Distribution and seasonal abundance of large cetaceans in the Durban whaling grounds off KwaZulu-Natal, South Africa, 1972–1975

Daily charts of the aerial search effort (432 206 nautical miles) of the Union Whaling Company and 1 099 sightings of 10 497 whales were available from 628 flights off Durban between 1972 and 1975. Densities of whales were analysed by month and water depth distribution over the four-year period. Low observed densities of blue Balaenoptera musculus, right Eubalaena australis, sei B. borealis and humpback Megaptera novaeangliae whales most likely resulted from earlier whaling pressure. Seasonality of blue, sei and humpback whales was bimodal, indicative of winter migrations to the north of the Durban whaling grounds, whereas the unimodal seasonality of fin whales B. physalus and minke whales B. bonaerensis or B. acutorostrata suggest the offshore region as the northern terminus of their migrations. Sperm whales Physeter macrocephalus migrate northwards offshore of the KwaZulu-Natal coast in autumn/early winter and southwards in late winter/spring, with larger males migrating later than the smaller males and females. Killer whale Orcinus orca presence was coincident with that of offshore minke whales and the southward migrations of other baleen whales, whereas densities of animals deemed as bottlenose whale Hyperoodon planifrons suggest strong early and late summer seasonal abundance in the offshore region. Such extensive surveys offshore of the KwaZulu-Natal coast are unlikely to be repeated; hence, data-extraction of whaling records provides a valuable source of seasonal and distributional information for marine management.

The seasonal distribution patterns of pelagic migratory species, although difficult and expensive to determine, remain crucial in ocean conservation management. Most Southern Hemisphere baleen whales make extensive annual migrations from summer polar feeding grounds to winter breeding grounds in lower-latitude waters, and despite extensive catches of these species by modern whaling in the twentieth century (Tønnessen and Johnsen 1982;Clapham and Baker 2002), the winter distribution patterns remain poorly understood at ocean basin scales.
Baleen whale species recorded within the coastal and offshore waters of KwaZulu-Natal on the east coast of South Africa include both pygmy Balaenoptera musculus brevicauda and Antarctic blue whales B. m. intermedia, fin whales B. physalus, sei whales B. borealis, both Antarctic B. bonaerensis and dwarf minke whales B. acutorostrata, Bryde's whales (either B. brydei or B. edeni, although the exact taxonomic status is uncertain), humpback whales Megaptera novaeangliae, southern right whales Eubalaena australis and pygmy right whales Caperea marginata. Such records include scientific survey-and incidental sightings, strandings and whaling industry catch and sighting records (e.g. Bannister and Gambell 1965;Gambell 1966Gambell , 1972Gambell , 1977. The majority of these baleen whales are highly migratory, with information on the distribution and seasonal abundance in the Southern Ocean in summer derived from whaling records (e.g. Morch 1911;Risting 1912;Mackintosh and Wheeler 1929;Matthews 1938aMatthews , 1938bMackintosh 1942;Omura 1953Omura , 1973Tønnessen and Johnsen 1982) and sighting surveys (Brown et al. 1995;Kasamatsu et al. 1996Kasamatsu et al. , 2000Murase et al. 2002;Gedamke and Robinson 2010;Thiele et al. 2000Thiele et al. , 2004. Information on the winter distribution patterns of many of the Southern Hemisphere species within lower-latitude temperate, subtropical and tropical waters is more limited (particularly in offshore regions), with a large proportion of the seasonality and distribution records coming from coastal whaling operations (e.g. Olsen 1914;Harmer 1929Harmer , 1931Matthews 1938aMatthews , 1938bMackintosh 1942;Angot 1951;Bannister and Gambell 1965;Gambell 1967Gambell , 1972Best 1974Best , 1982Williamson 1975).
