The first survey using high-throughput sequencing of cereal and barley yellow dwarf viruses in Irish spring and winter barley crops

Abstract

Yellow dwarf viruses (YDVs) are the most economically important plant viruses impacting cereal production worldwide and include viruses from the genus Luteovirus (e.g., barely yellow dwarf virus (BYDV)-PAV, BYDV-PAS, BYDV-MAV, BYDV-kerII, BYDV-kerIII) and Polerovirus (e.g., cereal yellow dwarf virus (CYDV)-RPV, CYDV-RPS). Until now, much of our knowledge on YDVs infecting Irish barley crops (Hordeum vulgare L.) has come from serological assays; however, due to cross-reactivity it can be difficult to discriminate between viruses of different species. In this study, we have carried out a high-throughput sequencing survey of symptomatic crops, positive with serological assays, to identify YDVs infecting Irish spring and winter barley crops and establish reference genomes to support further development of molecular surveillance tools. In total, RNA was extracted from 45 symptomatic crop samples that were collected across Ireland over 2 yr and sequenced following rRNA depletion. Three samples of barley plants from BYDV-infected aphid colonies were also included. BYDV-MAV was identified in all field samples sequenced. This confirms previous evidence based on serological assays that BYDV-MAV is the dominant YDV in Irish barley crops. We have also identified BYDV-PAS in 29% of symptomatic field samples, the first report of this species in Ireland. In addition, BYDV-PAV was also found, and crop samples with mixed infections were common; although in mixed infections the greatest proportion of YDV reads originated from BYDV-MAV. Finally, CYDV-RPS, the more severe variant of CYDV-RPV belonging to the genus Polerovirus, was identified in a single sample. The complete genomes, assembled from this first sequence-based survey, will enable the development of molecular surveillance tools with greater virus specificity, to further support the Irish aphid and YDV monitoring network.

Main article text

Introduction

Yellow dwarf viruses (YDVs) are one of the most detrimental infecting agents impacting cereal production worldwide. Yellow dwarf viruses are single-stranded RNA viruses belonging to two genera, the Luteovirus (barley yellow dwarf virus [BYDV]-PAV, BYDV-MAV, BYDV-PAS, BYDV-kerII, BYDV-kerIII, BYDV-GAV, BYDV-SGV) and the Polerovirus (cereal yellow dwarf virus [CYDV]-RPV, CYDV-RPS, maize yellow dwarf virus [MYDV]-RMV, maize yellow mosaic virus [MaYMV], Barley virus G [BVG], wheat yellow dwarf virus [WYDV]-GPV) (Miller & Lozier, 2022). Viruses of these different species are transmitted by more than 25 species of aphids worldwide, reviewed in Van den Eynde et al. (2020). However, aphid species differ in virus transmission specificity and efficiency, with some species being extremely efficient at acquiring and transmitting the virus, whereas others are relatively inefficient (Peters et al., 2022). The aphids feed on grasses, and transmit these viruses to various grain crops; over 150 non-crop grass species in the family Poaceae are also known to be hosts (Miller & Rasochová, 1997). In Europe, two of the most economically important YDV vectors are the bird cherry-oat aphid (Rhopalosiphum padi L.) and the English grain aphid (Sitobion avenae F.).

Plants infected with these viruses often exhibit symptoms of stunted growth and discolouration due to chlorosis of the leaves. Susceptible species of Poaceae may become infected with the virus at any time throughout their life cycle, but early infection is especially most damaging when inoculated before growth stage 31 (GS 31), the beginning of stem elongation (reviewed in Walls et al., 2019). Yellow dwarf viruses vary strongly in the emergence of symptoms, aphid transmission efficiency and host preferences.

Symptom severity can also vary greatly among viral species (Choudhury et al., 2017). Barley yellow dwarf virus-PAV is currently the most studied species and the symptoms are considered more severe than an infection by BYDV-MAV. Yellow dwarf viruses cause substantial yield losses of up to 80% in crops, reducing grain quality (reviewed by Peters et al., 2022). Like symptom severity, yield loss varies greatly between virus species, the host cultivar, time of infection and other environmental/climate factors (Perry et al., 2000).

Given that the determination of YDV incidence and its effects on yield are based principally on visual symptoms, the accuracy of these estimates may be grossly underestimated (Perry et al., 2000). Recent Irish field trials in winter barley have shown that YDV symptoms do not necessarily correlate with an impact on yield (Walsh et al., 2022). In addition, YDVs frequently co-infect host plants in the field, which can lead to more severe symptoms, which further complicates the issue (Choudhury et al., 2017; Malmstrom et al., 2017; Minato et al., 2022).

Barley yellow dwarf virus-PAV is considered the most widespread and abundant YDV worldwide due to the prevalence of its primary vector R. padi (Aradottir & Crespo-Herrera, 2021) and the high efficiency in transmitting BYDV-PAV (Du et al., 2007). In Ireland, past research indicated that BYDV-MAV is the key species, with S. avenae being its primary vector (Kennedy & Connery, 2005). This work was based on enzyme-linked immunosorbent assays (ELISAs) that target only a few species and are prone to cross-reactivity between virus species. The advent of high-throughput sequencing (HTS) now offers new opportunities to characterise the genetic diversity of YDVs across multiple regions (see Sõmera et al., 2021 for example), which enables the ability to rapidly distinguish different virus species and their varieties and determine infection/co-infection status of both host plant and aphid vector. Such tools are necessary for the development of integrated pest management (IPM) strategies and local and national monitoring programmes, based on the rapid and effective identification of YDV species and their distribution and abundance in addition to providing information on their principal aphid vectors and the potential tolerance/susceptibility levels of the plant cultivars themselves.

In the current study, RNA sequencing was carried out on samples collected from barley crops suspected of carrying YDV infection across Ireland in 2021 and 2022. The results showed that BYDV-MAV was by far the most prominent species in Irish barley crops, but we also identified BYDV-PAV and BYDV-PAS, and one incidence of CYDV-RPS. The sequence data generated here will support the development of more targeted molecular surveillance tools to determine the YDV status of Irish aphids and barley crops as part of wider IPM programmes.

Material and methods

Sampling and serological testing

Leaf samples were collected from spring and winter barley crops throughout eight different counties in Ireland in 2021 and 2022. Within a transect of 20–50 m (depending on field size) three to four leaves displaying symptoms of C/BYDV infection were randomly sampled. Three non-overlapping transects were sampled per field, to give a total sample size between 9 and 12 leaves per field. Samples were stored at −80°C until further processing. Samples were placed in extraction bags (Bioreba, Reinach, Switzerland) with 2 mL of phosphate-buffered saline (PBS, pH 7) and ground with the aid of a tissue homogeniser attached to a motorised drill. Aliquots of the homogenised solution were used for double antibody sandwich enzyme-linked immunosorbent assay (DAS-ELISA) and for nucleic acid extraction. DAS-ELISA (Adams & Clark, 1977) was performed to ensure virus presence prior to sequencing. For this, commercially available kits for BYDV (BYDV-B complete kit 960; BYDV-F complete kit 960; BYDV-RPV complete kit 960; Bioreba, Reinach, Switzerland) were used following the protocol from Bioreba (DAS-ELISA test specifications, version: 4 – 18.10.2017). For this, 100 μL of 1:1,000 diluted antigen/coating buffer was pipetted into each well of a Nunc F96 Maxisorp microtiter plate (Thermo Fisher Scientific, Waltham, MA, USA). After a 3.5-h incubation at 37°C, the plates were washed three times with the addition of 150 μL of washing buffer into each well. Into each well, 100 μL of the plant extract and 100 μL of positive and negative controls (included in Bioreba kits) were added and the plates were incubated overnight at 4°C. After another washing step, 100 μL of 1:1,000 diluted antibody/conjugate buffer was added into each well and the plates were incubated at 37°C for 4 h, followed by a final washing step. Into each well, 100 μL of pNPP substrate buffer (1 mg/mL concentration) was pipetted, and following an incubation time of 30 min, 1 h and 2 h the plates were scanned using a microplate reader (Tecan Trading AG, Switzerland, Männedorf, Switzerland) at a wavelength of 405 nm. Where necessary, threshold levels were manually calculated following the Bioreba protocol. For this, a cut-off (“step”) value was determined and the mean of all values below this step value was calculated. Three times the standard deviation was added to this value, and finally, the value was multiplied by 1.1 as described in the DAS-ELISA data analysis from Bioreba (Version: 4 – 11.07.2014) and according to the following formula:

