Reproduction is energetically expensive for males as well as for females, but evolutionary biologists have typically regarded the energy costs of sperm production as trivial compared to the energy costs of overt reproductive behaviours, such as mate-searching, courtship, copulation and male–male combat. Adders (Vipera berus) offer an ideal model system in which to quantify the relative costs of spermatogenesis (and associated physiological preparations for reproductive activity) versus the costs of overt reproductive behaviours, because (i) these two activities occur sequentially rather than simultaneously, and are separated by a clear indicator (sloughing of the skin); and (ii) males do not feed during either of these phases. Hence, the rate of mass loss by males can be used as an index of energy costs. Surprisingly, the rate of mass loss is at least as high during the phase when males are immobile, building up sperm supplies, as when they move about and engage in reproductive behaviour. Rates of mass loss are not significantly correlated with male size or measures of the male's participation in reproductive activities. Our data suggest that sperm production may be a major energy cost to reproducing male adders, and that this species offers a useful system in which to further investigate this possibility.

MALES of most animal species will enhance their reproductive success if they mate often and with many different partners, whereas promiscuous mating is unlikely to increase a female's reproductive success. Why then is multiple copulation by females so common^1–6? Many theoreticians have suggested that multiple copulations might enhance the viability of a female's offspring, either because of inadequate quantities of sperm from the first mating^1,7, additional nutrients derived from the seminal fluid^7,8 or some genetic advantage^9–14. Our field studies on Swedish adders provide the first empirical evidence that multiple copulations, with different partners each time, increase offspring viability. This advantage apparently results from more intense sperm competition in the female's reproductive tract, resulting in a higher proportion of her ova being fertilized by genetically superior males.

In mice and man, females prefer males with a major histocompatibility complex (MHC) genotype different to their own. We tested whether this phenomenon also occurs in the Swedish sand lizard (Lacerta agilis). Females in a laboratory experiment preferred to associate with odour samples obtained from more distantly related males at the MHC class 1 loci. Data on free-ranging lizards suggest that associations between males and females are nonrandom with respect to MHC genotype. However, male spatial distribution and mobility during the mating season suggest that the non-random pairing process in the wild may also be driven by corresponding genetic benefits to males pairing with less related females.