Can a genetic correlation with seed mass constrain adaptive evolution of seedling desiccation tolerance in wild barley?

Very young seedlings of wild barley Hordeum spontaneum have the ability to survive extended periods of severe drought. This desiccation tolerance is considered an adaptation to the rain-limited and unpredictable habitats that the species occupies. Genetic variation has been observed for this trait, but the limited evidence to date does not consistently show that genotypes from more xeric populations have better desiccation tolerance. As large seed endosperm volume may buffer the desiccating seedling from drought stress, we explored the hypothesis that tolerance to temporary but severe drought is affected by seed mass. We crossed a big-seeded with a small-seeded H. spontaneum genotype and measured seed mass and seedling survival rates after 6-ddesiccation in 140 F4 progeny families. A positive correlation was observed between family mean scores of the two traits. At the genetic level, this indicated that at least some genes underlying variation in the two traits overlapped (or show close chromosomal linkage), which is inevitable if seed mass causally affects desiccation tolerance. Quantitative Trait Loci (QTL) analysis, which can identify chromosome regions responsible for trait variation by exploring associations between trait expression and genetic markers dispersed over the genome, did not confirm this hypothesis of shared genetic control. Despite the lack of QTL evidence, the observed genetic association between seed mass and desiccation tolerance indicates that the two traits are not fully independent, implying that selection targeting one trait will cause a correlated response in the other. We discuss this finding in the context of adaptive differentiation of H. spontaneum populations, which is thought to involve selection of increasingly smaller seeds along a gradient of decreasing predictability and amount of annual precipitation. [KEYWORDS: Hordeum spontaneum; adaptation; drought; QTL analysis; correlated selection