting the need for further laboratory experiments to understand behavioral parameters and enhance larval survival [6,20,21].However, the movement behavior of E. marginatus during its early life stages remains largely unexplored.
Larval fish exhibit specific swimming patterns when searching for food and capturing prey [14], and such patterns respond to changing water quality conditions such as temperature, turbidity, and light intensity [22].Understanding these phenomena can help in the design of fish feeds that match with the larval behavior, ensuring better nutrition and higher survival rates in aquaculture facilities.Insights into larval movement can also inform tank design, water flow management, and the placement of feeding stations to maximize food accessibility and minimize waste [23].By aligning aquaculture practices with the natural behavior of fish larvae, we can enhance growth rates, improve health, and increase the overall efficiency and sustainability of fish farming operations [24].
This study aims to assist in filling this gap by investigating the movement behavior of E. marginatus larvae aged 3 to 10 days post hatching, focusing on understanding changes in routine speed, burst speed, caudal beat frequency, C-start duration, and rest duration in the presence and absence of prey.To achieve this goal, we employed highspeed video recordings using an optical system tailored for small-sized fish larvae and analyzed the acquired image sequences based on 2D Fourier and holography techniques.

Materials and Methods
A horizontal optical system (Figure 1) was configured on a Newport optical table, consisting of a matched spatial filter system (in the optical Fourier transform plane) combined with a bright-field system (with the frequency filter removed) [25].The setup included a laser beam with a wavelength of 657.5 nm, an objective lens (20× magnification), a pinhole (10 µm), collimator and focusing lenses (f 150 mm), and a high-speed digital camera, specifically the Photron Fastcam SA2 86K C3, which can capture up to 2000 frames per second, with a resolution of 2048 × 2048 pixels and a pixel size of 10 µm.
Another optical system was arranged both horizontally and vertically for in-line digital holography [26].This system closely resembled the previous setup but utilized a different collimator lens (f 400 mm) and a PhotonFocus camera model MV1-D1312-160-CL12, running at 30 frames per second, with a resolution of 1312 × 1082 pixels and an 8 µm pixel size.The positions of the camera and aquarium were determined using geometrical optics formulas [27], considering the image sensor and optical magnification ranging from 1 to 3.5 times.Holographic reconstruction was performed according to the methodology described by Ghiglieno et al. [28], utilizing a MATLAB routine.
Fish larvae of E. marginatus, aged between 3 and 10 days post hatching (dph), were sourced from Redemar Alevinos Aquaculture located in Ilhabela, on the north coast of São Paulo State, Brazil.The larvae were transported in plastic bags and maintained under controlled conditions of 25 °C temperature and 35 salinity using water from the cultivation system to minimize handling stress and mortality [29].The same conditions were kept until the onset of the experiments.
In the optics laboratory, the windows were sealed to prevent interference from natural light, being illuminated only with low-intensity, indirect fluorescent lighting.All larvae were kept in dark containers to reduce stress from potential external disturbances.Sound, physical, and mechanical disturbances were minimized in the laboratory to highlight only the influence of experimental stimuli on the behavior of the fish larvae, in this case, the presence of prey.About an hour before the experiments, the larvae were placed in the filming aquarium for acclimatization to the conditions to which they would be exposed.
The experimental design involved carefully analyzing E. marginatus fish larvae (n = 151) across all treatments and recording videos daily to track larval development.Five fish larvae were individually placed in small-volume aquariums (50 mL, 6.0 × 5.5 × 1.5 cm) filled with filtered seawater, for the side-view recordings.Additionally, a single fish larva was placed in a Petri dish (50 × 12 mm) filled with seawater to the top margin, to enable topside view imaging of body and fin movements during swimming.
Experiments with larvae aged 3 to 5 dph were divided into three sessions: (1) without feeding; (2) with 10 rotifers, Brachionus sp.; and (3) with 10 copepod nauplii.Larvae aged 6, 7, 8, and 10 dph were studied exclusively with Brachionus sp., as this is the food source routinely offered in the aquaculture facility.Additionally, we limited the use of copepod nauplii in our experiments to the early larval fish stages (3-5 dph) due to challenges in providing nauplii at the same developmental stage to grouper larvae beyond a 5-day timeframe.The concentration of organisms in the aquarium and prey density (ranging from 100 to 200 per liter) were determined based on studies by Hunter [30], Doi et al. [31], Russo et al. [20,32] and Cunha et al. [6] to optimize larval survival in the laboratory.
After each experimental session, fish larvae were preserved in a formalin solution (4%), following the guidelines of Conselho Nacional de Ética em Pesquisa (CONEP) and Código Brasileiro de Experimentação Animal (COBEA).Larval size (total length from snout to the final portion of the notochord) was measured using a micrometric scale under a stereomicroscope.
Data analysis involved recording the movements of individual fish and their interactions with prey using ad libitum and focal-animal sampling techniques [33].Swimming modes, velocity changes, and body movements were described according to established literature [14,[34][35][36][37], including routine and burst swimming behavior, caudal peduncle beat, C and S-turns, backward movement, and rest periods.
Instantaneous speed (mm s −1 ) was calculated as the ratio of distance traveled to time, expressed in millimeters per second.Relative speed (body lengths per second, BL s −1 ) was used to characterize swimming modes, with routine behavior below and burst swimming above 3.0 BL s −1 , based on established studies [14,[37][38][39].Reynolds numbers (Re) were calculated considering seawater kinematic viscosity, larval size (notochord length, mm), and swimming velocity.
Caudal beat frequency during swimming and maintenance of position was determined by lateral movements of the caudal peduncle and fin, expressed as beats per second (b s −1 ).Resting periods were classified as instances when larvae remained motionless with a complete interruption of movement [14].
Statistical analysis was conducted using non-parametric methods, specifically the Kruskal-Wallis test (p < 0.05) and Dunn's post hoc test, to assess differences in routine speed, burst speed, caudal beat frequency, C-start duration, and rest duration between treatments (presence and absence of prey) and with respect to the ontogenetic development of the larvae.The Past software version 4.03 was used for the statistical analysis and data plotting.