$$\text{cut-off}\hspace{0.17em}=\hspace{0.17em}(\text{mean}\hspace{0.17em}+\hspace{0.17em}\mathrm{3s})\hspace{0.17em}\times \hspace{0.17em}1.1,$$

where “mean” refers to the mean of values up to the “step” value, and s is the standard deviation of values up to the “step” value. All values above the cut-off were considered as BYDV-positive; however, in most cases, the sample value was already 5–10 times higher than the control, making the cut-off calculation obsolete.

Total RNA isolation and RNA sequencing

The Spectrum™ Plant Total RNA extraction kit (Merck KGaA, Darmstadt, Germany) was used to isolate total RNA from the homogenised solution of samples confirmed to have YDV infection via DAS-ELISA (Table 1), following a slightly modified protocol. For this, 100 μL of the homogenised solution was transferred to a 1.5-mL microcentrifuge tube and 400 μL of lysis buffer was added and vortexed. The lysate was transferred onto a (blue) filter column, centrifuged at 14,000 × g and an equal volume of 70% ethanol was added and vortexed. After this, the lysate was transferred onto a (red) binding column and spun down, to allow binding of RNA onto the column. After a washing step and DNAse digestion with 80 μL of DNAse I/Digest buffer for 15 min at room temperature, the column was washed twice with washing buffers. The total RNA was eluted in 50-μL elution buffer and the concentration was measured using NanoDrop^® (Thermo Fisher Scientific, Waltham, MA, USA). In addition to the samples collected in the field, three barley samples were collected from aphid colonies carrying the virus, which are maintained in the laboratory at Oak Park. The RNA was shipped on dry ice to GENEWIZ (Berlin, Germany) for library preparation with rRNA depletion and sequencing on an Illumina platform (San Diego, USA) (PE 150 base pairs [bp]).

Table 1:

Selection of yellow dwarf viruses in Irish barley crops using DAS-ELISA testing

Sample ID	Year sampled	County	Crop	Type B (PAV-like)	Type F (MAV-like)	CYDV-RPV
M3	2021	Meath	Winter barley	+	+	−
SAC 1	2021	NA	NA	+	+	−
Wi1	2021	Wicklow	Winter barley	+	+	−
Ti5	2021	Tipperary	Winter barley	+	+	−
Ki2	2021	Kildare	Winter barley	+	+	−
SBKST	2021	Kilkenny	Spring barley	+	+	−
La2a	2021	Laois	Winter barley	+	+	−
C6	2021	Cork	Winter barley	+	+	−
RPC 1	2021	NA	NA	+	+	−
SBKnST	2021	Carlow	Spring barley	+	+-	−
FC 1	2021	NA	NA	+	+	−
Ki5	2021	Kildare	Winter barley	+	+	−
C4	2021	Cork	Winter barley	+	+	−
C7	2021	Cork	Winter barley	+	+	−
Ki1	2021	Kildare	Winter barley	+	+	−
KK3	2021	Kilkenny	Winter barley	+	+	−
La1a1c	2021	Laois	Winter barley	+	+	−
La3a	2021	Laois	Winter barley	+	+	−
M5	2021	Meath	Winter barley	+	+	−
SBK	2021	Kilkenny	Spring barley	+	+	−
SBW	2021	Wexford	Spring barley	+	+	−
WB20	2020	Carlow	Winter barley	+	+	−
Ti3	2021	Tipperary	Winter barley	+	+	−
Wi6	2021	Wicklow	Winter barley	+	+	−
C1	2022	Cork	Winter barley	+	+	−
C10	2022	Cork	Winter barley	−	+	−
Ki4	2022	Kildare	Winter barley	+	+	−
Ki9	2022	Kildare	Winter barley	+	+	−
Wi6RD3	2022	Wicklow	Winter barley	+	+	−
Wi2J	2022	Wicklow	Winter barley	+	+	−
Ti4J	2022	Tipperary	Winter barley	+	+	−
Ti2J	2022	Tipperary	Winter barley	+	+	−
La3RD3	2022	Laois	Winter barley	+	+	−
La11J	2022	Laois	Winter barley	+	+	−
Kk3RD3	2022	Kilkenny	Winter barley	+	+	−
M2	2022	Meath	Winter barley	−	+	−
P3F2	2022	Wexford	Spring barley	+	+	+
P3F9	2022	Waterford	Spring barley	+	+	−
P4A8	2022	Laois	Spring barley	+	+	−
P1A7	2022	Carlow	Spring barley	+	+	−
P4E2	2022	Kilkenny	Spring barley	+	+	−
P4D1	2022	Cork	Spring barley	+	+	+
P3E1	2022	Louth	Spring barley	+	+	−
P3H3	2022	Tipperary	Spring barley	+	−	−
P3G3	2022	Wexford	Spring barley	+	+	−
P4B3	2022	Kilkenny	Spring barley	+	+	−
P4B9	2022	Meath	Spring barley	+	+	−
P1H1	2022	Kildare	Spring barley	−	+	−

¹Leaf samples were taken from colonies maintained in laboratories at Oak Park – these colonies were SAC (Sitobion avenae colony positive for BYDV-MAV); RPC (Rhopalosiphum padi colony positive for PAS) and FC (Sitobion avenae colony positive for PAS).

CYDV-RPV = cereal yellow dwarf virus; DAS-ELISA = double antibody sandwich enzyme-linked immunosorbent assay.

De novo assembly of RNA sequencing data

The sequencing data were evaluated using FastQC and the remaining adaptor contamination was removed using Cutadapt (Martin, 2011). These data have been deposited in the Sequence Read Archive (SRA) at the National Center for Biotechnology Information (NCBI) under BioProject PRJNA918968. The cleaned data for each sample were then aligned to the barley reference genome (IBSC-v2) (Mascher et al., 2017) using BWA-MEM (Li, 2013), and unmapped reads were captured. Subsamples of 1 million (M) read pairs were randomly sampled without replacement from the non-host reads of each sample in Geneious Prime (2022.0.2). In the case of one sample with a lower proportion (0.24) of non-host reads, a subsample of 5M read pairs was taken (sample RPC). These non-host reads were assembled de novo using SPAdes (Prjibelski et al., 2020) version 3.15.5 with RNA assembly pipeline (Bushmanova et al., 2019) with k-mer sizes of 33 and 49. CAP3 (Huang & Madan, 1999) was then used to further assemble contigs based on near-identical overlaps (at least 98%), with a minimum overlap length of 40 bp. Assembled contigs greater than 1,000 bp were then used in a BLAST search (Altschul et al., 1990) (megablast) against a nucleotide (nt) database and contigs with significant sequence similarity to YDVs were identified.

Mapping-based assemblies of RNA sequencing data

The subsampled non-host reads were then aligned against reference genomes of closely related accessions using the Geneious mapping algorithm in the map to reference function (Geneious Prime V2022.0.2). The sensitivity was set to medium-low, the minimum mapping quality was set to 30, and paired reads were only aligned where pairs mapped in the correct orientation and within the insert size ±0.5 times the insert size, and with five iterations of fine-tuning (involves mapping of reads to the consensus sequence generated from the previous round of mapping). The mean coverage and percentage coverage of reference were determined and consensus sequences were extracted when there was greater than 99% coverage of reference genome. Consensus sequences were extracted after trimming to the reference sequence and using a threshold of 60% for determining which base is called in the consensus.