Larger toothed whales (greater than 6 m in length) recorded from the KwaZulu-Natal coastal region include sperm whales Physeter macrocephalus, Arnoux's beaked whales Berardius arnouxii, Cuvier's beaked whales Ziphius cavirostris, bottlenose whales Hyperoodon planifrons and Introduction killer whales Orcinus orca. Like the baleen whales, the provenance of such records includes sightings (e.g. Findlay et al. 1992;Sekiguchi et al. 1993), strandings (e.g. Ross 1984Findlay et al. 1992;Peddemors 1999) and, in the case of sperm whales, whaling industry catch and sighting records (e.g. Gambell 1966Gambell , 1972. However, apart from information on sperm whales, little has been documented on the seasonal abundance or distribution patterns of toothed whales off the KwaZulu-Natal coast. Whaling operations were carried out from Durban, on the east coast of South Africa, each year from 1908 to 1975 (Best and Ross 1989). The current study investigated the seasonal abundance and distribution patterns of sightings of nine large cetacean species, or 'forms', sighted by aerial spotter crews within the Durban whaling grounds between 1972 and 1975, including blue (probably both Antarctic and pygmy) whales, fin whales, sei whales, minke (possibly both dwarf and Antarctic) whales, humpback whales, southern right whales, sperm whales, killer whales and large unidentified beaked whale species, termed 'bottlenose whales' by spotter aircraft personnel. The seasonal presence of the baleen whale species in the area results from their breeding migrations, because little feeding happens in the region (Mackintosh 1942;Bannister and Baker 1967). It should be noted that, by the 1970s, a number of these species had been whaled to extremely low abundance levels (Chapman 1974) and few sightings of blue, sei, southern right and humpback whales are recorded in the dataset. Sighting data from whaling operations do not show species selectivity biases inherent in catch data. For example, Best (1982) noted that minke whales were never a species of choice in the Durban whaling grounds, because they were not only of low value, but shark predation on the carcasses meant that they had to be taken alongside immediately, rather than flagged for later collection, and such practice slowed catcher vessels in their pursuit of larger, more valuable species. Sighting data are not constrained by such limitations and consequently provide a valuable body of information on both distribution and seasonal abundance patterns.

Material and methods
The Union Whaling Company, based in Durban (29°53′ S, 31°03′ E) on the east coast of South Africa, undertook flights over the Durban whaling grounds (Figure 1) to search for whales during the whaling season each year between 1955 and 1975. Daily flight paths of search effort and associated sightings are available (held by the Mammal Research Institute Whale Unit, University of Pretoria) for the austral late summer/autumn and winter (February-October inclusive) between 1972 and 1975. The modi operandi of the aerial sighting surveys have been fully described by Bannister and Gambell (1965). The altitude and speed of the aircraft (a Cessna 310H) were usually 500 ft (150 m) and 135 knots (250 km h -1 ), respectively, although these were weather-dependent in that in poor visibility the altitude was reduced to 400 ft (120 m), whereas in good visibility it was increased to 600 ft (180 m). The aircrews usually consisted of a pilot and a single spotter and the area searched usually depended on where the catcher fleet was operating, because the aircraft would normally remain within 60 nautical miles of the nearest catcher vessel. The search effort usually comprised parallel legs 10 nautical miles apart and could extend up to 180 nautical miles from the coast. Daily durations of searching could be up to 8 hours, so that more than 1 000 nautical miles could be surveyed in a day. Bannister and Gambell (1965) state that in 'good to fair' conditions, visibility averaged approximately 8 miles, although it was reduced to half of this in 'moderate to poor' conditions. They report further that, although whale blows were visible at up to 14-16 nautical miles from the aircraft, breaches were visible at greater distances, and that the distance at which blows could be seen was less for sperm whales than baleen whales. During flights, positions of the aircraft and all sightings were measured as distances and bearings from Durban Harbour using a radio beacon.
Daily charts of the flight paths searched and associated sightings were available for 628 flights between 1972 and 1975 and, for purposes of the current study, were processed as described below. Daily flight paths were digitised from the charts (as startpoints, endpoints, and waypoints for each flight) and sighting data recorded in sighting logs were entered into data retrieval files. Recorded sighting data included date, species identification, group size, latitude and longitude (to the nearest nautical mile), and in the case of sperm whales, estimated body size. All sperm whales sighted were categorised by the aerial spotter crews into four size classes: (a) large (>13.7 m); (b) medium (11.8-13.7 m); (c) small (<11.8 m); and (d) undersize (less than the minimum catch size of 35 ft [10.5 m]), with classifications having been made in imperial feet without access to reference objects of known size. Although there was some feedback from catcher crews who observed the same sightings, the size classification was not always entirely accurate. The term 'bottlenose whale' was used by spotter aircraft personnel to describe any large unidentified beaked whale, although these records have been ascribed to southern bottlenose whales Hyperoodon planifrons (Sekiguchi et al. 1993). The bathymetry of the region was digitised from South African hydrographic charts (scale 1:600 000), which showed the coastline and the 50 m, 100 m, 200 m, 500 m, 1 000 m, 2 000 m and 3 000 m depth contours (see Figure 1). Custom-written application programmes allowed for: 1. the plotting of contour maps, and flight tracks and associated sightings, and 2. the calculation of the search effort and relative sighting densities on the whaling grounds by: 3. 10-minute latitude and longitude blocks on the whaling grounds (because navigation accuracy did not warrant finer scale analyses), 4. each of 0-50 m, 50-100 m, 100-200 m, 200-500 m, 500-1 000 m, 1 000-2 000 m, 2 000-3 000 m and >3 000 m depth intervals, and 5. month.
