Effect of blackleg and sclerotinia stem rot on canola yield in the high rainfall zone of southern New South Wales, Australia

CSIRO PUBLISHING

www.publish.csiro.au/journals/ajar Australian Journal of Agricultural Research, 2006, 57, 201–212

Effect of blackleg and sclerotinia stem rot on canola yield in the highrainfall zone of southern New South Wales, Australia

John A. KirkegaardA,E, Michael J. RobertsonB, Peter HamblinC, and Susan J. SpragueA,D

ACSIRO Plant Industry, GPO Box 1600, Canberra, ACT 2601, Australia.BCSIRO Sustainable Ecosystems/APSRU, Queensland Biosciences Precinct, 306 Carmody Rd,

St Lucia, Qld 4067, Australia.CAgritech Crop Research, PO Box 678, Young, NSW 2594, Australia.

DSchool of Botany, The University of Melbourne, Vic. 3010, Australia.ECorresponding author. Email: [email protected]

Abstract. Canola (Brassica napus) yields have been declining in the medium–high rainfall (550–650 mm) areas ofsouthern NSW during the 1990s. This decline is not related to seasonal conditions and is not evident in wheat. Leadinggrowers and consultants suspect that increasing levels of the diseases blackleg, caused by Leptosphaeria maculans,and sclerotinia stem rot, caused by Sclerotinia sclerotiorum, may be responsible. We investigated the effect ofthese diseases on otherwise well-managed canola crops at 2 sites in southern NSW from 2001 to 2003. Blacklegwas present at both sites in all 3 years, with 8–10% of plants of variety Rainbow lodged in the untreated plots atGalong in all years, and 1–10% at Wallendbeen. Sclerotinia stem rot was predominately confined to 2001 when7–42% of plants were infected in the untreated plots at the sites. Individually, or in combination, the diseases causedsignificant yield losses at both sites in every year of the study, ranging from 0.39 t/ha to 1.54 t/ha. These findingssuggest that for each 1% reduction in the incidence of blackleg lodging, yield was increased by 5%, whereas each1% reduction in plants infected with sclerotinia stem rot increased yield by 1.3%. The ability of the crops to yieldto their simulated water and N-limited potential when these diseases were either absent or controlled, indicatesthat disease may be the key yield-limiting factor in the area. Cost-effective control strategies for both diseases arecritical if canola is to remain an important and profitable break crop in the farming system in the high-rainfall zoneof southern NSW.

Additional keywords: Leptosphaeria maculans, Sclerotinia sclerotiorum, Brassica napus, Phoma stem canker.

IntroductionCanola (Brassica napus) production in Australia expandedrapidly during the 1990s and the area had stabilised ataround 1.6 M ha by the end of the decade. The expansionfollowed the release of better adapted varieties resistantto the major disease blackleg, caused by the fungusLeptosphaeria maculans (Desm) Ces. et de Not., whichdevastated the industry during the 1970s. The new varietieswere profitable as crops in their own right, but also providedan excellent break crop for the control of diseases and weedsin cereal rotations (Norton et al. 1999). The inclusion ofcanola in the rotation in the medium- and high-rainfall areasof the Australian wheatbelt has underpinned unprecedentedwheat yield improvements in many areas during the 1990s(Kirkegaard et al. 2004).

During the late 1990s, there was growing concern fromseveral sources that the yield of canola was declining inthe medium- and high-rainfall areas of southern NSW.

Evidence to support these concerns came from a compilationand analysis of paddock-based survey data (CanolaCheck)collected from central and southern NSW from 1991 to 2001(around 150 paddocks/year across the area), which reportedan estimated average yield decline of 9% for dryland canolacrops in the area (Mead et al. 2004). A similar trend wasalso evident in the more extensive census and survey datacollected by the Australian Bureau of Statistics (ABS 2005)for the same period. Further analysis of the ABS datasuggests that the trend cannot simply be attributed to poorseasonal conditions. Canola yield decline was clear when acomposite of the Harden, Young, and Cootamundra statisticallocal areas (SLAs) was made (Fig. 1a). In contrast thewheat yield trend was flatter over the same period. Thedecline in canola yield relative to wheat, particularlythe shift after 1998, is apparent, the yield ratio falling fromaround 0.60 to remain around 0.46 after 1998. This isbelow the long-term simulated mean ratio of 0.51 reported

