Deadly interaction between a swordfish Xiphias gladius and a bigeye thresher shark Alopias superciliosus

Interspecific aggressive behaviour has been reported for a number of fish species, including not only coral reef fishes that compete for territory (Myrberg Jr and Thresher 1974; Eurich et al. 2018), but also large pelagic fishes, such as billfishes (Istiophoridae and Xiphiidae), which are known to exhibit aggressive behaviours towards other species. The swordfish Xiphias gladius, which is a highly mobile, predatory fish is known to attack sharks (e.g. Starck 1960; Penadés-Suay et al. 2017, 2019; Romeo et al. 2020), whales (Machida 1970), sea turtles (Frazier et al. 1994), humans (e.g. Gooi et al. 2007; Georgiadou et al. 2010), and even inanimate objects, e.g. boats and submarines (Zarudski and Haedrich 1974; Romeo et al. 2017). The resulting impact can lead to a transverse fracture of the bill, with the distal segment often remaining embedded in the victim. Aggressive behaviour towards thresher sharks (Lamniformes, Alopiidae), however, was only suggested once before in a conference contribution (Vacchi et al. 1999). Here, we present the first verified illustrated documentation of a stranded bigeye thresher Alopias superciliosus that was fatally wounded by a swordfish and discuss this incidence within the context of a possible accidental collision or interspecific aggression.

In April 2020, a dead female thresher shark was found washed ashore on the Libyan coast near the town of Brega (30°25′18.1”N, 19°37′38.9”E). Videos and photos of the animal were taken on April 1st and April 4th and sent to the citizen science initiative “Marine biology in Libya”. The species was identified based on the following features (sensu Ebert et al. 2013): (1) long curving upper tail lobe nearly as long as the rest of the shark; (2) body grey-brown on the dorsal surface and light grey to white ventrally, not extending above pectoral fin bases; (3) huge eyes, orbits expanded onto dorsal head surface; (4) deep horizontal groove above gills; (5) very long narrow pectoral fins (Figs. 1 and S1).

Fig. 1.

Female bigeye thresher Alopias superciliosus (TL = 445 cm) stranded on the Libyan coast (Mediterranean Sea), with a swordfish Xiphias gladius rostrum embedded deep in the branchial chamber. Scale bars indicate 50 cm (b) and 10 cm (c and d). Photo (a and b) and video content (c and d) courtesy of Faraj Habrisha and Abdalhakim Ahmed Al sebaihe

A total length (TL) of 445 cm was measured by one of the citizen scientists, which was supported by analysing the images and comparing the dimensions of the shark with an object of known size (i.e. the right hand of the observer). Female bigeye threshers mature at a TL of 330-350 cm (Ebert et al. 2013). Therefore, the examined specimen is regarded to represent an adult individual. All measurements were taken digitally to the closest mm.

The animal did not display any external injuries, except for a single penetrating trauma (7.9 cm in width and 5.0 cm in height) on the dorsal surface at the level of the pectoral girdle with the tip of a rostrum stuck in it (Figs. 1c and S2). The rostrum was identified as belonging to a swordfish X. gladius based on the flattened appearance in cross section (i.e. depth less than half the width) and the absence of denticles on the surface (Fierstine and Voigt 1996). The fragment was 30.1 cm long and had a distal and proximal width of 1.4 and 5 cm, respectively (Fig. 1d).

To reconstruct the total length of the rostrum, the ratio of maximum width to length of the fragment was calculated (R = 0.166) and compared with the dimensions of a previously published intact rostrum (Penadés-Suay et al. 2017) to determine the breaking point. Assuming an isometric relationship, the distance from the tip of the rostrum to the fracture point was ca. 29–32% of the total rostral length (TRL), resulting in a TRL of ca. 94–103 cm. Swordfish are known to have the longest rostra among all billfish species, making up one-third and more of the TL (Hoey et al. 1989). Under this assumption, the swordfish involved in the attack was estimated to have a TL of ca. 280–310 cm and a body length without rostrum of ca. 190–210 cm, corresponding to an adult individual (Macías et al. 2005).

The rostrum penetrated the shark in an acute angle of approximately 70° to the anterior–posterior axis, directed towards the branchial apparatus, revealing that the sword-fish must have been positioned dorsally from the shark and pierced it from behind (Figs. 1c, 2 and S2). The internal damage caused by the impact cannot be reconstructed, because the citizen scientists did not perform any necropsy. However, based on the location, the length and the angle of the impaling rostral portion, it can be that the shark was injured at or near the vertebral column and the branchial apparatus, probably severely injuring branchial nerves, branchial arteries and gill arches (Fig. 2). The injury was inflicted recently when the shark was found dead on the beach on April 1st, because the epaxial musculature showed no signs of wound healing or other alterations but still was supplied with blood and exhibited a pink-reddish colouration (Fig. 1c). This colouration was faded by the date of the second inspection three days later (Fig. S1). The location of the injury, timing of wound infliction and lack of other apparent injuries lead us to the conclusion that the impalement was fatal and the ultimate cause of death for the thresher shark.

