Increase of niche filling with increase of host richness for plant-infecting mastreviruses

Abstract

Now that it has been realized that viruses are ubiquitous, questions have been raised on factors influencing their diversity and distribution. For phytoviruses, understanding the interplay between plant diversity and virus species richness and prevalence remains cardinal. As both the amplification and the dilution of viral species richness due to increasing host diversity have been theorized and observed, a deeper understanding of how plants and viruses interact in natural environments is needed to explore how host availability conditions viral diversity and distributions. From a unique dataset, this study explores interactions of Mastrevirus species (family Geminiviridae) with Poales order hosts across 10 sites from three contrasting ecosystems on La Réunion. Among 273 plant pools, representing 61 Poales species, 15 Mastrevirus species were characterized from 22 hosts. The analysis revealed a strong association of mastreviruses with hosts from agroecosystems, the rare presence of viruses in coastal grasslands, and the absence of mastreviruses in subalpine areas, areas dominated by native plants. This suggests that detected mastreviruses were introduced through anthropogenic activities, emphasizing the role of humans in shaping the global pathobiome. By reconstructing the realized host-virus infection network, besides revealing a pattern of increasing viral richness with increasing host richness, we observed increasing viral niche occupancies with increasing host species richness, implying that virus realized richness at any given site is conditioned on the global capacity of the plant populations to host diverse mastreviruses. Whether this tendency is driven by synergy between viruses or by an interplay between vector population and plant richness remains to be established.

Keywords: Mastrevirus; Poales; agroecosystem; ecology; host–pathogen interaction network; viral metagenomics.

Figure 1.

Presence/absence matrix of the Poales species (rows) in the different sites (columns). The Poales species are organized by families (Cyperaceae, Juncaceae, and Poaceae) and colored and organized according to the classification of tribes. Origin and crop status (N: uncultivated native; W: uncultivated introduced; C: cultivated introduced) and life cycles (A: annual; P: perennial; A/P: annual or perennial; NA: unknown) are indicated in the first and second columns, respectively. Light colored squares indicate the presence of the Poales species. Black squares indicate the detection of mastreviruses for the corresponding species.

Figure 2.

Bipartite interaction graph representing the association between Poales species (left side of the diagram) and viral species and strains (right side of the diagram) build from all the available data from this work and previous studies. Poales species are colored and organized by tribe.

Figure 3.

Proportion of realized Mastrevirus species richness as a function of plant host richness. Sites are presented as colored dots (red for AE sites, blue for CG sites, and green for SG sites) whose sizes are proportional to the maximum number of Mastrevirus species that could have been observed knowing the plant virus diversity of the site. This maximum number of species is also indicated on each dot. For each site, the proportion is obtained by dividing the realized number of Mastrevirus species with the maximum number of Mastrevirus species achievable. Linear regression is indicated with the dotted line. The P-value for the linear regression slope to be different from zero is indicated on the top left.