© CSIRO 2006 10.1071/AR05179 0004-9409/06/020201

202 Australian Journal of Agricultural Research J. A. Kirkegaard et al.

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001

Yie

ld (

t/ha)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Can

ola:

whe

at y

ield

rat

io

0.4

0.5

0.6

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001

Can

ola

yiel

d (t

/ha)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Potential yield (t/ha)

0 1 2 3 4

Yie

ld g

ap (

t/ha)

0.0

0.5

1.0

1.5

2.0

(a)

(b)

(c)

Fig. 1. Trends in (a) yield of canola (�) and wheat (◦) and thecanola : wheat yield ratio (•), and (b) comparisons of measured yield(N ) and simulated potential yield (•) for the composite of Harden,Young, and Cootamundra statistical local areas (SLAs) from 1992to 2001. The relationship between potential yield and the yield gap(potential yield – measured yield) is shown in (c). Yield data obtainedfrom Australian Bureau of Statistics Census and Survey data (ABS2005). Trend lines are fitted in (a) excluding 1994 data (canola r2 = 0.57,P = 0.02 slope 0.055 t/ha.year; wheat trend not significant). Potentialyield was estimated using APSIM-Canola validated in the region,initiated at each site in each year using a sowing rule, and run usingBureau of Meteorology data for each of the 3 locations (Robertson andKirkegaard 2005).

for Wagga Wagga, further to the west in southern NSW(Holland et al. 1999). Comparisons of average canola yieldsfor these SLAs and estimates of potential canola yieldsfor the same locations predicted using the APSIM-Canolasimulation model (Robertson and Kirkegaard 2005) indicateno predicted decline in the potential canola yield for theseareas for the 1998–2001 period (Fig. 1b). In addition, therewas no yield difference (‘gap’) between the predicted andactual yields in drier years (1994, 1997). However, this gapincreased linearly when potential yield rose above 2–2.5 t/ha,

suggesting yield constraints increased with higher potentialyield (Fig. 1c). The yield gap cannot be attributed to latesowing or the amount and distribution of rainfall in theseseasons as these factors were both accounted for in thesimulation model.

At a more local scale, consultant and grower groups in thearea also reported a decline in water-use efficiency (WUE)for canola crops. Based on April–October rainfall, 150 clientsfrom around Harden reported a reduction in WUE fromaround 8.5 kg/ha.mm in 1997 to around 5.5 kg/ha.mm in2001 (client data provided by Harden District Rural AdvisoryService 2001). By 2001, the weight of evidence that canolawas not yielding to its potential (particularly in betterseasons), and might even be declining in the traditionalhigh-rainfall canola areas, prompted investigations into thepossible causes.

Consultation among leading growers and agronomists inthe area during 2000 highlighted the increased incidence andseverity of disease, particularly blackleg and sclerotinia stemrot caused by Sclerotinia sclerotiorum (Lib.) de Bary, as alikely cause of the failure of otherwise well-managed cropsto achieve higher yields (relative to wheat) in favourableseasons. The intensity with which canola was grown rosefrom approximately 8 to 20% of the area during the 1990sand extended sequences of alternate wheat and canolacrops were common. Both blackleg and sclerotinia stemrot inoculum are dispersed primarily through wind-blownascospores released from the stubble of previous crops and sothese diseases are likely to increase in incidence and severitywith an intensification of canola cropping within a region (seeSalisbury et al. 1995; Hind et al. 2003). In addition, bothpathogens tend to rely on wet periods for infection, whichmakes the higher rainfall areas of southern NSW particularlyconducive to disease development.

