The evolution of androdioecy
The fact that some animal and plant species are dioecious (with separate male and female individuals), while others are cosexual (with male and female functions in the same individual) has been explained in terms of a trade-off between investing resources in male or in female reproduction. It is predicted that evolution will favour dioecy in circumstances where an individual's fitness is be maximised by specialising in either male or in female reproduction, and will favour cosexuality in circumstances where fitness is highest among the individuals that devote resources both to male and to female reproduction. These ideas have proved hard to test empirically, because it is difficult to make direct comparisons between cosexual and dioecious species. However, the existence of certain species in which both unisexual and cosexual individuals occur within the same population provides a potential opportunity to do so (Pannell 1997).
Gynodioecy, a situation in which a population comprises females and cosexual individuals, can arise from cosexuality through the occurrence of a 'male sterility' allele that renders individuals female. Such an allele may increase fitness by increasing overall fecundity (due to the allocation of increased resources to female reproduction) or by eliminating inbreeding depression (since unisexual individuals cannot self-fertilise). If the male-sterility factor is cytoplasmically (and thus maternally) inherited, it need only produce a small increase in female fitness in order to spread through a population (Pannell 1997).
The evolution and maintenance of androdioecy (in which there are males and cosexuals) is more difficult to account for. Theory predicts that males will only invade a cosexual population if the fitness of a male is greater than twice that of the male component of a cosexual individual. High self-fertilisation rates will further reduce the advantage of maleness over cosexuality, by reducing the opportunity for males to fertilise cosexual individuals, so the evolution of androdioecy (unlike that of gynodioecy) cannot be due to the benefit of avoiding inbreeding depression (Fritsch & Rieseberg 1992, Pannell 1997).
Fritsch & Rieseberg (1992) suggest that the theoretical requirement for the evolution of androdioecy is difficult to satisfy in nature: it seems unlikely that the gain in male fecundity resulting from the redistribution of resources from female to male reproduction will normally be great enough to increase male fitness to more than twice the level found in hermaphrodites. This may explain why examples of androdioecy are much rarer in nature than examples of gynodioecy.
However, Pannell (1997) claims that Fritsch & Rieseberg and other previous authors failed to consider the implications of a situation in which cosexuals do not devote equal resources to male and female reproduction. If the proportion of resources that a cosexual allocates to male function is 1/r, loss of female reproduction will lead to an r-fold increase in male fertility, assuming complete re-allocation of resources from female to male functions. Therefore, if the proportion of resources devoted to male function is anything less than 1/2, male fertility will be more than twice as high in males as in cosexuals, fulfilling the condition necessary for androdioecy to be maintained. According to Pannell, this situation is found in androdioecious populations of the plant species Mercurialis annua. Stamen biomass (a measure of male fecundity) is between 4 and 10 times higher in males of this species than in cosexual individuals, which appear to devote a disproportionate share of their resources to female function (especially at higher population densities).
The theoretical proportion of male individuals in an androdioecious population can be predicted mathematically from the relative male fecundity of male/cosexual individuals, the outcrossing rate, and the relative fitness of inbred/outbred offspring. Fritsch & Rieseberg (1992) performed this calculation using data gathered on two androdioecious populations of the plant Datisca glomerata, and confirmed that the theoretical requirements for the maintenance of androdioecy are satisfied in this species. Their prediction of the frequencies of males in the two studied populations of D. glomerata was "surprisingly close" to the observed frequencies (although the observed values fell slightly outside the standard error limits of the theoretical values in both cases).
Genetic analysis showed that the outcrossing rate was higher in one of the two populations of D. glomerata studied by Fritsch & Rieseberg than in the other. This was probably due to differences in population density, with outcrossing more likely where plants were densely clustered together. The more outbred population was found to have a higher frequency of males, which concurs with the theoretical idea that the spread of males in a cosexual population is more likely when there is a high rate of outcrossing.
In a comparison of two populations of Mercurialis annua, Pannell (1997) also found male frequency to be correlated with the stand density of the plants. In this case the difference in male frequency is due to differential sex expression, with at least some genetic males able to function as cosexual individuals, and more likely to do so at lower stand densities. This sexual lability - the ability of an individual to develop either as a male or as a hermaphrodite, permitting self-fertilisation in the absence of a suitable partner - is suited to the colonising nature of M. annua. Pannell suggests that selection for the ability to self-fertilise at low population densities, together with the high reproductive success enjoyed by males at high population densities, accounts for the persistence of androdioecy in this species.
Theories involving trade-offs between male and female reproduction predict that the evolution and maintenance of androdioecy requires a specific, and perhaps unlikely, set of circumstances. The studies by Fritsch & Rieseberg (1992) and Pannell (1997), which confirm that these circumstances are present in two species in which androdioecy is known to occur, provide empirical evidence in support of such theories.
Fritsch, P. & Rieseberg, L. H., 1992. High outcrossing rates maintain male and hermaphrodite individuals in populations of the flowering plant Datisca glomerata. Nature 359, 633-636.
Pannell, J., 1997. Variation in sex ratios and sex allocation in androdioecious Mercurialis annua. Ecology, 85 57-69.
This was originally written as a university biology essay