COMMON DOLPHINS DELPHINUS DELPHIS IN SOUTHERN AUSTRALIA: A MORPHOMETRIC STUDY

Transcription

1 COMMON DOLPHINS DELPHINUS DELPHIS IN SOUTHERN AUSTRALIA: A MORPHOMETRIC STUDY CATHERINE H. BELL, CATHERINE M. KEMPER AND JOHN G. CONRAN Bell CH, Kemper CM and Conran JG, Common dolphins Delphinus delphis in southern Australia: a morphometric study. Australian Mammalogy 24: Examination of 211 Delphinus specimens from the coasts of Western Australia to New South Wales, including Tasmania, was conducted using 62 quantitative and 11 qualitative variables. After refining the dataset, multivariate analyses were performed on 130 cranially mature specimens using 21 cranial variables. MANOVA showed males to be slightly larger than females, but with substantial overlap, allowing analyses to combine genders. UPGMA Cluster Analysis and MDS Ordination showed three largely overlapping groups based on a size gradient. K-means analysis of these groups found no significant differences and confirmed a size gradient. Discriminant analysis of specimens grouped by geography and water depth showed a tendency for large skulls to be from coasts adjacent to deep water and small skulls from shallow water coasts. Cranial measurements were significant, postcranial measurements and features were not. Tooth counts were within the range for D. delphis for all specimens examined. This study confirms genetic evidence for a single continuously variable species (Delphinus delphis Linnaeus, 1758) in southern Australian waters. Compared with either D. delphis or D. capensis from the eastern North Pacific, the skulls of D. delphis in southern Australia were more variable for many characters. Key words: common dolphin, Delphinus delphis, skull, cranial, geographic variation, southern Australia. CH Bell, PO Box 3454, BMDC, Belconnen, ACT 2617, Australia. catherine_hobart_bell@hotmail.com. CM Kemper, South Australian Museum, North Terrace, SA 5000, Australia. JG Conran, Department of Environmental Biology, University of Adelaide, Adelaide, SA 5005, Australia. Manuscript received 4 September 2001; accepted 8 June COMMON dolphins (Delphinus spp.) are in need of a worldwide systematic review due to their cosmopolitan distribution and geographic variability resulting in over 20 nominal species (Evans 1994). Heyning and Perrin (1994) studied the genus in the eastern North Pacific and reviewed the nomenclature for Delphinus, finding evidence for two species, both sexually dimorphic: Delphinus delphis, the short-beaked common dolphin, and Delphinus capensis, the long-beaked common dolphin. A parallel genetic study (Rosel et al. 1994) also found evidence for two species in that region. D. delphis was found both inshore and offshore, while D. capensis was mostly coastal (Heyning and Perrin 1994). They suggested that worldwide there were two species, with the possibility of a third, D. tropicalis, an extra long-beaked form that could be part of a cline of D. capensis in the Indian Ocean. Jefferson (in press) investigated the taxonomy of D. tropicalis, suggesting that the tropicalis-form is a long-beaked subspecies of D. capensis. His investigation found evidence for clinal variation, with hybridisation or integration between the tropicalis-form and D. capensis possibly occurring in South-east Asia and along the east coast of Africa. Jefferson describes the subspecies as D. capensis tropicalis. A pilot study by Kemper and Gibbs (1997) noted morphological variation among the Delphinus skulls from South Australia (SA). However, as the number of specimens examined in that study was small and restricted to SA, a more in-depth study was required into the nature of the cranial variation and possible taxonomic implications. A genetic study of Delphinus from southern Australia using the mitochondrial DNA control region and cytochrome b (White 1999) found no strong genetic evidence for two species and, in contrast to the results of Rosel et al. (1994), none for two species worldwide. White (1999) suggested the cranial variation noted by Kemper and Gibbs (1997) may be due to resource polymorphism. The present study investigated the nature of the differences exhibited in Delphinus skulls and skeletons from southern Australian waters, and compared them with the findings of Heyning and Perrin (1994). The study investigated the