Blackleg remains the most significant disease of canolain Australia (West et al. 2001), although disease severitycan be minimised (to less than 10–15% yield loss) usingbroad-based genetic resistance (Salisbury et al. 1995), andfurther reduced by avoiding close proximity to the stubbleof previous crops (Marcroft et al. 2004) and using seedand/or fertiliser fungicides in high-risk situations. Despiteconsiderable research on the management of blackleg incanola, there are no reliable data available to quantify theyield loss caused by the disease under current managementguidelines in southern NSW. Previous studies in Victoriahave suggested that for each 1% increase in the number ofplants with severe blackleg stem cankers, yield is reducedby 0.7% (McGee and Emmett 1977), although these datawere for older varieties and are possibly now outdated.In addition, other less obvious symptoms of the diseasewithout basal canker or lodging are now observed. Thesesymptoms include internal necrosis in the stem (Marcroftet al. 2004) and roots (Sosnowski et al. 2001) in associationwith subsequent premature senescence of otherwise healthy

Effect of disease on canola yield in a high-rainfall zone Australian Journal of Agricultural Research 203

plants. Consultants in southern NSW have reported anincrease in the incidence of these symptoms in recent seasons,but their effect on crop yield is uncertain. The release ofvarieties with major gene resistance for total immunity toblackleg in 2000 provided the first opportunity to betterquantify the yield loss attributable to blackleg, although thisresistance had subsequently broken down in some areas by2002 (Li et al. 2003; Sprague et al. 2006).

Less research has been conducted in Australia onsclerotinia stem rot in canola than on blackleg, althoughprevalence and associated yield loss in southern NSW inseasons favourable for disease development was documentedin the survey of Hind et al. (2003). That survey, conductedfrom 1998 to 2000, showed increased incidence in southernNSW over those years, with stem infections ranging from0 to 37.5% and an average of 10% of plants infected.No resistant canola varieties are available, and althoughfungicides are available, they are expensive ($50/ha) andonly recently registered for use in Australia. Despite anumber of forecasting systems available to assist with sprayrecommendations overseas (e.g. Turkington et al. 1991),there are currently no clear guidelines available for theireffective and economic use under Australian conditions(Hind et al. 2003).

The extent to which blackleg and sclerotinia stem rot maybe contributing to the declining yield trends in otherwise well-managed crops in southern NSW remains unclear. The aim ofthe experiments reported in this paper was to determine theeffect of these 2 diseases on canola yields in the high-rainfallarea of southern NSW as a first step to developing better

Table 1. Site and experimental management detailsRates for fertiliser application are given in kg/ha of element

Galong Wallendbeen

Location 34.67◦S, 148.63◦E 34.53◦S, 148.16◦EElevation (m) 450 550Soil typeA Red Kandosol Red/brown DermosolpH (0–10 cm) (CaCl2) 5.3 6.0C% (0–10 cm) 1.77 2.20

2001 2002 2003 2001 2002 2003HistoryB Pa, Pa, Pa, Pa Pa, Pa, C, W Pa, Pa, C, W Pa, Pa, Pa, Pa C, W, C, W C, W, C, WIsolationC (m) 200 1000 200 800 1600 50Sow date 1 4 April 30 April 1 May 30 April 30 April 1 MaySow date 2 15 May 16 May 13 MayD 15 May 15 May 13 MaySow date 3 29 May 30 May 11 June 28 May 30 May 30 MayPre-sow N 337 144 151 382 77 192Sowing NPS 71, 22, 25 32, 20, 26 58, 26, 19 20, 35, 30 10, 47, 27 58, 53, 27Topdress N 0 46 62 0 92 46Harvest 1 22 Nov. 18 Nov. 29 Nov. 3 Dec. 23 Nov. 5 Dec.Harvest 2 22 Nov. 18 Nov. 5 Dec. 3 Dec. 23 Nov. 13 Dec.Harvest 3 27 Nov. 22 Nov. 12 Dec. 7 Dec. 26 Nov. 15 Dec.

AIsbell (1996).BPa, Pasture; C, canola; W, wheat.CDistance to nearest paddock with canola stubble from previous season.DEmerged late on 3/6 due to dry soil.

control guidelines. The yield of crops treated and or resistantto the diseases was compared with untreated crops over a3-year period, and a simulation model (APSIM-Canola) wasused to provide a benchmark of the potential yield (assumingno disease) for each crop.

