Whale killers: Prevalence and ecological implications of killer whale predation on humpback whale calves off Western Australia

Humpback/Killer Whale Interactions

We compiled a total of 35 separate interactions between killer whales and humpback whales that occurred along the coast of central WA, including 20 (57%) witnessed by one or more of the authors and 15 (43%) from other sources (Fig. 1). These comprised 24 (69%) attacks, 5 (14%) approaches (i.e., the humpback[s] responded to the presence of killer whales, but there was no evident attack), and 6 (17%) interactions of indeterminate nature (i.e., we were unable to determine if an approach resulted in an attack; Table 1 and Supplemental material). Of the 24 attacks, 14 (58%) resulted in kills, 8 (33%) were unsuccessful and 2 (8%) had unknown outcomes (Table 1). Overall, when attack outcomes were known (n = 22), 64% resulted in kills and 36% were unsuccessful.

When killer whales attacked humpbacks, they targeted only calves; there were no observations of attacks on adults or smaller animals traveling alone (presumably yearlings). When killer whales approached humpbacks and no calf was present (n = 6), the killer whales left almost immediately and did not attack; recorded interaction durations in these instances included 1–2 min (event #12), <1 min (# 21), <1 min (#22), “within a few minutes” (#24), and <1 min (#25). All of the 10 calves that we saw attacked in 2013 were neonates born during the northbound migration: all were very small (ca. 5 m; Fig. 3-6), pale gray, and at least two still had folded-over dorsal fins.

Killer whales appeared to target the throat and lower jaw of calves during at least four of the attacks (#4, 11, 19, 30), as evidenced by broken or dislocated jaws (Fig. 3). It was not clear if jaw injuries were the result of killer whales deliberately ramming the throat in an attempt to kill or injure the calf, or if the killer whales were attempting to feed on the tongue and lips (Fig. 6; see also Discussion).

The median group size for killer whales that attacked humpbacks (n = 17) was 6 (range 3–11). Median group size for successful attacks (6, range 3–8, n = 12) was not significantly different compared to unsuccessful attacks (8, range 5–11, n = 5; Wilcoxon-Mann-Whitney rank sum test, P = 0.09).

Although adult male killer whales are substantially larger than females (Barrett-Lennard and Heise 2006, Fearnbach et al. 2011), their presence did not appear to affect the outcome of attacks. Of 17 known-outcome attacks where exact group composition was known, a single adult male was present during 2 of 5 (40%) unsuccessful attacks and only 1 of 12 (8%) successful attacks. Killer whale groups without adult males were also successful in attacking humpback cow/calf pairs even when they were accompanied and defended by an “escort” (defined as one or more humpback adults accompanying a humpback mother and calf; see below). Of the 11 successful attacks by groups without an adult male killer whale present, at least 6 (46%) were on humpbacks with escorts present (#4, 14, 16, 19, 27, 30).

We observed two specific humpback defensive behaviors in response to killer whale attacks. The first was movement toward shallow water, reef edges, or nearby boats, which appeared to be an effective deterrence on at least five occasions. One attack ended after 70 min when the killer whales left after the mother moved her calf through a gap in the reef and onto the flats (#26). Another attack ended unsuccessfully after only “several minutes” when a mother, calf, and escort moved into shallow waters near the reef edge (#33). Event #11 also ended abruptly after a humpback mother led her calf into shallow reef waters, even though the calf appeared to be mortally wounded (Fig. 3). Another unsuccessful attack lasted 20 min, during which time the mother and calf humpback used a nearby boat for cover (#3). Finally, in event #20, the killer whales broke off an attack just as the mother, calf, and escort passed within a few m of our boat, although the aggressive defense of the escort may also have been a factor (see below).

The second humpback defensive behavior was direct interference or aggression towards killer whales, by the mother and/or an accompanying escort. When killer whales approached a humpback mother and calf, the mother typically moved her calf close by her side (#1, 4, 20), and sometimes lifted the calf out of the water using her head or upper back (#9, 11, 20; Figs. 4, 5). Other aggressive behaviors by the mother included loud and forceful exhalations (“trumpeting”; #8, 19, 27), lunging or charging at the surface (#8, 16, 19, 27), and slashing or slapping of flukes and flippers (#19).

