ABSTRACT

The Dall’s porpoise, Phocoenoides dalli (True 1885), the largest species in the family Phocoenidae, inhabits the northern North Pacific and adjacent seas It consists of two major pigmentation types, the dalli-type and the truei-type, and a rare black variation The dalli-type has a white flank patch extending from the anal region to the level of the dorsal fin and is called in Japanese ishi-iruka or ishi-iruka-type The truei-type has a larger flank patch that reaches to the base of the flipper and is called rikuzen-iruka or rikuzen-iruka-type Caution is required because ishi-iruka is also used as a species name for the Dall’s porpoise Japanese fishermen along the Pacific coast of northern Japan call a Dall’s porpoise of any pigmentation kamiyo or kamiyo-iruka and use hankuro (half black) for dalli-type individuals, which are less common in the region Suzume-iruka or kita-iruka has also been used for the species

Among the three species of Phocoenidae found in Japanese waters, the Dall’s porpoise is the only offshore inhabitant and swims at high speed, raising rooster-tail splashes The finless porpoise, another member of the family, inhabits temperate and tropical waters, usually within the 50 m isobath, and is known to the south of 38°24′N on the Pacific coast and 37°30′N on the Sea of Japan coast The third species of the family, the harbor porpoise, usually inhabits the continental shelf area north of 35°N on the Pacific coast and north of 37°N on the Sea of Japan coast Dall’s porpoises are easily identified by a distinct large white patch on the flank and a profile of small head and elevated thoracic portion The dorsal fin is almost triangular as in the case of the harbor porpoise, which is in contrast with the falcate dorsal fin of many delphinid species, and the trailing edges of the dorsal fin and flukes are marked with a gray or white margin

The average body length of full-grown adults is 1981 cm in males and 1897  cm in females from the central North Pacific (Ferrero and Walker 1999), but there seems to be large individual variation and some geographical variation The skull of the Dall’s porpoise is the largest among those of the three Phocoenidae species off Japan, and the rostrum is more pointed than in the other two species (Table 81) Among the Phocoenidae, the Dall’s porpoise has the largest number of vertebrae, 92-98 in total, which are compressed anteroposteriorly and have high neural processes

The Dall’s porpoise has a history of being thought to comprise two species, but it is currently dealt with by the Society for Marine Mammalogy as a single species containing two subspecies represented by the truei-and dalli-types This

classification assumes greater distances between the trueitype population and several dalli-type populations than those between the dalli-type populations The species was first described by True (1885) as Phocaena dalli based on a specimen collected by DH Dall in waters west of Adak Island in the Aleutian Islands He placed this new species in the genus together with the harbor porpoise Instead of the genus name Phocoena, which was created by G Cuvier for the harbor porpoise in 1817, Gray used Phocena in 1821 or Phocaena in 1828, but the original genus name Phocoena is currently in use (Hershkovitz 1966) The specimen described by True in 1885 had the pigmentation of the so-called dalli-type Later, in 1911, Andrews described a new species based on a specimen collected off the Sanriku Region (37°54′N-41°35′N) on the Pacific coast of northern Japan and placed it in a newly created genus Phocoenoides together with Ph. dalli (True 1885) to yield the two species P. dalli (True 1885) and P. truei Andrews 1911 The specimen described by Andrews was of the truei-type A slightly different spelling, Phocaenoides, was once used for the genus, but it has been rejected

The English names of these two species (or color morphs) were True’s porpoise and Dall’s porpoise Japanese cetacean science in its infancy was confused on this Nagasawa (1916, in Japanese) proposed the Japanese common name toru-iruka (True’s porpoise) for the former type, but Kishida (1925, in Japanese) used rikuzen-iruka for it after personal communication with N Kuroda Then, Kuroda (1938, in Japanese) used toru-iruka for P. dallii truei (sic) and Wogawa-iruka (Ogawa’s porpoise) for P. dallii dallii (sic), as two subspecies within the single species P. dalli However, Matsu-ura (1943, in Japanese) dealt with them as different species and used the Japanese name rikuzen-iruka (or toru-iruka) for P. truei and ishi-iruka (or Wogawa-iruka or kita-iruka) for P. dalli These scientists were aware of various vernacular names used for the genus by fishermen along the Sanriku coast (kamiyo-iruka, suzume-iruka, and hankuro) but were apparently reluctant to use them as species name (Kasuya and Yamada 1995, in Japanese), presumably because these names did not usually distinguish between the two nominal species

No objection was raised against moving Dall’s porpoise from Phocoena to a new genus Phocoenoides, but there were various opinions expressed on the taxonomy of the species This was probably due to the absence of identifiable taxonomic difference between the nominal two species except for the geographically segregating pigmentation pattern

Kuroda (1954, in Japanese) cruised along the Kuril Islands from off southern Hokkaido at around 42°N to the western Aleutian Islands area and observed that (1) dalli-types were present in the whole range, but truei-types were found only along the Japanese coast and (2) the two types did not form

mating season However, later in July off southern Hokkaido, he observed that the two types rode bow waves together and obtained a truei-type fetus from a dalli-type female Kuroda (1954, in Japanese) interpreted this to mean that reproductive isolation of the two types was incomplete and the pigmentation pattern was controlled by two alleles at one locus; he concluded that they should be dealt with as separate subspecies

Kuroda (1954, in Japanese) did not clarify whether the two pigmentation types joined on the bow of their ship or were in a group before arriving at the bow Another scientist F Wilke was on the same vessel with Kuroda and stated that mixed schools of the two types were not common even off Hokkaido (Wilke et al. 1953) The interpretation of the genetics of the pigmentation pattern by Kuroda (1954, in Japanese) seems to be too simplistic (Kasuya 1978), and the taxonomic conclusion is questionable even if his interpretation is accepted Our current understanding of the species is that the parturition season is followed by the mating season, which is from midsummer to early autumn and differs between populations (Sections 943 through 945), and that weaned juveniles are segregated from breeding individuals at least during the season of parturition/mating but possibly also in other seasons (Sections 933 and 942)

Miyazaki et al. (1984) reported as results of their sighting survey that (1) only dalli-types occurred in July and August in the Bering Sea and (2) that both types were seen in the Pacific area along the Kuril Islands, where 15 schools were of dallitypes, 24 schools were of truei-types, and 5 schools contained both types The proportion of mixed schools was only about 10% even in waters where both types occurred It seems to me that these mixed schools were formed in the area of coexistence after the porpoises arrived there Such combinations could be considered aggregations as discussed in Section 861

Kasuya (1978, 1982) observed in a 1005  cm full-term fetus obtained from a pregnant dalli-type in the Bering Sea that the anterior lateral portion, which was expected to be black on a typical dalli-type or white on a typical truei-type, had a uniform intermediate gray pigmentation and suggested that this anterior gray area darkens during a period from the late fetal stage to early postnatal stage I stressed the need for further examination of development of the pigmentation pattern during the perinatal period before accepting the hybrid hypothesis of Kuroda (1954, in Japanese) and Wilke et al (1953) After this observation, Szczepaniak and Webber (1992) published the photographs of two newborn individuals, a 97 cm truei-type and 103 cm dalli-type, and stated that distinction of the two types was not difficult at that stage Distinction of the two types was clearer on their two photographs than for the full-term fetus presented by Kasuya (1978), which confirmed that the two pigmentation patterns become distinct in the early stage of postnatal life

Houck (1976) questioned the taxonomic significance of the morphological characters used by Andrews (1911) to distinguish P. truei (Andrews 1911) from P. dalli (True 1885) He pointed out that (1) the greater height of the tail peduncle was a character of the adult male common to both species,

tail flukes, and tail peduncle) varied greatly individually and had little taxonomic value, and (3) only the 5th character, size of the white area on the flank, should be considered as important in evaluating the validity of the two species He further noted on the variation of the flank patch that entirely black individuals had been reported from both sides of the North Pacific and that truei-types occurred only in the western North Pacific and had various degree of individual variation intermediate between truei-and dalli-types (note: a rare example of the truei-type in the eastern North Pacific was later reported by Szczepaniak and Webber (1992)) He concluded that the two types are not reproductively isolated (note: he referred here to the observation of Wilke et al. (1953) on the truei-type fetus found in a dalli-type female) but the two types could not be considered as two subspecies because they were not geographically isolated Therefore, he concluded that the two types should be dealt as only two color morphs in one species It seems to me that he probably overemphasized the absence of reproductive isolation If reproductive isolation were proven between the two pigmentation types in sympatric distribution, the taxonomic argument could proceed in the opposite direction, that is, to classify them as two distinct species

Shimura and Numachi (1987) evaluated the genetic distance among toothed whale species based on the variation in 19 loci identified from isozyme analysis of skeletal muscle and liver Their Dall’s porpoise sample included truei-types from the Sanriku Region and dalli-types from the North Pacific south of the Aleutian Islands at 165°–175′E They found that the two types were genetically different by only 04% and concluded that they should be dealt with as a single species They also found that the genus Phocoenoides was genetically closer to Neophocaena than Phocoena and that these three genera constitute a single group of Phocoenidae

Escorza-Treviño et al. (2004) reached a similar conclusion by analyzing mitochondrial DNA (mtDNA) and microsatellite DNA Their specimen series consisted of 23 truei-types and 113 dalli-types from nearly the entire range of the species They identified 66 mtDNA haplotypes, many represented by minor differences Forty-six of the haplotypes were limited to the dalli-type and 8 to the truei-type and 12 were common to both types The 66 haplotypes were grouped into two major lineages, each represented by both dalli-and truei-types Thus, the evolution of mtDNA did not accord with pigmentation type They identified several populations of the dallitype but only one of the truei-type (Section 936) The genetic distance between the two pigmentation types was similar to those between populations of the dalli-types Based on these findings, Escorza-Treviño et al. (2004) concluded that the two pigmentation types had a long history of genetic exchange and that the absence of genetic distinction between them was due to genetic exchange, which is still continuing or terminated recently The high genetic variability of the species, 66 haplotypes represented in 136 individuals, identified by them supported the earlier result of Shimura and Numachi (1987)

Dalli-types are found in the entire range of the species, but truei-types are found only in waters extending from the

the Pacific coast of northern Japan (a rare exception from the eastern North Pacific is ignored as an outlier; see the preceding text in this section) The density and proportion of trueitypes within the range vary geographically and seasonally, reflecting different migration patterns (Section 93) There is about a 1-month difference between the mating seasons of nearby populations of the two pigmentation types, and adults from the populations aggregate in different breeding grounds This functions to suppress free interbreeding between the pigmentation types, but it is questionable if it is sufficient to explain the observed geographical distribution of the two types and the genetic similarity between them There might be some additional behavioral or physiological elements that restrict successful reproduction between the two types and survival of genetic traits thus introduced About half of the truei-type porpoises landed on the Sanriku coast had various degree of mottling on the anterior part of the white flank patch, but incidents of mottling on the flank patch of dallitype individuals and cases of black forms appeared to be less frequent (Kasuya 1978) Information on the genetic basis of these variations in pigmentation will contribute to understanding the mechanism apparently limiting interbreeding between the dalli-and truei-types

The Dall’s porpoise has several populations, which are also called stocks Free genetic mixing is expected within a typical population but is suppressed between populations For management purposes, the identification of populations should be evaluated with more weight on independence of population dynamics rather than on genetic independence The case of a metapopulation seems to be intermediate, where a group of individuals maintain their own population dynamics but maintain an occasional limited level of exchange with nearby groups of genetic materials or individuals Such a system enables a species in several patchy habitats to maintain stability as a whole and is expected for cetacean species inhabiting coastal habitats such as Tursiops aduncus (see Section 113)

Population structure has been rigorously studied in the Dall’s porpoise One of the reasons was international concern about the management of populations of this species incidentally killed in the high-seas salmon drift-net fishery of Japan Other reasons include the large-scale hunting of the species off Japan and its easily accessible distribution in the northern part of the North Pacific and adjacent seas

The Japanese mother-ship salmon drift-net fishery started operations in 1952 in the Bering Sea and south of the Aleutian Islands A group of scientists led by K Mizue first reported that 10,000-20,000 Dall’s porpoises were killed annually in the fishery (Mizue and Yoshida 1965; Mizue et al 1966, both in Japanese) Later, the US government became concerned about the mortality of the species occurring in its EEZ and started investigating the effect of the fishery on the Dall’s porpoise population A scientific meeting was convened in Seattle in January 1978, and some Japanese university scientists

reluctant to investigate the problem, but with increasing pressure by interested countries, it started cooperation with US scientists in the summer of 1978 The activity was transferred to the International North Pacific Fisheries Commission (INPFC), which was established by the International Convention for the High Seas Fisheries of the North Pacific Ocean between Canada, Japan, and the United States The research activity expanded to the large-mesh drift-net fishery, which targeted tunas and marlins and caused incidental mortality of the Pacific white-sided dolphins, right whale dolphins, and seabirds (INPFC 1993; Sano 1998, in Japanese)

The international research activity aimed to determine whether the incidental mortality was within an allowable level and included studies of population structure, abundance, reproduction, and mortality At the beginning of the program, the Fisheries Agency of Japan offered the opportunity for US scientists to work on Japanese mother ships to examine porpoise carcasses brought by catcher boats that operated the drift-net fishery, but it did not offer its own scientists for the program, pleading unavailability This policy was altered in 1982, when Japanese scientists were placed on research vessels (not on mother ships) to collect biological specimens and sighting data Following a resolution by the United Nations in 1989, Japan in 1992 closed the large-scale drift-net fisheries in the high seas (see Section I) This was the end of all the research activities under international cooperation

The identification of Dall’s porpoise populations was attempted using pigmentation type, level of pollutants, breeding ground, parasite load, and DNA, as outlined in the following

9.3.1.1 Identification of Truei-Type Population The Sea of Japan is inhabited only by dalli-types; the southern limit in the winter season is at around latitude 35°N-36°N (see this section) The southern limit along the Japanese coast moves northward from spring to summer (to around 40°N off Akita Prefecture in June and 43°N off Otaru in Hokkaido in July), and the species almost disappears from the Sea of Japan in midsummer No Dall’s porpoises were recorded in October during sighting surveys that covered the eastern half of the Sea of Japan (Miyashita and Kasuya 1988) Dall’s porpoises that wintered in the Sea of Japan are believed to move to the Okhotsk Sea through Soya Strait (La Perouse Strait) or to the Pacific through Tsugaru Strait (see the following text in this section)

Kasuya (1978, 1982) reported on pigmentation patterns of Dall’s porpoises taken by fisheries off the Sanriku coast on the Pacific coast of northern Japan and found (1) that the incidence of the typical truei-type, which has no mottling or an extremely low level in the white flank patch, was only about 50%; (2) that many of the truei-types had mottling of varying density in the white patch (most of the mottling occurred in an area that should be black on the typical dalli-type but white on the typical truei-type); (3) that some of the dalli-types, which are scarce off the Sanriku coast, had black spots in the white

exhibited individual variation, particularly in the pigmentation in the inguinal region An albino individual was reported off the Kuril Islands (Joyce et al. 1982)

Kasuya (1982) detailed the variation of pigmentation in 537 Dall’s porpoises taken in the winter fishery off Sanriku and recorded the frequency of types: (1) typical truei-types, 514%; (2) intermediate types, 443%; (3) typical dallitypes, 39%; and (4) black types, 04% It should be noted that all the individuals of the intermediate types had black spots in the white background, and none of them had white spots in the black background They could be classified as truei-type interpreted broadly This agrees with visual impressions and allows the conclusion that truei-types made up 957% of the Dall’s porpoises taken by the hand-harpoon fishery off Sanriku in winter

Classifying the intermediate types as truei-types is also reasonable because such intermediate types have been reported only within the known range of the typical trueitypes and not in waters where dalli-types predominantly occur Dall’s porpoises of the truei-type in the broad meaning are now believed to constitute one local population of the species that winters off the Pacific coast of northern Japan, and those of the dalli-type are thought to be temporal migrants from other populations Thus, the proportions of the two pigmentation types vary by season and location

Kasuya (1978) interpreted the then available limited information on the distribution of the species to suggest that dallitypes in the southern Okhotsk Sea and those in the offshore North Pacific are separated by the Kuril Islands and by the habitat of truei-types along the Pacific coast of northern Japan and the Kuril Islands I constructed a hypothesis on the population structure of Dall’s porpoises around Japan that assumed three populations: two dalli-type populations, one in the Okhotsk Sea and Sea of Japan and another to the east of the Kuril Islands and northern Japan, and a truei-type population off the east coast of northern Japan and the Kuril Islands Dalli-types that winter in the Sea of Japan are now known to migrate to the southern Okhotsk Sea through two routes: (1) Soya Strait and (2) Tsugaru Strait and southern Kuril Islands We have no information on the distribution of the species in Tatar Strait, which has a water depth of about 8 m

Discussion among the INPFC scientists on population structure did not move toward improving the hypotheses but at least temporarily moved in quite a different direction The scientists from the Japanese Fisheries Agency and the fishing industry accepted the presence of a trueitype population off Japan but were against the idea of multiple populations of dalli-types They believed that there was insufficient information supporting multiple populations and proposed to deal with all dalli-types in the North Pacific as a single population The hypothesis of Kasuya (1978) was a plausible argument against the Japanese hypothesis If multiple populations are managed under an erroneous single-population hypothesis and fishery mortality is geographically biased, then there is a greater risk of depleting a particular population However, it allows less

of controversy still continues in the management of other cetacean populations

Kasuya (1978) observed Dall’s porpoises from the Ginseimaru No.2 that operated for minke whales in the southern Okhotsk Sea and confirmed the presence of only dalli-types in the area, but my data were limited only to the southern part of the Okhotsk Sea within about 100 km from the coast This deficit was later covered by T Miyashita of the Far Seas Fisheries Research Laboratory He surveyed the whole Okhotsk Sea area during August and September in 1989 and 1990, which agreed with a presumed mating season after parturition and was considered suitable for studying distribution to be used for the study of population structure Results of his survey were presented to the Scientific Committee of the International Whaling Commission (IWC) (Miyashita and Doroshenko 1990; Miyashita and Berzin 1991), used for abundance estimation (Miyashita 1991), and cited in Yoshioka and Kasuya (1991, in Japanese) These surveys confirmed two breeding areas of dalli-types in the northern and southern Okhotsk Sea, separated by a breeding area of truei-types in the central Okhotsk Sea This result leads to a hypothesis of three Dall’s porpoise populations in summer in the Okhotsk Sea (Figures 91 and 92)

The range of the truei-type population, which is one of the three Dall’s porpoise populations summering in the Okhotsk Sea, extends from the coasts of central and northern Sakhalin Island (latitudes 50°N-56°N) toward the southeast and reaches to the northern and central part of the Kuril Islands (latitudes 47°N-50°N) A total of 24 (89%) mother-calf pairs of truei-types out of 27 such pairs found in the Okhotsk Sea were in this central Okhotsk Sea region, and the remaining 3 pairs were further north of the range or in the area occupied by dalli-types (Figure 92) This was used as evidence of a breeding ground of the truei-types in the area The range of truei-types not accompanied by a calf was broader than the range of mother-calf pairs and extended through the Kuril Islands to waters off the coasts of eastern and southern Hokkaido (Figure 91) Mother-calf pairs were not found in these Pacific waters From the similarity in segregation pattern to that between juveniles and breeding population in the Aleutian Islands area of the species (see Section 933), I believe it is likely that the truei-types found in summer in the Pacific waters off Hokkaido are likely to be mostly immature individuals Truei-types do not normally occur in the Pacific waters off the middle and northern Kuril Islands, and the eastern limit of the regular range of truei-types is around 153°E (Miyazaki et  al 1984; Miysashita and Doroshenko 1990; Miyashita and Berzin 1991)

Since most of the Okhotsk Sea is covered with ice in the winter, leaving only a small southeastern area ice-free, trueitypes that summer in the sea are believed to move to the Pacific and winter off the Pacific coast of northern Japan where the hand-harpoon fishery for the species operates The southern limit of their winter range is off Choshi (35°42′N, 140°51′E) (Miyashita and Kasuya 1988) Truei-types do not occur in the Sea of Japan in any season, and the Pacific coast of Hokkaido

is predominantly occupied by dalli-types in October and presumably also in winter (see Figures 94 and 95)

Some dalli-types occur in the central part of the Okhotsk Sea, which is used by truei-types as a breeding ground, but the density is lower than that of the truei-types as well as those of dalli-types in the northern and southern Okhotsk

Sea, that is, a density hiatus of dalli-type matches with the breeding ground of truei-types (Miyashita 1991) This observation suggests the presence of two separate dalli-type populations summering in the northern and southern Okhotsk Sea Sighting surveys by Miyashita identified 48 mother-calf pairs of dalli-types in 1989 and 43 in 1990 in 3 clusters in

area south of the Kamchatka Peninsula or east of the central and northern Kuril Islands in 154°E-163°E These dalli-type breeding grounds are further discussed in the following

