Spatial modelling of small mammal distributions in relation to parasite transmission in western China

Marston, C

Abstract

This research investigates the spatial distributions of small mammal species which act
as intermediate host vectors for the parasitic tapeworm Echinococcus multilocularis, which
causes a significant burden of human disease in western China. Small mammal distributions
are modelled in relation to landscape characteristics derived from multiple Landsat TM,
Landsat ETM+ and MODIS satellite-derived datasets. Statistical models are used to identify
the landscape variables influencing the small mammal spatial distributions, and to describe
the nature of these relationships. They are then used predictively to determine probable small
mammal distributions over large areas. Assessment of landscape change in two study areas,
over a thirty year period, and its impact on small mammal communities is also modelled.
Results indicate that some small mammal distributions are significantly related to the
spatial distribution of the degraded grassland habitat, with landscape metrics analysis
indicating that these distributions are also positively related to degraded grassland patch size.
Small mammal distributions are also shown to be negatively related to single-date vegetation
index values, although the strength of these relationships is lower. When modelled against
time-series MODIS vegetation index data, small mammal distributions show stronger
relationships than when using the single-date Landsat ETM+ imagery. These validated timeseries
models are used to predict small mammal distributions to a high level of accuracy.
Although successful for locations where both MODIS and small mammal transect datasets are
available, these models are not temporally or geographically transferable. Long-term
landscape change assessed using Landsat MSS and ETM+ imagery showed that large scale
landscape degradation occurred at the Serxu site between 1977 and 2001, increasing the area
of degraded grassland, and the probability of small mammal presence. At the Yili site no
overall landscape change trends were displayed. Probability maps of small mammal
distributions have been produced, allowing the identification of potential Echinococcus
multilocularis transmission foci. This has potentially significant applications in aiding the
development of future disease control programmes.

• Thesis Type: Thesis
• Deposit Date: Aug 18, 2021
• Award Date: Nov 1, 2008