Interestingly, escorts also vigorously defended the calf. During 6 of the 12 attacks when escorts were present, they were described as actively defending the calf (#1, 16, 19, 20, 27, 30); during 3 other attacks, the behavior of the escort was unrecorded (#14, 31, 33), and during 1 (#4), the escort apparently left as the killer whales approached the mother and calf. The defensive behaviors of the escorts were similar to that of the mothers, including flanking the calf opposite the mother (#20), flailing its flukes or flippers (#1, 19, 20, 27), charging toward the attacking killer whales (#1, 16, 19, 27), and trumpeting (#19, 20, 27).

Despite the efforts of the humpback mothers and escorts, killer whales were usually successful in their attacks, but less so when escorts were present. Of the 21 attacks where the outcome was recorded and the presence/absence of escorts noted, escorts were present during 10 (48%) and not present during 11 (52%; Table 1). Overall, calf kills recorded during 14 of the 21 (67%) attacks included 6 attacks (43%) when an escort was present and 8 attacks (57%) when there was no escort. Escorts may have been instrumental in driving off killer whales during at least five attacks (#1, 20, 31, 33, 34), although the calf was killed during at least six attacks when an escort was present (#4, 14, 16, 19, 27, 30).

Other Prey

The only prey other than humpbacks that we observed killer whales taking was spinner dolphins (Stenella longirostris), on one occasion. On 6 August 2013, we followed a group of 7–8 killer whales (including the tagged female, see below) as it moved slowly northward, 1 km out from the main channel at Coral Bay. At 1205, after encountering a small school of spinner dolphins, the killer whales split up and after some spirited chase lasting only a few minutes, they killed and ate at least two of the dolphins. By 1215 the group was heading northeast again and angling in towards the reef edge. Two hours later, the same group killed and ate a humpback calf (#30).

During our 2013 surveys, we encountered schools of spinner dolphins on four survey days. The schools typically numbered in the hundreds of individuals, with one group estimated to be on the order of 1,000 individuals; they were milling in waters 50–100 m deep during the daytime and appeared to be resting. This was by far the most numerous cetacean species that we encountered on the continental shelf adjacent to Ningaloo Reef and was potentially a readily available prey source for killer whales.

Satellite Tracking

On 20 July 2013, we deployed a satellite tag on an adult female killer whale (WA14) just seaward of Ningaloo Reef; her group killed a humpback calf just as we were approaching them (#19). The most detailed subset of our killer whale/humpback interaction data (included in the behavioral summary section above) came from the recorded movements and behavioral observations of this female and her associates, during the time her tag was operational.

The tag transmitted for 22 consecutive days, providing a total of 452 location estimates with a mean error radius of 1,558 m (minimum = 168 m, maximum = 31,060 m). Over this period, the tagged whale (and presumably her group) traveled an estimated 1,964 km, at an average speed of 4 km/h (maximum 12 km/h). Estimated displacement velocities identified two regions along the coast where the group slowed its movement and spent significant time (Fig. 2). On 21 July, the day after being tagged, the whale left our survey area and traveled 400 km southward to an area north of Shark Bay, where she spent 3 d close to shore near Carnarvon, WA (Fig. 2). By 29 July, the group was heading north again, and was back near Tantabiddi, where we tagged her, by 1 August; they remained in that area until at least 10 August when the tag quit transmitting.

While the tag was transmitting, we located the tagged female and associated whales at sea on 6 separate days (20 July; 1, 2, 3, 6, and 8 August, 2013) and followed them for a total of 20.3 h. During these focal follows, the group had 12 interactions with humpbacks (#23, 24, 26–35) and attacked 8 different calves: 4 attacks were unsuccessful, 3 resulted in kills, and 1 had an uncertain outcome—a 43% success rate for attacks with known outcomes.

Shark Interactions

Despite the fact that most observers usually had limited underwater viewing opportunities, large numbers of sharks were reported among killer whales that were attacking humpbacks, especially after kills. For example, on one occasion killer whales with a freshly killed humpback calf were attended by approximately 30 sharks (Carcharhinus spp.), estimated to be between 1.5 m and 2.5 m in length, and a 3 m tiger shark (Galeocerdo cuvier); as one of the killer whales was swimming around with the humpback calf in its mouth, one shark was filmed tearing flesh off the carcass (#16). Reported numbers of sharks at other kill sites included an estimated 12 (#15), an unspecified number present (#19), 8–12 (#27), and “numerous” sharks (#35).

