Yield-limiting potential of Beet western yellows virus in Brassica napus

CSIRO PUBLISHING

www.publish.csiro.au/journals/ajar Australian Journal of Agricultural Research, 2007, 58, 788–801

Yield-limiting potential of Beet western yellows virus in Brassica napus

R. A. C. JonesA,B,C, B. A. CouttsA, and J. HawkesA,B

AAgricultural Research Western Australia, Locked Bag No. 4, Bentley Delivery Centre, WA 6983, Australia.BCentre for Legumes in Mediterranean Agriculture, University of Western Australia, 35 Stirling Highway,Crawley, WA 6009, Australia.

CCorresponding author. Email: [email protected]

Abstract. Losses in seed yield and quality caused by infection with Beet western yellows virus (BWYV) alone or incombination with direct feeding damage by Myzus persicae (green peach aphid) were quantified in field experiments withBrassica napus (canola, oilseed rape) in the ‘grainbelt’ region of south-western Australia. Plants infected with BWYVand infested with M. persicae were introduced into plots early to provide infection sources and spread BWYV to B. napusplants. Insecticides were applied as seed dressings and/or foliar applications to generate a wide range of BWYV incidencesin plots. Colonisation by vector aphids and spread of BWYV infection were recorded in the plots of the different treatments.At sites A (Medina) and B (Badgingarra) in 2001, foliar insecticide applications were applied differentially at first, but,later, ‘blanket’ insecticide sprays were applied to all plots to exclude any direct feeding damage by aphids. When BWYVinfection at sites A and B reached 96% and 100% of plants, it decreased seed yield by up to 46% and 37%, respectively.Also, variation in BWYV incidence explained 95% (site A) and 96% (site B) of the variation in yield gaps, where for each1% increase in virus incidence there was a yield decrease of 12 (site A) and 6 (site B) kg/ha. At both sites, this yield declinewas entirely because fewer seeds formed on infected plants. At site B, BWYV infection significantly diminished oil contentof seeds (up to 3%), but significantly increased individual seed weight (up to 11%) and erucic acid content (up to 44%);significant increases in seed protein content (up to 6–11%) were recorded at both sites. In field experiments at sites B and C(Avondale) in 2002, insecticides were applied as seed dressings or foliar sprays. At site B, when BWYV incidence reached98%, the overall yield loss caused by BWYV and direct M. persicae feeding damage combined was 50%. At site C, whenBWYV incidence reached 97%, the overall combined yield decline caused by BWYV and direct feeding damage was 46%.This research under Australian conditions shows that, when aphids spread it to B. napus plantings such that many plantsbecome infected at an early growth stage, BWYV has substantial yield-limiting potential in B. napus crops. Although theresults represent a worst case scenario, the losses were greater than those reported previously in Europe and are causefor concern for the Australian B. napus industry. When applied at 525 g a.i./100 kg of seed, imidacloprid seed dressingcontrolled insecticide-resistant M. persicae and effectively suppressed spread of BWYV for 2.5 months and increasedseed yield by 84% at site B and 88% at site C. Therefore, provided that mixing the insecticide with seed is sufficientlythorough, dressing seed with imidacloprid before sowing provides good prospects for control of BWYV and M. persicaein B. napus crops.

Additional keywords: BWYV, Myzus persicae, virus, aphids, canola, oilseed rape, symptoms, insecticide, yield loss,seed quality, economic loss, control, Australia.

Introduction

Brassica napus (canola, oilseed rape) is an important oilseedcrop grown widely in many countries. The name ‘canola’denotes B. napus cultivars with erucic acid seed concentrations<2% of oils and glucosinolate contents <40 µmol/g of mea1(Salisbury et al. 1999). B. napus was first grown commerciallyin 1969 in Australia. However, the early cultivars sown allhad oils with high erucic acid concentrations and mealswith high glucosinolate contents, and were susceptible toblackleg disease caused by the fungus Leptosphaeria maculans.Subsequently, cultivars were bred with improved oil quality,low glucosinolate levels, resistance to blackleg disease, triazineherbicide tolerance, and improved yields. Once these becameavailable, the area sown increased rapidly from 150 000 ha in1991 to 1.2 million ha in 1998 (Salisbury et al. 1999). In the

‘grainbelt’ agricultural region of south-western Australia, whichhas a Mediterranean-type climate, the area sown to B. napusnow averages 0.45 million ha/year, fluctuating between 0.25 and0.92 million ha annually. The crops are sown in late April to earlyJune (late autumn–early winter) following the first substantialrains and harvested in November or December (late spring–earlysummer). There is little summer rainfall, which is insufficientto support rain-fed crop production at this time of the year.However, there is sufficient moisture at isolated sites such asroadside ditches, the edges of creeks, and soakaways for smallpopulations of volunteer B. napus plants to persist over summer(Coutts et al. 2006).

Beet western yellows virus (BWYV, family Luteoviridae,genus Polerovirus) is an important pathogen of B. napus andother Brassica species worldwide. The virus is spread in a

© CSIRO 2007 10.1071/AR06391 0004-9409/07/080788

Yield losses from BWYV in Brassica napus Australian Journal of Agricultural Research 789

persistent manner by 10 or so different aphid species, but Myzuspersicae (green peach aphid) is the principal natural vector.Brevicoryne brassicae (cabbage aphid), which transmits it lessefficiently, may also act as a minor BWYV vector in B. napuscrops (Duffus 1972; Walsh and Tomlinson 1985; Brunt et al.1996; Smith and Barker 1999). Recently, the virus isolatesreferred to as BWYV were reclassified into 4 distinct virusspecies: BWYV, Turnip yellows virus (TuYV), Beet chlorosisvirus, and Beet mild yellowing virus (Graichen and Rabenstein1996; Hauser et al. 2000, 2002; D’Arcy and Domier 2005).BWYV and TuYV both infect Brassica species under naturalconditions but the other 2 viruses do not. However, BWYVand TuYV cannot be distinguished readily by immunologicalmeans because they are closely related serologically (Duffus andRussell 1972; Hauser et al. 2000). When BWYV isolate WA-1from south-western Australia was analysed by O. Lemaire andM. Beuve (pers. comm.), its sequence resembled that of TuYVisolates from Europe (Coutts et al. 2006). However, insufficientAustralian isolates have been sequenced to determine whetherthis or both viruses infect Brassica species here. Therefore, untilmany more Australian isolates are sequenced and a phylogeneticanalysis is done with them, it seems appropriate to continue usingthe traditional nomenclature for the present, so in this article onlythe name BWYV is used.

In single infected B. napus plants, BWYV diminished seedyield by up to 50% (Schroder 1994). In experiments usinginsecticides to determine the effect of BWYV on B. napus plotsin the UK, Smith and Hinckes (1985) recorded increases of 10%in seed yield and 3% in oil content from plots with low comparedwith high amounts of BWYV infection, while Jay et al. (1999)obtained seed yield losses of up to 26%, a 2% decrease inoil content, and an 11% increase in glucosinolate levels dueto BWYV. In similar field experiments in Germany, seed yieldlosses of 12–34% were reported from plots with low comparedwith high amounts of BWYV infection (Graichen 1995, 1997;Graichen and Schliephake 1999).

In large-scale spring surveys to determine the incidence ofvirus infection in B. napus crops and Rapanus raphanistrum(wild radish) weeds in the south-western Australian grainbelt,BWYV was detected in 59% (1998) and 66% (1999) of crops,with incidences within individual infected crops of 1–65%(1998) and 1–61% (1999). R. raphanistrum was infected at 9of 12 sites sampled in 5 of 6 districts (1998) and 10 sites in7 districts (1999), with incidences of up to 48% (1998) and96% (1999) at individual sites. BWYV-infected samples ofB. napus and R. raphanistrum came from all rainfall zones inthe grainbelt region (Coutts and Jones 2000). BWYV infectionis also common in the region in spring in Vicia faba (faba bean),Pisum sativum (field pea), and Cicer arietinum (chickpea) crops,and occurs in Medicago sativa (lucerne), Trifolium subterraneum(subterranean clover), and Ornithopus sativus (pink serradella)pastures (McLean and Price 1984; Latham and Jones 2001;Hawkes et al. 2002; Jones 2004). During the summer periodsof 2000–02, large-scale tests on samples from broad-leafedweed species and volunteer crop plants persisting in isolated,damp sites in the grainbelt region detected BWYV at lowincidences in Citrullus lanatus (Afghan or wild melon), Conzyaspp. (fleabane), Navarretia squarrosa (stinkweed), Solanumnigrum (blackberry nightshade), and volunteer B. napus. The

virus was found in high- and medium-rainfall zones and smallpopulations of aphids were over-summering in all rainfall zones,mostly infesting volunteer B. napus and R. raphanistrum; thepredominant aphid species was B. brassicae, with M. persicaepresent occasionally (Coutts et al. 2006). Outside the growingseason BWYV also persists at low incidences in some perennialnative plant species, and perennial pastures of M. sativa (Jones2004; Coutts et al. 2006). Despite the widespread occurrenceof BWYV infection in B. napus, information is lacking onits effect on seed yield and quality in Australia. This paperquantifies the yield losses caused by BWYV in B. napusunder south-western Australian conditions, and investigates theextent to which seed quality is affected. It also examines themagnitude of the yield losses that result when BWYV infectionand direct feeding damage by M. persicae occur together,and the potential for using imidacloprid seed dressing asa control measure.