Miyashita (1991) observed an absence of mother-calf pairs in waters in the Kuril Islands area and stated that the breeding grounds of dalli-types in the Okhotsk Sea and the one in the Pacific south of the Kamchatka Peninsula are discontinuous (Figure 92) The two breeding grounds of dalli-types in the Okhotsk Sea are separated by that of truei-types, and they seem to represent different populations of dalli-types My interpretation of the population structure of Dall’s porpoises in  the Okhotsk Sea and adjacent waters is the existence of (1)  a truei-type population, which breeds in the central Okhotsk Sea and migrates to the wintering ground off the Pacific coast of northern Japan through the central and southern Kuril Islands (southern limit is in around 35°N-36°N, and presence in winter off east coast of Hokkaido north of 41°30′N is left to be confirmed); (2) a Sea of Japan dalli-type population, which breeds in the southern Okhotsk Sea and migrates to the wintering ground in the Sea of Japan through Soya Strait and via the Pacific coast of eastern Hokkaido and then Tsugaru Strait (presence in winter along the Pacific coast of Hokkaido is still to be determined); (3) a dalli-type population, which breeds in the northern Okhotsk Sea and presumably winters in the western North Pacific (summer range in the Pacific is yet to be confirmed); and (4) a dalli-type population that breeds in waters south of the Kamchatka Peninsula (summer range and wintering ground are yet to be confirmed) The basis of the reasoning is summarized in the following

The distribution of dalli-types in the northern Okhotsk Sea is apparently continuous into the western North Pacific through the Kuril Islands However, this cannot be interpreted to necessarily mean there is only a single population It is more reasonable to pay attention to the presence of two isolated breeding grounds in the area and to assume that two dalli-type populations exist (nos 3 and 4) The distribution of dalli-types in the southern Okhotsk Sea continues into the northern Sea of Japan via Soya Strait in the southwestern part of the Okhotsk Sea (Figure 94) and also to an area dominated by dalli-types off the Pacific coast of eastern Hokkaido (Figures 91 and 95) through the southern Kuril Islands Analysis of the white flank patch of dalli-types by Amano and Hayano (2007) allocated some of the dalli-types off the Sanriku coast to the Sea of Japan population, and a pollutant study by Subramanian et al. (1986) also suggested the strong possibility that dalli-types off the Pacific coast of southern Hokkaido were migrants from the Sea of Japan (see Section 934)

Numerous observations on distribution and pigmentation types of Dall’s porpoises in the Sea of Japan have accumulated (Miyashita and Kasuya 1988) Dall’s porpoises in the sea are represented only by the dalli-type As shown in Table 91, the sea surface temperature at the positions of sightings of these individuals was below 13°C in winter and spring, rising to below 15°C in June and 16°C in July (insufficient data for October) This seasonal change in habitat temperature is smaller than general oceanographic change in the

Sea of Japan because Dall’s porpoises move north following a suitable temperature environment This feature is different from that of the finless porpoise in the Inland Sea, which remains in the almost landlocked sea enduring the stress of large seasonal temperature change between 6°C and 29°C (Section 8452, Figure 86 and Table 101) The climax of the southbound movement of Dall’s porpoises in the Sea of Japan is in March, when they reach a line that connects Hamada (35°N) in Shimane Prefecture, in Japan, and Pohang (36°N) at the southeastern corner of the Korean Peninsula Sea surface temperature at this southern limit was 11°C in winter (Miyashita and Kasuya 1988)

Dall’s porpoises in the Sea of Japan begin a northward movement in spring This is evidenced by fishery data There was a hand-harpoon fishery for small cetaceans off the Tajima coast (c 35°40′N, 134°20′E-135°30′E) (Noguchi and Nakamura 1946, in Japanese) The catch of Dall’s porpoises started in February, peaked in April, and ended in May, while

Sea Surface Temperature (in Celsius) at the Position of Sighting of Dall’s Porpoise Schools in the Eastern Sea of Japan

had a peak in May, and continued to June The starting date of the fishery could have been affected by weather and the presence of other fishing targets, but the shift from one smallcetacean species to another must reflect a seasonal faunal change Thus, the last northbound group of Dall’s porpoises could have left the Tajima coast in May Sighting vessels operated by the Far Seas Fisheries Research Laboratory recorded the southern limit of the species at Yamagata (40°N) in early June and at Otaru (43°N) in July (Miyashita and Kasuya 1988)

One of the cruises to which Miyashita and Kasuya (1988) referred started its survey off Yamaguchi Prefecture in the southwestern Sea of Japan in early October, cruised northward in zigzag track lines, reached Soya Strait (45°30′N) after 23 days, and cruised the southern Okhotsk Sea toward Abashiri (44°01′N, 144°16′E) Sightings of Dall’s porpoises were recorded only twice in the Sea of Japan, at c. 41°N off Aomori Prefecture and at c. 42°N off the Oshima Peninsula in southern Hokkaido The remaining sightings occurred in the Okhotsk Sea This observation suggests that Dall’s porpoises are almost absent during summer and early autumn in the Sea of Japan The sea surface temperatures at the two sighting positions were 19°C-20°C and 13°C-14°C, respectively, and to the west of the surveyed area off Hokkaido and Aomori there was an area of cold water below 14°C Therefore, these observations alone cannot exclude the possibility of distribution of Dall’s porpoises in the western part of the northern Sea of Japan in this season However, another sighting cruise that covered the northern half of the Sea of Japan from the Japanese coast to the EEZ of the USSR from May to June 1969 recorded only one school of Dall’s porpoises, in the central part of the sea (c. 39°N, 135°E) (Kasuya and Kureha 1971) This also supports the conclusion that most Dall’s porpoises that winter in the Sea of Japan leave the sea in the summer

Miyashita and Kasuya (1988) analyzed the distribution of Dall’s porpoises off the Pacific coast of Japan Sea surface temperatures there fluctuate seasonally between 3°C and 28°C, but most sightings of Dall’s porpoises have occurred in waters below 15°C (Table 92) In summer, when surface temperatures are 12°C-28°C, sightings of Dall’s porpoises were limited to temperatures below 24°C and mostly below 18°C There was no difference in temperature preference between dalli-and truei-types in these waters of the Pacific Pacific Dall’s porpoises also tended to avoid high sea surface temperatures as in the Sea of Japan, but the upper bound tended to be higher in the Pacific than in the Sea of Japan In summer, the warm Kuroshio Current expands northward over the cold Oyashio Current, raising the surface water temperature, but the feeding activity of Dall’s porpoises is probably affected also by subsurface temperatures under the influence of the Oyashio Current

According to Miyashita and Kasuya (1988), out of 299 Dall’s porpoise schools sighted off the Pacific coast of northern Japan during May to October, 96 (321%) were schools of dalli-types, 195 (652%) were schools of truei-types, and only 8 schools (28%) contained both types Mixed schools were uncommon and limited to waters where pure schools of both

types were present (Figures 93 through 95) Mixed schools of two distinct species of cetaceans have been recorded with even slightly higher frequencies For example, Kasuya and Jones (1984) sighted 27 schools of five delphinid species during a sighting cruise in the western North Pacific from August to September 1982, and 10 of them (37%) contained more than one species This suggests that the mixed Dall’s porpoise schools were formed temporarily in waters where the two types overlap as aggregations of two or more schools of pure pigmentation types Such schools are likely to dissolve into pure schools

I estimated the proportion of the two pigmentation types  using school composition and mean school size in Miyashita and Kasuya (1988), which covered the months from May to October in the Pacific off Sanriku and Hokkaido By dividing the 8 mixed schools equally into those of each

Sea Surface Temperature (in Celsius) at the Position of Sighting of Dall’s Porpoise Schools in the Western North Pacific

type (4 to each type), I got 100 dalli-type schools and 199 truei-type schools I multiplied them by the respective mean school sizes (95 individuals/dalli-type school and 64 individuals/truei-type school) to obtain the proportions of the two pigmentation types in the surveyed waters, which were 43% dalli-types and 57% truei-types The proportion of dalli-type is greater than the corresponding value of less than 10% obtained from the catch of the hand-harpoon fishery off the Sanriku coast during December to April (Kasuya 1978; Amano et  al. 1998a,b, both in Japanese) The proportions

of the two pigmentation types vary by area and season as discussed in the following

Miyashita and Kasuya (1988) recorded Dall’s porpoise schools of unknown type in March about 50  km east of Choshi Point on the Pacific coast of Japan The sea surface temperature at this position was about 13°C, which was reasonable for the distribution of the species A truei-type school was recorded in May southeast of the earlier sighting at about 34°N, 142°E This sighting occurred in a water mass of 18°C-19°C surrounded by warmer water of 20°C-21°C (Figure 93) Thus, the porpoises could have been isolated from the ordinary range of the species by a seasonal change in oceanography Other sightings of the species occurred to the north of 36°N at surface water temperatures of less than 16°C Therefore, Miyashita and Kasuya (1988) considered the southern limit of the species off the Pacific coast of Japan to be around 35°N-36°N In May, truei-types occurred almost continuously, within a range of about 300 km from the coast, from this southern limit to near the Nemuro Peninsula (43°20′N, 145°50′E) (Figure 93) However, this distribution pattern was apparently different from that observed in July, August, and October, when truei-types occurred in farther offshore waters (Figures 94 and 95) The pattern of seasonal

northbound movements

The distribution of dalli-types off the Pacific coast is quite different from that of truei-types We know that dalli-types were only several percent of Dall’s porpoises in the catch of the hand-harpoon fishery off Sanriku operating in the winter Miyashita and Kasuya (1988) noted a high density of dalli-types along the Pacific coast of Hokkaido from Tsugaru Strait (which connects the Sea of Japan and the Pacific) to the Nemuro Peninsula at the southern end of the Kuril Islands Dalli-types were present in this area in early May and apparently increased in density toward a peak in September and October, when dalli-types represented almost all the Dall’s porpoises within 100  km of the Pacific coast of Hokkaido (Figure 95) In the 3 months of May, September, and October for which we have sighting data, their distribution extended from Tsugaru Strait to the Nemuro Peninsula along the Pacific coast of Hokkaido and then to the southern Okhotsk Sea coast This suggests that they are the dalli-type Dall’s porpoises that winter in the Sea of Japan (Kawamura et al. 1983, in Japanese; Miyashita and Kasuya 1988), which is supported by the analysis of the white flank patch (see Section 9312) In September and October, the density of truei-types off Sanriku coast is still low They have not arrived at the Sanriku ground but appear to be dispersed in more offshore waters (Figure 95)

9.3.1.2 Geographical Variation of the White Flank Patch of the Dalli-Type

The distribution pattern suggests that dalli-types found along the Pacific coast of Hokkaido are of the Sea of Japan origin: this has gained some support from a pollutant study (see Section 934) The question remained of the origin of dalli-types hunted off the Sanriku coast; the answer was obtained from the analysis of the white flank patch This study was started by Amano and Miyazaki (1996) who compared the white flank patch of dalli-types among porpoises from (1)  the Sea of Japan, (3) Bering Sea, (5) offshore western North Pacific west of 165°E, and (6) offshore western North Pacific between 165°E and 180°E (numerals in parentheses correspond to those in Figure 96) and found that the anterior margin of the white patch was located slightly more posteriorly on the Sea of Japan individuals than on those in other locations This meant that Sea of Japan Dall’s porpoises could be differentiated from those in other regions by the size of the white patch This finding was reinforced by Amano et al. (2000) with additional data from the eastern North Pacific (4) and by Marui et al. (1996, in Japanese) for the southern Okhotsk Sea (2)

This information was used for the identification of the origin of dalli-types hunted off the Sanriku coast and the results published in Marui et al. (1996, in Japanese) and Amano et al. (1998a,b, both in Japanese) These three studies were combined, with no significant changes in the analyses and conclusions, into one scientific publication by Amano and Hayano (2007) Marui et al. (1996, in Japanese) first plotted the distance from the tip of the rostrum to the anterior margin of the flank patch and confirmed that most of the samples were separated into two groups: porpoises from the Sea of Japan and

southern Okhotsk Sea and those from other areas including the Bering Sea, eastern North Pacific, offshore western North Pacific west of 165°E, and offshore western North Pacific between 165°E and 180°E Using this method, Amano et al (1998a,b, both in Japanese) and Amano and Hayano (2007) confirmed that about 90% of dalli-types were correctly classified into those in the geographical locations of the Sea of

Japan-southern Okhotsk Sea and offshore North Pacific They then tested dalli-types taken by the hand-harpoon fishery off Sanriku and found that about half of them were classified into the Sea of Japan-southern Okhotsk Sea group and the remaining into the offshore western North Pacific group (Figure 97) Their analysis did not cover dalli-types from the northern Okhotsk Sea

The analyses of the flank patch of dalli-types show that the current Japanese hand-harpoon fishery in the Sea of Japan and southern Okhotsk Sea targets dalli-types of the Sea of Japan-southern Okhotsk Sea population during spring to autumn The catch of the similar fishery off Sanriku, which operates during the winter, includes about 95% truei-types and about 5% dalli-types (see Section 9311), and the studies by Amano and Hayano (2007) and others have shown that the dalli-types (about 5% of the total catch of Dall’s porpoise off Sanriku) are divided equally between porpoises from the Sea of Japan-southern Okhotsk Sea population and some other population(s), that is, the northern Okhotsk Sea, North Pacific, and Bering Sea populations Our current understanding of the distribution pattern of Dall’s porpoise populations around Japan is summarized in simplified form in Figure 98

Amano et al. (1998b, in Japanese) made comparisons of sex ratio and maturity between populations of dalli-type Dall’s porpoises taken in the hand-harpoon fishery off Sanriku The male proportion was 56% among dalli-types from the Sea of

Japan-southern Okhotsk Sea population, which was similar to 66% observed for dalli-types from other population(s) on the same ground Male maturity identified from testicular weight varied annually between 5% and 27% for dalli-type males from the Sea of Japan-southern Okhotsk Sea population but from 32% to 47% for other dalli-types Averages for the 3 years’ data were 149% and 407%, respectively Female maturity had to be estimated from body length because the internal organs were often discarded by the fishermen at sea Female dalli-types from the Sea of Japan-southern Okhotsk Sea population showed annual maturity of 0%–10% (33% average), and those from other dalli-type populations had 29%–67% mature (429% average) Thus on the Sanriku ground, dalli-type porpoises of both sexes migrating from the Sea of Japan-southern Okhotsk Sea population had apparently lower sexual maturity rates than those from other population(s) However, there remained some uncertainty in the absolute figures, particularly for females

Before accepting these results, the maturity criterion for females needs to be reexamined In the analysis, female porpoises of body length of 187 cm or more were judged sexually mature in the Sea of Japan-southern Okhotsk Sea population (Amano and Kuramochi 1992), but 170 cm was used for other population(s) The former figure was based on observations on hand-harpoon vessels, and the latter on incidental mortality in the salmon drift-net fishery in the Bering Sea (Newby 1982) (it was close to the figure of 1720 cm later published by Ferrero and Walker (1999)) These figures represent body lengths where the probabilities of being immature and mature are expected to be equal, but they are not necessarily

or from different habitats Dall’s porpoises respond differentially to hand-harpoon vessels by growth stage and also segregate geographically by growth stage A sample obtained from the hand-harpoon fishery, which underrepresents sexually mature individuals due to differential response to the vessels, is likely to yield a greater average body length at the attainment of sexual maturity compared with the corresponding estimate for samples from less selective fishing methods such as drift-net fisheries Segregation by growth stage also has a similar effect Ferrero and Walker (1999) reported the body-length range of over 5000 individuals as 84-222  cm for males and 86-211 cm for females Amano and Miyazaki (1992) reported the body-length range of dalli-types captured in the hand-harpoon fishery from the Sea of Japan-southern Okhotsk Sea population as 215-220 cm for males and 205210 cm for females (greater body size to 230 cm was reported by Walker (2001) and Yoshioka et al. (1990) for the southern Okhotsk Sea Dall’s porpoises; see Sections 932 and 935) Even if the maximum body sizes are similar between the two sample sets, there is a difference of 17 cm between the average body lengths of females at attainment of sexual maturity (170 cm for the drift-net sample vs 187 cm for the hand-harpoon sample) This can be attributed to gear selection or geographical differences in growth The mean body length at sexual maturity calculated for samples from the drift-net fishery in the Bering Sea is likely to overestimate to some unknown degree the proportion of sexually mature females in the catch of the hand-harpoon fishery off Japan See Section 932 for a possible bias in a similar direction due to a longitudinal cline in body size

9.3.2.1 Skull Morphology Amano and Miyazaki (1992) carried out multivariate analysis of 27 measurements of 289 Dall’s porpoise skulls At least 25 of the specimens were truei-types, which were caught off the Pacific coast of Japan, and most of the remaining were dalli-types covering a broad area including the Sea of Japan, southern Okhotsk Sea, Bering Sea, and almost the entire longitudinal range in the North Pacific Specimens from around Japan were obtained from the catch of the hand-harpoon fishery, which included truei-types and a small number of dalli-types taken off Sanriku in winter and dalli-types taken in other areas during spring through autumn Samples from the Bering Sea and offshore North Pacific were obtained during cruises for scientific purposes using hand harpoons The sample also included some stranded individuals

Amano and Miyazaki (1992) found sexual dimorphism in some measurements They did not detect sexual dimorphism in skull length, which was probably due to limited sample size combined with large individual variation Their analysis rejected a hypothesis of “no geographical differentiation,” which means that the skull morphology exhibited some sort of geographical variation However, they failed to diagnose the geographical origin of specimens using

result can be obtained when individual variation is great compared with difference between populations or when a single sample represents multiple populations, both of which are very likely in this case

9.3.2.2 Body Length Kasuya (1978) compared body-length compositions of Dall’s porpoises taken in winter by hand-harpoon fisheries and suggested that dalli-types in the Sea of Japan were likely larger than truei-types off the Pacific coast of northern Japan A finding of Amano and Miyazaki (1992) was geographical variation in body size of the species in a broader Pacific area Dall’s porpoises of both types taken in Japanese coastal waters west of 155°E had an average skull length of about 33-34 cm (range approximately 31-36 cm), which was similar to the corresponding figure from the California coast However, those in the central North Pacific between 165°E and 145°W had an average skull length of about 32-33 cm (range approximately 30-35  cm); they were about 1  cm smaller As the authors did not mention possible body-length difference expected from the different skull size, I attempted to use the equations of Amano and Miyazaki (1993) for this, that is, Y = 100056(X0584) for males and Y  =  100004(X0616) for females, where X is the body length in cm and Y is the length from tip of the rostrum to the ear in cm These equations suggest that a body-length difference of 15 cm between 205 and 225 cm is equivalent to about 1  cm difference in length from tip of the rostrum to the ear

Amano and Miyazaki (1992) interpreted their findings on geographical variation in skull size to suggest that Dall’s porpoises in coastal waters were larger in body size and that this reflected high productivity of the habitats The results can also be examined assuming a linear east/west cline The Bering Sea sample, 8 males and 18 females, offers key information for evaluating the two hypotheses The sample can be classified as a coastal sample, but it belongs to the central part of the longitudinal range in the North Pacific Skull length ranged from 305 to 345 cm, which was similar to the range obtained for the central North Pacific Thus in my view, the finding of Amano and Miyazaki (1992) could be better interpreted by saying that Dall’s porpoise populations on the eastern and western sides of the North Pacific are likely to have larger body size than those in the central portion of the longitudinal range

The earlier calculation suggests that Dall’s porpoises along the Japanese coasts are about 15 cm larger than those in the central Pacific However, a greater geographical difference in body length was suggested by Yoshioka et al. (1990), who analyzed dalli-types (one black type included) taken by the single method of hand harpoon, in the commercial hand-harpoon fishery in the southern Okhotsk Sea and in the harpoon operation for scientific purposes in the offshore North Pacific and Bering Sea outside of the US EEZ during 7 summers from 1982 to 1988 They grouped the sample into four longitudinal strata: (1) the southern Okhotsk Sea, (2) western North Pacific west of 160°E, (3) western-central North Pacific between

160°E and 180°E and the Bering Sea, and (4) eastern North Pacific between 137°W and 180°W They compared the mean body lengths between these strata They found that body length was greater in the western samples, at least within the longitudinal range covered by the strata Mean body lengths of sexually mature individuals in the westernmost stratum were greater than in the easternmost stratum by 30  cm in males and 25 cm in females, and the values for the 2 strata between them were intermediate (Figure 99) Comparison of immature individuals suggested a similar trend This study did not have materials from the Californian coast, which were included by Amano and Miyazaki (1992), but showed a similar geographical cline as that detected by Amano and Miyazaki (1992) within the latitudinal range analyzed

