Why do RNA viruses recombine?

Abstract
RNA viruses are able to undergo two forms of recombination: RNA recombination, which (in principle) can occur in any type of RNA virus, and reassortment, which is restricted to those viruses with segmented genomes. Rates of RNA recombination vary markedly among RNA viruses. Some viruses, particularly those with negative-sense single-stranded genomes, exhibit such low rates of recombination that they are effectively clonal. By contrast, some positive-sense single-stranded RNA viruses and some retroviruses such as HIV exhibit high rates of recombination that can exceed the rates of mutation when measured per nucleotide. Although recombination is often argued to represent a form of sexual reproduction, there is little evidence that recombination in RNA viruses evolved as a way of creating advantageous genotypes or removing deleterious mutations. In particular, there is no association between recombination frequency and the burden of a deleterious mutation. Similarly, there is little evidence that recombination could have been selected as a form of genetic repair. The strongest association for rates of recombination in RNA viruses is with genome structure. Hence, negative-sense single-stranded RNA viruses may recombine at low rates because of the restrictive association of genomic RNA in a ribonucleoprotein complex, as well as a lack of substrates for template switching, whereas some retroviruses recombine rapidly because their virions contain two genome copies and template switching between these copies is an inevitable part of the viral replication cycle. We therefore hypothesize that recombination in RNA viruses is a mechanistic by-product of the processivity of the viral polymerase that is used in replication, and that it varies with genome structure.