Materials and methodsVirus isolates, plants, inoculations, and antiserumPlants were grown in insect-proofed, air-conditionedglasshouses and maintained at 15–20◦C. Plants of B. napuscv. Pinnacle and B. juncea (mustard) cv. Tendergreen weregrown in a steam-sterilised potting mix containing soil, sand,and peat (1 : 1 : 1). For aphid inoculations, M. persicae werestarved for 2 h, and then placed on infected leaves for 2 daysbefore feeding on healthy plants (10 aphids/plant) for 1–2 daysbefore being killed with aphicide. For sap inoculations, infectedleaves were ground in 0.1 M phosphate buffer, pH 7.2, andthe sap mixed with ‘celite’ (diatomaceous earth) before beingrubbed onto the leaves of B. napus plants. BWYV isolateWA-1 (Coutts and Jones 2000) was maintained by aphidinoculation to B. napus. Turnip mosaic virus (TuMV, familyPotyviridae, genus Potyvirus) isolate WA-Ap (Coutts and Jones2000) was maintained by sap inoculation to B. juncea plants.Cauliflower mosaic virus (CaMV, family Caulimoviridae, genusCaulimovirus) was obtained, under licence from the AustralianQuarantine and Inspection Service (AQIS), as freeze-driedinfected sap from Sanofi Phyto-Diagnostics, France. Samplesfrom the cultures of BWYV and TuMV, and freeze-driedmaterial of CaMV, were used as positive controls in serologicaltests. Tissue blot immunoassay (TBIA) was used for BWYV,and enzyme-linked immunosorbent assay (ELISA) for TuMVand CaMV. ‘Infector plants’ of B. napus cv. Pinnacle wereproduced by placing viruliferous M. persicae from coloniesgrowing on B. napus plants infected with BWYV isolate WA-1onto B. napus cv. Pinnacle seedlings (5 aphids/plant) growingin peat pots, and leaving them in place to colonise. Polyclonalantisera to BWYV, CaMV, and TuMV were obtained fromSanofi Phyto-Diagnostics, France, or DSMZ, Germany.

Tissue blot immunoassay (TBIA)B. napus stems were tested singly or bound in bundles of up to10 with Parafilm. A scalpel was used to cut the ends off single-stem or bundle samples. The cut surfaces were pressed twiceonto 0.45-µm pore size nitrocellulose membrane (Schleicher andSchuell, Inc., Keene, NH, USA). The procedure for TBIA wasas described by Coutts and Jones (2000).

790 Australian Journal of Agricultural Research R. A. C. Jones et al.

Enzyme-linked immunosorbent assay (ELISA)B. napus leaf samples were extracted singly or in groups of2–10 leaves in phosphate buffered saline (10 mM potassiumphosphate, 150 mM sodium chloride), pH 7.4, containing5 mL/L of Tween 20 and 20 g/L of polyvinyl pyrrolidone(1 g leaf/20 mL), using a leaf press (Pollahne, Germany). Theextracts were collected in labelled, plastic sample tubes andtested by double antibody sandwich ELISA using paired wellsin immunoplates as described by Clark and Adams (1977).The substrate used was 0.6 mg/mL of p-nitrophenyl phosphatein 100 mL/L of diethanolamine, pH 9.8. Absorbance values(A405 nm) were measured in a Titertek Multiskan immunoplatereader (Flow Laboratories, Finland). Absorbance values morethan twice those recorded for healthy leaf sap were consideredto represent infected plants. Virus incidence was estimated fromgrouped sample test results using the formula of Gibbs andGower (1960).

Field experimentsGeneral site and experimental detailsDepartment of Agriculture Research Stations at Medina

(32◦14′S, 115◦48′E), Badgingarra (30◦20′S, 115◦32′E), andAvondale (32◦06′S, 116◦52′E) were used for the fieldexperiments. Experiment 1 was at Medina in 2001, Expts 2 and3 at Badgingarra in 2001 and 2002, respectively, and Expt 4at Avondale in 2002. Appropriate herbicides were applied tothe soil to prepare ‘weed-free’ seed beds. B. napus cv. Pinnaclewas sown and Agflowr (100 g/ha) treated with fungicide drilledwith the seed. The day after sowing, insecticide and furtherherbicide were applied to provide early pest and weed protection,respectively, according to standard practice. All plots were top-dressed with nitrogen fertiliser (usually 120 kg/ha sulfate ofammonia and 50 kg/ha urea) at regular intervals. Grass andbroad-leafed weeds within growing plots were controlled usingappropriate selective herbicides and rigorous hand weeding;buffers and plot margins were also kept weed-free. To initiateearly virus spread, BWYV infector plants also infested withM. persicae were introduced into all plots of some treatmentsin each experiment. They were planted within each plot in aregular arrangement. To avoid sampling leaf/stem material fromthem, the infector plants were identified by tagging them withcoloured surveyors’ tape. Sampling was by walking in a ‘Z’- or‘W-shaped pattern through each plot. Every few paces, a leaf(before stem elongation) or a shoot (after stem elongation) wasselected from one canola plant for removal. Samples were thenstored in a cooler box for transport back to the laboratory wherethey were tested for virus presence. In general, colonising aphids(winged, non-winged, and nymphs) were counted before andafter spray applications, and the aphid species identified in situ.One 10-cm growing shoot tip and one old leaf wereexamined/plant. Depending on the aphid populations present,these counts were done on 25 or 50 plants/plot.

Experiment 1On 7 June 2001, 2 strips of B. napus, 6-m wide, were sown

3-m apart at a seedling rate of 10 kg/ha and a row spacingof 8–9 cm, using a ‘seeding spider’. The seedlings were fullyemerged by 32 days after sowing (DAS), when all plots were

hand-thinned. From each strip, sections 3-m wide were thensprayed out with herbicide to generate plots 3 m by 6 m in size,separated by bare-earth buffers 3-m wide, arranged along eachstrip of 18 plots. There were 4 treatments and 8 replications anda randomised block design was used. The individual treatmentswere: plots with or without added infector plants sprayed withinsecticide at emergence, and then 2, 4, 6, 8, and 10 weeksafter emergence; and plots with or without added infectorplants that were left unsprayed. Infector plants were introducedinto plots at time of sowing (10 plants/plot). Combined foliarinsecticide applications of α-cypermethrin (500 mL/ha Fastacr )and imidacloprid (170 mL/ha Confidorr ) in 100 L/ha were used.The treatment insecticide applications started at 9 DAS andfinished at 81 DAS, but subsequently 2 blanket insecticidesprays of the above insecticides were applied to all plots at93 and 106 DAS to prevent substantial late spread of BWYVinfection by aphids from plots without insecticide to those withinsecticide. Plant density was counted in each plot at 39 DAS,within 5 quadrats/plot (0.5 by 1.4 m) each laid across 6 rowsof plants.

Colonising aphid counts were made within replicates 1, 3,5, and 7 at 39, 55, 67, and 82 DAS. To assess if the blanketsprays applied to all plots had been successful, further countswere done at 95 and 109 DAS within 1 plot of each treatment.Leaf or shoot samples (50 per plot) collected at 55, 67, 82, 95,109, and 123 DAS were tested for BWYV. On the final samplingdate, separate tests for CaMV and TuMV were also done butneither was ever detected.

The first 11 plots were harvested by machine but seed wasspilt in this process. The remaining 21 plots were hand-harvestedavoiding their edges by taking a 4 m by 3 m area within each,threshing into bins, and whole-plot yields weighed. Only thehand-harvest weights were used to analyse yield data, but theanalyses for 1000-seed weight, oil content, and protein anderucic acid levels involved samples of seed from each of all32 plots.

Experiment 2The experiment was sown with B. napus seed (5 kg/ha), using

a cone seeder on 8 May 2001, and the plants were fully emergedby 11 DAS. Row spacing was 17.5 cm and the plots were each4.6 by 10 m. Triticum aestivum (wheat) buffers 4.6-m wide weresown around each plot. The 6 individual treatments were: plotswith or without added infector plants sprayed with insecticide atemergence, and then 4 and 8 weeks later; plots with or withoutadded infector plants sprayed only at 8 weeks; and plots withor without added infector plants that were left unsprayed. Theoriginal experimental design was a randomised block but due toincorrect initial spraying it subsequently became an unbalancedrow–column design with 6 plots per treatment, except for thetreatment with infector plants and insecticide applications atemergence, and 4 and 8 weeks later, which had 12 plots, andthe treatment with these insecticide sprays that lacked infectorplants, which had 4 plots. Infector plants were introducedinto plots at 6 DAS (12 plants/plot). As in Expt 1, combinedfoliar insecticide applications of α-cypermethrin (500 mL/haFastacr ) and imidacloprid (170 mL/ha Confidorr ) in 100 L/hawere used. The treatment insecticide applications started at11 DAS (emergence) and finished at 67 DAS, but subsequently,


blanket sprays were applied to all plots at 91 DAS and thereafteron 3 more occasions at fortnightly intervals. Plant density wascounted in each plot at 79 DAS within 8 quadrats (0.5 by 1.4 m),each laid along 3 rows of plants.

Colonising aphid counts were done at 29, 50, 63, and 91 DASin 4 plots of each treatment. To assess if the blanket sprays hadbeen successful, similar counts were also done in 1 plot of eachtreatment at 105, 119, and 133 DAS. Cylindrical traps consistingof plastic jars covered with yellow sticky paper (‘Contact’, NylexCorp., Melbourne, Australia) mounted 1.5 m above the groundon stakes were used to monitor aphids flying above the plantings,as described by Bwye et al. (1997). A trap was positioned atopposite ends of the field experiment at 14 DAS, and the trapsremained in place until 148 DAS. The sticky paper (14 by 43 cm)used in each trap was changed every 1–2 weeks, labelled and theaphids caught counted. The alatae were identified by referralto Blackman and Eastop (2000). Leaf or shoot samples werecollected at 50, 63, 78, 91, 105, 117, and 133 DAS and tested forBWYV. The samples from 117 DAS were also tested for CaMVand TuMV but neither was ever detected.

The plots were swathed before seed harvest, which was doneby taking two 10 by 1.8 m runs through each plot, avoiding theiredges. Plot yields were weighed. Further analysis of samples ofthe seed from each plot involved 1000-seed weight, oil content,and protein and erucic acid levels.

