Female choice for high roaring rates in red deer, Cervus elaphus

Red deer males roar loudly and repeatedly during the breeding season. A series of playback experiments, carried out on farmed red deer in New Zealand, revealed that females preferred a high to a low roaring rate, but not low-pitched to high-pitched roars. In free-ranging red deer, fundamental frequency (the main determinant of roar pitch) is not related to body weight or reproductive success, whereas roaring rate is positively associated with both reproductive success and fighting ability. Female choice for males with high roaring rates may either reflect a selection pressure on females to mate with males that are successful in contests with other males or show that females mate with males that are easy to locate. Females also preferred the roar bout that was delivered first in a sequence, suggesting that they may prefer stags that initiate bouts of roaring. This is the first study to demonstrate experimentally that vocal characteristics of male mammals can influence the female's choice of mating partner. Empirical evidence for sexual selection through female mating preferences has been slow to accumulate. While female choice for non-heritable benefits such as territory quality is now well established (e.g. Howard 1978; Pleszczynska 1978; Warner 1987), the most controversial issue is still whether females ever choose males solely on the basis of genetic differences that will influence the attractiveness (Fisher 1930) or viability (Andersson 1982) of their offspring. There are few well documented cases of female choice for male phenotypic characteristics (reviews by: Halliday 1983a; Partridge & Halliday 1984; Majerus 1986; Kirkpatrick 1987; see also: Balmford, in press). Competition between males for access to mates can result in the same patterns of assortative mating as female choice and it is difficult to separate the two processes except by using experiments (Halliday 1983a). Vocal displays are one aspect of the male phenotype that can be reproduced accurately in the absence of other male stimuli and some of the best evidence for female choice has been generated by playback experiments. There is good evidence that male vocal characteristics influence female choice in some insects (Hedrick 1986; Simmons 1986; Zuk 1987), anurans (Ryan 1980; Robertson 1986; *Fieldwork address: Serengeti Lion Project, Serengeti Wildlife Research Centre, Box 3134, Arusha, Tanzania. Klump & Gerhardt 1987) and birds (Searcy & Andersson 1986; Catchpole 1987). No similar evidence exists for mammals, although it has often been suggested that the loud calls of male mammals may serve to attract females (some ungulates: Kiley 1972; hammer-headed bats, Hyposignathus monstrosus: Bradbury 1977; old world monkeys: Gautier & Gautier 1977; humpback whales, Megaptera novaeangliae: Tyack 1981; pikas, Ochotona princeps: Conner 1985). However, Mitani (1985) found that female orangutans, Pongo pygmaeus, did not move towards playbacks of male long calls. While a negative result may, in this case, be a consequence of the relatively stable nature of orangutan social groups, it emphasizes the need to test such suggestions experimentally. Only male red deer, roar (i.e. produce loud, lowpitched vocalizations) and they do so only during the autumn breeding season when they gather and defend harems. Roaring rates commonly reach a maximum of around 8 roars per min during the roaring contests that precede fights, but stags continue to roar when competitors are absent, often seeming to direct their roars at females, and maintain overall roaring rates of 2 roars per min throughout the 24 h (Clutton-Brock & Albon 1979; McComb 1988). There is evidence that high roaring rates confer success in contests between males (Clutton-Brock & Albon 1979) and that roaring advances ovulation in females (McComb 1987). 0003-3472/91/010079 + 10 $03.00/0 9 1991 The Association for the Study of Animal Behaviour