Humpback whales provide a typical example of a migration pattern between breeding and feeding grounds (Clapham 2002), where most individuals regularly travel between a particular summer feeding ground and a winter breeding ground (also called wintering ground) where they almost do not feed Individuals that summer in a feeding ground, for example, Southeast Alaska, depart in autumn for their breeding grounds, which differ among individuals, for example, Hawaii or the Bonin Islands, and those that winter in a breeding ground return to their own feeding grounds A suckling calf accompanies its mother to her feeding ground in the first summer and is usually weaned there Calves thus learn their feeding ground from their mother and maintain the round trip for almost their whole life, but they are believed to have more

breeding population contains individuals from several feeding grounds, and a feeding population, whales from several breeding populations Currently, we do not have such an example in the toothed cetaceans, but Dall’s porpoises are known to have some seasonal shift of habitats (see above) and to form breeding aggregation in the summer as described in the following

We know that the distribution pattern of pigmentation types has revealed the presence of three Dall’s porpoise populations around Japan, each of which has its own breeding ground (Figure 98) and slightly different breeding season (Section 945) Several additional breeding grounds of the species are known to the east of the three populations where only dalli-types occur and are interpreted as representing local populations

International cooperation started in the 1970s for studies on the management of Dall’s porpoises under the framework of the INPFC (Sections 62 and 933) One of the research activities included collecting biological data from porpoises that were killed in the Japanese salmon drift-net fishery and brought to the mother ships in the summer, but scientists suspected that age structure and reproductive status obtained from such material could be biased To address this question, scientists conducted several annual cruises beginning in the summer of 1982 for sighting and hand-harpoon sampling in waters outside the drift-net fishery Scientists on the first cruise were myself, a student Y Fujise from Japan, a Canadian student and two technical assistants, a hand-harpoon fisherman hired from the Sanriku coast, and a whale spotter hired from a whaling company Participation of a US scientist was canceled before the cruise due apparently to domestic criticism of killing animals for scientific purposes The first cruise surveyed waters between the western Aleutian Islands (west of 174°E) and the southern limit of the Dall’s porpoise range excluding the EEZs of the United States and USSR The cruise was on the Hoyo-maru No. 12 of the Hoyo Suisan Co Ltd chartered by the Fisheries Agency of Japan It was an extremely enjoyable and memorable cruise for me, a university scientist without any previous knowledge on the customs or system of such chartered cruises, but later I realized that it was an extremely low-cost cruise at least partly supported by the crew For example, the bathtub was always full of warm seawater from the engine, which was nice, but the supply of freshwater for rinsing the body was ice cold, which helped save freshwater We Japanese scientists were happy if we were able to catch one Dall’s porpoise before breakfast, because it offered the only fresh dish available on the vessel, of sashimi (uncooked sliced pieces) of skeletal muscle and blubber The staple food items on the vessel were cooked rice, salted vegetables, salted uncooked saury or chunks of tuna, dry snacks, and miso soup with seaweed that never left the cooking oven during the cruise We scientists thought that the supply of stored food was short, but on returning to Kesen-numa (38°55′N, 141°35′E) after a month, we witnessed unloading of large amounts of frozen food including pork and chickens After returning to my office, I received a bill for meals on the vessel, which was so big that I suspected it covered the salary

However, I heard that some people paid the bill

In this environment our research team worked hard, along predetermined track lines We approached every cetacean sighting to confirm species and school composition We sighted 710 Dall’s porpoises and took 80 of them by hand harpoon equipped with an electric shocker (50 V) (Figure 21) Our strategy in the operation was to chase the porpoises windward while harpooning them one by one and then retrieve the carcasses while running leeward The electric shocker was sometimes ineffective; the porpoises resumed swimming with the harpoon line and buoy The electric shocker often caused a fracture of the vertebrae through muscle convulsions, which also destroyed the muscle tissue and rendered it unsuitable for sashimi The porpoises could have drowned rather than been killed directly by electric shock

This cruise was in a season of peak sea surface temperatures in August and September and was able to confirm the southern limit of Dall’s porpoises in summer at around 42°N in the offshore waters in longitudes 155°E-175°E Dall’s porpoises occurred in water temperatures below 20°C, while warmwater species such as the short-beaked common dolphin, striped dolphin, common bottlenose dolphin, and short-finned pilot whale occurred in water temperatures above 17°C The Pacific white-sided and right whale dolphins occurred in the intermediate temperatures of 12°C-19°C Combining these data with those obtained from subsequent cruises for the similar purpose, Miyashita (1993) determined that the last two species had the same temperature preference at 11°C-25°C and are distributed in latitudes of 40°N-50°N of the offshore North Pacific and estimated their abundance

During the cruise, we noticed a distinct geographical difference in the behavior of dalli-type Dall’s porpoises as detailed in Section 96 The difference was felt to be apparently related with surface water temperature Most individuals in waters above 11°C approached the research vessel and rode the bow wave, but most below that temperature were not attracted to the vessel but avoided it The area where we encountered the latter type of behavior agreed with the area in which we identified mother-calf pairs, which also avoided the vessel This means that Dall’s porpoises segregate in summer by reproductive or growth stages (Kasuya and Jones 1984) As their mating season comes soon after the parturition season (Sections 943 and 944), the area of low water temperature inhabited by mother-calf pairs as well as other presumably adult individuals could also be an area of mating and can be called the breeding ground The breeding ground identified during this cruise ranged from 45°N to 50°N and from 167°E to 175°N, but its northern and eastern limits were not confirmed

Kasuya and Ogi (1987), combining data from cruises of similar purpose in 1982, 1983, and 1985, identified a total of three areas inhabited by dalli-type mother-calf pairs: the central Bering Sea, south of the Kamchatka Peninsula, and south of the western Aleutian Islands The cruise of 1984 operated in May and June, or before the parturition season, and did not collect information on the distribution of mother-calf pairs Sea surface temperature profiles of these three breeding

grounds were different from each other, which made it clear that it is incorrect to conclude that a particular surface water temperature determines the location of breeding grounds of the Dall’s porpoise

Yoshioka and Kasuya (1991, in Japanese) made a further effort to collect locations of mother-calf pairs and identified eight breeding grounds in the North Pacific and adjacent seas: (1) the central Okhotsk Sea, (2) the southern Okhotsk Sea, (3) the northern Okhotsk Sea, (4) the central Bering Sea, (5) south of the Kamchatka Peninsula, (6) south of the Aleutian Islands, (7) the central Gulf of Alaska, and (8) coastal waters of North America near Vancouver Island (Figure 910) These areas of mother-calf pairs may shift annually, but they are apparently isolated from each other Breeding ground (1) is used by truei-types, (2) by the Sea of Japan-southern Okhotsk Sea population of dalli-types, and (3) by the northern Okhotsk Sea population of dalli-types The remaining five breeding grounds (nos 4-8) are used by dalli-types and are believed to represent local populations We still do not know where the wintering grounds are for the six populations that breed in grounds (3) to (8)

Now, we consider further details on why an area inhabited by mother-calf pairs can be considered a place of mating as well as parturition Ambiguity remains about the breeding cycle of Dall’s porpoises because of the difficulty of obtaining year-round information Ferrero and Walker (1999) reported that for Dall’s porpoises around the Aleutian Islands females mate for conception soon after parturition that occurs in early June to early August, which means that most females conceive annually This leads to an assumption of 10-11 months gestation Nursing lasts over 2 months and less than 5 months on average (see Section 946) It is difficult to assume that parturition and mating occur in different locations; rather it would be reasonable to assume that pregnant females and adult males of a population gather in a particular area for parturition and subsequent mating This is the reason why the area of mother-calf pairs is considered a breeding ground

represents a particular population, it is necessary to at least show that each dalli-type Dall’s porpoise breeds in a particular breeding ground There is no direct evidence for this, but I feel it likely because there are geographical differences in body size (see Section 932), parasite load (Section 935), pollutant level (Section 934), and genetics (Section 936) Additional support is available from truei-type Dall’s porpoises, which adhere to their own breeding ground in the central Okhotsk Sea located between two breeding grounds of dalli-types; mother-calf pairs of the two dalli-type groups tend to avoid the truei-type area (Figures 92 and 93) Similar behavior can be assumed for dalli-types in the Pacific and in the Bering Sea

If a low level of mixing does occur between the breeding grounds, it will be difficult to detect genetic differences between two neighboring breeding groups even if most of the members adhere to their original breeding ground However, they should be managed as independent populations, as long as the process of population dynamics remains almost independent and the effect of immigration is smaller than anthropogenic effect on the population

Subramanian et  al. (1986) analyzed geographical variation in PCB and DDE in Dall’s porpoises DDE is DDT and its derivatives Evaluation of their concentration levels is complicated, because it is a function of total intake, dilution through growth, decomposition in the body, and excretion The total intake depends on pollutant level in the food and amount of food consumption, dilution occurs in growing young animals, and decomposition and excretion depend on concentration level in the body In addition, adult females can off-load these pollutants to their offspring through pregnancy and subsequent lactation In order to avoid the complications of such effects, Subramanian et al. (1986) selected for their analysis only adult males taken during 1979-1985 They obtained the following results (concentration is expressed as ppm/wet weight of blubber)

They showed that the Bering Sea samples (n = 5) had PCB levels of 291-600 and DDE levels of 413-997, which were lower than for other samples Offshore specimens (n = 15) from south of the western Aleutian Islands and east of the Kamchatka Peninsula had similar pollutant levels (PCB 705160, DDE 664-152), but these values were slightly higher than those for the Bering Sea samples The most interesting was the pollutant load of 8 truei-types from off the Sanriku coast and 3 dalli-types taken off the Pacific coast of southern Hokkaido between Tsugaru Strait (41°30′N, 141°E) and the Erimo Peninsula (41°50′N, 143°E) Their pollutant levels, PCB 112-226 and DDE 124-368, were close to figures for one sample from the Sea of Japan (PCB126, DDE 324)

Subramanian et al. (1986) also looked at the ratio of PCB/ DDE, which was calculated for each individual and then averaged for geographical areas, and obtained the following figures: Bering Sea, 060; offshore western North Pacific, 096; truei-types off Sanriku, 091 (individual variation 071-115);

ation 037-041); and dalli-types in the Sea of Japan, 039 They found that the PCB/DDE ratio was low for the latter two groups, southern Hokkaido and the Sea of Japan The low figure for the Sea of Japan specimen could be a reflection of high DDE contamination of the sea due to continuing discharge of DDT in China (Tatsukawa et al 1979)

Subramanian et al. (1986) proposed two possible explanations for the low PCB/DDE ratios of three specimens from the Pacific coast of Hokkaido The Tsugaru Current runs through Tsugaru Strait to the Pacific, after being diverted from the Tsushima Current in the eastern Sea of Japan They thought that this current might reach the southern Hokkaido coast and could have polluted Dall’s porpoises living there or that dallitypes in the Sea of Japan might have migrated to the southern coast of Hokkaido We do not know what the pollutant levels are of seawater or of prey species consumed by the porpoises in this area The Tsugaru Current turns southward soon after passing Tsugaru Strait and runs along the Sanriku coast Therefore, if local seawater pollution is affecting the PCB/ DDE level of seasonally migrating Dall’s porpoises, the effect could also appear for the truei-type individuals that winter off the Sanriku coast, but this was not observed in the data (Subramanian et al. 1986), so I am of the opinion that the latter interpretation is correct

O’shea et al. (1980) reported extremely high pollutant levels in a dalli-type Dall’s porpoise off California: PCB 94 ppm and DDE 256 ppm These figures represent an extreme case of animals in locally polluted waters

Species of the genera Phyllobothrium and Monorygma are tapeworms that use cetaceans as an intermediary host Their larvae are found in the blubber as a cyst of about soybean size They are particularly numerous in the blubber in the inguinal region (Dailey and Brownell 1972) Their life history is incompletely known, but sharks are presumably one of their final hosts, receiving larvae through consuming carcasses of dolphins and porpoises (Walker 2001) Nothing is known about their route from the final host to cetaceans Experiments at sea have revealed that sharks prefer to eat the inguinal region of a dead porpoise Thus, concentration of the larvae in the region is considered as an adaptation by the parasite The larvae can survive in the blubber for at least 13 years waiting for a chance to be eaten by sharks Therefore, if other conditions are the same, the density of the parasite larvae in a porpoise correlates to some degree with the age of the intermediary host

Walker (2001) investigated the geographical difference in loads of Phyllobothrium cysts in dalli-type Dall’s porpoises, most of which were taken by the Japanese salmon drift-net fishery from June to July in 1983-1986 The sample size was 432 porpoises from the southwestern Bering Sea and 1957 from the western North Pacific adjacent to the Aleutian Islands in 46°N-59°N and 168°E-180°E (Figure 911) Additional small series (56 individuals) were obtained from the catch of hand-harpoon fishing by a small-type whaler

in July and August, 1988, in the southern Okhotsk Sea He removed a piece of blubber from the lateral side centered at the genital area, 35  cm in anterior-posterior length and 175 cm in midventral to lateral width, sliced it, and counted the number of cysts

Most of the drift-net specimens of Dall’s porpoise ranged from 160 to 200 cm in body length, but most of the Okhotsk Sea hand-harpoon samples were in the range of 190-230 cm with the modal length range of 180-220  cm The drift-net sample lacked individuals of 220-229 cm, and the hand-harpoon sample from the Okhotsk Sea had no small individuals below 180 cm, which can be explained mostly by gear selection and to some degree by geographical difference in growth (see Sections 932 and 962) The body-length compositions of drift-net samples were the same between north and south of the Aleutian Islands Approximate body lengths by age in the species are the following: 85-120 cm in the first year,

135-165 cm in the second year, and 140-175 cm in the third year Dall’s porpoises over 160  cm are mostly 2  years old or older, and those below 160 cm are mostly 4 years old or younger (Ferrero and Walker 1999) Walker excluded porpoises below 160  cm from the subsequent analysis because cysts rarely occurred in them

The proportion of Dall’s porpoises with cysts was zero in the Okhotsk Sea, 14% in the Bering Sea and Aleutian Islands area, and 227% in the Pacific Ocean The infection rate increased with body length, suggesting correlation with age The parasite load also increased from north to south: 05% (2/378) in the Bering Sea (north of 54°N), 108% (18/167) in the Aleutian Islands area in 52°N-54°N, and the highest figure 227% in the Pacific (south of 52°N) (Figure 911) This latitudinal cline suggests overlap of two populations in the Aleutian Islands area, which agreed with results of isozyme analysis by Winans and Jones (1988) mentioned in the

tiple years or inaccuracy of catch positions, which we are not given, can exaggerate the real level of geographical overlap of two populations in a single year Walker (2001) speculated that while not denying additional contribution of other intermediary hosts, the distribution pattern of sharks would be behind the geographical difference in parasite load He noted that only one species of shark Somniosus pacificus is known from the Bering Sea, Sea of Japan, and Okhotsk Sea as a species that attacks or scavenges dolphins and porpoises and that the shark species inhabits more coastal waters than the Dall’s porpoise and in low density However, there are more predatory sharks in the offshore waters south of the Aleutian Islands, including Isurus oxyrinchus and Prionace glauca, which are abundant in latitudes 20°N-50°N and surface water temperature of 7°C-16°C Other species of sharks abundant in warmer waters also certainly host the parasite, and dolphins in such waters are known to have high load of the cysts

The findings indicate that Dall’s porpoises that summer in the southern Okhotsk Sea or in the Bering Sea spend every summer there and do not move, at least during the summer, to the Pacific

DNA exists mostly within the nucleus, but a smaller amount resides in the mitochondria of animal cells, maintains genetic information, and functions to synthesize various proteins including enzymes It contains numerous minor variations that have no or limited effects on function, which are used for various purposes including individual identification, population structure, and taxonomy Variations in DNA may appear as minor structural variations of the enzymes produced, called isozymes Before the technology of studying DNA become available, isozymes were used as a mean of detecting genetic variation

To illustrate the usefulness and limitations of mtDNA for population study of Dall’s porpoises, I will start with a study by Hayano et al. (2003) on the species around Japan, which compares population structure viewed through mtDNA and from pigmentation pattern Two populations of Dall’s porpoises have been identified in the waters around Japan, a truei-type population ranging from the Pacific coast of northern Japan to the central Okhotsk Sea and a dalli-type population in the Sea of Japan-southern Okhotsk Sea area The latter is distinguished from the rest of the dalli-type populations by a smaller flank patch The Pacific waters of northern Japan are a major wintering ground of the truei-type population and also used by a few members of the Sea of Japan-Okhotsk Sea dalli-type population as well as dalli-types from other population(s) (Figure 98) Hayano et al. (2003) analyzed the mtDNA of a total of 103 specimens including 35 truei-types taken by the hand-harpoon fishery of the Sanriku Region (believed to have been taken in winter) on the Pacific coast of northern Japan, 35 dalli-types with small flank patch taken by the same fishery (presumably during late spring to early summer) off the west coast of Hokkaido in the northern Sea of Japan, and 33 dalli-types of

(see Section 933) in the western North Pacific south of the Kamchatka Peninsula (42°N-47°N, 155°E-162°E) Compared with the first two samples, evidence seems to be weaker for the last sample to be represented by a single population

Hayano et al (2003) identified a total of 49 mtDNA haplotypes or 20-24 for each sample Such large genetic variation found in a small sample is evidence of large genetic variation within the population and suggests the population was large in the past The number of haplotypes unique to each sample (represented by a total of 40 individuals) was also similar among the three samples: 14 for the truei-type sample, 12 for the dallitype sample from south of the Kamchatka Peninsula, and 11 for the dalli-type sample from the Sea of Japan A neighbor-joining dendrogram and minimum spanning network of the 49 haplotypes revealed two major clusters and some additional minor ones Members of the three samples were scattered in these clusters, and the haplotypes unique to each sample were also found intermingled in various clusters and did not constitute clusters that represented particular samples A similar result was reached by Escorza-Treviño et al (2004) The authors concluded that separation of these populations occurred rather recently or after the origin of the observed genetic polymorphisms

Hayano et al (2003) carried out an analysis of molecular variance to measure genetic differentiation among the three samples, which took account of genetic distance between haplotypes and frequency of haplotypes in each sample The results indicated some difference among the three samples They found significant difference between the Sea of Japan dalli-type sample and the remaining two samples but no significant difference between the truei-type sample and the southof-Kamchatka dalli-type sample This does not necessarily imply that the latter two samples were identical in haplotype composition Such a result can be reached due to insufficient sample size, insufficient time after separation of two populations, or continuing low level of genetic flow Hayano et al. (2003) thought from observation of the geographical distribution of pigmentation patterns that current genetic flow among the populations in Japanese waters is restricted, which is not counter to the conclusion of Escorza-Treviño and Dizon (2000) It seems to me that the observed mtDNA variation has nothing to do with hereditary or behavioral mechanisms that maintain particular pigmentation patterns within the populations; this remains as an important question to be answered for understanding the population structure of the species

One of the interesting points in the results is the abundance of haplotypes common among samples: 3 haplotypes (represented by 18 individuals) were common among the 3 samples and another 11 haplotypes (represented by 45 individuals) were common in 2 samples The number of individuals representing the common haplotypes, 4-6 individuals/haplotype, is greater than the corresponding figure 11 for 42 individuals representing the 37 unique haplotypes Common haplotypes are in the majority in the three samples (populations) and unique haplotypes are in the minority At least some, not necessarily all, of the latter are likely to have evolved after the populations separated

ecological mechanisms that may limit gene flow between neighboring populations Dall’s porpoises segregate by growth and reproductive stage within each population, and the breeding seasons also differ between neighboring populations (Section 945) These factors will function to diminish opportunities for interbreeding between populations from what might be expected from their overlapping ranges of distribution However, such behavioral barriers do not seem to be strong enough to explain the apparent genetic isolation between the two major pigmentation types of the Dall’s porpoise If the white flank patch of the Dall’s porpoise has a social function, such as identifying mating partners, about which we know nothing, the pigmentation pattern will function to suppress interbreeding between populations, and genes imported through limited interbreeding will be less likely to survive in the population

Hayano et  al. (2003) estimated separation time for the Sea of Japan-southern Okhotsk Sea dalli-types and southof-Kamchatka dalli-types at 30,000-40,000 years before present and that between the former and truei-types at 10,000-15,000 Their genetic analyses could not distinguish between the south-of-Kamchatka dalli-type sample and the truei-type sample from off Sanriku These estimates depend on an assumption of mutation rate, and a low level of gene flow after separation of the populations will bias downward the estimate of time after separation The latter figure, 10,000-15,000 years before present, agrees with the end of the last glacial age and start of the current warm-climate era During the glacial period, the Sea of Japan lost direct connection with the Okhotsk Sea, but it is uncertain whether the Sea of Japan was connected to the Pacific through Tsugaru Strait or to the East China Sea through Tsushima Strait Hayano et al. (2003) did not reach conclusions on the process of formation of the two populations wintering off the east and west coasts of Japan