In addition to scavenging, a white shark (Carcharodon carcharias) reportedly attempted to attack a humpback calf. On 24 July 2013, we interviewed Corey Hann who earlier that day had witnessed an apparent attack by a white shark on a humpback whale calf off Coral Bay, WA. He and a colleague were underwater and filming three humpbacks (mother, calf, and escort), when a 4–5 m white shark approached the whales. The divers immediately got out of the water but continued filming above water, from their boat. The escort began spyhopping when the shark approached from directly behind the whales. The mother humpback, apparently recognizing the danger, lifted the calf out of the water on her head. A few minutes later the calf breached several times. The escort remained behind the mother/calf pair, and began slapping its flukes vigorously when the shark approached, while the mother moved away from the calf. The shark disappeared and all three whales joined up again shortly afterward. According to Hann, the escort was instrumental in driving off the white shark.

Humpback Behavioral Responses to Killer Whale Predation

Predators can profoundly influence the behavior of their prey, even if they rarely kill them (Ripple and Beschta 2005, Wirsing et al. 2008). For example, Corkeron and Connor (1999, Connor and Corkeron 2001) suggested that the threat of killer whale predation alone could provide the impetus for large whales to migrate to low latitudes for calving. Clapham (2001) argued that this was not likely because predation on large whale calves had rarely been documented, but that could also be cited as further evidence that migration is an effective antipredator strategy.

It appears that killer whales now occur annually along coastal areas of Western Australia to prey on a flush of seasonally available humpback whale calves—an abundant, predictable, easily taken prey source that could represent a significant contribution to their annual energy intake and perhaps a growing threat to breeding humpbacks. We do not have enough data yet to offer more than a crude estimate of number of calves taken annually, but the group with the tagged female that we followed took three humpbacks during approximately 1 d of foraging (i.e., 20.3 h over 6 d). If this is extrapolated over the larger killer whale population that occurs off Ningaloo, for the 5 mo of humpback whale presence off the coast there, it seems likely that as a minimum, dozens of calves are taken there annually.

Stone et al. (1987) pointed out that humpbacks in the Northern Hemisphere migrate mainly through open oceans, while those in the Southern Hemisphere tend to travel along continental coastlines. Following coastlines in areas like in WA should make the timing and routes of humpback migration even more predictable to predators and increase their vulnerability. Humpbacks produce, on average, only one calf every 2 yr (Clapham 2000), so that each one is a substantial reproductive investment. Killer whales took those calves off the coast of WA with relatively little effort, and not surprisingly, humpbacks appear to have developed behavioral responses to counter that threat. We discuss two of those below.

Comparisons with Gray Whales

Gray whales feed in the Bering and Chukchi Seas during the northern summer; then during the winter they undertake a round-trip journey of up to 20,000 km, to the lagoons of Baja California, Mexico, where they calve and mate (Rice and Wolman 1971). Gray whale calves migrating with their mothers to their northern feeding grounds for the first time are subject to intense killer whale predation pressure. For example, during May and June each year from 2003 to 2006, over 170 individual mammal-eating killer whales have been documented aggregating near Unimak Pass, in the eastern Aleutian Islands, Alaska, where they preyed upon northbound gray whale calves (Barrett-Lennard et al. 2011, Matkin and Durban 2013). This is the largest assemblage of mammal-eating killer whales recorded anywhere in the world and is an important source of gray whale calf mortality (Matkin and Durban 2013).

