Testosterone and competitive ability in male house mice , Mus musculus : laboratory and field studies

Testosterone enhances aggression and can facilitate dominance in mammals, but most studies have been conducted on laboratory strains of mice and in simple environments. Unlike previous studies, laboratory and field tests were conducted using wild house mice, Mus musculus, to determine whether males with high, but physiological, testosterone levels (High T) would achieve dominance compared with males with low testosterone levels (Low T). All males were castrated and androgen was replaced with a single injection of testosterone enanthate, a long-acting androgen. At 8 days post-treatment the mean testosterone concentration in High T males was 5. 1 ng/ml, whereas the level in Low T males was 1 .3 ng/ml. Contests in the laboratory revealed that High T males achieved dominance over Low T males in the presence of females. In the field, equal numbers of High and Low T males, and a corresponding number of females, were released on three highway islands at densities equivalent to the highest density reported on similar sites. Periodic retrapping produced a capture history for males in each group and differences in survival were analysed using the program SURVIV. High T males had a higher probability of survival than Low T males on one island, but on the other two islands the higher capture rates of High T males were not statistically greater than those for Low T males. Of those that survived to the first trapping episode, the High T males had greater subsequent longevity than Low T males. On all three islands High T males were caught more frequently at feeding stations that were established to enhance food competition. There were no differences among the home range sizes of High T males, Low T males and females. However, on two of the islands the ranges of High T males overlapped more of the ranges of resident females than those of Low T males. The findings from this study, using testosterone-treated subjects, suggest that natural variation in basal testosterone also may affect competitive ability in wild house mice. Testosterone regulates spermatogenesis and mating behaviour in males. However, testosterone also mediates male-male aggression, presumably due to its aromatization to oestrogen in the brain (Naftolin et al. 1971; Dessi-Fulgheri et al. 1976; Brain 1983). Socially dominant males are often the most aggressive animals and are assumed to be capable of acqui ring more resources than less aggressive individuals. Thus, one might expect individual males with higher t estos terone levels to be more aggressive, to command greater control over resources, including females, and to have the highest fitness. However, a relationship between plasma testosterone, dominance and fitness has been difficult to establish. While testosterone appears to promote aggression, this behaviour does not necessarily facilitate dominance (e.g. Mazur 1976; Dixson 1980; Caldwell et al. 1984; Sachser & Prove 1984; Berkovitch & Coy 1990). Moreover, dominance does not always guarantee high relative fitness; a myriad of other factors influence reproductive success. Even where a relationship between androgens and dominance has been discovered it is often unclear whether a difference in testosterone caused the dominance pattern or resulted from it (Beach 1965). Luteinizing hormone and testosterone levels rise during a male ‘challenge’ (Wingfield et al. 1990) or after exposure to a receptive female (Batty 1978; Bronson & Desjardins 1982) and can fall quickly after a defeat (Bronson & Desjardins 1971; Bronson et al. 1973; Martinez et al. 1989). Testosterone levels also fluctuate greatly even *Present address: US. Forest Service, Redwood Sciences without social stimulation (Bartke et al. 1973; Laboratory, 1700 Bayview Drive, Arcata, CA 95521, Bartke & Dalterio 1975; Coquelin & Desjardins U.S.A. 1982). Fluctuations in basal testosterone levels 0003-3472/93/050873 + 19 $08.00/0