Factors influencing the sexual behaviour of the guppy, Poecilia reticulata.

The active courtship, bright colouring and polymorphism of males, and the cryptic colouring and passive behaviour of females bear witness to the combined effects of selection for survival and sexual selection pressures on the morphology and behaviour of the guppy. The aim of this study was to investigate guppy sexual behaviour in the light of these pressures during its evolution. Without knowledge of the sensory modalities involved, a study of the mechanisms by which selection acts on sexual behaviour cannot be complete. In particular, there is evidence that olfaction plays a significant part in signals from the female to the male. The role of different sensory modalities in the development of normal courtship was studied by rearing individuals of both sexes in conditions of restricted access to conspecifics from birth until maturity, and then observing their behaviour during courtship. Inexperienced virgin females are highly receptive at first exposure to courtship, while males reared apart from females tend to court other males as well as females. By systematically varying the type of access (visual, olfactory etc.) allowed during development, it was hoped to assess in which modalities stimuli ensuring normal courtship were important. In neither the experiments on the development of courtship, nor on the execution of courtship (in which quantitative and qualitative variation in male courtship with and without olfactory stimuli from the female was investigated) could the role of pheromonal communication be ascertained. However, two findings emerged from the studies of the development of courtship, which were investigated in greater depth: 1) The highly receptive virgin females actively sought out males. This finding was substantiated by additional experiments in which the virgin females were found to choose to be near males rather than females, and this choice appears to be based on both the morphology and the greater activity of males than females. In normal circumstances, male guppies court females a large amount of the time, including when females are not receptive. It is possible that males are able to detect a metabolic by-product from non-receptive females and use this to locate them. The receptive females' seeking out of males may be a mechanism by which females ensure fertilisation at the optimum point of their receptivity cycle - i.e. when conditions for successful fertilisation and gestation are best. 2) Even males which had been totally isolated from conspecifics were found to be able to recognise females, although the pattern of their sexual behaviour was abnormal. The explanation of this discrepancy with the finding that males reared apart from females later courted only males (Liley, 1966) probably lies in the manner of rearing. Liley's fish were reared in a group of males, while those in the present were reared in total isolation. It seems that although males do recognise females innately, there also exists some sort of critical period in which recognition of females is substantiated or, if no females are present, imprinting on other males occurs. Bateson (1966) notes that existing unlearned preferences may be affected by imprinting. It is suggested that the imprinting process provides the optimum balance between the costs associated with inflexible recognition of conspecific males in a polymorphic species, and the costs of energy and sperm wastage associated with an indiscriminate direction of sexual behaviour. The sexual behaviour of isolated males was abnormal both in intensity and form. This was shown to be due to a motivational change, rather than a developmental pathology. Isolated males perform an exceptionally large number of Thrusts, which Liley has shown are less efficient per single act than is copulation during normal courtship. However if Thrusting takes less time than courting, and males have a finite amount of time available for sexual behaviour, then a highly motivated (and therefore sexually energetic) male can achieve a greater number of inseminations by Thrusting than by courting. The predictions of a speculative model taking these constraints into account were largely borne out. Observations of an established group of normal adult guppies indicate that mating is not random, and that some males are more successful than others at gaining access to popular females. It may be that the hierarchy of males arises from females' preferences for particular males. The form of individual males' courtship varies according to their rank in the hierarchy, and also which female they are courting. Males which are able to gain access to popular females tend to court, while lower hierarchy males perform relatively more Thrusts. This may represent a type of conditional strategy in which preferred males tend to monopolise the courtship of females, but low-ranking males can 'make the best of a bad job' by using a less profitable means of insemination which does not require female co-operation.