INTRODUCTION
Anecdotal evidence and reported observations suggest that the presence and occurrence of killer whales (Orcinus orca) is increasing in the eastern Canadian Arctic (Higdon, Reference Higdon2007; Higdon et al., Reference Higdon, Hauser and Ferguson2012). Inuit hunters have expressed concern regarding effects on marine mammal prey populations, including a number of marine mammal species of socio-economic and cultural importance (Higdon, Reference Higdon2007; Ferguson et al., Reference Ferguson, Higdon and Westdal2012a, Reference Ferguson, Kingsley and Higdonb; Westdal et al., Reference Westdal, Higdon and Ferguson2013). There has been little directed scientific study on killer whales in the Canadian Arctic (Matthews et al., Reference Matthews, Luque, Petersen, Andrews and Ferguson2011; Young et al., Reference Young, Higdon and Ferguson2011) and the ecology of this population is poorly known. Killer whales may be capable of significant impacts on prey populations and ecosystem structure (Springer et al., Reference Springer, Estes, Van Vliet, Williams, Doak, Danner, Forney and Pfister2003, Reference Springer, Estes, Van Vliet, Williams, Doak, Danner and Pfister2008), and increased knowledge of Arctic killer whale distribution and ecology is needed for effective conservation and management. The Arctic region is undergoing considerable environmental and ecological changes (e.g. Serreze et al., Reference Serreze, Holland and Stroeve2007; Higdon & Ferguson, Reference Higdon, Ferguson, Ferguson, Loseto and Mallory2010), and information on changing distribution and/or abundance of killer whales may help improve understanding of ecosystem dynamics and changes.
Increased sighting frequency may be related to a growing population, range expansion with decreasing ice extent, increased effort and reporting, or some combination thereof (Higdon, Reference Higdon2007; Higdon & Ferguson, Reference Higdon and Ferguson2009; Higdon et al., Reference Higdon, Hauser and Ferguson2012). Whatever the reason(s), increased presence of marine mammal-eating killer whales is likely to influence the abundance, distribution and behaviour of Arctic cetaceans and pinnipeds (Ferguson et al., Reference Ferguson, Higdon and Westdal2012a, Reference Ferguson, Kingsley and Higdonb). In Nunavut, killer whales prey on a variety of species, including bowhead (Balaena mysticetus), narwhal (Monodon monoceros) and beluga (Delphinapterus leucas) whales, and ringed (Pusa hispida) and bearded (Erignathus barbatus) seals (Reeves & Mitchell, Reference Reeves and Mitchell1988; Higdon, Reference Higdon2007; Ferguson et al., Reference Ferguson, Higdon and Westdal2012a, Reference Ferguson, Kingsley and Higdonb; Higdon et al., Reference Higdon, Hauser and Ferguson2012; Matthews & Ferguson, Reference Matthews and Ferguson2013 (this volume)). All these species are important to Inuit subsistence and culture, and negative opinions towards killer whales often relate to perceived competition for resources (Westdal et al., Reference Westdal, Higdon and Ferguson2013).
Reported sightings peak during the ice-free season, and killer whales appear to be relatively common in some areas of the eastern Canadian Arctic (Reeves & Mitchell, Reference Reeves and Mitchell1988; Forney & Wade, Reference Forney, Wade, Estes, Brownell, DeMaster, Doak and Williams2006; Higdon et al., Reference Higdon, Hauser and Ferguson2012). There are no data on population size and growth or stock structure, however. In comparison to other Arctic cetaceans, killer whale population size is small and densities are low, making standard survey techniques (aerial and vessel-based) less effective methods for estimating population size and trends (Forney & Wade, Reference Forney, Wade, Estes, Brownell, DeMaster, Doak and Williams2006; Higdon, Reference Higdon2007). One valuable source of information is the traditional ecological knowledge (TEK) and observations of Inuit harvesters, which represents the cumulative body of knowledge a person gains, through observation, experience and cultural transmission, about their local environment and the ecological relationships of the species residing there (Berkes, Reference Berkes1999; Usher, Reference Usher2000).
There has been growing interest in the collection and use of TEK for co-management of Arctic resources and it is being increasingly used for wildlife management purposes (Wenzel, Reference Wenzel2004; Tester & Irniq, Reference Tester and Irniq2008). A combination of information derived from TEK and modern science techniques is useful as they can provide different but complementary types of information (Huntington et al., Reference Huntington, Suydam and Rosenberg2004; Gagnon & Berteaux, Reference Gagnon and Berteaux2009; Lewis et al., Reference Lewis, Hammill, Power, Doidge and Lesage2009; Westdal, Reference Westdal2009; Westdal et al., Reference Westdal, Richard, Orr, Ferguson, Loseto and Mallory2010). Information from TEK surveys can also provide resource managers with a long temporal and wide spatial record (Ferguson et al., Reference Ferguson, Williamson and Messier1998), and can be especially valuable in remote locations where scientific data are logistically difficult to collect. A number of studies have shown the value of Inuit observations in informing wildlife management and conservation (e.g. NWMB, 2000; Mallory et al., Reference Mallory, Fontaine, Akearok and Johnston2006; Lewis et al., Reference Lewis, Hammill, Power, Doidge and Lesage2009). Inuit harvesters are dedicated and reliable observers, and their observations can be particularly valuable for little known, rare and distinct species that are hard to survey using traditional scientific methods (Mallory et al., Reference Mallory, Gilchrist, Fontaine and Akearok2003, Reference Mallory, Fontaine, Akearok and Gilchrist2008). Inuit harvesters have considerable knowledge on killer whales (e.g. Ferguson et al., Reference Ferguson, Higdon and Westdal2012a), despite this species generally being a non-harvested species in Canada. Here, we summarize information on the spatiotemporal distribution and abundance of killer whales collected using semi-directed interviews with Inuit hunters and elders throughout Nunavut.
