Content area
Abstract
The evolution of self-fertilization and inbreeding depression (ID) are intertwined. ID is thought to inhibit the shift from outcrossing to selfing. The shift to self-fertilization may purge some ID. To understand why outcrossing is maintained and predict how and whether it might evolve to lower levels we can evaluate genetic variation for the level outcrossing and ID. In this thesis, I describe the genetic variation in outcrossing and ID within self-incompatible (SI) populations of Leavenworthia alabamicato answer the question: why is self-incompatibility and ID maintained?
In Chapter 1, I studied genetic variation for self-compatibility in four SI populations of L. alabamica. I measured the heritability of post-selfing pollen tube number and tested for the co-segregation of post-selfing pollen tube number and S-alleles. Ninety percent of plants exhibited some level of self-compatibility across all four populations. The heritability of the magnitude of self-compatibility ranged from 0.38-0.56. Self-compatibility was only linked to one of the seven S-alleles in the most highly self-compatible families. Most of the self-compatibility we detected was intermediate (not fully self-compatible) and not linked to the S-locus. I refer to this as pseudo-self-compatibility (PSC) because plants were somewhat self-compatible but carried fully functioning S-alleles. I conclude that PSC is widespread in this species, and likely due to genetic variation in the signalling cascade downstream of SI reaction that renders it only partially functional.
In Chapter 2, I tested if ID can be purged from self-incompatible plants by three generations of forced selfing. An experimental (“Ancestral”) treatment was first created from self-incompatible plants of L. alabamica. Lines derived from an experimental “Ancestral” population were propagated by self-pollination for three generations in the attempt to create a “Purged” population. ID as well as individual fitness components in the "Purged" and "Ancestral" populations were largely the same. There was a small reduction in ID at the germination life stage, but no change in lifetime ID. This suggests that ID in this species is largely caused by genetic variation that is difficult to purge.
In Chapter 3, I mapped ID loci using offspring from three families each produced by forcing self-incompatible L. alabamicaplants to self-fertilize. I obtained maximum likelihood estimates of selection and dominance coefficients and tested for their significance with Monte Carlo simulations. One unlinked recessive deleterious allele was found in family 1. In family 1, I identified multiple examples of ID that is difficult to purge including sheltered load around the S-locus and an underrepresentation of the least frequent homozygous genotype across the genome that suggests that ID includes many deleterious recessive alleles, which could cause selective interference. No purgeable inbreeding depression was found in the other three families. These results are concordant with those obtained in Chapter 2, that ID is likely caused by widespread, weakly deleterious alleles, that are difficult to purge.
Overall, the results of this thesis suggest that while genetic variation for the level of outcrossing exists in populations of Leavenworthia that a contain a functional S-locus (Chapter 1), ineffective purging ID presents a barrier to the evolution of high levels self-fertilization (Chapters 2 and 3). These findings align with the observation that self-incompatibility and inbreeding depression is maintained in most L. alabamicapopulations.





