AbstractThe reproductive success of competing male guppies, Poecilia reticulata, was measured and compared to various features of social behavior and secondary sexual coloration. The most important determinant of reproductive success was the rate at which a male courted females relative to those of other males. Males with higher display rates have a greater chance of encountering a receptive female and are preferred by females. Males adjust their display rate in such a way as to be just noticeably more active than a competitor, and inbred strains differ in the maximum rate at which males can court females. No other factor is able to offset the disadvantage of displaying at a relatively low rate. Sexual selection has resulted in the maximization of courtship activity in natural populations. If males court females with equal frequencies, those which also inseminate females through gonopodial thrusting without female cooperation have a selective advantage. Because poeciliid females store sperm, inseminations through gonopodial thrusting can reduce the reproductive success of competitor males which copulate only following a display. A mixed strategy of displaying and gonopodial thrusting is more successful than either pure strategy. The result is a mating system which partially ignores female choice mechanisms. Intermale aggresssion was found to be maladaptive. Males which displayed at higher rates than competitors were less successful if they were also more aggressive than the competitors, than when they were non-aggressive or the competitors were more aggressive. Males were not able to reduce a competitor's courtship display rate through aggression. It was hypothesized that the low level of aggression in natural guppy populations is attributable to the fact that variance in size of males is low and fights would be lengthy before a winner could be determined. This would subtract from time available for courtship, and female preference for high-displaying males would select against aggressive phenotypes. There was little evidence that conspicuousness of male coloration influences female choice of males. Dull males with high courtship display rates were significantly more successful than conspicuous males with low display rates. It was concluded that females prefer conspicuous males only if all males exhibit equal courtship display rates. The frequency of male color patterns in a population did affect reproductive success. Males with rare color patterns sired more offspring than expected given their frequency. The mechanism by which a rare male effect was achieved depended on the relative mating success of phenotypes in control populations (all phenotypes occurring with equal frequency). If a normally preferred phenotype was rare, females continued to prefer that phenotype over the common phenotype. If a normally unpreferred phenotype was rare, females mated with that phenotype in addition to the preferred phenotype, and the rare male effect was thus achieved by multiple inseminations. Females also tended to mate with more than one male in polymorphic populations, and it was concluded that certain female choice patterns with frequent multiple inseminations can maintain a polymorphism in addition to a pure rare male effect. It was hypothesized that courtship displays and conspicuous male coloration are sexually selected characters which evolved in response to the occupation of invariable habitats or specialized ecological niches by a sexually monomorphic ancestor with only gonopodial thrusting as a means of inseminating females. A subsequent reinvasion of variable habitats resulted in female choice mechanisms which maximize the level of heterozygosity of their offspring, thus resulting in polymorphic populations. The coloration of male guppies is a phenotypic cue which influences female choice in such a manner that they mate with those males with whom they most probably have the fewest genes in common.
Sexual selection and the evolution of preferential mating in ladybirds I. Selection for high and low lines of female preference