MATERIALS AND METHODS
We used semi-directed interviews to document information on killer whale distribution and ecology provided by Inuit hunters and elders throughout Nunavut (11 communities in the Qikiqtaaluk (Baffin Island, including Foxe Basin) and Kivalliq (western Hudson Bay area) regions between July 2007 and March 2010: Figure 1; Table 1). Semi-directed interviews allow for the collection of TEK in an open and flexible way that avoids the rigidity of questionnaires (Huntington, Reference Huntington1998, Reference Huntington2000). We developed a list of questions in advance but the interview remained open-ended, giving each interviewee the option to elaborate on matters that they considered important. Using this process, the interviewee does not always address every topic that was included in the original list of questions, but in many cases allowed the interviewee to provide information not anticipated by the researcher.
Fig. 1. Map showing Nunavut (darker grey) and surrounding area, major marine water bodies, locations of Nunavut communities (black circles: large circles with labels indicate communities interviewed, smaller circles are communities that were not included in the interview study), and killer whale sightings reported and mapped by interviewees (open circles, N = 189).
Table 1. Summary information for 105 semi-directed interviews conducted in 11 Nunavut communities, including region and demographic characteristics (sex and age (birth decade)) of interviewees.
Most interviewees were identified from lists of knowledgeable hunters provided by the local Hunters and Trappers Organization (HTO) (reputational sampling), augmented using snowball sampling (suggestions of potential interviewees from other interviewees or interpreters). The most knowledgeable subjects identified by the HTOs tended be the active and former male hunters of middle age and older (see Results). Interviews were conducted using paper copies of maps from the Nunavut Land Use and Occupancy Study (Freeman, Reference Freeman1976), at varying spatial scales. Multiple maps were used in each community to capture sightings from different locations where interviewees have lived or travelled (e.g. Foxe Basin interviewees who formerly lived in Pond Inlet or Arctic Bay, and vice versa). In each interview, the maps were used to document spatial information, including specific observations of killer whales, general locations where killer whales are seen, migration routes, the interviewee's hunting locations and travel routes, and the locations of outpost camps.
Interviews were conducted in each community with the aid of a local interpreter, and most were conducted in Inuktitut (some using a mix of English and Inuktitut or English only). As such, for many interviews, our analyses are based on the statements provided by the translator and not the actual interviewee. We sought clarification on statements whenever warranted, and made efforts to ensure that the translator understood both our questions and the interviewee responses, to the extent possible. Interviews often consisted of extensive back-and-forth dialogue between the interviewer, translator and interviewee. Additionally, we sought to hire interpreters with experience of doing this kind of work and known to have an extensive knowledge of Inuktitut terminology. There were times however when interpreters had difficulty understanding some of the terminology used by elder interviewees. We trust in the honesty and accuracy of the interpreters when it comes to their descriptions of interviewee responses regarding general patterns (e.g. seasonal occurrence, relative abundance), but acknowledge that some important nuances were likely lost in translation (see Freeman, Reference Freeman1976: p. 53). We informed all interviewees that the interviewer would be available at the local hotel on the last day of the community visit, to allow interviewees to review their transcripts if they desired. Additional details on the interview process and questions, research permitting and approvals, and reasons for choosing particular communities are described in Ferguson et al. (Reference Ferguson, Higdon and Westdal2012a) and Westdal et al. (Reference Westdal, Higdon and Ferguson2013).
Data summary and analyses
The type of information provided varied among interviewees given the open-ended format, and a mix of qualitative and quantitative survey data on killer whales was analysed. Qualitative data were analysed using an interpretive approach to categorize results, where interviewee statements were manually grouped into related categories (e.g. distribution, abundance, seasonality) and patterns then summarized (Kitchin & Tate, Reference Kitchin and Tate2000). Results were summarized within and between communities and four different regions (Table 1), and across eastern Nunavut as a whole. All individual sighting reports were added to a larger killer whale sightings database (Higdon, Reference Higdon2007), and observations from nine communities (pre-2010 interviews) were included in a recent analysis (Higdon et al., Reference Higdon, Hauser and Ferguson2012). In this study we report on information related to killer whale abundance and distribution in Nunavut waters. Inuit interviewees also provided extensive information on killer whale prey items and predation behaviour (Ferguson et al., Reference Ferguson, Higdon and Westdal2012a), and on human dimensions in killer whale management and conservation in Nunavut (Westdal et al., Reference Westdal, Higdon and Ferguson2013).
The information provided by each interviewee varied due to the semi-directed nature of the interview process, but all provided some indication of the number of killer whale sightings they had over the course of their lifetimes (also see Westdal et al., Reference Westdal, Higdon and Ferguson2013). This included a mix of quantitative (e.g. seen only once or twice, seen 5–10 times, N = 29) and qualitative (e.g. seen a few times, many times, regularly, annually, N = 75) descriptions. The number of lifetime sightings was given a categorical ranking, with interviewee responses ranked on a four-point scale: (1) None; (2) Little (one or two sightings in their lifetime); (3) Few (3–5 sightings, plus qualitative descriptions of ‘a few’ and ‘some’); and (4) Many (one interviewee with 5–10 sightings, plus all those noting ‘many’ sightings). For statistical analyses, the None (N = 3) and Few (N = 23) categories were combined. The categorical ranking was tested against region, age (birth decade) and harvester status (full-time or part-time harvester) using χ2 analyses of contingency tables. Community and regional results were summarized as an index of relative killer whale abundance. Interviewee age was divided into two categories for statistical analyses: those born in the 1910s–1940s and those born in the 1950s–1980s (i.e. evenly split into four decades per category, also see Westdal et al. (Reference Westdal, Higdon and Ferguson2013)).