Seasonal abundance by month was calculated as the total count of individuals of each species (and size class in the case of sperm whales) sighted during each month as a function of the total distance searched that month over all four years. Relative densities by depth interval were calculated as the sum of individuals sighted as a function of search effort (with inputs measured as sightings or nautical miles searched in 10-minute squares of latitude or longitude whose centre point fell within the depth interval). All sighting effort was assumed to be of equal intensity, so that effective search widths of the flight paths were not determined and the relative densities calculated here reflect encounter rates by 10-minute square. Although effort biases might be apparent, such aerial survey-derived densities do not reflect the catch selectivity often apparent in relative densities derived from whaling catch-series.

Results
Flight paths totalling some 432 206 nautical miles were searched over the four-year period, the distribution of which is shown by area in Figure 2 and by month and water depth in Figure 3. It should be noted that the relatively low search effort in October means that the relative densities for this month should be regarded with caution. In addition, the search effort during October tended to be more confined to coastal waters than in other months. A total of 1 099 sightings of some 10 497 individuals was made during these flights (Table 1), the distribution of which is shown in Figure 4. The individual species distributions by month are provided in the Supplementary Material, available online. The nine species or types recorded included blue, fin, sei, minke, humpback, sperm, southern right, killer and 'bottlenose whales', suggesting that most species of smaller cetaceans remained unrecognised or unrecorded.

Blue whales
Although the summer Antarctic distribution of blue whales is well documented, little is known of their winter migratory destinations (see Mackintosh 1942;Brown 1954;Kasuya and Wada 1991). Branch et al. (2007) noted that blue whales are generally absent from the central regions of each of the major Southern Hemisphere ocean basins, and that records are completely absent in the southcentral Indian Ocean. However, Samaran et al. (2013) recorded northern Indian Ocean blue whale call types to the north-east of the Amsterdam Islands and Stafford et al. (2004) and McDonald et al. (2006) reported Antarctic blue whale calls from the central Indian Ocean. Yochem and Leatherwood (1985) reported that blue whales occur as far north as Madagascar off the African east coast, and Angola and occasionally Gabon on the African west coast. Although few species-specific sighting records are available for the south-western Indian Ocean (apart from those recorded by Best et al. 2003), acoustic recordings (Stafford et al. 2004;McDonald 2006) suggest both Antarctic and pygmy blues can be found throughout the region.
Five sightings of blue whales (four single individuals and one pair) were made in the Durban whaling grounds between 1972 and 1975, the distribution of which is shown in Supplementary Figure S1 (available online). No distinction between pygmy or Antarctic blue subspecies could be made. All of the blue whale sightings were in water depths of >500 m ( Figure 5), but no differences in the observed and effort-based expected densities were recorded by depth (χ 2 = 1.466042, df = 6, p = 0.962), although this is assumed to reflect the few sightings of this species. Branch et al. (2007) noted that blue whales were generally associated with waters deeper than the continental shelves and that shallow-water records were typically from regions with narrow continental shelves. Frequencies of recorded group sizes are shown in Figure 6. Sightings were too few for seasonality to be determined, although the crude bimodal distribution of sightings by month ( Figure 7) might reflect northward (March-May) and southward (September) migrations through the region, with blue whales present in the Durban whaling grounds relatively early in the season compared with other species. Branch et al. (2007) noted that there is good evidence (including size distribution, seasonal distribution of catches, and a steeply declining catch per unit effort) for the winter migration of Antarctic blue whales to the west coast of South Africa, Namibia, and Angola, and to the east coast off Durban. However, they noted that the more recent (1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975) seasonal records off Durban were primarily in March-May compared with the June-July peak in earlier catch records. This, together with both the single pygmy blue whale record by Gambell (1964) and the incidence of pregnant blue whales shorter than 75 ft caught at Durban, suggests that during the more recent period (coincident with the records analysed here), the majority of blue whales encountered off Durban were pygmy blue whales.