Materials and methodsSites and experimental design

Field experiments were conducted in each of 3 years (2001–2003) onfarms at Galong and Wallendbeen in the high-rainfall zone (650 mm)of southern NSW (Table 1). The experiments at each farm wereconducted on different, but nearby paddocks in each year, with similarsoil types. The data in Table 1 represent mean values for each site.The sites were representative of the acidic loam soils of the higherrainfall canola-growing areas in south-eastern NSW where canola(formerly rapeseed) has been grown since the 1970s. The Wallendbeensite, located 20 km to the south-west of Galong, has a higher yieldpotential with deeper loam soils with higher water-holding capacity,and cooler temperatures. Canola crops generally mature 7–10 days laterat Wallendbeen than at Galong.

The treatments used were designed to assess the effect of blacklegand sclerotinia stem rot on canola yield, rather than to develop economicdisease-control strategies. Accordingly, the recommended doses offungicides to achieve maximum disease control were applied, ratherthan the most economic rates. The treatments included 3 sowingdates to maximise the chances of encountering optimum conditionsfor disease development at each site, 2 canola varieties differing inblackleg resistance, 2 (+/−) seed and fertiliser fungicide treatmentsto control blackleg, and 2 (+/−) fungicide spray treatments to controlsclerotinia stem rot.

The sowing dates were targetted within the normal late-April tolate-June sowing window for canola in the area, but varied from year toyear, depending on the timing of the first rains of the season (Table 1).The 2 varieties chosen were the widely grown variety Rainbow with an

204 Australian Journal of Agricultural Research J. A. Kirkegaard et al.

Australian Blackleg Rating (ABR) of 6–6.5 (moderately resistant), andthe variety Hyola 60 which had the highest resistance of commerciallines (ABR 9.0) when released in 2001. The high level of resistancein Hyola 60 was derived from the introgression of a major resistancegene from Brassica rapa subsp. sylvestris, which initially providedtotal immunity to leaf infection (Li et al. 2003; Sosnowski et al. 2004).This major gene resistance to blackleg had broken down in some areasof Australia by 2003, although Hyola 60 retained an ABR of 9.0 inNSW throughout the period of this experiment (Canola Association ofAustralia, Anon. www.canolaaustralia.com.au).

The fungicide treatments for blackleg control included a seed-dressing treatment using Jockey (167 g/L fluquinconozole) applied at20 L/t of seed, and a fertilizer-dressing treatment using Impact (250 g/Lflutriafol) added with the starter fertiliser at sowing at an equivalentrate of 400 mL/ha. A treatment using both fungicides was included toprovide the maximum protection from leaf infection during germinationand seedling emergence through to the early vegetative stage. In mostinstances, the combined treatment provided better blackleg control andsimilar or higher yields than each fungicide alone, and was used as the‘treated’ data. There were isolated instances of yield depression causedby the combined seed and fertiliser treatment. In such cases, only theImpact treatment was used as the treated data because we were moreinterested in reporting potential yield in the absence of disease, ratherthan the merits of individual products or product combinations. Theperformance of the treated plots was compared with an untreated control.

The fungicide treatment for sclerotinia stem rot control in 2001and 2002 was a mixture of Folicur 430 SC (430 g/L tebuconazole) at290 mL/ha and SpinFlow (500 g/L carbendizim) at 500 mL/ha appliedwith 0.5% Uptake. In 2003, the fungicide treatment used was SumisclexBroadacre SC (500 g/L procimidone) at 1 L/ha with 0.1% BS1000.The treatment was applied as an overhead spray using 100 L/ha ofwater, the first applied at around 20% flowering, and subsequentsprays (when used) 10–12 days apart. In 2001 and 2002, 3 sprays wereapplied to the first and second sowing and 2 sprays to the third sowing,due to the shorter flowering period. All treatments in 2003 receivedonly 2 sprays.

