Population genetics and phylogeography of alfalfa mosaic virus in China and a comparison with other regional epidemics based on the cp gene

Abstract

Alfalfa mosaic virus (AMV) is the most pervasive epidemic virus affecting alfalfa production. However, detailed investigations on the molecular population genetics and evolutionary dynamics of AMV are scarce. This study aimed to report on a large-scale long-term survey of genetic variability in AMV populations from China and perform a comparative analysis of AMV population genetics in the three most thoroughly studied countries to date: China, Iran, and Spain. The study was based on the analysis of the coat protein gene (cp) using two analytical approaches: an analysis of molecular variance (AMOVA) and a Bayesian Markov Chain Monte Carlo approach that investigates the association between geographical origin and phylogeny. Both analytical approaches found significant genetic differentiation within localities, but not among localities nor among provinces. This observation might result from inappropriate agronomical practices involving extensive exchange of plant materials followed by rapid viral diversification within localities. In the Chinese population, both methods found that genetic diversification in AMV was strongly associated with different bioclimatic zones. Rates of molecular evolution were similar in the three countries. The estimated epidemic exponential population size and growth rate suggest that the epidemics grew faster and with higher incidence in Iran, followed by Spain and China. Estimates of the time to the most recent common ancestors suggest that AMV was first seen in Spain by the beginning of the twentieth century and later on in eastern and central Eurasia. After ruling out the existence of recombination breakpoints within the cp gene, a codon-based selection analysis per population was performed and identified many codons under significant negative selection and a few under significant positive selection; the latter varied among countries, suggesting regional differences in selective pressures.