Fin whales
As in the case of blue whales, little is known of the wintering grounds of Southern Hemisphere fin whales (Mackintosh 1942). Gambell (1985) suggested that the Indian and Atlantic Ocean winter destinations of migratory fin whales lie to the north of South African waters. In all, 52 sightings of 84 fin whales were made in the Durban whaling grounds between 1972 and 1975 (Supplementary Figure S2). Group size ranged between one and three individuals, with a mean of 1.6 (SD 0.7) ( Figure 6). All groups sighted were in waters deeper than 500 m ( Figure 5), although there were no differences in the observed and effort-based expected densities recorded by depth (χ 2 = 12.0948, df = 6, p = 0.0675). The distribution of sightings of fin whales in this study appeared to show an association with the 1 000-3 000 m depth range with a clear east-west band in the region of 30°-31° S where the shelf edge has an eastwest orientation (Supplementary Figure S2). Furthermore, the sightings showed an increasing alignment to the shelf edge as the season progressed, indicative of an association with the shelf-edge bathymetry during migrations in a northeasterly direction until mid-July and in a south-westerly direction thereafter; thus the 'northward' and 'southward'  migrations were in fact north-easterly and south-westerly migrations, respectively, through the whaling grounds. However, Mackintosh (1966) reported that catches of fin whales in the Durban whaling grounds largely comprised immature individuals and proposed that adults are found mainly in deeper waters at the limit of the catcher range. In the current study, the seasonal abundance of fin whales off Durban showed a unimodal pattern with peak abundance in July (Figure 7), signifying that the Durban whaling grounds might represent the northern terminus of their migrations in the south-western Indian Ocean, which is contrary to Gambell (1985). Horwood (1987) maintained that sei whales are found in warm oceanic waters near the equator in winter and towards polar regions in summer. The fact that only four sightings of six sei whales were made in the Durban whaling grounds between 1972 and 1975 presumably reflects the collapse of the Southern Hemisphere sei whale stock in 1967/1968 (Best and Gambell 1968), and little information on distribution or seasonality can be inferred from these sightings. The distribution of two single individuals and two pairs of individuals is shown in Supplementary Figure S1. All were in water deeper than 1 000 m ( Figure 5) and, as with blue whales, the lack of difference in the observed and effort-based expected densities recorded by depth (χ 2 = 2.756, df = 6, p = 0.8387) is presumed to reflect the paucity of sightings and inshore search effort.    [1972][1973][1974][1975] Although the sightings are too few to draw any conclusions on seasonal abundance (Figure 7), they conform to a northward (April-June) and southward (September) migration through the study region, in accordance with the June and September seasonality trends found by Bannister and Gambell (1965). Matthews (1938b) found a similar northward migration in June. Stewart and Leatherwood (1985) described minke whales as common in Antarctica during summer, occurring off the west coast of South Africa and Angola (although no seasonality is given) and present throughout the year off Durban. Best (1982) described the seasonal abundance, migration, group composition, feeding, age and reproduction of minke whales off Durban, whereas Best (1985) described two colour phases of minke whales within the Durban whaling grounds, namely a 'bonaerensis type', now recognised as the Antarctic minke whale B bonaerensis, and a smaller type, now recognised as the dwarf minke whale, a conspecific of the northern minke whale B. acutorostrata (Pastene et al. 1994). Densities of the dwarf form were, however, very low off Durban. In all, 155 groups that totalled 258 minke whales were sighted in the Durban whaling grounds between 1972 and 1975, though no differentiation by species was recorded. Group sizes ranged between one and six individuals (Figure 6) Figure  S3) shows a general orientation parallel to the coast in the region of the continental shelf edge, which suggests that minke whales follow the shelf edge during their migrations to and from the Southern Ocean. The majority of these sightings were in waters deeper than 500 m ( Figure 5), and although there was a marked difference in the observed and effort-based expected densities recorded by depth (χ 2 = 51.2126, df = 6, p < 0.01), fewer sightings were in the inshore region where catches of the dwarf form were made (Best 1985). The seasonal abundance of minke whales showed an increase in abundance as the whaling season progressed (Figure 7), with peak abundance between July and September. Unlike many of the other species recorded, there was no autumn peak.

Humpback whales
Only 21 sightings of a total of 38 humpback whales were made in the Durban whaling grounds between 1972 and 1975 (Supplementary Figure S4). The low incidence (as with blue and sei whales) reflects the collapse of the Western Indian Ocean populations as a consequence of modern whaling earlier in the 20th century (Findlay 2001). Sizes of these groups ranged between one and five individuals ( Figure 6), with a mean of 1.8 individuals (SD 0.9). The relative abundance by depth interval (Figure 5) shows that humpback whales occurred in two depth regions; inshore of the 500 m isobath and offshore of the 1 000 m isobath.
