MARINE MAMMAL SCIENCE, 19(3):484–501 (July 2003)

Ó 2003 by the Society for Marine Mammalogy

CRANIAL SEXUAL DIMORPHISM AND

GEOGRAPHIC VARIATION IN FRASER’S
DOLPHIN, LAGENODELPHIS HOSEI
WILLIAM F. PERRIN
Southwest Fisheries Science Center,
National Marine Fisheries Service, NOAA,
8604 La Jolla Shores Drive,
La Jolla, California 92037, U.S.A.
E-mail: william.perrin@noaa.gov

M. LOUELLA L. DOLAR
Tropical Marine Research,
6363 Lakewood Street,
San Diego, California 92122, U.S.A.

MASAO AMANO
Otsuchi Marine Research Center,
Ocean Research Institute, University of Tokyo
Akahama, Otsuchi, Iwate 028-1102, Japan

AZUSA HAYANO
Department of Zoology, Kyoto University,
Kitashirakawa-Oiwake, Sakyo, Kyoto 606-8502, Japan

ABSTRACT
Knowledge of geographic variation is important to questions of population
assessment and management. Fraser’s dolphins have been exploited in two regions
in the western Pacific. Analysis of 137 skulls from the Philippines, Japan, Taiwan,
Australia, Indonesia, Malaysia, South Africa, France, the U.S., St. Vincent and the
Grenadines, and the eastern tropical Pacific revealed sexual dimorphism in 5 of 26
measurements (difference of 1.9%–5.8% between males and females), similar to
levels of cranial dimorphism in other small pelagic delphinids. Males had a larger
braincase and temporal fossae and smaller external nares than females. Sexually
dimorphic characters were excluded, and male and females samples were pooled to
examine geographic differences in the remaining characters. Multivariate analyses
yielded significant differences between the Philippine and Japanese series within
the North Pacific and between a pooled North Pacific series and a North Atlantic
series. The Japanese skulls were on average broader and had a wider rostrum,
larger orbit, larger internal nares, and longer braincase than the Philippine skulls.
These differences suggest that Fraser’s dolphins exploited in Japanese and
Philippine waters in directed fisheries or as bycatch belong to different
populations and should be assessed and managed separately.
Key words: Cetacea, Delphinidae, Lagenodelphis hosei, Fraser’s dolphin, sexual
dimorphism, geographic variation, Philippines, Japan, skull, discriminant analysis.

484
 PERRIN ET AL.: GEOGRAPHIC VARIATION IN FRASER’S DOLPHIN 485

The population biology of Fraser’s dolphin, Lagenodelphis hosei, is poorly known.

Although described in 1956 (Fraser 1956), its external appearance and overall
distribution remained unknown until the 1970s (Perrin et al. 1973, Perrin et al.
1994a, Dolar 2002). It is now known to occur regularly in pelagic equatorial waters
and occasionally in warm-temperate waters around the world (Jefferson et al. 1993,
Jefferson 1994, Dolar 2002). Abundance in the eastern tropical Pacific has been
estimated at about 289,000 (Wade and Gerrodette 1993) and in the eastern Sulu Sea
at about 8,700 (Dolar 1999). However, nothing has been published on population
structure or geographic variation. This is of concern because the species has been
exploited directly or as bycatch in a number of fisheries (Perrin et al., 1994a),
including in Japan (Amano et al. 1996) and the Sulu Sea (Dolar 1994, Dolar et al.
1994), and assessment of the impact and sustainability of the catches requires
knowledge of population range and size. The primary purpose of the study this paper
reports was to compare a series of skulls from the Sulu Sea with a series from Japan,
to examine the question of geographic variation and possible stock difference
between the two regions. We also examined smaller series from the North Atlantic,
southwestern Pacific, and southeastern Indian oceans and single specimens from
other regions, including the holotype specimen from Sarawak. Because of the
limited sample sizes, a necessary first step was to evaluate sexual dimorphism.

METHODS
The Sample
The sample included 137 skulls (Fig. 1, Appendix 1). For the Philippines (locality
1 in Fig. 1), 46 were from the bycatch in a tuna driftnet fishery at Siaton in Negros
Oriental (Dolar 1994) and one was from a dolphin fishery at Pamilacan in Bohol
(Dolar et al. 1994); these collectively are referred to hereinafter as the Philippine
series. The sample from the Japanese main islands (localities 2 and 3) consisted of 34
specimens taken in a drive fishery at Taiji in 1991 (Amano et al. 1996) and one from
a stranding. The North Atlantic sample included 10 from a mass stranding in Florida
(locality 4; Hersh and Odell 1986), nine from a mass stranding in France (locality 5;
van Bree et al. 1986), two from a dolphin fishery in St. Vincent (locality 6; Caldwell
et al. 1976), and one from a stranding in North Carolina (locality 7). Six specimens
from the eastern tropical Pacific (localities 8 and 9) were killed incidentally during
tuna purse-seining operations (Gerrodette 2002). The geographic series of eight from
the east coast of Australia (localities 10–12) and 13 from the Indian Ocean coast of
South Africa (localities 13 and 14) were from single strandings. In addition, we
examined two specimens from Indonesia (localities 15 and 16) and single specimens
from Iwo Jima in offshore Japan (locality 17), Taiwan (locality 18), north of Papua
New Guinea (locality 19), and Sarawak in Malaysia (locality 20).

Data Collection and Analysis

All skulls were measured by MLLD, WFP, or both; measurement techniques
were cross-checked for the two observers. Measurements were taken as described in
Perrin (1975) with the addition of a measure of tooth width; all measurements were
to the nearest mm except for tooth width, which was to the nearest 0.1 mm. Sex
was determined by AH for five of the specimens from Japan from bone samples
through multiplex PCR reaction with three zinc-finger primers as described by
486 MARINE MAMMAL SCIENCE, VOL. 19, NO. 3, 2003

Figure 1. Approximate collection localities for specimens included in study. Details in

text and Appendix 1.

