Family level phylogenies reveal modes of macroevolution in RNA viruses

Abstract

Despite advances in understanding the patterns and processes of microevolution in RNA viruses, little is known about the determinants of viral diversification at the macroevolutionary scale. In particular, the processes by which viral lineages assigned as different "species" are generated remain largely uncharacterized. To address this issue, we use a robust phylogenetic approach to analyze patterns of lineage diversification in five representative families of RNA viruses. We ask whether the process of lineage diversification primarily occurs when viruses infect new host species, either through cross-species transmission or codivergence, and which are defined here as analogous to allopatric speciation in animals, or by acquiring new niches within the same host species, analogous to sympatric speciation. By mapping probable primary host species onto family level viral phylogenies, we reveal a strong clustering among viral lineages that infect groups of closely related host species. Although this is consistent with lineage diversification within individual hosts, we argue that this pattern more likely represents strong biases in our knowledge of viral biodiversity, because we also find that better-sampled human viruses rarely cluster together. Hence, although closely related viruses tend to infect related host species, it is unlikely that they often infect the same host species, such that evolutionary constraints hinder lineage diversification within individual host species. We conclude that the colonization of new but related host species may represent the principle mode of macroevolution in RNA viruses.

Fig. 1.

Maximum-likelihood tree of the genus Alphavirus. Viral taxa are denoted in black, alongside their probable reservoir host range (in red) and categories used in the association analysis (in red and bracketed). The tree is midpoint rooted with quartet puzzling, and Bayesian posterior support values ≥70% are given above branches (quartet puzzling/Bayesian posterior).

Fig. 2.

Maximum-likelihood tree of the family Caliciviridae. Tree labels and description of rooting and node support values as in Fig. 1. Genera within the family are indicated by black bars to the right of the tree.

Fig. 3.

Maximum-likelihood tree of the genus Flavivirus. Tree labels and description of rooting and node support values as in Fig. 1. Vector class (mosquito, tick, or unknown) are indicated by black bars to the right of the tree.

Fig. 4.

Maximum-likelihood tree of the family Paramyxoviridae. Tree labels and description of rooting and node support values as in Fig. 1. Genera within the family are indicated by black bars to the right of the tree.

Fig. 5.

Maximum-likelihood tree of the family Rhabdoviridae. Tree labels and description of rooting and node support values as in Fig. 1. Genera within the family are indicated by black bars to the right of the tree.