Barrett-Lennard et al. (2011) summarized their observations of killer whale attacks on gray whale calves in the eastern Aleutians and cited seven “key observations.” Our observations at Ningaloo reveal some striking similarities and important differences in the way that gray whales and humpbacks respond to the threat of killer whale attack on their calves. Below we list and address each of the key points presented by Barrett-Lennard et al. (2011):

Scavenger Hunt

Scavenging is an often-underrated energy pathway in food webs (Wilson and Wolkovich 2011). During our survey, not only were humpback calves regular prey for killer whales off WA, but their carcasses appeared to be a major food source for scavengers, especially sharks, which readily feed on whale carrion (Dudley et al. 2000, Curtis et al. 2006, Dicken 2008, Fallows et al. 2013). In Alaska, Barrett-Lennard et al. (2011) suggested that gray whale calf remains left by killer whales were potentially an important food source for sleeper sharks (Somniosus pacificus), and brown bears (Ursus arctos), which fed on carcasses that washed ashore. Red foxes (Vulpes vulpes) and various species of scavenging birds also benefitted.

The shark population of Ningaloo Reef is large and diverse. Carcharinid sharks are the most abundant and diverse forms, but tiger sharks and hammerheads (Sphyrna spp.) are also common, and white sharks occur regularly (Last and Stevens 1994; authors’ personal observations). During and after at least 5 of the 14 (36%) humpback calf kills that we documented off WA, numerous large (2–4+ m) sharks gathered within minutes at the attack site to feed on the carcass (#15, 16, 19, 27, 35).

Another important factor for scavengers off WA is the consumption efficiency of feeding killer whales and the possibility of “surplus killing” (Kruuk 1972, Jefferson et al. 1991). Although surplus killing typically refers to killing more individual potential prey animals than a predator will actually eat (e.g., Stacey et al. 1990, Williams et al. 1990, Gaydos et al. 2005, Barrett-Lennard and Heise 2006), it can refer to any kind of wasteful killing. For example, when prey is abundant, killer whales tend to selectively feed on certain body parts, e.g., the tongues, lips, and ventral grooves of baleen whales or just the breast muscles of penguins (Lowry et al. 1987; Flood 2000; Whitehead and Reeves 2005; Pitman and Durban 2010, 2012; Ferguson et al. 2012), and discard the rest of the carcass, leaving more for the scavengers. Evidence that Ningaloo killer whales may at times selectively feed on the tongue and throat of humpback calves is shown in Figure 6 and in a report of a freshly-dead humpback calf found on the beach near Yardie Creek on 17 September 2013, which was missing its lower jaw, throat, and some of its underside,4 presumably from a successful killer whale attack. A 10 m humpback whale found on the beach at Warnbro Sound, just south of Perth, WA, in June 2008, also had its tongue missing and damage to its lower jaw and throat.5

Killer whales in Alaska often returned to gray whale calf kill sites to feed on submerged carcasses for up to at least 5 d, a behavior that Barrett-Lennard et al. (2011) referred to as “prey caching.” Off WA, however, prey caching is unlikely because a carcass there would be consumed (or carried off) very quickly due to the large number of large sharks present and the fact that WA humpback calves are much younger and smaller than gray whale calves taken off Alaska. As an indication of this scavenger efficiency off central WA, Chittleborough (2003: fig. 1) showed a photograph of a 12.8 m humpback that had been harpooned off Pt. Cloates, Australia, in 1952; by the time the carcass was pulled ashore 4 h later, little more than a skeleton remained. More recently, on 20 May 2014, a live humpback, estimated to be 8–9 m in length (possibly a yearling), was seen being attacked by “dozens” of sharks off Coral Bay, WA. Identified sharks included “mostly tigers, though also bull [Carcharhinus leucas] and the ever-present whalers.” Less than 60 h later, this shallow “whale fall” (Smith 2006) was little more than a skeleton on the ocean floor (Fig. 7).

Fallows et al. (2013) recently suggested that scavenging on dead large whales could be a critical, if underappreciated, aspect of white shark foraging ecology, especially for mature sharks. Although they considered a whale carcass to be a temporary “pulse” resource, a prolonged, seasonal occurrence of large numbers of humpback calf carcasses provided by killer whales along the coast of WA each year could offer a more predictable food source, with possible implications for the abundance and movements of several species of sharks, including whites (Jorgensen et al. 2010).

