Hunting behavior of a solitary sailfish Istiophorus platypterus and estimated energy gain after prey capture

We used a custom designed biologging tag package with onboard video to describe a 3D high-resolution pursuit between a solitary sailfish and an individual small tuna in open water, representing the first time such an interaction has been documented. The sailfish was tagged at 09:53 on 18 October 2019, and the tag package remained attached to the sailfish for 67 h. However, analyses here are limited to the 24 h period in which the predation event took place (19 October–20 October; ~ 14 h after tagging and ~ 9 h after post-release recovery¹⁸) because this coincides with the time period the video camera was recording during daylight hours (on at 0600, off at 1800, sunrise and sunset, respectively) enabling us to ground-truth acceleration signals. Biologging data and accompanying video show the sailfish performing oscillatory dives between the surface and depths of 40–50 m during daylight hours. At night, fewer dives were performed and the sailfish generally remained within the top 10–20 m of the water column (Fig. 1a), leading to a greater range of temperatures experienced during the day (day 20.9–27.9 °C; night 26.5–28.2 °C). Due to the temperature dependence of the estimated active metabolic rate (AMR_E), the cooler temperatures at depth led to a reduced AMR_E during daylight hours (212.9 ± 89.1 mgO₂ kg⁻¹ h⁻¹) compared to night (224.7 ± 44.4 mgO₂ kg⁻¹ h⁻¹). Additionally, AMR_E initially increases with depth due to increased swim speeds during diving (Fig. 1b), until the thermocline is reached in the 30–40 m depth bin, at which point AMR_E decreases with further increased depth (Fig. 1b, c). However, due to thermal inertia of large-bodied fishes^19,20,21, it is possible that the sailfish’s body retained heat during the short (14.7 ± 1.7 min) excursions below the thermocline and did not drop to ambient temperature. As such, the metabolic rate calculated at depth may be underestimated with the temperature correction performed here. For example, during the dive in which the predation event occurred (Fig. 1; Table 1), if body temperature was assumed equivalent to surface temperature throughout the dive, estimated metabolic rates would increase by 18% compared to if the metabolic rates were temperature corrected according to the tag’s external temperature reading (Table S2). Yet, because the majority (> 90%) of time over the 24 h was spent above the thermocline, the temperature correction has little impact on the daily calculated AMR_E and subsequent energy expenditure (< 1% difference; Table S3).

The sailfish exhibited greatly reduced tailbeat activity and swimming speeds (≤ 0.25 ms⁻¹; 0.14 body lengths [BL] s⁻¹) when near the surface (Fig. 1), characteristic of basking behavior exhibited by swordfish and other istiophorid billfishes^22,23. Basking is believed to serve thermoregulatory purposes²³, but would also serve to reduce energy expenditure for billfishes facilitated by their swim bladders²⁴. Indeed, basking behavior observed here led to a significant reduction in AMR_E (186.6 ± 3.1 mgO₂ kg⁻¹ h⁻¹; T₃₇₅₂₀ = − 158.3, p < 0.001), when compared to active swimming behavior with strong and sustained tailbeats during dives (261.1 ± 91.1 mgO₂ kg⁻¹ h⁻¹; mean swimming speed 0.56 ± 0.2 ms⁻¹; 0.3 ± 0.1 BL s⁻¹).

