Large‐scale multivariate analysis reveals sexual dimorphism and geographic differences in theGray's beaked whale

Abstract

AbstractGray's beaked whales (Mesoplodon grayi) are medium‐sized odontocete (toothed) cetaceans that are members of the family Ziphiidae. These animals inhabit deep oceanic waters and are rarely seen at sea. Most information is derived from stranded animals and there has been no systematic study of their morphology. We present a multivariate analysis of the morphology ofGray's beaked whales using 80 cranial measurements from 22 individuals and 13 external measurements from 50 individuals. Sparse principal component and linear discriminant function analyses were used to classify samples into sexes. Males and females have markedly different cranial morphology. In particular, females have longer skulls with longer more slender rostra (beaks) in comparison to males. Two variables, depth of the rostrum at mid‐length and tip of rostrum to the right temporal fossa, can classify sex with 100% accuracy. The external body measurements used in this study are more prone to error as they were recorded by a number of observers on carcasses in differing states of decomposition and this is reflected in the level of variance in most measurements. However, analyses of these measurements showed a significant difference between sexes in the distance between (1) the tip of the rostrum to the genital slit, (2) the tip of the rostrum to the blowhole, as found in the cranial analyses and (3) tail fluke width where males have absolutely wider tail flukes than females. Differences in these same measurements were also found between animals stranded on the east and west coasts suggesting a degree of population separation across New Zealand. Finally, we present two linear models that enable the assignment of sex from either skull or external measurements. These models will be useful for future studies as well as the management of these whales and can be applied to archived data where genetic sex assignment is not possible.