Phylogenetic analysis

Consensus sequences representing each of the YDV species identified in the field samples were aligned with reference genomes used in the phylogenetic grouping of BYDV species (Sõmera et al., 2021) and other YDV species from the Luteovirus and Polerovirus genera used in phylogenetic analysis in (Miller & Lozier, 2022). Alignments were carried out with MUSCLE (V5.1) (Edgar, 2022) implemented in Geneious Prime (2023.0.1), and identities of nucleotide alignments were determined for the nucleotide sequences of the complete genome and individual open reading frames (ORFs). Phylogenies were determined from these sequence alignments using PhyML (version 3.3.20180214) (Guindon et al., 2010) with a general time reversible (GTR) model of nucleotide substitutions and 1,000 bootstrap replications.

Results

Confirmation that symptomatic plants were infected with yellow dwarf viruses

Of the 48 samples selected for HTS, 45 tested positive with DAS-ELISA for F-type antisera (MAV-like), and 43 tested positive with DAS-ELISA for B-type antiserum (PAV-like). Only two samples tested positive with DAS-ELISA for RPV antiserum (CYDV-RPV) (Table 1).

Identification of yellow dwarf viruses by HTS

Extracted RNA from 48 samples were sent for RNA sequencing following rRNA depletion and between 28.2M and 58.6M read pairs were retained following trimming and quality control (Table 2). The proportion of data from each sample that did not align with the barley host genome ranged between 24% and 84% (median of 64%). The fact that a large proportion of the sequence data did not originate from the host genome indicates that viral enrichment via rRNA depletion was successful. Sequences not originating from the host genome were subsampled (1M reads per sample) and assembled into contigs (using SPAdes and CAP3) and compared to sequences in NCBI using BLASTn for their identification. Given the purpose of this study, we only focused on contigs with significant sequence similarity to YDVs. We were able to identify contigs matching YDVs in all samples sequenced (Table 2), in line with the visual assessments of plants and results of serological testing. To confirm the presence of YDVs, reads from a sample were mapped to reference genomes of the YDVs with the highest bit-score to contigs assembled from data of that sample, and the proportion of reads mapped and percentage coverage were calculated (Table 2).

Table 2:

Summary of HTS results and identification of yellow dwarf viruses

Sample	Read pairs (M)	Non-host read pairs (M)	Percentage of non-host reads	NCBI BLASTn Yellow dwarf virus hits 1	Reads mapped to each reference genome (as percentage)	Mean coverage (percentage coverage) 2
M3	58.6	37.1	63	BYDV-PAV (KY593458.1)	37,365 (3.7%)	931.0 (99.6%)
				BYDV-GAV (MK012662.1)	157,141 (15.7%)	3,924.1 (100%)
SAC	31.7	19.2	61	BYDV-GAV (MK012662.1)	629,207 (62.9%)	16,672.3 (99.9%)
Wi1	37.1	19.7	53	BYDV-GAV (MK012662.1)	31,554 (3.2%)	144.0 (97.6%)
Ti5	46.4	24.7	53	BYDV-GAV (MK012662.1)	101,387 (10.1%)	2516.1 (100%)
Ki2	31.6	16.1	51	BYDV-PAV (EF521844)	10,312 (1.0%)	268.1 (67.7%)
				BYDV-GAV (MK012662.1)	66,106 (6.6%)	1,663.1 (98.0%)
				BYDV-PAS (MN128940)	1,036 (0.1%)	27.6 (78.7%)
SBKST	29.8	15.5	52	BYDV-GAV (MK012662.1)	435,882 (43.6%)	11,715.6 (99.5%)
				BYDV-PAV (MK962883.1)	87,438 (8.7%)	2,405.1 (95.2%)
La2a	29.3	22	75	BYDV-GAV (MK012662.1)	220,844 (22.1%)	5,507.1 (99.8%)
C6	28.7	14.5	51	BYDV-GAV (MK012662.1)	144,142 (14.4%)	3,804.1 (99.2%)
				BYDV-PAS (MK012654.1)	1,888 (1.9%)	49.7 (97.9%)
Ki5	33.3	23.2	70	BYDV-GAV (MK012662.1)	222,197 (22.2%)	5,542.8 (100%)
SBKnST	32.9	23.8	72	BYDV-PAV (MK962883.1)	22,832 (2.3%)	568.8 (81.8%)
				BYDV-GAV (MK012662.1)	222,642 (22.3%)	5,571.6 (97.9%)
FC	36.3	28.3	78	BYDV-PAS (MK012654.1)	447,490 (44.7%)	11,358.2 (100%)
RPC	28.2	6.8	24	BYDV-GAV (MK012662.1)	26,657 (2.7%)	701.6 (98.9%)
				BYDV-PAS (MK012652.1)	15,773 (1.6%)	415.2 (99.4%)
C4	34.5	13.8	40	BYDV-GAV (MK012662.1)	200,316 (20.0%)	4,865.4 (99.9%)
				BYDV-PAS (MK012654.1)	1,808 (1.8%)	42.2 (94.4%)
C7	33.1	16.5	50	BYDV-GAV (MK012662.1)	118,980 (11.9%)	2,863.9 (100%)
				BYDV-PAS (MK012654.1)	1,378 (1.4%)	32.9 (76.4%)
Ki1	44.8	21.4	48	BYDV-PAV (KY621333)	26,870 (2.7%)	664.7 (99.9%)
				BYDV-GAV (MK012662.1)	187,854 (18.8%)	4,644.6 (100%)
KK3	29.1	18	62	BYDV-GAV (MK012662.1)	88,380 (8.8%)	2,148.8 (100%)
La1a1c	32.5	19.3	59	BYDV-GAV (MK012662.1)	14,614 (1.5%)	363.0 (96.4%)
La3a	34.5	16.7	48	BYDV-PAV (AJ810418)	46,196 (4.6%)	141.7 (98.4%)
				BYDV-GAV (MK012662.1)	5,050 (0.5%)	125.7 (97.1%)
				BYDV-PAS (MK012654)	90,342 (9.0%)	2,217.7 (100%)
				CYDV-RPS (MK012664)	55,998 (5.6%)	1,399.7 (98.4%)
M5	37.9	22.5	59	BYDV-GAV (MK012662.1)	427,060 (42.7%)	10,455.7 (100%)
SBK	36	23	64	BYDV-GAV (MK012662.1)	483,390 (48.3%)	12,012.8 (100%)
SBW	33.1	18.6	56	BYDV-GAV (MK012662.1)	356,072 (35.6%)	8,847.9 (99.9%)
				BYDV-PAS (MK012652.1)	926 (0.09%)	22.7 (93.3%)
WB20	34.4	21.9	64	BYDV-GAV (MK012662.1)	436,058 (43.6%)	10,832.4 (99.9%)
Ti3	41.7	18.8	45	BYDV-GAV (MK012662.1)	60,690 (6.1%)	1,464.3 (98.2%)
Wi6	42.8	21.6	50	BYDV-GAV (MK012662.1)	49,042 (4.9%)	1,226.3 (99.4%)
C1	32.8	27.3	83	BYDV-PAV (MK012661)	24,802 (2.5%)	622.8 (91.3%)
				BYDV-GAV (MK012662.1)	145,552 (14.6%)	3,585.1 (99.7%)
				BYDV-PAS (MK012654)	28,582 (2.9%)	707.0 (100%)
C10	34.3	25.4	74	BYDV-PAV (EF521841)	16,803 (1.7%)	421.0 (82.1%)
				BYDV-GAV (MK012662.1)	157,831 (15.8%)	3,910.1 (99.7%)
Ki4	34.1	28.6	84	BYDV-PAV (AJ810418)	76,134 (7.6%)	1,840.4 (93.9%)
				BYDV-GAV (MK012662.1)	70,896 (7.1%)	1,759.6 (100%)
				BYDV-PAS (MK012654)	52,654 (5.3%)	1,315.4 (100%)
Ki9	35.2	29.6	84	BYDV-PAV (AJ810418)	18,178 (1.8%)	453.7 (100%)
				BYDV-GAV (MK012662.1)	157,342 (15.7%)	3,978.3 (99.9%)
Wi6	40.8	31.7	78	BYDV-PAV (AJ810418)	25,316 (2.5%)	636.1 (99.4%)
				BYDV-GAV (MK012662.1)	127,630 (12.8%)	3,142.4 (100%)
Wi2J	35.3	27	76	BYDV-PAV (AJ810418)	15,919 (1.6%)	389.2 (100%)
				BYDV-GAV (MK012662.1)	137,605 (13.8%)	3,407.9 (100%)
Ti4J	37.1	27.3	74	BYDV-PAV (AJ810418)	26,334 (2.6%)	674.7 (100%)
				BYDV-GAV (MK012662.1)	149,650 (15.0%)	3,712.4 (99.8%)
Ti2J	35.5	29	82	BYDV-PAV (AJ810418)	42,994 (4.3%)	1,024.9 (95.6%)
				BYDV-PAS (MK012654)	89,330 (8.9%)	2,205.0 (98.7%)
				BYDV-GAV (MK012662.1)	65,700 (6.6%)	1,653.1 (99.6%)
La3RD3	38.7	30.9	80	BYDV-GAV (MK012662.1)	229,153 (22.9%)	5,674.2 (100%)
La11J	31.5	24.3	77	BYDV-GAV (MK012662.1)	189,643 (19.0%)	4,668.2 (100%)
Kk3-RD3	37.3	30.1	81	BYDV-PAV (AJ810418)	49,536 (4.9%)	1,184.1 (97.8%)
				BYDV-GAV (MK012662.1)	121,220 (12.1%)	3,005.5 (99.9%)
				BYDV-PAS (MK012654)	63,956 (6.4%)	1,583.8 (99.6%)
M2	31.6	15.8	50	BYDV-GAV (MK012662.1)	74,947 (7.5%)	1,855.9 (100%)
P3F2	33.7	18.1	54	BYDV-GAV (MK012662.1)	111,529 (11.2%)	2,765.4 (99.9%)
P3F9	30	18.3	61	BYDV-PAV (AJ810418)	21,590 (2.2%)	535.1 (99.9%)
				BYDV-GAV (MK012662.1)	135,124 (13.5%)	3,430.8 (98.1%)
P4A8	32.8	24.1	73	BYDV-GAV (MK012662.1)	165,271 (16.5%)	4,122.2 (99.9%)
P1A7	34.7	27.4	79	BYDV-PAV (KY593458.1)	23,026 (2.3%)	598.4 (76.9%)
				BYDV-GAV (MK012662.1)	146,078 (14.6%)	3,728.9 (100%)
P4E2	29.7	20.7	70	BYDV-GAV (MK012662.1)	126,848 (12.7%)	2,863.9 (100%)
				BYDV-PAS (MK012654.1)	1,038 (0.1%)	32.9 (76.4%)
P4D1	37.8	31	82	BYDV-PAV (AJ810418)	16,041 (1.6%)	410.1 (99.6%)
				BYDV-GAV (MK012662.1)	110,331 (11.0%)	2,740.3 (100%)
P3E1	34.5	21.2	61	BYDV-GAV (MK012662.1)	128,212 (12.8%)	3,220.3 (99.9%)
				BYDV-PAS (MK012654.1)	870 (0.09%)	22.2 (94.2%)
P3H3	30.3	22.9	76	BYDV-PAV (AJ810418)	45,100 (4.5%)	1,113.9 (93.6%)
				BYDV-GAV (MK012662.1)	45,694 (4.6%)	1,145.9 (100%)
				BYDV-PAS (MK012654)	55,682 (5.6%)	1,462 (100%)
P3G3	33.8	18.8	56	BYDV-PAV (AJ810418)	19,856 (2.0%)	510.3 (92.4%)
				BYDV-GAV (MK012662.1)	112,472 (11.2%)	2,820.4 (97.3%)
P4B3	33.1	26.6	80	BYDV-GAV (MK012662.1)	195,187 (19.5%)	4,820.7 (100%)
P4B9	35.4	24.2	68	BYDV-PAV (MK962883)	18,192 (1.8%)	462.1 (86.5%)
				BYDV-GAV (MK012662.1)	132,626 (13.3%)	3,290.0 (99.2%)
P1H1	37.9	24.6	65	BYDV-GAV (MK012662.1)	124,593 (12.5%)	3,150.7 (99.9%)

