Virus resistance of Australian pea (Pisum sativum) varieties

Abstract

Australian pea varieties were evaluated for virus resistance using spreader plots sown with Pea seed-borne mosaic virus (PSbMV)-infected seed and aphid inoculations with Bean leafroll virus (BLRV). Natural infections by Bean yellow mosaic virus (BYMV) and Soybean dwarf virus (SbDV) allowed also screening for these viruses. Complete PSbMV resistance was limited to a few varieties, but BYMV resistance was relatively frequent. No consistent ranking of partial PSbMV resistance or PSbMV seed transmission rates were found. Selection for partial resistance to PSbMV is therefore not a practical breeding objective, while incorporating complete resistance is relatively easy. A wide range of resistance was found for BLRV and confirmed in field and greenhouse tests against local BLRV strains in Syria, but infection levels varied over years and no complete resistance was identified. Genotypes with a high level of partial resistance to BLRV also appeared to show resistance to the closely related SbDV.

Keywords:

• resistance breeding
• pulses
• food legumes
• luteovirus
• potyvirus

Introduction

Field pea (Pisum sativum) is an important pulse crop in Australia with a 5-year (2007–2011) average production of 328,100 ton on 274,600 ha (Pulse Australia 2007–2011). Growing peas as a rotation crop improves the sustainability and profitability of broadacre farming through higher soil fertility, better weed control and less stubble and soil-borne disease in the following wheat crops (Stevenson & van Kessel 1996). However, a further increase of pulses in Australia's grain belt, either through shorter rotations between pulse crops or through expansion in new areas, will aggravate its own disease problems. Viral diseases are of particular concern as curative control is not possible and, unlike most fungal diseases, pulse viruses are generally not species specific. Particularly extending field pea cultivation in Australia's northern grain zone (northern New South Wales and southern Queensland), a sub-tropical region characterised by a short and relatively warm winter growing season, will aggravate virus problems. Rain can occur throughout the year in this region and, together with fertile soils with good water holding capacity, allows cultivation of a range of summer crops, rainfed or under irrigation. These summer crops, as well as weeds and volunteers, provide a ‘green bridge’ for pathogens and pests of winter crops between growing seasons. This is especially of importance to virus diseases, as most require living plant tissue to survive and need insect vectors to spread.

Surveys in Australia's main field pea growing regions in the south-east and west indicated the wide spread occurrence of viruses, even though their Mediterranean climate with dry summers is less favourable for virus survival. Pea seed-borne mosaic virus (PSbMV, genus Potyvirus, family Potyviridae) was the most widespread, but infections by Bean leafroll virus (BLRV, Luteovirus, Luteoviridae), Beet western yellows virus (BWYV, Polerovirus, Luteoviridae), Bean yellow mosaic virus (BYMV, Potyvirus, Potyviridae), Cucumber mosaic virus (CMV, Cucumovirus, Bromoviridae) and Alfalfa mosaic virus (AMV, Alfamovirus, Bromoviridae) were reported as well (Latham & Jones 2001a; Freeman et al. 2013). Natural infection of garden and canning peas by Soybean dwarf virus (SbDV, Luteovirus, Luteoviridae, syn. Subterranean clover red leaf virus, SCRLV) has been reported in Tasmania (Johnstone 1978) and by Subterranean clover stunt virus (SCSV, Nanovirus, Nanoviridae) on the central tablelands of New South Wales (Grylls 1972).

An assessment of virus resistance in current Australian pea varieties is needed to support efforts by the field pea breeding program to improve virus resistance in newly developed germplasm. Some anecdotal information is present in industry publications, but apart for studies by Latham & Jones (2001b) on AMV and PSbMV and Coutts et al. (2008) on PSbMV, no systematic screening of Australian pea varieties for virus resistance has been published.

This paper presents data on virus resistance in established Australian pea varieties, as well as in a limited number of advanced breeding lines. The evaluation relied on field screening in a site with a high incidence of natural occurring virus epidemics and was aided by introducing target viruses through inoculation and growing of virus-infested seed lots. Emphasis was given to the two viruses that are considered to be the most important in north-eastern Australia, BLRV and PSbMV. BLRV has only been recently reported in Australia (Schwinghamer et al. 1999), but is known to be destructive in BLRV-susceptible field pea varieties (Hampton 1983). Because of the relatively small area under field pea in the north-east currently, no data on its occurrence in commercial sowings are available. However, within this region, it is widespread in commercial faba bean fields (van Leur et al. 2003), while its presence in lucerne pastures (van Leur & Kumari 2011) will provide a continuous source of inoculum and vectors to neighbouring pulse crops (Hampton 1983). Unlike BLRV, PSbMV does not need to depend on living plants to survive between cropping seasons, as the very high levels of seed transmission found in commercial seed lots of susceptible varieties (Freeman et al. 2013) ensures an abundance of starting inoculum. While not causing crop death like BLRV, PSbMV can cause serious losses to grain yield and quality (Coutts et al. 2009). Because of the importance of seed-borne inoculum in PSbMV epidemiology (Coutts et al. 2009), varieties were as well tested for differences in seed transmission rates. Virus diagnostics were extensively used in the evaluation as virus symptoms can be inconspicuous and easily confused with symptoms caused by root rot, water logging or drought stress.