Bérubé and Palsbøll (1996). Of these, three were cranially adult females (NSM
M29579, 29584, and 29585) and two were cranially immature males (M29580 and
29583). Specimens were judged cranially mature if distal fusion had occurred
between the premaxillae and maxillae in the rostrum; this provides a rough
assessment of maturity based on homology with other small delphinines sensu
LeDuc et al. 1999 (Perrin and Heyning 1993). Only cranially mature specimens
were included in the metric analyses. The cranially mature sample totaled 120
specimens from all localities. As small terminal teeth may be lost during
preparation, the higher of left and right counts were used as indices of tooth count.
If only the left or right count was available, it was used as an estimate of the highest
count. The measurement data are available from the authors in an Excel 2000
spreadsheet (Microsoft Corporation 1999).
Statistical analyses were performed with the software package SYSTAT 9 (SPSS
Inc. 1999). Comparisons of means with t-tests used pooled variance and were run
with and without Bonferroni adjustment when multiple comparisons were carried
out simultaneously. For the multivariate analyses, missing values were estimated by
multiple regression after eliminating variables and cases with large numbers of
missing values. In this culled sample, variables were reduced from 26 to 21 used in
the analyses (11 of remaining variables with 1–6 missing values) and cases reduced
from 120 to 115 included in the analyses (18 of remaining cases with 1–4 missing
values). Reliability of discriminant analyses was assessed with jackknifed
classifications (cases omitted from the analysis one at a time and classified with
the resulting function[s]).
RESULTS
Sexual Dimorphism
Sex was known or genetically determined for 36 males and 41 females in the
culled sample of cranially mature specimens (Table 1). The sex composition of the
geographic samples was not statistically different from the overall ratio at P 5 0.95.
 PERRIN ET AL.: GEOGRAPHIC VARIATION IN FRASER’S DOLPHIN 487

Table 1. Sex composition of geographic samples of cranially mature specimens. Sex of

three females from Japan determined by genetic analysis (see text).

Sample Males Females Total

Philippines 7 5 12
Japan 13 19 32
North Atlantic 7 6 13
Australia 1 3 4
South Africa 6 4 10
Other 2 4 6
Total 36 41 77

For example, chi-square for the largest and most skewed sample (Japan: 13 males
and 19 females) was 0.534 (df 5 1).
Comparison of cranially adult males and females in the unculled sample using
t-tests with a Bonferroni adjustment indicated no differences at P > 0.05. However,
in the interest of avoiding possible type-II statistical error caused by the recognized
tendency of the Bonferroni adjustment to inject positive bias toward accepting the
null hypothesis (Perneger 1998), both adjusted and unadjusted P values are given
in Table 2, and males and females are treated separately for measurements where
the unadjusted P , 0.05. (On average, only one or two measurements of the 27
would be expected to exhibit difference due to chance alone at a 5 5%.)
The unadjusted analyses showed average difference for five measurements: width
of external nares, parietal width, height of braincase, length of temporal fossa,
and height of temporal fossa. Males had a larger braincase and temporal fossae
and smaller external nares than females. In a discriminant analysis (stepwise,
backwards), the jackknifed performance of the discriminant function was relatively
poor (72% of males and 69% of females correctly classified). Consequently the sex
of specimens of unknown sex was not estimated; rather the dimorphic measure-
ments were excluded from the multivariate geographical analyses described below.

Geographic Variation
The Philippine and Japanese series did not differ in length of skull or rostrum,
but the Japanese skulls were on average broader, with wider rostrum, larger orbit
and internal nares, and longer braincase (Table 2, Fig. 2). In a discriminant analysis
(stepwise, backwards), the jackknifed classification function correctly assigned 85%
of the Japanese and 83% of the Philippine specimens (Fig. 3); approximate F 5
11.04; df 5 8, 60; P , 0.001.The most important discriminating variable was
zygomatic width. Number of teeth did not differ between the series (Table 2).
Comparison of a combined western North Pacific sample (WNP: Philippines,
Japan, Taiwan, and Iwo Jima; n 5 71) with the North Atlantic sample (NA:
France, Florida, North Carolina, and St. Vincent; n 5 22) revealed differences
between means of only one measurement, width of internal nares, at P < 0.05 after
Bonferroni adjustment. However, a discriminant analysis (stepwise, backwards,
with prior distributions of 0.76 for WNP and 0.24 for NA, jacknifed) correctly
classified 83% of WNP and 73% of NA specimens (Fig. 4) and indicated difference
at P , 0.0001 (Wilks’ lambda: approximate F 5 7.91; df 5 8, 84). The most
important discriminating variable was rostrum length, which was on average about
 Table 2. Means, standard deviations (for sample sizes >5), minimum and maximum values, sample sizes (in parentheses), and results of t-test
488

comparisons (with and without Bonferroni adjustment) for selected cranial measurements (numbers in parentheses follow Perrin [1975]) of cranially
adult specimens and highest upper and lower tooth counts for series of Lagenodelphis hosei from the Philippines, Japan, North Atlantic, Australia,
South Africa, and for all locations pooled (mean, range, and sample size only). Includes specimens not included in geographic series for multivariate
analyses. Males and females listed separately for characters exhibiting sexual dimorphism in sample pooled over localities. Measurements in mm
except tooth width (to nearest 1/10 mm). Asterisks indicate difference (with Bonferroni adjustment) between Philippines and Japan at P 5 0.05(*),
0.01(**), or 0.001(***). Degrees of freedom not given for Philippines vs. Japan comparison (can be calculated from sample sizes; 5 total n 2 2). See
Appendix 1 for list of specimens included.

Results of t-tests (with/without

Bonferroni adjustment)
males vs. Phil. vs. WNP vs.
females Japan NA
Measurement Philippines Japan N. Atlantic Australia South Africa All locat. P df P P df
Condylobasal 419.4 6 9.78 424.9 6 14.27 422.0 6 11.71 412.1 6 11.32 424.6 6 12.45 421.6 0.834/ 73 0.063/ 0.960/ 92
length (1) 396–440 397–460 403–453 400–431 410–449 396–460 1.000 1.000 1.000
(36) (34) (22) (7) (12) (116)
Length of 234.8 6 7.66 234.9 6 10.52 230.5 6 7.89 226.4 6 8.85 235.8 6 5.23 233.4 0.694/ 73 0.982/ 0.047/ 92
rostrum (2) 213–248 219–259 217–249 219–240 228–243 213–259 1.000 1.000 0.839
(36) (34) (22) (7) (12) (116)
Width of 116.9 6 5.14 121.1 6 5.62 117.0 6 5.57 113.1 6 4.74 121.7 6 4.21 118.2 0.314/ 75 0.002/ 0.233/ 92
rostrum at: 108–126 111–137 109–129 105–119 116–131 105–137 1.000 0.043 1.000
MARINE MAMMAL SCIENCE, VOL. 19, NO. 3, 2003

base (3)* (36) (34) (22) (7) (12) (118)