The total read pairs sequenced for each sample and proportion of reads not originating from barley host are shown. A random sample of 1M read pairs from each library was then used for de novo assembly as described, and yellow dwarf virus genomes with significant similarity (highest bit-score) to assembled contigs are shown. The percentage of reads mapped, mean read depth and reference coverage after mapping reads back to these reference genomes are shown.

¹Yellow dwarf virus genomes with significant sequence similarity (highest bit-score) to contigs assembled from non-host reads.

²Mean read depth across reference genome, and the percentage of reference genome covered with sample reads.

BYDV = barley yellow dwarf virus; HTS = high-throughput sequencing; NCBI = National Center for Biotechnology Information.

The most common YDV species identified was BYDV-GAV/MAV, which was observed in all field samples and all but one of the 48 samples. While the sequences had the highest bit-score to BYDV-GAV, the percent identity was highest with the single genome in GenBank characterised as BYDV-MAV. However, given the genome representing BYDV-MAV is truncated at the 3-prime end, the BYDV-GAV genome was used as the reference for mapping-based assemblies. The one sample without the BYDV-MAV-like sequence originated from plants in an aphid colony maintained at Oak Park, Carlow, Ireland, and with data from this sample we were able to extract a consensus sequence representing the complete genome of a BYDV-PAS-like sequence. BYDV-PAV-like and BYDV-PAS-like sequences were also identified in many samples as co-infection with the BYDV-MAV-like sequences, and a CYDV-RPS-like sequence was identified in one sample (Table 2). There were some disagreements between serological and HTS results; for example, DAS-ELISA with RPV antisera identified two positive samples and these were then prioritised for HTS. However, in both cases CYDV-RPV or CYDV-RPS genomes were not assembled from these samples. The DAS-ELISA also shows samples as positive for both F- and B-type antisera, despite not always finding BYDV-PAV in the HTS. It is also known that the F- and B-type reagents can cross-react. We extracted consensus sequences in cases where coverage of the reference genome was greater than 99% and the final consensus sequences are available in GenBank under accession numbers OQ686645, OR771726-OR771729 and OQ686648-OQ686695.

