SOIL PROPERTIES INFLUENCING THE SPATIAL DISTRIBUTION OF ANTHRAX IN NORTHWEST MINNESOTA
Statistical models, like Maximum Entropy (Maxent) and logistic regression, can use environmental data along with known locations of a species to define the ecological niche of that species. However, many models are limited by coarse spatial data, which compromises their ability to predict which soil variables best support B. anthracis. Maxent models using coarse resolution data from government agencies has shown that high sand, organic matter, calcium content, and high pH, are the strongest predictors for anthrax outbreaks in soils in northwest Minnesota.
We collected 99 soil samples around known anthrax infection locations in northwest Minnesota and measured soil properties, including pH, particle size distribution, total organic matter and Ca, Na, P, Mn and Sr concentrations, to determine their effect on the spatial distribution of B. anthracis in soil. An additional 26 soil samples were taken from an area with no known anthrax cases as a control. The resultant data will be used to determine which of the soil parameters are mostly closely associated with B. anthracis using Maxent and logistic regression and provide tentative thresholds for anthrax to occur in a given area.