The technique employed by Hayanao et al. (2003) is used widely in studies of population structure of cetaceans and is perhaps the most relied upon in the field However, their results provide a valuable example of the limitation of the methodology If a founder effect is ignored, we may have to wait at least several thousand years before sufficient genetic differentiation is accumulated between newly formed populations Our damage to the marine environment can proceed rather fast; populations may decline at an annual rate of several percent Such rapid decline in abundance can result in reduction of the population range and subsequent split of the distribution into more than one geographical area Conservation will need to handle these newly formed geographical populations of almost identical genetic structure as independent populations Thus, the absence of genetic evidence for population structure should not be considered as an indication of no population structure for purposes of management

Winans and Jones (1988) and Escorza-Treviño and Dizon (2000) studied the genetics of dalli-type Dall’s porpoises in the vast area of the offshore North Pacific and adjacent seas not covered by Hayano et al. (2003) Information on breeding

ence of several populations of Dall’s porpoises in this region Winans and Jones (1988) analyzed isozymes of 360 dallitype Dall’s porpoises killed in the Japanese salmon drift-net fishery from June to September 1962 The sample came from the area indicated in Figure 911 plus the southern area surrounded by 42°N-47°N, 158°E-170°E They analyzed 26 loci that controlled 14 enzymes and found polymorphisms at 11 loci The fit to Hardy-Weinberg equilibrium was better if the sample was split into two geographical groups south and north of the Aleutian Islands This suggested that the sample was not from a uniform population They determined that this result was influenced by specimens from the Aleutian Islands area at 50°N-52°N and concluded that the ranges of two populations, to the north and south of the Aleutian Islands, overlapped in those latitudes Their results coincided exactly with those obtained from parasite load analysis by Walker (2001)

Escorza-Treviño and Dizon (2000) used mtDNA to examine the population structure of dalli-type Dall’s porpoises in the North Pacific and adjacent seas from the Okhotsk Sea to the coast of North America Their results are briefly presented in the following using the geographical terminology of this chapter They identified 58 haplotypes and compared their frequencies between the three geographical regions of (1) the Okhotsk Sea, (2) the Bering Sea and central and western North Pacific, and (3) off the coast of North America The commonest two haplotypes occurred in all three areas, but minor haplotypes, which were thought by them to be derived from the common haplotypes, occurred only in one or two of the three geographical areas Later, Hayano et al. (2003) made a similar interpretation that common haplotypes were closer to the ancestral type The common haplotypes were more frequent in the western area than in the east, which was considered by Escorza-Treviño and Dizon to suggest possible expansion of distribution from west to east As expected, haplotype composition was not uniform in the three areas

They further divided the sample into smaller geographical areas and found that haplotype structure was different among seven areas, suggesting the presence of different populations: (1) western Bering Sea, (2) eastern Bering Sea, (3) around the central Aleutian Islands, (4) south of the Kamchatka Peninsula, (5) south of the Aleutian Islands, (6) off the coast of North America, and (7) Okhotsk Sea The genetic difference between the western Bering Sea and the eastern Bering Sea was inconclusive (p = 0050) Evaluation of such difference should be done in reference to other biological information Out of the these seven putative populations, the two populations south of the Kamchatka Peninsula and south of the Aleutian Islands apparently match the south-of-Kamchatka population and south-of-western Aleutian Islands populations of Kasuya and Ogi (1987) and Yosjhioka and Kasuya (1991), respectively

Escorza-Treviño and Dizon (2000) did not find significant genetic difference between dalli-types in the northern Okhotsk Sea and those in the southern Okhotsk Sea Amano and Hayano (2007) did not examine the flank patch of dallitypes inhabiting the northern Okhotsk Sea in summer, and

morphologically separable or not However, distribution of pigmentation types and of mother and calf pairs (Figures 91 and 92) is sufficient to justify an assumption of three populations in the Okhotsk Sea in summer

Some questions still remain on the population structure of Dall’s porpoises in the Bering Sea Escorza-Treviño and Dizon (2000) thought that there were three populations: (1) western Bering Sea, (2) eastern Bering Sea, and (3) around the central Aleutian Islands The relationship between the third population and individuals that breed in the central Bering Sea (Kasuya and Ogi 1987) is unclear, because Kasuya and Ogi (1987) did not survey the Russian EEZ (western Bering Sea) and US EEZ (Aleutian Islands area and eastern Bering Sea), and Escorza-Treviño and Dizon (2000) did not have material from the central Bering Sea If we assume they are all different, there can be four populations in the Bering Sea and central Aleutian Islands area, but it is also likely that the 3rd population breeds in the central Bering Sea (three populations in this case) The question needs to be resolved EscorzaTreviño and Dizon (2000) did not include in their analysis porpoises from the central Gulf of Alaska where Yoshioka and Kasuya (1991) recorded a breeding ground of the species, thus it seems to be reasonable to expect a population in the central Gulf of Alaska

Escorza-Treviño and Dizon (2000) presented another interesting result on sexual difference in dispersal pattern of the Dall’s porpoise They found that geographical difference in haplotypes was greater among females than males, which suggests that males move greater distances, perhaps into the range of other nearby populations However, as inheritance of mtDNA is matrilineal, the analyses do not tell whether such traveling males breed with local females The answer will be available through analysis of nuclear DNA

The small-cetacean subcommittee of the Scientific Committee of the IWC concluded in 2001 that there are at least 11 populations of the Dall’s porpoise in the North Pacific (IWC 2002) (Figure 912): (1) a population that breeds in the northern Okhotsk Sea, (2) a truei-type population that breeds in the central Okhotsk Sea and winters off the Pacific coast of northern Japan, (3) a population that breeds in the southern Okhotsk Sea and winters in the Sea of Japan, (4) a population in the northwestern North Pacific that breeds south of the Kamchatka Peninsula, (5) a population in the western Bering Sea, (6) a population in the eastern Bering Sea, (7) a population in the western Aleutian Islands area, (8) a population in the central North Pacific that breeds south of the Aleutian Islands, (9) a population that breeds in the central Gulf of Alaska, (10) a population that breeds along the Oregon coast, and (11) a population along the California coast

Among the 11 Dall’s porpoise populations, only the second is represented by truei-types; other populations are represented by dalli-types Both wintering ground and breeding

ground are known only for the 2nd and 3rd populations The relationship between the 7th population and individuals that breed in the Central Bering Sea needs further study It should be noted that the western Aleutian Islands area was shown as an area of mixing of populations

This conclusion mostly depends on information from pigmentation pattern, mtDNA, and breeding grounds, and its weak point is the absence of linkage of the latter two methodologies It will be useful for further improvement of our understanding of population structure of the Dall’s porpoise to survey distribution of mother-calf pairs in the whole range of the species, including the EEZs of the range countries, and to collect genetic samples from individuals on the breeding grounds, although the tasks will experience some difficulties due to ship-avoiding behavior in the breeding ground Analysis of nuclear DNA will be useful for the purpose of understanding possible gene flow and the mechanism of genetic isolation between populations

Age is key information for understanding the life history of animals There have been attempts to age Dall’s porpoises taken by the hand-harpoon fishery off Japan and those killed incidentally in the offshore salmon drift-net fishery by reading growth layers in their teeth The teeth of Dall’s porpoise are small, with length of about 1 cm and diameter of 1-2 mm In adults the crown is often lost through abrasion, and the remaining portion of the tooth is hidden between horny ridges of the gum The teeth are evolutionarily in the process of degeneration; the function of grasping prey seems to be achieved with the horny surface of the gum This was interpreted by Miller (1929) to be similar to the process in which ancestral baleen whales evolved baleen plates

Growth layers are deposited in both the dentine and cementum of dolphins and porpoises, but the layers in dentine are thicker and more regular in spacing and are more suitable

tinal layers are accumulated on the pulp wall, the volume of the cavity decreases and dentine deposition finally ceases in most toothed cetaceans, leaving a small space for blood vessels This stage usually occurs at around the attainment of sexual maturity or soon after when the body growth slows in dolphins and porpoises After this stage, age has to be determined by reading layers in the cementum The teeth of Dall’s porpoise are probably the most unsuitable for age determination I have ever experienced Dentine deposition stops at an early age, presumably reflecting the small tooth size or early attainment of physical maturity, and the cemental layers to be relied upon thereafter are irregular These factors raise a question on the reliability of ages thus determined Some attempts in the 1970s to use growth layers in the mandible and tympanic bones were unsuccessful, because older layers in the bone tissue are resorbed while new layers are accumulated (Shirakihara and Kasuya unpublished)

In 1984, several scientists attempted to compare their readings on a set of decalcified and hematoxylin-stained tooth sections of Dall’s porpoises The materials were prepared by some of the participants using materials obtained either by US scientists on board Japanese vessels in the salmon drift-net fishery or by Japanese scientists on board the scientific hunting vessels operated in the offshore North Pacific (see Section 933) The season was summer, or around the parturition season of the species Readers could not refer to the biological data but were able to identify the origin of the specimen from the specimen number

Results of the cross reading were reported to the science section of the INPFC meeting in March of the next year (Jones et al. 1985a) but have remained otherwise unpublished Believing that the experiment is useful for understanding the difficulty of age determination in the species, I have extracted the readings by myself and two anonymous scientists and compared them in Tables 93 through 95 Table 93 compares average ages by body length (sexes combined) Interreader difference in age estimated was unclear for porpoises over 180 cm in body length, which did not mean that interreader difference was negligible There were greater interreader differences for body lengths of juveniles of 140-159  cm

the other two readers thought them to be about 3 years old Porpoises at 160-169 cm were aged at 2-3 years by one reader and 3-4 years or 4 years by the other readers

In order to examine further details of this disagreement, estimated ages are compared between readers in Table 94, where only the results for female porpoises are given If two readers agreed in their reading, the point should fall on the diagonal line This did not often happen; the points are scattered over a rather broad range The only features common among readers were that (1) larger individuals tended to be aged older and (2) all readings were within the range of 0-13  years Anonymous readers B and C showed better agreement in the readings than comparison between them and reader A Reader A aged 21 specimens at 1 year, but only 1-2 specimens were aged at 1 year by readers B and C, which suggests that readers B and C counted finer layers compared with reader A, who picked up rather coarse layers

The next analysis I made was to calculate average body length for each age group in the female sample (Table 95) No correlation was found between age and average body length for ages 1-4  years as determined by reader B, which casts some doubt on the age reading Reader C detected only about 5 cm growth between ages 2 and 3 years; however, it was not accompanied by a reasonable increase in body length below and above this interval Compared with these readings, the results by reader A gave a better correlation between age and body length, but this alone does not prove that reader A’s ages were correct, only that they were correlated with length

We know from other sources that Dall’s porpoises are born at around 100 cm body length, but such newborns were not included in the cross-reading samples Information on growth during the first year after birth would be a key to evaluate the age readings discussed here Newborns of several dolphins and porpoises are known to reach 155%–163% of neonatal length in the first year after birth (Section 8524; Kasuya and Matsui 1984; Kasuya et al. 1986) Thus, the average body length of 1595 cm obtained by reader A for animals aged at 1 year seems to be reasonable If we accept this age reading, then it leads to the conclusion that 1-year-old porpoises were taken in the salmon drift-net fishery in the Bering Sea and Aleutian Islands area and 2-year-olds in the waters south of the Aleutian Islands area of 40°N-45°N where Japanese scientific cruises sampled porpoises with hand harpoons This further led to the hypothesis of Kasuya and Jones (1984) and Kasuya and Shiraga (1985) that Dall’s porpoises segregate by reproductive status as well as growth stage More importantly, there was general agreement among the INPFC scientists that even if aging by particular readers were accepted, age determination of Dall’s porpoises at ages over 6 years would be extremely unreliable

One of the reasons why the results of the cross-reading experiment were only reviewed at the NPFC meeting and not followed by action would be the absence of firm evidence for choosing one of the several readers who participated Another reason was the disappointment of the scientists who wanted to use age readings for the analyses of population dynamics

TABLE 9.3 Among Readers Disagreement in Age Determination of Dall’s Porpoises Experienced in the 1980s

If we were unable to reliably age animals over 6 years, the age data would be almost useless for such purposes

Ferrero and Walker (1999) recently published a biological study of Dall’s porpoises using large samples from the salmon drift-net fishery collected through the activities of the INPFC

This was very welcome because the only study previously available on the subject was a thesis by Newby (1982) The method of tooth preparation seems to have been identical to that in previous studies, and the analysis is a useful interpretation of the species’ life history

Cross Reading of Growth Layers in Dall’s Porpoise Teeth, Illustrating Trouble Experienced by Scientists in the 1980s

The body-length composition of a sample of Dall’s porpoises is influenced by sampling season and gear selectivity, but it still offers useful information on behavior and geographical segregation within a population This source of information is free from the problems attendant on age determinations as described earlier

Ferrero and Walker (1999) presented the body-length composition of Dall’s porpoises taken in the Japanese salmon drift-net fishery in the Pacific and Aleutian Islands area of 46°N-53°N, 170°E-175°E during the months of June and July in 1981-1987 (Figure 913) It had two major modes One was at body length 84-120 cm with a peak around 105 cm for both sexes The other mode was at 132-220 cm, where length frequencies differed between sexes There was a female peak at 180-187 cm and two less distinct male peaks at around 158 and 195  cm The right-hand peaks, 185  cm for females and 195 cm for males, are close to average body length at which growth ceases (see Section 9410) The difference between the sexes of 10 cm is close to that between asymptotic body lengths, which is about 12 cm These features agree with those identified by Newby (1982) Because the sample was obtained during the parturition season, the smaller individuals constituting the 105 cm peak represent newborns of the year The hiatus around 115-140 cm represents 1 year’s growth

Kasuya and Shiraga (1985) reported body-length composition of Dall’s porpoises taken during sighting and handharpooning cruises that covered latitudes 40°N-50°N (most of the samples were obtained in latitudes 42°N-47°N) (Figure 921), which was south of the area where Ferrero and Walker (1999) and Newby (1982) obtained their samples The collection period for the hand-harpoon sample, August and September, was about 2  months behind that of the salmon drift-net sample, but the difference was not considered significant in the analyses The body-length composition in Figure 914 lacks calves of the year Mother-calf pairs were sighted in the northern part of the area covered by the cruise, which was north of 47°N and close to the salmon drift-net ground It was not possible to harpoon the mother-calf pairs, as well as other individuals in the area due to strong shipavoiding behavior, which contrasted with the reverse behavior of individuals in the south (Kasuya and Jones 1984) Figure 914 compares body-length composition of Dall’s porpoises in the breeding ground in the western Aleutian Islands area (ie, salmon drift-net ground) with that south of the breeding ground (where harpooning was carried out) Similar ship-avoiding behavior by adult individuals has been noted in striped dolphins through the comparison of catches by drive and hand-harpoon fisheries off Taiji (33°36′N, 135°57′E), that is, the latter fishery took mostly weaned immature individuals of both sexes (Kasuya 1978)

Another important point to be noted about body-length composition of the hand-harpoon sample from the southern part of the species’ range is a body-length peak at around 165-190 cm in females and 170-195 cm in males, which corresponds to a hiatus in the body-length composition of the drift-net samples of Newby (1982) and Ferrero and Walker (1999) Females over 190  cm and males over 195  cm were less frequent in the hand-harpoon sample than in the drift-net

Among Readers Comparison of Early Growth of Dall’s Porpoises Estimated by Reading Growth Layers in the Teeth

sample (Figures 913 and 914) Combining the two sets of samples better documents the body-length composition of the population

These observations indicate that Dall’s porpoises tend to segregate by sex and growth stage Inhabitants of the summer breeding ground are mostly calves of the year accompanying their lactating mothers, yearlings, and sexually mature males, while inhabitants of the area south of the breeding ground do not include mother-calf pairs but include numerous juveniles aged 2 years and above (not yet sexually mature) and some adult males and nonlactating adult females (see Section 946) This was supported by the fact that 88% of the 156 specimens obtained from the latter area (south of the breeding ground) by the dedicated sampling vessel were aged 2-5  years and there were only 2 yearlings and no young of the year (Kasuya and Shiraga 1985) The sex ratio in the southern sample was 24 males/female (Kasuya and Shiraga 1985), but the corresponding figure from the northern sample (in the breeding ground) was the reverse at 063 males/female (Ferrero and Walker 1999) (Figure 918) However, the higher female proportion in the breeding ground could be a reflection of females attaining sexual maturity about 1 year earlier than males

The earlier discussion on dalli-type samples is worth comparing with the body-length composition of truei-types off the Sanriku coast, which have been hunted in winter or about half a year after the dalli-type samples The body-length composition of 884 truei-types examined by Kasuya (1978) had a mode with a peak at 170-190 cm and was composed of individuals aged 15 years and above The sex ratio was 12 males/ female, which was similar to that in the earlier mentioned dalli-type sample obtained from waters south of the breeding

1144) landed at the Otsuchi Fish Market (39°21′N, 141°54′E) on the Sanriku coast during November to April in 1995-1996 and examined by Amano et al. (1998c) also showed a similar structure (Figure 915) The difference in the peak lengths, 6-7 cm, is a reflection of difference in growth between the sexes The sex ratio was 21 males/female (n = 1979) Dallitypes landed at the same market during the same period and examined by Amano et al. (1998c, in Japanese) had the same body-length composition and a similar sex ratio of 15 males/ female (n = 175) It seems to be true that within-population segregation by growth stage and sex observed in the summer is likely retained in the winter Amano et al. (1998c, in Japanese) examined Dall’s porpoises at the Otsuchi Fish Market nearly throughout the year during 1994-1996 and confirmed that landings from May to October were most likely to have been taken in the northern Sea of Japan and Okhotsk Sea but that landings in winter were from catches off the Sanriku coast as in the past operation examined by Kasuya (1978)

9.4.3.1 Neonatal Body Length Measuring the body length of apparent newborns is the most direct way of obtaining neonatal length, but such an opportunity is limited for cetacean biology and the method is likely to result in some bias due to the time between birth and measurement Another way is to identify a body length where the proportions of fetuses and neonates are equal in the sample This method is subject to a bias due to availability differences between fetuses and neonates

Ferrero and Walker (1999) used materials collected in June and July from the salmon drift-net fishery south of the western Aleutian Islands They classified the healing of the umbilicus into four stages and calculated the average body length for each stage The healing stages were (1) umbilicus open with fragment of umbilical cord attached, (2) umbilicus open but without fragment of umbilical cord, (3) progressed stage of healing, and (4) fully healed umbilicus They presented mean body length for each of these stages (no length distribution available): 990 cm at stage 1 (n = 102), 1027 cm at stage 2 (n = 80), 1106 cm at stage 3 (n = 31), and 1141 cm at stage 4 (n = 88) The body-length increment was about 15 cm over all the umbilical stages, but time between the stages was not available It seems to me that the first figure, 990 cm, is closest to the mean neonatal length in the population

Ferrero and Walker (1999) stated that no individuals in the four categories had deposited postnatal dentine in their teeth as observed in decalcified and hematoxylin-stained preparations A similar observation has been made on short-finned pilot whales and interpreted as follows (Kasuya and Matsui 1984) The boundary between fetal dentine and postnatal dentine is identified as the neonatal line or neonatal layer It is a dentinal layer of poor mineralization stained lightly with hematoxylin Identification of postnatal dentine in a tooth preparation needs to await formation of dark-stained dentine next to the lightly stained neonatal layer It takes some time for the deposition of

postnatal predentine to be followed by mineralization The lag time between birth and full mineralization of postnatal dentine would be sufficiently long for newborn Dall’s porpoises to attain a body length of over 114 cm

The second method used by Ferrero and Waker (1999) was to calculate a point where the proportions of fetuses and neonates were equal in their sample Fitting two different mathematical models to their data, they calculated mean neonatal length at 1011 cm with a 95% confidence interval of 1006-1016  cm and 1030  cm (two figures derived from the two models) They also attempted to calculate mean body

length of full-term fetuses, which underestimates true neonatal length It seems to be true that Dall’s porpoises in the western Aleutian Islands area are born at an average body length of about 100 cm

Before the study of Ferrero and Walker (1999), Mizue et  al. (1966, in Japanese) examined fetuses and neonates of Dall’s porpoises taken in the salmon drift-net fishery around the western Aleutian Islands and reported that body length at birth was between 92 and 109 cm, with an average of about 100  cm Kasuya (1978) added some supplementary data to those of Mizue et al. (1966, in Japanese) and obtained 100 cm

were equal Thus, there is good agreement among the published estimates of neonatal length of Dall’s porpoises

Adult body size of Dall’s porpoises differs between regions (Section 932) However, it is generally understood that correlation between mean adult size and mean neonatal size is weak among small cetaceans, which can be explained by an interspecies relationship between mean body length at attainment of sexual maturity (X, m) and mean neonatal length (Y, m):

Y = 0532X0916 Ohsumi (1966)

This suggests a 5 cm difference between two species with body lengths at sexual maturity of 2 and 21 m The difference in neonatal length will be smaller between two populations of one species such as Dall’s porpoise

9.4.3.2 Parturition Season As cetaceans usually deliver their young in waters remote from human habitats, recording seasonal frequency is often impractical Examined Dall’s porpoises were killed in hunting activities or incidental to other fishing activities that were seasonally limited Thus, materials from such activities did not offer sufficient information on parturition season