Nucleotide identities of yellow dwarf viruses

The BYDV-MAV-like consensus sequences extracted in this study were highly similar to one another with pairwise nucleotide identities of greater than 95%. The BYDV-PAS-like sequences extracted were also highly similar, with pairwise nucleotide identities of greater than 98%. In the case of the BYDV-PAV-like sequences extracted, the pairwise nucleotide identities were lower, with one isolate having less than a 90% pairwise identity with other isolates. We also looked at pairwise nucleotide identities across different genomic regions: open reading frame 1-2 (ORF1-2) (encoding a replication-associated protein P1 from ORF1, and ORF2 fused to ORF1 encoding RdRp), ORF3a (encodes P3a protein), ORF3-5 (encodes coat protein from ORF3; aphid transmission protein from ORF3 fused to ORF5 in a read-through domain; movement protein from ORF4), and ORF6 (encodes P6 protein), with reference to luteoviruses and using sequences representative of those identified in this study (Figure 1). The BYDV genome encodes proteins from seven ORFs and the regions selected above cover these seven ORFs. This includes ORF6, which is found in the luteoviruses and not the poleroviruses, and was the ORF with the greatest sequence dissimilarity (Chalhoub et al., 1994). In two regions (P6 protein and aphid transmission protein) with the greatest sequence dissimilarity, the sequence identified in this study as BYDV-MAV-like shared the greatest sequence identity with the BYDV-MAV reference sequence (Figure 1B and D). In the case of the aphid transmission protein region, the sequence was also highly similar to the BYDV-GAV reference (Figure 1B). BYDV-GAV is considered a subspecies of BYDV-MAV, although not listed by the international committee on taxonomy of viruses (ICTV) as an official species (Miller & Lozier, 2022).

Figure 1.

Nucleotide identity matrix for yellow dwarf viruses from the genus Luteovirus and regions encoding the (A) P1-RdRp fusion protein, (B) aphid transmission protein, (C) P3a protein and (D) P6 protein. The final consensus sequences extracted in this study are shown in blue (representative sequences for each Luteovirus species identified are included).

Phylogenetic grouping of yellow dwarf viruses

Representative sequences for each YDV species identified in this study were grouped and aligned alongside representative genome sequences of YDVs used in the phylogenetic grouping in other studies (Sõmera et al., 2021; Miller & Lozier, 2022) (Figure 2). In addition to complete genomes, this focused on the region encoding the P1-RdRp fusion protein (ORF1-2) and the region encoding the read-through protein (ORF3-5) involved in movement and aphid transmission. These regions were selected to build on previous phylogenetic analysis (Sõmera et al., 2021). In the case of BYDV-PAV, two representative sequences were included to capture the BYDV-PAV M3 isolate that showed sequence dissimilarity with other BYDV-PAV isolates. These largely agreed with phylogenetic analysis from other studies (Sõmera et al., 2021; Miller & Lozier, 2022). As previously observed, the results for the phylogenetic analysis with the ORF1-2 and ORF3-5 regions showed different groupings (Figure 2). In the case of ORF1-2, BYDV-PAV was on a branch with BYDV-MAV and BYDV-GAV, and in the case of ORF3-5, BYDV-PAV was on a branch with BYDV-PAS. This is not unexpected given the important role that recombination has played in the evolution of BYDV (Boulila, 2011; Wu et al., 2011).

Figure 2.

Maximum likelihood (ML) phylogenetic tree for yellow dwarf viruses (YDVs). The final consensus sequences extracted in this study are shown in blue and reference YDV sequences are in black. The trees are based on MUSCLE multiple sequences alignments of (A) complete genome, (B) open reading frame 1-2 (ORF1-2) and (C) ORF3-5. The ML phylogenetic trees were constructed with PhylML v3.3.20180214 with a general time reversible (GTR) nucleotide substitution model, and bootstrap values after 1,000 replicates are shown. Branch lengths indicate the number of substitutions per site.

Phylogenetic analysis also confirmed that the single polerovirus identified in this study grouped with CYDV-RPS, as expected from sequence homology (Figure 2). Interestingly, this occurred in a sample where we also detected BYDV-MAV, BYDV-PAV and BYDV-PAS and was the only field sample where BYDV-MAV was not the most abundant YDV (based on sequence reads). In addition, a phylogenetic analysis was carried out using all sequences generated in this study and YDV genomes downloaded from NCBI. In the case of the single CYDV sequence assembled, this confirmed that the sequence grouped with other available CYDV-RPS genomes (Supplementary Figure 1). In the case of the BYDV genomes, all the Irish BYDV-MAV sequences assembled in this study grouped together in a single clade and were separated from BYDV-GAV sequences and the single BYDV-MAV genome (Supplementary Figure 2).

Discussion

Yellow dwarf viruses are one of the most economically important groups of viruses impacting cereal production worldwide. Yellow dwarf viruses include the BYDVs that are transmitted by aphids and can cause yield losses of up to 80%. With the loss in the availability of insecticides, emerging insecticide resistance, ambitions to half pesticide usage in the EU, and a changing climate, there is a greater focus on developing improved IPM tools. Any IPM approach for BYDV must be built on a clear understanding of the tripartite interaction between virus, vector and plant and the resulting impact on yield.

A first step in developing robust IPM approaches is understanding what species of YDV are transmitted to and infecting Irish barley crops. Until now in Ireland, our knowledge on what species are infecting crops has been based on serological assays, which can have limitations with specificity (Bouallegue et al., 2014). This was observed in this study where the majority of samples testing positive for BYDV-MAV also tested positive for BYDV-PAV, and it is known that the antiserum for type F (BYDV-MAV) can cross-react with type B (BYDV-PAV) (Plumb, 1974). In order to move to more sensitive and specific molecular assays that can be incorporated into national surveillance programmes, we need to better understand the viruses present and generate reference genomes for Irish isolates, upon which we can develop new molecular assays.

This study represents the first HTS survey targeting YDVs in Irish barley crops. As expected based on sampling symptomatic plants and serological confirmation, YDVs were identified in all 48 samples (45 field-collected samples and 3 leaf samples from infected aphid colonies). A BYDV-MAV-like sequence was identified in all field samples. This is in agreement with previous studies carried out in Ireland using serological approaches on samples of spring barley from both Carlow and Cork (Kennedy & Connery, 2005). These were tested by ELISA using F- and B-type antisera which detect MAV- and PAV-type virus (Plumb, 1974), respectively. Leaf samples (8,047) were tested between 1990–1993 and 1996–2001, with 52.8% testing positive for the F-type (MAV) virus. Of these, 1.5% also tested positive with a mixture of antisera for the PAV- and RPV-type virus (Kennedy & Connery, 2005). Even in cases where co-infection was found, in the majority of these cases sequence reads associated with the BYDV-MAV-like sequence were most abundant. Using nucleotide sequence approaches, the BYDV-MAV-like sequences assembled in this study were identified as BYDV-MAV, as opposed to the BYDV-GAV subspecies. In particular, in the region (ORF6) with the greatest sequence dissimilarity among BYDV species, the sequences shared the greatest nucleotide identity with the reference BYDV-MAV sequence. The ORF6 region, which encodes the P6 protein and is specific to YDVs from the Luteovirus genus, is interesting given that it is the genomic region of highest sequence variability (Chalhoub et al., 1994). While initially thought not to encode a functional protein, evidence based on alignments showed variation was largely restricted to the third base of ORF6 codons, supporting translation into a functional protein (Smirnova et al., 2015). A phylogenetic analysis showed that all BYDV-MAV sequences from this study grouped together and were separated from a grouping of BYDV-GAV sequences, and a clade consisting of the single BYDV-MAV genome sequence represented in GenBank (from a US isolate [Ueng et al., 1992]). The grouping of BYDV-GAV consisted of genome sequences of isolates from China and Estonia.

