1 Transmission dynamics of Echinococcus multilocularis in a patchy environment

We illustrate some of the challenges and potential benefits of modelling parasite population dynamics in a spatially explicit manner, using the example of Echinococcus multilocularis in semi-arid areas of Kazakhstan. We show that by taking a parasite-focussed approach to modelling population dynamics, rather than the usual host-focussed approach, we gain new insights into the complex factors that affect parasite transmission. However, restrictive assumptions are required in order to use RAMAS GIS for modelling parasite ecology. For example, because the host is modelled as habitat, parasite-induced mortality must be negligible. Although these assumptions are probably valid in our case, they are unlikely to apply to many host-parasite systems. Introduction There has been an increasing realisation of the importance of incorporating spatial dynamics into models of the transmission dynamics of parasites and infectious diseases, particularly when hosts are free-living in the wild (Mollison & Levin 1995, Hess 1996, Kennedy 2001, Hudson et al., 2002). Spatial variation in parasite prevalences can reflect complex interactions between climatic suitability for free-living stages and host ecology (including host movement patterns, population dynamics and the extent to which primary and secondary host distributions overlap). Echinococcus multilocularis is a tapeworm which has wild carnivores as its definitive hosts and small mammals as its intermediate hosts (Eckert et al. 2001, Fig. 1). It is widespread throughout the Northern hemisphere, and causes fatal disease in humans if eggs are accidentally ingested. In Kazakhstan E. multilocularis has been found in 18 rodent species and 4 carnivore species (mostly in foxes). Its prevalence depends strongly on the biome (Fig. 2). This is chiefly because egg survival is strongly related to humidity and temperature (Eckert et al. 2001); under optimal conditions (4°C and high humidity) eggs can remain viable for up to 16 months. Temperatures in desert areas of Kazakhstan can reach 40°C in the summer, with very low humidity; under these conditions eggs are not likely to survive for more than a few hours. Hence extensive field surveys in arid areas of Kazakhstan only found infected rodents in patches with a humid microclimate, such as depressions or river valleys (Shaikenov & Torgerson, 2001). In mountains or river valleys, up to 8% of rodents examined were infected, and around 30% of carnivores. By contrast, in desert areas <1% prevalence was found in both rodents and foxes.