To get around this difficulty and utilize information from the salmon drift-net fishery that operated during the breeding season, Newby (1982) observed seasonal alternation of full-term fetuses, newborns, and lactating females in the incidentally taken samples in order to estimate the seasonality of reproduction His materials were collected in the US EEZ where the Japanese salmon drift-net fishery was licensed, that is, around the western Aleutian Islands, 46°N-59°N, 168°E-175°E, which was almost the same as the sampling area of Winans and Jones (1988) and Ferrero and Walker (1999) and overlapped with that of Mizue et al (1966, in Japanese) Newby (1982) obtained his material in the period from June 5 to July 25 of three seasons from 1978 to 1980 Since the United Sates-Japan agreement on the Dall’s porpoise project gave first priority to collection of carcasses within the US EEZ, sampling activity was less when the salmon fleets moved outside of the EEZ, including during the previously mentioned period

Newby (1982) noted that the proportion of pregnant females with full-term fetus decreased in July but that they still occurred through the sampling period, which indicated that the parturition season lasted until late July Postpartum females started to occur around June 10, the number became equal with that of near-term females around July 10, and then postpartum females exceeded pregnant females in number From this observation, he concluded that the average parturition date was around July 10 He further stated that the parturition peak was in the last 10 days period of July when he observed a rapid increase in number of newborns (data were grouped for each 10-day period) The difference between average parturition date and parturition peak is unclear to me His study could not identify the end of the parturition season

Ferrero and Walker (1999) analyzed a greater sample obtained in 1981-1987, or after Newby (1982) collected his

They were apparently able to retrieve most of the incidental takes without distinguishing between inside and outside of the US EEZ as experienced during earlier activities This sampling scheme improved the quality of the data as well as the quantity of materials Ferrero and Walker confirmed that almost all the sexually mature females were pregnant or early postpartum Postpartum females that were lactating started to appear during June 5-9, and pregnant females occurred until July 20-24 and were absent on July 25 and later A sigmoid curve fitted to frequency of postpartum individuals gave the proportion of such females at 5% on June 11, 50% on July 3, and 95% on July 24 This result indicated that the parturition season of Dall’s porpoises around the western Aleutian Islands area lasted for about 50 days from June 5 to July 24, with an average date as well as the peak date of parturition on July 3 This agrees with our general understanding that species in higher latitudes tend to have shorter breeding seasons than those in lower latitudes

The further analyses of lactating females by Ferrero and Walker (1999) are also worth mentioning While parturition started in the population on around June 5, females secreting colostrum were already present on June 2 when sampling started and continued to appear on July 24 when the last parturition was recorded The peak was during June 10-24, which was before the parturition peak on July 3 calculated from a model Secretion of colostrum starts before parturition and continues until shortly after parturition Colostrum can be identified through histological examination of mammary glands for the presence of colostrum bodies in the milk (Kasuya and Tai 1993), but it can reasonably be identified as colostrum if milk is secreted by a pregnant female with a nearterm fetus At least at the early stage of the United StatesJapan cooperative activities on the Dall’s porpoise, there was no instruction for the microscopic identification of colostrum

9.4.4.1 Mating Season The Japanese high-seas salmon drift-net fisheries operated from 1952 to 1992 with two modes: mother-ship (factory ship) fishery and land-based fishery The operation area changed several times by international agreement, and in the 1980s, mother ships were allowed in the area indicated in Figure 911 and the land-based fishery in an area further to the southwest (Sano 1998, in Japanese) The US government placed observers on Japanese mother ships operating in its EEZ to examine Dall’s porpoise carcasses brought in by the catcher boats The results revealed that most of the adult females were pregnant with a full-term fetus or just postpartum, indicating that the parturition season was during the fishing and sampling season of June and July (Composition of 1061 sexually mature females analyzed by Ferrero and Walker (1999) was 50% postpartum, 47% pregnant, and 3% resting)

Ferrero and Walker (1999) observed a seasonal change in Graafian follicles in the samples The diameter of the largest follicle in most porpoises was below 8 mm during the period

denly started to see females having follicles of 10-18  mm around July 20 and this continued until July 25, the last day of sampling This suggested that follicle size at ovulation was not below 10 mm and was around 10-18 mm and that estrus or the mating season started around July 20 The starting day of the mating season was about 45 days after the start of the parturition season (June 5) As the length of the mating season will be almost the same as that of the parturition season, estimated earlier at about 50  days, the mating season lasts approximately from July 20 to September 10

To further confirm the seasonality of mating, Ferrero and Walker (1999) examined seasonality in spermatogenesis in the testicular tissue The proportion of inactive males was 20%–30% during early June and declined to less than 5% in the second 10-day period of July, while the proportion of males with testis having very active spermatogenesis increased from 10% in early June to around 40% in the second and third 10-day periods of July The rest of the males were classified as being at a moderately active stage and constituted 50%–60% of the observed adult males This observation does not disagree with the mating season deduced earlier, because (1) it takes some time for spermatozoa in the seminiferous tubule to be transported to the epididymis and (2) it is biologically reasonable for some males to be ready for mating before the start of female estrus A biological explanation needs to be investigated for the 50%–60% of males that were classified as at intermediate activity during the mating season

9.4.4.2 Gestation Period Gestation in most cetacean species lasts close to 1 year, with a possible maximum of 15-17 months observed or estimated for killer and sperm whales (Kasuya 1995) The estrous season of Dall’s porpoises in the Aleutian Islands area started about 45 days after the start of the parturition season, which suggests that gestation lasts for 105 months

Kasuya (1978) examined the fetal growth of truei-types using 39 fetuses measuring 20-50 cm in January-March and obtained 33 mm as an estimate of average daily growth rate in the linear phase of fetal growth By extending this regression line to the right, I obtained a mean parturition date of August 28 when the extended fetal growth reached 100 cm, the estimated mean neonatal length The time between the mean parturition date and the date when left-side extension of the regression line crossed the axis of time was 300 days This estimate is downwardly biased because fetal growth in the early stage of gestation is slower Assuming the bias at 135% of the total gestation time for this species (Section 8522), I estimated gestation time of 114 months and mean conception date at September 17 for truei-types off Japan This figure is based on some assumptions and on a small and seasonally limited sample and is of low reliability If the gestation time of 105  months estimated by Ferrero and Waker (1999) is assumed for the truei-type population, peaks of parturition and conception would be expected in middle August and early October, respectively, by shifting both of the dates (August 28 and September 17) by about 05 month

length (X, cm) and gestation time (Y, month) was proposed by Perrin et al. (1977):

LogY = 01659 + 04586LogX

This equation and neonatal length of 100 cm suggest a gestation time of 121 months, which does not agree with the seasonality of reproduction observed earlier and appears to be an overestimate

Amano and Kuramochi (1992) thought that the Sea of Japansouthern Okhotsk Sea dalli-type population had a parturition season in May and June, different from that reported for the western Aleutian Islands area This conclusion was based on observations on board hand-harpoon vessels from June 15 to July 1, 1988, when the vessels encountered numerous mothercalf pairs in an area from Soya Strait to the southern Okhotsk Sea (44°N-46°N) but were unable to capture them because of their ship-avoiding behavior They also noted a high proportion of immature individuals, particularly among females, in the catch made in late May 1988 and 1989 by the same vessel in Tsugaru Strait and the nearby eastern Sea of Japan (41°N-43°′N), which suggests a north-south segregation by growth stage within the population The reproductive data they obtained in these cruises are in Table 96

TABLE 9.6 Reproductive Status of Females in the Sea of Japan-Southern Okhotsk Sea Dalli-Type Population

ported, as detailed in the following, by biological data for 31 dalli-types collected by myself and my colleague Aoki during July 13 to August 26, 1988, from the catch of the Yasu-maru No. 1, a small-type whaling vessel then operating in the handharpoon fishery for Dall’s porpoises in the southern Okhotsk Sea Reproductive data for the 31 carcasses were used in Yoshioka et al (1990) and are included in Table 96 I stayed on the vessel for 4  days in August and recorded numerous Dall’s porpoises that could not be approached by the vessel and thus were unable to be harpooned They were of adult size and did not include mother-calf pairs

Biological information on dalli-types caught in the southern Okhotsk Sea in June through August provides evidence of the breeding season, which is different from that of some other populations of the species (Table 96) The only pregnant female in June reported by Amano and Kuramochi (1992) had a full-term fetus of 96 cm, while 7 fetuses in July and 6 fetuses in August (Yoshioka et al. 1990; Kasuya unpublished) were 07-30  cm (mean: 20  cm) and 38-90  cm (mean:  66  cm) long, respectively The average fetal size increased 46  cm in the 1-month period This suggests only that the parturition season ends by June and the mating season has already started in July However, an earlier start of the mating season is suggested by the presence of females with a corpus luteum of ovulation (and without identifiable embryo) in the latter half of June The proportion of such sexually mature females was 94% in the latter half of June and 53% in July (Table 96) All of these females must have experienced recent estrus, and many of them could have been in an early stage of pregnancy Histological examination of endometrial tissue (see Section 1244) was not conducted for these females

This information allows the conclusion that the Sea of Japan-southern Okhotsk Sea dalli-type porpoises mate from mid-June to late July (a 50-day mating season in the Aleutian population is assumed without evidence) The mating season is about 15  months earlier than that for a population in the Aleutian Islands area (from July 20 to September 10) Although there are insufficient data to determine the parturition season of the Sea of Japan-Okhotsk Sea population, this mating season, the presence of numerous mother-calf pairs in middle and late June, and an assumption of the same gestation time as in the western Aleutian Islands area (105  months) suggest that the parturition season there starts in the middle of April and continues to June, although Amano and Kuramochi (1992) concluded that the population had parturitions in May and June

Kasuya (1978) fitted a fetal growth curve calculated for the truei-type off the Pacific coast of northern Japan to fetal records of dalli-types in the western Aleutian Islands area and found about a 1-month difference in the parturition peaks between the two populations, that is, July-August for the Aleutian sample and August-September for the trueitypes off Japan Although these absolute dates of parturition may be subject to some uncertainty due to the fetal growth curve having been fitted to small and seasonally limited samples, the resultant difference in mating seasons can be significant An improved estimation of breeding season also

(Sections 943 and 944), that is, dalli-types in the Aleutian Islands area had a parturition peak in early July and a conception peak in middle August, while truei-types off Japan had a parturition peak in mid-August and a conception peak in early October The two populations have breeding seasons that are about 1 month apart

Dalli-type porpoises of the Sea of Japan-southern Okhotsk Sea population probably give birth in the northern Sea of Japan off Hokkaido and southern Okhotsk Sea from mid-April to mid-June and mate during the period from mid-June to late July in the southern Okhotsk Sea (off northern Hokkaido), while truei-type porpoises mate presumably in September and October (with an estimated conception peak in early October) Although the mating grounds are located close to each other (southern Okhotsk Sea and central Okhotsk Sea), the mating seasons have a difference of about 1 month This gap between the mating seasons likely functions to limit the chance of interbreeding between the populations The mating season of a dalli-type population that breeds in the northern Okhotsk Sea is still to be investigated I expect it is likely to have a mating season that is different from that of nearby truei-types

Sea surface temperature in temperate and subarctic waters of the northern hemisphere usually peaks in late August and September Thus, it can be generalized that Dall’s porpoises give birth in early summer and that most of the postpartum females experience estrus in midsummer while nursing the calf The mating season and gestation time are believed to be so arranged as to maximize the survival of the offspring, but the mating season will also be influenced by maternal physiology that reflects the seasonality of food availability Marine productivity is high during the short summer in high latitudes where Dall’s porpoises summer and offers a suitable feeding environment to the offspring, which are believed to start taking solid food at 2-3 months of age, as well as to their mothers starting the next gestation while nursing It is reasonable to have parturition just prior to the peak of productivity, because the energetic cost to the mother is believed to be greater for nursing than for maintaining gestation (Lockyer 1981a,b, 1984)

The breeding seasonality and migration pattern of the Sea of Japan-southern Okhotsk Sea population are likely to have a relationship with oceanography in the Okhotsk Sea The Okhotsk Sea ice floes are formed in the northern part of the sea in November, cover about 75% of the sea in the maximum period, and gradually retreat from the south and entirely disappear in July They remain off the coast of the southern Okhotsk Sea from January to April Dall’s porpoises that winter in the Sea of Japan arrive at the southwestern and southeastern entrances to the Okhotsk Sea (Soya Strait and Nemuro Pass) in late May, but they have yet to pass into the major part of the southern Okhotsk Seas (Figure 93) They enter the southern Okhotsk Sea after retreat of the ice floes that occurs from April to May Sea surface temperature at the time is around 5°C-6°C, which is close to the minimum sea surface temperature known for the species Giving birth while they are in the Sea of Japan, that is, before entering

spring (Amano and Kuramochi 1992) The winter sea surface temperature off the coasts of Yamaguchi and Shimane (35°N-36°N, 131°E-134°E) where the population is known to winter is around 11°C-13°C, and the temperature off the west coast of Hokkaido where they give birth in May and June is 8°C-13°C; in both cases warmer than the temperature, Dall’s porpoises meet when they enter the southern Okhotsk Sea in late May and June (Miyashita and Kasuya 1988)

One way to work out the reproductive cycle of female cetaceans is to observe individually identified animals for years The method documents individual variation in the cycle but is feasible for only to a limited number of species Another method is to estimate the average cycle from the proportion of pregnant females in a sample of sexually mature individuals using gestation length estimated separately as a time calibration Caution is required because the proportion may change within a year If there is an ideal situation where we are able to obtain a certain number of unbiased monthly samples from a population for 1 year, we can calculate mean reproductive parameters of the population “Apparent pregnancy rate” is the proportion of pregnant females in the sample of sexually mature females and may contain biases of various sources Dividing the “unbiased” apparent pregnancy rate by the gestation period in years gives the annual pregnancy rate, which is a probability of a mature female becoming pregnant in a year, and the reciprocal of the annual pregnancy rate is the mean calving interval In this method, fetal mortality is incorrectly ignored The mean lactation period and mean resting period are also calculated assuming that the number of females at each reproductive stage is proportional to the average time in the stage Such an ideal situation will never happen in our cetacean studies, particularly when samples come from fisheries, because they can be biased due to fishing season, geographical limitation of fishing operation, and gear selection

Dall’s porpoise materials from the western Aleutian Islands area were interpreted to indicate that most of the females give birth during June 5 to July 24 and undergo estrus during July 20 to September 10, or about 45 days after parturition (Section 943) This suggests that most of the females reproduce annually Although this will be close to the truth, it has to be remembered that the conclusion was drawn from samples covering only a 2-month period and only part of the geographical range of the population and that it may require correction if some adult nonestrous females are segregating outside the sampled area Investigation of such a possibility was one of the objectives of the Japanese special cruises for sighting and hunting Dall’s porpoises (Section 933)

To the south of the breeding area around the western Aleutian Islands, these cruises found an area that was inhabited mostly by male Dall’s porpoises but contained about 30% females About 60% of the females, or 17% of the total sample, were sexually mature and were neither pregnant nor lactating, that is, resting This finding means that some proportion of

during the summer outside the breeding ground However, it was not possible to estimate the proportion of such females and estimate annual pregnancy rate for the population

Kasuya (1978) reported the reproductive status of 71 adult truei-type females taken off the Sanriku coast from January to March, which was after the mating season and before the parturition season There were only 6 in lactation (not pregnant), 5 lactating and simultaneously pregnant, and 58 in pregnancy (not lactating) Ignoring gear selection and segregation by reproductive stage, the proportion of females that became pregnant in the last estrous season was calculated as [5 in lactation and simultaneous pregnancy plus 58 in pregnancy]/71 = 089, which is an estimate of the annual pregnancy rate of the truei-type population The remaining 11% represent females that have not become pregnant in the last mating season

In addition to the 71 females, Kasuya (1978) used other biological information from gutted carcasses to estimate that an additional 34 females were sexually mature and not lactating These females have a very high probability of having been pregnant, because the fishermen are likely to discard pregnant uteri at sea together with other internal organs If we assume that these 34 females were pregnant, then the upper limit of their annual pregnancy rate would be given by (63 + 34)/ (71 + 34) = 0923 It is likely that porpoises in the truei-type population off Japan reproduce with an annual pregnancy rate of around 90%

These data allow speculation about the lactation period for the truei-type population The parturition season of this population is assumed to extend for about 45 days centered in mid-August (Section 944) It was assumed for simplicity that the earlier mentioned 105 likely pregnant females were lactating at the time of conception Out of the 105 females, 94 females had already ceased lactation by the fishing season from January to March, but 11 females were still lactating Since parturition of these 11 females likely occurred in the later part of the season (probably in September), they could have been lactating for 4  months (if killed in January) or 6 months (if killed in March) The majority ceased lactation before January Thus, lactation in truei-type population is likely to last 4-6 months at the longest Amano et al (1998c) recorded 8 lactating females in 50 sexually mature females killed in the same fishery, which does not alter the earlier conclusion on lactation time

Ferrero and Walker (1999) found that all the 69 calves of the year collected in the western Aleutian Islands area had only milk in their stomach and concluded that they lived only on milk for a minimum of 2 months after birth

Harbor porpoises, another species of Phocoenidae, inhabit cold waters of the northern hemisphere and are known to have a life history that is similar to that of Dall’s porpoises, that is, early sexual maturity (3-4 years), short longevity (less than 23  years), high annual pregnancy rate (74%–99%), parturition season in summer, and gestation period of 10-11 months (Lockyer 2003) Their lactation length was once thought to last over 8 months based on the presence of lactating females in March, but the possibility of shorter lactation is suggested

(Olafsdottir et al. 2003) Both Dall’s porpoise and the harbor porpoise exhibit precocity, short longevity, and high reproductive rate

The age of first parturition is important for the population dynamics of wild animals, but cetacean biologists have used age at the first ovulation as a sign of sexual maturation The possible reasons could have been because a trace of ovulation, the corpus albicans, remains in the ovary for a considerable length of time, perhaps for life (see also Section 1056) and offers a useful key for identifying maturation and because first ovulations are usually followed by conception in most cetaceans The age at first parturition can be about 1 year (the gestation time of Dall’s porpoise) greater than the age at first ovulation

Ferrero and Walker (1999) examined materials from 1911 females taken by the salmon drift-net fishery in the western Aleutian Islands area and identified 1061 as sexually mature This figure may be an overestimate of the maturity rate of the population because it does not take into account individuals segregating south of the salmon drift-net ground (see the following text in this section) The smallest sexually mature female was 147 cm in body length and was pregnant with a near-term fetus; it was considered to have first ovulated in the previous summer The proportion of sexually mature individuals in the sample increased with increasing body length, and the largest immature female was 193 cm long The body length at which 50% were sexually mature was 1797 cm (with 95% confidence interval of 1796-1798  cm) This is often stated as the average body length at sexual maturity (Table 97) The relationship between sexual maturity and age was analyzed in the same manner The youngest sexually mature female was aged at 3 years and the oldest immature female at 8 years Ferrero and Walker (1999) obtained 38 years as the unbiased estimate of mean age at the time of sexual maturation using the method of DeMaster (1978) and the age at 50% mature at 44 years by fitting a sigmoid curve to the data The

ent in a strict sense Considering the example of a Japanese finless porpoise that was born in an aquarium and underwent first estrus when she was 2 years old and gave birth at 3 years of age (Section 8526), we can accept such early maturation for the Dall’s porpoise

Kasuya and Shiraga (1985) analyzed hand-harpoon samples obtained in 1982 and 1983 mostly in waters 42°N-48°N, 155°E-180°E, located to the south of the locality of the sample analyzed by Ferrero and Walker (1999) They were unable to harpoon Dall’s porpoises that avoided the vessel in waters between the two sampling areas, which included mother-calf pairs and was considered as an extension of the breeding ground in the western Aleutian Islands area studied by Ferrero and Walker (1999) Their sample (185) contained 20 immature and 29 mature females (including two animals taken in latitudes 48°N-50°N, which were females), for a maturity rate of 59%, which was similar to the figure for the western Aleutian Islands area However, their sample had fewer young and old animals Female body lengths ranged between 160 and 190 cm, and most were 2-4 years old; only one female was aged at 1 year and the oldest female was 17  years old (Figure 919) These females had completed weaning but were most under or around the age at attainment of sexual maturity

Kasuya and Shiraga (1985) reported that ages of the youngest mature female and oldest immature female were 2 and 4 years (38 females in the age range), respectively, and that the age at 50% maturity was between 2 and 3 years These figures were slightly lower than the corresponding figures from the western Aleutian Islands area However, in view of the small sample size of our study and my observation that growth layers in their teeth start to be irregular at the age of 3-4 years, making determination by means of the tooth layers less reliable, a firm conclusion about the significance of the difference between the two estimates is not possible

Kasuya and Shiraga (1985) analyzed the relationship between female maturity and body length using the handharpoon sample from the southern area The range from the smallest mature female to the largest immature female was 164-187 cm (n = 40), which was similar to the range of