Experiment 3The experiment was sown with B. napus seed (5 kg/ha) using

a cone seeder on 11 June 2002, and the plants were fullyemerged by 15 DAS. Row spacing was 17.5 cm and the plotswere 4.6 × 10 m each. The bare earth buffers that surroundedthem were 4.6 m wide. A randomised block design was usedwith 6 individual treatments and 4 replications. The treatmentswere: plots with added infector plants that were sown withuntreated seed and sprayed with insecticide at 3 and 7 weeks afteremergence; plots with added infector plants that were sown withinsecticide dressed seed and either left unsprayed or sprayed at3 and 7 weeks; control plots without added infector plants thatwere sown with dressed seed and sprayed at 3, 7 and 11 weeks;and control plots with or without infector plants that were sownwith untreated seed and left unsprayed. At 5 DAS, infector plants(12/plot) were planted into plots of treatments with these, andhealthy B. napus transplants (12/plot) infested with M. persicaeinto the plots of the other treatments; all of these introducedtransplants were removed at 49 DAS. The insecticidal seeddressing used was immidacloprid (Gauchor ) at 525 g a.i./100 kgof seed; the chemical was mixed with 0.5-kg batches of seed ina 2-L glass jar. The foliar insecticide spray was α-cypermethrin(Fastacr at 250 mL/ha in 100 L/ha of water), and it was appliedto the plots at 35, 62, and 92 DAS. Plant densities were countedat 22 DAS within 5 quadrats (1.4 by 0.5 m) placed at randomwithin each plot. Colonising aphid counts were done in all plotsat 29, 43, 57, 71, 85, 99, and 101 DAS. Leaf or shoot sampleswere collected at 29, 43, 57, 71, and 85 DAS from all plotsbelonging to 4 replicates. The plots were swathed before seedharvest, which was done by taking two 10 by 1.8 m runs througheach plot, avoiding their edges. The seed yield from each plotwas weighed.

Experiment 4The experiment was sown with B. napus seed (5 kg/ha),

using a cone seeder on 7 June 2002, and the plants were fullyemerged at 18 DAS. Row spacing, plot size, experimental design,and insecticide treatments were as in Expt 3, except that therewere 7 replicates instead of 4 and the insecticide sprays wereapplied at 3, 7, 11, and 15 weeks after emergence (at 40, 68,96, and 124 DAS). At 4 DAS, infector plants (12/plot) weretransplanted into plots of treatments with these, and healthyB. napus transplants (12/plot) infested with M. persicae into theplots of the other treatments; all of these introduced transplantswere removed at 53 DAS. Plant densities were counted withineach plot at 26 DAS within 5 quadrats (1.4 by 0.5 m) placed atrandom within each plot. Colonising aphid counts were done onthe plots in replicates 1, 3, 5, and 7 of all treatments at 39, 50, 63,75, 89, and 102 DAS. Leaf or shoot samples were collected fromall plots at the same times (excluding 102 DAS). The plots wereswathed before seed harvest, which was done by taking two 10by 1.8 m runs through each plot, avoiding their edges. The seedyield from each plot was weighed.

Seed quality measurementsFrom each plot in Expts 1 and 2, 1000 seeds were countedand weighed to provide a measurement of seed size. Then,samples of 200 seeds/plot were sent to the Western AustralianState Government Chemistry Centre where each sample wasanalysed to determine its moisture, oil, and protein and erucicacid contents. Moisture, oil, and protein contents of seeds weredetermined simultaneously by near infrared spectroscopy (NIR)(Aksouh et al. 2001). From each seed sample, 0.25 g was placedinto a micro-cup sample holder, which was placed into an NIRanalyser (FOSS NIRSystems 6500, Copenhagen, Denmark) andthe sample scanned an average of 32 times. The NIR analyserhad been calibrated to measure these parameters. The method ofFirestone and Horwitz (1979) was used to measure erucic acidcontent. In brief, extracted oil (10–20 mg) was hydrolysed underreflux with sodium hydroxide-methanol. Next, the free fattyacids were esterified with boron trifloride methanol complexunder reflux. After addition of hexane and saturated sodiumchloride solution, the methyl esters were transferred to theorganic phase, which was dried over anhydrous sodium sulfate.Erucic acid content was then determined by gas chromatographyusing a carbowax fused silica capillary column and flameionisation detection. Results for contents of protein (in oil-freemeal), oil, and erucic acid in seed are reported as percentages atseed moisture contents of 6.3% (Expt 1) and 5.9% (Expt 2).

Statistical analysisData for each plot for aphid numbers, area under the pathogenprogress curve (AUPPC), angular transformed percentageBWYV incidence, and seed yield were subjected to ANOVAusing GENSTAT for Windows, release 6.1. ANOVA forrandomised block designs was used with the data analyses inExpts 1, 3, and 4, except with seed yield data from Expt 1 wherea regression procedure was used to fit the ANOVA model to theyields from the plots harvested by hand. In Expt 2, because of theunbalanced row–column design, an ANOVA model that adjustedfor row and column effects was used instead. In Expts 1 and


2, linear regression was used to relate final BWYV incidenceassessments (x) to data for seed yield, yield gaps, 1000-seedweight, and oil, protein, and erucic acid contents, using bothindividual plot results and mean results for each treatment (y);this was not done with Expts 3 and 4 because no late blanketinsecticide sprays were applied to remove aphid direct feedingdamage as a secondary factor potentially contributing to theiryield losses. Coefficients of determinations (R2), F-tests, andstandard errors of the y-estimate were determined. Yield gapswere measured by taking the maximum potential yield in eachexperiment (i.e. the best individual plot yield) and subtractingthe yield obtained from each of the other individual plots asdescribed by McKirdy et al. (2002) and Thackray et al. (2005).

Results

Plant densities

Plant establishment was an average of 110 plants/m2,58 plants/m2, 55 plants/m2, and 91 plants/m2 in Expts 1, 2, 3, and4, respectively. Better plant establishment in Expt 4 reflected soiltype, which was loamy (Expt 4) or sand (Expts 1–3).

Aphid numbers and species

In Expt 1, populations of colonising aphids remained lowthroughout. The highest count for M. persicae in plots withoutinsecticide before blanket spraying was only 1.3/plant sample,while its numbers in plots with insecticide never exceeded0.1/plant sample (Fig. 1a). A few B. brassicae and Lipaphiserysimi (turnip aphid) were also present but in very lownumbers. No aphids were found after blanket insecticide sprayscommenced.

Fig. 1. Mean counts of numbers of non-winged Myzus persicae/plantsample in the different treatments in each experiment. (a) Expt 1: �, withinfector plants (IP) and insecticide; �, with IP without (w/o) insecticide; N ,w/o IP or insecticide; (×), w/o IP with insecticide. The 6 fortnightly treatmentfoliar insecticide applications started at 9 DAS and finished at 81 DAS.Blanket insecticide sprays to all plots were at 93 and 106 DAS. (b) Expt 2:�, with IP and insecticide sprayed at emergence, then at 4 and 8 weeks afteremergence; �, with IP and insecticide sprayed at 8 weeks after emergence;N , with IP w/o insecticide; (◦), w/o IP or insecticide; (×), w/o IP withinsecticide sprayed at emergence, then at 4 and 8 weeks after emergence;•,w/o IP with insecticide sprayed at 8 weeks after emergence. The 3 treatmentfoliar insecticide applications were at 11, 39, and 67 DAS. The 4 fortnightlyblanket insecticide sprays to all plots commenced at 91 DAS. [� and (×) notseen in figure because of other symbols at same positions along horizontalbaseline.] (c) Expt 3: �, with IP and foliar insecticide sprayed at 3 and7 weeks after emergence; �, with IP and insecticide seed dressing usedalone; N , with IP, insecticide seed dressing, and foliar insecticide sprayed at3 and 7 weeks after emergence; (×), with IP w/o insecticide; (◦), w/o IP orinsecticide;•, w/o IP, with insecticide seed dressing, and foliar insecticidesprayed at 3, 7, and 11 weeks after emergence. Treatment foliar insecticideapplications were at 35, 62, and 92 DAS. (d ) Expt 4: �, with IP andfoliar insecticide sprayed at 3 and 7 weeks after emergence; �, with IP andinsecticide seed dressing used alone; N , with IP, insecticide seed dressing,and foliar insecticide sprayed at 3 and 7 weeks after emergence; (×), withIP w/o insecticide; (◦), w/o IP or insecticide; •, w/o IP, with insecticideseed dressing, and foliar insecticide sprayed at 3, 7, 11, and 15 weeks afteremergence. Treatment foliar insecticide applications were at 40, 68, 96, and124 DAS.

In Expt 2, M. persicae and L. erysimi colonised theplots but M. persicae was the predominant species, withL. erysimi found on only one occasion (91 DAS). In plotswith infector plants, insecticide treatment effectively suppressedaphid numbers (Fig. 1b). For example, the mean numbers ofM. persicae increased from 15/plant sample at 63 DAS to62/plant sample at 91 DAS where no insecticide was applied(excluding subsequent blanket sprays), but where insecticidewas applied at 8 weeks (67 DAS) the respective figures were11 and 0.01/plant sample. Aphids were very slow to spreadto the plots without infector plants (i.e. where no aphids wereintroduced), only 0.6/plant sample being present at 63 DAS inplots of the treatment without insecticide until blanket spraying.After the aphid counts done earlier on the day when the blanketsprays with insecticide commenced (91 DAS), a few aphids were

0

20

40

60

80

100

29 43 57 71 85 101

0

10

20

30

40

50

60

70

29 50 63 91

No.

of n

on-w

inge

d ap

hids

/pla

nt

0

0.2

0.4

0.6

0.8

1

1.2

1.4

39 55 67 82

0

10

20

30

40

39 50 63 75 89 102

Days after sowing

(a)

(b)

(c)

(d)


occasionally found, predominantly winged L. erysimi. Only fewaphids were caught flying over the experiment before 100 DAS,but the numbers caught increased rapidly thereafter (Table 1).Only few colonising aphids (M. persicae and L. erysime) werecaught. Larger numbers of non-colonising species were caught:Acyrthosiphon kondoi (blue-green aphid), Rhopalosiphum padi(oat aphid), and Brachycaudus rumexicolens (dock aphid).