There was a marked difference in the observed and effortbased expected densities recorded by depth (χ 2 = 46.0944, df = 6, p < 0.01). The observed seasonal abundance is somewhat surprising, given the current understanding of a bimodal abundance in KwaZulu-Natal waters as the animals migrate between the calving grounds off Mozambique and, presumably, the Southern Ocean (Findlay et al. 2011a). Although humpback whales were present throughout the whaling season, a peak abundance was evident in October (presumably corresponding to a southward migration, but the low effort in October makes confirmation difficult), but there was no marked early (July) peak, corresponding to a northward migration (Figure 7). It is possible that the northward migration occurred inshore of the southwardflowing Agulhas Current off Durban and was therefore inside the more offshore search effort of the spotter aircraft. Furthermore, the greater amount of inshore search effort during October might accentuate the relative abundance of the southward migration in these water depths.

Southern right whales
Two sightings of single southern right whales were made in the Durban whaling grounds (Supplementary Figure  S1). Although sightings are obviously too few to determine any patterns in distribution or seasonal abundance, the offshore distribution of both of the sightings in water deeper than 500 m ( Figure 5) and the early sighting in March (Figure 7) are surprising, given the species' association with nearshore waters on the southern Cape coast in late winter and spring (Best 1990).   [1972][1973][1974][1975] small, and undersize size classes are shown in Figure 5. There were differences in the observed and effort-based expected densities recorded by depth for all size classes (large sperm whales: χ 2 = 74.5622, df = 6, p < 0.01; medium sperm whales: χ 2 = 150.6109, df = 6, p < 0.01; small sperm whales: χ 2 = 943.5509, df = 6, p < 0.01; and undersize sperm whales: χ 2 = 597.2587, df = 6, p < 0.01). Although there was no significant difference between size classes in the distribution with water depth (F = 0.7492; df = 3, p > 0.5), small and undersize whales appeared to occur closer to shore than large or medium-sized whales. Figure 6 clearly shows the incidence of large, single individuals, which would most likely have been males, because females congregate into larger groups (Best 2007). Small sperm whales were most abundant early in the season, with a large peak centred around April and a smaller peak in July, whereas both large and medium sperm whales showed peaks of abundance in May and July/August (Figure 7). Both Gambell (1967) and Best (1981) found that females and small males were most abundant from February to May, declining thereafter, whereas large and medium-sized sperm whales peaked from May to August.

'Bottlenose whales'
In all, 138 sightings of a total of 584 'bottlenose whales' (considered by Sekiguchi et al. [1993] to have been southern bottlenose whales) were recorded by spotter aircraft personnel in the Durban whaling grounds between 1972 and 1975 (Supplementary Figure S9). Group sizes of these sightings ranged between one and 12 individuals, with a mean of 4.34 individuals (SD 2.28) (Figure 6). A marked difference in the observed and effort-based expected densities was recorded by water depth (χ 2 = 236.3723, df = 6, p < 0.01), because all sightings were in waters deeper than 500 m (Figure 5), and a strong austral spring and summer seasonality between October and February is implied (Figure 7), although it cannot be confirmed because of the absence of data from December and January.

Killer whales
In all, 22 sightings of a total of 161 killer whales were recorded in the Durban whaling grounds between 1972 and 1975 (Supplementary Figure S10, Table 1). Group sizes ranged between one and 25 individuals (Figure 6), with a mean group size of 7.32 individuals (SD 4.68). Kasamatsu et al. (1988) found a large variation in the sizes of killer whale groups in Antarctic waters, ranging from 1 to 380, although they noted the difficulty in assessing killer whale school size because of group dispersal. The strong seasonality in abundance of killer whales (Figure 7) mirrored the late seasonal abundance of minke whales and/or the southward calf-accompanied migration of other species, such as humpback whales. Pitman et al. (2015) reported on killer whale predations on humpback whale calves off Western Australia and noted that, despite the low incidence of successful predations, the occurrence of presumed killer whale tooth-rake marks on the flukes of nearly every large whale species (particularly humpback whales) provides evidence that attacks could be common during migrations to and from breeding grounds.