The distribution of interviewee responses regarding various facets of killer whale abundance (sighting frequency, annual presence, population trends) by region was tested using χ2 goodness of fit tests, and χ2 tests of independence were employed to compare responses among different variables (e.g. dependence of population trend statements on past experience seeing killer whales). Fisher's exact tests were used in cases where χ2 was inappropriate due to expected values less than five (McDonald, Reference McDonald2009). The index of qualitative variation (IQV) (Gibbs & Poston, 1973; Wilcox, Reference Wilcox1973) was used to measure variability among responses, both within response categories and by region. The index is based on the ratio of the total number of differences in the distribution to the maximum number of possible differences within the same distribution. The IQV can vary from 0.00, when all cases are in the same category and there is no variation, to 1.00, when the cases are distributed evenly across the categories and there is maximum variation.
RESULTS
Interviewees
Of the 105 semi-directed interviews conducted in 11 communities (5–17 interviews per community, mean = 9.5), the majority of interviewees were male (91%, Table 1). We made an effort to interview older hunters to collect the longest temporal record possible. Information on age (year or birth decade) was provided by 89 interviewees, and the majority of these (72, 81%) were born in the 1950s or earlier (42 born in the 1940s or earlier, and the oldest interviewee was born in the 1910s) (Table 1). Younger hunters were interviewed on occasion, but only four interviewees were born in the 1970s or 1980s, and only one was less than 30 years old. The two age groups defined for analyses were evenly divided (42 born in the 1910s–1940s, 47 born in the 1950s–1980s).
Harvester status was provided by 92 interviewees, with varying levels of detail. Some interviewees identified themselves as active full-time (N = 18) or active part-time (N = 10) harvesters, and another 13 stated that they were active part-time harvesters but were formerly full-time harvesters with reduced effort due to age, health or work. Nineteen elders were no longer active, due to age and health, but were full-time harvesters for most of their lives (or in the case of female interviewees were married to full-time harvesters). The remaining 32 interviewees simply noted that they were active hunters, without stating if they were full-time or part-time. The data were divided into two categories for statistical analysis–full-time and part-time. Full-time included those who identified themselves as active full-time, active part-time but formerly full-time, and no longer active but formerly full-time (N = 50) and part-time included the ten who identified themselves as part-time hunters plus the 32 who simply identified themselves as ‘active’ and were assumed to be part-time based on our experience in Nunavut communities. The majority of these interviewees were middle-aged or younger (in their 30s to 50s), and there are few full-time harvesters in this age bracket as many people have wage employment for at least part of the year (there are some younger full-time hunters, and some interviewees identified themselves as such) (also see Westdal et al., Reference Westdal, Higdon and Ferguson2013).
Killer whale presence in Nunavut
Nearly all participants (97%) had seen killer whales at least once, although not necessarily near their current community. Killer whales are widely distributed in Nunavut and have been observed in the general vicinity of all 11 communities, often in the same general areas and typically close to the community where the majority of harvesting activity is now focused (e.g. Priest & Usher, Reference Priest and Usher2004) (189 individual observations were compiled from the interviews, mapped in Figure 1 to indicate the areas where whales are generally observed). However, the magnitude of their presence varies, and interviewees in some communities report seeing killer whales more regularly than others. Most interviewees (N = 78, 74%) provided an opinion on whether or not killer whales were present annually in their local area. In all four regions, the majority of interviewees reported that killer whales were present on an annual basis (Figure 2A). Patterns for yes/no responses did not vary by region (Fisher's exact test, df = 3, N = 78, P = 0.313), although analyses are confounded by non-responses, particularly in the Hudson Bay region. A small number of interviewees in all regions thought killer whales were not present every year, particularly in Foxe Basin and South Baffin.
Fig. 2. Summary of interviewee responses by region (Table 1) to questions on killer whale occurrence: (A) whether killer whales are observed annually; (B) sighting frequency of individual interviewees; (C) observed trends in sighting frequency or population size.
Seasonal distribution
Killer whales are generally observed during the ice-free summer season, and interviewees in all communities noted that occurrence in a particular summer depends on ice conditions. Most interviewees provided information on the month(s) or season(s) when killer whales are usually seen. Over half the interviewees in each region reported that killer whales were usually seen in August (Figure 3), with July and September reported about equally often. Reports of killer whale presence in May occurred only in the South Baffin region, and this region also had the highest percentage of reports indicating June. Interviewees in Kimmirut noted that killer whales are usually observed migrating past in both spring and autumn.
Fig. 3. Summary of interviewee responses by region (Table 1) on the months when killer whales are observed.
