CRC for Legumes in Mediterranean Agriculture, University of Western Australia, Nedlands, WA 6907, Australia, and Agriculture Western Australia, Locked Bag No. 4, Bentley Delivery Centre, WA 6983, Australia

Background and objectives
Epidemiological studies are vital to establish the risks of introducing seedborne pathogens to new localities. The knowledge gained helps us not only to understand the key factors determining success or failure of pathogen establishment, but also to develop appropriate cultural, chemical and/or host-resistance strategies to eradicate, contain or manage them. The best information is provided by field experiments which demonstrate the consequences of introducing pathogens to new sites by sowing disease-carrying seed, preferably as sowings with different levels of seed infection. Such experiments need to be done in different localities and years so that information is obtained on the impacts of different climatic, soil and biological factors impacting on pathogen survival and spread. For quarantine reasons, these field investigations are dfficult to undertake with exotic pathogens, but the general principles involved can be studied using common seedborne pathogens. Over the past 10 years, such studies have been done with two common aphid and seedborne viruses, cucumber mosaic (CMV) and alfalfa mosaic (AMV), in grain and annual pasture legumes. This work was done in the mediterranean-type climate of the agricultural region of southwest Western Australia, where the growing season extends from late autumn to spring and the dry summer makes it difficult for viruses that are not seedbome to survive between growing seasons.

Materials and methods
Replicated field experiments were sown in 1987-97 at diverse sites. With CMV, infected seed of narrow-leafed lupin (Lupinus angustifolius) and subterranean clover (Trifolium subterraneum) with different infection levels was used. With AMV, infected seed of annual burr medic (Medicago polymorpha) was sown. Plots were separated by wide non-host buffers of cereal (lupins) or ryegrass (pasture swards). Pasture swards were grazed by sheep and allowed to regenerate annually for up to 7 years. Measurements taken from crop and pasture experiments included virus disease progress, numbers of aphid vectors, herbage and seed yields, and levels of infection in newly produced seed. Leaf samples were tested by ELISA to confirm visual diagnoses based on symptoms or to determine virus levels directly.

Results and conclusions
In the field experiments, the actual infection and yield loss scenarios that unfolded at individual sites varied greatly from year to year. The key factors determining whether virus spread was substantial, whether economic grain or herbage yield loss occurred, and whether infection of newly produced seed was high, were the level of infection in the seed originally sown and the time of arrival and abundance of aphid vectors. The extent of rainfall before and at the start of the growing season was critical in determining (i) the ability of seed-infected plants to survive well and be available to act as sources for virus spread to healthy plants by aphids; and (ii) when aphids would arrive and initiate virus spread. Higher rainfall before and at the start of the growing season resulted in greater virus spread, yield losses and infection of newly produced seed. AMV persisted for at least seven growing seasons in self-regenerating, mixed-species swards originally sown with infected burr medic seed, but lateral spread was slow. In mixed-species swards originally sown with infected subterranean clover seed, CMV was no longer detected after the fifth year from sowing. For lupins, a critical threshold level of <0.5% CMV infection in seed sown to produce crops was established for all but the highest 'risk zones' where <0.1% was more appropriate. A threshold level of <0.1% was established for AMV in medic seed used to sow new pastures, but subterranean clover seed with higher levels of infection could be sown. These findings will be used to establish underlying principles enabling development of an improved understanding of (i) the risks associated with introducing seedborne pathogens to new locations; (ii) the factors likely to result in their successful establishment and subsequent dispersion; and (iii) the prediction of likely infection scenarios based on different sets of circumstances.