Distribution patterns
The austral winter destinations (and therefore the breeding grounds) of the Southern Hemisphere baleen whale species, apart from those of humpback and southern right whales, are assumed (because of a lack of coastal records or data) to lie in deep oceanic waters, an assumption supported by relatively recent advances in acoustic monitoring (Stafford et al. 2004). The sightings of blue whales were too few to define adequately any distribution or seasonal abundance patterns, although those recorded were in deep offshore waters. Catches of 2 986 blue whales recorded at Durban between 1908 and 1964 comprised immature and adult blue whales (Branch et al. 2007). These authors reported that early catches comprised Antarctic blue whales, but that by the end of the blue-whaling era the majority of catches and sightings were probably pygmy blue whales. Gambell (1964) reported a pygmy blue whale taken off Durban in September 1963 andBranch et al. (2007) noted that the abundance of pygmy blue whales south of Madagascar, as recorded by Best et al. (2003), peaks in summer, when Antarctic blue whales are in the Southern Ocean. Pygmy blue whales in the northern Indian Ocean are believed to form a resident population (Yochem and Leatherwood 1985;Anderson 2005), although the subspecies identity of these and other pygmy blue whale forms remains speculative. Year-round autonomous acoustic recordings have detected Antarctic blue whale type calls in the austral autumn and winter in the central Indian Ocean (Stafford et al. 2004;McDonald et al. 2006). Furthermore, three distinct blue whale call types have been recorded in the Indian Ocean (and tentatively ascribed to pygmy blue whales): a Sri Lankan call off north-east Sri Lanka (Alling et al. 1991) and off Diego Garcia (Stafford et al. 2004;Branch et al. 2007); a Madagascar call (Ljungblad et al. 1998), which has also been recorded at Diego Garcia (Stafford et al. 2004) and off Crozet Island (Samaran et al. 2006); and a West Australian call recorded in south-western Australian waters (McCauley et al. 2004cited in Branch et al. 2007Stafford et al. 2004Stafford et al. , 2011. Mackintosh (1966) found 10 of 11 fin whale sightings made off the Patagonian coast to be closely associated with the shelf edge and reported that fin whales follow a migration route in deep water not far from the continental shelf. Bannister and Gambell (1965) noted that the aerial spotters off Durban reported that the migrations of fin whales followed the continental shelf, with whales moving out of the Durban whaling grounds in a north-easterly direction until mid-July, and into the grounds from the east and out of the grounds to the south-west from the end of July onwards. Such a deep-shelf-edge association was also clearly evident in the current study.
Minke whale sightings showed a general orientation parallel to the coast in the region of the continental shelf edge, suggesting some orientation of the migration by ocean topography as noted for fin whales (Bannister and Gambell 1965). A difference in the distribution of the two forms of minke whale was recorded in the Durban whaling grounds by Best (1985), who found the dwarf minke whale to occur inshore of the Antarctic minke whale; hence the nearshore records analysed here might represent this form.
Seven breeding stocks of humpback whales (termed A-G) are recognised from the Southern Hemisphere (IWC 1998(IWC , 2011, each of which migrates between Antarctic feeding grounds and warm tropical breeding grounds within coastal waters of Southern Hemisphere continents or archipelagos. Breeding Stock C of the south-western Indian Ocean overwinters in four broad coastal regions, namely off Mozambique and possibly southern Tanzania, off the islands in the Mozambique Channel, off Madagascar, and off the islands of the Mascarene archipelago (Best et al. 1998;IWC 2011). The distribution of humpback whales on their Mozambican breeding grounds has been described by Findlay et al. (1994Findlay et al. ( , 2011b. Migration links of humpback whales between the KwaZulu-Natal and Mozambique coasts have been identified through recoveries of lost harpoons (Olson 1914) and through photo-identification studies (Cerchio et al. 2008), and links between the KwaZulu-Natal coast and Plettenberg Bay (34.07° S, 23.37° E) have been identified by Banks (2013). Discovery mark returns (see Best 2007 for details) link high-latitude feeding grounds with Madagascar (Rayner 1940). Humpback whales utilise the nearshore waters of Southern Hemisphere continents (within the 200 m isobath) as migratory corridors once they reach the coast on the northward migration, and before they diverge from the coast on the southward migration (Findlay et al. 2011b). Off Durban, the two peaks in the humpback whale distribution by water depth possibly reflect (a) an inshore migration to and from the Mozambique breeding grounds (Findlay et al. 2011b), and (b) an offshore migration, either en route to more easterly breeding grounds in the Mozambique Channel, or Madagascar, or an offshore migration to Mozambique. However, the timing of the offshore sightings coincides with that on the northward migration as found by Findlay and Best (1996) off Cape Vidal, some 300 km to the north of Durban, and is assumed to involve individuals that have yet to converge with the coast.
The offshore distribution of the two southern right whales in our study is surprising, given that Best (1990) found 90% of right whales off the southern Cape coast to be within 2 km of the shore. Richards and du Pasquier (1989) discussed three centres of right whaling in southern African waters, namely Walvis Bay, the Cape coast and Maputo (formerly Delagoa) Bay, and it is probable that the right whales sighted in the current study were en route to Maputo Bay. Banks et al. (2011) reviewed recent records of southern right whales from southern Mozambique, noting that no records had been published from Mozambican waters since those of open-boat whaling where records of catches were from Delagoa Bay and Sofala Bay in the early 1800s (Wray and Martin 1983).
