Ectomycorrhizal associations function to maintain tropical monodominance: Studies from Guyana.

Tropical rain forests are the epicenters of tree diversity. Nonetheless, tropical monodominant forests, where one tree species comprises more than 50% of the tree species, co-occur in matrices of high-diversity, mixed rain forest. Several alternative mechanisms could produce this pattern, but one frequently cited observation is that most tropical monodominant trees form ectomycorrhizal (ECM) associations. The majority of other trees in mixed rain forest form arbuscular mycorrhizal (AM) associations, suggesting that ECM associations provide advantages to their monodominant trees; however, the mechanisms underlying this hypothesis have not been fully explored. Using a combination of field and laboratory experiments, I tested whether ECM colonization conveys a competitive advantage on Dicymbe corymbosa that results in the development of monodominant stands in Guyana. First, D. corymbosa seeds planted in the monodominant forest had 70% higher seedling survival and 86% higher ECM colonization after one year compared to seedlings planted in mixed forest. Connection of seedling roots to the common ECM network was found to be the mechanism for higher D. corymbosa seedling survival in the monodominant forest, potentially providing seedlings with photosynthate from overstory individuals of the same species. Seedlings with access to an ECM network had greater growth (73% greater), leaf number (55% more) and survivorship (47% greater) than seedlings without such access. Averaged over 2 yrs, aboveground litter production was higher in the mixed forest (11,510 kg ha-1 yr-1) relative to monodominant forest (8282 kg ha-1 yr-1), an observation that slower decomposition in the monodominant forest was responsible for a greater floor mass. A reciprocal litter bag transplant experiment demonstrated that leaf litter decomposition was slower in the monodominant forest (k = 0.44 yr-1) compared to the mixed forest (k = 0.93 yr -1), and that leaf type (D. corymbosa or mixed species leaf litter) had no effect on decomposition rate, verifying my contention that slow decomposition creates greater forest floor in the monodominant forest. Moreover, I found microbial biomass, bacterial-fungal ratios and biomarkers for broad saprotrophic bacteria were higher in the mixed forest compared to the monodominant forest. DGGE analysis revealed that fungal community composition was different between forest types and that forest type explained about 80% of the variation. Together, these observations suggest that ECM associations maintain monodominance by increasing seedling survival via connections to the ECM network and that ECM fungi were suppressing saprotrophic decomposition in the monodominant forest to gain preferential access to nutrients contained in forest floor.Ph.D.Biological SciencesEcologyUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/126775/2/3276244.pd