Hantavirus Research in South America
Much of the current research in the emergence and spread of infectious disease is focused on the role anthropogenic impacts (especially land cover change) on the pathogen-host-landscape system. This ongoing project investigates the role of land cover change in hantavirus dynamics in South America. Ultimately, this research will result in an improved capability for understanding and predicting outbreaks of hantavirus disease. Since the dynamics of hantavirus diesase are similar to other zoonotic diseases (e.g. Lassa Fever, Rift Valley Fever, Bolivian Hemorrhagic Fever), findings from this research can be applied to general disease predictive models. Zoonotic pathogens cause hundres of thousands of deaths per yar world-wide. Predictive capability may mitigate this suffering, directly benefiting society.
There are as many as five different strains of hantavirus in Paraguay, each host specific to a species of rodent. Goodin's previous research into patterns of deforestation in eastern Paraguay (funded in part by KNEP) has shown that human activities often result in a fragmented landscape. The hypothesis of this research is that human-induced landscape fragmentation favors the rodents that host hantaviruses, thus increased occurence of the virus within rodent populations should be associtated with anthropogenic land cover change.
This project funds a GRA who has responsibilities including the development and implementation of a spatial database for archiving and accessing geospatial data, as well as assisting in the analysis of the data. The analysis uses data from Ikonos, Landsat ETM+, and SPOT-VGT satellites, which is "fused" using a wavelet transformation in the IHS chromatic plane. This fused dataset allows for mapping and analysis of host habitat fragmentation at a scale more appropriate to the rodent viral hosts. These nested resolution data sets will be used in a "downscaling" model to predict conditions at the finer resolution using the coarser resolution data. This will enable the development of an effective multiresolution monitoring technique, in which large spatial areas can be observed at coarser resolution to indentify potential "hotspots" of viral activity. Any identified hotspots can then be observed at progressively finer resolutions and more refined estimates of viral activity can be made. This multiresolution technique will ultimately improve the capability to predict when and where a hantavirus outbreak might occur.
Because this research generates an enormeous quantity of spatial data which must be made available to a diverse team of researchers, the GRA funded by this proposal is also responsible for assisting in the access of the data. This is made possible through the use of an ArcIMS server, where there data can be accessed remotely. The database resides on KSU servers, so it is important that there is a maintence and support system in place at KSU. The database will also encompass integrated tools for epidemiological (SIR) and individual-based spatial modeling using the archived data.