Twenty-one interviewees discussed killer whale arrival in the spring and/or summer, and most noted that arrival depends on ice conditions. Half (N = 11) reported that killer whales arrive once the ice has cleared, while eight indicated that they are present after break-up starts, when there is still some loose ice present. One Repulse Bay interviewee noted that they usually arrive after the ice is gone, but in some years arrive when some ice is still present. One from Igloolik indicated that killer whales do not come to northern Foxe Basin in heavy ice years. Springtime killer whale occurrence, or lack thereof, at the floe edge (edge of the landfast ice) was mentioned by 28 interviewees. Most (N = 20) indicated that killer whales are not seen at the floe edge, as far as they were aware. Interviewees had never seen them there, despite considerable time spent hunting at the ice edge. Eight interviewees did note floe edge sightings, although only two were direct observations (once seen during springtime by an Igloolik hunter, and one from Pond Inlet noted that they were sometimes seen in June and July at the floe edge). The others were second-hand reports: two Pangnirtung interviewees noted that other hunters had seen three at the floe edge in April 2009 (also reported independently to DFO; S. Sowdlooapik, Government of Nunavut, personal communication, April 2009), and eight were reportedly observed at the Igloolik floe edge in June 1999 (a year when killer whales were reportedly more numerous than usual, Cosens & Blouw, Reference Cosens and Blouw2003). One Igloolik interviewee stated that killer whales were present in July at the floe edge, and another noted that they had been seen going by the floe edge in Repulse Bay in June. A Repulse Bay interviewee had also heard of floe edge sightings in the spring. Overall, however, there were was very little evidence of killer whales being observed early in the season, before most of the ice is gone.
Six interviewees (North Baffin, Hudson Bay and South Baffin regions) noted that killer whale depart again before freeze-up, soon after the ice starts to form. Eleven interviewees stated that killer whales are never seen during the winter, and another 11 suggested that they went south to the open ocean where there is no ice. Five interviewees indicated that they did not know where killer whales went during winter. One from Arviat suggested that they might overwinter in the area between Greenland and Baffin Island (Davis Strait, Baffin Bay). An interviewee in Qikiqtarjuaq reported regular winter sightings while working on fishing trawlers off the coast of Labrador. Some groups would be small, and others would be larger groups of up to ~20 whales.
Relative abundance of killer whales throughout Nunavut
SIGHTING FREQUENCY OF INTERVIEWEES
The categorical index of killer whale sighting frequency was available for all 105 interviewees. Only one-quarter of interviewees had none or few lifetime killer whales sightings (N = 26), and a third had many sightings (N = 31). Sighting frequency patterns varied across region (χ2 = 24.412, df = 6, N = 105, P = 0.0004) (Figure 2B). This was most pronounced for interviewees with many killer whale sightings, as they were predominantly in the North Baffin region (63% of the total, IQV for region = 0.737). Sighting frequency categories were more variably distributed in the other regions (IQV range 0.828–0.992). The distribution of interviewees with some killer whale sightings was evenly distributed across all four regions, and the distribution of those with few or no sightings was somewhat uneven, with few in the North Baffin region. The proportion of interviewees with many, and many or some (i.e. both categories combined) killer whale sightings is mapped in Figure 4 as a measure of relative killer whale abundance in eastern Nunavut waters.
Fig. 4. Map showing frequency of killer whale sightings among interviewees in the 11 Nunavut communities visited and Inuit observations and knowledge of killer whale movements and suspected migration routes. Size of community symbols indicates the proportion of interviewees that had ‘many’ killer whale sightings (black inner circles) and the proportion with ‘many’ or ‘some’ combined (open outer circles). Arrows summarize killer whale movement patterns, with thickness approximately proportional to the number of interviewees identifying each area as a movement or migration corridor.
Sighting frequency also varied by community, with nearly all Arctic Bay residents (9 out of 11, 88%) having many killer whale sightings, and the other two North Baffin communities having half their interviewees with many sightings. Eight of the Arctic Bay interviewees reported that they see killer whales every year, and four from Pond Inlet reported seeing killer whales frequently or regularly (e.g. almost every year). Three interviewees also reported that they do not see them often, or only once in a while, however. In Qikiqtarjuaq, two interviewees reported seeing whales every year, two more reported seeing them regularly, whereas others (N = 2) reported either rarely seeing them or having only one lifetime sighting. Further south, over half (6 out of 11) the Pangnirtung interviewees also had many sightings, including five who reported that they see killer whales every year, or did historically (prior to 1977, when 14 were killed after becoming trapped in a saltwater lake, Reeves & Mitchell, Reference Reeves and Mitchell1988), and another reported that he sees them ‘frequently’.
In contrast, Arviat and Kimmirut had no interviewees with many killer whale sightings, and Rankin Inlet, Hall Beach, Igloolik and Iqaluit had very few (≤14%). Kimmirut interviewees had never seen killer whales in the local area (N = 2), or had seen them once (N = 1) or rarely (N = 1), although the fifth interviewee reported occasional observations from the 1960s to the 2000s. Four Arviat interviewees reported only a single sighting in the local area, despite a lifetime of travelling and boating there (interviewees born between the 1930s and 1950s), and the fifth (born in the 1960s) had never seen killer whales. One Rankin Inlet interviewee reported seeing killer whales on ‘many occasions’, but the other interviewees had seen killer whales in the local area only once (N = 3), twice (N = 2) or never (N = 3) (with most reporting additional sightings in other areas). In Igloolik, most interviewees reported seeing killer whales in Foxe Basin rarely or occasionally (N = 11 noting less than five sightings in their lifetime) or having never seen them there (N = 2), although one reported seeing ‘lots’. Three Igloolik elders declined an interview, and two indicated they had never seen killer whales before. Four Hall Beach interviewees had only seen killer whales once, another had never seen them in Foxe Basin, and another had seen them four times. Another 17 people who were contacted declined an interview, including nine who stated that they had never seen killer whales before and four who said they had only seen them once, at a distance. Four Iqaluit interviewees reported only seeing killer whales once, despite living there for 40-plus years. The youngest Iqaluit interviewee (born in the 1970s) reported sightings every summer in recent years, but always outside of Frobisher Bay. Three Iqaluit elders declined an interview, stating that they do not see killer whales in the bay. Results for Repulse Bay were variable, with one interviewee reporting that he saw killer whales every year, others with occasional sightings (N = 8) and others having never seen killer whales or only seen them in other areas.