1/4 84.0 6 4.07 88.3 6 4.33 84.9 6 3.24 81.7 6 4.27 87.2 6 5.10 85.6 0.888/ 72 ,0.001/ 0.323/ 91
lengtha** 76–93 79–101 79–92 78–90 80–100 76–101 1.000 0.001 1.000
(36) (33) (22) (7) (12) (115)
1/2 length 67.5 6 3.53 73.7 6 4.92 70.4 6 4.86 67.9 6 5.18 69.8 6 3.98 70.2 0.964/ 73 ,0.001/ 0.952/ 92
(5)*** 61–76 66–89 65–88 62–78 63–80 61–89 1.000 ,0.001 1.000
(36) (34) (22) (7) (12) (116)
 Table 2. Continued.

Results of t-tests (with/without

Bonferroni adjustment)
males vs. Phil. vs. WNP vs.
females Japan NA
Measurement Philippines Japan N. Atlantic Australia South Africa All locat. P df P P df
3/4 length 51.5 6 5.00 56.9 6 4.43 52.7 6 4.04 51.9 6 2.61 53.2 6 4.55 53.6 0.616/ 73 ,0.001/ 0.287/ 92
(7)*** 45–69 50–72 41–58 49–57 47–65 41–72 1.000 ,0.001 1.000
(36) (33) (22) (7) (12) (116)
Width of 31.2 6 2.55 32.7 6 3.23 32.1 6 2.75 29.0 6 2.94 32.0 6 3.00 31.8 0.524/ 69 0.035/ 0.820/ 91
PMXs at 1/2 27–37 25–39 28–39 26–35 27–38 25–39 1.000 0.729 1.000
length (6) (36) (33) (22) (7) (9) (112)
Tip of rostrum 291.0 6 7.55 291.6 6 12.29 288.9 6 9.43 280.7 6 10.18 291.5 6 7.45 289.9 0.960/ 71 0.798/ 0.340/ 88
to external 273–306 268–321 272–308 272–295 281–305 268–321 1.000 1.000 1.000
nares (8) (32) (34) (22) (7) (12) (112)
Tip of rostrum 292.1 6 8.97 293.9 6 11.01 289.0 6 1.31 287.2 6 2.55 291.3 6 8.7 290.9 0.310/ 64 0.482/ 0.113/ 89
to internal 271–313 269–317 266–316 274–303 278–301 208–317 1.000 1.000 1.000
nares (9) (34) (33) (22) (6) (11) (111)
Preorbital 205.5 6 6.21 212.5 6 7.43 208.8 6 6.67 203.3 6 6.37 214.3 6 6.64 208.9 0.466/ 75 ,0.001/ 0.907/ 89
width (10)** 195–221 200–235 199–221 190–208 204–224 190–235 1.000 0.002 1.000
(33) (34) (22) (7) (12) (116)
Postorbital 224.7 6 7.39 234.0 6 6.66 230.7 6 6.11 223.7 6 6.82 236.8 6 5.73 229.8 0.956/ 68 ,0.001/ 0.563/ 85
PERRIN ET AL.: GEOGRAPHIC VARIATION IN FRASER’S DOLPHIN

width 208–241 221–253 221–241 210–233 228–247 208–253 1.000 ,0.001 1.000
(11)*** (32) (33) (21) (7) (8) (106)
Zygomatic 220.3 6 7.05 230.0 6 5.71 225.4 6 5.73 218.4 6 4.83 232.7 6 8.60 225.3 0.498/ 69 ,0.001/ 0.866/ 84
width 203–234 220–246 216–236 208–222 216–247 203–247 1.000 ,0.001 1.000
(14)*** (32) (32) (20) (7) (11) (108)
489
 Table 2. Continued.
490

Results of t-tests (with/without

Bonferroni adjustment)
males vs. Phil. vs. WNP vs.
females Japan NA
Measurement Philippines Japan N. Atlantic Australia South Africa All locat. P df P P df
Width of 44.3 6 1.80 45.3 6 2.06 42.6 6 3.21 42 47.0 6 2.00 44.9 0.006/ 75 — — —
external nares 42–47 42–50 38–46 (1) 44–50 38–50 0.186
(13): males (7) (13) (7) (6) (36)
females 47.4 6 2.97 47.2 6 3.38 45.5 6 7.00 45.7 46.0 46.7 — — — — —
43–51 39–52 42–50 44–47 45–47 39–52
(5) (19) (6) (3) (4) (41)
Maximum 73.4 6 3.18 75.6 6 4.83 74.0 6 3.43 73.6 6 4.50 77.5 6 3.00 74.7 0.719/ 75 0.026/ 0.586/ 90
width of pre- 68–81 64–87 68–80 67–80 73–82 64–87 1.000 0.538 1.000
maxillae (15) (34) (34) (22) (7) (12) (116)
Parietal width 169.4 6 5.00 170.1 6 3.97 173.8 6 3.66 165 177.5 6 5.13 171.7 0.027/ 74 — — —
(16): males 163–177 163–175 168–178 (1) 173–187 163–187 0.826
(7) (13) (6) (6) (35)
females 169.0 6 7.07 171.2 6 6.97 168.2 6 5.00 163.7 162.8 168.5 — — — — —
159–177 163–188 164–178 160–169 159–165 159–188
MARINE MAMMAL SCIENCE, VOL. 19, NO. 3, 2003

(5) (19) (6) (3) (4) (41)

Height of 115.9 6 4.22 114.4 6 3.20 115.4 6 4.72 114 116.2 6 3.06 115.0 0.008/ 75 — — —
braincase 109–121 111–123 110–123 (1) 112–121 108–123 0.242
(17): males (7) (13) (7) (6) (36)
females 114.6 6 2.70 113.7 6 4.43 114.0 6 3.74 107.7 106.8 112.3 — — — — —
111–118 106–124 109–120 105–110 102–112 102–124
(5) (19) (6) (3) (4) (41)
 Table 2. Continued.

