Absence of complementary sex determination in the parasitoid wasp genus Asobara (Hymenoptera: Braconidae)

Abstract

An attractive way to improve our understanding of sex determination evolution is to study the underlying mechanisms in closely related species and in a phylogenetic perspective. Hymenopterans are well suited owing to the diverse sex determination mechanisms, including different types of Complementary Sex Determination (CSD) and maternal control sex determination. We investigated different types of CSD in four species within the braconid wasp genus Asobara that exhibit diverse life-history traits. Nine to thirteen generations of inbreeding were monitored for diploid male production, brood size, offspring sex ratio, and pupal mortality as indicators for CSD. In addition, simulation models were developed to compare these observations to predicted patterns for multilocus CSD with up to ten loci. The inbreeding regime did not result in diploid male production, decreased brood sizes, substantially increased offspring sex ratios nor in increased pupal mortality. The simulations further allowed us to reject CSD with up to ten loci, which is a strong refutation of the multilocus CSD model. We discuss how the absence of CSD can be reconciled with the variation in life-history traits among Asobara species, and the ramifications for the phylogenetic distribution of sex determination mechanisms in the Hymenoptera.

Figure 1. Secondary offspring sex ratio, brood size, male and female offspring numbers over generations of inbreeding.

(a): Asobara tabida, (b): A. japonica, (c): A. citri, (d): A. pleuralis, OC: outcross. Open and grey bars denote male and female offspring number respectively. Black triangles represent mean sex ratio, and error bars represent standard error.

Figure 2. Simulation of the proportion diploid males.

(a): Asobara tabida, (b): A. japonica, (c): A. citri and (d): A. pleuralis. 10 000 replicates of each experiment were simulated, assuming different numbers of unlinked csd loci, n _loci = {1, 2, 5 and 10}. Blue shading with solid lines represent predicted proportion diploid males under CSD with one locus; green shading with dashed lines represent the trend under CSD with two loci; red shading with dotted lines for five loci, and pink shading with dot-dashed lines for ten loci. Each shaded polygon represents the 95% confidence intervals of the proportion diploid males for a particular number of csd loci. Black dots are the observed proportion diploid males in our experiments.

Figure 3. Simulation of secondary offspring sex ratios.

(a): Asobara tabida, (b): A. japonica, (c): A. citri and (d): A. pleuralis. 10 000 replicates of each experiment were simulated, assuming different numbers of unlinked csd loci, n _loci = {1, 2, 5 and 10}. Blue shading with solid line represents predicted offspring sex ratio under CSD with one locus; green shading with dashed line represents the trend under CSD with two loci; red shading with dotted line for five loci, and pink shading with dot-dashed line for ten loci. Each color-shaded polygon represents the 95% confidence intervals of offspring sex ratio for a particular number of csd loci n _loci, which is listed on the left side of the polygons. Black dots represent observed mean offspring sex ratio per generation, and corresponding error bars represent 95% confidence intervals of the observed mean offspring sex ratio. Note that in A. pleuralis, the low initial brood size (on average five) in the 1^st generation of the B-S cross makes the stochastic effects more pronounced, resulting in overlapping confidence intervals for model predictions during the first generations (the same effect also occurs in simulations for proportions of diploid males in Figure 2.).