Skull morphometry of the common dolphin, Delphinus sp., from New Zealand waters : a thesis presented in partial fulfilment of the requirements for the degree of Master of Science in Conservation Biology at Massey University, Auckland, New Zealand

The short-beaked, Delphinus delphis, and long-beaked, D. capensis, common dolphin, two morphotypes of the Delphinus genus, are recognized as different species. However, to date, species status of the New Zealand common dolphin, among other geographic populations, remains unclear, owing to morphometric and genetic uncertainty. This lack of taxonomic knowledge is one of the reasons preventing adequate threat status assessment. The main objective of the current skull morphometric study, the first to solely focus on New Zealand Delphinus sp., was therefore, to obtain further information regarding taxonomy and life history for conservation purposes. In particular, the study aimed (1) to determine age at cranial maturity through the computation of a suture index; and (2) to assess the validity of several cranial parameters as cranial maturity indicators through the determination of a misclassification index. Furthermore, (3) presence of cranial sexual dimorphism was investigated in (i) metric characters through ANOVA and ANCOVA analyses and in (ii) non-metric characters through Chi-Square tests. (4) The taxonomic status was assessed based on the rostrum length to zygomatic width (RL/ZW) ratio, tooth counts, and the Kalya Index. Moreover, (5) Potential regional differences between Hauraki Gulf (HG) and non-HG specimens were investigated through MANOVA analyses (metric characters), Chi-Square tests, and the computation of the mean measure of divergence (non-metric characters). In addition, (6) measurement error of two metric data acquisition methods (callipers versus microscribe) was compared through the computation of three precision estimates (variance, mean absolute difference (MAD), and relative error magnitude (REM)). A total of 67 common dolphin skulls from stranded and by-caught individuals were available for analyses. The majority of skeletal material (73.1%, n = 49), had been archived frozen as intact heads following necropsies at Massey University. Those heads were prepared as part of the present study via applying the manure decomposition method. The remaining 26.9% (n = 18) of skulls were cleaned specimens housed at the Museum of New Zealand Te Papa Tongarewa. Sex was known for 88.1% (n = 59) of specimens (males: 40.7%, n= 24; females: 59.3%, n = 35). Based on age data and the suture index, 46.3% (n = 31) and 53.7% (n = 36) of specimens were regarded as cranially immature and mature, respectively. Sex ratio of immatures was approximately 1:1 (males: n = 16, females: n = 13), while that of mature specimens was almost 1:3 (males: n = 8, females: n = 22). The suture index suggested that New Zealand Delphinus sp. obtain cranial maturity at approximately 11 years. Specimens with ≤ 6.8 % of partly worn teeth were between 1 to 3 years and cranially immature, while specimens with any number of rostral teeth worn down to the gum line were physically mature. Sexual size dimorphism, with larger sizes recorded for males, were detected in total body length (TBL) and in 22.7% (n = 15) of cranial characters analyzed, of which 86.7% (n = 13) were width measurements. In total 70.0% (n = 7) of size dimorphic characters that could be allocated to a cranial functional complex were related to the feeding apparatus. RL/ZW ratio (mean: 1.49 ± 0.06 (SD); range: 1.39 - 1.61) and upper tooth counts (45 - 56) of cranially mature New Zealand specimens assessed (pooled for both sexes) overlapped with values published for both the short-beaked and long-beaked form. Values of TBL, condylobasal length (CBL), rostrum length (RL), and zygomatic width (ZW) were also of intermediate status in both sexes. Findings reported herein suggest that New Zealand Delphinus sp. should be regarded as a large form of D. delphis until further morphometric and genetic data becomes available. No evidence of regional differences between HG and non-HG specimens was detected in either metric or non-metric characters, however, sample sizes were small. Variance of repeated measures was lower in the calliper (range: 0.1 to 0.7%) than in the microscribe (range: 1.1 to 10.7%) data set for all characters assessed (n = 33). High precision between both data sets was detected for 69.7% (n = 23) of characters (MAD below the 1 mm threshold) and REM of 93.9% (n = 31) of character was deemed excellent or good, indicating high compliance between both methods for the majority of characters assessed