TABLE 9.7 Age (Years) and Body Length (cm) at Attainment of Sexual Maturity of Female Dall’s Porpoises

sample from the northern area Kasuya and Shiraga concluded that 50% were sexually mature at around 168-171 cm, which was smaller than the corresponding figure, 1797 cm, reported by Ferrero and Walker (1999) It was inconclusive whether the disagreement was due to the small sample size of Kasuya and Shiraga (1985) or to a biological factor

Kasuya (1978) analyzed body length and female sexual maturity using truei-type sample obtained from the handharpoon fishery off the Pacific coast of northern Japan I reported that the porpoises attained sexual maturity at body length 172-203  cm and that the length at 50% mature was 1867  cm These figures are about 6-7  cm greater than the corresponding figures from the western Aleutian Islands area It is already known that Dall’s porpoises in the western and eastern Pacific are larger than those in the central Pacific (Section 932) I also analyzed the relationship between sexual maturity and age and found that sexual maturity was reached at 25-115 years Such broad individual variation in maturation age throws doubt on the age determinations It could be attributed to erroneous interpretation of tooth layers that overestimated ages of some animals, particularly those at ages of over 3 or 4 years

Of the 399 truei-type females examined by Kasuya (1978), only 35% (139) were sexually mature, which was lower than the figure from the drift-net samples from the western Aleutian Islands area The sample contained extremely few individuals below age 2 years and had a peak age at 3-5 years As there were only two individuals below 160 cm, which were aged at 05 year, this skewed age composition, or scarcity of juveniles, is attributed to the real absence of juveniles and not to erroneous age determination I interpreted this almost complete lack of juveniles born in the previous summer in the trueitypes sample to indicate that the hand-harpoon fishery hunted mostly weaned individuals before they matured sexually, but it is more correct to say that the fishery targeted mostly individuals between their second winter and sexual maturity Such a skewed catch can happen due either to geographical segregation by age or to different responses of animals to the harpoon vessel, but the former seems to be more likely because of the similarity to samples in waters south of the western Aleutian Islands area analyzed by Kasuya and Shiraga (1985)

Amano and Kuramochi (1992) reported the sexual maturity of dalli-types taken by the hand-harpoon fishery in the Sea of Japan in May and June (38 immature and 35 mature) The proportion of sexually mature was 48%, which was higher than the 35% of the hand-harpoon sample from off Sanriku The range between the smallest mature and largest immature females was 180-199 cm, and 50% of individuals were sexually mature at 1870 cm These figures are close to those for off Sanriku and greater than those for the western Aleutian Islands area

Uncertainty remains about the determination of male sexual maturity It is more a question of behavioral investigation to judge whether a male has the ability to mate and produce

mate it through histological examination of carcasses taken in fisheries Maturation of testicular tissue is a gradual process spanning perhaps years, and dividing animals into two categories of mature and immature is a problem on which there is probably no common understanding among scientists Seasonal breeders such as the Dall’s porpoise change in anatomical feature of the gonads with season, which makes it difficult to apply a single criterion to materials collected in different seasons

Kasuya (1978) analyzed sexual maturity in 486 male trueitypes taken in winter, or the nonmating season, by the handharpoon fishery off Sanriku on the Pacific coast of northern Japan I sampled tissue from midlength of a testis centrally and peripherally from each male and examined the samples for the presence of spermatids and spermatocytes A testis was classified as “mature” if either of the cell types was found in both central and peripheral samples, as “maturing” if they were found only in the central sample, and “immature” if they were not found There were no cases where they were found only in the peripheral sample I also examined smears from the testis at midlength and from the epididymis for the presence of spermatozoa

Spermatozoa were present in only 30% of the epididymal smears of mature testes and 10% of maturing testes Spermatids or spermatocytes in such testes that have no spermatozoa in the winter will be transformed into spermatozoa by the beginning of the next mating season in the autumn About 3% of testes classified as “immature” by the earlier histological criteria had spermatozoa in the testicular smear, which shows the limitation of detecting small amount of such cells in ordinary histological preparations These results were in contrast with the summer sample of dalli-types obtained by the same hand-harpoon method in waters south of the western Aleutian Islands area (Kasuya and Jones 1984) Twentyfive individuals (806%) of 31 histologically mature males had copious spermatozoa in the epididymis, 3 had fewer spermatozoa, and 2 (6%) had no spermatozoa in the epididymis This difference, that is, 94% of mature testes had sperm in the epididymis in summer compared with 30% in winter, reflects seasonal change in gonadal activity The maturity criteria of Kasuya (1978) differed slightly from those used by Kasuya and Jones (1984), which classified testicular maturity into four stages of mature, late maturing, early maturing, and immature following a method used for the bottlenose dolphin (Section 1147), short-finned pilot whale (Section 12433), and Baird’s beaked whale (Section 1342) However, this does not change the earlier conclusion that many mature males do not have spermatozoa in the epididymis during winter

Mature males almost double their testicular weights in the mating season, while average testicular weight at attainment of sexual maturity does not show such great seasonal change (Table 98) The weight of a single mature testis was around 30-100 g in winter off Sanriku (truei-types), but it was 40-340 g in the summer sample of dalli-types in the western Aleutian Islands area (Figure 916), which is the reverse of the geographical cline in body size Ferrero and Walker (1999)

weighed the testis with epididymis attached, so their analyses were not directly comparable with those of other studies

The diameter of seminiferous tubules of mature males was approximately in a range of 68-120 μm The relationship between the diameter (Y, μm) and testicular weight (X, g) was shown by a single equation for the entire range of the data for truei-types off Sanriku in winter (Kasuya 1978):

Y = 3379 X0284, 5 < X < 100

However, the corresponding figure for dalli-types seemed to be almost doubled, 125-210 μm, in summer off the western Aleutian Islands area (Ferrero and Walker 1999) (Figure 917) The testis-weight data of these two studies are not directly comparable because Kasuya (1978) excluded the epididymis while Ferrero and Walker (1999) weighed the testis together with the epididymis

Kasuya (1978) obtained 2577 g and 3285 g as the weight of a single testis at the attainment of the maturing stage and the mature stage, respectively However, although the maturing males were in general located intermediate between immature and mature individuals in terms of testicular weight and tubule diameter, the overlap was extremely broad with the two other maturity categories of immature and mature This casts doubt on his classification of the

Weight of Single Testis (g), Body Length (cm), and Age (Years) at Attainment of Sexual Maturity in Male Dall’s Porpoises

the maturity of reproductively inactive males in winter The average of the two figures, 293 g, did actually separate mature and immature individuals with almost no exception, which suggests that the weight of 293 g can be an adequate criterion for classifying winter testes into the two categories of mature and immature The range in body length of the smallest mature male and the largest immature male was 180-215  cm, and 50% of males were sexually mature at 1959 cm The minimum age of a mature male was 35 years, and the maximum age of an immature male was 155 years Although the former figure is not unreasonable, the latter should be questioned because of the reason mentioned earlier (problems in aging)

Kasuya and Jones (1984) applied the maturity criteria of testicular histology established for short-finned pilot whales (see Section 12433) to the dalli-type porpoises hand-harpooned in waters south of the area sampled by Ferrero and Walker (1999) They observed that “mature” and “immature” individuals were separated at around 40 g testis weight and that “early maturing” and “late maturing” males were found near that weight They concluded that a threshold of 40 g was adequate for separating males into the two large categories of mature and immature (Figure 916) This threshold weight was about 10 g greater than the corresponding figure of 293 g, for a winter sample of truei-types off Sanriku Kasuya and Shiraga (1985), combining the dalli-type samples of 1982 and 1983 from the same area and applying the 40 g criterion to the total of 118 males, found 168-199 cm as the range in length of the smallest mature and largest immature males They estimated that 50% of males were sexually mature at around 184 cm The same sample gave an age range of sexual maturation of 2-7 years and age at 50% mature between 4 and 5 years (Table 98)

Ferrero and Walker (1999) analyzed the dalli-type sample obtained in summer in the western Aleutian Islands area, 46°N-53°N and 170°E-175°E, which was located north of the area sampled by Kasuya and Shiraga (1985) By examining testicular tissue taken from the center at the midlength of the testis, they classified the degree of maturity into three categories: “mature” having spermatozoa, “maturing” having spermatids or spermatocytes but not spermatozoa, and “immature” having none of the three cell types, where the proportion of tubules undergoing spermatogenesis was not taken into account It should be noted that they weighed the testis together with the epididymis, which I call gonad weight Tubule diameter correlated positively with gonad weight before maturation, but the correlation was lost in many mature individuals Tubule diameter ranged from 30 to 80 μm in immature testes (most of the gonads weighed less than 50 g) and increased with gonad weight until about 150 g (Figure 917) Gonads of mature males weighed between 70 and 350 g, and tubule diameter was in the range of 120-220 μm in mature males with gonads over 150 g This diameter is almost twice that (65-120 μm) of mature truei-type males in the winter off Sanriku Mention has been made previously of a similar seasonal change in testis weight

between body length and sexual maturity based on the large sample obtained from the drift-net fishery Their figures indicate that the body-length range for smallest mature male and largest immature male was approximately 166-194 cm, which was similar to the corresponding figure of 168-199 cm for the sample from further south, but body length at 50% mature, 1797 cm, was about 4 cm smaller than for the southern sample (Table 98) This disagreement probably could be attributed to selection by hand-harpoon sampling, where mature individuals are less likely to be attracted to the bow wave and are underrepresented compared to immature individuals of the same body length The age range of the youngest mature male and oldest immature male was 3-6 years, age of 50% mature was calculated at 45 years by fitting a sigmoid curve, and average age at sexual maturation was 50 years by the method of DeMaster (1978) (Table 98)

Amano and Kuramochi (1992) analyzed the sexual maturity of male dalli-types taken by the hand-harpoon fishery in the Sea of Japan in May and June, or shortly before the mating season that was believed to be in mid-June to late July (Sections 944 and 945), using a method almost identical with that of Kasuya and Jones (1984) and Kasuya and Shiraga (1985) They obtained 40 g as the average weight of the testis at attainment of sexual maturity This figure was about 10 g greater than the corresponding figure for off the Sanriku coast from January to March but similar to the figure for the offshore population during the period of parturition and mating They found that immature and mature individuals coexisted in a body-length range of 180-209 cm and half of the individuals were sexually mature at 192 cm, which was about 10 cm greater than in the western Aleutian Islands area and to the south

Kasuya (1978) believed that sexually mature individuals and juveniles below 2 years of age were underrepresented in the hand-harpoon fishery off Sanriku in the winter This could have been due to differential response to fishing vessels by growth stages or to segregation by growth and reproductive stage Adult individuals and mother-calf pairs tended to avoid vessels and were found segregated in the western Aleutian Islands area, including in the US EEZ; chase of a maximum of 48 min by a vessel at 10 knots failed to close with any for harpooning However, behavior was quite different in waters further south, where many approached the bow and offered a chance for sampling with hand harpoon (Kasuya and Jones 1984; Kasuya and Ogi 1987)

The Japanese hand-harpoon fishermen usually give up chasing such shy porpoises and look for other targets, but some of the fishermen recently started using a different strategy, in which they equipped their vessel with powerful engines and continued to chase mother-calf pairs until they were exhausted and then harpooned them (Amano et al. 1998c, in Japanese) This method was started by some fishermen who operated in the Okhotsk Sea in the summer season In 1995, Amano et al. (1998c, in Japanese) examined the catch

(175%) out of 325 sexually mature females, but the rates for 3 vessels included in the 57 were in a range of 30%–60% or 25 out of 54 adult females lactating Such an operation will inflict greater damage on the population

The Japanese salmon drift-net fishery operated in the western Aleutian Islands area, which was a breeding ground of the population and was inhabited by dalli-type porpoises that were not attracted to the bow It is unclear if the porpoises did not notice the presence of the net and got entangled or were attracted to prey animals in the net and therefore got entangled Newby (1982) was the first to analyze Dall’s porpoises caught in the salmon drift nets Ferrero and Walker (1999) carried out similar analyses using larger samples Their bodylength composition data are shown in Figures 913 and 914 and age composition in Figures 918 and 919 The abundance of zero-age individuals is quite different between these two studies as shown in Figures 918 and 919, presumably reflecting a difference in their sampling procedure or a difference in overlap of the sampling period with the parturition season However, the compositions of ages 1 year and above are similar between them The frequency decreased with increasing age until the highest age of 15 years Even allowing some possible uncertainty in age determination of old individuals, the longevity of this species does not exceed 20 years

Porpoises in the western Aleutian Islands area are believed to reach sexual maturity at the mean age of 38  years in females and 50 years in males (see Sections 947 and 948) The sex ratio was almost 1:1 for ages below 3 years, but the male proportion declined to about 1:2 or less at ages 4 and above, while longevity seemed to be similar for the two sexes at around 14-15 years (Figure 918) This general trend is also observed in the age composition reported by Newby (dotted line in Figure 919) This was different from the pattern of longevity observed in short-finned pilot whales off Japan, which exhibited a rapid decline of the male proportion after the age

of sexual maturation, accompanied by a great difference in observed longevity between sexes, that is, females lived longer It follows that the earlier mentioned age structure cannot be attributed to higher male mortality after sexual maturity as occurs in the short-finned pilot whales It is reasonable to assume that adult males were less vulnerable to the salmon drift-net fishery than adult females and juveniles below the age of 3 years One of the explanations for this is segregation of males to outside the drift-net ground

Kasuya and Shiraga (1985) reported the age composition of a hand-harpoon sample obtained in latitudes 42°N-47°N, which is to the south of the salmon drift-net ground off the western Aleutian Islands Their sample contained far more males, both adult and immature; the adult sex ratio was 55 males to 29 females and immature sex ratio 63 males to 20  females Figure 919 compares age compositions of this hand-harpoon sample with those obtained from the salmon drift-net fishery (Newby 1982) Newby and myself cross-checked our age readings, so the two age readers in Figure 919 are expected to have followed similar principles in interpreting tooth layers and are considered comparable The two drift-net samples are similar in age composition except for the frequency of zeroage animals as mentioned earlier (see Figures 918 and 919), but the hand-harpoon and drift-net samples show quite different age compositions (see Figure 919) Age classes with lower frequency in the northern drift-net sample are abundant

lier, this allows us to deduce that young individuals of both sexes around the age of sexual maturity tend to segregate in southern waters (note that males mature at a greater age than females)

Similar sex-biased distribution or migration patterns have been reported for humpback whales Recent nonlethal studies (Brown et al. 1995; Smith et al. 1999) in the breeding ground and on the migration route along the Australian coast identified a high male ratio (24 males/female) This was similar to 21:1 obtained from the past whaling data in lower latitudes but differed from 074:1 obtained from commercial catches in the Antarctic Brown et  al. (1995) deduced from these data that about half the female humpback whales remained in the feeding ground in the Antarctic during the season of parturition and mating Females of the species breed every 2-3 years, so those females who are not expecting estrus or parturition remain in the higher latitudes for more feeding and recovery for subsequent conception Such an explanation cannot be applied to the segregation in Dall’s porpoise population, because they feed both in the summer breeding ground and wintering grounds

Kasuya (1978), Newby (1982), Kasuya and Shiraga (1985), and Ferrero and Walker (1999) presented mean growth curves of Dall’s porpoises, using materials collected in a relatively short period and assuming that the growth pattern did not change for the years covered by the animals’ ages, that is, at least for the previous 15 years in the case of Dall’s porpoises Violation of this assumption of a stable growth pattern would make the obtained mean growth curve unreliable (Section 15322) Age determination is another key element in the growth analysis, but it has been my experience that the teeth of Dall’s porpoise are the most unsuitable for age determination among those of the odontocetes I have studied Among the earlier four growth analyses, that of Kasuya (1978) should be considered of less value due to greater uncertainty in age determination, particularly for individuals above the average age at sexual maturation

Ferrero and Walker (1999) determined physical maturity for 692 Dall’s porpoises and made a great contribution to understanding their growth Physical maturity is a sign of cessation of growth in body length and is characterized by fusion of all the epiphyses to the vertebral centra The state of maturity can be identified by examining the posterior thoracic and anterior lumbar vertebrae where fusion occurs last On the vertebrae of physically immature individuals, there is a layer of cartilage between the centrum and the epiphysis; this is replaced by bone tissue in vertebrae that have ceased increasing in length Ferrero and Walker (1999) showed that the minimum body length of physically mature males was 182 cm and that of the largest immature male 220 cm The corresponding range for females was 180-205 cm Mean body length of physically mature males was 1981 cm (n = 83, SE = 08566) and that of females 1897 cm (n = 164, SE = 04002), which

Ferrero and Walker (1999) did not find sexual difference in the age at attainment of physical maturity The youngest physically mature individual was aged at 5 years and the oldest immature individual at 8 years The age when 50% of individuals were physically mature was 716 years in males with 95% confidence interval of 57-86 years and 724 in females with a confidence interval of 63-81 years The interval between average age at sexual maturity and that at physical maturity was about 2-3 years in both sexes

Ferrero and Walker (1999) fitted Laird-Gompertz models to age-at-body-length data separately for two maturity stages: (1) sexually immature individuals (males <6  years, females <8), forced to pass through average birth length of 101 cm, and (2) sexually mature individuals (males >3 years, females >2  years) with the earlier mentioned body length at attainment of physical maturity as the asymptotic length The mean growth for each sex is shown by two equations in Figure 920 I am unable to understand the value of fitting growth curve separately for two maturity stages of overlapped age ranges The equations have little meaning for the overlapped age range (males, 3-6 years; females, 2-8 years) However, the equations reasonably suggest that neonates, which are born at a mean length of 101  cm (with a range of 86-118  cm), reach 151 cm in 1 year and 161 cm in 2 years The average

(see Sections 941 and 942) Ferrero and Walker (1999) also calculated the relationships between body weight and age, which suggest that a calf born with an average weight of 17 kg attains 158 kg (males) or 123 kg (females) at the time of physical maturation Plots of body weight on body length suggest that the correlation is almost lost at around the age of 8 years This should agree with the age when all the individuals have attained physical maturity

Dall’s porpoise males exceed females by 44% in average body length and 28% in average weight Behind the greater weight of males relative to length is sexual dimorphism in body shape (Jefferson 1989) First, volume of dorsal muscle is greater in the thoracic region, which looks vertically swollen when mature males are seen from the side A second difference is the high tail peduncle of males These differences contribute to weight difference between the sexes Other sexual dimorphisms that will not affect weight are shape of the dorsal fin and the tail flukes The tip of the dorsal fin is located farther forward on adult males than on females, giving an impression that the fin is tilted anteriorly The posterior margins of the tail flukes expand posteriorly on adult males The posterior margins of the tail flukes are located anterior to a line connecting the tips of the flukes on juveniles, are nearly on a line with the tips of the flukes on adult females, and are located posterior to the line on adult males Jefferson (1988) concluded that such sexual dimorphism could function in fighting or deciding rank between males The large body muscle might have such a function, but further information on behavior is needed to evaluate such a hypothesis

Koga (1969, in Japanese) obtained the following equation between body weight (W, kg) and body length (L, cm) for dalli-types of both sexes taken by the salmon drift-net fishery off the western Aleutian Islands area:

Log W = 2441 LogL − 3435

This equation can also be expressed as W = 0000367L2441

Dall’s porpoises can make a dash at 55 km/h (Jefferson 1988) and dive to a maximum of 180 m (Berta and Sumich 1999) Such a high activity level and the offshore habitat could have been behind the difficulty of maintaining the species in captivity In the 1960s, a facility of the US Navy succeeded in keeping Dall’s porpoises in captivity for over 2  years, and physiological information thus obtained from 5 porpoises was compared with that for bottlenose dolphins (n = 5), a coastal species, and Pacific white-sided dolphins (5), which inhabit intermediate waters (Ridgway and Johnston 1966) The authors found that the blood volume of Dall’s porpoises (143 mL/kg of body weight) was higher than for Pacific white-sided dolphins (108 mL/kg) and bottlenose dolphins (71  mL/kg) and that other parameters followed a similar order, that is,

(g/100 mL) 203, 170, and 144; and heart weight as % of body weight 131, 085, and 054 They concluded that these physiological parameters suggest that the Dall’s porpoise has the highest physical ability among the three species compared, as evidenced by observations in the ocean Ridgway and Johnston (1966) further reinforced this conclusion from the daily food requirement to maintain body weight, 14, 85, and 6 kg, with the highest for the Dall’s porpoise and the lowest for the bottlenose dolphin, and from blubber thickness measured at the same position on the body that had a reverse correlation with activity rank, that is, 1 cm on the Dall’s porpoise, 2 cm on the Pacific white-sided dolphin, and 3 cm on the bottlenose dolphin Generally speaking, Dall’s porpoises are the smallest and bottlenose dolphins the largest among the three species Dall’s porpoises seem to have the highest heat production and depend least on the insulation ability of blubber in their thermoregulation The authors noted high swimming activity of Dall’s porpoises during capture operations, as evidenced by “rooster-tail” swimming