The experimental design was identical at both sites. In 2001,experiments were arranged in 3 adjacent (unreplicated) blocks foreach sowing date, within which the variety and fungicide treatmentswere arranged as completely randomised plots (16 by 1.44 m) with3 replicates. In 2002 and 2003 the design was modified so that sowingdate was replicated within the experiment. The treatments werearranged in 3 blocks, and each block comprised a split–split-plot designwith the fungicide and variety treatments (16 by 1.44 m) completelyrandomised within the 3 sowing date main plots. Sclerotinia stem rottreatments were applied to half of each replicate from 20% floweringin each year.

Crop management and measurements

Details of crop management for each site are shown in Table 1. Pre-sowing soil cores (8/site) were taken to a depth of 1.8 m each year atthe 2 sites and divided into 20 cm segments for soil water and mineralN measurement. Seed and fertiliser fungicide treatments were appliedprior to sowing at the recommended rates and the seed sown at 3 kg/ha in18 cm rows, using narrow points and press wheels. Crops were managedby applying recommended rates of fertilisers (based on pre-sowing soiltests) and appropriate recommended herbicides to ensure that nitrogen,other nutrients, and weeds did not limit growth.

Established plant population was measured in 4 by 1 m of rowselected at random within each plot at the 3rd true-leaf stage.Blackleg leaf infection levels were estimated in winter during the earlyvegetative stage (4–6-leaf) using a visual rating (% leaf infected or leaflesions/plant). Leaf lesions were identified by the typical black fruitingbodies (pycnidia) visible within the necrotic tissue (Hammond et al.

1985). The level of lodging caused by L. maculans was assessed1–2 weeks prior to desiccation, by counting the number of plants andlodged plants in 4 pairs of adjacent rows containing approximately12 plants/row (total of 100 plants). Each plant along the row was assessedfor lodging associated with the typical basal girdling associated withblackleg canker (West et al. 2001), and the percentage of plants cankeredrecorded for each plot.

In 2001, the level of sclerotinia stem rot infection was assessedimmediately after harvest by counting the number of plants and stem-infected plants along the entire length of 2 adjacent rows in eachplot. Infected plants were those that had lesions characteristic ofsclerotinia stem rot on the main stem (Rimmer and Buchwaldt 1995).A considerable number of stems contained sclerotia when dissected.Disease incidence was expressed as the percentage of plants with steminfection. In addition, in 2001, the percentage of plants with infection inhigher branches was also measured in the hand quadrats (2 by 0.4 m2)taken for biomass and harvest index measurement prior to desiccation(see below). In 2002 and 2003, stem assessments were carried out priorto desiccation on the same 100 plants/plot assessed for blackleg.

The plots were desiccated using Sprayseed (paraquat 135 g/L, diquat115 g/L) when 60–70% of seeds had turned from green to black. Plantswere harvested 9–10 days later. Seed yield was measured using amechanical plot harvester, and also from bordered quadrat cuts (2 by0.4 m2) taken at ground level by hand at physiological maturity (justprior to desiccation). The quadrat cuts were used to determine harvestindex and total biomass.

Data analysis

The yield data were analysed by ANOVA using Genstat 6 withappropriate models to assess the main effect and interactions betweenthe experimental treatments. As a result of significant site and yearinteractions often associated with variation in the levels of the 2 diseasesof interest, the data from each site were analysed separately in each year.In 2001, the sowing dates were analysed separately due to the lack ofreplication for sowing date, and in 2002 and 2003 both overall analysesand separate analyses for each sowing date were conducted. Data fordisease incidence were checked for the need for transformation prior toanalysis. Generally, treatments were considered to differ when P < 0.05.For selected data, correlation and regression analyses using Genstat 6were carried out to investigate relationships between disease incidenceand yield.

