Inbreeding depression in artificial selection lines of Ipomoea purpurea

Inbreeding depression is a central parameter underlying mating system variation in nature and one that can be altered by environmental stress. Although a variety of systems show that inbreeding depression tends to increase under stressful conditions, we have very little understanding across most organisms how the level of inbreeding depression may change as a result of adaptation to stressors. In this work we examined the potential that inbreeding depression varied among lineages of Ipomoea purpurea artificially evolved to exhibit divergent levels of herbicide resistance. We examined inbreeding depression in a variety of fitness-related traits in both the growth chamber and in the field, paired with an examination of gene expression changes. We found that, while inbreeding depression was present across many of the traits, lineages artificially selected for increased herbicide resistance often showed no evidence of inbreeding depression in the presence of herbicide, and in fact, showed evidence of outbreeding depression in some traits compared to non-selected control lines and lineages selected for increased herbicide susceptibility. Further, at the transcriptome level, the resistant selection lines had differing patterns of gene expression according to breeding type (inbred vs outcrossed) compared to the control and susceptible selection lines. Our data together indicate that inbreeding depression may be lessened in populations that are adapting to regimes of strong selection.

Methods
Plants were artificially selected for increased or decreased resistance (resistant and susceptible lines, respectively) to herbicide; control line was randomly mated. After 2 generations of selection, seeds were created by either outcrossing (within the line) or selfing. These plants were assayed to determine inbreeding depression in a variety of traits. Survival and biomass post-herbicide was collected from plants grown in growth chamber with varying levels of herbicide. Reproductive trait data was collected from field plants, either sprayed or not sprayed with herbicide. For a subset of those field-grown plants, transcriptomic response was measured at 8 and 32 hours post herbicide spray.