Results of t-tests (with/without

Bonferroni adjustment)
males vs. Phil. vs. WNP vs.
females Japan NA
Measurement Philippines Japan N. Atlantic Australia South Africa All locat. P df P P df
Internal 124.2 6 4.06 130.1 6 5.02 126.6 6 3.51 123.6 6 4.08 127.6 6 2.43 126.8 0.320/ 74 ,0.001/ 0.660/ 90
length of 117–134 120–139 121–136 117–130 124–132 117–139 1.000 ,0.001 1.000
braincase (35) (34) (21) (7) (12) (116)
(18)***
Length of 70.0 6 3.06 75.6 6 4.43 80.0 6 3.06 71 76.0 6 4.82 75.2 0.037/ 74 — — —
temporal fossa 65–75 69–83 76–84 (1) 71–83 65–84 1.000
(19): males (7) (13) (7) (6) (36)
females 73.6 6 5.77 72.8 6 4.39 75.3 6 5.57 68.3 70.3 72.6 — — — — —
67–82 66–80 67–80 62–75 65–78 62–82
(5) (19) (6) (3) (3) (40)
Height of 48.0 6 4.24 47.2 6 3.33 55.7 6 4.27 50 51.3 6 48-59 50.2 0.003/ 75 — — —
temporal fossa 41–53 45–54 51–61 (1) (6) 41–61 0.105
(20): males (7) (13) (7) (36)
females 48.8 6 5.12 47.2 6 4.89 47.5 6 3.62 46.7 43.0 47.0 — — — — —
41–55 41–60 44–53 42–53 40–44 40–60
(5) (19) (6) (3) (4) (41)
,0.001/
PERRIN ET AL.: GEOGRAPHIC VARIATION IN FRASER’S DOLPHIN

Length of 53.8 6 2.23 55.7 6 1.94 56.4 6 3.47 52.6 6 3.82 55.8 6 2.56 54.9 0.226/ 71 0.010/ 86
orbit (25)** 49–58 52–60 51–68 45–56 53–60 45–68 1.000 0.009 0.171
(34) (32) (21) (7) (12) (110)
Length of 58.7 6 4.30 60.4 6 3.74 58.5 6 3.45 60.6 6 4.93 61.3 6 4.32 59.6 0.901/ 68 0.086/ 0.228/ 90
antorbital 50–69 53–72 50–64 54–70 54–69 50–72 1.000 1.000 1.000
491

process (26) (34) (34) (22) (7) (11) (115)

 Table 2. Continued.
492

Results of t-tests (with/without

Bonferroni adjustment)
males vs. Phil. vs. WNP vs.
females Japan NA
Measurement Philippines Japan N. Atlantic Australia South Africa All locat. P df P P df
Width of 64.9 6 3.21 68.4 6 3.28 69.5 6 4.60 67.7 6 3.99 70.8 6 3.59 67.7 0.549/ 68 ,0.001/ 0.002/ 89
internal nares 58–75 62–77 60–75 63–73 66–78 58–78 1.000 0.001 0.044
(27)** (34) (33) (22) (7) (12) (115)
Length of 199.8 6 7.16 199.1 6 9.58 199.4 6 8.73 192.9 6 7.47 196.9 6 7.82 198.6 0.861/ 73 0.716/ 0.973/ 91
upper 183–213 178–216 186–220 184–206 185–213 178–220 1.000 1.000 1.000
toothrow (32) (35) (34) (22) (7) (12) (115)
Length of 200.0 6 6.80 195.4 6 9.50 198.4 6 8.58 193.2 6 6.91 195.7 6 6.62 197.4 0.521/ 68 0.030/ 0.867/ 80
lower 186–219 172–217 180–214 184–202 182–207 172–219 1.000 0.635 1.000
toothrow (37) (31) (33) (16) (6) (10) (101)
Length of 356.4 6 9.97 358.6 6 12.29 355.2 6 7.83 353.2 6 12.12 361.0 6 9.85 357.1 0.499/ 69 0.438/ 0.399/ 80
ramus (38) 330–378 335–395 344–371 343–371 347–376 330–395 1.000 1.000 1.000
(32) (33) (15) (6) (10) (101)
Height of 67.2 6 3.69 68.4 6 2.71 65.9 6 3.02 64.0 6 1.79 67.5 6 3.11 67.2 0.156/ 68 0.124/ 0.343/ 83
MARINE MAMMAL SCIENCE, VOL. 19, NO. 3, 2003

ramus (39) 60–77 62–75 60–72 62–67 60–71 60–77 1.000 1.000 0.946
(33) (34) (16) (6) (11) (107)
Width of 3.67 6 0.623 3.93 6 0.604 2.5 3.95 3.50 3.81 0.654/ 46 0.151/ — —
tooth (at 2.4–5.0 2.7–5.0 (1) 3.7–4.5 3.0–4.0 2.4–5.0 1.000 1.000
midlength of (19) (33) (4) (2) (60)
row, transverse
at alveolus)
 Table 2. Continued.

Results of t-tests (with/without

Bonferroni adjustment)
males vs. Phil. vs. WNP vs.
females Japan NA
Measurement Philippines Japan N. Atlantic Australia South Africa All locat. P df P P df
Highest 41.2 6 1.69 41.0 6 2.26 40.7 6 2.20 40.6 6 1.30 40.4 6 1.62 41.0 0.629/ 78 0.647/
upper 37–45 34–45 37–45 39–42 38–43 34–46 1.000 1.000
toothcount (40) (35) (20) (8) (12) (122)
Highest 40.5 6 1.71 40.1 6 2.31 40.4 6 1.71 39.7 6 0.76 39.9 6 1.62 40.4 0.377/ 77 0.374/
lower 37–44 34–47 37–44 39–41 37–43 34–47 1.000 0.747
toothcount (40) (34) (16) (7) (9) (115)
a
Taken at ¼ length of rostrum rather than at 60 mm as in Perrin (1975).
PERRIN ET AL.: GEOGRAPHIC VARIATION IN FRASER’S DOLPHIN
493
494 MARINE MAMMAL SCIENCE, VOL. 19, NO. 3, 2003

Figure 2. Scatterplot of zygomatic width, braincase length, and rostrum width at

midlength for 34 specimens of Lagenodelphis hosei from the Philippines and 35 from Japan.