RELATIVE ABUNDANCE AS REPORTED BY INTERVIEWEES
Many interviewees formerly lived in other areas of Nunavut, travelled extensively when younger, or often visit relatives in other communities, and thus have knowledge of killer whales in multiple areas. In total, 22 interviewees provided information on the relative abundance of killer whales in different areas of Nunavut (range 2–9 interviewees per region). Killer whales are reported to be less abundant in Foxe Basin than they are in Pond Inlet (N = 8), Arctic Bay (N = 6) and Repulse Bay (N = 2). Repulse Bay Inuit also see more killer whales than those in Rankin Inlet (N = 3), Iqaluit and Kimmirut (N = 1 each). Both the Pond Inlet and Pangnirtung areas were reported to have more killer whales than near Arviat (N = 1 each). One interviewee also indicated that there are more killer whales in Admiralty Inlet (near Arctic Bay) than Pond Inlet, and another suggested that there were more near Pangnirtung than Pond Inlet (historically at least). These observations further reinforce the pattern of relative abundance measured by interviewee sighting frequency. One Qikiqtarjuaq interviewee, a commercial fisherman, also reported that killer whales are more abundant offshore in Baffin Bay and Davis Strait, an area where there has been no directed study, than along coastal Baffin Island.
Movements and Migration
Interviewees provided extensive information on killer whale movements and suspected migration routes, summarized in Figure 4. Interviewees in all regions noted that killer whales are usually seen migrating past, although the whales sometimes stay around for a few days to several weeks to hunt (N = 18; Foxe Basin = 6, Hudson Bay = 3, North Baffin = 5, South Baffin = 4). Killer whales can be fast swimmers that are always moving around, or staying in different areas for only a short period of time (N = 24; Foxe Basin = 6, Hudson Bay = 2, North Baffin = 9, South Baffin = 7). Twelve interviewees reported that killer whales generally avoided shallow water, migrating through the deeper areas. This included seven people from Foxe Basin, who all noted that killer whales are not seen in the eastern side of the basin because it is too shallow. Five interviewees (Foxe Basin and North Baffin regions) reported that more killer whales migrate offshore than in shallow coastal waters. Reported major travel routes for killer whales are through Hudson Strait and Lancaster Sound (Figure 4). These whales follow similar routes as other marine mammals, as they may be migrating to follow their food. Nine Foxe Basin interviewees noted a relationship between bowhead population growth and increased killer whale presence, suggesting that killer whales were following bowheads to their summering areas.
Killer whales moved into Admiralty Inlet from Lancaster Sound as the ice breaks up, and were seen most frequently on the west side of the inlet where narwhal congregate (N = 6). They have also been observed near the community, which was reported to be a recent occurrence (N = 6). Five Pond Inlet interviewees discussed killer whale migration (both east and west) through Eclipse Sound. Killer whales move throughout Cumberland Sound, and sometimes even enter Pangnirtung Fjord, close to the community (N = 6). Three Iqaluit interviewees noted that killer whales are rarely seen inside Frobisher Bay, and are usually seen outside the bay. After migrating through Hudson Strait, killer whales enter Repulse Bay through Frozen Strait (N = 3), and they will also follow narwhal into Lyon Inlet (N = 3). One interviewee noted that killer whales will also move south through Roes Welcome Sound, and reported a sighting there, whereas another noted that he had never heard of killer whales there. Five interviewees noted that killer whales move north and south along the western Hudson Bay and Foxe Basin coastlines. One Repulse Bay interviewee noted that killer whales are often seen in that area around five days after being reported in Igloolik or Rankin Inlet. Foxe Basin interviewees suggested that killer whales came north from Hudson Bay and Hudson Strait, moving through either western (N = 4) or central (N = 3) Foxe Basin. Four interviewees suggested that killer whales do not migrate through Fury and Hecla Strait, but three others indicated that they do, following bowhead whales into the Gulf of Boothia in summer, and following narwhal in the opposite direction in autumn.
Absolute abundance
Interviewees in all four regions provided estimates of the number of killer whales thought to be present in the local area, although these numbers were highly variable (Table 2). Most interviewees (N = 84) provided at least some information, although 22 of these simply noted that they did not know or it was a topic to which they had given no thought. Two interviewees noted that abundance varied year to year, depending on food supply and/or ice conditions. Seventeen interviewees provided a numerical estimate of local abundance, with seven suggesting a local abundance of less than 20 whales (Table 2). Thirteen suggested a local population of less than 100 whales, although four indicated larger numbers: an estimate of 400–500 from Repulse Bay, a report from Pond Inlet that hundreds would be seen in offshore Baffin Bay in summer during the 1960s, an Arctic Bay interviewee who estimated over 200 killer whales were in Admiralty Inlet in late August 2004, and another Arctic Bay interviewee who felt that there might be hundreds in the area every year. Sixteen interviewees did not explicitly estimate local abundance but did note the largest aggregations that they had seen or heard of. Reports were again highly variable, with most reporting maximums of less than 30 whales, although several reported large aggregations of 100-plus animals. Qualitative estimates of abundance were provided by 24 interviewees, with the majority (15) stating that killer whales are rare (or few in number, hardly any, not many). Conversely, eight interviewees stated that there were lots, quite a few, or too many (Table 2). Three North Baffin interviewees indicated that killer whales are more abundant in offshore regions than they are near the coast.
