Tuesday, August 5, 2008 - 5:00 PM

OOS 11-11: Sequencing the boreal forest: What do 70,000 new sequences tell us about fungal ecology?

D. Lee Taylor1, Jack W. McFarland2, Michael G. Booth2, Ian C. Herriott2, Teresa N. Hollingsworth2, Niall J. Lennon3, and Chad Nusbaum3. (1) University of Alaska, (2) Institute of Arctic Biology, (3) Broad Institute of MIT & Harvard

Background/Question/Methods

Fungi are key players in nutrient cycling in the boreal forest, yet their diversity and specific roles are completely undescribed.  Boreal forest soils contain roughly one quarter of Earth's labile organic carbon, but climate warming and climate-driven changes in disturbance regimes are likely to substantially alter carbon dynamics. We are characterizing fungal community structure across habitats, successional stages, and soil horizons, primarily within the Bonanza Creek LTER site of interior Alaska.  Here we present an overview of results from high throughput sequencing of 14 PCR clone libraries constructed from 2100 soil cores collected across 21 sites sampled twice in consecutive years.  Nine of the sites represent early (willow-aspen), mid (birch-aspen) and late (white spruce) successional stages in Bonanza Creek upland ecosystems, while 12 of the sites represent four distinct black spruce community types. 

Results/Conclusions

To date, the project has generated roughly 70,000 bidirectional clone sequences which reveal that boreal forest soil fungal communities are remarkably diverse in species numbers and in deep phylogenetic breadth.  Approximately 2000 operational taxonomic units (OTUs) were recovered from the black spruce sites and diversity was even higher in the upland sites.  Fifty-eight percent of the black spruce OTUs occurred once in the dataset (i.e. “singletons”), suggesting that even with our massive sequencing efforts, we have not fully captured the diversity present.  Fungal community composition varied relatively little between years, while shifts in species composition through successional stages were apparent.  We also observed differences among communities in acidic and non-acidic black spruce sites, with higher diversity indices (due to greater evenness of dominants) in the non-acidic sites.  However, community composition was much more strongly correlated with soil horizon (litter vs. humic vs. mineral) than any other factor in every study site.  Because we sequenced a rapidly evolving marker, the nuclear ribosomal ITS plus LSU, we were able to identify many of the OTUs to species or genus and infer the trophic guilds to which they belong.  Plant-associated taxa (especially ericoid and ectomycorrhizal) were far more abundant (typically >60% of clones) than classical decomposers, i.e. “saprophytes” (typically <15% of clones) in each study site.  Because clone numbers are roughly proportional to biomass, our results strongly suggest that plant-associated fungi account for a much greater fraction of the living mycelium in soil than do decomposer fungi.  This observation has implications for plant nutrition and the dynamics of carbon sequestration and release from the massive pools present in boreal forest soils.