Material and methods

Field testing at the Liverpool Plains Field Station

The pea genotypes reported in this study were tested over 2–4 years at the NSW DPI Liverpool Plains Field Station (LPFS) at Breeza, northern New South Wales (31.85°S, 150.47°E) as part of an on-going virus resistance screening program of breeding lines and germplasm accessions, generally 150–200 entries per year. Randomised complete block designs with two replicates using plots of four 3.0 m long rows with 0.25 m row spacing within and 1.05 m between plots were used in each season. Plots were sowed with 25 g seed (equals 100–120 seed) using a tractor mounted cone seeder. Breeding materials in different stages of development (established varieties, advanced breeding material, early generation breeding material) were blocked. Plots were separated within sowing runs by 1.5 m wide alleys. These alleys were used to hand-sow virus spreader hill plots (10–15 plants closely sowed together) next to each test plot. Seed for the virus spreader plots consisted of an equal mixture of faba beans, a highly BLRV-susceptible variety (‘Fiord’) and field peas, a mixture of PSbMV-infected seed lots. Tests on PSbMV strains isolated from these pea seed lots indicated that the P4 pathotype was predominant (van Leur unpubl. data). Check varieties were repeated systematically (randomly assigned within equally sized blocks) to monitor the virus distribution through the trial. Seed lots with low levels of PSbMV infection were used for the PSbMV-susceptible checks. Sowing dates and other management details of the trials during 2006–2009 are given in Table 1.

Table 1  Pea virus screening trial management at Liverpool Plains Field Station, 2006/09 seasons.

	2006	2007	2008	2009
Total number of plots	432	528	624	480
Number of test entries (2 replicates)	200	252	296	222
Check varieties	8 (4×replicated)	3 (8×replicated)	4 (8×replicated)	6 (6×replicated)
Sowing dates	17 May	30 May	2 June	26 May
Scoring dates	22 Sep, 9 Oct	20 Sep, 5 Oct	29 Sep	14 Sep, 16 Oct
Scoring system	1–9 scale	1–9 scale	1–9 scale	% symptomatic plants
Harvest dates	20 Nov	30 Oct	6 Nov	5 Nov

Only artificial inoculations with BLRV were carried out. Pea aphids (Acyrthosiphon pisum) were reared in the greenhouse on faba bean and field pea. The aphids were placed on BLRV-infected faba bean in a different greenhouse to acquire the virus 1 week before release in the field. A mixture of BLRV isolates collected over the previous years in northern NSW was used for the inoculations. Starting from early July each year viruliferous aphids were released on the hill plots (around 50 aphids/plot) several times. The spreader plots were monitored with the diagnostic tools described below for virus infection and hoed out after re-infections were detected (generally late August), thereby forcing colonising, viruliferous, aphids into the neighbouring test plots. In each year, an aerial application of a synthetic pyrethroid insecticide (Karate) to control Helicoverpa (H. armigera and H. punctigera) was carried out in October, which would have also decimated any colonising aphid colonies.

Powdery mildew (Erysiphe pisi) was present every year in the trial, but occurred too late and at too low level to warrant the application of fungicides. No other diseases were noted during the 4 years of testing.

Trials were scored for luteovirus symptoms on a plot basis. In 2006, 2007 and 2008, a 1–9 scoring scale was used (1: no or only few plants with possible symptoms; 2: no clear symptoms; 3: <5% plants with clear yellowing or stunting symptoms; 4: 5–10% symptomatic plants; 5: 10–25%; 6: 25–50%; 7: 50–75%; 8: 75–90% symptomatic plants, plants dying; 9: >90% plants dead or dying). This scoring system was changed in 2009 by estimating the percentage of luteovirus symptomatic plants/plot, as percentage scores are more suitable for statistical analysis. The trial was scored twice (at early and late pod set) in 2006, 2007 and 2009, but only once in 2008.

Plots were harvested with a plot combine (Kingaroy Engineering Works, Qld) on dates listed in Table 1.

Virus diagnostics

Tissue blot immunoassay (TBIA) was used for all virus diagnostics, as it provides a reliable, fast and cost efficient methodology to process large numbers of individual plant samples (Makkouk & Comeau 1994; Kumari et al. 2001). Samples were blotted in the laboratory at Tamworth on nitrocellulose membranes (Schleiger & Schuell Protran, 0.45 µm pore size), cut to 9.5×5 cm size (allowing 150–250 blots/membrane). In order to allow testing of the samples to a range of viruses, samples were generally blotted over several membranes (duplicates). The blot position of each individual plant was kept the same on each duplicate membrane, so its reaction to different antibodies could be compared. Blotted membranes were processed at the virology laboratories at ICARDA (Aleppo, Syria) or DPI-Victoria (Horsham, Australia) using the antisera listed in Table 2.

Table 2  Names and abbreviations of virus species and monoclonal (MAb) or polyclonal (PAb) antibodies used for detection of pea viruses 2006–2009.

		Antibodies used for detection^1
Virus species		ICARDA		DPI Victoria, Horsham
Name	Abbrev.	Type	Source	Type	Source
Bean leafroll virus	BLRV	MAb	BBA (6G4)	MAb	DSMZ
Soybean dwarf virus	SbDV	MAb	Agdia (56H1)
Pea seed-borne mosaic virus	PSbMV	PAb	ICARDA (SP9-88)	PAb	DSMZ (Loewe up to July 2008)
Bean yellow mosaic virus	BYMV	PAb	ICARDA (SV205-85)	PAb	Loewe
Alfalfa mosaic virus	AMV	PAb	ICARDA (SC10-86)	PAb	Loewe
Note: ^1BBA, Biologische Bundesanstalt für Land- und Forstwirtschaft, Braunschweig, Germany; DSMZ, Deutsche Sammlung von Mikroorganismen und Zellkulturen.

Sampling was either done randomly (generally 15–20 plants/plot) to determine virus incidences, or selectively to relate symptoms to specific viruses. In both cases, the youngest fully grown leaf or tendril was used for blotting. Time of sampling differed between seasons, with limited sampling until symptoms appeared. Tests for BLRV and PSbMV were made in each year, while testing for other viruses depended on the noticing of typical symptoms or results of preliminary, limited tests.