As in Expt 2, in Expt 3, M. persicae and L. erysimi bothcolonised the plots. However, M. persicae was the predominantaphid species; L. erysimi was only present from 57 DAS onwardsat <1–5/plant sample and no B. brassicae were found. Fromthe first count at 29 DAS until 85 DAS when aphid numbersreached 100/plant sample, the treatment with infector plants butno insecticide had significantly larger numbers of M. persicaethan all the others (Fig. 1c, Table 2). During the same period,the numbers of M. persicae in the treatments without infectorplants with insecticide, with infector plants with seed dressingalone, and with infector plants and both seed dressing andfoliar insecticide were not significantly different from eachother, but were significantly smaller than those in the treatmentswith infector plants and foliar spays alone, with infector plantswithout insecticide, and without infector plants or insecticide.However, by the final assessment date at 101 DAS there wereno longer any significant differences in aphid numbers amongtreatments (10–21 aphids/plant sample).

In Expt 4, again M. persicae predominated throughout, withL. erysimi present in low numbers (<1/plant sample) from63 DAS onwards and no B. brassicae being found. From thefirst count at 39 DAS until the final one at 102 DAS, whennumbers of M. persicae reached 36/plant sample, the treatmentwith infector plants but without insecticide had the largestnumber of non-winged M. persicae present overall, but onlyonce were the numbers for each count significantly differentfrom those in the treatment with infector plants and foliarinsecticide (Table 2, Fig. 1d). The treatment without infectorplants with insecticide always had the smallest number ofnon-winged M. persicae/plant, but these values were neversignificantly different from those in the treatments with infectorplants and seed dressing alone or with infector plants and both

seed dressing and foliar sprays. Between 63 and 102 DAS, therewere significantly fewer aphids in the treatments with infectorplants with seed dressing alone and with infector plants, seeddressing, and foliar sprays than in the treatments with infectorplants and no insecticide or with infector plants and foliarsprays alone.

BWYV spread

Experiment 1

Plots with infector plants that were left unsprayed up untilblanket spraying soon developed expanding patches of stunted,poorer growing plants, showing symptoms of leaf reddening,chlorosis, and premature senescence of lower leaves. As theplants grew upwards, these reddening symptoms spread tosome middle leaves and affected plants remained short withpale leaves, ‘spindly’ growth, and reduced branching andflowering; plot appearance remained noticeably uneven andstunted throughout. In contrast, plots sprayed repeatedly withinsecticide were noticeably taller and more even with lush-growing plants, mostly large dark green lower leaves, and betterbranching and flowering. Plots without infector plants that wereleft unsprayed up until blanket spraying were intermediate inappearance.

In the treatment with infector plants without insecticide upuntil blanket spraying there was very rapid spread of BWYV toyoung seedlings such that 68% of plants were already infectedwith BWYV by first assessment at 55 DAS (Fig. 2a). In contrast,BWYV spread was much slower in the other treatments, reachingonly 0–10% at this stage. Final BWYV incidence was 96%in the treatment with infector plants without insecticide upuntil blanket spraying but only 10% in the treatment withoutinfector plants with regular insecticide application. The differentfinal % BWYV incidence values were all significantly differentfrom each other (Table 3). AUPPC values for the 2 treatmentswith regular insecticide applications were not significantlydifferent from each other. However, these AUPPC values weresignificantly different from those for the 2 treatments withoutinsecticide until blanket spraying. Also, the respective AUPPC

Table 1. Numbers of different aphid species caught flying over Expt 2Mp, Myzus persicae; Le, Lipaphis erysimi; Bb, Brevicoryne brassicae; Ak, Acyrthosiphon kondoi; Rp, Rhopalosiphum padi;

Br, Brachycaudus rumexicolens; Unid, unidentifiable; Unkn, unknown

Date trap No. of Species Total no. No./collected days out Mp Le Bb Ak Rp Br Unid Unkn Other caught week

31 May 13 0A 0 0 0 0 0 0 0 0 0 006 June 6 0 0 0 0 0 0 0 0 0 0 021 June 15 0 0 0 0 0 0 0 0 0 0 026 June 5 0 0 0 0 0 1 0 0 0 1 103 July 7 0 0 0 0 0 0 0 0 0 0 010 July 7 0 0 0 0 0 0 0 0 0 0 024 July 14 0 2 0 1 0 0 0 0 0 3 207 Aug. 14 0 2 0 0 0 0 1 0 0 3 221 Aug. 14 1 15 0 6 6 0 1 0 1 30 1504 Sept. 14 0 28 0 50 38 0 2 0 3 121 6118 Sept. 14 5 22 0 175 137 4 10 6 17 376 18803 Oct. 15 10 12 0 279 181 30 234 1 30 777 363

AFigures are numbers of winged aphids caught on 2 sticky traps placed at opposite ends of the field experiment.


Table 2. Statistical analysis of data for numbers of M. persicae colonising the different treatments in Expts 3 and 4Treatment foliar insecticide applications were at 35, 62, and 92 (Expt 3) and 40, 68, 96, and 124 (Expt 4) days after sowing; a.e., after seedling emergence.Counts are mean numbers of M. persicae (non-winged adult and nymphs) from the top 10 cm of one shoot tip and one old leaf from each of 25–50 plants/plot.

Mean aphid numbers followed by the same letter are not significantly different at P < 0.05. n.s., Not significant

Treatments Assessment time (days after sowing)Expt 3 29 43 57 71 85 101

Foliar sprays at 3 and 7 wks (a.e.) + infectors 0.8b 0.04a 1.3ab 6.4a 24.1b 21.4Seed dressing + infectors 0.1a 0.005a 1.7ab 2.1a 11.8a 17.3Seed dressing + foliar sprays at 3 and 7 wks (a.e.) + infectors 0.005a 0a 0.2a 0.8a 7.4a 11.0Untreated + infectors 1.1b 0.41b 6.8c 25.5c 100d 16.4Untreated – infectors 0.1a 0.09a 3.4b 18.6b 28.5c 20.1Seed dressing + foliar sprays at 3, 7 and 11 wks (a.e.) – infectors 0.005a 0.005a 0.2a 0.9a 7.8a 9.7P <0.001 0.03 <0.001 <0.001 <0.001 n.s.l.s.d. (d.f. = 15) 0.43 0.25 2.47 6.42 8.3

Expt 4 39 50 63 75 89 102

Foliar sprays at 3 and 7 wks (a.e.) + infectors 1.3b 0.6a 5.1bc 3.5bc 38.4b 30.7bSeed dressing + infectors 0.08a 0.05a 0.1a 0.4a 1.5a 9.9aSeed dressing + foliar sprays at 3 and 7 wks (a.e.) + infectors 0.07a 0.005a 0.1a 0.06a 1.5a 6.7aUntreated + infectors 1.4b 1.7b 8.1c 5.2c 35.9b 36.0bUntreated – infectors 0.2a 0.3a 2.0ab 2.5b 28.7b 31.9bSeed dressing + foliar sprays at 3, 7, 11 and 15 wks (a.e.) – infectors 0a 0a 0.8a 0.04a 0.6a 5.3aP 0.009 <0.001 0.002 <0.001 0.009 <0.001l.s.d. (d.f. = 15) 0.92 0.72 3.78 1.74 25.75 8.24

values for the latter 2 treatments were significantly different fromeach other.

Experiment 2

Obvious large patches of stunted plants with reddening andearly senescence of lower leaves soon appeared in plots withinfector plants, contrasting with lush, taller, even growth inplots without infector plants. Thereafter, the plots with infectorplants and 1 or no sprays before blanket spraying remainedshorter and more uneven with poorer growing, patchy areas.The affected plants in these patches showed reddening, chlorosis,and premature senescence of lower leaves and some reddeningand pallor of middle leaves, and were shorter with more spindlygrowth, smaller leaves, and fewer branches and flowers. Incontrast, although some plants developed such symptoms inthem, the plots without infector plants that received insecticidesprays before blanket spraying remained more even and tallerwith lush growth. Plots of the treatments with infector plantsand 3 sprays, or without infector plants and insecticide beforeblanket spraying, were more intermediate in appearance.

BWYV spread very rapidly to young seedlings in plots ofthe 3 treatments with infector plants, reaching incidences of48–58% on the first sampling date (50 DAS) (Fig. 2b). Incontrast, at this stage in plots without infector plants, eitherno virus was detected or the incidence found was only 1%.In treatments with infector plants, final BWYV incidence was100% in plots without insecticide or that lacked the 2 earlierfoliar applications and 78% where all sprays were applied.BWYV spread was slower in the 3 treatments lacking infectorplants, reaching 4–36% at final assessment (Table 3). In the3 treatments with infector plants, the AUPPC and % finalincidence values for the treatment that received all sprays wereboth significantly smaller than those for the other 2 treatments,which had mean values that were not significantly different from

each other. However, these values were all significantly largerthan the respective values for the 3 treatments without infectorplants. Among the latter, the % final incidence values for the3 treatments were significantly different from each other buttheir AUPPC values were not.

