Tomato spotted wilt virus (TSWV; genus Tospovirus, family Bunyaviridae) genetic diversity was evaluated by sequencing parts of the three RNA genome segments of 224 isolates, mostly from pepper and tomato crops in southern Europe. Eighty-three per cent of the isolates showed consistent clustering into three clades, corresponding to their geographical origin, Spain, France or the USA, for the three RNA segments. In contrast, the remaining 17 % of isolates did not belong to the same clade for the three RNA segments and were shown to be reassortants. Among them, eight different reassortment patterns were observed. Further phylogenetic analyses provided insights into the dynamic processes of the worldwide resurgence of TSWV that, since the 1980s, has followed the worldwide dispersal of the western flower thrips (Frankliniella occidentalis) tospovirus vector. For two clades composed essentially of Old World (OW) isolates, tree topology suggested a local re-emergence of indigenous TSWV populations following F. occidentalis introductions, while it could not be excluded that the ancestors of two other OW clades were introduced from North America contemporarily with F. occidentalis. Finally, estimation of the selection intensity that has affected the evolution of the NSs and nucleocapsid proteins encoded by RNA S of TSWV suggests that the former could be involved in the breakdown of resistance conferred by the Tsw gene in pepper.

No abstract

Analysis of an RNA-seq strand-specific library revealed a complete ge-nome of Hardenbergia mosaic virus (HarMV) from RNA extracted from a native wiste-ria (Hardenbergia comptoniana) plant from southwest Australia. We compared it with eight other complete HarMV genomes. It most resembled (85.8% nucleotide identity) the genome of HarMV isolate MD4-D. I n July 2016, an apical leaf sample showing obvious mosaic and deformation symptoms was collected from a plant of the indigenous southwest Australian species native wisteria (Hardenbergia comptoniana; family Fabaceae). The plant sampled also exhibited symptoms typical of phytoplasma infection (leaf chlorosis, proliferation of axillary shoots, and stunting). It was growing on a fence line bordering a playing field area in the Victoria Park suburb of Perth in southwest Australia. Hardenbergia mosaic virus (HarMV; genus Potyvirus, family Potyviridae) causes a conspicuous disease in native wisteria plants, is spread nonpersistently by aphids, and, like native wisteria itself, is indigenous to the region (1-5). It invades introduced lupin species (Lupinus spp.) at the interface of the local ancient ecosystem and recent agroecosystem (6, 7). Analysis of polyadenylated transcripts derived from RNAseq-stranded libraries (8-17) prepared from RNA extracted from the collected sample (designated VPK) detected one complete HarMV genome. RNA was extracted from the VPK sample using a ZR Plant RNA MiniPrepTM kit (Zymo Research) and treated with RNase-free DNase (Invitrogen). The extract was subjected to library preparation using a TruSeq-stranded Ribo-Zero plant kit (Illumina, catalog no. RS-122-2401) and was subsequently subjected to quality control (8-17). The library was sent to Macrogen, Inc. (South Korea), where sequencing was done using the HiSeq 2500 platform with a TruSeq SBS version 4 kit (Illumina) with 151 cycles of paired-end reads. Reads were then assembled and genomes annotated using CLC Genomics Workbench version 6.5 (CLC bio) and Geneious version 8.1.7 (Biomatters) (18, 19). The VPK sample yielded 13,609,056 reads and, after trimming, 12,972,958 remained. De novo assembly generated 25 contigs and 622,179 reads mapped to the contig of interest with a coverage of 9,107. The complete genome obtained was named VPK-1. It consisted of 9,621 nucleotides (nt) and coded for 10 proteins, which is similar to other potyviruses (20, 21). There were eight other complete HarMV genomes already in GenBank (3, 7). A BLAST-based search (22) revealed that sequence VPK-1 most resembled the sequence of HarMV isolate MD4-D (KJ152157) with an 85.8% nt identity. In addition, the analysis revealed a partial sequence (6,002 nt in length) named VPK-2,

Specialization of species in interaction networks infl uences network stability and ecosystem functioning. Spatial and temporal variation in resource availability may provide insight into how ecological factors, such as resource abundance, and evolutionary factors, such as phylogenetically conserved morphological traits, infl uence specialization within mutualistic networks. We used independent measures of hummingbird abundance and resources (nectar), information on hummingbird traits and plant -hummingbird interactions to examine how resource availability and species ' morphology infl uence the specialization of hummingbirds in three habitat types (forest, shrubs, cattle ranch) sampled over 10 sessions across two years in the southern Andes of Ecuador. Specialization of hummingbird species in the networks was measured by three indices: d ' (related to niche partitioning), generality (related to niche width) and PSI (related to pollination services). Specialization indices d ' , generality and PSI of hummingbird species were infl uenced by resource availability. All indices indicated that specialization of hummingbirds increased when the availability of resources decreased. Variation in d ' was also explained by an interaction between resource availability and bill length; hummingbirds with a long bill switched from being more specialized than other species when resource availability was low to being similarly specialized when availability was high. Overall, our results highlight the importance of ecological and evolutionary factors determining the specialization of species in interaction networks. We demonstrate in particular that ecological gradients in resource availability cause substantial changes in consumers ' foraging behavior contingent on their morphology. Changes in pollinator specialization along resource gradients can have impacts on ecosystem functions, such as pollination by animals.