2 2 AUSTRALIAN MAMMALOGY possibilities of sexual dimorphism, extent and nature of the morphological variation, and trends associated with geographical/oceanographic regions. MATERIALS AND METHODS Skulls and skeletons of 211 specimens from the coasts of Western Australia to New South Wales, including Tasmania, were examined (Fig. 1, Appendix 1). Specimens were obtained from live strandings, entanglements and beach-washed carcasses collected between 1879 and mid-2000 (mostly since 1985 in SA). The total sample contained 122 complete and 89 incomplete skulls. A total of 62 quantitative (47 cranial and 15 postcranial measurements) and 11 qualitative variables for each specimen (Fig. 2, Appendix 2) were chosen based on the following studies: tympanic bulla and periotic bone measurements (Kasuya 1973); cranial measurements (Ross 1984; Heyning and Perrin 1994; Kemper and Bell unpubl. data) and postcranial measurements (Benke 1993). As a convention the left side was measured, except where features were asymmetrical and both sides were measured (see Amano et. al. 2000). All measurements were made by Bell using metal vernier callipers (< 310 mm) or spreading callipers and retractable metal tape measure (> 310 mm). Where the measurement was not in a straight line, as for the length of the top perimeter of the skull (TopPer, Fig. 2) a 0.5 mm wide, flexible nylon tape measure was used. In order to test measurement accuracy, repetitions of measurements 120 were made on several specimens until a level of precision of ± 0.5 mm was obtained. Cranial maturity was identified in 165 specimens. Cranial maturity was assessed by noting the degree of maxillary and premaxillary fusion (see Appendix 2). Skulls rated 3 or 4 were considered cranially mature (Perrin and Heyning 1993). Where available, epiphyseal fusion of the vertebrae and tooth development were used to assist in relative age of specimens. Data were analysed using PATN (Belbin 1994) with Gower s general coefficient of similarity and unweighted pair group mean association clustering (UPGMA). Gower s general coefficient of similarity allows for missing data and mixed character datasets, and is generally preferred for taxonomic studies (Gower 1967; Podini 1999). Due to the amount of missing data from incomplete specimens and the large number of measurements, the dataset was refined using Kruskal-wallis box whisker plots by PATN. Only variables found to differ significantly (α ) between the UPGMA dendrogram groups were included in subsequent analyses. All postcranial and qualitative measurements (Appendix 2) were therefore excluded, while many skull measurements were significantly variable (Fig. 2). The dataset was further refined by the elimination of either variables with data missing for several specimens or specimens with data missing for several variables, resulting in a complete dataset of 130 cranially mature specimens and 21 cranial variables. 140 WA SA NSW WA SA1 Vic. NSW 40 S N Deep Water Shallow Water Kilometres SA2 TAS BS Tas E 140 Fig. 1. Geographical zones and locations for Delphinus specimens from southern Australia. Geographical zones (geo-zone) codes as follows BS: Bass Strait (Victorian coast and north coast Tasmania, King Is. and Flinders Is.); NSW: south and central New South Wales coast; SA1: Great Australian Bight, South Australian gulfs and the Coorong; SA2: South East South Australia and western Victoria; TAS: Tasmanian coast; WA: southern Western Australia. Symbols may represent more than one specimen.

3 BELL ET AL.: COMMON DOLPHINS IN SOUTHERN AUSTRALIA Fig. 2 Measurements for D. delphis skulls. Measurement codes shown in parentheses, those used in the analyses are underlined. 