Table 2. Summary of information provided by interviewees on abundance of killer whales in their local areas.
Population trends
Most interviewees (75, 71.1%) provided an opinion on changes in killer whale abundance, local occurrence, and/or sightings over time (four others (one per region) noted that they were unsure if the population size was changing, and were removed from statistical analyses of regional patterns given the small sample size). Interviewees often distinguished between changes to sighting frequency and changes in population size (Table 3). Of 53 interviewees who reported ‘increases’, 27 specifically stated that the population was increasing, and 13 noted that sighting rates were increasing but did not mention anything about population size. Another eight reported increases in both sighting frequency and population size, and five suggested that sighting rates were increasing but the population was stable (i.e. killer whales spending more time in the local area, but no growth in numbers). Eight interviewees suggested that the killer whale population in their area was stable. For statistical analyses, reports of increasing numbers or increasing sightings were combined into one category, except for the reports of increasing sighting rates of a stable population, which were combined with reports that the population size was stable.
Table 3. Summary of interviewee opinions and observations of changes in killer population size and/or sighting frequency.
There was a significant difference in responses among regions (Fisher's exact test, I = 75, df = 6, I < 0.0001). Most interviewees thought that killer whale presence was increasing in the Foxe Basin, Hudson Bay and North Baffin regions (Figure 2B). No interviewees in the Hudson Bay or Foxe Basin regions felt that killer whale numbers (or sightings) were decreasing (but note the high degree of non-response for Hudson Bay: 17 of 30 with no data), and only three North Baffin interviewees provided the opinion that killer whale numbers were decreasing. In comparison, 11 of the 17 South Baffin interviewees that provided a response to this question suggested that the killer whale population was declining (every Pangnirtung interviewee). All 11 interviewees reported that killer whales are seen less frequently and in smaller groups than in the past, although one said that they have started to see more of them again in the last decade. Ten interviewees discussed an incident in 1977 as a possible link to this decline: 14 killer whales became trapped in a saltwater lake and were killed by local hunters (Reeves & Mitchell, Reference Reeves and Mitchell1988). Some participants felt that these killer whales were possibly part of the same group that returned to this area each year, and there have not been as many frequenting the area since this hunting event. Two said that killer whales did not return to the area for a long time after the incident, and one thought that in the future more killer whales will return again.
DISCUSSION
Interviews with Inuit hunters and elders provided extensive information on the distribution, abundance, and movements of killer whales, and these data represent a significant contribution to the body of knowledge on this species in the Canadian Arctic. Semi-directed interviews allowed an open and flexible format that avoids the rigidity of questionnaire surveys, but also resulted in variation in the types of information collected. Inuit harvesters are reliable observers with good recall abilities, and the semi-directed interviews provided information at a long temporal and wide spatial record. Interviewing elders allowed us to extract a long temporal record and a wider spatial record (from a period when Inuit were more mobile, prior to social changes such as movement into permanent communities). Information reliability is always a concern with historic recollections provided by older interviewees, as reliability can decline with time since an event was observed (e.g. Lien et al., Reference Lien, Stenson, Carver and Chardine1994). Inuit informants, however, may be ‘more reliable’ than many survey or interview participants (e.g. commercial fishermen). Freeman (Reference Freeman1976: vol. II) noted that people in non-literate cultures are trained from early childhood to have accurate memories, and for young men in hunting societies the greatest emphasis is on accurate recall of environmental information. Smith (Reference Smith1991: p. 153) considered the memories of his Inuit informants to be ‘very keen’, with an accuracy of recall that was ‘extremely high’. Arima (Reference Arima and Freeman1976: 35) concluded that Inuit recall was highly reliable within living memory from at least young adulthood, i.e. for approximately half a century.
Nonetheless, some degree of memory attrition was inevitable among older interviewees. Freeman (Reference Freeman1976: vol. II, p. 53) noted that for land use interviews conducted in the early 1970s, it was apparent that incomplete data was being provided for the early years of the study (the years prior to the local arrival of traders, generally pre-1910s–pre-1930s throughout Nunavut). The information provided for this period was typically in the form of general statements rather than specific, personal information. Reliable information was thus provided for a period of ~50 years, in agreement with the conclusions of Arima (Reference Arima and Freeman1976) noted above. Some elder interviewees also were no longer active harvesters due to age/health concerns, and thus were not as knowledgeable about recent trends in killer whale occurrence (many of whom acknowledged this as well). Any recent bias from elder interviewees would have a negligible effect on the overall patterns and trends however, given that approximately half (43 of 89 with age information) of the interviewees were middle-aged (born in the 1950s or 1960s) active harvesters who, while living in a much different socio-economic environment than their parents, were able to provide a current perspective on killer whale occurrence near their communities. These results would suggest that, given the age distribution (Table 1) of our interviewees, we have reliable (as perceived by interviewees and interpreters) information on killer whale occurrence in Nunavut from the present and extending back at least to the mid-1900s.