Results
We described the behavioral patterns observed in E. marginatus larvae through frameby-frame analysis encompassing a total of 90,336 frames, which corresponded to 442 behavioral events.The larval behavior included various types of swimming activities, such as routine (intermittent swimming and constant velocity) and burst (large magnitude and short duration) swimming, along with variations in velocity, caudal peduncle beats, body contractions, and rest periods.
With respect to larval development, we observed an increase in values for routine and burst swimming in dusky grouper larvae, considering age, size, and the presence of prey.Specifically, we noted significantly higher velocity values in larvae aged 6, 7, 8, and 10 days post hatching (dph) in the treatment with rotifers (Kruskal-Wallis, p < 0.05).
The body length of the larvae ranged from 1.88 to 4.20 mm across the 3 to 10 dph time interval.Relative routine swimming speed ranged from 1.0 to 2.9 body lengths per second (BL s −1 ), while relative burst speed ranged from 3.0 to 25 BL s −1 (Table 1).Median values for absolute routine swimming speed ranged from 2.0 to 6.0 mm s −1 , while burst speed median values varied between 10 and 30 mm s −1 , observed in the presence of copepod nauplii, rotifers, and no prey (Figure 2A,B).
Grouper larvae exhibited burst behavior in the absence of prey and in the presence of rotifers.The maximum absolute speed recorded was 76.7 mm s −1 , corresponding to 25 BL s −1 , observed in 8 dph larvae in the presence of rotifers, although this increase was not statistically significant (Kruskal-Wallis, p > 0.05).
Backward movements, characterized by the opposite direction from forward swimming and movement of pectoral fins, were observed more frequently in treatments with prey, regardless of larval age, but without significant differences (Kruskal-Wallis, p > 0.05).The absolute speed recorded during these movements ranged from 0.65 to 2.9 mm s −1 .During routine swimming, the average beats of the caudal peduncle were 40 ± 20 beats per second (b s −1 ), ranging between 11 and 120 b s −1 .Significantly, higher values (Kruskal-Wallis, p < 0.05) were observed with respect to larval development in the rotifer treatment (Figure 3A).For burst swimming, caudal beat frequency ranged from 50 to 222 b s −1 , with the maximum value observed in 3 dph larvae in a single event.Caudal beating was also observed during position maintenance, with higher values noted in the presence of prey and according to larval development, although without significant differences (Kruskal-Wallis, p > 0.05), with a mean of 38 ± 15 b s −1 (Figure 3B).The Reynolds number (Re) analysis revealed the expected increase in Re according to larval development, with a significant difference observed between burst swimming and routine swimming (Kruskal-Wallis, p < 0.05).During routine swimming, Re varied from 1 to 45, while during burst swimming, it ranged from 14 to 242.
The methods employed in this study, including 2D-matched filters and digital holography in-line, facilitated the observation of pectoral fins (Figures 4 and 5) and dorsal spine (Figure 6) movements during swimming behavior and position maintenance in E. marginatus larvae.Pectoral fins exhibited alternating contraction and rotational movements, while larvae with 8 and 10 dph showed developed dorsal spines, with associated movements observed during larval displacement.Additionally, rapid contractions of the body transitioning from S to C shapes were observed, such as in the example shown in Figure 7, depicting a sequence of frames capturing the initial bend of the body in S-shaped, followed by a change in head position and a C-shaped curvature.C-turns were more frequent and lasted longer in the early larval stages, in the presence of prey (Figure 8A).The duration of body contractions decreased with larval development, and significant differences were noted in treatments with rotifers and copepod nauplii (Kruskal-Wallis, p < 0.05).The duration of rest periods (Figure 8B) had a significant decrease with age (Kruskal-Wallis, p < 0.05), whether prey items were present or absent.These findings elucidate distinctive behavioral patterns in Epinephelus marginatus larvae, ranging from routine swimming to complex maneuvers, which differ significantly with the larva's age and are notably influenced by the presence or absence of prey.Younger larvae exhibit more frequent burst and rest swimming behaviors, which are crucial for escaping predators and capturing prey, whereas older larvae show more advanced and strategic movements.

