Do genetic differences in growth thermal reaction norms maintain genetic variation in timing of diapause induction?

An optimal timing for diapause induction through the sexual production of dormant propagules is expected in populations of annual organisms. Yet, experimental work typically finds high within-population genetic variation in the sexual production of such propagules. Thus, high genetic variation in timing for diapause induction should be a common feature of annual organisms. Here, we hypothesize that genetic variation in the propensity to produce diapause propagules, P d , is maintained by a genotype-by-environment interaction in growth performance, where fast-growing genotypes within an environment should delay diapause relative to slow-growing genotypes. From this, we derive two predictions. First, if there is ecological crossover in growth performance, the genetic correlation of P d between environments should be negative. Second, the correlation between absolute plasticities of growth and P d should be negative. We tested these predictions by quantifying ephippia production in genotypes of a population of the facultative sexual cladoceran Daphnia magna at two temperatures. The population biomass at the onset of ephippia production was used as a measure of P d , whereas somatic growth rate was used to quantify growth. Plasticity for both measurements was derived from thermal reaction norms. Our results did not support either prediction, as neither the genetic correlation of P d between environments, nor the correlation between absolute plasticity of growth and P d were found to be significant. Our results suggest that genetic variation in the timing of diapause is not maintained by genetic differences in thermal growth reaction norms. We propose as an alternative hypothesis that if there is across year variation in how stochastically the environment deteriorates, fluctuating selection may favor genotypes with different P d between years.