Environmental Science and Engineering Seminar

Pollution from agricultural and urban areas fuels excessive algae and cyanobacteria
growth, resulting in low oxygen dead-zones during decomposition. These microbial
processes deteriorate water quality, reduce the habitat of many economically important
aquatic animals and drive biogeochemical processes that alter nutrient cycling and
generate potent greenhouse gases. Since population growth and climate change are
expected to exacerbate these problems, understanding the dynamic chemical and
microbial changes that impact aquatic dead-zones will aid modeling efforts that guide
remediation strategies. I will present work to 1.) improve our understanding of the
relationship between genes, populations and biogeochemical processes to improve
predictive biogeochemical models and 2.) identify viral infections that contribute to
cyanobacteria mortality with a novel high-throughput, culture-independent method,
epicPCR. To investigate the relationship between microbial genes, populations and the
biogeochemical processes they mediate, we used genome reconstruction from
metagenomic data and a previously developed biogeochemical model to identify
microbial populations implicated in major biogeochemical transformations in a model
lake ecosystem. By reconstructing microbial genomes from complex assemblages of
microorganisms, we gained insight into microbial processes in the lake and identified
additional biogeochemical processes previously omitted from the model that could
significantly alter the predicted biogeochemistry of the lake if active. We are also
investigating the relationship between microbes, their genes and model predictions in a
more complex ecosystem, the Chesapeake Bay. Viral infections will also be identified in
the Chesapeake Bay through epicPCR. Identifying populations under the most viral
pressure in the environment can improve models of biogeochemical cycling, providing a
holistic picture of viruses in the trophic structure of marine environments. Yet, these
efforts are stalled because the specific host a virus infects remains largely unknown for a
majority of observed viruses. We hope to identify infections that contribute to ecological
shifts and alter biogeochemical processes with our to high-throughput, cultureindependent
approach. Although this method is currently under development, preliminary
data suggests the approach can identify specific infections in the environment and reveal
the complex network of viral infections in natural microbial communities.