Discussion
Epinephelus marginatus exhibits specific behavioral patterns that vary with ontogenetic development and in the presence of prey.The mean size of larvae during their first feeding (3 dph) was 2.02 mm, slightly smaller than the 2.7 mm reported by Russo et al. [32] under similar environmental conditions (25 °C temperature and 34 salinity).This variation in size could be influenced by a combination of endogenous factors such as genetic characteristics and maternal contributions, as well as exogenous factors like temperature, salinity, oxygen concentration, food availability, and predation pressure [39].
The relative routine swimming speeds ranged from 1 to 3 body lengths per second (BL s −1 ), with pauses during displacement and absolute values between 4.0 and 6.0 mm s −1 .Similar data were found in other species at the same ontogenetic stage, such as Engraulis mordax [14], Gadus morhua [40], Trachurus japonicus [41], and Anchoa mitchilli [42].There was a notable increase in routine swimming speed with larval development, likely attributed to the development of locomotion structures, including rays and spines of fins, muscle growth, and overall growth [13,[43][44][45].The increase in sensory organ development, such as free neuromasts, also contributes to the larvae's ability to detect prey stimuli and adjust their behavior accordingly [46,47].
Routine swimming behavior was consistently observed in all treatments, regardless of the presence or absence of planktonic prey.This behavior indicates that fish larvae, including Epinephelus marginatus, actively search and explore for food, even when prey is not immediately available.On the first feeding day (3 dph), there was a decrease in routine swimming speed, which may be related to strategies for capturing food efficiently, such as reducing disturbances and conserving energy when prey availability is not limited.This behavior aligns with observations in other species like Epinephelus malabaricus, as reported by Mahjoub et al. [16] at 21 dph.
Similar behavioral patterns to those of E. marginatus larvae have been observed in various marine fish species.For instance, clown fish larvae (Amphiprion perideraion) [48]), cod larvae (Gadus morhua), and turbot larvae (Scophthalmus maximus) [49] exhibit variations in swimming velocity depending on the availability of food.Dusky grouper larvae, when exposed to rotifers, showed a significant increase in routine swimming behavior as they developed, a behavior noted in coral reef fish larvae by Fisher and Bellwood [50,51].These differences in behavior highlight the capacity and flexibility of organisms in adjusting their behavior based on the characteristics of the available prey [52].
The maximum value of burst swimming speed observed in E. marginatus larvae was 25 body lengths per second (BL s −1 ), with higher values recorded in older larvae.This burst swimming behavior is integral to the swimming patterns of E. marginatus and aids in capturing prey.Hunter [14] recorded burst swimming speeds of 15 BL s −1 in anchovy larvae (Engraulis mordax), while mackerel larvae (Scomber japonicus) exhibited velocities ranging from 5 to 20 BL s −1 [53].Remarkably, burst swimming behavior was observed in E. marginatus larvae across all treatments, even in the absence of prey.This suggests that the presence of conspecifics in the observation vessel may have influenced this behavior, as burst swimming is often associated with predator avoidance and attempts to catch prey [54].
Lower Reynolds number (Re) values indicate that viscous forces play a predominant role in the movements of E. marginatus larvae during routine swimming as they explore the environment and search for food.Conversely, higher Re values suggest that inertial forces are more influential during burst swimming, particularly when the larvae are engaged in rapid movements.Re values vary among species and are influenced by factors such as temperature and viscosity of the surrounding medium.For instance, Coughlin et al. [48] calculated an Re of 38 for Amphiprion perideraion larvae at 24 °C, while Fuiman and Batty [37] reported an Re of 121 for small Clupea harengus larvae at temperatures between 10 and 14 °C.However, they noted that viscous forces can extend up to Re 300 for herring larvae with a body length of 9.6 mm.
The movements of fins and the caudal peduncle during both routine and burst swimming are described in detail in this study for E. marginatus larvae.Small larvae are capable of rapidly beating their caudal fin in short intervals, generating significant force that enhances their swimming performance.The beat frequency can increase with speed, showing a positive linear relationship regardless of larval size, as observed in E. marginatus (with a maximum value of 222 beats per second in 3 dph larvae).The caudal beat frequency of E. marginatus larvae was found to be higher than that of other species such as Engraulis mordax [14], Pleuronectes platessa, and Clupea harengus [34].
The pectoral fins of dusky grouper larvae exhibit an alternate abduction and adduction movement, where one fin contracts and remains close to the larval body while the other moves rotationally.Using digital particle image velocimetry (DPIV), Drucker and Lauder [55] observed vortexes in the fluid that support the displacement of bluegill sunfish (Lepomis macrochirus), through their pectoral fin movements.Pectoral fins play a crucial role in foraging behavior, including backward movement, and acting as oars, aiding in food acquisition and collision avoidance.The backward movement observed in E. marginatus larvae, driven by pectoral fin movements, occurs with velocities ranging from 0.25 to 9.6 mm s −1 .This movement strategy is characteristic of specialized fish that maneuver through environments such as caves and orifices, contributing to their ability to search for food and occupy specific habitats.
The rays and spines of fin fishes play a crucial role in displacement and position maintenance, primarily due to the increase in surface area.In E. marginatus larvae at 8 and 10 days post hatching (dph), movements of the dorsal spines (expansion and contraction) were observed during displacement.When contracted, these spines reduce friction in a viscous medium, while their expansion helps decrease the sinking rate by increasing the body surface area, also serving as a defense mechanism against predators, as reported by Russo et al. [32] and Leis and Yerman [56].
Fast turns of the larval body are considered an initial response mechanism to predator attacks, involving a short latent period where the fish accelerates rapidly and changes its position.The curvature of the body typically resembles a letter C, S or J shape [57], possibly developed through evolutionary processes under selective pressures on fishes.Both C and S shapes were observed in E. marginatus larvae, including their swimming pattern.Body flexion occurred across all treatments, with or without food, contributing to prey attack strategies and escape maneuvers, as extensively reported in the literature for other fish species [58,59], and across the animal kingdom [60].
The S-shaped curvature has been observed in organisms with elongated bodies, such as eels [61], larvae of Engraulidae and Clupeidae [14,30], and cod larvae, Gadus morhua [52].In E. marginatus, S-turns had a shorter duration (0.17 ± 0.53 s) compared to Engraulis mordax (0.82 s) [14].Russo et al. [20] also recorded S-contractions of E. marginatus during prey capture.In this study, S curvature was observed preceding C-shaped contractions in all movements, with a higher occurrence in treatments involving prey.This emphasizes the relationship between body muscle contractions and feeding behavior, indicating a coordinated response in capturing prey [57,59].
C-shaped contractions in E. marginatus larvae are linked to the presence of Mauthner cells in the brain and their interaction with specialized axons, as discussed by Eaton et al. [62] and Eaton and DiDomenico [35].This coordinated behavioral pattern prepares the fish for swimming, protects the vulnerable head region, optimizes escape efficiency against predators [63], and potentially contributes to increased survival in larvae [64].A change in the head's direction in E. marginatus larvae is preceded by a C-turn, leading to repositioning and propulsion during displacement.This constant change in direction can aid predators with high maneuverability, such as groupers, in their search for food, similarly to the observed behavior in cod larvae [52] and clownfish larvae [59], which exhibit turns during the attack position after visualizing prey.
Resting behavior in fish larvae is commonly observed during early life stages [65].The rest periods in dusky grouper larvae decreased as the larvae developed, aligning with observations in anchovy larvae, E. mordax, which almost completely cease resting behavior by 4 dph [14].
Fish larvae selected for this study were in a critical period, as described by Hjort [66], also known as the first feeding stage, which is associated with a high mortality rate.E. marginatus initiates first feeding at 3 days post hatching [20].Consequently, a high mortality rate in the laboratory setting (70-100%) led to gaps in the experiments, with 9 dph larvae not included in the study.Additionally, the mortality of copepod nauplii affected the experiments, leading to 5-day post-hatching larvae not being tested with nauplii.
The prey detection ability of fish varies according to species and ontogenetic stages, depending on the amplitude of stimuli, including angle, intensity, light absorption, and the distance between predator and prey [67,68].The relationship between fish larvae and prey is complex and nonlinear, influenced by factors such as vertical migration, environmental clues, and match-mismatch of fish and zooplankton occurrences [69,70].The behavioral patterns observed in dusky grouper larvae, as described in this study, reveal a high level of coordinated muscle control and complexity compared to previous findings by Glamuzina et al. [8] for the same species.Contrary to earlier beliefs that newly hatched larvae exhibit minimal movement and control, merely drifting passively in sea currents, as proposed in the Simplification Hypothesis by Roberts [71], this research highlights the active and purposeful behavior of fish larvae.Studies by Leis [72] and Houde [73] have also contributed to challenging the notion of fish larvae as passive particles.