Testing seed for PSbMV seed-to-plant transmission

Seed harvested of selected entries was tested for PSbMV seed-to-plant transmission (SPT) rates by incubating seed on wet filter paper at 22 °C in the dark and blotting plumules or radicles of the germinated seed for TBIA testing (van Leur et al. 2012). Seed lots were tested in batches of 30 seed together with five seed of a check seed lot with a high, known, PSbMV-SPT rate. A sequential sampling approach was taken: After results were obtained on a single batch of 30 seed, further tests were made on lots that appeared to have low infection levels.

Statistical analysis

Data were analysed using the CropStat V7.2 Statistical Package (International Rice Research Institute, Manilla, Philippines). Least significant differences at P=0.05 are listed if a significant (P<0.05) difference was found among the entries tested.

For the PSbMV-STP transmission rates a 95% confidence range was calculated as described in detail by van Leur et al. (2012).

Greenhouse testing for PSbMV and BYMV resistance

Selected entries were grown in a commercial potting soil (Premium Potting Mix, Searle Pty Ltd, Kilcoy, Qld) in an aphid proof, temperature-controlled (18–22 °C) greenhouse. PSbMV and BYMV isolates used for the inoculation were multiplied on the faba bean variety ‘Fiesta’. The PSbMV isolate originated from a ‘Bluey’ field pea seed lot and belonged to the P4 pathotype according to its virulence on the PSbMV differentials PI 269774 and Dark Skin Perfection. The P4 pathotype is the most virulent among the known PSbMV pathotypes and resistance to this pathotype will act against other known pathotypes as well (Johansen et al. 2001). The BYMV isolate was isolated from a volunteer faba bean at the Tamworth Agricultural Institute and can be classified as BYMV-S as it causes a severe mosaic on faba bean, but only a local necrosis without systemic infection on the Phaseolus vulgaris variety ‘Hawkesbury Wonder’ (Randles et al. 1980).

Young virus-infected faba bean leaves were homogenised in a cold neutral phosphate buffer using a mortar and pestle. A minimum of 10 plants of the entries to be tested were inoculated 10–14 days after sowing by dusting first and second leaves with carborundum powder (silicon carbide #400) after which the virus suspension was rubbed in the leaves. The inoculation was repeated on the third leaves after 1 week. Two weeks after the second inoculation the youngest emerged leaves of the inoculated plants were TBIA tested for virus presence.

Testing at the International Centre for Agricultural Research in the Dry Areas (ICARDA), Syria

Lines selected from the 2008 field trial were included in a BLRV screening trial for pea germplasm sowed at ICARDA's main research station at Tel Hadya, 30 km south of Aleppo, Syria. The lines were tested in two replicates, using plots of single 1.0 m length sowed with 15 seed on 15 December 2008. Two months after sowing, all emerged plants were inoculated with a local Syrian BLRV strain (SV64-95). Aphids acquired the virus by feeding for 2 days on BLRV-infected faba bean plants in the greenhouse after which 10–15 viruliferous pea aphids were placed on each plant. After an inoculation period of 48 h the aphids were killed by the application of an aphicide (Thiamethoxam). Six weeks after inoculation all plants were tested for BLRV presence by TBIA. Progenies of single, virus-free, plants was used for re-testing in the following year. Field testing during the 2009/10 season was similar to the preceding season. A test on juvenile plants in a temperature-controlled greenhouse (18–20 °C) was as well made in the same season. Each entry was tested in two pots with five plants/pot, sowed on 20 February 2010. On 3 March, each plant was inoculated with 10–15 viruliferous aphids. BLRV presence in the youngest fully grown leaf was tested by TBIA 27 days after inoculation.

Results

Field testing at the Liverpool Plains Field Station

2006 season

First sampling (15 randomly selected plants in each of four plots) of the virus-susceptible check varieties Bluey, Excell and Kaspa on 12 July showed no BLRV and PSbMV only in Kaspa (average 7%), likely through seed infection. By 21 August, the PSbMV infection in Kaspa had increased to 38%, while 6% infected plants were found in Excell and Bluey was still PSbMV free. BLRV infection was 8%, 10% and 18% for Kaspa, Excell and Bluey, respectively. Virus infection spread rapidly after this with 90%, 43% and 33% PSbMV and 28%, 56% and 55% BLRV for Kaspa, Excell and Bluey, respectively, on 11 September. Sampling on 22 September showed increased levels for both PSbMV and BLRV (Table 3). The last sampling on 12 October of the four more resistant checks showed PSbMV and BLRV incidences of 90% and 0% for Cressy Blue; 0% and 43% for Yarrum; 100% and 13% for Boreen; and 86% and 28% for Moonlight, respectively.

Table 3  Virus incidences¹ in pea varieties and advanced breeding lines, Liverpool Plains Field Station, 2006/09 seasons².

