<i>Arabidopsis thaliana</i> as a model for the study of plant–virus co-evolution

Understanding plant-virus coevolution requires wild systems in which there is no human manipulation of either host or virus. To develop such a system, we analysed virus infection in six wild populations of Arabidopsis thaliana in Central Spain. The incidence of five virus species with different life-styles was monitored during four years, and this was analysed in relation to the demography of the host populations. Total virus incidence reached 70 per cent, which suggests a role of virus infection in the population structure and dynamics of the host, under the assumption of a host fitness cost caused by the infection. Maximum incidence occurred at early growth stages, and co-infection with different viruses was frequent, two factors often resulting in increased virulence. Experimental infections under controlled conditions with two isolates of the most prevalent viruses, cauliflower mosaic virus and cucumber mosaic virus, showed that there is genetic variation for virus accumulation, although this depended on the interaction between host and virus genotypes. Comparison of Q(ST)-based genetic differentiations between both host populations with F(ST) genetic differentiation based on putatively neutral markers suggests different selection dynamics for resistance against different virus species or genotypes. Together, these results are compatible with a hypothesis of plant-virus coevolution.