Conclusions
We found that the highest values of routine swimming speed in grouper larvae occurred in the absence of prey, suggesting active food searching behavior.In the presence of prey, grouper larvae showed a reduction in routine swimming speed and more frequent events of burst swimming behavior, with C-shaped contractions towards nauplii, and routine swimming towards rotifers.
The observed swimming patterns indicate the ability and flexibility of fish larvae to adjust their behavior to the different characteristics of prey, involving muscle contractions, movements of the caudal complex, specialized movement of pectoral fins, and movement of dorsal spines that provide the undulating body movements during displacement and changes from resting periods.Food search and capture movements indicated a relationship with swimming behavior patterns, especially variations in swimming speed and muscle contractions in C and S shapes.
Investigations into the behavior of fish larvae reveal important information about ontogeny, ecological interactions, survival strategies, feeding, and vertical migration patterns [74], which can assist in improvements for the cultivation of potential species, such as groupers.The findings presented here contribute to the knowledge held on specific biological aspects of E. marginatus, including the control of movements and interactions with the environment from early life stages.Behavioral studies in fish are crucial for the implementation of best practices in marine aquaculture, although there are still many challenges in this field.The genomic uniqueness of the Epinephelinae subfamily, which includes groupers, compared to other serranids and perciforms, as indicated by Zhuang et al. [75], highlights the importance of continued research on this group of fish across all biological aspects.
Epinephelus marginatus is considered a vulnerable species [76], mainly due to predatory fishing, but factors such as global warming and ocean acidification significantly contribute to reducing the growth of fish larvae and consequently interfere with the success of feeding and the survival of fish species in general.To optimize the conservation and cultivation of this species, understanding their unique larval behavior and movement is essential.These insights are crucial for implementing customized aquaculture practices that not only enhance grouper production but also ensure the replenishment and sustainability of their populations in natural habitats.