	BLRV				SbDV	PSbMV				AMV		BYMV
Variety	2006	2007	2008	2009	2009	2006	2007	2008	2009	2007	2009	2007
Established varieties
Alezan	nt^3	nt	7	3	0	nt	nt	100	100	nt	0	nt
Alma	nt	63	58	87	27	nt	73	87	97	17	0	97
Bluey	97	27	33	80	3	73	60	83	90	7	3	93
Boreen	3	0	57	57	0	73	77	87	93	3	7	90
Bundi	nt	79	53	73	17	nt	76	71	70	23	0	0
Cooke	nt	3	53	37	3	nt	85	73	93	21	7	93
Cressy Blue	3	10	30	17	0	37	50	97	100	13	0	0
Dun	nt	nt	25	83	27	nt	nt	100	93	nt	0	nt
Dundale	40	83	53	70	47	70	87	80	77	20	7	97
Dunwa	nt	37	30	63	13	nt	83	97	89	20	3	97
Early Dunn	nt	30	18	83	13	nt	97	75	87	13	3	97
Excell	80	43	51	77	20	80	70	77	90	13	3	97
Glenroy	nt	10	73	80	0	nt	43	33	87	30	7	0
Greenfeast	23	nt	53	20	3	77	nt	97	80	nt	0	nt
Helena	30	30	77	79	14	37	87	100	96	7	0	93
Jupiter	nt	nt	30	30	0	nt	nt	97	100	nt	0	nt
Kaspa	77	67	70	83	17	97	90	100	80	23	0	97
Laura	nt	17	70	60	50	nt	77	90	90	20	3	100
Maki	3	7	37	27	0	4	10	10	3	30	0	57
Moonlight	0	3	53	30	10	73	87	100	100	37	10	7
Morgan	nt	80	100	93	27	nt	63	52	97	10	3	97
Mukta	10	7	43	47	0	47	73	87	87	23	3	0
Parafield	7	37	23	37	0	67	100	100	87	23	7	100
Santi	nt	23	17	47	10	nt	47	67	90	7	0	0
Snowpeak	33	nt	nt	43	3	70	nt	nt	80	nt	0	nt
Soupa	nt	31	65	67	0	nt	24	63	53	17	10	3
Sturt	0	14	52	40	7	63	100	68	100	17	3	100
Wirrega	nt	13	54	27	0	nt	70	85	90	10	7	97
Yarrum	13	13	33	30	0	0	0	0	0	23	3	7
Advanced breeding and parental lines
G-1000	nt	0	23	17	3	nt	0	3	0	60	7	0
OZP805	nt	nt	43	33	0	nt	nt	0	0	nt	7	nt
OZP819	nt	7	37	33	3	nt	83	97	87	60	20	0
OZP913	nt	nt	40	nt	nt	nt	nt	40		nt	nt	nt
LSD (5%)	29	25	43	30	ns	37	34	41	23	25	ns	11
Note: BLRV, Bean leafroll virus; SbDV, Soybean dwarf virus; PSbMV, Pea seed-borne mosaic virus; AMV, Alfalfa mosaic virus; BYMV, Bean yellow mosaic virus; LSD, least significant difference. ^1Average% virus positive plants as determined by tissue blot immunoassay in a random sample of 15 plants in each of two replicates. ^2Sampling dates; 22 Sep 2006, 20 Sep 2007, 29 Sep 2008, 10 Sep 2009. ^3nt, variety not tested in this year.

The trial was twice (22 September and 9 October) scored for virus symptoms (Table 4). Yields of lines with low BLRV incidences and/or low scores were far superior to those of the BLRV-susceptible lines (Table 4).

Table 4  Virus symptoms scores and yields of pea varieties and advanced breeding lines, Liverpool Plains Field Station, 2006/09 seasons.

	2006^1			2007^1			2008^1		2009^2
Variety/breeding lines	22 Sep	9 Oct	Yield (kg/ha)	20 Sep	5 Oct	Yield (kg/ha)	29 Sep	Yield (kg/ha)	14 Sep	16 Oct	Yield (kg/ha)
Established varieties
Alezan			nt^3			nt	3.0	2366	22	69	702
Alma			nt	6.0	8.0	762	8.5	0	100	100	0
Bluey	7.3	8.8	270	9.0	9.0	38	9.0	0	100	100	12
Boreen	2.0	6.3	1775	3.5	7.5	1211	7.0	665	38	97	233
Bundi			nt	5.0	8.0	655	9.0	0	99	100	0
Cooke			nt	1.0	7.0	2020	6.5	1662	50	93	1092
Cressy Blue	1.5	4.3	2156	1.0	5.0	3000	4.0	2083	3	38	2483
Dun			nt			nt	9.0	0	100	100	0
Dundale			nt	6.5	8.0	472	8.0	0	95	100	0
Dunwa	3.0	7.0	1321	3.5	7.0	1599	7.0	1516	85	90	586
Early Dunn			nt	5.5	7.0	903	9.0	0	99	100	0
Excell	6.5	8.5	292	5.5	8.0	807	8.0	92	98	100	0
Glenroy			nt	1.5	5.5	1992	7.5	282	95	100	34
Greenfeast	1.5	7.5	972			nt	4.5	1526	25	90	656
Helena	6.0	7.5	1022	8.0	9.0	150	8.5	0	99	100	0
Jupiter			nt			nt	7.0	438	30	100	282
Kaspa	7.3	8.0	395	7.0	8.0	516	8.5	142	100	100	4
Laura			nt	4.5	6.5	1383	7.5	388	95	100	52
Maki	1.0	3.0	1784	1.0	4.0	3188	3.0	3016	1	33	1680
Moonlight	1.5	5.0	1482	1.5	5.5	1838	6.0	1086	15	88	672
Morgan			nt	6.0	8.0	452	9.0	0	97	100	0
Mukta	2.0	7.5	645	1.0	6.5	1226	7.5	694	53	95	173
Parafield	1.5	5.0	2187	6.0	7.0	953	7.5	1048	88	100	321
Santi			nt	2.5	7.0	1666	8.5	167	83	100	0
Snowpeak	4.0	8.0	504						63	90	352
Soupa			nt	3.5	6.0	2279	7.0	991	75	100	459
Sturt	4.0	5.5	2002	4.5	7.0	1776	7.5	542	50	97	735
Walana			nt	1.5	5.5	2447	4.5	3947	15	55	1919
Wirrega			nt	4.5	6.5	1658	6.5	1471	30	73	1119
Yarrum	1.0	4.8	2398	1.5	6.0	2242	4.5	1828	6	55	1626
Advanced breeding and parental lines
G-1000			nt	1.5	5.0	1868	1.5	1437	3	50	1072
OZP805			nt			nt	1.5	4719	4	43	2399
OZP819			nt	1.0	3.5	2261	3.5	3088	5	55	1744
OZP913			nt			nt	1.5	4362	10	38	2666
LSD (5%)	2.3	1.6	724	1.2	1.2	431	1.8	835	22	27	634
Note: ^11–9 plot score, averaged over two replicates. ^2% virus symptomatic plants/plot, averaged over two replicates. ^3nt, variety not tested in this year. LSD, least significant difference.