Simulation of potential yield

Potential yield was estimated for each sowing date at both sites from2001 to 2003 using APSIM-Canola, which has been validated in thisregion (Robertson and Kirkegaard 2005). The model predicts the yieldof crops assuming no yield limitation associated with factors suchas disease, weeds, frost, and header loss, but accounts for the dailywater balance and nitrogen supply to the crop. A summary of themodel initialisation values used for each site is given in Robertsonand Kirkegaard (2005). The lower limit and drained upper limit forsoil water were either measured under appropriate field conditions (e.g.lower limit was measured following harvest in the 2002 drought), orwere measured directly using pressure plate (wilting point) and tensiontable (field capacity) apparatus. The model was initiated using themeasured pre-sowing soil water and N data measured at each site.Daily rainfall data were recoded at each site and the other climaticdata required for simulation were obtained from the Australian Bureauof Meteorology through the SILO database (www.bom.gov.au/silo).Management information (date of sowing, established plant population,dates and amounts of N fertiliser) was also model inputs. Simulationswere run from the date of the pre-sowing soil sampling until theobserved date of canola physiological maturity. The simulated yieldswere compared with the measured yields from plots treated for both of

Effect of disease on canola yield in a high-rainfall zone Australian Journal of Agricultural Research 205

the diseases in order to reveal factors other than blackleg and sclerotiniastem rot that may be limiting yield in the region.

Results

Seasonal conditions and the incidence of disease

During the 2001 season, Galong and Wallendbeen receivedbelow-average rainfall, although good spring rainfallprovided ideal conditions for the development of sclerotiniastem rot (Table 2). In 2002, much less total and spring rainfallwas received and in 2003, although monthly rainfall was closeto average, there were extended periods without rain duringlate flowering and pod-fill (Table 2). As a consequence, theoverall levels of sclerotinia stem rot were much higher atboth sites in 2001 than in 2002 or 2003 (Table 3), and weregenerally higher at Wallendbeen than at Galong. Overall,there was a higher incidence of sclerotinia stem rot withearlier sowing. In contrast, blackleg was present at both sitesin the 3 years, although at much lower levels at Wallendbeenin 2001 and 2003. The levels of blackleg lodging were highat both sites in 2002 (8–10%), but over the 3 years, levelswere much higher at Galong than at Wallendbeen (Table 3).In 2001 and 2003, the levels of blackleg generally increasedwith later sowings at both sites.

Table 2. Monthly rainfall (January to December) (mm) at Galong and Wallendbeen from2001 to 2003 and the long-term means (LTM)

GSR, Growing-season rainfall from April to October; TOT, total annual rainfall

Year J F M A M J J A S O N D GSR TOT

Galong2001 33 81 86 32 13 81 42 57 63 44 51 7 332 5902002 4 128 35 16 28 40 30 32 59 2 4 14 207 3922003 7 67 12 17 7 67 29 62 43 72 35 45 297 463LTM 50 42 51 49 51 62 62 61 56 63 52 51 404 650

Wallendbeen2001 33 72 62 67 17 67 24 47 59 54 35 13 333 5462002 6 126 26 17 50 37 46 40 78 0 3 9 268 4382003 10 74 17 50 10 67 63 79 43 62 35 83 374 593LTM 53 42 47 51 53 65 66 62 57 64 51 52 419 663

Table 3. Mean and range (parentheses) for sclerotinia stem rot (% plants with stem lesion) in the unsprayed treatmentsfor canola varieties Rainbow and Hyola 60, and mean blackleg levels (% lodged plants) measured in the untreated

Rainbow plots at the sites in each year

Year Galong WallendbeenSowing 1 Sowing 2 Sowing 3 Sowing 1 Sowing 2 Sowing 3

Sclerotinia stem rot2001 14.8 (12.7–7.3) 5.7 (4.0–6.7) 5.0 (2.6–8.7) 24.3 (20.0–29.3) 23.7 (14–42) 6.8 (5.3–7.3)2002 0.0 0.0 0.0 0.0 0.0 0.02003 0.6 (0–1.3) 0.1 (0–0.3) 0.1 (0–0.3) 1.8 (1.3–3.0) 1.1 (1.0–1.3) 0.3 (0–0.7)

Blackleg (var. Rainbow only)2001 3.3 8.0 7.7 1.0 2.7 2.32002 8.7 3.7 8.3 10.0 6.7 9.32003 5.3 11.3 10.3 0.0 0.3 1.3

Treatment effects on yield

The significance of treatment effects (sowing date, variety,blackleg and sclerotinia fungicides) on yield for theexperiments from overall analyses carried out for each seasonat both sites is summarised in Table 4. As disease incidencewas different across the sowing dates for some sites andseasons, separate analyses were also carried out for eachsowing date at both sites. In the following section, referenceis made to both the overall effect of treatments (shownin Table 4) and to the analysis of individual sowing datespresented in the figures that follow. For clarity, data for sometreatments are also pooled for presentation where there wereno significant effects or interactions.

