2.2.1S
HOW PATHOGEN POPULATIONS CHANGE: UNDERSTANDING THE EVIDENCE

LM KOHN

Department of Botany, University of Toronto at Mississauga, 3359 Mississauga Road North, Mississauga, Ontario, Canada L5L 1C6

Background and objectives
The rationale in organizing this symposium on 'Population genetics: structure and dynamics' has been that we are still early on the road to rigorously demonstrating how plant pathogen populations change. The structure of pathogen populations can be described. In systems studied to date, key determinants of population structure have been the relative contributions of genetic exchange with recombination and asexual reproduction, of highly dispersive and more spatially localized propagules, and of temporal stability or instability of pathogen genotypes. The history of pathogen populations can be reconstructed. Phylogenetic methods offer far more than the depiction of relative similarity in a group of genotypes or isolates. Because they indicate the mutational and recombinational history of populations, gene genealogies are being used to test hypotheses on how populations change. Topology and branch lengths in the genealogies will reflect the formative evolutionary process, such as a bottleneck that occurs when a genotype is introduced and a subsequent diversification of genotypes. In our studies on Sclerotinia, gene genealogies indicate the importance of host jumps and bottlenecks in speciation. While the phylogenetic approach reconstructs history from an existing sample, experimental studies provide the empirical data needed to test hypotheses of selection or drift on the population dynamic. These may be in vitro experimental evolution studies, such as those that we are conducting on the human pathogen, Candida albicans, or field experiments, such as those on pathosystems to be discussed in this symposium.

Materials and methods
Hierarchical sampling and other aspects of experimental design will be discussed in this symposium. Our methods for inference of gene genealogies are based on DNA sequence data of known genomic regions. There are three caveats. First, the suite of nuclear ribosomal DNA sequences found in genetic databases may not be adequate or sufficient, especially at the intraspecific level. We utilize genetic databases to design primers sets that can amplify other, specific genomic regions across many different fungal species. Second, the type of character, e.g. single base substitutions, indels, and repeated motifs, such as micro satellites, may alter the topology or branch length. We investigate this by comparing genealogies. Third, to understand evolutionary history, the genealogy should be rooted but the outgroup in a population study should not be too unrelated to the ingroup. We have applied a coalescence-based method to determine the probabilities for each real and putative lineage being the root.

Results and conclusions
Congruence of gene genealogies within samples from Canadian and Norwegian canola of Sclerotinia sclerotiorum indicates a predominantly clonal population structure and is consistent with previous genotyping by DNA fingerprinting with a transposable element [1,2]. Expanding the investigation to a sibling species on a wild plant, Ranunculus ficaria, to other agricultural population samples, to S. minor, Sclerotium cepivorum and to other species on wild plant hosts, we test the hypothesis that host jumps and bottlenecks have been key in speciation. While our work investigates processes at the population-species interface, other work to be presented in this symposium will focus on population processes.

References
1. Kohn LM, 1995. Canadian Journal of Botany 73(Suppl. 1), S1231-S1240.
2. http://www.erin.utoronto.ca/~kohn/