2007 season

An unusual high incidence of BYMV symptoms was observed in neighbouring faba bean trials early in the season. No clear mosaic symptoms were noticed in the field pea trial, but random sampling of both faba bean and pea hill plots (30 plants/species) on 28 August showed 17% BYMV and 30% PSbMV in the peas and 30% BYMV in the faba beans. Subsequent random sampling on 10 September of the hill plots (100 plants/species) showed considerable increase in virus incidences with infection levels of 2% and 65% PSbMV, 29% and 64% BYMV and 5% and 25% BLRV for faba bean and field pea, respectively.

A random sampling on 14 Augustus of virus-susceptible varieties showed a low BYMV incidence (<3%) in Kaspa, Excell, Bluey and Dundale. No PSbMV was found in these highly PSbMV-susceptible varieties, indicating that clean seed lots were used and no spread of this virus from the hill plots had occurred so far. BLRV infection was low with only a single infected plant in one Bluey and one Excell plot. However, BLRV was spreading rapidly in the second half of August with a sampling of over 50 symptomatic plants from different plots on 28 August and 10 September showing all stunted plants to be BLRV positive.

Virus incidence had increased by 20 September, with BLRV incidences (averaged over two replicates) of over 60% in Alma, Bundi, Dundale and Kaspa (Table 3). BYMV had spread even faster and a clear distinction between resistant and susceptible entries was found, with nine out of 25 tested varieties showing complete or near complete (not more than two BYMV positive plants out of 30 sampled) resistance and all others having a BYMV incidence of >90%. PSbMV resistance was as well clearly identified, but PSbMV spread was slightly less than BYMV and some lines (Cressy Blue, Glenroy, Santi and Soupa) were showing a reaction (<50% infection) that could indicate a partial type of resistance. AMV was found in all sampled entries, with four entries Bluey, Helena, Santi and Boreen having <10% infected plants. Most plants in the highly BLRV-susceptible Bluey and Helena were already dead well before sampling, which could explain the low AMV level in these lines.

Eight of the pea varieties listed in Table 3 were again sampled on 12 October; Minor differences were found for BLRV, BYMV and PSbMV, but AMV incidence had nearly doubled to 33%, 23%, 57%, 63%, 100%, 50%, 63% and 68% for Cressy Blue, Excell, Glenroy, Kaspa, Mukta, Sturt, Moonlight and Yarrum, respectively.

The trial was twice (20 September and 5 October) scored for virus symptoms (Table 4). A close relationship between the first score and the presence of BLRV in the 25 lines sampled for TBIA test in the same day. Only the two lines with the highest scores, Helena and Bluey (with a score of 8.0 and 9.0, respectively), were atypical with 30% and 27% BLRV positive plants, respectively. However, nearly all plants of these two lines were close to death at the time of sampling, which would have affected the ability of TBIA to detect viruses. Exclusion of these two lines from the analysis increased the correlation coefficient between TBIA score and yield from 0.64 to 0.79.

2008 season

BLRV and PSbMV spread early, with sampling on 17 July (15 random plants/plot) of the checks Kaspa and Excell (four plots/check) as well as the varieties Helena and Bluey (two plots/entry) showing BLRV infection of 5%, 7%, 17% and 10%, respectively and PSbMV infection of 5%, 17%, 20% and 43%, respectively. A sampling of the same plots as well as the two virus-resistant checks Yarrum and Maki on 25 August (four plots/check) showed a rapid progress of infection with BLRV incidences on Kaspa, Excell, Helena, Bluey, Yarrum and Maki of 57%, 42%, 47%, 83%, 12% and 8%, and PSbMV incidences of 70%, 70%, 80%, 40%, 0% and 15%, respectively. The same sampling showed as well a relatively high natural infection by SbDV of 45%, 35%, 40%, 10%, 0% and 2%, respectively. However, the SbDV infection showed a large variability between plots (range in Kaspa 0–93%, in Excell 0–86%). There was no indication that the earlier BLRV infection had an effect on the later SbDV infection: of the total 327 randomly selected plants from 22 plots, 125 were BLRV and 72 SbDV positive. Of these 29 were positive for both viruses.

Sampling on 29 September showed a wide range of BLRV incidences among 32 entries listed in Table 3, with only one variety, Alezan, having less than 10% BLRV positive plants. The samples were not tested for SbDV. There was a clear distinction between PSbMV-resistant and -susceptible lines (Table 3). Some of the new varieties and breeding lines showed a low level of PSbMV indicating genetic heterogeneity.

Scores, taken on 29 September, as well as grain yields were clearly related to BLRV incidence and showed the severe impact of the early season infection (Table 4).

2009 season

Random sampling of pea spreader plots on 21 July showed 4% BLRV and 52% PSbMV. Symptoms were by then also clearly visible in plots of the susceptible check variety Kaspa, with 69% of the virus symptomatic plants BLRV positive and none of the non-symptomatic plants. PSbMV incidences between symptomatic and non-symptomatic Kaspa plants did not differ with 31% and 38% positive plants, respectively. The sampling on 10 September showed a clear distinction between PSbMV-resistant and -susceptible lines, but as well differences among the lines in BLRV resistance (Table 3). SbDV infection reached high levels in some plots with up to 87% positive plants in one plot of the variety Laura. However, the natural infection by this virus was clustered and differences between entries were not statistically different. As in 2008, there was no indication that the two luteoviruses excluded each other: of a total of 1498 plants that were randomly sampled from 100 plots (two replicates of each of 50 lines, among which the 36 varieties and advanced breeding lines listed in Table 3), 834 were BLRV and 121 were SbDV positive. Of these 89 were positive for both viruses. AMV incidences were lower and highly variable between replicate plots with no significant differences between entries (Table 3).

