Sex allocation is the allocation of resources to male versus female reproduction in sexual species (Charnov 1982). Sex allocation depends upon the breeding system of a species, as well as how reproduction is carried out within each breeding system. Breeding systems can be categorised as dioecious, in which individuals are either male or female for their entire lifetime (e.g. birds and mammals) or hermaphroditic, in which the same individual can produce both male and female gametes. Hermaphrodites can be either sequential or simultaneous. Sequential hermaphrodites, or sex changers, function as one sex early in their life, and then switch to the other (e.g. some reef fish such as anglefish, and some invertebrates such as Pandalid shrimps). Simultaneous hermaphrodites are capable of both female and male reproduction at the same time (e.g. most flowering plants).
Given the above scheme, the fundamental problems of sex allocation are as follows (Charnov 1982).
Darwin (1871, 1874) realised that the preponderance of unbiased sex ratios posed a problem for his theory of natural selection. He made a start at developing possible explanations, but was unsatisfied and left the problem for future generations. This problem was solved decisively by Fisher (1930), who showed that selection for an unbiased sex ratio follows from the fact each offspring has a mother and father, and so males and females make equal genetic contributions to the next generation. Importantly, Fisher clarified the frequency dependent nature of selection on sex allocation that is at the centre of all subsequent developments.
Modern research on sex allocation began with Hamilton (1967), which made five pivotal contributions to the field of sex allocation, and evolutionary biology more generally. First, he showed how competition between relatives can select for biased sex allocation. When populations are structured such that brothers compete for mates, this leads to selection for a female biased sex allocation by a process that Hamilton termed local mate competition (LMC). This insight has led to one of the most productive areas of evolutionary biology. Second, Hamilton showed how the sex ratio can be modelled using game theory. His approach for determining the ‘unbeatable strategy’ was very similar to, and laid the ground-stones for the technically superior evolutionary stable strategy (ESS) approach that was later formalised by Maynard Smith & Price (1973). Third, he showed that simple mathematical models could be used to make comparative predictions that could be easily tested. His specific example was to show that selection favours more female biased sex ratios when less females lay eggs on a patch, and that this could be tested either by comparing across species or by looking at how individuals vary their behaviour under different conditions. The use of comparative predictions is taken for granted today, because they form the daily bread of evolutionary and behavioural ecology research programmes. However, it should be remembered just how astounding it was at the time, to suggest that a few lines of simple maths could make testable predictions about how organisms should behave. Fourth, he showed how different genes within a genome can be selected to pursue their own selfish interests, to the detriment of other member of the genome, and the way in which meiotic drive fitted into this framework. Fifth, by emphasising the costliness of male production, and the evolution of parthenogenesis, it helped to initiate the debate over the adaptive function of sex.
The next major step was made by Trivers & Willard (1973) who showed that individuals could be selected to adjust the sex of their offspring in response to environmental conditions. They discussed their prediction in the context of mammals such as caribou, and why offspring sex ratios might be adjusted in response to maternal condition. Charnov and colleagues built upon this work by showing how the same principal could be applied more widely to a huge range of issues in both dioecious and hermaphroditic species. For example, whether host size should influence offspring sex ratios in parasitoid wasps, the age and direction of sex change in sequential hermaphrodites and when different breeding systems such as simultaneous hermaphroditism or environmental sex determination (ESD) should be favoured (Charnov 1982). Importantly, these predictions clearly lend themselves to empirical testing, which has helped make The Trivers & Wilard hypothesis and its various extensions, one of the two most productive areas of sex allocation, alongside LMC theory.
Another major strand of sex allocation research was initiated when Trivers & Hare (1976) examined conflict over sex allocation in the social hymenoptera (ants, bees & wasps). This paper made two key contributions. First, it combined Fishers (1930) theory of equal investment with Hamilton’s (1964) inclusive fitness theory, to show how the ES sex allocation differed from the point of view of the queen and their workers. Research on sex allocation conflict within the social hymenoptera has since become the third most productive area in the field of sex allocation (chapter 9). Second, they showed how parent-offspring conflict and inclusive fitness (kin selection) theory could generate predictions that could be tested with empirical data. This was at a time when these topics where still contentious, and to this day, sex allocation still provides some of the clearest support for inclusive fitness theory.
