Working with partners from Italy, Kazakhstan, and the US (Louisiana State University), Drs. Jocelyn
Mullins and Jason Blackburn developed ecological niche models of Bacillus
anthracis, the causative agent of anthrax, to predict the geographic
distribution of the disease across the landscape of each country. The work was
published in PLOS ONE (http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0072451).
Previous studies have built models of the pathogen using outbreak data, which
represent multiple genetic lineages of the pathogens. In Dr. Mullins recent
work, we showed that within Kazakhstan, the A1.a lineage of B. anthracis had a
different ecological envelope than models built using all outbreak data. The
A1.a lineage is a broadly geographically distributed lineage with a large
number of representative samples in the global pathogen collections that have
been used in global phylogenetic analyses. In this study, we focused on
building ecological niche models of the pathogen using only occurrence data
that had been genotyped using the MLVA-8 genotyping technique to assign each
outbreak location to the A1.a lineage. We built models for each country and
then transferred those models onto the other countries. The goal was to
determine if we could successfully predict B. anthracis suitable environments
across the landscapes using any given country. If the lineage had a single or
very broad ecological envelope, then transferred models should predict all
three landscapes well. To the contrary, we found that each country was best
predicted by the native locations defining pathogen presence. In this way, we
were not able to develop a single model that predicted all three landscapes,
suggesting that B. anthracis may have adapted to different regional or local
ecological conditions on each landscape.
This study presents an exciting direction forward linking ecological niche
modeling approaches with phylogenetics to improve our understanding of the
geography and ecological conditions associated with pathogen persistence across
different landscapes. The results also suggest that these modeling approaches
may be best interpreted on native landscapes and future work is needed to
identify better approaches for predicting unknown landscapes. Ecological niche
models provide researchers and policy makers with tools for estimating the
geographic potential for a species on the landscape. From an epidemiological
point of view, such models may inform policy by identifying conditions and
locations on the landscape where the pathogen may be persisting, providing maps
for surveillance planning and educational outreach. This study suggests that
such modeling should be coupled with high resolution genetic analyses
(improving upon the resolution limits of this current paper) to best predict
the geographic potential for the pathogen and anthrax outbreaks. Such modeling
remains an important task for anthrax, particularly as we learn more about the
persistence of the disease in wildlife populations where vaccination control is
untenable. It is also important to understand this disease in agrarian and
developing countries where the disease burden in humans is highest.
This work was completed by Dr. Jocelyn Mullins (DVM, PhD) as part of her
doctoral dissertation work within the Department of Geography. Dr. Mullins is
currently a fellow with the US CDC’s Epidemiological Intelligence Service,
stationed in Connecticut’s Department of Health. This work was partially funded
by the US Defense Threat Reduction Agency and the the UF Emerging Pathogens
Institute.