Our results show that probability of cookiecutter bites on resident pilot whales varies by season, lunar phase, and water temperature. In this study we obtain additional understanding of cookiecutter shark foraging ecology (and their interaction with a marine mammal prey species) by using bite wounds on free-swimming pilot whales. Unlike previous studies, which relied on dead pelagic fish brought to port, by observing free-swimming pilot whales we know the approximate geographic area where bites took place (Papastamatiou et al. 2010). We estimate that fresh bites observed on pilot whales likely occurred within one week of the individual being photographed, and therefore interactions almost certainly took place waters off the west side of Hawaiʻi Island. By concentrating on a resident cetacean population, we could determine how the probability of fresh bites varied based on both abiotic conditions and season.
Stable isotopes analyses of the livers and muscles tissues of cookiecutter sharks suggest that cookiecutter sharks exhibit seasonal shifts in diet, or in the location where they forage (Carlisle et al. 2021). We observed a general seasonal occurrence in fresh bites. While there was an overall increase in bite probability in April, the number of bites were not consistent, showing zero fresh bites in September and lower numbers of bites in May and August (Table 2). Cookiecutter shark bite probability peaked in October (Julian day 290) but was lowest during the winter months (Table 2, Fig. 3a). There were troughs between the peaks where virtually no pilot whales were observed. Why these reductions in pilot whale sightings occurred is unclear, but it was impossible to measure bite probability during these periods. The probability of fresh cookiecutter bites on bigeye tuna (Thunnus obesus), the majority of which were caught within the Hawaiʻi EEZ, similarly peaks from October-December (Papastamatiou et al. 2010). Swordfish (Xiphias gladius) are primarily caught outside the Hawaiian EEZ and show a peak in fresh bites from March-May (Papastamatiou et al. 2010). Combined these results suggest either that cookiecutter sharks show seasonal shifts in their diet or display seasonal migrations moving away from the main Hawaiian Islands in the winter and returning in the late spring.
The diet of cookiecutter sharks in Hawaiʻi includes squid (Papastamatiou et al. 2010; Carlisle et al. 2021), which is also prey for pilot whales, suggesting the potential for competitive interactions between sharks and pilot whales (Seagars and Henderson 1985; Sinclair 1992). A high biomass of vertical diel migrating micronekton, including squid, occurs off the west coast of Hawaiʻi Island, and the distance from shore also changes in relation to day and lunar phase (Benoit-Bird et al. 2001). Pilot whales in Hawaiʻi dive deeper during the day (mean = 666 m) than at night (mean = 415 m), although they perform more dives at night, and this variation may increase the rate at which pilot whales are encountering cookiecutter sharks as they move into shallower waters at night following the vertical diel migrating micronekton. Our data shows an increase in shark bite probability peaking in the transition from crescent to quarter moon (Fig. 3); this may either be a product of overlapping use of the vertical water column, or, as pilot whales in Hawaiʻi have been observed to move farther offshore as lunar illumination increases, an indication of movement into an area occupied by more cookiecutter sharks (Owen et al. 2019).
Pilot whale social structure may influence the dynamics of interaction between the two study species, and model predictions had a high standard deviation, suggesting that there is substantial variation in the effect across social cluster. Pilot whales in Hawaiʻi are known to have a social structures characterized by strong, long-term associations, suggesting individuals from the same social groups are exposed to cookiecutter sharks at similar rates and should may therefore have similar numbers of bites (Alves et al. 2013; Mahaffy et al. 2015). This may suggest cookiecutter sharks show heterogeneity in their distribution and that pilot whales within a group are more likely to get bitten when they move through a high-density cookiecutter shark area.
Sea surface temperature showed a significant effect on the probability of pilot whales being bitten, with bite probability decreasing as SSTs increased. Based on trawl data from the north Pacific, cookiecutter sharks were caught at temperatures ranging from 18oC to 26oC (Nakano and Tabuchi 1990). We recorded no fresh bites on pilot whales in September, the month with the highest average temperatures (27 ± 0.35 oC), despite 125 individuals being photographed from nine different sightings. Water temperatures off Hawaiʻi are relatively stable but these results further suggest that cookiecutter sharks may avoid surface waters > 26–27ºC, although this assumes that sharks are biting pilot whales at the surface. The probability of fresh bites also changes with lunar illumination, with a peak occurring between crescent and gibbous moon phases (Fig. 3c). The increases in bite prevalence on pilot whales during these times may be attributed to changes in the lunar movement of micronekton or the pilot whales themselves. Micronekton in the mesopelagic boundary community increase their depth and move farther offshore during periods of high lunar illumination (Benoit-Bird et al. 2009; Abecassis et al. 2015; Prihartato et al. 2016; Comfort et al. 2017). Increased surface illumination may also impact cookiecutter shark ability to hunt in surface waters or attract prey (Widder 1998). Finally, pilot whales dive shallower and for shorter durations during the Quarter and Crescent lunar phases, which may cause them to spend more time in the near-surface waters where they may be more susceptible to cookiecutter shark bites (Owen et al. 2019). We would predict that bites are more likely to occur at the surface due to the amount of time cookiecutter sharks and pilot whales spend there relative to deeper depths (as pilot whales are only diving below the surface for relatively short periods of time).
Cookiecutter sharks appear to show some selection for biting the head and dorsal areas of pilot whales, while avoiding or randomly biting the lateral and peduncle areas. However, in studies involving both odontocetes and mysticetes there were higher numbers of cookiecutter bites on the peduncle of mysticetes in comparison to odontocetes (Best and Photopoulou 2016). In other species, such as rough-toothed dolphins (Steno bredanensis), examination of the ventral side was facilitated through their aerial behaviour and this area was often covered in cookiecutter scars (Baird 2016). Unfortunately, pilot whales rarely leap out of the water and examination of the ventral side was therefore not possible, although some records of underwater sightings were obtained. Hence, our estimates of bite probabilities on short-finned pilot whales are conservative.
We provide new insight into the foraging dynamics of cookiecutter sharks, an incredibly versatile pelagic predator whose bites are ubiquitous on pelagic predators in tropical waters. By using bites on free-swimming pilot whales, we can approximate the geographic location where bites occurred, unlike previous studies with pelagic fishes which could only approximate bite location to the Hawaiian Islands (Papastamatiou et al. 2010). While cookiecutter shark bites on pilot whales may not be fatal (although some small dolphins may die from bites that penetrate into the abdominal cavity, Baird 2016), they may still reduce pilot whale fitness. Due to their potential high abundance, and ability to play an important ecological and economic role (e.g., cookiecutter shark bites reduce the price of market fish at auction, Papastamatiou et al. 2010), additional studies of cookiecutter shark foraging dynamics are warranted.