Debbie Thackray

An Epidemiological Model for Externally Sourced Vector-Borne Viruses Applied to Bean yellow mosaic virus in Lupin Crops in a Mediterranean-Type Environment

A hybrid mechanistic/statistical model was developed to predict vector activity and epidemics of vector-borne viruses spreading from external virus sources to an adjacent crop. The pathosystem tested was Bean yellow mosaic virus (BYMV) spreading from annually self-regenerating, legume-based pastures to adjacent crops of narrow-leafed lupin (Lupinus angustifolius) in the winter-spring growing season in a region with a Mediterranean-type environment where the virus persists over summer within dormant seed of annual clovers. The model uses a combination of daily rainfall and mean temperature during late summer and early fall to drive aphid population increase, migration of aphids from pasture to lupin crops, and the spread of BYMV. The model predicted time of arrival of aphid vectors and resulting BYMV spread successfully for seven of eight datasets from 2 years of field observations at four sites representing different rainfall and geographic zones of the southwestern Australian grainbelt. Sensitivity analysis was performed to determine the relative importance of the main parameters that describe the pathosystem. The hybrid mechanistic/statistical approach used created a flexible analytical tool for vector-mediated plant pathosystems that made useful predictions even when field data were not available for some components of the system.

An epidemiological model for externally acquired vector-borne viruses applied to Beet western yellows virus in Brassica napus crops in a Mediterranean-type environment

A hybrid mechanistic/statistical model developed previously to predict vector activity and epidemics of vector-borne viruses was modified to simulate virus epidemics in the Beet western yellows virus (BWYV) – Brassica napus pathosystem. BWYV, which is persistently aphid-borne, spreads to B. napus crops from external sources and causes substantial yield losses when there is widespread infection of young plants. Risk that such losses may occur depends on the magnitude and availability of viral inoculum in the external source, the amount of biomass available to support aphid vectors, its duration before crop emergence, and the time of arrival of vector aphids in the crop. The model uses daily rainfall, temperature, and evaporation data from over 450 sites in the grainbelt of south-western Australia to track biomass levels throughout the growing season. This information is then used to simulate aphid vector populations and virus incidence, initially in the external source environment, then in the crop, and ultimately to provide risk forecasts. The model predicted BWYV spread successfully for 10 of 12 different datasets from 3 years of field observations on B. napus blocks at 4 sites representing different rainfall and geographic zones of the grainbelt. Sensitivity analysis was used to determine the relative importance of the main parameters that describe the pathosystem and to predict which control measures are likely to be useful. An analysis of timing of predictions v. their accuracy was also done to establish optimum timing of forecasts for BWYV epidemics in B. napus crops.