Whaling–Impact and Recovery

Humpback whale numbers have increased rapidly off WA since whaling ended there in 1963. Bannister (1964) estimated that the population had been reduced to 568 individuals by the end of 1963, and this did not even take into account the illegal whaling that continued on this population in Antarctica until at least 1968 (Yablokov 1994, Clapham et al. 2009). By 2008, however, population estimates for WA humpbacks ranged from 21,750 (95% CI = 17,550–43,000, Hedley et al. 2011) to 26,100 (95% CI = 20,152–33,272, Salgado Kent et al. 2012). Subsequently, Branch (2011) estimated that the population was probably >30,000, and Salgado Kent et al. (2012) further suggested that it could was as high as 33,300. More recently, the IWC (2014) estimated that the population in 2012 was 19,264.

The preexploitation population size of WA humpbacks was estimated by Chittleborough (1965) to be 12,000–17,000, although he was also unaware of the huge illegal Soviet catches. That estimate was revised to 21,686 by the IWC (2014), which when compared with the population estimate of 19,264 cited above would indicate that the WA humpback population could now be at or above 90% of the estimated carrying capacity (probability interval: 74%–98%).

Regardless of the exact figures, WA is now home to the largest humpback breeding population known, having increased annually at an estimated 10.15% (Bannister and Hedley 2001) or 13% (Salgado Kent et al. 2012), which are just below and above, respectively, the maximum plausible rate posited for this species (i.e., 11.8%, Zerbini et al. 2010). Clearly, killer whale predation has not been a significant factor in the recovery and survival of WA humpbacks—at least until recently.

If there had been a sizeable population of killer whales that preyed extensively on WA humpback whale calves prior to commercial whaling, then it is possible or even likely that when the humpback population collapsed in the 1960s, the local killer whale population could have collapsed along with it. In fact, there is some evidence that, historically, the killer whale population off WA was larger than it is today, that it declined after the whaling era, and that it might be increasing today. Chittleborough (1953) reported five sightings of killer whales during aerial surveys for humpbacks off Pt. Cloates, WA, conducted during just two months (August/September) in 1952; his group size estimate for three of the sightings was >30, ca. 40, and ca. 50–60 (the other two were 5 and “several”). Also in September 1952, the pilot for the same spotter plane reported a sighting of “at least 150” killer whales in Exmouth Gulf. By contrast, the largest killer whale group size in our survey of recent records was 11 (Table 1).

After the whaling ended and up until fairly recently, there had been very few reported sightings of killer whales off Ningaloo (Table 1). For example, Sleeman et al. (2007) conducted a series of systematic aerial surveys for whales and dolphins off Ninglaoo Reef (as far south as Pt. Cloates), during every month except December, from June 2000 to April 2002, and they did not record a single sighting of killer whales. The first record of an attack on a humpback calf since the end of the whaling era that we are aware of was in 2006. Craig Kitson, captain of an Exmouth whale shark tour boat, told us: “I moved to Exmouth in 2001 and the first time I saw an attempted kill by the orca pod was on the 10th of August 2006 in Lighthouse Bay [#3]. Up until this time, orcas were not seen that regularly. We did see the odd one occasionally, however.”

One plausible interpretation of these events is that after whaling depleted WA humpbacks, killer whales that preyed on them either declined in numbers or were extirpated. Furthermore, any subsequent recovery of those killer whales would be expected to lag far behind the humpbacks. For example, the maximum population growth rate for killer whales is considerably lower than for humpbacks (3.5% vs. 11.8%, respectively; Matkin et al. 2014, Zerbini et al. 2010), and once killer whales resumed hunting humpbacks off WA, it would be many years, perhaps decades, before they recovered their former population size and realized their full predatory potential on the local humpbacks. This could also apply to populations of killer whales that prey on large whales elsewhere.

Killer Whales as Predators of Large Whales

The debate about whether killer whales were important predators of large whales prior to the advent of industrial whaling, and, if they were, how they responded to the worldwide depletion of whale stocks, could have important implications for marine food web dynamics. For example, Springer et al. (2003) argued that mammal-eating killer whales were important predators of large whales, and that after whale stocks were decimated in the North Pacific, killer whales targeted and serially depleted populations of other, increasingly less-profitable species of marine mammals—a trophic cascade that they referred to as the “sequential megafaunal collapse” hypothesis (SMC; Springer et al. 2003, 2008; Williams et al. 2004, Estes et al. 2009). Since it was published, the SMC has met with numerous criticisms about the assumptions and data used for the analyses (DeMaster et al. 2006; Trites et al. 2007; Wade et al. 2007, 2009), with some authors further contending that large whales are not now and probably never were important prey for killer whales (Mizroch and Rice 2006, Mehta et al. 2007, but see Reeves et al. 2006).

