The effectiveness of various rabies spatial vaccination patterns in a simulated host population with clumped distribution

In Israel's Negev Desert, periodic episodes of rabies epidemics erupt every 5-7 years despite low densities of the main reservoir species of rabies in this area. Canids in this desert are spread non-uniformly over space, with high densities in and around human related waste sites and low densities (presumably below the disease threshold) in the rest of the region. The persistence of the disease in the Negev desert is due likely to the high host density patches around human waste sites. We used an individual-based spatially explicit model to test whether a non-uniform spread of oral rabies vaccination could be more effective in eradicating the disease than the commonly used uniform distribution. We evaluated two models based on a 20 x 20 grid: the first with 17 high-density patches distributed evenly over the grid and the second with 16 high-density patches clumped in four groups of four. We tested six patterns of spreading the oral vaccination under four different host dispersal ranges. Immunizing only the high-density patches was ineffective in all cases. Immunizing only low host-density patches eradicated the disease within 5 years, on average. Immunizing only the areas surrounding the high-density patches was ineffective except when high-density patches were not clumped and dispersal distance of the host was short. We conclude that in desert environments. where host densities vary over space, non-uniform spreading of oral rabies vaccination may, under certain circumstances, be more effective than the commonly used uniform spread. However, this requires good knowledge of the host's dispersal patterns and its distribution over space. (C) 2002 Elsevier Science B.V. All rights reserved.