97th ESA Annual Meeting (August 5 -- 10, 2012)

COS 144-5 - Phylogeographic structure of the Neotropical epiphytic orchid Epidendrum firmum corresponds to differences in symbiotic fungal communities

Thursday, August 9, 2012: 9:20 AM
B117, Oregon Convention Center
Tyler R. Kartzinel, Ecology and Evolutionary Biology, Princeton University, Princeton, NJ, Richard P. Shefferson, General Systems Studies, University of Tokyo, Meguro-ku, Japan and Dorset W. Trapnell, Plant Biology, University of Georgia, Athens, GA
Background/Question/Methods

Historical biogeography, habitat suitability, the distribution of essential mycorrhizal fungi, and the dispersal of pollen and seeds across a landscape all contribute to spatial genetic structure of rare orchid species. We investigated spatial genetic structure of the Neotropical epiphytic orchid Epidendrum firmum in montane forests of Costa Rica. We first hypothesized that strong phylogeographic breaks occur between mountain regions, as a result of small effective population sizes for a species that is restricted to a narrow elevation range of suitable habitat, infrequent dispersal among mountains, and/or recent colonization following a period of intense volcanic activity c0.6Ma. Like all orchids, the lifecycle of E. firmum completely depends on symbiosis with mycorrhizal fungi. Thus, our second hypothesis is that phylogeographic breaks are associated with differences in communities of fungal symbionts. We sampled thirteen populations from four mountain regions in Costa Rica, then used chloroplast DNA (cpDNA) sequences and nuclear microsatellites to investigate levels and partitioning of genetic variation. A network algorithm and Bayesian clustering analysis were to identify genetic boundaries in our cpDNA and microsatellite datasets, respectively. Finally, we used genetic barcoding techniques to quantify the phylogenetic community dissimilarity of fungal symbiont communities associating with different populations of E. firmum.

Results/Conclusions

We found high genetic diversity in cpDNA (24 haplotypes) and nuclear microsatellites (mean expected heterozygosity = 0.822 ± 0.011). Consistent with our first hypothesis, we found three genetic barriers for cpDNA (>50% bootstrap support) that correspond to four significantly different mountain regions. However, Bayesian clustering assigned individuals in most populations to a single genetic cluster. Analysis of molecular variance (AMOVA) found 42.35% (P≤0.001) of cpDNA variation was partitioned among these regions while only 2.17% (P≤0.001) of microsatellite variation was partitioned among these regions. Nuclear microsatellites are dispersed by both pollen and seeds whereas cpDNA is dispersed only by seeds, which could account for differences in genetic structure for the two marker types. Genetic barcoding of fungal symbionts identified diverse communities from the orchid-mycorrhiza families of Tulasnellaceae and Sebacinaceae. Different taxa from the Tulasnellaceae, but not the Sebacinaceae, associated with E. firmum on either side of a strong phylogeographic break in northwestern Costa Rica, providing mixed support for our second hypothesis. Experiments are needed to distinguish between the possibilities that orchid populations adapt to different communities of symbiotic fungi or that the orchid demonstrates flexibility to associate with an unusually broad array of fungi in different ecological contexts.