Sergeant (1969) addressed the relationship between daily food consumption and heart weight He noted that a Dall’s porpoise weighing 120 kg (2 m in body length) consumed as much as 15 kg of mackerel daily, or 1135% of body weight (125% when calculated from data in the original text of Sergeant 1969), which was much greater than the corresponding figures for Pacific white-sided dolphins (78%) and bottlenose dolphins (42%) He found a correlation between daily food intake (kg/body weight) and heart weight/body weight ratio and estimated daily food intakes for other species with known heart weight and body weight Later, Ridgway and Kohn (1995) with additional data obtained the following relationship between body weight (B, kg) and heart weight (H, kg) of adult Dall’s porpoises:

logH = −1614 + 0808logB

They calculated similar equations for adult individuals of the genus Lagenorhynchus and bottlenose dolphins and obtained −1729 and −1927, respectively, as the intercept The slope of 0808 remained the same for the three equations Their equations gave the heart weight of individuals with an assumed body weight of 190  kg at 169  kg for the Dall’s porpoise, 130 kg for the Pacific white-sided dolphin, and 082 kg for the bottlenose dolphin

Lockyer (1981a) obtained the following single relationship between body weight (B, kg) and heart weight (H, kg) for all species of cetaceans from the harbor porpoise weighing tens of kg to the blue whale weighting over 100 tons:

H = 000588B0984

This relationship can also be expressed as logH = −2230 + 0981 logB This expresses the general trend among cetacean species but can be quite different from equations that represent the relationship within smaller taxonomic groups or for single species

the food consumption rate estimated by Sergeant (1969) They observed for dalli-type Dall’s porpoises taken in May in the northern Sea of Japan off the west coast of Hokkaido that the maximum contents of the forestomach was 168 kg and that the contents of the forestomach tended to decline with time from early morning They also referred to information for other species suggesting that it took 8 h for a full forestomach to be emptied and calculated that if the stomach was filled three times a day the daily food intake could be only 504% of body weight They concluded that the estimated daily food consumption of 15 kg or 1135% of body weight reported by Sergeant (1969) represented a case of overfeeding in the captive environment I would question whether a Dall’s porpoise could continue to eat twice as much in captivity as in the wild for over 2 years as reported by Ridgway and Johnston (1966) It should be noted that some Dall’s porpoises taken by the salmon drift-net fishery had an amount of undigested soft remains of prey in their stomach that could have represented 4 kg in fresh condition (see Kuramochi et al. 1991)

It is true that there are difficulties in estimating the food requirements of small cetaceans using individuals in captivity, because their activity is different from that of wild animals and change in body weight must be monitored and adjusted carefully However, if we accept a daily food consumption of 504% of body weight proposed for Dall’s porpoises by Ohizumi and Miyazaki (1998) and reject the observed daily food consumption of 125%, then we also have to reject the corresponding figures 78% for Pacific white-sided dolphins and 42% for the bottlenose dolphin However, the last figure was supported by another independent experiment on 11 bottlenose dolphins in a Japanese aquarium, which concluded that the daily food intake of 35%–61% of body weight was necessary to maintain body weight for 21  months (Section 11521) Further, if we accept that Dall’s porpoises examined by Ohizumi and Miyazaki (1998) fed less in daytime, which is likely in certain circumstances, then it would be difficult to fill their stomach three times a day at an interval of 8 h A key problem in their calculation is the assumption that Dall’s porpoises fill the forestomach when it is emptied by digestion

The stomach of dolphins and porpoises has four compartments: (1) forestomach, (2) main stomach, (3) connecting channel, and (4) pyloric stomach The connecting channel may be divided into two compartments in some species (Harrison et al. 1970) The main function of the forestomach, the largest of the four compartments and lacking in digestive glands, is temporary storage of ingested food, although there is some progress in digestion by powerful mechanical activity and digestive juices migrating up from the main stomach Food stored in the forestomach gradually moves to the main stomach, the second largest compartment and equipped with digestive glands, then to the connecting channel and pyloric stomach, and finally to the duodenum for subsequent absorption The feeding environment for dolphins and porpoises is quite different from that of cows feeding in a meadow For Dall’s porpoises, the chance to meet prey animals is less predictable, so they will make an effort to ingest as much food

to store ingested food in the forestomach The main stomach and subsequent compartments continue digestion as materials are received from the forestomach With such feeding habits, the amount of food found in the forestomach at any particular time of the day does not have value for estimating daily food consumption in cetaceans

Dall’s porpoises consume various epipelagic and mesopelagic squids, fishes, and crustaceans The species composition apparently varies seasonally and geographically, which is considered a reflection of both availability and food preference Morejohn (1979) analyzed the stomach contents of carcasses stranded year-round in Monterey Bay Numerous species of squids and fishes were present in the stomachs throughout the year, but seasonal trends were not clear due to limited sample size He concluded that the fish species preferred were hake, herring, juvenile rockfish, and anchovy, and the preferred squid species were Doryteuthis (then Loligo) opalescens and Gonatus sp A larger sample would perhaps have identified crustaceans year-round He reported that the Dall’s porpoise occurred year-round in waters outside the 200 m isobath, but in winter the range expanded to waters inside the 100 m isobath, which he suggested reflected seasonal distribution of prey

Mizue et al. (1966, in Japanese) and Koga (1969, in Japanese) examined the stomach contents of Dall’s porpoises killed in the salmon drift-net fishery in the Bering Sea and noted that over half of the individuals had only squid in the stomach and the remaining individuals had squid and fish, squid and crustaceans, or squid, crustaceans, and fish From these results, they concluded that squid are the main nutritional source for Dall’s porpoises (and of salmon) in the region Mizue et al. (1966) stressed that there was only one Dall’s porpoise found with salmon in the stomach even though their sample was obtained in the salmon drift-net fishery

Mizue et  al (1966) examined food preference by reproductive status of Dall’s porpoises killed in the salmon driftnet fishery and found that 53% of 17 pregnant females (with fetuses over 70 cm) had fish in their stomach but only 27% of other individuals had fish in the stomach From this observation, they concluded that pregnant females prefer fish because they require more nutrition They did not clarify the sex or reproductive status of the “other individuals,” which makes evaluation of their conclusion difficult Ohizumi et al. (2003) offered more information relating to this question; they found that sexually mature Dall’s porpoises of either sex depended significantly more on myctophids than immature individuals of the same sex caught in the salmon drift-net fishery in the Bering Sea A difference in food preference between sexes was also significant but of less magnitude The possible factors behind such differences in food preference are still to be investigated Weaning short-finned pilot whales were found with smaller squid beaks in their stomach than those found in adults in the same school (Section 12443)

ments are greater during lactation than in pregnancy

Kuramochi et al. (1991) compared the contents of the forestomach between the Bering Sea and offshore western North Pacific based on 32 Dall’s porpoises hand-harpooned during research cruises in May through September, 1984 and 1985 Thirty-one had both fish and squid in the forestomach, but the proportions were different between the two geographical areas, north and south of the US EEZ around the Aleutian Islands The Bering Sea sample had almost equal numbers of squid and fish (45:55), while fish predominated in the Pacific with a ratio of 3 squid to 97 fish Ohizumi et al. (2003) also found a similar result, that is, 80%–94% of the prey was fish in the offshore western North Pacific west of 135°W while 69% was squid in the Bering Sea These results are useful in judging geographical difference in prey taken, but the relative degree of nutritional contribution between fish and squid is difficult to evaluate because we do not know how squid beaks and fish otoliths differ in time of passage through the stomach

The average number of squid (represented by beaks and undigested bodies) found in a single forestomach was 63 in the Bering Sea and 27 in the offshore North Pacific, and squid beaks represented 77% of the Bering Sea total and 91% of the Pacific total (Kuramochi et al. 1991) The squid belonged to 6 families, the majority (89%–97% by the regions) to the Gonatidae The remaining families were Enoploteuthidae, Cranchiidae, Chiroteuthidae, Onychoteuthidae, and Histiotheuthidae, which comprised only 3%–11% in the total number of individuals identified and were considered almost negligible in the overall diet The most important family was the Gonatidae; seven species in three genera were identified in the stomachs, in decreasing order of abundance Gonatopsis borealis, Gonatus onyx, G. pyros, G. berryi, G. middendorffi, Berryteuthis anonychus, and B. magister Thus, G. borealis was the most important species, comprising 23% (Pacific) or 63% (Bering Sea) of squid identified to species Fiscus and Jones later (1999) reported cephalopod composition in the stomachs of Dall’s porpoises in the Bering Sea and western North Pacific The sample was obtained from the Japanese salmon drift-net fishery in the area indicated in Figure 911 and the Japanese hand-harpoon cruise for scientific purpose in 1982 (Figure 921) The results were similar to those reported by Kuramochi et  al. (1991) in both the most dominant family being Gonatidae and species composition within the family Fiscus and Jones (1999) did not find significant difference in the composition of major squid species between the Pacific area and the Bering Sea plus Aleutian Islands area

Kuramochi et al (1991) estimated the weight of squid consumed by Dall’s porpoises Using the length of lower beaks, they calculated the mantle length of G. borealis consumed by the porpoises to be in a range of 44-405 mm, with 89% in 50-150 mm The total weight of these squid was estimated at 21-6,460 g/individual in the Pacific and 21-11,413 g/individual in the Bering Sea They also studied the number and

weight of undigested soft parts of squid present in the stomach of 17 Dall’s porpoises The number of squid represented by the soft tissue ranged between 1 and 111 squid per porpoise, and their estimated weight at the time of ingestion ranged between 5 and 4387 g per porpoise Some of these porpoises could have had over 4  kg of squid, or about 29% of body weight in their stomachs at some time before capture (150 kg is assumed for body weight), but we do not know how long it remained in the stomach or how many feeding incidents were represented It should also be noted that Dall’s porpoises often vomit their stomach contents when killed by hand harpoon

US scientists collected the stomach contents of Dall’s porpoises killed in the Japanese salmon drift-net fishery beginning in 1978 and analyzed them subsequently In 1982, Japan and US scientists planned a cruise for sighting and lethal sampling of the species outside the drift-net operation area, but at a late stage, the United States withdrew from the lethal sampling program (see Section 933) The stomach contents collected during the cruise from August to September 1982 were transferred to US scientists following an earlier agreement, and it was reported to the INPFC meeting that the cephalopod beaks were analyzed by Cliff Fiscus and fish otoliths by Thomas Crawford (Jones et  al. 1985b) The analysis of cephalopod beaks was published by Fiscus and Jones (1999), which is cited previously However, the results of the study of fish otoliths remain unpublished The only results of the study available to me are in a table that was handed to me by Crawford’s superior at an IWC Scientific Committee meeting and is now deposited in the Zoological Section of the National Science Museum in Japan It contains the only data on fish species consumed by Dall’s porpoises in the offshore North Pacific south of the western Aleutian Islands area

As Crawford had already retired from his position and I was unable to contact him, I will summarize his results in the following to prevent the valuable record from being lost His

the Hoyo-maru No. 12 in August and September 1982, mostly in waters south of the western Aleutian Islands, 40°N-50°N, 158°E-174°E (Figure 921), but could have included samples from three individuals killed in the nearshore waters of Hokkaido during the cruise (Fig 1 in Kasuya and Shiraga 1985) Dall’s porpoises in offshore waters are known to consume more fish than cephalopods (see the preceding text in this section) An average of 760 fish otoliths were found in the 27 stomach samples (with a range of 4-2834) As these figures did not apparently distinguish between the left-and right-side otoliths, the number of fish represented by the otoliths could be smaller The number of fish species in the stomachs was between 1 and 13 (73 on average) A total of 19,329 fish otoliths were identified to family: in decreasing order Myctophidae (806%), Gadidae (48%), Microstomatidae (33%), Notosudidae (08%), others (5%), and unidentified (58%) The top three families coincide with the results of Ohizumi et al. (2003) for Dall’s porpoises in northern waters

Wilke et al. (1953) were probably the first to describe the stomach contents of Dall’s porpoises from off the Pacific coast of northern Japan They examined 54 truei-types and 4 dalli-types caught in latitudes between 38°15′N and 42°N during March-May, 1949 and 1950 The southern part of this range could have been inhabited mostly by truei-types and the northern part by dalli-types They identified the prey species using undigested food remains; they did not use squid beaks and fish otoliths for the analyses The stomach contents of the 54 truei-types were represented mostly by Myctophidae (70% of total prey animals), small squid (18%), and one gadid Myctophidae were not found in the stomachs of the 4 dalli-types, but they had squid (98% of total prey animals) and some gadids Details on the location of capture of these animals are not available The apparently distinct difference in prey items between truei-and dalli-types could be due to a latitudinal difference and small sample size for the latter Wilke et al. (1953) also reported the stomach contents of striped dolphins that resembled those of the truei-types and were represented mostly by Myctophidae and some squid Striped dolphins usually occur south of the Dall’s porpoise range

Walker (1996) reported the stomach contents of Dall’s porpoises in the Okhotsk Sea based on samples obtained from the hand-harpoon fishery operated by a small-type whaler Yasu-maru No.1 in July and August 1988 Out of 88 dallitypes examined, he obtained the stomach contents for 73 He identified fish species using left and right otoliths separately and used the larger figure for his analysis In the same way, lower and upper beaks were separately identified and the greater figure was used He found fish remains in all the 73 stomachs and squid remains in 54 (74%) A total of 2916 individual fish were represented by 13 species in 9 families, which were in decreasing order Clupeidae (901%, represented by the single species the Japanese sardine Sardinops melanostictus), Gadidae (74%, represented by the single species Alaska pollock Theragra chalcogramma), Zoarcidae (3 species identified), Engraulidae (the Japanese anchovy

ilies Microstomatidae, Ammodytidae, and Hexagrammidae, each represented by a single fish Squid in the 54 porpoises comprised 733 individuals of 6 species in 3 families The most abundant was Gonatidae (965% in number of individuals), of which the schoolmaster gonate squid Berryteuthis magister was the most abundant (869% of the total cephalopods) Thus, the major food items in the Okhotsk Sea in the summer of 1988, in terms of number of individuals, were the three species Japanese sardine, schoolmaster gonate squid, and Alaska pollock Walker (1996) stated that the order remained the same on a nutritional basis Time of capture and degree of digestion suggested that Japanese sardines were eaten during the day, while the schoolmaster gonate squid were consumed at night Ohizumi (2008, in Japanese) reviewed recent information on the food habit of Dall’s porpoises

At an early stage of the study of food habits of Dall’s porpoises, there was a general understanding that they were nocturnal feeders preying on Myctophidae and cephalopods that surface at night Later, it became known that they were likely to change feeding behavior to meet the habits of prey species, for example, by feeding on Japanese sardines in the Okhotsk Sea Amano et  al. (1998d) attempted to confirm this feeding flexibility based on direct observation in the ocean They assumed that Dall’s porpoises were feeding if they frequently changed swimming direction while swimming at the surface and recorded this behavior on board the research vessel Hoyo-maru No. 12 in the offshore North Pacific in 1986 and 1987 and on the hand-harpoon vessel Man-ei-maru No. 5 in the Okhotsk Sea in 1988 In the offshore North Pacific, the feeding behavior was most frequent around sunrise (no observations available in night) and decreased with time during the day On the other hand, in the Okhotsk Sea they encountered such feeding behavior at high frequency during the day

Ohizumi et al. (2000) reported yearly change in food items of dalli-type Dall’s porpoises taken by the hand-harpoon fishery off Hokkaido in the Okhotsk Sea and the Sea of Japan The Japanese sardine was their major prey item in both the seas in the 1980s, but it almost disappeared from their stomachs in the 1990s and was replaced by Alaska pollock in the Sea of Japan and by Japanese anchovy and schoolmaster gonate squid in the Okhotsk Sea Ohizumi et al. (2000) believed that this change was a response of Dall’s porpoises to the decline of the sardine population around Japan and that Dall’s porpoises were obliged to switch their prey items from epipelagic species to those in deeper waters

We learned that Dall’s porpoises probably change their prey with growth and with change in reproductive status It is also evident that Dall’s porpoises have the ability to switch prey items responding to change in abundance of prey species and to modify their feeding behavior to match the behavior of the prey species However, it will require further study on the abundance of the earlier mentioned prey species and feeding energetics of Dall’s porpoises before evaluating the effects on their physiology of the change in feeding environment after the decline of the Japanese sardine

School size in cetaceans not only reflects social structure but is also influenced by environmental factors such as size and density of schools of prey species and chances of meeting predators According to data obtained during the cruise of the Hoyo-maru No. 12 in the area south of the western Aleutian Islands form August to September 1982 (Figure 921), Dall’s porpoises sighted in waters between 44°N and 50°N and between 163°E and 174°E, which was close to the Aleutian Islands rarely rode the bow wave of the vessel; many of them suddenly disappeared after the first sighting, offering no opportunity for resighting Within this ship-avoiding area, there was a slightly smaller area (46°N-50°N, 166°E-174°E) where we encountered numerous mother-calf pairs, which also avoided vessels Thus, the northern part of the surveyed area near the Aleutian Islands was judged to be inhabited mostly by adult breeding individuals To the southwest of this breeding ground, there was an area with approximate range of 41°N-47°N and 158°E-174°E, which was inhabited by young individuals that rode the bow wave and were harpooned (Kasuya and Jones 1984) The age composition of the sample obtained in 1982 and 1983 in the same area (Kasuya and Shiraga 1985) showed that there were almost no individuals of 1 year of age and that the peak frequency was at 2-4 years of age, which were of lower frequency in the drift-net sample from the Aleutian Islands area Similar geographical differences in behavior have been reported in other parts of the range of Dall’s porpoise (Sections 933 and 945; Kasuya and Ogi 1987; Yoshioka and Kasuya 1991; Amano and Kuramochi 1992)

The size of Dall’s porpoise schools encountered during the Hoyo-maru No. 12 cruise in 1982 ranged from 1 to 14 and had a modal frequency of 2 (298% to the north of 45°N and 274% to the south of that latitude) The mean school size, 351 (SD = 25) and 377 (SD = 21), was not different between the two latitudinal ranges where we observed great behavioral difference These school sizes were obtained during the mating season (Section 944)

On the Hoyo-maru No.12, I encountered 28 schools of Dall’s porpoises, mostly truei-types, off the Pacific coast of Hokkaido on September 16-18, 1982, before her arrival at the port of Kushiro (42°59′N, 144°23′E) School size ranged from 1 to 10 and had a mode at 3 and 4 (both represented by 6  schools) The mean schools size was 425 (SD = 23) Slightly greater mean school size in Japanese coastal waters could be a reflection of high productivity

Miyashita (1991) provided information on school size of Dall’s porpoises in summer off Japan, covering the whole Okhotsk Sea and the western North Pacific extending to 170°E (Figures 91 and 92) This area is inhabited in summer by one truei-type population and several dalli-type populations and is known to include several breeding grounds (Figure 910) His survey was conducted in August and September, 1989 and 1990, or in the same season as the cruise of the Hoyo-maru No. 12 mentioned earlier, and reported mean school size of

Sea and Pacific area within 200 nm from the Kuril Islands, and 577 for the offshore dalli-types The figure for trueitypes was similar to the earlier figure, that is, 425, obtained off the Pacific coast of eastern Hokkaido, and slightly smaller than the mean school sizes for dalli-types

Kasuya (1978) compared school size between 32 schools of truei-types off Sanriku (Pacific) and 54 schools of dalli-types in the Sea of Japan and southern Okhotsk Sea The season was May-September for both Both samples had a mode at 2 individuals, but the mean school size was 54 for the trueitypes and 35 for the dalli-types The proportion of schools of 4 or more was 59% for the truei-types and 37% for the dalli-types The dalli-types had slightly smaller school sizes There are no explanations available for the minor seasonal and geographical variation in group size

The social structure of some cetaceans has been studied based on the long-term observation of individually identified groups Such study is suited to species that are individually identifiable, inhabit easily accessible coastal waters, and live in relatively small communities or populations Such a situation does not always exist for the Dall’s porpoise The following is some information from the cruise of the Hoyo-maru No.2 in 1982 that offers a glimpse of social structure in the species

The Hoyo-maru No.2 cruise was conducted in August and September, the season of peak sea surface temperature, when the southern limit of the Dall’s porpoise was around 42°N in longitudes of 155°E-173°E and sea surface temperature was at 20°C Warmwater species such as the common dolphin, striped dolphin, bottlenose dolphin, and short-finned pilot whale occurred in sea surface temperature over 17°C The two intermediate species, Pacific white-sided dolphin and northern right whale dolphin, occurred in temperature of 12°C-19°C Later, Myashita (1993) concluded, with additional data, that these two species in summer inhabit the same temperature range of 11°C-25°C and in latitudes of 40°N-50°N and estimated their abundance