The seminal assumption of the SMC hypothesis—that large whales were important prey for at least some populations of killer whales—has received some support recently from evidence that, in addition to WA humpbacks, another recovered large whale population may also be experiencing increased predation pressure from killer whales. Commercial whaling had depleted eastern Pacific gray whales by 1939, but under international protection their numbers rebounded, and by 1994 they were removed from the US Endangered Species List (Jones and Swartz 2009, Durban et al. 2013). Although, historically, few attacks on gray whales were ever reported (Jefferson et al. 1991, Barrett-Lennard et al. 2011), it has recently been suggested that killer whales could now be taking as much as half of the gray whale calf production in some years (Matkin and Durban 2013).

Research on killer whales only really began after many large whale populations had already been pushed to the brink of extinction by the 1960s and early 1970s (Baird 2002, Roman et al. 2014). By then, any killer whales that had preyed extensively on large whale calves would already have been forced to either starve and die, or switch to other, presumably smaller prey (the SMC hypothesis; Estes et al. 2009). Either way, it is likely that those killer whale populations would have declined or perhaps even became extirpated in some areas. In Alaska, for example, one wonders what would happen to the 170+ mammal-eating killer whales that currently prey on gray whale calves during spring and summer months if those gray whales became depleted, as they were by 1939 (Jones and Swartz 2009).

It is now well established that when killer whales attack large whales they target the calves (Reeves et al. 2006, Pitman et al. 2007, Barrett-Lennard et al. 2011, Higdon et al. 2012, this study). Mehta et al. (2007) and Steiger et al. (2008) both concluded that when killer whales attacked humpbacks, because they targeted mainly calves and in lower latitudes, that this somehow invalidated the SMC. But Springer et al. (2003) never specifically excluded calves as prey in their hypothesis (Springer et al. 2008) or ruled out the possibility of predation in lower latitudes. Furthermore, recent satellite-tracking studies in the North Atlantic (Matthews et al. 2011), North Pacific (Durban et al., unpublished data), and Southern Ocean (Durban and Pitman 2012, unpublished data) have shown that mammal-eating killer whales from high latitudes are capable of rapid, long distance movements (1,000s of km) into mid-latitudes, which means they could be capable of intercepting whales during migration when the calves are smaller and more vulnerable. Perhaps the best test of the Springer et al. (2003) hypothesis will come from monitoring killer whale responses to recovering large whale populations (Kareiva et al. 2006, Wade et al. 2007), an experiment that may already be unfolding off Western Australia, in the eastern Canadian Arctic, in the eastern Aleutian Islands of Alaska, and possibly elsewhere.

Migration and Predation—the Moveable Feast

Large whales are important components of marine ecosystems not only as in situ mega-consumers (Laws 1985, Kareiva et al. 2006, Croll et al. 2006), but also for the role they play in packaging and redistributing nutrients in the ocean (Roman and McCarthy 2010), which their extensive migrations can extend to an ocean-basin scale (Roman et al. 2014). We suggest that further study of the continental shelf waters along WA will show that the vast majority of the cetacean biomass that occurs there on an annual basis is represented by migrating humpback whales (Sleeman et al. 2007), which extract little and contribute much to the local ecosystem. Although they do essentially no feeding during migration (Chittleborough 1965, Dawbin 1966, Baraff et al. 1991), during almost 6 mo every year, 20,000–30,000 humpbacks may visit the nearshore waters of WA, where they slough skin, shed placentae, pass metabolic waste, and sometimes die; killer whales and white sharks prey on their calves, and scavenging sharks and benthic detritivores feed on their carcasses (Smith 2006). All of this largesse ultimately derives from the Southern Ocean where Antarctic krill (Euphausia superba) is shape-shifted into Area IV humpbacks and injected annually into the relatively oligotrophic, littoral-marine ecosystem of Western Australia, where killer whales their do their part in making it available to local consumers. The full impact of this Antarctic subsidy to the WA marine economy will not be realized, however, until that system has recovered, to the extent possible, from the ongoing legacy of commercial whaling.