Already at the first score on 14 September the severe impact of BLRV was evident with only few lines showing resistance. One month later most of the established varieties had all plants showing virus symptoms. The severity of the BLRV infection in this season was reflected in the yields with a large number of entries showing complete yield loss.

Greenhouse tests for PSbMV and BYMV resistance

The high frequency of BYMV resistance in the Australian field pea varieties was confirmed following mechanical inoculation in all the lines that showed resistance in the 2007 field test: Bundi, Cressy Blue, Glenroy, Moonlight, Mukta, Santi, Soupa, Yarrum, G-1000 and OZP819. Testing of lines that were not evaluated in the 2007 field trial showed Greenfeast, OZP805, Maki and Walana as well to be resistant. For most of these lines, the resistance was expressed as complete immunity; no reaction to BYMV antisera in TBIA tests of over 10 inoculated plants/variety. Yarrum, Walana, Maki and OZP805 showed a low incidence of BYMV positive plants, indicating that these (relatively new) varieties are not genetically pure for BYMV resistance. PSbMV resistance was confirmed in the four lines that were identified in the field testing; Yarrum, Maki, G-1000 and OZP805 as well as in the variety Walana, which was not TBIA tested in the field. Similarly to the BYMV testing, Maki and Walana both showed a low frequency of PSbMV-susceptible plants.

PSbMV seed-to-plant transmission

PSbMV-SPT rates for PSbMV-susceptible (not immune) lines data are listed in Table 5. Yield failures of BLRV-susceptible lines in 2008 and (especially) 2009 obviously affected the number of lines that could be tested. A wide range in PSbMV-SPT rates was found between entries and years. The variety Alezan reached the highest incidence both in 2008 and 2009. However, the seed source of this variety was found to have a high (7%) level of infection, while for other lines seed sources were used that were free of infection or with low (<1%) incidences. None of the varieties were completely free of infection over the complete testing period, although Cressy Blue, Boreen and Mukta showed low PSbMV-SPT rates over the 4 years.

Table 5   Pea seed-borne mosaic virus (PSbMV) seed-to-plant transmission rates in pea varieties and advanced breeding lines, Liverpool Plains Field Station, 2006/09 seasons.

	2006			2007			2008			2009
Variety	%PST	Range^1	Tested^2	%PST	Range^1	Tested^2	%PST	Range^1	Tested^2	%PST	Range^1	Tested^2
Alezan	nt			nt			29.3	19–43	58	48.3	43–56	180
Alma	nt			2.3	0–7	128	nh			nh
Boreen	0.0	0–3	132	0.0	0–3	131	3.3	0–12	60	0.8	0–5	118
Bundi	nt			0.8	0–4	127	nh			nh
Cooke	nt			0.0	0–3	131	10.0	4–21	60	5.5	1–15	55
Cressy Blue	3.1	1–8	129	0.0	0–3	131	0.0	0–6	59	1.8	0–5	170
Dundale	nt			2.3	0–7	130	nh			nh
Dunwa	2.4	1–7	123	0.0	0–3	124	11.9	5–23	59	3.3	0–12	60
Early Dunn	nt			0.0	0–3	125	nh			nh
Excell	12.3	7–19	130	1.6	0–5	129	28.1	18–42	57	nh
Glenroy	nt			0.8	0–4	131	11.7	5–23	60	nh
Greenfeast	2.4	1–7	123	nt			16.9	9–29	59	0.0	0–6	58
Helena	5.3	2–11	114	4.7	2–10	129	nh			nh
Jupiter	nt			nt			14.0	6–26	57	14.0	6–26	57
Kaspa	10.3	6–17	126	5.3	2–11	132	13.3	6–25	60	nh
Laura	nt			4.6	2–10	130	nh			nh
Moonlight	12.1	7–19	132	0.0	0–3	129	20.0	11–32	60	8.3	3–18	60
Morgan	nt			1.0	0–5	104	nh			nh
Mukta	1.5	0–8	66	0.0	0–3	127	0.0	0–6	55	5.3	2–10	169
Parafield	5.0	2–10	121	0.0	0–3	130	6.8	2–16	59	1.7	0–9	60
Santi	nt			1.5	0–5	131	nh			nh
Snowpeak	12.3	6–23	65	nt			nt			nh
Soupa	nt			0.8	0–4	129	10.5	4–22	57	6.5	1–18	46
Sturt	15.7	10–24	108	2.3	0–7	130	5.7	1–16	53	11.9	5–23	59
Wirrega	nt			0.0	0–3	132	10.2	4–21	59	3.4	0–12	59
Note: ^195% probability range. ^2Total number of germinated seed tested. ^3nt, variety not tested in this year. ^4nh, not enough or no seed harvested for testing.

BLRV screening at ICARDA, Syria

Resistance of the varieties Alezan, Walana and G-1000, as well as that of advanced breeding lines was confirmed in both field and greenhouse tests at ICARDA against a local Syrian BLRV strains (Table 6). However, lines like Maki and Yarrum, which performed well over several seasons in northern NSW, only showed a moderate level of resistance.

Table 6   Bean leafroll virus (BLRV) incidences¹ in pea varieties and advanced breeding lines, ICARDA, 2008/09 and 2009/10 seasons.