Barley yellow dwarf virus-MAV is efficiently transmitted by S. avenae and Metopolophium dirhodum but not by R. padi (Van den Eynde et al., 2020). There is limited research on the impact of BYDV-MAV on yield in artificially inoculated field trials, with most research available on the impacts of BYDV-PAV (Perry et al., 2000; Nancarrow et al., 2021). In general, BYDV has a global distribution, with BYDV-PAV being the most prevalent and abundant YDV species worldwide, likely due to its vectoring efficiency and the wide distribution of its primary vector R. padi (Aradottir & Crespo-Herrera, 2021). However, BYDV-MAV has been reported in studies carried out in many countries including Poland, Ukraine (Snihur et al., 2018), Tunisia (Hamdi & Najar, 2023), Australia (Nancarrow et al., 2018), Latvia and Sweden (Bisnieks et al., 2004). It was found that the species incidence in Australia appears to change between geographical regions, where BYDV-PAV is the most common in the Victorian belt, whereas BYDV-MAV is the most prevalent in New South Wales (Nancarrow et al., 2018). This variability in the abundance of particular species of BYDV in different areas may be influenced by a number of factors, including the presence and activity of alternative vectors, the types of alternative hosts for both the virus species and aphid vectors and the susceptibility of these alternative hosts to the different BYDV species and aphid colonisation (Nancarrow et al., 2018). Similarly, the abundance of different species of BYDV in the UK was found to vary in different studies, hosts and locations (Henry et al., 1993; Foster et al., 2004).

Knowledge of cereal aphid species distribution is important to assess the spread of YDVs. Previous work conducted between 1988 and 1995 in Ireland identified S. avenae as the more abundant species of aphid encountered, while the MAV-strain, which is transmitted most efficiently by S. avenae, was the only serotype of the virus found in these trials using ELISA (Kennedy & Connery, 2005). The occurrence of BYDV in spring barley was investigated in Ireland between 1990 and 2001. The most common aphid encountered was S. avenae, and BYDV-MAV was the most common YDV species found; meanwhile, only a small number of samples tested positive for PAV/RPV (Kennedy & Connery, 2005). In an Irish study of minimum-till and conventional-till winter barley and wheat, the majority tested positive for BYDV-MAV and only a single sample was positive for a mixture of PAV and RPV antisera (Kennedy et al., 2010). In this study the predominance of BYDV-MAV in Ireland has been confirmed using HTS, which may reflect the abundance of S. avenae. Previous research has shown that S. avenae is a “very efficient” transmitter of BYDV-MAV (75–100% transmission rate), an “efficient” transmitter of BYDV-PAV (25–75% transmission rate) and that it can transmit BYDV-PAS, but the transmission rates are unknown (Van den Eynde et al., 2020).

It is also possible that the predominance of BYDV-MAV can be explained by landscape composition. It is known that aphid incidence is higher where there is uncropped land such as grassland, moorland and wasteland, and there is evidence that virus levels are lower where arable land dominates (Foster et al., 2004). A French study found that R. padi colonisation was reduced when grassland cover in the landscape was high and increased when there was more maize (Zea mays) in the surrounding 1-km radius (Gilabert et al., 2017). Ireland has very low levels of maize, which may contribute to the lower levels of R. padi and BYDV-PAV. Work done in Kansas (USA) looked at the effect of environmental drivers and land cover on spring cereal aphid abundance. The abundance of BYDV-positive S. avenae was positively correlated with distance to forest or shrubland. S. avenae and R. padi were more prevalent at eastern sites where ground cover is more grassland than cropland, suggesting that grassland may provide over-summering sites for vectors and pose a risk as BYDV reservoirs (Enders et al., 2018). To understand the predominance of MAV in Ireland, the surrounding landscape and potential reservoirs need to be considered, particularly the levels of permanent grassland in the landscape (ca. 80% of the agricultural land area). Expanding this work from barley to other cereals is required to confirm the predominance of BYDV-MAV and S. avenae in Ireland.

In addition to BYDV-MAV, BYDV-PAV was also identified in a number of samples as co-infection with BYDV-MAV. BYDV-PAV is most efficiently transmitted by R. padi (Power & Gray, 1995; Du et al., 2007; Parizoto et al., 2013). The negative impact of BYDV-PAV on yield has been shown to be up to 84% in wheat (Triticum aestivum) and 64% in barley (Nancarrow et al., 2021). In the case of the BYDV-PAV sequences, one isolate had less than 90% nucleotide identity with other BYDV-PAV sequences assembled in this study. This isolate (M3) shared the greatest sequence similarity (90.6% nucleotide identity) with a YDV (BYDV-PAV isolate KS) identified in the United States (Laney et al., 2018). This was supported by the phylogenetic analysis where the BYDV-PAV M3 isolate is present in a clade with the BYDV-PAV KS isolate and an isolate from Pakistan. The remaining seven Irish BYDV-PAV isolates grouped together. The BYDV-PAV isolates used in the phylogenetic analysis (Supplementary Figure 2) were grouped into two branches, one branch had BYDV-PAV isolates predominantly from the United States, Pakistan and Europe, which included all the BYDV-PAV isolates from this study. The other branch had the BYDV-PAV isolates predominantly from China, together with all the BYDV-PAS and BYDV-OYV isolates. Barley yellow dwarf virus-OYV was identified as a potential novel species from Swedish meadow fescue samples and cereals from Estonia, and shares the closest relationship with BYDV-PAV isolates from China (Sõmera et al., 2021). The division of BYDV-PAV into two branches, one with BYDV-PAV only and one with BYDV-PAV and BYDV-PAS, was recently observed in a phylogenetic analysis of a large global collection of nucleotide sequences encoding BYDV coat and movement proteins (Wei et al., 2023).

The present study also found BYDV-PAS, which has not been previously reported in Ireland. BYDV-PAS was identified in 29% of the symptomatic field samples and is known to be transmitted by S. avenae, M. dirhodum and R. padi (Jaroá et al., 2013), although there is no information on the relative transmission efficiency of this virus by different aphid species. Plants infected with BYDV-PAS (originally identified as PAV-129) showed delayed development of yellowing but more severe symptoms in oats (Avena sativa) than BYDV-PAV; this included greater stunting and abnormal leaf development (Chay et al., 1996). This was also observed in a study of barley and oat cultivars in Morocco where isolates more closely related to PAV-129 (now BYDV-PAS) had more severe symptoms (Bencharki et al., 1999). However, studies in wheat found the opposite, with BYDV-PAV leading to greater stunting and more severe symptoms (Jaroá et al., 2013). Understanding how efficiently BYDV-PAS is transmitted by different aphid species and its impact on crop yield is important for a better understanding of the relative importance of this species.

Cereal yellow dwarf virus-RPS was the only polerovirus identified in our study, a more severe variant of CYDV-RPV and known to be transmitted by R. padi (Almasi et al., 2015), although its ability to be transmitted by other aphids is unclear. CYDV-RPS has previously been found with HTS in Estonia (Sõmera et al., 2021), the UK (Pallett et al., 2010), the United States (Malmstrom et al., 2017) and the Czech Republic (Singh et al., 2020). Initially, the YDVs from the genera Luteovirus and Polerovirus were assigned to the family Luteoviridae, but this was changed recently and both genera were assigned to separate families, the Tombusviridae and Solemoviridae (Walker et al., 2021). While YDVs from both genera encode highly related genes for movement and transmission, they do encode very different RNA replication mechanisms (Miller et al., 2002), which is the region used by ICTV for classifying RNA viruses. The significance of the YDVs from the genus Polerovirus in an Irish context is unclear and warrants further investigation.

In order to develop effective monitoring programmes, we require molecular tools for rapid and high-throughput detection and identification of YDVs in (i) aphids captured when inspecting fields, in-field traps and suction towers and (ii) from crop samples and potential viral reservoirs (grasslands, margins, volunteers). An ability to quantify viral load will also enhance efforts to better understand disease development. The HTS survey carried out by this study will support the development of these molecular tools using reference genomes for species typically found in Ireland. Armed with these new tools and an expansive monitoring programme, we begin to gain greater insights into the drivers of BYDV epidemiology and how YDV species, aphid vector and biotype, the timing of infection and host genetics interact to ultimately impact yield.