Charnov’s (1982) monograph, The theory of sex allocation, brought all this together, providing a masterly synthesis of theoretical and empirical work. He unified the different areas of sex allocation research into a single field. From a theoretical perspective, Charnov showed how the same underlying concepts and similar mathematical models could be applied to all of the problems of sex allocation. From an empirical perspective, Charnov’s monograph showed the power of ‘selection thinking’ and simple models to make predictions that could be tested with empirical data, and led to a surge of interest in sex allocation that continues to this day. The increase in interest in this area is demonstrated by the increasing number of citations per year – comparing 2007 with 1982, the number of citations produced by a search on the phrase “sex allocation” has increased 50 fold, and the number of citations produced by a search on the phrases “sex allocation” or “sex ratio” has doubled (Web of Knowledge; subject areas: zoology, genetics & heredity, evolutionary biology, behavioural sciences, plant biology). Charnov’s monograph also contained a wealth of leads to potentially useful biological systems, that remain underexploited to this day.
In the 1980’s, our theoretical understanding of LMC leaped forward. At a very general level, the reasons for the female biased sex ratio were clarified, disentangling the separate effects of competition between males, the availability of mates for those males, and inbreeeding (Taylor 1981; Frank 1985; Herre 1985; Frank 1986). In addition to settling a long running controversy, this work solved the debate over the level at which selection operates (Frank 1986), which sadly still persists in other areas. At a more specific level, a number of workers began extending LMC theory to fit the biology of specific systems. This generated a slew of new predictions, which allowed for some of the most elegant tests of LMC theory, in a wide range of organisms, and such work is still extremely active today.
Following Charnov’s monograph, there was a profusion of empirical studies testing the various forms of Trivers & Willard’s (1973) hypothesis. The most famous of these was Clutton-Brock and colleagues work on red deer, which provided support for both the assumptions and predictions of Trivers & Willard’s hypothesis, in response to maternal quality (Clutton-Brock et al. 1984). This work has inspired many researchers over the years and an extensive literature on sex allocation in ungulates has accumulated (Sheldon and West 2004). Equally impressive, were two long-term studies on species with environmental sex determination (ESD), by Conover and colleagues on a fish and by Adams and colleagues on a shrimp. These studies showed the pattern of ESD, the fitness consequences, and why the pattern of ESD should vary across populations.
Our understanding of selfish sex ratio distorters was revolutionised in the 1980’s and 1990’s. Relatively little was known about distorters at the time of Charnov’s (1982) monograph; they were assumed to be rare aberrations. Appreciation of their importance started to emerge, however, with Werren and Skinner’s discovery that three different sex ratio distorters occurred in the parasitoid wasp Nasonia vitripennis (Werren et al. 1981; Skinner 1982, 1985). This discovery was shocking because Nasonia had been intensively studied as a model species for understanding LMC and had provided some of the best evidence that individuals adjust offspring sex ratios in response to environmental conditions. The next major jump into the sex allocation limelight for sex ratio distorters was the discovery that endosymbiotic bacteria such as Wolbachia and Cardinium were responsible for many cases of sex ratio distortion, and that these endosymbionts were extremely widespread. There is now an extensive literature on sex ratio distorters, with recent work by G. Hurst and colleagues demonstrating how we can even follow their spread and suppression in natural populations.
The other major development of the 1980’s was an understanding of the population level consequences of individual level sex ratio adjustment. Frank (Frank 1987) showed that Trivers & Willard type sex ratio adjustment can lead to a bias in the population sex ratio or the overall population investment ratio. He also showed that the direction and magnitude of this bias could be hard to predict, depending upon biological details that could be hard or impossible to assess. A consequence of this, which is still rarely appreciated, is that population level patterns will often be useless for testing whether sex allocation is being adjusted facultatively in response to local conditions. Frank, Charnov and Bull also showed that an important exception to this is in sex changing organism, where we can make and test predictions about the population sex ratio.