	Field testing		Greenhouse tests
Variety	2008/09	2009/10	2009/10
Established varieties
Alezan	21	0	0
Boreen	16	30
Cooke	75
Excell	78
Glenroy	60
Helena	18	45^2	70^2
Kaspa	74
Maki^3		34	30
Moonlight	8	20^2	20^2
Soupa	77
Walana	4	0^2	0^2
Wirresa	50
Yarrum^3		51	47
Advanced breeding and parental lines
G-1000	27	15^2	0^2
OZP805^3		0	10
OZP819^3		5	10
OZP913	13	15^2	0^2
Note: ^1% virus positive plants as determined by TBIA averaged over two replicates. ^22010/11 testing done on single plant progenies of lines tested in the preceding season. ^3Only tested in 2009/10 season.

Discussion

Inoculation with BLRV viruliferous aphids and the use of PSbMV seed-infected spreader plots proved to provide reliable screening for both viruses at the Liverpool Plains field testing site. Natural BYMV and SbDV infection provided useful additional information on resistance to these viruses, while natural AMV infection was too late and too variable to detect differences among tested lines. While the evaluation relied primarily on visual scoring of symptoms caused by the luteoviruses BLRV and SbDV, the extensive use of diagnostics was essential to relate symptoms to infection by specific viruses or to detect viruses like PSbMV that do not cause clear symptoms.

High PSbMV incidences throughout the trials in each of the testing years showed how fast this non-persistently transmitted virus can spread from initial infection sources and its potential threat to pea cultivation in regions with high aphid activity. It also showed the importance of timing in sampling and surveying pea crops. Late sampling will only allow the identification of PSbMV immune lines, as heavy virus pressure will not allow the detection of partial resistance.

The PSbMV–P. sativum pathosystem is probably one of the best researched virus-pathosystems worldwide. After the first description of the virus (Musil 1966), three clearly distinct pathotypes (P1, P2 and P4) were identified (Alconero et al. 1986). Differential reactions of pea germplasm lines to these pathotypes lead to the postulation of four recessive resistance genes (Provvidenti & Hampton 1992). More recently large advances in the understanding of the genetic basis of this pathosystem have been made through the use of sophisticated molecular techniques, with Johansen et al. (2001) demonstrating that PSbMV pathotypes can be explained by the properties of two viral cistrons and predicting the existence of a fourth pathotype, P3. This pathotype was subsequently identified in a faba bean germplasm accession originating from Nepal (Hjulsager et al. 2002). Research by Gao et al. (2004) indicated the existence of only two resistance genes (sbm1 and sbm2): The sbm1 gene confers resistance to all four PSbMV pathotypes, while a different allele of the sbm1 gene, sbm1 ¹ , gives resistance to the P1 and P2 pathotypes only and the sbm2 gene only provides resistance to the P2 and P3 pathotypes. Both genes as well control, or are strongly linked to separate genes that control, resistance to a number of other potyviruses (Bruun-Rasmussen et al. 2007) with sbm2 linked to, or the same as, the mo resistance gene for BYMV.

So far P1 and P4 are the only PSbMV pathotypes reported in Australia (Torok & Randles 2007) and both were present in the PSbMV-infected seedlots that were used for the virus spreader in the field trials (van Leur unpubl. data). Our field results, confirmed by greenhouse tests on juvenile plants, indicated that only the recently released Australian varieties Yarrum, Maki and Walana posses the sbm1 resistance gene. Tests for homogeneity of PSbMV resistance were not made prior to their release and each of these lines still contain a low level of susceptible plants. The Australian field pea breeding program has made the incorporation of PSbMV resistance one of its priorities with a program that screens early generation breeding material for resistance (van Leur et al. 2007) and advanced breeding lines with PSbMV resistance (OZP805 and others) are now available.

Next to sbm genes coding for complete resistance, it is likely that there other genes that provide partial resistance to PSbMV or influence symptom severity (Hampton 1980). Coutts et al. (2008) reported a wide range of PSbMV infection levels among pea genotypes, with the varieties Dundale and Snowpeak the most susceptible. Our results failed to confirm consistent differences in infection levels over years, with the possible exception of Glenroy and Soupa, which showed lower incidences over 3 years of testing. However, partial resistance would only be useful if it also results in lower seed to plant transmission. Of the PSbMV-susceptible (non-immune) varieties tested, only Cressy Blue and Boreen showed consistently a lower (but not zero) PSbMV seed transmission. Varieties with a low PSbMV seed transmission rate have been reported (Wang et al. 1993) and PSbMV free seed will be as effective in the control of this virus as varieties with complete resistance (Jones 2000). However, the infection of pea seed embryo's by PSbMV is influenced by a large number of factors (Roberts et al. 2003) and in our tests even lines like Excell and Kaspa with high infection rates in commercial crops (Freeman et al. 2013), had low PSbMV-SPT rates in some years. Selecting genotypes with low PSbMV seed transmission rates is therefore not a practical objective for a breeding program, while selection of immune, sbm1, segregants is relatively easy.

The finding of a relatively large number of BYMV-resistant varieties was surprising as this virus is not considered to be of importance in the Australian field pea industry and no deliberate selection for BYMV resistance has been made in the past. A high frequency of BYMV resistance was found as well in pea varieties and germplasm in the USA (Kasimor et al. 1997) and BYMV resistance could have been introduced in the Australian pea breeding programs through the use of USA varieties. The fact that the sbm2/mo gene was retained during the local selection process could indicate an advantage for BYMV resistance, even in the absence of the virus. It is tempting to speculate on an effect of a ‘defeated’ sbm2 gene on infection by the sbm2-virulent P1 and P4 PSbMV pathotypes: The two varieties that were showing a (slightly) lower PSbMV incidence in 3 years of field testing, Glenroy and Soupa, are both BYMV resistant, as is one (Cressy Blue) of the two varieties that showed a low PSbMV seed transmission rate. However, both Glenroy and Soupa showed high PSbMV seed transmission rates and other BYMV-resistant varieties like Moonlight are highly PSbMV susceptible.