Killer whales were observed during summer (July–September) in all areas of Nunavut where interviews were conducted, most often in North Baffin communities, particularly Arctic Bay. Killer whales were described as fast swimmers (see Matthews et al., Reference Matthews, Luque, Petersen, Andrews and Ferguson2011) that are always moving around, staying in different areas for short periods of time. They are often seen passing through or migrating past sighting locations, but they sometimes remained in one area for short periods (days to weeks) hunting other marine mammals. Killer whales were reported to migrate through Hudson Strait and Lancaster Sound and follow the routes used by their marine mammal prey (Ferguson et al., Reference Ferguson, Higdon, Chmelnitsky, Ferguson, Loseto and Mallory2010a, Reference Ferguson, Higdon and Westdal2012a). Belugas, narwhals, and bowhead whales follow regular migration routes and show fidelity to summer ranges (Higdon et al., Reference Higdon, Hauser and Ferguson2012), providing predictable foraging locations for killer whales. Most interviewees thought that killer whale presence was increasing in their area, with South Baffin (Pangnirtung) a notable exception.
Interviewees in Foxe Basin felt that the increase in killer whale occurrence there was related to the increasing bowhead population (Ferguson et al., Reference Ferguson, Higdon and Westdal2012a). Northern Foxe Basin is a nursery ground occupied by cow–calf pairs and juvenile bowhead whales in spring and early summer (NWMB, 2000; Cosens & Blouw, Reference Cosens and Blouw2003), when ice cover provides a possible refuge from killer whale predation (Ferguson et al., Reference Ferguson, Dueck, Loseto and Luque2010b, Reference Ferguson, Kingsley and Higdon2012b; Higdon & Ferguson, Reference Higdon, Ferguson, Ferguson, Loseto and Mallory2010). When sea ice habitat is reduced in late summer and early autumn, bowhead calves and juveniles are likely more vulnerable to predation, and reports of predation attempts on bowhead whales peak during this time (Ferguson et al., Reference Ferguson, Higdon, Chmelnitsky, Ferguson, Loseto and Mallory2010a, Reference Ferguson, Higdon and Westdal2012a; Higdon et al., Reference Higdon, Hauser and Ferguson2012). Satellite-tagging studies confirm bowhead selection for heavy ice conditions in summer, and killer whale avoidance of the same (Ferguson et al., Reference Ferguson, Dueck, Loseto and Luque2010b; Matthews et al., Reference Matthews, Luque, Petersen, Andrews and Ferguson2011). Some groups of killer whales may preferentially hunt bowhead whales (Ferguson et al., Reference Ferguson, Kingsley and Higdon2012b; Matthews & Ferguson, Reference Matthews and Ferguson2013 (this volume)), and the importance of Foxe Basin as a summertime foraging area may increase in the future.
Killer whales were considered to be less abundant than other marine mammal species found throughout the eastern Arctic, although estimates of local population sizes were variable. In general, interviewees suggested that the local population of killer whales around their communities varied from tens to the low hundreds, and they were usually observed in small groups (<10) (the database analyses in Higdon et al., Reference Higdon, Hauser and Ferguson2012 show similar patterns). Young et al. (Reference Young, Higdon and Ferguson2011) identified 53 individual killer whales from photographs of nine different sightings in Hudson Bay and along Baffin Island. Only one whale was re-sighted (within a year), and no matches were discovered when photographs were compared to 63 individuals photo-identified in the north-west Atlantic (Lawson et al., Reference Lawson, Stevens and Snow2007; updated to 67 individuals in Lawson & Stevens, Reference Lawson and Stevens2013 (this volume)). Both photo-identification catalogues are known to be incomplete, as not all individuals were identified in each photographed encounter and numerous photographs were of insufficient quality for use (Young et al., Reference Young, Higdon and Ferguson2011; Lawson & Stevens, Reference Lawson and Stevens2013 (this volume)). The total number of killer whales found in the western North Atlantic and eastern Canadian Arctic is higher than the combined estimate of the two catalogues, but population size is not known with any certainty. Population structure is similarly uncertain, although analyses of stable isotopes in killer whale teeth indicates some dietary separation at both the individual and population levels (Matthews & Ferguson, Reference Matthews and Ferguson2013 (this volume)), suggesting that multiple ecotypes may exist in this region.
Forney & Wade (Reference Forney, Wade, Estes, Brownell, DeMaster, Doak and Williams2006) synthesized information on global killer whale abundance using a variety of sources and assigned abundance classes for different regions. They considered killer whales to be ‘common’ (0.20–0.40 whales per 100 km2) in parts of the Canadian Arctic, ‘uncommon’ (0.10–0.20 whales per 100 km2) in Newfoundland and Labrador waters, and ‘rare’ (0–0.10 whales per 100 km2) in more southerly areas such as the Bay of Fundy and the eastern United States. In contrast, killer whales were ‘abundant; (>0.40 whales per 100 km2) in parts of the north-east Atlantic such as northern Norway and the Norwegian Sea. COSEWIC (2008) assigned the killer whales found in the Canadian eastern Arctic and north-west Atlantic to a single population due to uncertainty in stock relationships (but see Matthews & Ferguson, Reference Matthews and Ferguson2013 (this volume)), with an area of occupancy of over 3.3 million km2. If we consider killer whales to be ‘uncommon’ (Forney & Wade, Reference Forney, Wade, Estes, Brownell, DeMaster, Doak and Williams2006) across the entire eastern Canadian range, resulting population size estimates would range from 3300 to 6600. COSEWIC (2008) estimated that the population size was likely <1000 mature animals (and possibly <250 mature animals), based on sighting effort and knowledge of Pacific killer whale population abundance. Population size is underestimated in the photo-identification catalogues and may also be underestimated by Inuit interviewees, particularly if killer whales are abundant in offshore regions, as was suggested by several interviewees with experience in that area. Of note, most interviewees who did provide an opinion on population size also indicated a high level of uncertainty in their estimate.
