Recombination in natural populations of watermelon mosaic virus: new agronomic threat or damp squib?

Abstract

Since their introduction in south-eastern France around 1999, new, ‘emerging’ (EM) strains of watermelon mosaic virus (WMV) coexist with the ‘classic’ (CL) strains present for more than 40 years. This situation constitutes a unique opportunity to estimate the frequency of recombinants appearing in the few years following introduction of new strains of a plant RNA virus. Molecular analyses performed on more than 1000 isolates from epidemiological surveys (2004–2008) and from experimental plots (2009–2010), and targeting only recombinants that became predominant in at least one plant, revealed at least seven independent CL/EM or EM/EM recombination events. The frequency of recombinants involving at least one EM parent in the natural populations tested was on the order of 1 %. No new recombinant was detected for more than 1 year, and none but one in more than one location. In tests comparing host range and aphid transmissibility, the new recombinants did not display a better fitness than their ‘parental’ isolates. No recombinant was detected from artificial mixed infections of CL and EM isolates of various hosts after testing more than 1500 subcultures obtained after single-aphid transmission. These results constitute one of the first estimations of the frequency of recombinants in natural conditions for a plant RNA virus. This suggests that although viable recombinants of WMV are not rare, and although recombination may potentially lead to new highly damaging strains, the new recombinants observed so far had a lower fitness than the parental strains and did not emerge durably in the populations.