Estimation of the effective number of founders that initiate an infection after aphid transmission of a multipartite plant virus.

The fecundity of RNA viruses can be very high. Thus, it is often assumed that viruses have large populations, and RNA virus evolution has been mostly explained using purely deterministic models. However, population bottlenecks during the virus life cycle could result in effective population numbers being much smaller than reported censuses, and random genetic drift could be important in virus evolution. A step at which population bottlenecks may be severe is host-to-host transmission. We report here an estimate of the size of the population that starts a new infection when Cucumber mosaic virus (CMV) is transmitted by the aphid Aphis gossypii, based on the segregation of two CMV genotypes in plants infected by aphids that acquired the virus from plants infected by both genotypes. Results show very small effective numbers of founders, between one and two, both in experiments in which the three-partite genome of CMV was aphid transmitted and in experiments in which a fourth RNA, CMV satellite RNA, was also transmitted. These numbers are very similar to those published for Potato virus Y, which has a monopartite genome and is transmitted by aphids according to a different mechanism than CMV. Thus, the number of genomic segments seems not to be a major determinant of the effective number of founders. Also, our results suggest that the occurrence of severe bottlenecks during horizontal transmission is general for viruses nonpersistently transmitted by aphids, indicating that random genetic drift should be considered when modeling virus evolution.