5 mm greater in the western North Pacific series (without Bonferroni adjustment,

statistically significant at P 5 0.047). Number of teeth did not differ between the
WNP and NA series.
In a discriminant analysis (stepwise, backwards, with priors input) of the four
geographic series WNP, NA, southwestern Indian Ocean (SIO: South Africa, n 5
12), and western South Pacific (WSP: Australia, n 5 7), the first canonical axis
maximized separation between NA and the other series, the second distinguished
between the northern hemisphere (WNP and NA) and southern hemisphere (SIO
and WSP) series, and the third between the South African (SIO) and Australian
(WSP) series. Correct classifications (jackknifed) comprised 89%, 59%, 17%, and
25% for the WNP, NA, SIO, and WSP specimens, respectively, indicating rela-
tively low reliability of the discrimination due to the small sample sizes for South
Africa and Australia.

DISCUSSION
Sexual Dimorphism
Although comparison of quantitative estimates across species must be tentative
because of the relatively small sample size in the present study and the scattered
provenance of the sample compared to those of other studies, dimorphism in cranial
dimensions was similar to that in the spinner dolphin (Stenella longirostris) in the
eastern Pacific: 1.9%–6.8% difference (difference divided by female mean) in five
(22%) of 26 measurements, pooled n 5 77, vs. 1.2%–5.8% in 13 (36%) of 36
measurements, pooled n 5 200 (Douglas et al. 1986) and the pantropical spotted
dolphin (S. attenuata) in the same region: 0.9%–5.9% in 23 (64%) of 36
measurements, pooled n 5 612 (Schnell et al. 1985). Sexual dimorphism in
external appearance is striking in Fraser’s dolphin; the adult male has a distinct
 PERRIN ET AL.: GEOGRAPHIC VARIATION IN FRASER’S DOLPHIN 495

Figure 3. Histogram of canonical scores from discriminant analysis of skulls of

Lagenodelphis hosei from the Philippines (left, n 5 34) and Japan (right, n 5 35).
Crosshatched specimens to left of zero are misclassified Japanese specimens (2) and those to
right of zero are misclassified Philippine specimens (6).

broad lateral black stripe, while this feature is muted in females and juveniles
(Jefferson et al. 1997). While the spinner dolphin is not dimorphic in coloration, it
exhibits marked dimorphism in external shape (dorsal fin and ventral hump) in
some geographic races/subspecies (Perrin 1990). The spotted dolphin is dimorphic
in external size but not in pigmentation or shape (Perrin 1975, Perrin et al. 1985).
Despite this discordant pattern of external dimorphism, all three species are
dimorphic in width and height of the braincase (parietal width and braincase
height) and in dimensions of the temporal fossa; of these two features, the temporal
fossa is consistently the most dimorphic (up to 6.8% difference in height of fossa in
Fraser’s dolphin). The temporal fossa houses the origin of the temporal muscle,
which functions in closing the gape. The male on average has a relatively larger
braincase and more powerful jaw musculature (as indicated by size of the temporal
fossa) than the female in all three species.
Fraser’s dolphin was not here shown to be dimorphic in widths of the rostrum or
preorbital and zygomatic widths, whereas both the spinner dolphin and the
pantropical spotted dolphin were dimorphic in all of these; however, this may be an
artifact of sample size. Fraser’s dolphin was dimorphic in width of the external bony
nares (females had larger nares), while the other two species (based on larger
samples) were not. Of the three species, only the pantropical spotted dolphin was
shown to be dimorphic in rostrum length, postorbital width, and width of the
internal bony nares (sample size was largest for this species—612). Examination of
larger series of L. hosei and the resulting increased statistical power may reveal
slighter sexual dimorphism in additional cranial measurements.
Geographic Variation
As would be predicted for a gregarious and far-ranging pelagic dolphin (Perrin
et al. 1994a), geographic variation within the North Pacific was comparable to that
in the offshore pelagic races of the pantropical spotted dolphin and the spinner
496 MARINE MAMMAL SCIENCE, VOL. 19, NO. 3, 2003

Figure 4. Histogram of canonical scores from discriminant analysis of skulls of

Lagenodelphis hosei from the western North Pacific (left: Philippines, Japan [Honshu and Iwo
Jima], and Taiwan; n 5 71) and the North Atlantic (right: France, Florida, North Carolina,
and St. Vincent; n 5 22).

dolphin in the eastern Pacific (Perrin 1975, Douglas et al. 1992, Perrin et al.
1994b), and it was less than that in the coastal Irrawaddy dolphin (Orcaella
brevirostris) and finless porpoise (Neophocaena phocaenoides) in Southeast Asia (Beasley
et al. 2002, Jefferson 2002). It was also less than between offshore and coastal races
of spotted and spinner dolphins in the eastern Pacific and Southeast Asia (Perrin
1975, 1984, 1998; Perrin et al. 1999).
It could be argued that because the Japanese sample came from a single drive in
the fishery, the sample may not be representative of the local/regional population
but rather of a single school. However, the sample exhibited more, rather than less,
variability than the other geographic samples that were collected over more time
and space; standard deviations were larger than for the other four samples for 15 of
the 25 non-dimorphic cranial measurements and counts (Table 2). The Philippine
sample was collected over a period of years by fishermen operating in dispersed areas
and can be assumed to represent the regional population of the central Philippines
rather than a single school or restricted locality.
The pattern of sharp division between regions within an ocean basin is consistent
with differentiation of local populations. The lack of adequately fine-grained
sampling from across the basin range does not permit us to determine whether the
differences between the present samples reflect differentiation of local/regional
populations or genetic isolation by distance. Nothing is known of the size of home
range in the species. However, a precautionary approach would be to assume that
the dolphins exploited in the different national fisheries (e.g., in the Philippines and
in Japan) come from different populations that should be assessed and managed as
separate stocks.