References

1. Adams AN, Clark MF. 1977. Characteristics of the microplate method of enzyme-linked immunosorbent assay for the detection of plant viruses. Journal of General Virology. Vol. 34:475–483. [Cross Ref]

2. Almasi R, Miller WA, Ziegler-Graff V. 2015. Mild and severe cereal yellow dwarf viruses differ in silencing suppressor efficiency of the P0 protein. Virus Research. Vol. 208:199–206. [Cross Ref]

3. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. Journal of Molecular Biology. Vol. 215:403–410. [Cross Ref]

4. Aradottir GI, Crespo-Herrera L. 2021. Host plant resistance in wheat to barley yellow dwarf viruses and their aphid vectors: a review. Current Opinion in Insect Science. Vol. 45:59–68. [Cross Ref]

5. Bencharki B, Mutterer J, El Yamani M, Ziegler-Graff V, Zaoui D, Jonard G. 1999. Severity of infection of Moroccan barley yellow dwarf virus PAV isolates correlates with variability in their coat protein sequences. Annals of Applied Biology. Vol. 134:89–99. [Cross Ref]

6. Bisnieks M, Kvarnheden A, Sigvald R, Valkonen JPT. 2004. Molecular diversity of the coat protein-encoding region of barley yellow dwarf virus-PAV and barley yellow dwarf virus-MAV from Latvia and Sweden. Archives of Virology. Vol. 149:843–853. [Cross Ref]

7. Bouallegue M, Mezghani-Khemakhem M, Makni H, Makni M. 2014. First report of Sugarcane yellow leaf virus infecting barley in Tunisia. Plant Disease. Vol. 98:1016. [Cross Ref]

8. Boulila M. 2011. Selective constraints, molecular recombination structure and phylogenetic reconstruction of isometric plant RNA viruses of the families Luteoviridae and Tymoviridae. Biochimie. Vol. 93:242–253. [Cross Ref]

9. Bushmanova E, Antipov D, Lapidus A, Prjibelski AD. 2019. rnaSPAdes: a de novo transcriptome assembler and its application to RNA-Seq data. GigaScience. Vol. 8:giz100. [Cross Ref]

10. Chalhoub BA, Kelly L, Robaglia C, Lapierre HD. 1994. Sequence variability in the genome-3′-terminal region of BYDV for 10 geographically distinct PAV-like isolates of barley yellow dwarf virus: analysis of the ORF6 variation. Archives of Virology. Vol. 139:403–416. [Cross Ref]

11. Chay CA, Smith DM, Vaughan R, Gray SM. 1996. Diversity among isolates within the PAV serotype of barley yellow dwarf virus. Phytopathology. Vol. 86:370–377

12. Choudhury S, Hu H, Meinke H, Shabala S, Westmore G, Larkin P, Zhou M. 2017. Barley yellow dwarf viruses: infection mechanisms and breeding strategies. Euphytica. Vol. 213:168[Cross Ref]

13. Du ZQ, Li L, Liu L, Wang XF, Zhou G. 2007. Evaluation of aphid transmission abilities and vector transmission phenotypes of barley yellow dwarf viruses in China. Journal of Plant Pathology. Vol. 89:251–259

14. Edgar RC. 2022. Muscle5: High-accuracy alignment ensembles enable unbiased assessments of sequence homology and phylogeny. Nature Communications. Vol. 13:6968[Cross Ref]

15. Enders LS, Hefley TJ, Girvin JJ, Whitworth RJ, Smith CM. 2018. Spatiotemporal distribution and environmental drivers of barley yellow dwarf virus and vector abundance in Kansas. Phytopathology. Vol. 108:1196–1205. [Cross Ref]

16. Foster GN, Blake S, Tones SJ, Barker I, Harrington R. 2004. Occurrence of barley yellow dwarf virus in autumn-sown cereal crops in the United Kingdom in relation to field characteristics. Pest Management Science. Vol. 60:113–125. [Cross Ref]

17. Gilabert A, Gauffre B, Parisey N, Le Gallic JF, Lhomme P, Bretagnolle V, Dedryver CA, Baudry J, Plantegenest M. 2017. Influence of the surrounding landscape on the colonization rate of cereal aphids and phytovirus transmission in autumn. Journal of Pest Science. Vol. 90:447–457. [Cross Ref]

18. Guindon S, Dufayard JF, Lefort V, Anisimova M, Hordijk W, Gascuel O. 2010. New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0. Systematic Biology. Vol. 59:307–321. [Cross Ref]

19. Hamdi I, Najar A. 2023. First report of BYDV-PAV and BYDV-MAV on triticale in Tunisia. Journal of Plant Pathology. Vol. 105:1219[Cross Ref]

20. Henry M, George S, Arnold GM, Dedryver CA, Kendall DA, Robert Y, Smith BD. 1993. Occurrence of barley yellow dwarf virus (BYDV) isolates in different farmland habitats in western France and South-west England. Annals of Applied Biology. Vol. 123:315–329. [Cross Ref]

21. Huang X, Madan A. 1999. CAP3: a DNA sequence assembly program. Genome Research. Vol. 9:868–877. [Cross Ref]

22. Jarošová J, Chrpová J, Šíp V, Kundu JK. 2013. A comparative study of the barley yellow dwarf virus species PAV and PAS: distribution, accumulation and host resistance: comparative study of BYDV-PAV and -PAS. Plant Pathology. Vol. 62:436–443. [Cross Ref]

23. Kennedy TF, Connery J. 2005. Grain yield reductions in spring barley due to barley yellow dwarf virus and aphid feeding. Irish Journal of Agricultural and Food Research. Vol. 44:111–128

24. Kennedy TF, McDonald JG, Connery J, Purvis G. 2010. A comparison of the occurrence of aphids and barley yellow dwarf virus in minimum-till and conventional-till autumn-sown cereals. The Journal of Agricultural Science. Vol. 148:407–419. [Cross Ref]

25. Laney AG, Acosta-Leal R, Rotenberg D. 2018. Optimized yellow dwarf virus multiplex PCR assay reveals a common occurrence of Barley yellow dwarf virus-PAS in Kansas Winter wheat. Plant Health Progress. Vol. 19:37–43. [Cross Ref]

26. Li H. 2013. Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. arXiv preprint arXiv:1303.3997. [Cross Ref]

27. Malmstrom CM, Bigelow P, Trębicki P, Busch AK, Friel C, Cole E, Abdel-Azim H, Phillippo C, Alexander HM. 2017. Crop-associated virus reduces the rooting depth of non-crop perennial native grass more than non-crop-associated virus with known viral suppressor of RNA silencing (VSR). Virus Research. Vol. 241:172–184. [Cross Ref]

28. Martin M. 2011. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet. Journal. Vol. 17:10[Cross Ref]

29. Mascher M, Gundlach H, Himmelbach A, Beier S, Twardziok SO, Wicker T, Radchuk V, Dockter C, Hedley PE, Russell J, Bayer M, Ramsay L, Liu H, Haberer G, Zhang XQ, Zhang Q, Barrero RA, Li L, Taudien S, Groth M, Felder M, Hastie A, Šimková H, Staňková H, Vrána J, Chan S, Muñoz-Amatriaín M, Ounit R, Wanamaker S, Bolser D, Colmsee C, Schmutzer T, Aliyeva-Schnorr L, Grasso S, Tanskanen J, Chailyan A, Sampath D, Heavens D, Clissold L, Cao S, Chapman B, Dai F, Han Y, Li H, Li X, Lin C, McCooke JK, Tan C, Wang P, Wang S, Yin S, Zhou G, Poland JA, Bellgard MI, Borisjuk L, Houben A, Doležel J, Ayling S, Lonardi S, Kersey P, Langridge P, Muehlbauer GA, Clark MD, Caccamo M, Schulman AH, Mayer KFX, Platzer M, Close TJ, Scholz U, Hansson M, Zhang G, Braumann I, Spannagl M, Li C, Waugh R, Stein N. 2017. A chromosome conformation capture ordered sequence of the barley genome. Nature. Vol. 544:427–433. [Cross Ref]