Marine mammal research in remote Arctic regions is expensive and logistically challenging, particularly for species like killer whales that exist at relatively low densities, range over large areas, and have variable movement patterns contingent on local sea ice conditions. Directed field research on Arctic killer whales has been successful (Matthews et al., Reference Matthews, Luque, Petersen, Andrews and Ferguson2011), but most published studies have been based on chance observations (e.g. Steltner et al., Reference Steltner, Steltner and Sergeant1984; Campbell et al., Reference Campbell, Yurick and Snow1988; Laidre et al., Reference Laidre, Heide and Orr2006) or compilations of sighting reports (e.g. Reeves & Mitchell, Reference Reeves and Mitchell1988; Higdon, Reference Higdon2007; Higdon & Ferguson, Reference Higdon and Ferguson2009; Higdon et al., Reference Higdon, Hauser and Ferguson2012). Observations by local harvesters compiled through semi-directed interviews have contributed much information on killer whale movements, relative abundance and distribution over a wide spatial range and a long temporal record, in addition to data on predation ecology (Ferguson et al., Reference Ferguson, Higdon and Westdal2012a) and human dimensions of Arctic killer whale management and conservation (Westdal et al., Reference Westdal, Higdon and Ferguson2013). Local observations, including those collected here, also contribute most of the available records in the Arctic killer whale sighting database (Higdon et al., Reference Higdon, Hauser and Ferguson2012). Information from different communities was often complementary, and there was consistency in interview comments related to the relative abundance of killer whales in different areas of Nunavut, seasonal distribution patterns, and trends.
Information derived from TEK can complement that obtained via modern science techniques, and long-term community observations can provide greater confidence in scientific results (Johannes et al., Reference Johannes, Freeman and Hamilton2000). The incorporation of TEK to address scientific research questions is relevant and valuable, and continued research on Arctic killer whales will be most effective if it integrates modern science approaches with the traditional skills, knowledge and experience of Inuit harvesters. Efforts are underway to develop a citizen science programme, where local residents can collect and contribute data (e.g. identifiable photographs), and a community-based monitoring programme. Many of the photographs in the photo-identification database (Young et al., Reference Young, Higdon and Ferguson2011) have been contributed on an opportunistic basis by both southern and Inuit scientists and photographers working on a variety of projects. Continued submissions from local photographers will be critical to developing an expanded database with photographs from around the region. Similarly, two recent killer whale strandings in Hudson Bay (Ferguson, unpublished data) were both discovered and reported by local Inuit, and residents of the local communities conducted the sampling.
Inuit are concerned about environmental change and killer whale effects on marine mammal species are a descriptive indicator of socio-economic and cultural changes happening in the Arctic. The importance of top-down predation results in negative opinions towards killer whales due to the perceived competition for similar resources (Westdal et al., Reference Westdal, Higdon and Ferguson2013). Killer whale numbers are likely increasing (there have been no harvests in the Canadian Arctic in over three decades), and increased predation on species such as ringed seals, narwhals and belugas will have implications for the management of subsistence harvests in the future. Increased presence of killer whales could become an additional stressor on marine mammal populations that are already experiencing changing climatic conditions and increased industrial development (Heide-Jorgensen et al., Reference Heide-Jorgensen, Hansen, Westdal, Reeves and Mosbech2013). Inuit observations on the behaviour, local abundance and seasonal distribution of killer whales and their marine mammal prey will be of critical importance to elucidating current and future changes to the Arctic marine system.
The use of TEK information provides a long temporal record that can facilitate retrospective analyses. For example, combining our recent interviews with other accounts of historic Inuit (and whaler and trader) knowledge and observations (Kumlien, Reference Kumlien1879; Soper, Reference Soper1944; Haller et al., Reference Haller, Foote, Cove and Anders1966; Reeves & Mitchell, Reference Reeves and Mitchell1988; Kilabuk, Reference Kilabuk1998) provides a time-series of relative killer whale abundance in the Cumberland Sound (Pangnirtung) area that spans approximately 150 years. Sighting effort could be estimated through analyses of hunting and travelling patterns and Inuit population growth and dispersion (e.g. Boas, Reference Boas1884; Haller et al., Reference Haller, Foote, Cove and Anders1966; Kemp, Reference Kemp and Freeman1976; Stevenson, Reference Stevenson1997 with respect to Cumberland Sound), and then used in conjunction with marine mammal catch histories (e.g. Stewart, Reference Stewart2004; Higdon, Reference Higdon2010), survey results, and population growth models to examine marine mammal population trends over time and the potential impacts of killer whale predation.
ACKNOWLEDGEMENTS
We thank the numerous Inuit participants (interviewees and interpreters) who shared information on Nunavut killer whales during our tours of 11 communities (listed in Ferguson et al., Reference Ferguson, Higdon and Westdal2012a). We would like to specifically acknowledge the assistance of Marius Tungilik who, sadly, passed away in late December 2012, for logistical support and interpreting in the community of Repulse Bay. Additional thanks to Tara Bortoluzzi and the community Hunters and Trappers Organizations for logistic support. This research was conducted under permits from the Nunavut Research Institute (License No. 0500808N-M) and the University of Manitoba (research ethics approval). Finally, we thank two referees and the Editor for their constructive comments on an earlier version of this manuscript.
FINANCIAL SUPPORT
This research was supported by the following funding agencies: Nunavut Wildlife Management Board Research Trust Fund, Fisheries & Oceans Canada, International Polar Year (Global Warming and Arctic Marine Mammals), the University of Manitoba, and Natural Sciences and Engineering Research Council.