2001 season

Blackleg. There was no significant response to the blacklegfungicide treatment at any sowing date at Galong in 2001, andno interactions of blackleg with other treatments (Table 4).This was surprising given the reasonably high levels oflodging (up to 8% in Sowings 2 and 3, Table 3), which wasreduced to 2.5% with fungicide treatment (data not shown).At Wallendbeen, there was an overall response to blacklegfungicide treatment in both varieties (Table 4), which resulted

206 Australian Journal of Agricultural Research J. A. Kirkegaard et al.

Table 4. Summary of treatment main effects and significant 2-way interactions oncanola yield at Galong and Wallendbeen from 2001 to 2003

Higher order interactions and 2-way interactions not shown were not significant.Experimental design in 2001 did not allow statistical analysis of sowing date effects

Source Galong Wallendbeen2001 2002 2003 2001 2002 2003

Sowing date – ∗ n.s. – n.s. ∗

Variety ∗∗ ∗∗ ∗∗ ∗∗ n.s. ∗∗

Sclerotinia fungicide ∗∗ n.s. n.s. ∗∗ ∗ ∗

Blackleg fungicide n.s. ∗∗ ∗∗ ∗ ∗∗ n.s.Sowing date × Variety – n.s. ∗ – n.s. ∗∗

Sowing date × Blackleg – ∗ n.s. – ∗ n.s.Variety × Blackleg n.s. ∗∗ n.s. ∗ n.s. n.s.Variety × Sclerotinia n.s. n.s. n.s. ∗ n.s. n.s.

n.s., Not significant (P > 0.05); ∗P < 0.05; ∗∗P < 0.001.

from a significant response in Sowings 2 and 3, but not inSowing 1 (data not shown). The yield increase associated withblackleg fungicide treatment was similar in Sowings 2 and 3,but averaged 0.51 t/ha for Hyola 60 and 0.85 t/ha for Rainbow,reflected in the significant Variety × Blackleg interaction(Table 4). These yield responses occurred despite thelow levels of blackleg lodging observed at the site forthose sowings (0–2.7%, Table 3). Some of the inconsistentresponses to blackleg treatment at the sites in 2001 mayhave been due to the over-riding effect of sclerotinia stemrot, as fungicide sprays did not completely control stem rotdevelopment (see below).

Sclerotinia. The major response in 2001 at both sites wasto sclerotinia stem rot treatment (Table 4). To visualise theresponse to sclerotinia stem rot, the data for +/− blacklegtreatment have been pooled for each variety and sowing time(Fig. 2a). There was a consistent yield increase in responseto sclerotinia treatment (ranging from 0.2 to 1.5 t/ha) acrossboth sites, varieties, and sowing dates, with Sowing 3 atGalong the only exception (Fig. 2a). The only significantinteraction within the individual sowing data analysesbetween any of the experimental treatments at either site wasa Variety × Sclerotinia treatment interaction at Sowing 2 atGalong, where the yield increase in response to sclerotiniatreatment was much larger in Rainbow (0.70 t/ha) than inHyola 60 (0.15 t/ha) (Fig. 2a). Generally, the larger yieldresponses to sclerotinia control measured at Wallendbeen(>1.0 t/ha) were consistent with higher levels of infectedstems at that site (Fig. 2b), although significant yieldresponses persisted in Sowing 3 at Wallendbeen despite themuch lower incidence of the disease.