Future Research
It was not possible in this study to examine material from recent strandings on
the Atlantic coast of South America (e.g., Praderi et al. 1992, Hucke-Gaete 2000,
 PERRIN ET AL.: GEOGRAPHIC VARIATION IN FRASER’S DOLPHIN 497

Di Beneditto et al. 2001). This should be done and comparisons made with the
series from the North Atlantic and the other oceans. Efforts should be made in the
future to collect tissue samples from all takes and strandings, so that genetic
variation can be examined in parallel with studies of morphology (e.g., as has been
done for the bottlenose dolphins (Tursiops spp.; Hersh and Duffield 1990, Mead and
Potter 1990, Curry 1997, Curry and Smith 1997, Hoelzel et al. 1998), spinner
dolphin (Galver 2002), striped dolphin (Stenella coeruleoalba; Archer 1996, Garcı́a-
Martı́nez et al. 1999), Dall’s porpoise (Phocoenoides dalli; Amano and Miyazaki 1992,
1996; Escorza-Treviño and Dizon 2000), finless porpoise (Yoshida et al. 1995,
2001), and humpback whale (Megaptera novaeangliae; Smith et al. 1999).

ACKNOWLEDGMENTS
This project depended for its success on the freely given help of many people in
collecting and preparing specimens, providing access to museum specimens and
unpublished data, obtaining CITES permits and shipping specimens, preparing the
figures, and assisting in other ways too numerous to mention. These include James G. Mead
and Charles W. Potter of the U.S. National Museum of Natural History; Laurie Wilkins of
the Florida Museum of Natural History; Paula Jenkins of the Natural History Museum,
London; Tadasu K. Yamada and Toshiaki Kuramochi of the National Science Museum,
Tokyo; Nobuyuki Miyazaki of the Ocean Research Institute, University of Tokyo, Haruka
Ito of the University of Tokyo; Peter B. Best of the South African Museum; Joan M. Dixon
and Mark Darragh of Museum Victoria; Sandy Ingleby and Tish Ennis of the Australian
Museum; Steve Van Dyck of the Queensland Museum; Noah Cooper of the Western
Australia Museum; Gillian Watson of Bayworld, Natal (formerly Port Elizabeth Museum);
Peter van Bree and Adri G. Rol of the Zoölogisch Museum Amsterdam; Anne Collet of the
Musée Océanographique, La Rochelle; the late William E. Schevill of the Museum of
Comparative Zoology; Roy Allen, Henry Orr, Nick Kellar, and Kelly Robertson of the
Southwest Fisheries Science Center; Scott Serena of the U.S. Fish and Wildlife Service;
Lawrence Garrad and Marvin Ramos of the U.S. Embassy, Manila; Carlos Dolar; Lory Tan of
Bookmark, Inc., Manila; fisherman Isidro Bellamide and his sons Fred, Ramon, Narcisso,
and Dario, daughter Heidi, and wife Oping of Siaton, Philippines; Nida Calumpong,
Moonyeen Alava, Erwin Dolumbal, Esther Himoy, and Adon Gaudiano of Silliman
University Marine Laboratory, Philippines; Dolly Felicitas and Linda Flores of the Silliman
Office, Manila; and others we may have missed. Jay Barlow, James G. Mead, and two
anonymous referees reviewed the manuscript. Support was provided by the Southwest
Fisheries Science Center of the U.S. National Marine Fisheries Service/NOAA, the U.S.
National Science Foundation (Grant 950339 to Gerald Kooyman, Scripps Institution of
Oceanography, University of California, San Diego), the International Foundation for
Science (Grant A/2202-1 to MLLD), the Whale and Dolphin Conservation Society, and the
Department of Environment and Natural Resources (DENR) of the Philippines. Isidro
Bellamide was especially kind in housing and feeding the field party in his home in the
fishing village at Siaton for weeks on end, liaising with the tuna driftnetting boats, and
assisting with handling and preparation of the dozens of specimens collected there; the
project would not have been possible without him.

LITERATURE CITED
AMANO, M., AND N. MIYAZAKI. 1992. Geographic variation and sexual dimorphism in the
skull of Dall’s porpoise, Phocoenoides dalli. Marine Mammal Science 8:240–261.
AMANO, M., AND N. MIYAZAKI. 1996. Geographic variation in external morphology of Dall’s
porpoise, Phocoenoides dalli. Aquatic Mammals 22:167–174.
498 MARINE MAMMAL SCIENCE, VOL. 19, NO. 3, 2003

