Hermaphroditic origins of anisogamy

Abstract

Anisogamy-the size dimorphism of gametes-is the defining difference between the male and female sexual strategies. Game-theoretic thinking led to the first convincing explanation for the evolutionary origins of anisogamy in the 1970s. Since then, formal game-theoretic models have continued to refine our understanding of when and why anisogamy should evolve. Such models typically presume that the earliest anisogamous organisms had separate sexes. However, in most taxa, there is no empirical evidence to support this assumption. Here, we present a model of the coevolution of gamete size and sex allocation, which allows for anisogamy to emerge alongside either hermaphroditism or separate sexes. We show that hermaphroditic anisogamy can evolve directly from isogamous ancestors when the average size of spawning groups is small and fertilization is relatively efficient. Sex allocation under hermaphroditism becomes increasingly female-biased as group size decreases and the degree of anisogamy increases. When spawning groups are very small, our model also predicts the existence of complex isogamous organisms in which individuals allocate resources equally to two large gamete types. We discuss common, but potentially unwarranted, assumptions in the literature that could be relaxed in future models. This article is part of the theme issue 'Half a century of evolutionary games: a synthesis of theory, application and future directions'.

Keywords: anisogamy; cosexual; hermaphroditism; homothallic; monoecy; monoicy.

Figure 1.

Illustration of the trioecy index T in four different anisogamous populations. Numbers indicate the proportions of each population that are female, male and monoecious. The trioecy index equals the proportion of males divided by the total proportion of males and females. For example, in the bottommost population, we have T = 0.1/(0.1 + 0.2) = 1/3. Note that populations with different proportions of monoecious individuals can nonetheless have the same trioecy index (the bottom two populations).

Figure 2.

The proportion of monoecious individuals shown against variation in the average group size, λ. Populations were classified roughly as ‘monoecious’ (circles) if more than a quarter of individuals were monoecious and otherwise as ‘dioecious’ (triangles). Blue circles represent monoecious population in which greater than 95% of individuals were monoecious. For populations with less than 95% monoecious individuals, colouring represents the trioecy index (i.e. the proportion of individuals producing the smaller gamete type among all those specializing in one gamete type; see inset legend and figure 1). Under anisogamy, T = 0 and T = 1 represent gyno- and androdioecy, respectively. Fertilization was either efficient (γ = 10; (a,c,e)) or inefficient (γ = 0.01; (b,d,f)). Selfing was either absent (B = 1; (a,b)); unrestricted with no inbreeding depression (B = 0, D = 0; (c,d)); or unrestricted with strong inbreeding depression (B = 0, D = 0.5; (e,f)). Median fertilization rates (i.e. the proportion of the larger gamete type that was successfully fertilized) across simulation runs were approximately 0.995 when fertilization was efficient (a,c,e) and approximately 0.5 when fertilization was inefficient (b,d,f). All other parameter values are as in table 1.

Figure 3.

The anisogamy ratio (i.e. the ratio of larger gamete size to smaller gamete size) shown against variation in the degree of organismal complexity β/α. Colouring represents the average group size, λ (see inset legend). Fertilization was either efficient (γ = 10; (a,c,e)) or inefficient (γ = 0.01; (b,d,f)). Selfing was either absent (B = 1; (a,b)); unrestricted with no inbreeding depression (B = 0, D = 0; (c,d)); or unrestricted with strong inbreeding depression (B = 0, D = 0.5; (e,f)). Populations were classified roughly as ‘monoecious’ (circles) if more than a quarter of individuals were monoecious and otherwise as ‘dioecious’ (triangles). All other parameter values are as in table 1.

Figure 4.

Sizes of the smaller and larger gamete types. The dashed red line indicates isogamy. Colouring represents the degree of organismal complexity, β/α (see inset legend). Fertilization was either efficient (γ = 10; (a,c,e)) or inefficient (γ = 0.01; (b,d,f)). Selfing was either absent (B = 1; (a,b)); unrestricted with no inbreeding depression (B = 0, D = 0; (c,d)); or unrestricted with strong inbreeding depression (B = 0, D = 0.5; (e,f)). Populations were roughly classified as ‘monoecious’ (circles) if more than a quarter of individuals were monoecious and otherwise as ‘dioecious’ (triangles). All other parameter values are as in table 1.

Figure 5.

Sex allocation (i.e. the average proportion of resources allocated to the smaller gamete type) shown against variation in the average group size, λ. The red dashed line indicates equal allocation to each gamete type. Colouring represents the anisogamy ratio (ratio of larger gamete size to smaller gamete size; see inset legend). Fertilization was either efficient (γ = 10; (a,c,e)) or inefficient (γ = 0.01; (b,d,f)). Selfing was either absent (B = 1; (a,b)); unrestricted with no inbreeding depression (B = 0, D = 0; (c,d)); or unrestricted with strong inbreeding depression (B = 0, D = 0.5; (e,f)). Populations were classified roughly as ‘monoecious’ (circles) if more than a quarter of individuals were monoecious and otherwise as ‘dioecious’ (triangles). All other parameter values are as in table 1.