Figure 1 .
Figure 1.Schematic representation of the in-line holographic system utilized for observing larval fish.The matched spatial filter system was additionally composed by a focusing lens and a frequency filter, both positioned between the vessel and the camera.Red and blue colors represent the laser beam and the seawater, respectively.

Figure 2 .
Figure 2. Routine (A) and burst (B) absolute swimming speed (mm s −1 ) of E. marginatus larvae (3 to 10 dph (days post hatching)) fed copepod nauplii and rotifers, and with no prey.Horizontal bars represent median values.Note difference in the scales of speed range in (A,B).

Figure 3 .
Figure 3. Caudal beat frequency (beats per second) during routine swimming (A) and position maintenance (B) fed copepod nauplii and rotifers, and with no prey.Horizontal bars represent median values.Note the difference in beat frequency range in (A,B).

Figure 4 .
Figure 4. Temporal sequence of pectoral fin (dorsal view) movements of a 6 dph (days post hatching) E. marginatus larva, as seen after holographic reconstruction, where t is the time in seconds.Schematic drawings illustrate frame by frame movements.

Figure 5 .
Figure 5. Temporal sequence of pectoral fin (dorsal view) movements of a 5 dph (days post hatching) E. marginatus larva, acquired by the 2D matching spatial filter technique.

Figure 6 .
Figure 6.Temporal sequence of dorsal spine (lateral view) movements of a 10 dph (days post hatching) E. marginatus larva.Dorsal spine expansion at t = 0 s in the rest position, followed by full contraction at t = 0.6 s during a displacement.

Figure 7 .
Figure 7. Temporal sequence of a 5 dph E. marginatus larva.S-turn is in t = 0 to t = 0.005 s followed by a C-shaped body curvature (t = 0.009 to t = 0.013 s).

Figure 8 .
Figure 8. C-turn (A) and rest time (B) duration of E. marginatus larvae from 3 to 10 dph, with and without prey.Horizontal bars represent median values.

Table 1 .
Mean of size, relative routine and burst swimming (body lengths per second, BL s −1 ) of E. marginatus larvae, according to age (3 to 10 dph (days post hatching)); # is the number of observations, considering all treatments.