The distribution patterns of 'bottlenose whales' in the current study appear widely distributed in, and strongly associated with, the deeper offshore waters of the warm southward-flowing Agulhas Current system. All sightings were in waters deeper than 500 m. This distribution is similar to that found for southern bottlenose whales in South African waters and the predilection of this species for the warmer waters in the core of the Agulhas Current (see Findlay et al. 1992) is in contrast to the species' known distribution in polar waters in summer (see Van Waerebeek et al. 2010). The type specimen of this species is from 20° S on the north-western coast of Australia (Flower 1882), although it should be noted that this finding was based on a beach-worn skull that might not have originated from the region.
Although a cosmopolitan distribution of killer whales in South African waters was found by Findlay et al. (1992), the records of killer whales in the current study all occurred in deep offshore waters beyond the 1 000 m isobath.

Seasonal abundance
Despite the paucity of information on seasonal abundance, blue whales appear to arrive first in the Durban whaling grounds, followed by sei whales and then fin whales. Although the data appear to show that all three species left the region at the same time (September), search effort in October was low and generally in more coastal waters than in other months. Bannister and Gambell (1965) found blue whales mostly between April and June, with a few occurring up to September, and they noted that, although the 56 whales taken between 1954 and 1966 were too few to draw conclusions about seasonal abundance, the migratory cycles of blue whales appear to precede those of baleen whales. Harmer (1931), however, found a unimodal trend in the seasonal abundance of blue whale catches off KwaZulu-Natal, with a maximum from June to August, and suggested that blue whales did not migrate as far north on the African east coast as they do on the west coast, as a consequence of the more southerly location of warmer waters in the southwestern Indian Ocean. Samaran et al. (2013), however, reported Antarctic blue whale call types off the Maldives, well to the north (and east) of the Durban whaling grounds, suggestive of migrations to this region in the austral winter, so that a marked bimodal seasonality pattern would be expected off Durban, if the Durban whaling grounds formed part of this migration corridor. Harmer (1931) found a unimodal distribution in seasonal abundance of catches of fin whales off Durban with a maximum in June and July and concluded that the species does not migrate farther north than the Durban whaling grounds on the east coast. Such a seasonality pattern is similar to that found by Bannister and Gambell (1965), with fin whales present off KwaZulu-Natal between April and October and a peak of abundance in July. Gambell (1985) noted that fin whales tend to migrate to and from the Antarctic feeding regions after blue whales, but before sei whales.
The seasonal abundance of minke whales reported here and by Best (1982) is generally in agreement with seasonal abundance patterns off Brazil (at 7° S), where Williamson (1975) found that the first minke whales arrive in late June and July, abundance increases to a maximum between September and November, and the last individuals leave in December. Similarly, Da Rocha (1980) found peak densities of minke whales off Brazil (6°-8° S during September and October. Best (1982) noted that the incidence of the cold-water skin diatom Benettella ceticola was high during the period June-September, suggesting a movement by minke whales into KwaZulu-Natal waters from the Southern Ocean at that time. Kasamatsu and Ohsumi (1981) found that Antarctic minke whale densities increased from November to a peak in January, declining thereafter, and Best (1982) noted that the majority of mature minke whales appear to be in Southern Ocean waters in summer. The direction of movement of minke whales found by Best (1982) was still predominantly northward at the end of the whaling season (September), so that southward movement must happen later than this or offshore of the area of survey. Furthermore, a marked difference in migrations by sex class was recorded by Best (1982), in that catches early in the season comprised a large proportion of sexually immature individuals; during the months of June to August males outnumbered females by more than 2:1, and the sex ratio become more equal only in September.
The seasonal abundance of humpback whales found in the current study is markedly different from seasonal abundance patterns in the Durban whaling grounds reported previously. Olsen (1914) reported on the seasonality of humpback whale catches over the period 1910-1912, although Best et al. (1998 suggested that the 1912 data might have been compromised by both an industry strike and depletion of the whale stock. The seasonality of catches in 1910 and 1911 was almost identical, with peaks in the last 10 days of July and in midto late September (Olsen 1914). Matthews (1938a) found a marked bimodal seasonality in the catches of humpback whales off KwaZulu-Natal between 1918 and 1928, with the northward and southward migrations being represented by peaks in July and September, respectively. Bannister and Gambell (1965), reporting on the period 1954-1963, showed humpback whales to be present off KwaZulu-Natal between April and September, with a peak abundance in July representing the northward migration and a slight peak in September representing the southward migration. Findlay and Best (1996) and Findlay et al. (2011b) described the bimodal seasonal migration patterns of humpback whales on the northern coast of KwaZulu-Natal. By contrast, Olsen (1914) found unimodal seasonality (with a peak in August) off Linga-Linga, Mozambique, i.e. to the north of KwaZulu-Natal, suggesting this region to be the terminus of the migration (although humpback whales are recorded to migrate well north of this, i.e. at least as far north as Kenyan waters [Cerchio et al. 2013]).
