Spatial-temporal dynamics of pseudorabies virus infections in a wild boar populationT. Selhorsta, T. Müllera and C. Staubacha
a Bundesforschungsanstalt für Viruskrankheiten der Tiere, aInstitut für Epidemiologie bInstitut für Epidomiologische Diagnostik, Seestrasse 55, 16868 Wusterhaussen / Dosse, Germany
Abstract - In order to get a greater insight into the epidemiology of PrV infections in the wild boar (WB), the time-spatial spread of PrV in a densely populated area was - for the first time - analysed over a period of 10 years using a total of 13 220 serological data (seroneutralisation tests) collected between 1985 and 1994. The study was conducted in the German federal state of Brandenburg covering about 30 000 km2 using a visualisation based on a mathematical time-spatial smoothing technique. In addition, the infection risk for the wild boar was estimated using a Kernel density function. The wild boar population continuously increased as shown by the hunting index of population density (HIPD) ranging from 0.96 to 1.45 WB shot per km2 in 1993 and 1994, respectively. The study area exhibits the highest density of wild boars in Germany. Between 1985 and 1994 the seroprevalence for the whole study area increased on the average from 0.98% to 6.02%. When an indirect ELISA was available, higher seroprevalences were determined ranging from 5.33% to 9.45% between 1991 and 1994. The time spatial pattern of PrV spread is shown to occur in stages. PrV first appeared in 1987 in small clusters located in the southeast showing seroprevalences of 6.9% and 7.43% within the foci (1st stage). In 1988, the infection rapidly spread north on a large scale to cover nearly half of the study area resulting in decreasing seroprevalences (2nd stage). In a 3rd stage (1989/90), the conquered area shrank slightly accompanied by a simultaneous increase of the seroprevalence inside reaching 16% to 18% in two connected clusters. In 1991/92, a further significant increase occurred up to 30% on the average within the core areas. This process was accompanied by a broad front-spread in northern and western directions. In 1993/94, the spread reached its largest expansion. Additionally, new areas close to the western border became newly infected (4th stage). When topographical information was linked with the simulation of the time-spatial spread, rivers and highways were found to temporally control the spread of the infection. The relative infection risk for WB was shown to decrease over a 96.35 km distance from the anchor point located in the highest infested area observed in 1987. As shown by time-spatial analysis, the PrV seroprevalence within the study area did not increase continuously but in cyclic stages. It seems that if the seroprevalence remains constant over a distinct period of time, then the PrV infection will spread and vice versa. There is also some evidence that PrV spread is influenced by natural and artificial barriers. Another important factor is given by local hunting pressure in the different areas. Based on the infection risk function the PrV infection must have invaded neighbouring regions.
Corresponding author: T. Selhorst Tel.: (49) 33979 800 0; fax: (49) 33979 80 200;
© INRA, EDP Sciences 2000