1 Condylobasal length (CBL); 2 Rostrum length (RL); 3 Distance from external nares to rostrum tip (ExN-RT); 4 Rostrum width base (RWB); 5 Rostrum width at 60 mm anterior (RW60); 6 Rostrum width at midlength (RWM); 7 Rostrum width at ¾ length (RW3/4); 8 Premaxilla width at midlength (PMWM); 9 Greatest premaxilla width (GPMW); 10 Greatest preorbital width (PreOrbW); 11 Greatest postorbital width (PostOrbW); 12 Zygomatic width (ZW); 13 Greatest parietal width (PW); 14 Least width braincase parietal borders (BCW); 15 Length top perimeter skull (TopPer); 16 External height braincase (BCH); 17 Internal length braincase (BCL); 18 Greatest length temporal fossa (PTFL); 19 Greatest height temporal fossa (PTFH); 20 Minor diameter temporal fossa (left) (TFMin) - not shown; 21 Major diameter temporal fossa (left) (TFMaj) - not shown; 22 Length orbit (left) (OL); 23 Length antorbital (lacrimal) (AOL); 24 Greatest width overhang supraoccipital crest (SOC) - not shown; 25 Greatest length pterygoids (PtL); 26 Greatest width internal nares (InNW); 27 Greatest width basioccipital (BOW); 28 Length upper tooth row to rostrum tip (UTR-T); 29 Length upper tooth row (UTR); 30 Length mandible tooth row to tip (ManTR-T); 31 Length mandible (ManL); 32 Length mandibular tooth row (ManTR) - not shown; 33 Length mandibular fossa (ManFL); 34 Depth mandible (ManD); 35 Length mandibular symphysis (ManSL) - not shown; 36 Greatest width bulla (BulW) - not shown; 37 Greatest length bulla (BulL) - not shown; 38 Greatest width periotic (PerW) - not shown; 39 Greatest length periotic (PerL) - not shown; 40 Tooth diameter (TD) not shown; 41 Distance nasals to left (NasL) - not shown; 42 Length right nasal (RNasL); 43 Width right nasal (RNasW); 44 Length left nasal (LNasL); 45 Width left nasal (LNasW); 46 Premaxillary asymmetry: greatest width left (PAWL); 47 Premaxillary asymmetry: greatest width right (PAWR). Top left C29589; top right C29590; bottom left M27971; middle M27971, bottom right M140. Sexual dimorphism was tested using MANOVA with known-sex specimens (37 females, 42 males) in SYSTAT10 (SPSS 2000). Cluster analysis was performed using PATN to produce an association matrix, which was used to create a UPGMA dendrogram and an ordination using multidimensional scaling (MDS). This allowed dendrogram groups to be determined and the relationships between the data points, groups and associated character states to be visualised in two dimensional space on the ordination plot. K-means analysis to examine group structure and discriminant analysis of geographical/ oceanographic zones and relative water depth were performed in SYSTAT10. Specimens were allocated to a geographical/oceanographic zone (geo-zone) based on coastal divisions from shore to the continental slope (ACIUCN 1986). A relative water depth category was also allocated to specimens (Fig. 1). Specimens were deep-water if located on a coast with a narrow continental shelf (WA, SA2, TAS and NSW) and shallow-water if located on coast adjacent to a broad continental shelf (SA1 and BS). There were more specimens from SA1 than any other geo-zone. RESULTS A MANOVA found considerable statistical overlap between males and females (classification accuracy 78%, for α0.002: P = ; Mahalanobis distance F = 2.200, Wilks Λ 0.552, Pillai s Trace and Lawley-Hotelling Trace 0.810). The biggest differences (> 5%) in average size occurred in width measures, particularly those of the rostrum

4 4 AUSTRALIAN MAMMALOGY (RW3/4, RWM, RW60). All major length measures (CBL, ExN-RT, RL) differed by < 2%. With an overall mean difference of 3.2%, females were only slightly smaller than the males and so all cranially mature specimens, of known and unknown sex, were combined in the remaining analyses. UPGMA cluster analysis produced two weaklydivided groups (larger versus smaller specimens) and MDS ordination placed these with the group representing the smaller specimens closer to the origin (Fig. 3). The group containing the larger specimens was divided in the ordination plot based on shape, with some specimens having larger rostrum measurements, but most individuals having relatively shorter rostrum measurements and generally larger braincase features. The stress level for the MDS (which measures the degree of visual distortion of inter-point distances) was 14.8%; the acceptable limit is up to 20% (Belbin 1994) MDS Axis RL ExN-RT RW3/4 CBL RW60 RWM RWB Stress = 14.8% OL PTFL UTR BCH PTFH AOL PW PMWM PreorbW PostorbW BCW ZW GPMW TopPer MDS Axis 1 Fig. 3. Ordination plot for the relationship between the two groups (closed and open circles) of southern Australian Delphinus, as found by UPGMA cluster analysis. Character codes (Fig. 2) show