Experiment 3

In plots of the 3 treatments without insecticidal seed dressing,BWYV spread very quickly in young B. napus plants, causingsymptoms resembling those seen in Expts 1 and 2. Patchesof BWYV-infected plants within plots without seed dressingsometimes also showed direct feeding damage due to M. persicae(severe stunting and leaf down-curling), sometimes resulting inplant death. The seed dressing dramatically increased growth,vigour, and height of the B. napus plants, whereas the foliar sprayapplied alone did not. In the treatments with infector plants withno insecticide or with foliar sprays alone, substantial BYWVinfection was detected by the second and third sampling dates (43and 57 DAS) (Fig. 2c). Thereafter, BWYV spread more rapidlyin these treatments and in the treatment without infector plantsor insecticides than in the 3 treatments with the insecticidal seeddressing. In the 3 treatments without seed dressings, BWYVinfection eventually reached 92–98%, and the values for AUPPCand % final assessment were all significantly larger than thecorresponding values for the 3 treatments with seed dressing,including those for the seed dressing applied without foliarsprays (Table 4). With AUPPC and the % final assessment,within each of these 2 categories (i.e. within the treatmentswith or without seed dressing), there were also some significantdifferences between the values for the individual treatments.

Experiment 4

As in Expt 3, in plots of the 3 treatments without insecticidalseed dressing, BWYV spread very quickly in young B. napus


plants, causing obvious symptoms resembling those seen inExpts 1 and 2. Patches of BWYV-infected plants within plots,without seed dressing, also often showed direct-feeding damagedue to M. persicae (severe stunting and leaf down-curling),sometimes resulting in plant death. Again, the seed dressingdramatically increased growth, vigour, and height of the B. napusplants, whereas the foliar spray applied alone did not. In thetreatments with infector plants with foliar sprays alone or noinsecticide, substantial BYWV infection was detected by thefirst and second sampling dates (39 and 50 DAS) (Fig. 2d). As

0

20

40

60

80

100

55 67 82 95 109 123

0

20

40

60

80

100

50 63 78 91 105 117 133

0

20

40

60

80

100

29 43 57 71 85

% P

lant

s w

ith B

WY

V

0

20

40

60

80

100

39 50 63 75 89

Days after sowing

(a)

(b)

(c)

(d )

Table 3. Statistical analysis of percentage Beet western yellows virusincidence and yield data from Expts 1 and 2

AUPPC, Area under the pathogen progress curve; a.e., after seedlingemergence. Values followed by the same letter are not significantly differentat P < 0.05. Values in parentheses show angular transformations upon whichthe statistical analyses are based. Values in square brackets show yield as apercentage of that within plots with insecticide seed dressing and 3 foliar

sprays without infector plants

Treatment AUPPC % Final BWYV Yieldincidence (t/ha)

Expt 1Foliar sprays + infectors 850a 25 (30.1)b 2.42 [8]cUnsprayed + infectors 5445c 96 (78.9)d 1.42 [46]aUnsprayed – infectors 2741b 76 (60.5)c 2.14 [20]bFoliar sprays – infectors 198a 10 (18.4)a 2.62dP <0.001 <0.001 <0.001l.s.d. 910.6 7.75 0.131d.f. 21 21 10

Expt 2Sprayed 3 times + infectors 5288b 78 (62.2)d 1.11 [26]bSprayed at 8 weeks (a.e.) 7372c 100 (88.1)e 0.95 [37]a

+ infectorsUnsprayed + infectors 7423c 100 (86.0)e 0.96 [36]aUnsprayed – infectors 1216a 36 (37.2)c 1.36 [10]cSprayed at 8 weeks (a.e.) 539a 23 (29.1)b 1.37 [9]c

– infectorsSprayed 3 times – infectors 126a 4 (11.9)a 1.51dP <0.001 <0.001 <0.001l.s.d. 1265 5.27 0.074d.f. 23 23 23

in Expt 3, BWYV spread more rapidly in these treatments and inthe treatment without infector plants or insecticides than in the3 treatments with insecticidal seed dressings. For the 3 treatments

Fig. 2. Incidence of BWYV in the different treatments in each experiment.(a) Expt 1: �, with infector plants (IP) and insecticide; �, with IP without(w/o) insecticide; N , w/o IP or insecticide;•, w/o IP with insecticide. The6 fortnightly treatment foliar insecticide applications started at 9 DAS andfinished at 81 DAS. Blanket insecticide sprays to all plots at 93 and 106 DAS.(b) Expt 2: �, with IP and insecticide sprayed at emergence, then at 4 and8 weeks after emergence; �, with IP and insecticide sprayed at 8 weeksafter emergence; N , with IP w/o insecticide; (×), w/o IP or insecticide; ∗,w/o IP with insecticide sprayed at emergence, then at 4 and 8 weeks afteremergence;•, w/o IP with insecticide sprayed at 8 weeks after emergence.The 3 treatment foliar insecticide applications at 11, 39, and 67 DAS. The4 fortnightly blanket insecticide sprays to all plots commenced at 91 DAS.(c) Expt 3: �, with IP and foliar insecticide sprayed at 3 and 7 weeks afteremergence; �, with IP and insecticide seed dressing used alone; N , withIP, insecticide seed dressing, and foliar insecticide sprayed at 3 and 7 weeksafter emergence; (×), with IP w/o insecticide; (◦), w/o IP or insecticide;•,w/o IP, with insecticide seed dressing, and foliar insecticide sprayed at 3, 7,and 11 weeks after emergence. Treatment foliar insecticide applications at35, 62, and 92 DAS. (d) Expt 4: �, with IP and foliar insecticide sprayed at3 and 7 weeks after emergence; �, with IP and insecticide seed dressingused alone; N , with IP, insecticide seed dressing, and foliar insecticidesprayed at 3 and 7 weeks after emergence; (×), with IP w/o insecticide; (◦),w/o IP or insecticide; •, w/o IP, with insecticide seed dressing, and foliarinsecticide sprayed at 3, 7, 11, and 15 weeks after emergence. Treatmentfoliar insecticide applications at 40, 68, 96, and 124 DAS.


Table 4. Statistical analysis of percentage Beet western yellows virus incidence and yield data from Expts 3 and 4AUPPC, Area under the pathogen progress curve; a.e., after seedling emergence. Values followed by the same letter are not significantly different at P < 0.05.Values in parentheses show angular transformations upon which the statistical analyses are based. Values in square brackets show yield as a percentage of

that within plots with insecticide seed dressing and all foliar sprays without infector plants

Treatment Expt 3 Expt 4AUPPC % Final BWYV Yield AUPPC % Final BWYV Yield

incidence (t/ha) incidence (t/ha)

Foliar sprays at 3 and 7 wks (a.e.) + infectors 2740c 92 (73.3)c 0.93 [26]abc 3536d 96 (78.8)d 0.86 [39]abSeed dressing + infectors 1320b 72 (58.1)b 1.14 [9]bc 1252b 35 (36.3)b 1.43 [0]cSeed dressing + foliar sprays at 3 and 7 wks 1120ab 48 (43.8)a 1.17 [6]c 1091ab 27 (31.4)ab 1.28 [9]bc

(a.e.) + infectorsNo insecticide + infectors 3356d 98 (82.5)c 0.62 [50]a 3646d 97 (79.8)d 0.76 [46]aNo insecticide – infectors 2418c 97(80.5)c 0.79 [37]ab 2383c 84 (66.1)c 1.13 [20]abcSeed dressing + 3–4 foliar sprays 749a 31 (33.9)a 1.25c 617a 17 (24.3)a 1.41c

(a.e.) – infectorsP <0.001 <0.001 0.01 <0.001 <0.001 0.006l.s.d. 498.2 (13.53) 0.352 494.8 (9.77) 0.427d.f. 15 14 18 30 30 30

without seed dressings, BWYV infection eventually reached84–97%, and the values for AUPPC and % final assessmentwere all significantly larger than the corresponding values forthe 3 treatments with seed dressing, including those for the seeddressing applied without foliar sprays (Table 4). With AUPPCand the % final assessment, within each of these 2 categories(i.e. within the treatments with or without seed dressing), therewere also some significant differences between the values for theindividual treatments.

Effect on seed yield

Experiment 1

The mean seed yields for each treatment were all significantlydifferent from one another. They were 46% and 20% lower,respectively, in plots with infector plants without insecticidebefore blanket spraying and plots without infector plants orinsecticide before blanket spraying than in the control treatmentwithout infector plants with regular insecticide application(Table 3). Early aphid numbers before plots were blanket sprayedwere insufficient (<1.5/plant sample in plots without infectorplants or insecticide) to have contributed directly to the seed yieldlosses (Fig. 1a). When data for the individual treatment meansfor the relationship between final % BWYV incidence and seedyield (t/ha) were plotted, 96% of the variation was explainedby % BWYV incidence (y = –0.0122x + 2.7582, R2 = 0.96,P = 0.02, SEEy = 0.12; Fig. 3a). When mean yield gap datawere substituted for mean seed yields, 95% of the variationwas explained by % BWYV incidence (y = 0.0123x + 0.2299,R2 = 0.95, P = 0.02, SEEy = 0.13; Fig. 3b), and for each 1%increase in virus incidence there was a yield decline of12.3 kg/ha.

Experiment 2

The mean seed yields for the treatments with all sprays withor without infector plants were significantly different from eachother and from those of all the other treatments (Table 3). Thecorresponding values for the other 2 treatments with infectorplants were not significantly different from each other, but weresignificantly smaller than the values for all the other treatments.

Their yields were 36–37% smaller than those for the controlplots without infector plants with all sprays. Yields for theremaining 2 treatments without infector plants did not differsignificantly from each other but were significantly differentfrom those of all other treatments. They were 9–10% smallerthan that of the treatment with all sprays without infectorplants. Early aphid numbers in June and July before plots wereblanket sprayed were insufficient (<15/plant sample in plotswith infector plants without insecticide) to have contributeddirectly to the yield losses (Table 2, Fig. 1b). When data forthe individual treatment means for the relationship betweenfinal % BWYV incidence and seed yield (t/ha) were plotted,97% of the variation in yield was explained by BWYV incidence(y = –0.006x + 1.5377, R2 = 0.97, P < 0.001, SEEy = 0.05;Fig. 3c). When mean yield gap data were substituted for meanseed yields, 96% of the variation in yield gaps was explained byBWYV incidence (y = 0.006x + 0.1114, R2 = 0.96, P < 0.001,SEEy = 0.05; Fig. 3d), and for each 1% increase in BWYVincidence there was a yield decline of 6.0 kg/ha.

