Abstract:
Little is known about the ecological and genetic factors that influence
plant-pathogen interactions in nature, although theories about the importance
of pathogens assume that certain factors are paramount. The idea that
pathogens regulate host population size assumes that pathogens respond
strongly to host plant density. The idea that pathogens favor sexual
reproduction in plants assumes that they respond strongly to host plant
genotypes. As microscopic organisms, it is possible that plant pathogens are
most affected by local environmental factors.
In this dissertation, the influences of some potentially important
factors were examined in a three-year spatial study of a Long Island pine
barrens population of the perennial rush Juncus dichotomus naturally
infected by the smut fungus Ciniractia junci, and in three multifactorial field
experiments. For the spatial study, the relationships between disease and density, nonhost plant density, soil water content, and host genetic variation
were examined by comparing surface plots and spatial autocorrelations,
and/or by calculating partial Mantel-test correlations. The experiments
consisted of plantings and inoculations of propagated replicates of
uninfected plants that were collected from the natural population and
genotyped electrophoretically. One experiment tested the effects of host
density, pathogen density, host genotype, and block heterogeneity; another
tested the effects of heterospecific neighbors, host genotype, and block
heterogeneity; the last tested the effect of soil water, host genotype, and
block heterogeneity.
In the natural population, disease had stable spatial pattern, and
disease and host density were positively correlated, with disease appearing
to thin the population in a density-dependent manner. No associations were
detected between disease and the other factors. Experiments showed that
plants exposed to higher pathogen loads were more likely to be infected,
that plants grown at high host density or in a dense stand of heterospecifics
were smaller and less likely to become infected, that plant genotypes were
variable for infection, that genotype X environment interactions were
evident, and that environmental heterogeneity influenced infection.