30. Miller WA, Rasochová L. 1997. Barley yellow dwarf viruses. Annual Review of Phytopathology. Vol. 35:167–190. [Cross Ref]

31. Miller WA, Lozier Z. 2022. Yellow dwarf viruses of cereals: taxonomy and molecular mechanisms. Annual Review of Phytopathology. Vol. 60:121–141. [Cross Ref]

32. Miller WA, Liu S, Beckett R. 2002. Barley yellow dwarf virus: Luteoviridae or Tombusviridae? Molecular Plant Pathology. Vol. 3:177–183. [Cross Ref]

33. Minato N, Hatori S, Okawa A, Nakagawa K, Hironaka M. 2022. Manipulation of insect vectors’ host selection behavior by barley yellow dwarf virus is dependent on the host plant species and viral co-infection. Life. Vol. 12:644. [Cross Ref]

34. Nancarrow N, Aftab M, Freeman A, Rodoni B, Hollaway G, Trębicki P. 2018. Prevalence and incidence of yellow dwarf viruses across a climatic gradient: a four-year field study in Southeastern Australia. Plant Disease. Vol. 102:2465–2472. [Cross Ref]

35. Nancarrow N, Aftab M, Hollaway G, Rodoni B, Trębicki P. 2021. Yield losses caused by barley yellow dwarf virus-PAV infection in wheat and barley: a three-year field study in South-Eastern Australia. Microorganisms. Vol. 9:645. [Cross Ref]

36. Pallett DW, Ho T, Cooper I, Wang H. 2010. Detection of cereal yellow dwarf virus using small interfering RNAs and enhanced infection rate with cocksfoot streak virus in wild cocksfoot grass (Dactylis Glomerata). Journal of Virological Methods. Vol. 168:223–227. [Cross Ref]

37. Parizoto G, Rebonatto A, Schons J, Lau D. 2013. Barley yellow dwarf virus-PAV in Brazil: seasonal fluctuation and biological characteristics. Tropical Plant Pathology. Vol. 38:11–19. [Cross Ref]

38. Perry KL, Kolb FL, Sammons B, Lawson C, Cisar G, Ohm H. 2000. Yield effects of barley yellow dwarf virus in soft red winter wheat. Phytopathology. Vol. 90:1043–1048. [Cross Ref]

39. Peters JS, Aguirre BA, DiPaola A, Power AG. 2022. Ecology of yellow dwarf viruses in crops and grasslands: interactions in the context of climate change. Annual Review of Phytopathology. Vol. 60:283–305. [Cross Ref]

40. Plumb RT. 1974. Properties and isolates of barley yellow dwarf virus. Annals of Applied Biology. Vol. 77:87–91. [Cross Ref]

41. Power AG, Gray SM. 1995. Aphid transmission of barley yellow dwarf viruses: interactions between viruses, vectors, and host plantsBarley Yellow Dwarf: 40 Years of Progress. APS Press. St. Paul, MN: p. 259–291

42. Prjibelski A, Antipov D, Meleshko D, Lapidus A, Korobeynikov A. 2020. Using SPAdes de novo assembler. Current Protocols in Bioinformatics. Vol. 70:e102. [Cross Ref]

43. Singh K, Jarošová J, Fousek J, Chen H, Kundu JK. 2020. Virome identification in wheat in the Czech Republic using small RNA deep sequencing. Journal of Integrative Agriculture. Vol. 19:1825–1833. [Cross Ref]

44. Smirnova E, Firth AE, Miller WA, Scheidecker D, Brault V, Reinbold C, Rakotondrafara AM, Chung BY, Ziegler-Graff V. 2015. Discovery of a small non-AUG-initiated ORF in poleroviruses and luteoviruses that is required for long-distance movement. PLoS Pathogens. Vol. 11:e1004868. [Cross Ref]

45. Snihur H, Petrenko S, Kot T, Shevchenko O, Polischuk V. 2018. Widespread viral diseases endangering cereal crops in Ukraine. Mikrobiolohichnyi Zhurnal. Vol. 80:103–116. [Cross Ref]

46. Sõmera M, Massart S, Tamisier L, Tamisier L, Sooväli P, Sathees K, Kvarnheden A. 2021. A survey using high-throughput sequencing suggests that the diversity of cereal and barley yellow dwarf viruses is underestimated. Frontiers in Microbiology. Vol. 12:673218. [Cross Ref]

47. Ueng PP, Vincent JR, Kawata EE, Lei CH, Lister RM, Larkins BA. 1992. Nucleotide sequence analysis of the genomes of the MAV-PS1 and P-PAV isolates of barley yellow dwarf virus. Journal of General Virology. Vol. 73:487–492. [Cross Ref]

48. Van den Eynde R, Van Leeuwen T, Haesaert G. 2020. Identifying drivers of spatio-temporal dynamics in barley yellow dwarf virus epidemiology as a critical factor in disease control. Pest Management Science. Vol. 76:2548–2556. [Cross Ref]

49. Walker PJ, Siddell SG, Lefkowitz EJ, Mushegian AR, Adriaenssens EM, Alfenas-Zerbini P, Davison AJ, Dempsey DM, Dutilh BE, García ML, Harrach B, Harrison RL, Curtis Hendrickson R, Junglen S, Knowles NJ, Krupovic M, Kuhn JH, Lambert AJ, Łobocka M, Nibert ML, Oksanen HM, Orton RJ, Robertson DL, Rubino L, Sabanadzovic S, Simmonds P, Smith DB, Suzuki N, Van Dooerslaer K, Vandamme AM, Varsani A, Zerbini FM. 2021. Changes to virus taxonomy and to the international code of virus classification and nomenclature ratified by the international committee on taxonomy of viruses (2021). Archives of Virology. Vol. 166:2633–2648. [Cross Ref]

50. Walls J, Rajotte E, Rosa C. 2019. The past, present, and future of barley yellow dwarf management. Agriculture. Vol. 9:23. [Cross Ref]

51. Walsh LE, Lacey S, Doyle D, Gaffney MT, McNamara L. 2022. Evaluating the timing of insecticide application to manage barley yellow dwarf virus and yield in winter barley. Irish Journal of Agricultural and Food Research. Vol. 61:372–383. [Cross Ref]

52. Wei S, Chen G, Yang H, Huang L, Gong G, Luo P, Zhang M. 2023. Global molecular evolution and phylogeographic analysis of barley yellow dwarf virus based on the cp and mp genes. Virology Journal. Vol. 20:130[Cross Ref]

53. Wu B, Blanchard-Letort A, Liu Y, Zhou G, Wang X, Elena SF. 2011. Dynamics of molecular evolution and phylogeography of Barley yellow dwarf Virus-PAV. PLoS One. Vol. 6:e16896. [Cross Ref]

Supplementary material

Supplementary Figure 1.

Neighbor-Joining consensus phylogenetic tree using the Tamura-Nei genetic distance model for the cereal yellow dwarf viruses (CYDVs) assembled in this study (shown in orange) and CYDV genome sequences downloaded from NCBI (shown in black). The GenBank accession numbers are shown alongside the organism name assigned in GenBank. Bootstrap values obtained from 1,000 replicates are indicated along the nodes. The consensus tree was built with a branch support threshold of 70%.

Supplementary Figure 2.

Neighbor-Joining consensus phylogenetic tree using the Tamura-Nei genetic distance model for all barley yellow dwarf viruses (BYDVs) assembled in this study (shown in orange) and BYDV genome sequences downloaded from NCBI (shown in black). The GenBank accession numbers are shown alongside the organism name assigned in GenBank. Bootstrap values obtained from 1,000 replicates are indicated along the nodes. The consensus tree was built with a branch support threshold of 70%.