direction of significance for each character. MDS = Multi Dimensional Scaling The ordination plot (Fig. 3) suggested that there may be three morphological groups in southern Australia, but K-means analysis showed no significant difference between them. For each group, each variable average was highest in cluster 1 and lowest in cluster 3 (Table 1). These groups overlapped considerably, so in reality were one morphological group with continuous variation. The strongest variables (F-ratio > 200) were braincase width measurements (Table 1). Condylobasal length and several rostrum measurements were also strong (F-ratio between 93 and 120). The remaining cranial variables all had F-ratios < 70. Comparisons of measurements between geozones using discriminant analysis detected significant differences (P < ; Wilks Λ 0.119; Pillai s Trace 1.160; Lawley-Hotelling Trace 2.995) with % classification accuracy of specimens based on location. The largest difference (Mahalanobis distance = 4.203) was between TAS and SA1. Braincase measurements (BCW, BCH, TopPer) ranked highly, as did some rostral and premaxillary width measurements (GPMW, RW60, RWB). The geo-zone discriminant analysis scores revealed a tendency for geo-zones near deep water to group towards the left of Fig. 4 and those near shallow water tended towards the right. Investigation using discriminant analysis of relative water depth showed that larger specimens tended to be found near deep water areas and smaller specimens near shallow water areas (P < ; Wilks Λ 0.494; Pillai s Trace 0.506; Lawley-Hotelling Trace 1.023). The discriminant score (Fig. 5) for relative water depth analysis showed two peaks, with the shallow water peak occurring at discriminant score (DS) = -2 and the deep water peak at DS = 2. The between-groups F- matrix gave a Mahalanobis distance of and width measurements (GPMW, BCW) ranking higher than length. As with the geo-zone analysis, braincase measurements (BCW, BCH) ranked highly. Specimens were classified accurately in 88% of cases. Tooth counts for southern Australian Delphinus were in the range for the upper left, with a mean of Examples of the extremes for shape and size of the skull are found in Fig. 6, the posterior view of the skull being described as umbrella and square by Kemper and Gibbs (1997). The umbrella-shape example (M18083), with a short and broad rostrum, was from SA2 (near deep water). The square-shape example (M18042), with a long and narrow rostrum, was from SA1 (near shallow water). M18083 was a more robust skull in comparison with M18042, with an overall perception of being larger despite being the same length. DISCUSSION The results of the present study showed that southern Australian Delphinus belong to a single morphologically-variable species. White s (1999) genetic study concluded that common dolphins from the same region were D. delphis. Morphologically there is a greater range of variation within southern Australian D. delphis than in either D. delphis or D. capensis from the eastern North Pacific (Tables 2a, 2b)

5 BELL ET AL.: COMMON DOLPHINS IN SOUTHERN AUSTRALIA 5 K-means Cluster Means Character F-ratio Cluster 1 (n=39) Cluster 2 (n=51) Cluster 3 (n=40) PostOrbW PreOrbW ZW CBL ExN-RT RWB RL RW Table 1. Cluster means for K-means analysis. Data have been range standardised. Refer to Fig. 2 for character codes. Only characters with an F-ratio > 90 are shown Discriminant Function Score BS NSW SA1 SA2 TAS WA Fig. 4. Discriminant analysis scores for Delphinus specimens assigned to geo-zones (refer to Fig. 1). Lines, drawn by hand, show approximate boundary for each geo-zone group. Based on skull morphology, no sexual dimorphism was found for the specimens from southern Australia. This finding differs from that of Heyning and Perrin (1994) who found that females were significantly smaller than males in the eastern North Pacific for both species. In the eastern North Pacific Discriminant Function Score 1 the average male body length is 5% longer than female body length and for many cranial variables males were on average > 10% larger than females. The maximum difference between male and female specimens from southern Australia was 7.4% (RW3/4), with an average of 3.2%. Wang et al.