AMANO, M., N. MIYAZAKI AND F. YANAGISAWA. 1996. Life history of Fraser’s dolphin,
Lagenodelphis hosei, based on a school captured off the Pacific coast of Japan. Marine
Mammal Science 12:199–214.
ARCHER, F. I., II. 1996. Morphological and genetic variation of striped dolphins (Stenella
coeruleoalba, Meyen 1833). Ph.D. dissertation, Scripps Institution of Oceanography,
University of California, San Diego, CA. 185 pp.
BEASLEY, I., P. ARNOLD AND G. HEINSOHN. 2002. Geographical variation in skull morphology
of the Irrawaddy dolphin, Orcaella brevirostris (Owen in Gray, 1866). Raffles Bulletin of
Zoology, Supplement 10:15–34.
BÉRUBÉ, M., AND P. J. PALSBØLL. 1996. Identification of sex in cetaceans by multiplexing
with three ZFX and ZFY specific primers. Molecular Ecology 5:283–287.
vAN BREE, P. J. H., A. COLLET, G. DESPORTES, E. HUSSENOT AND J. A. RAGA. 1986. Le
dauphin de Fraser, Lagenodelphis hosei (Cetacea, Odontoceti), espèce nouvelle pour la
faune d’Europe. Mammalia 50:57–86.
CALDWELL, D. K., M. C. CALDWELL AND R. V. WALKER. 1976. First records for Fraser’s
dolphin (Lagenodelphis hosei) in the Atlantic and the melon-headed whale (Peponocephala
electra) in the western Atlantic. Cetology 25:1–4.
CURRY, B. E. 1997. Phylogenetic relationships among bottlenose dolphins (genus Tursiops) in
a worldwide context. Ph.D. dissertation, Texas A&M University, College Station, TX.
137 pp.
CURRY, B. E., AND J. SMITH. 1997. Phylogeographic structure of the bottlenose dolphin
(Tursiops truncatus): Stock identification and implications for management. Pages
227–247 in A. E. Dizon, S. J. Chivers and W. F. Perrin, eds. Molecular genetics of
marine mammals. Special Publication 3, Society for Marine Mammalogy, Lawrence, KS.
DI BENEDITTO, A. P. M., R. M. ARRUDA RAMOS, S. SICILIANO, R. AGUIAR DOS SANTOS,
G. BASTOS AND E. FAGUNDES-NETTO. 2001. Stomach contents of delphinids from Rio de
Janeiro, southeastern Brazil. Aquatic Mammals 27:24–28.
DOLAR, M. L. L. 1994. Incidental takes of small cetaceans in fisheries in Palawan, central
Visayas and northern Mindanao in the Philippines. Reports of the International
Whaling Commission (Special Issue 15):355–363.
DOLAR, M. L. L. 1999. Abundance, distribution and feeding ecology of small cetaceans in the
eastern Sulu Sea and Tañon Strait, Philippines. Ph.D. dissertation, Scripps Institution
of Oceanography, University of California, San Diego, CA. 241 pages.
DOLAR, M. L. L. 2002. Fraser’s dolphin Lagenodelphis hosei. Pages 485–487 in W. F. Perrin,
B. Würsig and J. G. M. Thewissen. 2002. Encyclopedia of marine mammals. Academic
Press, San Diego, CA.
DOLAR, M. L. L., S. LEATHERWOOD, C. WOOD, M. N. R. ALAVA, C. HILL AND L. V. ARAGONES.
1994. Directed fisheries for cetaceans in the Philippines. Report of the International
Whaling Commission 44:439–449.
DOUGLAS, M. E., G. D. SCHNELL AND D. J. HOUGH. 1986. Variation in spinner dolphins
(Stenella longirostris) from the eastern tropical Pacific Ocean: Sexual dimorphism in
cranial morphology. Journal of Mammalogy 67:537–544.
DOUGLAS, M. E., G. D. SCHNELL, D. J. HOUGH AND W. F. PERRIN. 1992. Geographic
variation in cranial morphology of spinner dolphins Stenella longirostris in the eastern
tropical Pacific Ocean. Fishery Bulletin, U.S. 90:54–76.
ESCORZA-TREVIÑO, S., AND A. E. DIZON. 2000. Phylogeography, intraspecific structure and
sex-biased dispersal of Dall’s porpoise, Phocoenoides dalli, revealed by mitochondrial and
microsatellite DNA analyses. Molecular Ecology B9:1049–1060.
FRASER, F. C. 1956. A new Sarawak dolphin. Sarawak Museum Journal 7:478–503.
GALVER, L. M. 2002. The molecular ecology of spinner dolphins, Stenella longirostris: Genetic
diversity and population structure. Ph.D. dissertation, Scripps Institution of Oceano-
graphy, University of California, San Diego, CA. 192 pp.
GARCÍA-MARTÍNEZ, J., A. MOYA, J. A. RAGA AND A. LATORRE. 1999. Genetic differentiation
in the striped dolphin Stenella coeruleoalba from European waters according to
 PERRIN ET AL.: GEOGRAPHIC VARIATION IN FRASER’S DOLPHIN 499

mitochondrial DNA (mtDNA) restriction analysis. Molecular Ecology 8:1069–

1073.
GERRODETTE, T. 2002. Tuna-dolphin issue. Pages 1269–1273 in W. F. Perrin, B. Würsig
and J. G. M. Thewissen, eds. Encyclopedia of marine mammals. Academic Press,
San Diego, CA.
HERSH, S. L., AND D. A. DUFFIELD. 1990. Distinction between northwest Atlantic offshore
and coastal bottlenose dolphins based on hemoglobin profile and morphometry. Pages
129–139 in S. Leatherwood and R. R. Reeves, eds. The bottlenose dolphin. Academic
Press, San Diego, CA.
HERSH, S. L., AND D. K. ODELL. 1986. Mass stranding of Fraser’s dolphin, Lagenodelphis hosei,
in the western North Atlantic. Marine Mammal Science 2:73–76.
HOELZEL, A. R., C. W. POTTER AND P. B. BEST. 1998. Genetic differentiation between
parapatric ‘‘nearshore’’ and ‘‘offshore’’ populations of the bottlenose dolphin. Pro-
ceedings of the Royal Society of London 265:1177–1183.
HUCKE-GAETE, R. 2000. Review of the conservation status of small cetaceans in southern
South America. UNEP/CMS Secretariat, Bonn, Germany.
JEFFERSON, T. A. 1994. Lagenodelphis hosei. Mammalian Species 470:1–5.
JEFFERSON, T. A. 2002. Preliminary analysis of geographic variation in cranial morpho-
metrics of the finless porpoise (Neophocaena phocaenoides). Raffles Bulletin of Zoology,
Supplement 10:3–14.
JEFFERSON, T. A., S. LEATHERWOOD AND M. A. WEBBER. 1993. FAO species identification
guide. Marine mammals of the world. FAO, Rome.
JEFFERSON, T. A., R. L. PITMAN, S. LEATHERWOOD AND M. L. L. DOLAR. 1997. Developmental
and sexual variation in the external appearance of Fraser’s dolphins (Lagenodelphis hosei).
Aquatic Mammals 23:145–153.
LEDUC, R. G., W. F. PERRIN AND A. E. DIZON. 1999. Phylogenetic relationships among the
delphinid cetaceans based on full cytochrome b sequences. Marine Mammal Science
15:619–648.
MEAD, J. G., AND C. W. POTTER. 1990. Natural history of bottlenose dolphins along the
central Atlantic coast of the United States. Pages 165–195 in S. Leatherwood and R. R.
Reeves, eds. The bottlenose dolphin. Academic Press, San Diego, CA.
MICROSOFT CORPORATION. 1999. Microsoft Office 2000 Professional. Microsoft Corporation,
Seattle.
PERNEGER, T. V. 1998. What’s wrong with Bonferroni adustments? British Medical Journal
316:1236–1238.
PERRIN, W. F. 1975. Variation in spotted and spinner porpoise (genus Stenella) in the
eastern Pacific and Hawaii. Bulletin of the Scripps Institution of Oceanography 21:
1–206.
PERRIN, W. F. 1984. Patterns of geographical variation in small cetaceans. Acta Zoologica
Fennica 172:137–140.
PERRIN, W. F. 1990. Subspecies of Stenella longirostris (Mammalia: Cetacea: Delphinidae).
Proceedings of the Biological Society of Washington 103:453–463.
PERRIN, W. F. 1998. Stenella longirostris. Mammalian Species 599:1–7.
PERRIN, W. F., AND J. E. HEYNING. 1993. Rostral fusion as a criterion of cranial
maturity in the common dolphin, Delphinus delphis. Marine Mammal Science 9:195–
197.
PERRIN, W. F., P. B. BEST, W. H. DAWBIN, K. C. BALCOMB, R. GAMBELL AND G. J. B. ROSS.
1973. Rediscovery of Fraser’s dolphin Lagenodelphis hosei. Nature 241:345–350.
PERRIN, W. F, M. D. SCOTT, G. JAY WALKER AND V. L. CASS. 1985. Review of geographical
stocks of tropical dolphins (Stenella spp. and Delphinus delphis) in the eastern Pacific.
NOAA Technical Report NMFS 28:1–28.
PERRIN, W. F., S. LEATHERWOOD AND A. COLLET. 1994a. Fraser’s dolphin Lagenodelphis hosei
Fraser, 1956. Pages 225–240 in S. H. Ridgway and R. Harrison, eds. Academic Press,
San Diego, CA.
500 MARINE MAMMAL SCIENCE, VOL. 19, NO. 3, 2003

