Monitoring WMV and ZYMV dynamics at local scale

The results strongly suggest independent introductions even in plots less than 1km apart and very spatially restricted dissemination of non-persistently transmitted viruses in fragmented landscapes as well as a low number of primary introductions for ZYMV.

Monitoring WMV and ZYMV dynamics was conducted between 2009 and 2013 in four experimental plots distant from 0.5 to 4km in a highly fragmented landscape. This spatio-temporal study revealed a high heterogeneity of epidemics in spite of short distances between plots: in several cases, no ZYMV was detected in one of the plots during all the growing season, whereas infection in the three other plots reached 100%, 7 to 12 weeks after planting. Besides, some plots exhibited very similar epidemical dynamics for ZYMV while molecular studies revealed that strains in the different plots belonged to distinct molecular groups.

Monitoring WMV dynamics at larger scale

The team gathered a unique dataset through a 5-year study in interaction with farmers, farm advisers and seed companies. More than two thousand plant samples were collected over an area of about 10500 km² in the region of Avignon.

The analysis of 144 WMV haplotypes identified – out of 1583 positive samples - showed that:

• emergent strains have probably been introduced several times independently through importation of cucurbit fruits or plant material form south-eastern Asia
• resident strains of WMV were almost entirely replaced by emerging strains in the course of a few years
• emergent strains were highly structured in space with potential barriers to dispersal corresponding to geographic obstacles (e.g. the Rhône river)

Monitoring of  a potyvirus, endive necrotic mosaic virus (ENMV), highly frequent in a wild host, Tragopogon pratensis (salsify)

The results showed that within habitats where the hosts occur as small and relatively isolated patches (i.e. orchards, vineyards, mosaics of small parcels of diverse annuals and permanent crops), the virus is less present but variants between host patches are more diversified than within habitats (where the hosts are abundant and evenly distributed (i.e. meadows)). This strongly suggests that genetic drift, resulting from low host abundance, is one of the predominant factors driving the spatial organization of the genetic diversity of viral populations in wild plants at the landscape scale.

This study had methodological as well as more fundamental aims:

1. develop deep sequencing techniques to better assess plant virus genetic diversity, especially co-infections that cannot be estimated using “Sanger sequencing”,
2. determine the usefulness of a new exploratory method (MAPI), developed by members of the team and collaborators, for landscape genetics approaches on plant viruses,
3. assess which processes are predominant in shaping the spatial genetic structure of a plant virus within a non-cultivated host.

A quasi-systematic sampling based on a grid constituted of 5x5km cells was realized over a surface of about 3000 km2, centered on the city of Avignon during two years (more than 3000 plants collected). 1247 ENMV variants were identified over the 2-year survey and the maximum number of variants detected within a same host (co-infections) reached 6 and 9 in 2014 and 2015, respectively. The spatial genetic structure of the viral population analyzed using MAPI revealed a clear correlation between spatial variations of population size and genetic diversity of ENMV and spatial heterogeneity in host abundance and distribution.