Fig. 2.

Tentative graphical reconstruction of the potential damage inflicted by the distal segment of a swordfish Xiphias gladius rostrum to the anterior-most body including the branchial apparatus of a bigeye thresher Alopias superciliosus. The depth of the penetrating wound was reconstructed based on the length of the swordfish bill fragment (30.1 cm) which was fully embedded in the shark

Although impalements by swordfish are not uncommon and were reported to occur against a number of species and objects, the nature underlying these attacks, whether they are instances of accidents or were carried out by purpose, still is not fully understood. The spearing of inanimate objects like bales of rubber and ships that cause no apparent threat to the swordfish were usually regarded as hunting accidents and the result of attempts made by the billfish to capture prey that sought shelter around such objects (Stark 1960; Frazier et al. 1994). However, there is rising evidence that swordfish deliberately utilize their rostra also as weapon when provoked. Romeo et al. (2017) analysed swordfish attacks on harpoon vessels and observed aggressive behaviour as a response to prior provocation or when male swordfish wanted to defend females that had been harpooned during the breeding season. A defensive behaviour is also reported to be the cause of swordfish attacks on humans, of which most involved fishermen (Caravajal et al. 2002; Gooi et al. 2007; Mendonça-Caridad et al. 2008), while only two putative unprovoked attacks have been reported up to now (Georgiadou et al. 2010; Galarza et al. 2016).

Intentional impalement as a defensive strategy was also proposed to be the reason for collisions between billfishes and sharks. Although swordfish’ bills are thought to have important hydrodynamic and feeding related functions (e.g. Habegger et al. 2015) and their breaking could be seriously disadvantageous for swordfish, it was reported that individual billfishes with damaged, malformed, or even missing rostra were apparently still in good physical shape (Frazier et al. 1994), leaving room for speculation if swordfish experience a trade-off between averting a threat (e.g. sharks) and the potential impairment of high-speed locomotion. The shark species involved in these incidents are usually either the blue shark Prionace glauca (Penadés-Suay et al. 2017, 2019) or the mako shark Isurus oxyrinchus (Cliff et al. 1990; Fierstine et al. 1997), both of which are known to opportunistically prey on fish, including swordfish (e.g. Stillwell and Kohler 1982; Maia et al. 2006; Vaske Júnior et al. 2009). Penadés-Suay et al. (2017), 2019) reported the stranding of several adult blue sharks in the western Mediterranean Sea, which had swordfish rostra embedded in their head capsules close to the eyes. The fact that all rostra came from juvenile swordfish, which are especially vulnerable to large pelagic sharks, and the angle of piercing that indicates nearly horizontal strikes against the sharks led the authors to propose intentional attacks, allegedly to fend off the blue sharks. Injuries inflicted by billfish rostra on eyes also were reported for the shortfin mako shark (Cliff et al. 1990) and for the bigeye thresher (Vacchi et al. 1999). In all other reported incidents, the rostra were found in close approximation to the axial skeleton (Starck 1960) or even pierced the vertebral column (Fierstine et al. 1997).

A deep penetrating trauma close to the vertebral column is also reported here for the examined thresher shark. In contrast to previously reported attacks on blue sharks, the swordfish involved in the current attack was not a juvenile, but roughly the same size as the thresher shark and, therefore, does not fit into its prey spectrum, which makes an attack as a defensive response unlikely. Both species have specialized morphological [i.e. elongated rostrum and whip-like upper tail lobe, respectively, for slashing or stunning their prey (Oliver et al. 2013; Habegger et al. 2015)] and physiological features [i.e. regional endothermy (Weng and Block 2004; Fritsches et al. 2005)] that enable both to occupy the same ecological niche in upper trophic levels, preying on teleosts, cephalopods, and crustaceans (Romeo et al. 2009; Ebert et al. 2013; Li et al. 2016). The cause of the injury in the thresher shark reported here cannot be definitively identified, beyond the fact that the thresher shark was impaled by the rostrum of the swordfish. This fatal injury might have been the result of an accidental collision, e.g. while feeding on the same prey resource. However, the potential for the impalement to be a result from a directed attack against the shark to drive a competitor away from this resource cannot be excluded either.

We hypothesize based on the results presented here (similar size and ecology of both species, direction of trauma infliction) that competition could be a driving force for swordfish attacks on sharks or other possible competitors. However, due to the fact that the protagonists are pelagic species and lethal interactions may usually remain undetected, more data is needed to unambiguously support this interpretation and we therefore encourage future studies on stranded sharks to conduct necroscopies and thoroughly search for evidence of swordfish interactions.