The dive in which the predation event took place occurred roughly 31 h into the 67 h that the tag package remained attached to the sailfish (Fig. 2a). At 16:15 19 October 2019, the sailfish dove from the surface to a depth of 62.4 m with a mean (± SD) vertical velocity (VV) of 0.24 ± 0.21 ms⁻¹, where it remained for a short period before ascending to ~ 40 m (Fig. 2b). During the ascent, multiple possible prey items are seen in the video (Fig. 2c), and there was a brief increase in speed and tailbeat frequency (TBF), before the fish’s depth leveled off for ~ 2 min. The fish then dove again to 57.5 m, where visible light almost completely attenuated (Fig. 2d), and a change in locomotory mode from slow and steady swimming to rapid and forceful tailbeats occurred, beginning a rapid ascent (VV = 1.3 ± 0.43 ms⁻¹), with speeds reaching 3.1 ms⁻¹ (1.7 BL s⁻¹), and a body pitch of 54.6 ± 16.1° (maximum of 77.6°; Fig. 2b). It should be noted that due to the rapid ascent, the temperature readout of the tag lagged behind true ambient temperature (e.g., temperature of the descent compared to temperature of the ascent; Fig. 2b). Summary statistics for the dive in which the predation event took place are compared to all other daytime dives (Table 1).

The prey that was pursued during the predation event first became visible in the video when the sailfish reached the surface (Fig. 2e). The sailfish made several attempts to capture the prey, often breaking the surface of the water (supplemental video). From the video, the prey appeared to be a frigate or bullet tuna (Auxis thazard brachydorax or A. rochei eudorax), both of which are common in the region and known sailfish prey^5,25. During the rapid ascent and while at the surface, TBF and swimming speed remained high (1.6 ± 0.7 Hz and 1.7 ± 0.84 ms⁻¹, respectively, maximum of 2.92 ms⁻¹). At the surface, there were frequent changes in heading and the tuna appeared in the video several times (Fig. 2b; Fig. 3d; supplemental video). At one point, the tuna engaged in antipredator behavior presumably to ‘hide’, by swimming very close to the sailfish in front of the video camera along its right flank and out of its peripheral view (Fig. 2f; supplemental video). After roughly 60 s from the tuna’s first appearance on camera, the video and biologging data suggest that the sailfish caught the tuna or terminated the pursuit (Fig. 2b). Because the mouth of the sailfish was not in view of the camera, it is uncertain if the foraging attempt was successful; however, the tuna was last seen directly in front of the sailfish, immediately followed by a headshake (often characteristic of swallowing / prey manipulation for shallowing) and resumption of slow steady swimming by the sailfish, suggesting it was successful (Fig. 2b; supplementary video).

During the 24 h monitoring period, the mean estimated active metabolic rate (AMR_E; mgO₂ kg⁻¹ h⁻¹) of the sailfish for the median, 25th and 75th percentile of all 10,000 iterations were 218.9 ± 70.5, 156.6 ± 48.4 and 307 ± 102.9 mgO₂ kg⁻¹ h⁻¹, respectively (Table 2, Figure S5). Using the median iteration, during the dive where the predation occurred, mean AMR_E was 518.8 ± 586.3 (IQR 361.2–748.5) mgO₂ kg⁻¹ h⁻¹ (Fig. 3). From this median iteration, we estimate that 2.7 MJ (IQR 1.9–3.8 MJ) of energy was expended over the course of the day, where only 1% (0.04 MJ) was expended during the pursuit (Table 2). The estimated energy content of the tuna was 5.1 MJ [calculated from 26], and assuming a successful predation outcome, this encounter resulted in a net energy gain of 2.4 MJ (IQR 1.3–3.2 MJ). However, if this predation was unsuccessful, the cost of this pursuit was only 1% of the energy expenditure for the day. For a sailfish of this size, this daily energy expenditure equates to the required consumption of ~ 0.5 tuna d⁻¹ to sustain daily metabolic costs estimated for the median AMR_E (218.9 mgO₂ kg⁻¹ h⁻¹; see Sect. 3 of supplemental methods for details of calculation).

