Do virus-resistant plants pose a threat to non-target ecosystems? I. Evidence from an Australian pathosystem based on glasshouse challenge experiments

One key environmental risk associated with the release of novel disease-resistant plants is the potential for non-target host populations to acquire resistance genes and undergo enemy release, leading to damage to associated native plant populations in high conservation-value ecosystems. Unfortunately, the dynamics of most natural pathosystems are poorly understood, and risk assessment of disease-resistant plants remains a challenge. Here we describe the first stage of a multi-tiered risk assessment strategy aimed at quantifying potential ecological release in a model pathosystem (the weedy pasture species Trifolium repens infected with Clover yellow vein virus; ClYVV) in order to assess the level of risk posed by genetically modified and conventionally bred disease-resistant host genotypes to non-target plant communities in south-eastern Australia. Glasshouse inoculation and growth experiments using 14 ClYVV isolates and 20 wild T. repens lines collected from high conservation-value montane grassland and woodland communities show that viral infection reduces the survival and growth of host plants by on average 10–50%. However, T. repens lines exhibited variable levels of resistance and tolerance to virus infection and ClYVV isolates differed in infectivity and aggressiveness, with grassland isolates having a greater pathogenic effect on associated host plants than woodland isolates. We conclude that ClYVV potentially plays an important role in limiting the size of T. repens populations in some at-risk non-target ecosystems and that second-tier field experiments are required to adequately quantify the risk associated with the commercial release of V-R T. repens genotypes in Australia.