Risk analysis applied to biological control of a forest weed, using the gaussian plume model

Abstract

Abstract Biological control of Primus serotina, a forest weed in the Netherlands, by means of the silverleaf fungus Chondrostereum purpureum is studied. The fungus is a common saprophyte in dead wood. As a wound invader, it can cause silverleaf disease of P. serotina and fruit trees. Biological control of P. serotina within a forest leads to infection, sporulation and spore dispersal, and thus creates a risk to fruit trees outside the forest. To estimate this risk, spore immission was calculated by means of the Gaussian plume model (GPM). A short-term version of the GPM was used to calculate the ‘absolute risk’ as the spore density outside the forest due to biological control. Mean spore densities at 5000 m distance downwind of the forest were at least one order of magnitude lower than at 500 m. Maximum spore density at 5000 m downwind of the forest was calculated to be 32 spores.m-3, and at 500 m 141 spores.m-3. It was concluded that risk to fruit trees was appreciable at 500 m distance and negligible at 5000 m distance. The ‘relative risk’ was calculated as the ratio of added spore density, due to biological control, over natural spore density, due to normal presence of C. purpureum in nature. A long-term version of the GPM was used for this purpose. Calculations were performed for the district of Olst using literature values for emission of natural sources. Added spore densities during the day (10-1 spores.m-3) and at night (100 spores.m-3) were of the same order of magnitude as the natural spore densities. It was concluded that the added risk to orchards in a rural area was about equal to the natural risk. Assumptions and approximations in the risk analysis and the merits of the GPM are discussed in view of the acceptability of biological control of P. serotina.