6 6 AUSTRALIAN MAMMALOGY (2000) found no sexual dimorphism for Tursiops spp. (bottlenose dolphins) in Chinese waters, and reported that for different regions around the world the presence of sexual dimorphism is inconsistent for this genus. It is likely that for Delphinus the degree and significance of sexual dimorphism will vary geographically. Delphinus specimens from areas close to the continental slope (NSW, WA, TAS, SA2), and hence close to deep water, tended to be larger than those from near shallow water (SA1, BS). This size gradient could be correlated to water temperature (Ross and Cockcroft 1990). The deep-water areas are colder than the shallow- water areas, which 30 No. of animals deep water shallow water Discriminant Score Fig. 5. Discriminant analysis scores for Delphinus specimens from geo-zones (refer to Fig. 1) of different water depths. Southern Australia Character Deep water Shallow water Combined Rostrum length ± ( ) n = ± ( ) n = ± ( ) n = 130 Zygomatic width ± ( ) n = ± 9.25 ( ) n = ± ( ) n = 130 Table 2a. Mean ± standard deviation (range) and sample size for two skull measurements (mm) for cranially mature Delphinus specimens from southern Australian waters. Eastern North Pacific Males Females Character D. delphis D. capensis D. delphis D. capensis Rostrum length ± ( ) n = ( ) n = ± ( ) n = ( ) n = 12 Zygomatic width ± 5.27 ( ) n = ( ) n = ± 4.91 ( ) n = 51 Table 2b. Delphinus measurements (mm) for eastern North Pacific (Heyning and Perrin 1994) ( ) n = 13

7 BELL ET AL.: COMMON DOLPHINS IN SOUTHERN AUSTRALIA 7 A B C D Fig. 6. Extremes of skull variation. Scale bars = 10 cm. Both skulls are from adult females, housed in the South Australian Museum. A, posterior view M18083; B, posterior view M C, dorsal view M18083; D, dorsal view M M18083 collected 7 Nov 1994 from 2 km NE South End, SA S E. M18042, collected 7 Mar 1994 from 7 km SE Goolwa, SA S E. suggests that the tendency for larger specimens in deep waters could be an adaptive feature for cold. Ross and Cockcroft (1990) found evidence to suggest that bottlenose dolphins in Australia exhibited variation in accordance to Bergman s Rule i.e., that there is an optimum body size for the average temperature the population is exposed to and one would expect a larger body size in colder climates. The tendency for larger Delphinus to be found in deeper, and possibly colder, waters is strongly suggestive of a variation in accordance with Bergman s rule because the surface area to volume ratio is lower in cold water so that heat is retained. When compared with Heyning and Perrin s study (1994) (Tables 2a, 2b), the range of several variables measured in southern Australian Delphinus almost spans the ranges of both species found in the eastern North Pacific. The width variables of the cranium were the more influential, rather than length of the rostrum, as Heyning and Perrin found. For D. delphis, Heyning and Perrin found the ratio of rostrum length to zygomatic width was 1.37 (range: ) for males and 1.36 (range: ) for females. For D. capensis, they found 1.60 (range: ) for males and 1.64 (range: ) for females. Southern Australian Delphinus had a mean ratio of 1.52 (range ), which was just within the range reported by Heyning and Perrin (1994) for D. capensis males. If specimens from near deep and shallow water are considered separately, the mean is still 1.52 for both. The range for this ratio is greater for southern Australian Delphinus than in either species reported for the eastern North Pacific. The ratio of rostral length to zygomatic width for D. tropicalis was reported as 2.06 by van Bree and Gallagher (1978, as cited in Heyning and Perrin, 1994). Jefferson (in press) reports this ratio as 1.85 (range ). Evans (1994) showed that the potential third nominal species, D. tropicalis, had a rostrum length > 320 mm, which is outside the range for Delphinus in both southern Australia and the eastern North Pacific (Tables 2a, 2b). However, Jefferson (in press) reports the length of the rostrum for tropicalis-form to range mm (mean mm), which is on average greater in length than for capensis reported by both Jefferson and Heyning and Perrin, but with overlap, and outside the range for delphis as reported by Heyning and Perrin and in the present study.