Research on sex allocation conflict between individuals really took off in the 1990’s. Trivers & Hare’s (1976) paper had attracted much interest, but there are limitations on the testability of their predictions, using population level data. Boomsma & Grafen (Boomsma and Grafen 1990; Boomsma 1991; Boomsma and Grafen 1991) solved this, by showing that a range of more specific predictions could be made for how sex allocation should vary between colonies, within a population. In particular, they predicted that if workers were in control of sex ratio in a colony, we should observe split sex ratios, with some colonies producing predominantly male reproductives, and others predominantly female. Stunning support for their predictions rapidly followed from both observational and experimental studies on bees and ants. Since then, an impressive level of understanding has been obtained in this area by looking at: the underlying mechanisms; finer levels of within colony adjustment; mistakes; situations where the workers do not win. A new area of research on conflict was also opened up by the work of Strand and colleagues, showing the potential for sex allocation conflict in polyembryonic wasps, and how this might lead to the evolution of a sterile worker caste.
The 1990’s saw the conventional wisdom on sex ratio adjustment in vertebrates overturned. It had long been assumed that chromosomal (genetic) sex determination (CSD) in vertebrates such as birds and mammals would prevent adaptive control of offspring sex ratios. This conception was clearly blown out of the water by a number of studies, primarily on birds. Komdeur and colleagues showed that Seychelles warblers were capable of adjusting the proportion of males in a clutch from between 10% and 90%, depending upon environmental conditions (Komdeur et al. 1997). Sex allocation is adjusted in the Seychelles warbler in response to cooperation and competition with offspring. Another area of sex ratio adjustment in birds was opened up by Sheldon and colleagues, who showed that females, in species such as collared flycathers and blue tits, can adjust the sex of their offspring in response to mate quality, with females producing a higher proportion of sons when they mated more attractive males. This work was built upon previous findings by Burley (1981), that were so revolutionary in their time, that they had been effectively ignored for 15 years. The patterns of sex ratio adjustment in response to helping and male attractiveness have since been shown to be repeatable within and across species, proving clear evidence for control of offspring sex ratios in species with CSD (West and Sheldon 2002).
The final major development of the 1990’s was Frank’s (1998) reunification of sex allocation theory, in his mongraph Foundations of social evolution. Our understanding of sex allocation theory increased enormously during the 1980’s and 1990’s, thanks largely to the work of Taylor and Frank. They clarified the underling reasons for adjustment of sex allocation, linked different areas of research, and developed new methods for constructing theory, that were both simpler to apply and more general. Frank brought all this together in his 1998 monograph, which provided a guide on how to model sex allocation, as well as a unification of existing work. Taylor and Frank’s work was part of a more general programme on how to model inclusive fitness and social evolution, in which sex allocation theory has played a pivotal role.
The major development this millennium has been the attempt to explain broad taxonomic variation in the extent of sex ratio adjustment. This has united work in different conceptual areas on different taxa. One consequence has been to determine when vertebrates, with supposedly constraining CSD, really do show consistent patterns of sex ratio adjustment in the predicted direction. For example, birds adjust their offspring sex ratios in response to mate quality and the number of helpers on their patch (West and Sheldon 2002), but primates show no consistent pattern with maternal quality (Brown and Silk 2002). The other consequence of this work has been to show how variation in the extent of sex ratio adjustment across species can be explained by variation in the strength of selection. For example, birds show greater shifts of sex ratio in response to the number of helpers on their patch, when helpers provide greater benefits (Griffin et al. 2005), and wasps show greater shifts of sex ratio in response to host size, when host size better correlates with the resources that will be available for their offspring (West and Sheldon 2002). This work has emphasised that cases in which vertebrates show little or no sex ratio adjustment may just reflect a lack of selection, rather than the constraints of CSD, but also how sex allocation can be used to address very general issues on how adaptation may be limited.