Experiment 3

The seed yield for the treatment with infector plants withoutinsecticide was significantly smaller than each of those forthe 3 treatments with seed dressings, but not from those forthe treatment with infector plants and foliar sprays alone, orthe treatment without infector plants or insecticide (Table 4).However, the yield for the control treatment without infectorplants with seed dressing and foliar sprays was not significantlygreater than those for the 2 treatments with infector plants andseed dressing or from the treatment with infector plants andfoliar sprays alone, the yields of which were not significantlydifferent from each other. When the seed yield for the treatmentwithout infector plants with seed dressing and foliar sprays wascompared with those for the 2 treatments without insecticide, theyield losses were 50% (with infector plants) and 37% (withoutinfector plants). Plots with infector plants that received the seeddressing alone or the seed dressing followed by 2 foliar spraysof insecticide suffered only 6–9% yield losses and their yieldswere not significantly different from each other.


y = 0.0123x + 0.2299R 2 = 0.95

0

0.4

0.8

1.2

1.6

Yie

ld g

ap (

t/ha)

y = –0.0122x + 2.7582R 2 = 0.96

0

0.5

1

1.5

2

2.5

3

y = –0.006x + 1.5377R 2

= 0.970

0.4

0.8

1.2

1.6

y = 0.006x + 0.1114R 2 = 0.96

0

0.2

0.4

0.6

0.8

0 20 40 60 80 100

(a)

(b)

(c)

(d)

Yie

ld (

t/ha)

Yie

ld g

ap (

t/ha)

Yie

ld (

t/ha)

% Plants with BWYV

Fig. 3. Relationships between yields and yield gaps in t/ha, and finalpercentage BWYV incidence using experimental treatment means: (a) yieldand (b) yield gaps in Expt 1; (c) yield and (d) yield gaps in Expt 2.

Experiment 4

The seed yield for the treatment with infector plants andno insecticide was significantly smaller than those for the3 treatments with seed dressings, but not from those of thetreatment without infector plants or insecticide, or the treatmentwith infector plants and foliar sprays alone (Table 4). Theyield for the treatment with infector plants and foliar spraysalone was significantly smaller than those for the treatmentwith infector plants that received seed dressing alone andthe treatment without infector plants that received both seed

dressing and foliar sprays. When the seed yields for thecontrol treatment without infector plants with seed dressingand foliar sprays was compared with those for the treatmentswithout insecticide, the yield loss was 46% (with infectorplants) and 20% (without infector plants). Plots with infectorplants that received the seed dressing suffered no yield lossand the yield loss of 9% was not statistically significantfor plots with infector plants that received both types ofinsecticide treatments.

Effect on 1000-seed weight, oil, protein, and erucicacid contents

In Expts 1 and 2, the mean 1000-seed weight values forthe B. napus samples for the different treatments were2.77–2.95 and 3.19–3.58 g, respectively. Regression analysisusing individual treatment means for 1000-seed weight and final% virus incidence did not reveal any significant change in thesevalues due to BWYV infection in Expt 1. However, in Expt 2an up to 11% increase in seed weight due to BWYV infectionwas indicated when data for the individual treatment meanswere used to plot the relationship between 1000-seed weightand percentage final BWYV incidence; 73% of the variationwas explained by BWYV incidence (y = 0.0045x + 3.0864,R2 = 0.73, P = 0.04, SEEy = 0.0935).

For Expt 1, the mean treatment values for the seedmoisture, oil, protein, and erucic acid contents of the B. napusseed samples were 6.3%, 44.0–45.4%, 21.2–22.6%, and0.274–0.299%, respectively. In Expt 2, the corresponding figureswere 5.9%, 46.0–47.3%, 18.7–20.9%, and 0.178–0.318%.Regression analysis using final % virus incidence data revealedstatistically significant changes in these values due to BWYVinfection for protein content (Expts 1 and 2), oil content (Expt 2only), and erucic acid (Expt 2 only). An up to 6% (Expt 1) and11% (Expt 2) increase in seed protein values with increasingvirus incidence was indicated when data for the individualtreatment means were used to plot the relationship betweenpecentage final BWYV incidence and protein content: 97%(Expt 1) and 84% (Expt 2) of the variation was explainedby BWYV incidence (Expt 1: y = 0.0157x + 21.02, R2 = 0.97,P = 0.01, SEEy = 0.13; Expt 2: y = 0.019x + 19.094, R2 = 0.84,P < 0.001, SEEy = 0.37; Fig. 4a, b). In Expt 2, an up to3% decrease in oil content with increasing virus incidencewas indicated when data for the individual treatment meanswere used to plot the relationship between percentage finalBWYV incidence and oil content: 92% of the variationwas explained by BWYV incidence (y = –0.0126x + 47.223,R2 = 0.92, P = 0.002, SEEy = 0.16; Fig. 4c). Similarly, inExpt 2, an up to 44% increase in erucic acid content withincreasing virus incidence was indicated when data for theindividual treatment means were used to plot the relationshipbetween percentage final BWYV incidence and oil content:70% of the variation was explained by BWYV incidence(y = 0.0016x + 0.1713, R2 = 0.70, P = 0.04, SEEy = 0.0473;Fig. 4d).

Discussion

This research demonstrates that, when aphids introduce it earlyand infection becomes widespread in young plants, BWYVcauses substantial losses in seed yields of B. napus stands


y = 0.0157x + 21.02R 2 = 0.97

20.8

21.2

21.6

22

22.4

22.8

y = 0.019x + 19.094R 2 = 0.84

18.5

19

19.5

20

20.5

21

21.5

% P

rote

in c

onte

nt

y = –0.0126x + 47.223R 2 = 0.92

45.6

46

46.4

46.8

47.2

47.6

% O

il co

nten

t

y = 0.0016x + 0.1713R 2 = 0.70

0

0.1

0.2

0.3

0.4

0 20 40 60 80 100

% Plants with BWYV

% E

ruci

c ac

id c

onte

nt

(a)

(b)

(c)

(d)

Fig. 4. Relationships between percentage seed protein, oil, and erucicacid contents and final percentage BWYV incidence using experimentaltreatment means: (a) protein content in Expt 1, (b) protein, (c) oil, and (d)erucic acid contents in Expt 2.

growing in the field in the south-western Australian grainbelt.BWYV therefore has considerable yield-limiting potential underAustralian conditions. Indeed, the magnitude of the yield lossesfound in the absence of any direct aphid feeding damage (up to46%) was greater than that reported previously under northernEuropean conditions in autumn-sown B. napus plots (up to 34%;Graichen 1995, 1997; Graichen and Schliephake 1999), wherethe winter growing period is much colder and the growing periodis considerably longer. Our studies also revealed that for B. napusseed, although overall yield and oil content declines, individualseed weight, protein content, and erucic acid levels increasewith increasing BWYV incidence. Decreased oil content andincreased erucic acid levels both diminish B. napus seed quality.In addition, dressing seed with imidacloprid insecticide such that

the active ingredient reached all seeds sown, provided effectiveearly control of M. persicae and BWYV infection, resulting inconsiderable yield increases.

BWYV spread very quickly within populations of youngB. napus plants, causing symptoms of leaf purpling andpronounced plant stunting, which were absent in healthy plants.The affected plants then grew less vigorously, producing fewershoots, flowers, and seeds than healthy ones, the purplingsymptoms persisting on their lower leaves as they matured.Early-infected plots were noticeably dwarfed and uneven inappearance compared with the lush, even growth of plotsin which, due to insecticide application and absence ofBWYV inoculum introduced via infector plants, few plantswere infected. When plants became infected with BWYV atlater growth stages, symptoms were milder or infection wassymptomless. The different types of insecticide applicationsused (seed dressing and foliar) and range of timings of theapplications resulted in a wide range of final BWYV incidencesin the different treatments. Rapid spread of BWYV at the mostvulnerable early crop growth stage caused the greatest losses inseed yield.

In Expts 1 and 2, our experimental procedure excluded anypossibility that direct feeding damage by M. persicae (or otheraphid species) might have contributed to these losses. This isbecause repeated blanket insecticide sprays were applied acrossall plots well before aphid numbers built up sufficiently to havecontributed directly to the yield losses obtained. In the region,colonies consisting of >200 aphids/1-cm length of buddingtip or flowering head provide a threshold number above whichapplication of insecticides is considered economically viableto prevent yield loss arising from direct aphid feeding damage(Berlandier 1999). However, the M. persicae numbers/plantsample in plots without insecticide before blanket sprayingnever exceeded 1.3 (Expt 1) or 62 (Expt 2), and these samplesalso included a lower leaf in addition to a 10-cm (not 1-cm)-longgrowing tip. Moreover, when data for the individual treatmentmeans for the relationship between final percentage BWYVincidence and seed yield (t/ha) were plotted together, only 4%(Expt 1) or 3% (Expt 2) of the variation was explained byfactors other than BWYV incidence. In Expts 3 and 4, whereno late blanket insecticide sprays were applied to prevent directaphid feeding damage, it was impossible to know to what extentfeeding also contributed to their losses. Such direct feedingdamage was evident as patches of very stunted BWYV-infectedplants heavily colonised by M. persicae and showing severestunting and leaf down-curling within plots of both experiments,and the seed yield losses reached 50% in Expt 3 and 46% inExpt 4. However, its effect was probably considerably smallerthan that of BWYV infection as these severely stunted patcheswere limited in extent within plots and the losses induced byBWYV alone reached 46% in Expt 1 and 37% in Expt 2, whichwas at the same site as Expt 3. In Expts 1 and 2, as seed size(measured as 1000-seed weight) was not decreased by infection,the seed yield losses were all due to fewer seeds developingon BWYV-infected plants. When yield gaps were calculated,they were 460–1080 kg/ha (Expt 1) and 40–541 kg/ha (Expt 2)and, for each 1% increase in incidence of BWYV in B. napusplants, there was a yield decrease of 12 kg/ha (Expt 1) and6 kg/ha (Expt 2).


