Evolutionary dynamics of mating systems in populations of North American Arabidopsis lyrata

Hoebe, Petrus Nicolaas (2009) Evolutionary dynamics of mating systems in populations of North American Arabidopsis lyrata. PhD thesis, University of Glasgow.

Abstract

Plants can vary in their mating systems from completely inbreeding to completely outcrossing, with intermediate forms referred to as mixed mating systems. Arabidopsis lyrata is a strongly outcrossing perennial due to a sporophytic self incompatibility (SI) system. The species occurs in temperate regions of the Northern hemisphere where in Europe its SI system is fully working but around the Great Lakes of North America some populations of A. lyrata show a breakdown in SI. Consequently these North American populations are inbreeding or have a mixed mating system next to outcrossing populations with a working SI system. In this thesis I used North American A. lyrata to investigate the evolutionary consequences involving variation in mating systems.
First of all I was interested in the time that populations had been isolated from each other in the past that could explain differences in mating systems. In order to determine whether populations experienced a breakdown of SI independently or whether this originated from a single event I used chloroplast DNA (cpDNA) markers to reveal deep phylogeny and microsatellite markers to determine recent population genetic patterns. The results showed a loss of SI in populations from all three detected cpDNA haplotypes. Microsatellite data showed that predominantly inbreeding populations sharing one of these haplotypes showed high levels of homozygosity and that in all three haplotype lineages self-compatible individuals always had reduced heterozygosity compared to self-incompatible individuals. The data further showed that there had likely been at least two independent postglacial colonization routes to the north of the great lakes. This was consistent with phylogeographic studies of other organisms with limited dispersal such as reptiles and amphibians.
The next question was the role of inbreeding depression in the loss of SI. Inbreeding depression is defined as the decline of fitness after an inbreeding event. Inbreeding causes an increase in homozygosity that exposes recessive deleterious mutations, which would normally be sheltered in a heterozygous state, and causes a fitness decline. Individuals experiencing a loss of SI will have higher inbreeding levels and can result in inbreeding depression, which is thought to maintain the SI system. To gain more insight into the role of inbreeding depression in the shift from self-incompatibility to self-compatibility, I conducted an experiment in which I created outcrossed and selfed offspring from self-compatible and self-incompatible mothers from populations with different outcrossing histories. I monitored the offspring for early- and late acting fitness traits like germination rate, growth and time to flowering. I found inbreeding depression in only one late acting fitness trait, the increase in leaves 5 weeks after germination, to be significantly higher for self-incompatible than self-compatible individuals. I also conducted a regression analysis where relative fitness (the ratio of the fitness trait values of selfed and outcrossed offspring) per mother was regressed against population heterozygosity and found a significantly negative regression. This result suggested that individuals from a population with a relatively high heterozygosity suffered more from inbreeding depression than individuals from populations with a relatively low heterozygosity. This indicated that the history of outcrossing of a population, or purging, played an important role in the shift from outcrossing to inbreeding.
The detection of inbreeding depression could not be evident by only looking at life history traits under greenhouse conditions. But stressful environmental conditions like a pathogen infection could magnify inbreeding depression. I would expect that predominantly outcrossing populations would have a higher heterozygosity than predominantly inbreeding populations and therefore be able to show a higher fitness when exposed to a pathogen. To test this hypothesis I used four outcrossing and four inbreeding populations, which I infected with the crucifer pathogen Albugo candida and measured relative growth rates (RGR) and monitored resistance rates. The results showed that there were three infection phenotypes: resistant (no signs of infection), partially resistant (only the initially infected parts showed symptoms) and susceptible (symptoms present on the whole plant). The inbreeding populations showed a bimodal distribution of resistance as two populations showed a high rate of resistance and two showed a low rate of resistance. The outcrossing populations showed a much more uniform distribution of resistant individuals with a higher rate of partially infected individuals across populations than inbreeding populations. Resistant and partially resistant individuals did not differ significantly in their RGR from each other but both had a significantly lower RGR than the untreated control group and a significantly higher RGR than the susceptible individuals. This suggested a cost of resistance that was lower than a cost of being susceptible in the presence of a pathogen. There was no effect of mating system on RGR, which was primarily caused by the fact that two inbreeding populations contained a high amount of resistant individuals and an outcrossing population that showed a very low amount of partially resistant and resistant individuals. The difference in resistance to A. candida in A. lyrata differed much more between inbreeding than between outcrossing populations. This suggested that alleles responsible for resistance were concentrated in homozygous form in inbreeding populations and both homozygous and heterozygous form in outcrossing populations. This would mean that mating system plays a role in susceptibility, as resistance genes would be concentrated in certain individuals in inbreeding populations as opposed to a more modal distribution in outcrossing populations.
A shift in mating system often has an effect on floral traits, as there is a lack of necessity to attract pollinators. I wanted to test whether these changes were apparent in A. lyrata by comparing pollinator attractants and sexual floral traits between strongly outcrossing and strongly inbreeding populations. I hypothesized that individuals depending on pollinators for outcrossing would show a higher emission of volatiles and floral traits that had evolved to optimize pollen transmission to conspecifics. Autonomously selfing individuals would be independent of pollinators so should show a reduced volatile emission pattern, a floral trait composition that evolved to transmit pollen to their own stigma, and a reduction in floral display compared to outcrossers. My results showed a somewhat contradicting pattern as self-compatible individuals showed higher volatile emission than self-incompatible individuals but self-incompatible individuals showed larger petal size than self-compatible individuals. Pistil height and stamen length were strongly correlated but petal size seemed to co-vary relatively independent from pistil and stamen length. I found no effect of mating system on the evolvement of floral traits to optimize pollen to the stigma and contradicting patterns for pollinator attractant traits. Due to low sample sizes this study turned out to be a pilot study for further research so the results in this study were not conclusive at this stage.
Finally I conclude that SI has been lost independently several times and the low observed genetic load in the North American populations compared to the European populations could be responsible for that. There have probably been two independent colonization routes to the North of the Great Lakes following the last glaciation in which a Northern distributed cpDNA haplotype lineage seems to have a lower frequency of SC individuals than a southern cpDNA haplotype lineage. Inbreeding populations showed a bimodal distribution of infection phenotypes among individuals compared to outcrossing populations that showed a more evenly distributed of infection phenotypes among individuals which is thought to be caused by higher heterozygosity in outcrossing populations. Traits involved with pollinator attraction show a contradicting pattern with high volatile emissions for self-compatible individuals, which could be due to the dependence on pollinators for self-fertilization in self-compatible plants.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Keywords: plant, evolution, mating system, phylogeography, inbreeding depression, resistance, pollinator
• Subjects: Q Science > QK Botany; S Agriculture > S Agriculture (General); Q Science > Q Science (General)
• Colleges/Schools: College of Medical Veterinary and Life Sciences
• Supervisor's Name: Mable, Dr. Barbara
• Date of Award: 2009
• Depositing User: mr peter hoebe
• Unique ID: glathesis:2009-1405
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 18 Dec 2009
• Last Modified: 04 Feb 2013 15:58
• URI: https://theses.gla.ac.uk/id/eprint/1405