PERRIN, W. F., G. D. SCHNELL, D. J. HOUGH, J. W. GILPATRICK, JR. AND J. V. KASHIWADA.

1994b. Re-examination of geographic variation of the pantropical spotted dolphin,
Stenella attenuata, in the eastern Pacific. Fishery Bulletin, U.S. 92:324–346.
PERRIN, W. F., M. L. L. DOLAR AND D. ROBINEAU. 1999. Spinner dolphins (Stenella
longirostris) of the western Pacific and Southeast Asia: pelagic and shallow-water forms.
Marine Mammal Science 15:1029–1053.
PRADERI, R., G. PRADERI AND R. GARCÍA. 1992. First record of Fraser’s dolphin, Lagenodelphis
hosei, in the South Atlantic Ocean (Mammalia: Cetacea: Delphinidae). Comunicaciones
Zoológicas del Museo de Historia Natural de Montevideo 178:1–6.
SCHNELL, G. D., M. E. DOUGLAs AND D. J. HOUGH. 1985. Sexual dimorphism in spotted
dolphins (Stenella attenuata) in the eastern tropical Pacific Ocean. Marine Mammal
Science 1:1–14.
SMITH, T. D., J. ALLEN, P. J. CLAPHAM, P. S. HAMMOND, S. KATONA, F. LARSEN, J. LIEN,
D. MATTILA, P. J. PALSBØLL, J. SIGORJØNSSON, P. T. STEVICK AND N. ØIEN 1999. An ocean-
basin-wide mark-recapture study of the North Atlantic humpback whale (Megaptera
novaeangliae). Marine Mammal Science 15:1–32.
SPSS. 1999. SYSTAT 9. SPSS Incorporated, Chicago, IL.
WADE, P. R., AND T. GERRODETTE. 1993. Estimates of cetacean abundance and distribution
in the eastern tropical Pacific. Report of the International Whaling Commission 43:
477–493.
YOSHIDA, H., K. SHIRAKIHARA, M. SHIRAKIHARA AND A. TAKEMURA. 1995. Geographic
variation in the skull morphology of the finless porpoise Neophocaena phocaenoides in
Japanese waters. Fisheries Science 61:555–558.
YOSHIDA, H., M. YOSHIOKA, M. SHIRAKIHARA AND S. CHOW. 2001. Population structure
of finless porpoises (Neophocaena phocaenoides) in coastal waters of Japan based on
mitochondrial DNA sequences. Journal of Mammalogy 82:123–130.
Received: 8 April 2002
Accepted: 3 November 2002

APPENDIX 1
Specimens Examined
Eastern tropical Pacific: USNM 396079, 504411, 550022; MCZ 52979, 54379; SWFSC
60 (LGP179). Negros Oriental and Bohol, Philippines: SUML 7, 8, 19–31, 35, 52, 55–64,
121, 126, 137, 138, 705–712, 725, 726, field nos. HAB12, HAB24, WFP813 (not yet
assigned museum numbers); field nos. WFP785, 801, 806 (specimens not saved).
Kamogawa, Japan: NSMT M24904. Taiji, Japan: NSMT M29552–29585. Iwo Jima,
Japan: NSMT M24920. Taiwan: NSMT M24825. North of Papua, New Guinea (18359N,
1428049E): NSMT M24921. Lamalera, Indonesia: WAM M16294. Ambon, Indonesia:
ZMA 24.777. Sarawak, Malaysia: BMNH 1895.5.9.1 (holotype specimen). Queensland,
Australia: QM JM2749. New South Wales, Australia: AM M22837, 34166, 34299.
Victoria, Australia: NMV C24959, 24992, 24993. Western Australia: WAM M25789. Cape
Province, South Africa: SAM 36322, 36323, 40925. Natal, South Africa: PEM N341, 395,
675, 740, 827, 831, 959, 1773, 1958, 2179. North Carolina, USA: USNM 571619.
Marquesas Keys, Florida, USA: FMNH(UF) 25844–25853. Brittany, France: CRMM 1685–
1687, 1689–1692, 1694; ZMA 22.680.

APPENDIX 2
Museum and Collection Acronyms
AM, Australian Museum, Sydney; BMNH, Natural History Museum, London (formerly
British Museum [Natural History]); CRMM, Centre National d’Etude des Mamifères
Marins, Musée Océanographique, La Rochelle, France; FMNH(UF), Florida Museum of
 PERRIN ET AL.: GEOGRAPHIC VARIATION IN FRASER’S DOLPHIN 501

Natural History, University of Florida, Gainesville; MCZ, Museum of Comparative Zoology,

Cambridge, Massachusetts; NMV, Museum Victoria, Victoria, Australia; NSMT, National
Science Museum, Tokyo; PEM, Bayworld, Natal, South Africa (formerly Port Elizabeth
Museum); QM, Queensland Museum, Brisbane; SAM, South African Museum, Cape Town;
SUML, Silliman University Marine Laboratory, Dumaguete, Philippines; SWFSC,
Southwest Fisheries Science Center, La Jolla, California; USNM, U.S. National Museum
of Natural History, Washington, D.C.; WAM, Western Australia Museum, Perth; ZMA,
Zoölogisch Museum, Amsterdam.