Natural AMV infection was recorded in 2 years, but occurred too late in the season to allow reliable screening for resistance: Lines that showed a low level of infection in 2007 were highly susceptible to BLRV and already had a high frequency of dead or dying plants early in the season, which would have affected AMV infection. AMV is a potential threat to an expanding pea industry in Australia's northern grain region as the large area under lucerne provides a huge and continuous reservoir of AMV inoculum (van Leur & Kumari 2011). However, so far little progress has been made in identifying AMV resistance in the pea germplasm pool (Timmerman-Vaughan et al. 2001).

In terms of impact on yield, BLRV is arguably the most important field pea virus in northern NSW. Early infection of susceptible varieties resulted in severe stunting and plant death. In each of the four testing seasons, a clear relation between BLRV susceptibility and yield was found, although yield differences would have also reflected general adaptation. Timing on the start of the epidemic had a clear effect on final scores as well as yields: during the 2006 and 2007 seasons BLRV infection was not detected until the middle of August, while in 2008 and 2009 infections were already noticed in July. Symptoms were far more severe in the last two seasons and yields more affected. Differences in resistance among tested genotypes, including older varieties, were apparent, but none showed complete immunity. Variability in susceptibility, resistance and/or tolerance to BLRV among pea varieties was published (Hubbeling 1956) soon after the first description of BLRV (Quantz & Völk 1954). Only a limited number of studies on the genetic basis BLRV resistance in peas have been published. Both Drijfhout (1968) in the Netherlands and Baggett & Hampton (1991) in the USA concluded that the resistance was based on a single recessive gene, but commented on the difficulties in distinguishing between resistant and susceptible segregants. Crampton & Watts (1968) in New Zealand on the other hand concluded that resistance to pea leaf-roll (top-yellows) was based on additive and dominant genes. Wilson (1968) described studies of the vectors on ‘pea leaf roll virus’ in New Zealand, but did not confirm the identity of the causal organism with serology. It is more likely that the ‘top-yellows’/'pea leaf roll’ disease of peas in New Zealand is caused by SbDV (Ashby et al. 1979) in addition with, possibly, a BWYV strain, BWYV-NZ (Kyriakou 1984). Also, BLRV was not identified in a survey of legume crops in New Zealands's South Island, while SbDV was widespread (Fletcher 1993). In view of the absence of BLRV in New Zealand, it is interesting to note that some of the most BLRV-resistant varieties in our study (Yarrum, Maki, Walana) were selected in the virus prone environment of northern NSW, but bred in New Zealand (Adrian Russell, Plant Research (NZ) Ltd). SbDV infections were recorded in our trials in both 2008 and 2009. Although infection occurred late and incidences were variable between replicates, lines with low levels of BLRV infection also showed low SbDV incidences, indicating the possibility of common genes for resistance to these two closely related luteoviruses. Analysis of individual plant results demonstrated that the infection of the two viruses was not mutually exclusive, but that SbDV showed a more clustered infection pattern than BLRV. This could be caused by different vectors: the foxglove aphid (Aulacorthum solani) is considered the main vector of SbDV (Johnstone & Patten 1981), although strains specifically transmitted by the pea aphid (Acyrthosiphon pisum) have been reported in Australia (Srithongchai 1990; Schwinghamer et al. 2009).

Baggett & Hampton (1991) did not exclude the possibility of additional, additive genes for BLRV resistance. The variety Yarrum, extensively used in the breeding program as a donor of resistance to both BLRV and PSbMV, was the resistant parent in the cross from which the advanced breeding line OZP805 was selected. The generally better performance of OZP805 (symptom development, TBIA scores as well as yield) over its parent Yarrum could indicate additive genes playing a role in resistance or symptom expression. High levels of BLRV resistance in other advanced breeding lines selected from crosses that did not involve specific BLRV-resistant parents (OZP819 and OZP913) would also point to additive genes.

Among the different potential sources for BLRV resistance that we evaluated, G-1000 deserves a special mention. This vegetable type pea, bred at the Cornell University Agricultural Station (Geneva, New York), was developed to combine resistance to a range of potyviruses (Provvidenti et al. 1991). Its immunity to both BYMV and PSbMV was confirmed in our field and greenhouse trials, but it also consistently showed a high level of BLRV resistance, both in Australia and in Syria.

The clear relation between the BLRV screening data in Australia and in Syria would indicate that the BLRV resistance identified, although only partial, will be lasting. Syria is located in the centre of origin of P. sativum (as well as other legume and cereal food crops) and legume pathogens are likely to be more variable than in Australia with its recent history of legume cultivation.

BLRV resistance in field peas appears to be highly heritable, but the partial nature of the resistance identified so far complicates its identification and necessitates repeated field testing over sites and years. Earlier studies relied largely on symptom expression only, while our extensive use of TBIA helped at quantifying BLRV resistance. While the Australian field pea breeding program has been successful in incorporating BLRV resistance in new breeding lines, the source and genetic basis of the resistance is unknown. In view of its importance, more detailed studies on BLRV resistance are warranted and more research is needed to explore diverse pea germplasm collections for resistance to broaden the BLRV resistance gene pool. As current screening methods are costly and time consuming, the pea–BLRV pathosystem would be an ideal candidate for the development of molecular marker-based resistance screening.

Acknowledgements

We would like to acknowledge the technical assistance of Janine Sipple, Finn Fensbo and Merv Riley. We also thank the NSW DPI District Agronomists for providing pea seed lots. This study was financially supported by the Grains Research and Development Corporation, Australia.