8 8 AUSTRALIAN MAMMALOGY Tooth counts in southern Australian Delphinus ranged (mean 47.9) per upper tooth row, which is in the range for eastern North Pacific delphis but not capensis. Heyning and Perrin (1994) report tooth counts per upper tooth row to range (mean 49) for D. Delphis and (mean 53) for D. capensis. Jefferson found tooth counts for the tropicalis-form to range (mean 59.5) per upper tooth row, and the capensis-form to range from (mean 53.9). The ecological findings for southern Australian specimens are different from those that Heyning and Perrin (1994) described from eastern North Pacific, in that the larger animals in southern Australia were found near deep water, whereas in the eastern North Pacific the opposite was true. Mayr (1996) describes Multidimensional Species Taxa as conspecific and widespread species consisting of several different (polytypic) populations. D. delphis is a widespread species (Evans 1994) so it may be that eastern North Pacific and southern Australian animals are conspecific. Heyning and Perrin (1994:18) separated Delphinus based on complete morphological separation and the fact that the populations they studied occurred sympatrically over part of their ranges. The tendency for larger skulls to be found near deep water may agree with Bergman s Rule but more fresh carcasses from deep-water specimens are required to determine whole body size. External measurements and postcranial skeletons were not available for many deep-water classified specimens so these could not be investigated in the present study. Due to the environmental influence on the phenotype and apparent geographical variation, a long-term effort into the collection of Delphinus both southern and northern Australian specimens, would provide useful data for further comparison of morphological variation. The continued collection of data, either through dedicated study or opportunistic research, for use in colour pattern and external measurement analysis would allow further investigation into the variation found by the present study. ACKNOWLEDGEMENTS Russell Baudinette, Head of Department of Environmental Biology, University of Adelaide for his support; Norah Cooper and the Western Australian Museum; Sandy Ingleby and The Australian Museum; Joan Dixon and the Museum of Victoria; Debby Robertson, David Pemberton and the Tasmanian Museum and Art Gallery; Stephen Donnellan for his assistance with the genetics of Delphinus; Tom Jefferson for information regarding the taxonomy of D. tropicalis; Adam Bell for preparing the diagrams in Fig. 2 and the financial assistance of Sea World Research and Rescue Foundation, A.R. Riddle Scholarship, Walter and Dorothy Duncan Trust Grant, and Adelaide Honours Scholarship. REFERENCES ACIUCN (Australian Committee for the International Union for Conservation of Nature and Natural Resources), Australia's Marine and Estuarine Areas - A Policy for Protection. Occasional Paper No. 1. AMANO M, ITO H AND MIYAZAKI N, Geographic and temporal comparison of skulls of striped dolphins off the Pacific coast of Japan. Journal of Cetacean Research and Management 2: BELBIN L, PATN Pattern Analysis Package. CSIRO Division of Wildlife and Ecology: Canberra. BENKE H, Investigations on the osteology and the functional morphology of the flipper of whales and dolphins (Cetacea). Investigations on Cetacea XXIV: VAN BREE PJH AND GALLAGHER MD, On the taxonomic status of Delphinus tropicalis van Bree, 1971 (notes on Cetacea, Delphinoidea IX). Beaufortia 28:1-8. EVANS WE, Common dolphin, white-bellied porpoise Delphinus delphis Linnaeus, Pp in Handbook of marine mammals Volume 5: The first book of dolphins ed by S.H. Ridgeway and R. Harrison. Academic Press: London. GOWER JC, A comparison of some methods of cluster analysis. Biometrics 23: HEYNING JE AND PERRIN WF, Evidence for two species of common dolphins (genus Delphinus) from the eastern North Pacific. Contributions in Science 442: JEFFERSON TA, in press. The taxonomic status of the nominal dolphin species Delphinus tropicalis van Bree, Marine Mammal Science. KASUYA T, Systematic consideration of recent toothed whales based on the morphology of tympano-periotic bone. Scientific Reports of the Whales Research Institute 25: KEMPER CM AND GIBBS SE, A study of life history parameters of dolphins and seals entangled in tuna farms near Port Lincoln, and comparisons with information from other South Australian dolphin carcasses. Report to

