COS 108-2
Ecological and genetic causes of female-biased sex ratios in dwarf willows

Thursday, August 14, 2014: 1:50 PM
301, Sacramento Convention Center
Andrés J. Cortés, Evolutionary Biology Centre, University of Uppsala, Uppsala, Sweden
Xiaodong Liu, Evolutionary Biology Centre, University of Uppsala, Uppsala, Sweden
Christian Lexer, Department of Biology, University of Fribourg, Fribourg, Switzerland
Janosch Sedlacek, Department of Biology, University of Konstanz, Konstanz, Germany
Julia A. Wheeler, WSL Institute for Snow and Avalanche Research SLF, Davos, Switzerland
Sophie Karrenberg, University of Uppsala, Uppsala, Sweden
Background/Question/Methods

A 1:1 sex ratio is predicted as an evolutionarily stable strategy. Any deviations from that ratio are likely due to ecological selection, gametic selection or genetic determination. Willows are widespread shrubs and trees that present a strongly biased ratio usually towards females (~70:30); their sex determination mechanism is unknown. We explored the causes of female predominance in a common alpine dwarf willow (Salix herbacea). During three years, we sexed and recorded the performance of 227 females and 65 males growing naturally in three elevational transects in the Swiss Alps. A subset of 127 females and 38 males were genotyped with 99,497 genotyping-by-sequencing-derived SNP markers anchored in a female S. purpureareference genome.

Results/Conclusions

Temperature, altitude, microhabitat and biotic stress did not predict the female-biased sex ratio and females did not grow more nor were more vigorous than males. 24 genetic markers distributed in 11 regions and 10 chromosomes predicted sex with a specificity of 88%. Some of those regions blasted against expressed sequences in poplar male catkins. Interestingly, the expected female/male ratio (0.76) persisted after simulating an offspring with the allele frequencies of the sampled populations. This suggests that the sex determination mechanism in S. herbacea likely involves several loci, each with small effects. Together, these results indicate that the multi-loci sex determination system may explain sex bias, rather than unmeasured ecological differences between sexes, for instance at early life history stages.