Interpretation of the seasonality of southern right whales is compromised by the paucity of sightings. However, the March sighting is somewhat asynchronous with the late winter and spring seasonality of the species found elsewhere in South African waters (Best 1990) and with the August/September seasonality of 12 sightings in southern Mozambique reported by Banks et al. (2011).
All four of the size classes of sperm whales under study showed bimodal seasonal abundance, with relatively lower densities during June. Small and undersize sperm whales were most abundant early in the season, with a large peak in April and a lesser one in July, whereas both large and medium whales showed later seasonal abundance, with small peaks of abundance in May and larger peaks in July/ August (large whales) or August (medium-sized whales). Such seasonal abundance patterns of sightings are similar to those of catches of sperm whales off Durban between 1954 and 1963 (Gambell 1967), although the late-season sighting densities are lower in the current study (1972)(1973)(1974)(1975) data) than in the 1954-1963 data. These peaks were not apparent, however, in the sighting records between 1959 and 1963 (Gambell 1967) or between 1961and 1969(Gambell 1972, possibly as a result of the heavy exploitation of large sperm whales after 1961. The seasonal abundance patterns support a northward migration off the KwaZulu-Natal coast in autumn/early winter and a southward migration in late winter/spring, with larger males appearing later in the season than small males and females. Gambell (1967) suggested that sperm whale conception takes place in summer to the south of Durban, followed by a northward migration in autumn and a return southward migration the following spring. Because the majority of sperm whale births happen in April (Gambell 1966), sperm whales would be expected to be found off Durban during the subsequent northward migration.
Although killer whales are found in southern African waters year-round (Findlay et al. 1992), Mikhalev et al. (1981) found that migrations of Southern Hemisphere killer whales appeared to coincide with migrations of rorqual whales, particularly minke whales, and Best (1982) reported a possible attack by a group of killer whales on a minke whale within the Durban whaling grounds. The late-season abundance within the study area coincides with the seasonal abundance of minke whales or southbound migrations of other large baleen whales. All killer whales in South African waters have been assumed to be of Pitman and Ensor's (2003) Type A (Best 2007), although  reported the presence of a second morphotype in SA waters. Although no type information for those sighted in the current study is available, a genetic analysis of killer whales from South Africa (including material from 36 individuals taken from the Durban whaling grounds between 1971 and 1975) shows both high haplotype diversity and a high incidence of private haplotypes (Moura et al. 2014), compared with samples collected from elsewhere in the Southern Hemisphere.
The high seasonal abundance of 'bottlenose whales' in both February and October suggests a strong summer seasonality pattern in local waters, although the lack of midsummer search effort precludes confirmation of this. Nevertheless, such summer seasonality is surprising, given the numerous summer sightings of southern bottlenose whales in Antarctica, as reported by Van Waerebeek et al. (2010) and Sekiguchi et al. (1993). Sekiguchi et al. (1993) found squid beaks from four Antarctic species and four sub-Antarctic species in two bottlenose whale stomachs from South African waters, and Nemoto et al. (1980) recorded the presence of the Antarctic skin diatom Benettella ceticola from a specimen sampled in southern African waters off East London, suggesting some migratory movement between these regions. Both the near-total lack of sightings in the high-effort winter months, (which is suggestive of movement out of the Durban whaling grounds), and the high Antarctic densities recorded in summer (Sekiguchi et al. 1993), indicate a winter movement to the north of the Durban whaling grounds and a summer distribution that extends from south of Durban to Antarctica.
The distribution and seasonal migration patterns of many of the pelagic cetacean species remain poorly understood and the paucity of information means that the seasonality and distribution data examined here are extremely valuable, despite the fact that they are both dated and overlap an era of low abundance of a number of species that had declined during some 60 years of commercial whaling in the first part of the 20th century. The extensive survey coverage of 432 206 nautical miles is unlikely ever to be repeated at this scale in this area, and the records of 1 099 sightings of an estimated 10 497 whales provide an important legacy in the understanding of the seasonality and distribution patterns of the offshore large cetaceans off the KwaZulu-Natal coast. Baseline data extracted from such historic records are crucial starting points in the assessment of stock structure of populations and in the development of management plans for the mitigation of increasing anthropogenic marine and maritime pressures in the region.