What constrains directional selection on complex traits in the wild?

Abstract

The fact that abundant genetic variation persists within populations despite strong directional selection on complex traits is one of the unresolved conundrums in evolutionary biology. In this dissertation, I employed a multi-faceted approach combining classical and modern genomic methods with field studies to identify the factors that may reduce total selection on a complex trait. I investigated the causes and consequences of phenotypic and genotypic variation in flower size using the wild flower Mimulus guttatus (yellow monkeyflower) as the model system. Flower size in Mimulus guttatus exhibits abundant genetic variation amidst strong directional selection in the wild. To understand directional selection, we must consider the "invisible fraction" (the proportion of individuals that die before expressing the trait), which is typically unmeasured in correlative studies. Chapter 1 demonstrates that viability selection prior to trait expression can change the direction and magnitude of selection. In Chapter 2, I identified fitness trade-offs (viability and fecundity) and varying selection at the spatial and temporal scale as factors that can reduce the effect of directional selection in the wild. Using NILs (Nearly Isogenic Lines), I demonstrate that alleles increasing flower size also increase fecundity but they reduce survivorship. Furthermore, I also detected fluctuating selection by year and on a spatial scale of meters. Finally, correlated selection on corolla width may limit the evolutionary response to directional selection for flower size. I demonstrated in Chapter 3 using a multi-year phenotypic manipulation experiment that corolla width is under indirect selection due to its genetic correlation with other traits that influence fitness in the field (e.g. rate of development, reproductive capacity, vegetative size).