The selective and developmental maintenance of genetic variation in a natural population of Mimulus guttatus

Abstract

The maintenance of genetic variation for quantitative traits has long puzzled evolutionary biologists. Previous studies have effectively interrogated large genomic regions both in the greenhouse and field to understand the selective forces maintaining variation. However, subtler points can be difficult to tease out from these genomic regions. Here, we take advantage of advances in sequencing technology to understand how variation for flowering time and size is maintained within a single population of Mimulus guttatus at the nucleotide level. By focusing on the nucleotide level, we directly address questions of allelic effect and frequency, and interrogate polymorphisms found to be significantly associated with traits in the greenhouse in other contexts to determine their contributions to growth and fitness. Flowering time and flower size have been shown to be highly quantitative traits within the Iron Mountain population of M. guttatus, and it has been proposed that minor alleles are likely to increase trait values. We present here confirmation of this long-standing hypothesis. By creating a genome wide association (GWA) mapping panel, we have identified multiple significant site-trait associations within the IM population of M. guttatus, and found extensive evidence of pleiotropy and polygenic adaptation. To understand how developmental processes might contribute to variation in flowering time and flower size, we next investigate genetic variation for circadian period length. The circadian clock is known to be upstream of floral induction pathways, and in Arabidopsis thaliana over one third of the genome is suspected to be under circadian regulation. We find extensive variation for endogenous period length, and genetic correlations between endogenous period length and days to germination, flower size measurements, height, and leaf size. Despite a small GWA analysis sample size, we find abundant evidence of overlap of sites found to be significant for growth and flowering traits and period length. Finally, we investigate how genomic variants identified in the greenhouse contribute to variation in a natural context. By planting F1 crosses of the lines sequenced for the GWA mapping panel in the field, we are able to confirm significant impacts on field growth and fitness, and identify a genomic site significantly associated with survival to seed production. Taken together our results not only confirm previous quantitative trait locus work at the nucleotide level, but present a newly developed sequence resource for analysis of intra-population variation in M. guttatus.