During the Hoyo-maru No.2 cruise, the scientists noted that dalli-type Dall’s porpoises responded to the vessel differently by sea surface temperature in the surveyed area (Figure 921) Out of 432 individuals sighted in waters above 11°C, 245 or 57% of the sighted individuals rode the bow wave, but only 4 or 23% of 171 sighted in waters below 11°C approached the bow However, later surveys also identified the presence of a similar behavioral difference in other parts of the North Pacific and adjacent seas of different water temperature profiles, and it was found that the effect of sea surface temperature was not a primary factor (Section 933) Dall’s porpoises in the southern survey area killed by harpooning contained a high proportion of 2-4-year-old individuals under the age of attainment of sexual maturity and a high male proportion of 83% Currently, available information suggests that juveniles born in the summer are weaned at 4-6 months of age but stay at least until the second summer in the same waters inhabited

2 years, they are likely to segregate from breeding individuals The tendency of segregation from the breeding group is more pronounced in males, which has some similarity with the case in striped dolphins where weaned immature individuals, particularly males, tend to live in a segregated school of similar-age individuals (Section 10514)

As mentioned earlier, we scientists on the research vessel Hoyo-maru No.12 attempted to chase the dalli-type porpoises that avoided our vessel as long as possible, but we were unable to catch up with them They repeatedly displayed spurts in speed followed by a pause at a distance of 300-500 m from our vessel and finally increased swimming speed and disappeared from sight The vessel speed was about 10 knots (185 km/h), and duration of the chase was 12 min on average with a maximum of 48  min This result was disappointing given the purpose of our cruise to collect unbiased samples

One of the factors behind the geographical variation in behavior of the species was the presence of mother-calf pairs Apart from one mother-calf pair sighted off the east coast of Hokkaido, we sighted 41 mother-calf pairs in offshore waters, and 36 (86%) of them occurred during 6 days from August 25-30 in surface of 98°C-128°C The surface temperature in the remaining period was over 11°C The area with mother-calf pairs was a part of the area where porpoises tended to avoid the research vessel The whole area occupied by ship-avoiding individuals was tentatively designated as a “breeding area” (Section 933)

The Hoyo-maru No.12 encountered 100 dalli-type Dall’s porpoise schools during the 6-day period in the breeding area, of which only 29 schools were visually observed for their composition at a close distance (Kasuya and Jones 1984) Out of the 29 schools, 16 schools totaling 45 individuals did not contain mother-calf pairs, and 13 schools of 69 individuals contained a total of 36 mother-calf pairs Since it has been over 20 years since the publication of Kasuya and Jones (1984), I cannot recall how we selected the 29 schools Reexamination of original records for the 100 schools allows the conclusion that 22 schools of 70 individuals did not contain mother-calf pairs (mean school size was 557), that 14 schools of 78 individuals contained mother-calf pairs (mean school size was 318), and that the remaining 64 schools were not examined for the presence of mother-calf pairs

The earlier observations suggest that about half of the individuals in the breeding area are found in schools that contain mother-calf pairs and that the mean size of schools with mother-calf pairs is slightly greater than those without them encountered in the same area Only two Dall’s porpoises were captured from the 16 schools containing mother-calf pairs; they were aged 3 and 10 years and were neither pregnant nor lactating but had a corpus luteum and corpora albicantia in the ovaries These females could have been in estrus or in the very early stage of pregnancy Adult males would have aggregated around such females near estrus for mating opportunities A group of four individuals were visually identified as adults

In the breeding area, we also encountered Dall’s porpoise schools, which were composed only of smaller individuals,

size They were two schools each containing three small individuals and one school of five small individuals Such small individuals were also found in schools either with or without mother-calf pairs They were probably 1-year-old juveniles born in the previous summer that had already become weaned before the summer but still remained in the breeding area living with individuals of the same age or with some adults Many of them would segregate outside the breeding area (southern waters in this case) by the third summer or 2 years of age (Figure 919)

Thus, it appears that weaned calves tend to live together and increase distance from their mothers, and many of them (though not all) segregate outside the breeding area in the third summer Examples of similar dispersal after weaning have been reported in striped dolphins (Section 10514), bottlenose dolphins (Section 1144), and male sperm whales (Best 1979)

Each of the previously mentioned 13 schools that contained at least one mother-calf pair and allowed some detailed observations contained 1-7 mother-calf pairs (Kasuya and Jones 1984) Seven of these 13 schools consisted of only mother-calf pairs Their composition was 1 mother-calf pair (2  schools), 3  mother-calf pairs (2 schools), 4 mother-calf pairs (2 schools), and 7 mother-calf pairs (1 school) More surveys will find additional combinations The last case with seven mother-calf pairs requires additional explanation When it was first sighted, there were 2 schools close to each other, 1 school with 3 mother-calf pairs, and the other with 4 mother-calf pairs, but they joined to form a school of 7  mother-calf pairs when approached by our research vessel These observations suggest that mother-calf pairs tend to merge to form larger aggregations Mothers in these schools that contained no adult males are likely not in estrus

The remaining 6 schools in the 13 schools contained individuals of adult size in addition to mother-calf pairs:

1 mother-calf pair + 1 adult: 3 schools 2 mother-calf pairs + 1 adult: 1 school 2 mother-calf pairs + 1 adult + 1 of unknown age:

1 school 3 mother-calf pairs + 2 adults: 1 school

The adult individuals seen with mother-calf pairs were probably adult males waiting for the opportunity to mate Most females ovulate soon after parturition and enter into the next pregnancy while lactating (Sections 944 and 946)

Our current knowledge on social structure of Dall’s porpoises can be summarized as follows Calves are nursed for some period of more than 2 months and probably 4-6 months Weaned calves leave their mothers by their second summer or before their mothers again give birth, which is deduced from the facts that most females are believed to breed annually, and there has been no sighting of a mother accompanied by two juveniles These weaned juveniles spend the second summer in their mother’s summering ground, but many of them seem to segregate from their mothers in the third summer when they are 2 years old to an area dominated by

pronounced in males, judging from the high male proportion in the summering ground frequented by immature individuals and low male proportion in the summer breeding ground (Section 949) After these juveniles attain sexual maturity at ages around 4-5 years, they presumably return to their breeding ground, but some nonbreeding adults of both sexes are likely to summer with immature individuals

The majority of females with their newborn calves live in a temporary aggregation of several mother-calf pairs in the summer breeding ground The relatively small school size with a mode at two individuals and the earlier analysis of school structure suggest that it is unlikely that this species lives in cohesive schools that last for several years Rather, the mother-calf bond, which is likely to last less than a year, seems to be the longest stable bond between individuals The currently available limited observations do not support cooperation among adult males in gaining mating partners Males are more likely to approach females on their own

Large-scale commercial hunting of Dall’s porpoises started in Japan in the 1910s in Iwate Prefecture (in latitudes of 38°59′N-40°27′N) in the Sanriku Region on the Pacific coast of northern Japan, expanded geographically accompanied by increase in catches during the peri-World War II period, and since the 1950s withdrew to a local fishery along the Sanriku coast, particularly to the two prefectures of Iwate and Miyagi (37°44′N-38°59′N), as alternative winter fishing when other fishing objectives were limited The fishery operation at the last stage also exhibited some fluctuation The annual catch of around 10,000 in 1957-1965 declined to 5,000-6,000 in 1967-1975, accompanied by a decline in daily catch per vessel (Kasuya 1982) The cause of the decline in catch per vessel has not been clarified Subsequently, the catch underwent two temporary declines (1979-1980 and 1984-1986) followed by rebounds The first of the rebounds came with expansion of the winter fishery south to the coast off Ibaraki Prefecture (35°45′–36°50′N) and the spring to autumn operation off the Pacific coast and Sea of Japan coast of Hokkaido (41°30′– 43°30′N) and the second with extensive summer operation in the southern Okhotsk Sea (43°50′–45°30′N) Such changes generated the speculation that the fishery retained catch level through the expansion of fishing ground and targeted populations (Kasuya and Miyashita 1989, in Japanese)

The Fisheries Agency of Japan reported increase in catch to over 10,000 in 1982 and a higher figure of over 13,000 in 1987, which was reported to the IWC in May 1988 Under these circumstances, I had an opportunity in August 1988 to travel the Hokkaido coasts to collect information on the Dall’s porpoise hand-harpoon fishery and confirmed operations of 21 vessels in Abashiri (44°01′N, 144°16′E), a fishing port on the Okhotsk Sea coast, and 9 vessels in Kiritappu (43°05′N, 145°08′E), another fishing port on the Pacific coast

and the remaining 22 in Iwate Prefecture Local fishermen on the Okhotsk Sea coast who were not operating the Dall’s porpoise fishery stated that there were 40 vessels operating there in the summer Dall’s porpoise fishermen of the time told me that an ordinary vessel could catch 200 porpoises in a month and certainly over 1000 in a year, but that more catch was possible with a vessel of higher speed Even if the hunters’ estimate was halved, the estimated catch of these vessels exceeded the earlier mentioned official statistics for 1987, which threw doubt on the official statistics of 1987, unless it was assumed that there was a sudden expansion of the fishery in 1988 I reported this question to the Fisheries Agency and published it in Kasuya and Miyashita (1989, in Japanese) to be reviewed in the Scientific Committee Meeting of the IWC

In 1989, the IWC Scientific Committee discussed the discrepancy in statistics and expressed concern about the reliability of the Japanese statistics and on the sustainability of the Dall’s porpoise fishery (IWC 1990) Responding to the question on the government statistics, Kasuya (1992) examined landing records of fish markets in Iwate Prefecture, which were the basic source of the government statistics, and reached the conclusion (Kasuya 1992) that Iwate Prefecture underreported the catch of the 1987 season, presumably to avoid criticism of the rapid catch increase Some attempt at overreporting was also seen for the 1988 catch, when introduction of a quota system was discussed for the fishery, with the possible purpose of getting a large share of the quota and with probable consent of Iwate Prefecture The manipulation included use of an erroneous conversion factor that created a smaller-than-actual catch figure estimated from the weight of meat landed By correcting the identified manipulations, Kasuya (1992) estimated Dall’s porpoises taken in the Japanese hand-harpoon fishery at 37,200 in 1987, which was 145% of the official “revised” statistics for the year, and at 45,600 in 1988, which was 113% of the corresponding official figure (Table 29) The latter figure was believed to be the peak level for the fishery

Change in the system of collecting catch statistics for Japanese small-cetacean fisheries, or improvement of coverage in later statistics, caused some additional confusion in evaluating the government statistics The Offshore Division of the Fisheries Agency collected the statistics until 1987, when the task was transferred to the Coastal Division, which issued the 1988 statistics that were reported to the IWC in the spring 1989 and included levels much greater than reported for 1987 The Coastal Division, responding to the question raised by the Scientific Committee in 1989, in 1990 presented to the IWC the revised statistics for the years 1986 and 1987 (IWC 1991; Japan Progress Report 1991) Thus, there were two sets of catch figures for the 1986 and 1987 seasons, and the revised figures were all greater (Table 29) The disagreements between the two sets of statistics mainly came from incomplete geographical coverage by the earlier system, which did not include operation of the fishery in Hokkaido, Aomori, Ehime, and Oita Prefectures and catches of Iwate fishermen that were landed in Hokkaido (Kasuya 1992) The statistics of the Offshore Division could not

the reason for the failure is unclear

Another concern of the Scientific Committee was the population size and sustainability of the catch Miyashita and Kasuya (1988) made the first attempt to estimate the abundance using sighting data from May to August in the area of the northeastern Sea of Japan north of 40°N and the coastal western North Pacific between 34°N and 45°N and obtained estimates of 46,400 for the Sea of Japan stock dalli-types (31,800 in the Sea of Japan and 14,600 in the Pacific) and 58,000 for truei-types These figures certainly underestimated the abundance, because they did not include truei-types in the central Okhotsk Sea and dalli-types in the southern Okhotsk Sea, and were soon exceeded by total catches in several subsequent years in the late 1980s

Miyashita (1991) resolved the problem with the earlier abundance estimates by surveying the Okhotsk Sea and adjacent area including the Russian EEZ outside of territorial waters using 3 vessels during July 26 to September 22, 1989 and 1990 The Sea of Japan was excluded because the Sea of Japan-Okhotsk Sea population remaining there was considered negligible during the season of the survey (Miyashita and Kasuya 1988) For the purpose of estimating abundance, the survey area was divided into four strata based on existing knowledge on the distribution of the stocks (Figures 91 and 92): (1) range of a putative dalli-type population in the western North Pacific (south of 51°N and in longitudes 150°E-165°E), (2) range of a dalli-type population summering in the northern Okhotsk Sea, (3) range of a Sea of Japan-southern Okhotsk Sea dalli-type population in the southern Okhotsk Sea, and (4) range of a truei-type population summering in the central part of the Okhotsk Sea, centered at the boundary between the earlier 2nd and 3rd populations The boundary between the 2nd and 3rd population was the line connecting the north end of Sakhalin Island and the northern Kuril Islands (dotted line in Figure 92) The Pacific area west of 150°E contained both a truei-type population and Sea of Japan-Okhotsk Sea dalli-type population

Abundances estimated from the previous analysis are in Table 99 The confidence intervals are tentative figures calculated by myself using the coefficient of variation presented in Miyashita (1991) Miyashita et al. (2007, in Japanese) reported another abundance estimate using sighting data obtained in 2003, which was not entirely an independent estimate because

Miyashita (1991) The incomplete survey coverage occurred due to unavailability of a Russian permit for surveys in the Russian EEZ, which has always been the largest obstacle to the management of cetaceans off northern Japan

One of the criticisms of the abundance estimates had to do with the effect of porpoise response to survey vessels Dall’s porpoises are often attracted to the vessel If they approached the vessel before being sighted by observers, abundance would be overestimated, but if the animal fled from the vessel the abundance will be biased low We know that response of Dall’s porpoises toward vessels varies by region, reflecting behavioral difference and segregation by growth and reproductive status Correction for such plausible biases was not made by Miyashita (1991) and Miyashita et al. (2007) These abundance estimates assumed g(0) = 1 as usual

Among the 4 populations listed in Table 99, (1) and (2) have been subject to incidental kill in the salmon drift-net fishery since before World War II, (3) and (4) have been targeted by the Japanese hand-harpoon fishery, and (3) was once subject to incidental kill in the salmon drift-net fishery in the Sea of Japan A goal is to evaluate the effect of these mortalities on the populations and to understand the current levels of the populations relative to those before exploitation An attempt in this direction was made (Okamura et al. 2008), but further efforts are required for the construction of more realistic catch histories of the affected populations, including the large amount of hunting during the peri-World War II period and the 1980s

The catch quotas given to the Japanese hand-harpoon fisheries for the 2008 season were 8396 dalli-types and 7916 truei-types, which were slight modifications of basic figures used in 1993 as 4% of the central estimate of abundance by Miyashita (1991) (Tables 62 and 64) The idea of the potential biological removal (PBR) is a method often used to judge whether the current or expected level of artificial removal of small cetaceans is within a safe level (Wade 1998), and I feel it places great weight on the safety of a population but not on the maximum utilization of the resource This has been used in the United States to judge whether the known level of incidental mortality in fishing gear is acceptable PBR is calculated by the following equation:

PBR = [Minimum Abundance, Nmin] × Rmax × 05 × [Safety Factor, Fr]

TABLE 9.9 Abundance of Dall’s Porpoises around Japan

percentile of the estimate is used for Nmim, and Fr is arbitrarily selected from a range of 05-10 considering reliability of the catch statistics and the possibility of causes of additional mortality other than in fisheries (a smaller figure is used if such risk is considered great) Rmax is the reproductive rate expected for the population at close to zero density, and 4% is assumed as a default for small cetaceans The PBR, even with the most optimistic assumption of Fr = 1, would estimate as a safe level less than 3000 for either of the two Dall’s porpoise populations targeted by the Japanese hand-harpoon fishery, which is less than half of the current catch quotas If a smaller figure is assumed for Fr to compensate for possible underreporting, then the PBR value will be smaller The Fisheries Agency of Japan stated in its publicity materials that even the PBR supported the current quota for the Dall’s porpoise fishery However, it used 008 or 009 for Rmax, without presenting any supporting evidence for such an optimistic level Thus, the current quota for the fishery is not supported by the PBR as used in the United States

Rmax is the maximum rate of increase expected for a population in the ideal environment with minimum density, which is difficult to measure for an actual population I once attempted to determine the possible range of Rmax for Dall’s porpoises in collaboration with a population dynamist and obtained a range of 3%–9% The Fisheries Agency felt that our calculation could not exclude the possibility of Rmax being as low as 3%–4%, which was a figure at odds with the catch level of the time, and refused to allow our work to be presented to the Scientific Committee meeting of the IWC I agree that our estimate of Rmax is of limited scientific value but believe that such a biased attitude of the Fisheries Agency puts the rational management of fishery resources at risk It should be noted that the current catch quotas for the Japanese Dall’s porpoise fishery are supported by only very optimistic assumptions of Rmax

As a part of the activities of the INPFC to evaluate the effect of incidental mortality of Dall’s porpoises in various driftnet fisheries on their population, the United States placed observers on board Japanese fishing vessels and their own research vessels to collect sighting data for an abundance estimate Their data covered the months from June to August in 1987-1990 Using the data, Buckland et al (1993) estimated the abundance for each 5° latitude by 5° longitude square and obtained a total abundance of 1,186,000 Dall’s porpoises by summing the figures for each square (95% confidence interval was 991,000-1,420,000) This figure covered the range of the species in the Pacific Ocean from the Japanese coast to the US coast and most of the Bering Sea area but did not cover the Okhotsk Sea, the Sea of Japan, and the part of the Bering Sea outside of the drift-net fishery operation Combining the given figure with 550,000 that represents those areas 2, 3, and 4 in Table 99 results in a total of 174 million Dall’s porpoises, which still does not include the porpoises in the Russian EEZ on the east coast of the Kamchatka Peninsula This figure also

sel and assumes g(0) = 1 as usual

Abundance of the Pacific white-sided dolphin and northern right whale dolphin were estimated at 931,000 (with 95% confidence interval of 206,000-4,216,000) and 68,000 (20,000239,000), respectively (Buckland et  al 1993) These figures cover the entire range of the species in the North Pacific proper

Dall’s porpoises were killed in the offshore North Pacific incidentally in the salmon drift-net fishery, squid drift-net fishery, and large-mesh drift-net fishery Since a trial operation in 1914, Japan enjoyed free operation of the salmon driftnet fishery for a long period, but the International Convention for the High Seas Fisheries on the North Pacific Ocean signed in 1952 (Section 62) limited the Japanese operation to the international waters in the Bering Sea and western half of the North Pacific (Matsumoto 1980, in Japanese) Since 1956, a fishery agreement between Japan and USSR has required a USSR permit for the Japanese vessels operating in its EEZ Regulation continued to be tightened thereafter, and only a small-scale operation was left after 1992 in the Russian EEZ, which was operated jointly with Russia (Morita 1994, in Japanese) The squid drift-net fishery was also started by Japanese fishermen in 1978 and operated off the coast of Sanriku and Hokkaido and expanded to 160°W in 1980 (Yatsu et al. 1993) This fishery was joined by vessels of Taiwan since the late 1970s and Korea since 1979 (Gong et al. 1993; Yeh and Tung 1993) The large-mesh drift-net fishery targets marlin and tunas This fishery started in the 1840s and was operated in nearshore waters of Japan, expanded to the offshore waters of northern Japan in the early 1970s, and reached to 150°W in the 1980s (Nakano et al. 1993) Taiwanese fishermen also participated in this fishery, but details are not available These large-scale high-seas drift-net fisheries ceased operation by the end of 1992 following the UN resolution of December 1989 (Section I32)

It seems certain that the drift-net fisheries have influenced at least some of the numerous Dall’s porpoise populations discussed previously and also the Pacific white-sided dolphin and northern right whale dolphin For the latter two species, almost nothing is known about stock structure (Hobbs and Jones 1993)

At the occasion of a symposium in Tokyo convened by the INPFC in 1991, Hobbs and Jones (1993) estimated the incidental mortality of several dolphin species in the squid and large-mesh drift-net fisheries and their effects on the populations (Table 910) However, they did not make any significant mention of the mortality of Dall’s porpoises in the salmon drift-net fishery and its effect on the population, which has been the major task of the small-cetacean-related activity of the INPFC for nearly 10 years The salmon drift-net fishery operated since 1952 under international regulations, but the fishery deployed a several times greater number of nets than permitted by the regulations and underreported the resultant takes of salmon (Sano 1998, in Japanese) Under such circumstances, it must have been impossible to estimate with any certainty the number of Dall’s porpoises killed in the fishery It was also possible that the scientists had lost enthusiasm for

pursuing the truth under the circumstance of the fishery going to be closed in accord with the UN resolution The INPFC symposium in Tokyo concluded its scientific activity on Dall’s porpoise mortality in the drift-net fishery

Incidental Mortality of Dolphins and Porpoises in Drift-Net Fisheries in the Northern North Pacific Estimated for 1990