Although our field experiments involved only one B. napuscultivar (Pinnacle), we do not believe that its sensitivity toBWYV infection is atypical for most Australian B. napuscultivars as, when 2 others (cvv. Grace and Hyden) were usedin field experiments with BWYV in 2003–05, they developedsimilar foliage symptoms (Coutts and Jones 2005). Also, whena range of B. napus genotypes were evaluated in the field in 2006,only a small proportion developed milder symptoms (B. Couttsand R. Jones, unpublished). With European cultivars of B. napus,although those grown are all highly susceptible, they sometimesdiffer in sensitivity to BWYV, which in turn influences themagnitude of the yield losses suffered. For example, in Germany,when a field experiment with cvv. Falcon and Zeus was repeatedannually from 1993 to 1995, the yield losses ranged from 12 to34%, with an average yield reduction of 20%. In 1993 they were12% for both cultivars, but in the other 2 years they were greaterfor Zeus (22–34%) than for Falcon (18–19%) (Graichen 1995,1997; Graichen and Schliephake 1999).

Previous research into BWYV-induced yield losses inB. napus came from temperate climates in northern Europerather than from Mediterranean ones such as that of south-western Australia. In the former, aphid immigration into B. napuscrops normally ceases in late autumn, not starting up again untilspring. Substantial virus spread over winter is rare and, unlessconsiderable early infection occurs in autumn, the main virusinfection period is spring. Also, the growing period is muchlonger. In contrast, in our experiments, aphid activity continuedthroughout the mild winter period (June–August), allowingBWYV spread to continue. Moreover, we introduced the virusand vector to plots before seedling emergence, using infectorplants, which resulted in very rapid BWYV spread in youngplants. European researchers introduced viruliferous aphidspreviously raised on BWYV-infected plants in the glasshousedirectly into plots (e.g. Graichen and Schliephake 1999; Jay et al.1999), or transplanted BWYV-infected B. napus plants infestedwith M. persicae at their windward edges (e.g. Walsh et al.1989), both of which procedures were apparently less effectiveat introducing inoculum sources, or relied on natural BWYVspread without any introduced inoculum (e.g. Smith and Hinckes1985). The different winter temperatures, growing periods, andmethods of applying inoculum may account for the greatermagnitude of the yield losses recorded in B. napus in south-western Australia. Two additional factors potentially contributeto the greater yield losses measured in B. napus in south-western Australia: (1) the Australian ‘spring-type’ B. napusgenotype used might be more sensitive to BWYV infectionthan the ‘winter-type’ genotypes they used; (2) the newergeneration neonicotinoid insecticide we applied (imidacloprid)is likely to have been more effective at suppressing insecticide-resistant M. persicae in control plots and, in consequence,BWYV spread in them, than the older generation insecticidesthey used (carbamates and/or pyrethroids). In south-westernAustralia, M. persicae is often resistant to the latter insecticidesbut none has been reported with resistance to imidacloprid(Thackray et al. 2000a).

As mentioned above, there was no evidence of decreasedseed size resulting from BWYV infection of B. napus. Indeed,regression analysis demonstrated the opposite in Expt 2, withseed size being increased significantly by infection (up to 11%).This agrees with the earlier findings of Jay et al. (1999) who

attributed the greater seed size they found to presence of fewerseeds/pod on BWYV-infected plants, resulting in increasednutrients per developing seed being available at pod-filling time.Such a scenario might also explain why protein content of seedsincreased with BWYV infection in Expts 1 and 2 (by up to11%). Although virion accumulation within seeds (excludingthe embryo) might also explain increased protein content,this seems unlikely for a phloem-limited virus like BWYV.Increased protein content resulting from BWYV infection isnot unexpected as Thackray et al. (2005) recorded significantlyincreased seed protein content of wheat caused by infectionwith another related, phloem-limited virus, Barley yellow dwarfvirus (BYDV, family Luteoviridae, genus Luteovirus). Jay et al.(1999) also reported increased glucosinolate levels in B. napusseeds due to BWYV infection. Although we did not assay forglucosinolates, we are apparently the first to assay for alterederucic acid contents resulting from BWYV infection, and inExpt 2 our regression analysis demonstrated that this is increasedsignificantly in B. napus seeds (by up to 44%). We also recorded asignificant decrease in seed oil content due to BWYV infectionin Expt 2 (up to 3% less), which agrees with earlier findings(2–3% decline) with B. napus in England (Smith and Hinckes1985; Jay et al. 1999). As mentioned in the Introduction,increased contents of glucosinolates and eurcic acid anddecreased oil contents are all undesirable quality characteristicsin B. napus seed, diminishing its market value.

In the grainbelt region of south-western Australia, 2 earlyfoliar applications of the pyrethroid insecticide α-cypermethrinprovided effective control of the related, persistently aphid-transmitted virus BYDV in autumn-sown cereal crops (McKirdyand Jones 1996, 1997; Thackray et al. 2005). When 2 earlyapplications of this chemical were made to B. napus plots atsimilar foliar application rates in Expts 3 and 4, they wereineffective in controlling M. persicae or BWYV. This is due tothe widespread occurrence of insecticide-resistant M. persicae inthe region (Thackray et al. 2000a). In contrast, in the same 2 fieldexperiments, imidacloprid seed dressing used alone and appliedat 525 g a.i./100 kg of seed effectively controlled insecticide-resistant M. persicae and suppressed spread of BWYV for2.5 months. Moreover, growth, vigour, and height of the B. napusplots were dramatically increased compared with those in plotswithout this seed dressing, and seed yield was increased by84–88%. Thus, when they arrive in B. napus crops just afteremergence and spread the virus rapidly, the early control ofM. persicae and BWYV achieved from seed dressing withimidacloprid is sufficiently thorough to increase seed yieldsdramatically. Moreover, dressing B. napus seed with it has thepotential to provide effective multi-purpose control not only ofaphid vectors, including insecticide-resistant M. persicae, andBWYV, but also of other damaging early insect and mite pestsof B. napus. However, a word of caution is required here becauselater studies involving similar field experiments to Expts 3 and4 have revealed that this level of control of M. persicae andBWYV is not obtainable unless the chemical is mixed withB. napus seeds in such a way that it reaches them all, whichcurrent commercial, large-scale methods of application are notachieving. In 2003–05, 6 field experiments (2/year) with thisseed dressing applied at the recommended rate (240 g a.i./100 kgof seed) by commercial procedures gave disappointing levelsof control of M. persicae and BWYV, not providing anything


like the magnitude of the yield increases anticipated (Coutts andJones 2005; unpublished). Examining the ability of M. persicaeto survive on seedlings grown from B. napus seed dressedwith imidacloprid commercially, revealed that the chemical wasabsent from a relatively high proportion of the treated seeds. Theaphid was able to multiply readily on the seedlings that grewfrom them but not on the others (unpublished). Of course, ourexperimental design represents a worst case early aphid-arrivalscenario, so current commercial seed-dressing procedures mightbe expected to prove more effective in less extreme scenarioswhere flights of M. persicae are delayed and less BWYV isintroduced to young crops. Such considerations, however, shouldnot delay the search for improved, large-scale commercial seed-dressing procedures for use with crops with small-sized seedssuch as those of B. napus.

This study has important practical implications for helping todevise a predictive model for BWYV and its vectors in B. napuscrops in the grainbelt. We have quantified the magnitude oflosses in seed yield and quality for the worst case scenariowhen viruliferous M. persicae arrive in large numbers in newlyemerged crops, and provided relationships between differentfinal BWYV infection incidences and seed yield, yield gaps,and seed quality parameters. This information can be combinedwith information on over-summering and within growing-season reservoirs of BWYV, and crop aphid colonisation andBWYV incidence data (Coutts and Jones 2000, 2005; Thackrayet al. 2000b, 2001, 2002; Coutts et al. 2006) to develop aforecasting and decision support system for BWYV and aphidsin B. napus such as those developed previously for BYDV incereals and Cucumber mosaic virus (family Bromoviridae, genusCucumovirus) in lupin (Thackray and Jones 2003; Thackrayet al. 2004). Such forecasting models use rainfall during summerand early autumn to calculate an index of aphid build-up ineach district before the growing season starts. The index is thenused to forecast the timing and magnitude of aphid immigrationinto the crop, the subsequent aphid build-up and movementwithin the crop, virus spread, and related seed yield and qualitylosses. Once an improved insecticidal commercial seed-dressingprocedure is achieved for B. napus, the predictive model can beincorporated into a decision support system accessible via theinternet, such as those produced for the other 2 pathosystems(www.agric.wa.gov.au/bydv; www.agric.wa.gov.au/cmv).

Acknowledgments

The authors thank Lisa Smith and Rohan Prince for technical assistance,staff at Avondale, Badgingarra, and Medina Research Stations for help withfield experiments, the Western Australian Chemistry Centre for the seederucic acid, moisture, oil, and protein analyses, Mario D’Antuono for helpwith statistical analysis, and Bayer Crop Science for the seed dressings.The Grains Research and Development Corporation supported this researchfinancially.