9 BELL ET AL.: COMMON DOLPHINS IN SOUTHERN AUSTRALIA 9 Australian Nature Conservation Agency: Canberra. MAYR E, What is a species, and what is not? Philosophy of Science 63: PERRIN WF AND HEYNING JE, Rostral fusion as a criterion of cranial maturity in the common dolphin, Delphinus delphis. Marine Mammal Science 9: PODINI J, Extending Gower's general coefficient of similarity to ordinal characters. Taxon 48: ROSEL PE, DIZON AE AND HEYNING JE, Genetic analysis of sympatric morphotypes of common dolphins (genus Delphinus). Marine Biology 119: ROSS GJB, The smaller cetaceans of the south east coast of southern Africa. Annals of the Cape Provincial Museum 15: ROSS GJB AND COCKCROFT VG, Comments on Australian bottlenose dolphins and taxonomic status of Tursiops aduncus (Ehernberg, 1832) Pp in The bottlenose dolphin ed by S. Leatherwood and R.R. Reeves. Academic Press: San Diego. SPSS, SYSTAT10. SPSS Inc.: Chicago. WANG JY, CHOU L-S AND WHITE BN, Osteological differences between two sympatric forms of bottlenose dolphin (genus Tursiops) in Chinese waters. Journal of Zoology, London 252: WHITE C, Molecular systematics of the common dolphin, Delphinus delphis. B.Sc. (Hons) thesis, University of Adelaide, Adelaide. APPENDIX 1 Specimens examined, bolded specimens were included in the statistical analyses. Australian Museum: M22, M29, M30, M32, M137, M139, M140, M141, M12400, M12407, M18551, M22899, M23173, M26999, M27971, M33524, M33571, M33618, M33755, P276, P286, P287, P289, P290, S266, S360, S472, S473, S1912 Museum of Victoria: C11360, C14518, C14519, C16073, C23561, C24941, C24942, C24943, C24964, C24969, C24970, C24973, C25800, C26193, C26628, C28103, C28894, C28895, C28907, C29501, C29522, C29589, C29590, C29592 South Australian Museum registered: M1065, M1066, M1067, M1068, M1070, M2297, M3087, M4976, M5295, M10096, M10630, M10637, M12797, M14467, M16259, M16267, M16268, M16269, M16389, M16390, M16391, M16429, M16430, M16965, M16968, M16969, M16974, M16975, M16979, M17421, M17593, M17598, M17809, M18042, M18043, M18044, M18049, M18050, M18054, M18056, M18063, M18083, M18090, M18091, M18092, M18094, M18096, M18903, M18906, M18907, M18908, M18909, M18910, M18913, M18936, M19955, M19956, M19957, M19959, M19960, M19961, M19970, M19976, M19980, M19981, M19982, M19983, M19984, M19985, M19986, M19988, M19993, M19994, M19996, M19997, M19998, M20715, M20717, M20718, M20721, M20725, M20726, M20729, M20730, M20732, M21246, M21247, M21248, M21249, M21250, M21251, M21252, M21253, M21254, M21256, M21257, M21258, M21259, M21260, M21261, M21262, M21263, M21264, M21276, M21283, M21284, M21285, M21287, M21288, M21289, M21291, M21292, M21293, M21295, M21296, M21310, M21311, M21312, M21313 Temporary Accession Number: , , , Tasmanian Museum and Art Gallery: A194, A195, A196, A197, A365, A366, A367, A784, A785, A787, A928, A929, A1260, A1291, A1292, A1293, A1350, A1351, A1359 Western Australian Museum: M3321, M4007, M4363, M6844, M8541, M9099, M11182, M11380, M15251, M16239, M19853, M34698, M48700, M48702 APPENDIX 2 Postcranial Measurements. 1 Scapula height (SH); 2 Scapula width (SW); 3 Scapula glenoid cavity width (SGCW); 4 Scapula coracoid process length (SCPL); 5 Scapula coracoid process width (SCPW); 6 Scapular acromion length; 7 Scapula acromion width; 8 Humerus length (HL); 9 Caput humeri width (CHW); 10 Caput humeri depth (CHD); 11 Humerus middle width (HMW); 12 Humerus middle depth (HMD); 13 Humerus distal width (HDW); 14 Humerus proximal width (HPW); 15 Tuberculum margos humeri width (TMHW). Counts. Tooth counts = premaxillary and maxillary tooth alveoli. Vertebral: cervical, thoracic, lumbar and caudal. Number of the vertebra (caudal) with the first perforating foramen. Counts of phalanges, metacarpals and carpals made using radiographs of flippers.

10 10 AUSTRALIAN MAMMALOGY Qualitative Observation Codes. Nasal size (relative to each other): 1. L < R; 2. R < L; 3. R = L. Pterygoid shape (posterior view): 1. Arched; 2. Angular. Pterygoid relative position (posterior view): 1. Touching surface; 2. Off surface. Pterygoid shape (ventral view): 1. No ridge (rounded); 2. Slight ridge or ridge just at anterior end; 3. Distinct ridge all along pterygoid (sharp/angular). Parasite damage to pterygoids: 1. None; 2. Slight; 3. Moderate; 4. Extensive. Physical Maturity Codes. Epiphyseal fusion in vertebrae. 1. None; 2. Some, not thoracics; 3. Some thoracics; 4. All epiphyses fused to centra. Epiphyseal fusion in flipper. 1. None; 2. Some, but not distal ulna; 3. Distal ulna epiphyses fused; 4. All flipper epiphyses fused. Maxillary/premaxillary fusion: 1. No fusion, bones move freely or are disarticulated; 2. Loosely fused with some movement; 3. Fused and secure, suture line closing but still visible; 4. Very well fused, suture line disappeared or almost so. Tooth Development:1. Open (pulp cavity is % open at base); 2. Closing (< 50% open); 3. Pinhole (fine hole at base); 4. Closed (cavity closed); 5. Closed with some wear.