We observed a willingness for this individual sailfish to attack from both sides of the prey. The alternating pattern of positive, negative, positive, negative (Fig. 4a–d) in the degrees of rotation s⁻¹ immediately prior to and during each strike suggests that the sailfish attacked from different sides of the prey in each successive strike. This behavior is in contrast to Kurvers et al.⁸, who found that during bait-ball hunting aggregations with multiple sailfish, individual sailfish were strongly lateralized and would tend to strike from the same side each time they entered the bait-ball. Attacking from different sides in succession is a novel finding for sailfish and suggests behavioral plasticity within different hunting scenarios (i.e., group vs solitary hunting). In one-on-one hunting situations when neither the predator nor prey are in a group setting, it would benefit the predator to avoid lateralization because the prey can quickly learn any tendencies the predator may have^8,34.

One noteworthy finding was the relatively low maximum speed attained by the sailfish (3.1 ms⁻¹) during the encounter with the prey. Sailfish are believed to be one of the fastest swimming fish^24,27 with recent estimates suggesting maximum speeds of 8.2–8.3 ± 1.4 ms^−19,28, and predator–prey interactions might be expected to be events where maximal speeds are exhibited by both predator and prey⁹. However, because it is the prey that sets the speed, timing of accelerations, decelerations and turns, the predator is either reacting to or predicting what the prey will do to enable trajectory interception and capture, which culminates in lower than maximal predator speeds during pursuits²⁹. Additionally, theoretical models predict that if prey are slower than their predators, as is the case here³⁰, prey should avoid the predator by turning rather than trying to increase separation by travelling as fast as possible²⁹. For example, Wilson et al.²⁹ demonstrated that if a prey animal is moving as fast as possible, it cannot accelerate forward and must either turn or continue straight, making its movements more predictable and interceptable, compared to a slow moving prey that has more escape options (speed up, slow down, turn) and is therefore less predictable. As such, sailfish and their prey are likely avoiding maximum speeds during one-on-one encounters in open water³¹. Additionally, previous studies have noted that cursorial and avian predators will slow down in the moments prior to an attack to increase maneuverability when in close proximity to prey^29,32,33, which was also observed here (Fig. 4e–h). Furthermore, morphological adaptations of sailfish (i.e., the bill) can be moved through the water more rapidly than the whole body⁹, potentially allowing sailfish to rely on these morphological ‘weapons’, rather than speed during one-on-one pursuits to increase capture success rates. We also observed a tendency of the sailfish to approach the prey from below in each capture attempt (Fig. 4i–l).

Direct observation of natural predation events by marine predators are rare^35,36, particularly for pelagic fish predators where visual observation is difficult, prey is sparse, and feeding rates are low compared to that of marine mammals and seabirds³⁶. For the predation event presented here, based on the footage of potential prey items near the thermocline (Fig. 2c; supplementary video), the observed increase in speed and TBF immediately prior to the rapid ascent (Figs. 2b; 3a), and the shorter than average dive time (Table 1), we propose that the prey item was encountered at depth, and chased to the surface⁶. Given the shallow thermocline and co-occurring oxycline present in the Eastern Tropical Pacific³⁷, and the potential for these features to concentrate prey^38,39,40, we hypothesize that oscillatory dives in the mixed layer are prey-searching dives to increase foraging opportunities⁴¹. Due to their unique metabolic biochemistry suited to life in the open ocean, tunas and billfishes have long been described as energy speculators, gambling high and continual energy output, expecting to offset the costs with periodic high energetic gain events^42,43,44. The estimated energy gain of 2.4 MJ resulting from the prey encounter in the 24 h period described here is consistent with energy speculation behavior, but also suggests this sailfish would need to regularly forage on high energy prey to support its metabolic requirements. Because the video camera was only recording for the daylight hours of the 24 h period analyzed here, we cannot say if any other feeding events occurred during the remainder of the track; however, there were two other similar bursts of activity identified in the acceleration data during the following day prior to the tag releasing from the fish. As such, energy obtained from individual foraging events like that described here may be an important energy supply for the routine energy requirements of these high metabolic performers between opportunistic, large energy gain, group hunting bait-ball foraging events where multiple prey can be consumed in a short amount of time.

Source: Ecology - nature.com