References

Aksouh NM, Jacobs BC, Stoddard FL, Mailer RJ (2001) Response of canolato different heat stresses. Australian Journal of Agricultural Research 52,817–824. doi: 10.1071/AR00120

Berlandier F (1999) Managing aphids in canola. Agriculture WesternAustralia Farmnote No. 45/99.

Blackman RL, Eastop VF (2000) ‘Aphids on the world’s crops–anidentification and information guide.’ (John Wiley and Sons Ltd: WestSussex, UK)

Brunt AA, Crabtree K, Dallwitz MJ, Gibbs AJ, Watson L, Zurcher EJ (Eds)(1996) ‘Viruses of plants: descriptions and lists from the VIDE database.’(CABI: Oxford, UK)

Bwye AM, Proudlove W, Berlandier FA, Jones RAC (1997) Effects ofapplying insecticides to control aphid vectors and cucumber mosaic virusin narrow-leafed lupins (Lupinus angustifolius). Australian Journal ofExperimental Agriculture 37, 93–102. doi: 10.1071/EA96041

Clark MF, Adams AN (1977) Characteristics of the microplate method ofenzyme-linked immunosorbent assay for the detection of plant viruses.Journal of General Virology 34, 475–483.

Coutts B, Jones R (2005) Controlling aphids and Beet western yellowsvirus in canola using imidacloprid seed dressing. In ‘Crop UpdatesConference. 2005 Oilseeds updates’. (Ed. D Hamilton) pp. 19–20.(Department of Agriculture, Western Australia: Perth, W. Aust.)

Coutts BA, Hawkes JR, Jones RAC (2006) Occurrence of Beet westernyellows virus and its aphid vectors in over-summering broad-leafedweeds and volunteer crop plants in the grainbelt region of south-westernAustralia. Australian Journal of Agricultural Research 57, 975–982.doi: 10.1071/AR05407

Coutts BA, Jones RAC (2000) Viruses infecting canola (Brassica napus)in south-west Australia: incidence, distribution, spread and infectionreservoir in wild radish (Raphanus raphanistrum). Australian Journalof Agricultural Research 51, 925–936. doi: 10.1071/AR00014

D’Arcy CJ, Domier LL (2005) Luteoviridae. In ‘Virus taxonomy. VIIIthReport of the International Committee on Taxonomy of Viruses’.(Eds CM Fauquet, MA Mayo, J Maniloff, U Desselberger, LA Ball)pp. 343–352. (Academic Press: New York)

Duffus JE (1972) ‘Beet western yellows virus.’ CMI/AAB Descriptions ofplant viruses, No. 89. (Association of Applied Biologists: Wellesbourne,UK)

Duffus JE, Russell GE (1972) Serological relationship between beet westernyellows and turnip yellows viruses. Phytopathology 62, 1274–1277.

Firestone D, Horwitz W (1979) IUPAC gas chromatographic methodfor determination of fatty acids composition. Collaborative study.Journal of the Association of Official Analytical Chemists 62,709–721.

Gibbs AJ, Gower JC (1960) The use of a multiple transfer method in plantvirus transmission studies—some statistical points arising in the analysisof results. Annals of Applied Biology 48, 75–83.

Graichen K (1995) Zur bedeutung von virusbefall fur den anbauvon winterraps und leindotter. Mitteilungen aus der BiologischenBundesanstalt 310, 102–108.

Graichen K (1997) Wasserrubenvergilbungsvirus – ertrags- undqualistsminderungen beim winterraps. Raps 4, 156–159.

Graichen K, Rabenstein F (1996) European isolates of beet western yellowsvirus (BWYV) from oilseed rape (Brassica napus L. ssp. napus) are non-pathogenic on sugar beet (Beta vulgaris L. var. altissima) but representisolates of turnip yellows virus (TuYV). Journal of Plant Diseases andProtection 103, 233–245.

Graichen K, Schliephake E (1999) Infestation of winter oilseed rape by turnipyellows luteovirus and its effect on yield in Germany. In ‘Proceedings of10th International Rapeseed Congress—New horizons for an old crop’.(Eds N Wratten, PA Salisbury) pp. 131–136. (International ConsultativeGroup for Rapeseed Research: Canberra, ACT)

Hauser S, Stevens M, Beuve M, Lemaire O (2002) Biological properties andmolecular characterization of beet chlorosis virus (BChV). Archives ofVirology 147, 745–762. doi: 10.1007/s007050200023

Hauser S, Stevens M, Mougel C, Smith HG, Fritsch C, Herrbach E,Lemaire O (2000) Biological, serological, and molecular variabilitysuggest three distinct polerovirus species infecting beet or rape.Phytopathology 90, 460–466. doi: 10.1094/PHYTO.2000.90.5.460

Hawkes J, Thackray D, Jones R (2002) Incidence of virus diseases inchickpea. In ‘Crop Updates Conference. 2002 Pulse research andindustry development in Western Australia’. (Eds K Regan, P White)pp. 101–102. (Department of Agriculture, Western Australia: Perth,W. Aust.)


Jay CN, Rossall S, Smith HG (1999) Effects of beet westernyellows virus on growth and yield of oilseed rape (Brassicanapus). Journal of Agricultural Science, Cambridge 133, 131–139.doi: 10.1017/S0021859699006711

Jones RAC (2004) Occurrence of virus infection in seed stocks and3-year-old pastures of lucerne (Medicago sativa). Australian Journalof Agricultural Research 55, 757–764. doi: 10.1071/AR04011

Latham LJ, Jones RAC (2001) Incidence of virus infection in experimentalplots, commercial crops and seed stocks of cool season croplegumes. Australian Journal of Agricultural Research 52, 397–413.doi: 10.1071/AR00079

McKirdy SJ, Jones RAC (1996) Use of imidacloprid and newer generationsynthetic pyrethroids to control the spread of barley yellow dwarfluteovirus in cereals. Plant Disease 80, 895–901.

McKirdy SJ, Jones RAC (1997) Effect of sowing time on barley yellowdwarf virus infection in wheat: virus incidence and grain yieldlosses. Australian Journal of Agricultural Research 48, 199–206.doi: 10.1071/A96073

McKirdy SJ, Jones RAC, Nutter FW Jr (2002) Quantification of yield lossescaused by Barley yellow dwarf virus in wheat and oats. Plant Disease86, 769–773. doi: 10.1094/PDIS.2002.86.7.769

McLean GD, Price LK (1984) Virus, viroid, mycoplasma and rickettsiadiseases of plants in Western Australia. Technical Bulletin No. 68,Department of Agriculture, Perth, W. Aust.

Salisbury PA, Potter TD, McDonald G, Green AG (Eds) (1999) Canolain Australia: the first thirty years. Organising Committee of the 10thInternational Rapeseed Congress, Canberra, ACT

Schroder M (1994) Investigations on the susceptibility of oilseedrape (Brassica napus L. spp. napus) to different virus diseases.Zeitschrift fur Pflanzenkrankheiten und Pflanzenschutz 101,576–589.

Smith HG, Barker H (Eds) (1999) ‘The Luteoviridae.’ (CABI Publishing:Wallingford, UK)

Smith HG, Hinckes JA (1985) Studies on beet western yellows virusin oilseed rape (Brassica napus ssp. oleifera) and sugar beet (Betavulgaris). Annals of Applied Biology 107, 473–484.

Thackray D, Hawkes J, Jones R (2000b) Forecasting aphid and virus riskin canola. In ‘Crop Updates Conference. 2000 Oilseeds updates’.(Ed. A Cox) pp. 36–37. (Agriculture Western Australia: Perth, W. Aust.)

Thackray D, Hawkes J, Jones R (2001) Further developments in forecastingaphid and virus risk in canola. In ‘Crop Updates Conference. 2001Oilseeds updates’. (Ed. C Zaicou-Kunesch) pp. 55–57. (AgricultureWestern Australia: Perth, W. Aust.)

Thackray D, Hawkes J, Jones R (2002) Influence of climate on aphidoutbreaks and virus epidemics in canola. In ‘Crop Updates Conference.2002 Oilseeds updates’. (Ed. D Ecksteen) pp. 44–46. (Department ofAgriculture, Western Australia: Perth, W. Aust.)

Thackray DJ, Diggle AJ, Berlandier FA, Jones RAC (2004) Forecastingaphid outbreaks and epidemics of Cucumber mosaic virus in lupincrops in a Mediterranean-type environment. Virus Research 100,67–82. doi: 10.1016/j.virusres.2003.12.015

Thackray DJ, Jones RAC (2003) Forecasting aphid outbreaks andepidemics of Barley yellow dwarf virus – a decision support systemfor a Mediterranean-type climate. Australasian Plant Pathology 32,438[Abstr.].

Thackray DJ, Jones RAC, Bwye AM, Coutts BA (2000a) Further studieson the effects of insecticides on aphid vector numbers and spread ofcucumber mosaic virus in narrow-leafed lupins (Lupinus angustifolius).Crop Protection 19, 121–139. doi: 10.1016/S0261-2194(99)00093-9

Thackray DJ, Ward LT, Thomas-Carroll ML, Jones RAC (2005) Role ofwinter-active aphids spreading Barley yellow dwarf virus in decreasingwheat yields in a Mediterranean-type environment. Australian Journalof Agricultural Research 56, 1089–1099. doi: 10.1071/AR05048

Walsh JA, Perrin RM, Miller A, Laycock DS (1989) Studies on beet westernyellows virus in winter oilseed rape (Brassica napus spp. oleifera) andthe effect of the insecticidal treatment on its spread. Crop Protection 8,137–143.

Walsh JA, Tomlinson JA (1985) Viruses infecting winter oilseed rape(Brassica napus ssp. oleifera). Annals of Applied Biology 107, 485–495.

Manuscript received 11 December 2006, accepted 2 May 2007

http://www.publish.csiro.au/journals/ajar

Documents

Yield-limiting potential of Beet western yellows virus in Brassica napus