Network Analysis of the Papaya Orchard Virome from Two Agroecological Regions of Chiapas, Mexico

Abstract

The study of complex ecological interactions, such as those among host, pathogen, and vector communities, can help to explain host ranges and the emergence of novel pathogens. We evaluated the viromes of papaya orchards, including weed and insect viromes, to identify common viruses in intensive production of papaya in the Pacific Coastal Plain and the Central Depression of Chiapas, Mexico. Samples of papaya cultivar Maradol, susceptible to papaya ringspot virus (PRSV), were categorized by symptoms by local farmers (papaya ringspot symptoms, non-PRSV symptoms, or asymptomatic). These analyses revealed the presence of 61 viruses, where only 4 species were shared among both regions, 16 showed homology to known viruses, and 36 were homologous with genera including Potyvirus, Comovirus, and Tombusvirus (RNA viruses) and Begomovirus and Mastrevirus (DNA viruses). We analyzed the network of associations between viruses and host-location combinations, revealing ecological properties of the network, such as an asymmetric nested pattern, and compared the observed network to null models of network association. Understanding the network structure informs management strategies, for example, revealing the potential role of PRSV in asymptomatic papaya and that weeds may be an important pathogen reservoir. We identify three key management implications: (i) each region may need a customized management strategy; (ii) visual assessment of papaya may be insufficient for PRSV, requiring diagnostic assays; and (iii) weed control within orchards may reduce the risk of virus spread to papaya. Network analysis advances understanding of host-pathogen interactions in the agroecological landscape.IMPORTANCE Virus-virus interactions in plants can modify host symptoms. As a result, disease management strategies may be unsuccessful if they are based solely on visual assessment and diagnostic assays for known individual viruses. Papaya ringspot virus is an important limiting factor for papaya production and likely has interactions with other viruses that are not yet known. Using high-throughput sequencing, we recovered known and novel RNA and DNA viruses from papaya orchards in Chiapas, Mexico, and categorized them by host and, in the case of papaya, symptom type: asymptomatic papaya, papaya with ringspot virus symptoms, papaya with nonringspot symptoms, weeds, and insects. Using network analysis, we demonstrated virus associations within and among host types and described the ecological community patterns. Recovery of viruses from weeds and asymptomatic papaya suggests the need for additional management attention. These analyses contribute to the understanding of the community structure of viruses in the agroecological landscape.

Keywords: bipartite networks; network analysis; phytobiome; plant viruses; viral ecology; viral metagenomics.

FIG 1

Two physiographic regions in the Federal State of Chiapas in Mexico. Maps showing the Central Depression (at 700 masl; light green), the Pacific Coastal Plain (at sea level; light blue), and the municipalities from which samples were collected: Acalá (a), Villa Corzo (b), La Concordia (c), Acapetahua (d), Mazatán (e), and Suchiate (f).

FIG 2

Distribution of DNA and RNA viruses in papaya, weeds, and insects. The frequency of log-transformed relative sequence abundances of viruses from the papaya orchard virome by host and location: papaya plants (A), weeds (B), and insects (C). The inner rings indicate the proportions of viruses from the samples collected from the Pacific Coastal Plain (light green) and the Central Depression (light blue). The middle ring (papaya only) indicates the proportion of viruses for each of three different types of papaya symptoms: visually asymptomatic (yellow), non-PRSV symptoms (orange), and PRSV symptoms (red). The outer rings indicate the proportions of putative RNA (light red) and DNA (blue) virus species.

FIG 3

The log-transformed relative abundances of DNA and RNA viruses by genus and by host and location in the papaya orchard virome of Chiapas. (A) The diversity and relative abundances of viruses by host and location for weeds, insects, and papaya plants by physiographic region for the Central Depression (CD) and Pacific Coastal Plain (PC). (B) The diversity and relative abundances of viruses in papaya divided by symptoms: visually asymptomatic (VA), non-PRSV symptoms (OS), and PRSV symptoms (PS). Cold colors represent DNA viruses, and warm colors indicate RNA viruses.

FIG 4

Papaya orchard virome, represented in a bipartite network with two types of nodes: viruses (circles) and host-location combinations (squares) for two physiographic regions, the Pacific Coastal Plain (light blue) and Central Depression (light green). The size of circular nodes varies with the sum of the relative abundance by virus species. Links indicate that a virus is present in a host and location, and their width and color (yellow to red) represent relative abundance by host and location. Alternate hosts are weeds (W), insects (I), papaya with PRSV symptoms (PS), papaya with non-PRSV symptoms (OS), and visually asymptomatic papaya (VA). Viruses are indicated by their acronyms (Table S3). Node color represents the status of the virus: previously reported in papaya (red), new report for Chiapas in nonpapaya hosts (orange), previously reported in Chiapas in nonpapaya hosts (yellow), and novel viruses not yet reported anywhere else on any hosts (purple).

FIG 5

A one-mode network extracted from the bipartite network of associations. Nodes represent host types from the Central Depression and the Pacific Coastal Plain. Node size is proportional to the betweenness centrality of the node. Links represent associations of viruses, and link thickness is proportional to the number of viruses shared within (blue) and among (red) regions.

FIG 6

Bipartite network displaying only the nodes that have two or more links. Viruses (circles) and host-location combinations (squares) from two physiographic regions, the Pacific Coastal Plain (light blue) and Central Depression (light green) are represented. Node sizes are proportional to the relative abundance of virus species. Links indicate associations between nodes, and their width and color (yellow to red) represent low and high relative abundances, respectively, by host and location. Hosts are weeds (W), insects (I), papaya with PRSV symptoms (PS), non-PRSV symptomatic papaya (OS), and visually asymptomatic papaya (VA). Viruses are indicated by their acronyms. Node color represents the status of the virus: reported in papaya (red), new report for Chiapas in nonpapaya hosts (orange), previously reported in Chiapas in nonpapaya hosts (yellow), and novel viruses not yet reported anywhere (purple).