In addition to the stem infection, significant numbers ofplants with sclerotinia were infected in the upper branches atboth of the sites, although the sclerotinia fungicide treatmentwas less effective in controlling this symptom than incontrolling stem infection (Table 5). In the untreated plots atGalong, the overall incidence of branch infection was higherthan that of stem infection, but branch infection showed a

Galong

Yie

ld (

g/m

2 )

0

100

200

300

400

500

600

Wallendbeen

30/4 15/5 28/5 30/4 15/5 28/5Scl

erot

inia

(%

infe

cted

ste

ms)

0

10

20

30

40

Hyola-60 Rainbow

30/4 15/5 28/5 30/4 15/5 28/5

Sowing date

VarietyHyola-60 Rainbow

ns

ns

ns

ns

(a)

(b)

Fig. 2. Effect of sclerotinia fungicide treatment on (a) canola yield and(b) incidence of sclerotinia at Galong and Wallendbeen in 2001. Dataare means of blackleg-treated and untreated plots. ns, No significantdifference between the untreated (solid) and treated (hatched) treatmentswithin each sowing date.

similar decline with sowing date (cf. Table 5 and Fig. 2b). AtWallendbeen, the levels of branch infection were similar tothe levels of stem infection at Sowings 1 and 2 (20–30%), butunlike stem infection, remained high at Sowing 3.

The relationship between grain yield and the level ofsclerotinia stem infection was investigated at both sitesusing regression analysis. There were significant regressions(P < 0.001) at both sites when data from all sowing dateswere included. However, the relationship explained only1.4% of the variation in yield at Galong and 30.3% at

Effect of disease on canola yield in a high-rainfall zone Australian Journal of Agricultural Research 207

Wallendbeen (data not shown). Exclusion of Sowing 3data at both sites, where infection levels were relativelylow, increased the amount of variation explained by therelationship to 18.1% at Galong and 46.5% at Wallendbeen.The regression equations relating yield (t/ha) and sclerotiniastem infection (% infected plants) had similar slopes of−0.048 t/ha at Galong (s.e. = 0.008) and −0.046 t/ha atWallendbeen (s.e. = 0.004), indicating that for these sites in2001, each 1% increase in sclerotinia stem infection reducedyield by 46–48 kg/ha:

Yield (t/ha) = −0.048 (% stem infection)+4.62 (P < 0.001) for Galong

Yield (t/ha) = −0.046 (% stem infection)+5.94 (P < 0.001) for Wallendbeen

The significant amount of yield variation unaccounted forby sclerotinia indicates that other factors were important ininfluencing yield at the sites.

At Galong, Hyola 60 yielded significantly more thanRainbow at all sowing dates (Table 4, Fig. 2a), with asimilar yield advantage (0.70 t/ha) at Sowings 1 and 2, and asmaller advantage at Sowing 3 (0.25 t/ha). At Wallendbeen,a similar trend was evident but was significant only atSowing 2 (yield increase 0.72 t/ha) (Fig. 2a). The effect ofsowing time in the experiment, and its interaction with othertreatments cannot be compared statistically due to a lack ofreplication, but the trend in the disease treated plots was fora decline in yield with later sowing at Galong, but not atWallendbeen (Fig. 2a).

2002 season

Sclerotinia. The low level of sclerotinia stem rot observedat the sites in 2002 (Table 3) was reflected in the absenceof an overall response to sclerotinia fungicide spray at bothsites, although there was a small negative response (0.3 t/ha)to sclerotinia spray at Wallendbeen in Sowing 1 (Table 4). Asa result, and as there were no other interactions between thesclerotinia treatments and other treatments, data presented

Table 5. The effect of sclerotinia stem rot treatment on theincidence of sclerotinia branch infection (% canola plants infected)

at Galong and Wallendbeen in 2001Data are means for the 2 varieties

Site/sowing Untreated Treated P value

GalongSowing 1 28.2 15.9 0.05Sowing 2 16.3 7.9 0.03Sowing 3 9.0 11.5 n.s.

WallendbeenSowing 1 25.5 10.4 n.s.Sowing 2 17.0 4.3 0.05Sowing 3 20.7 20.2 n.s.

n.s., Not significant.

for 2002 have been pooled for +/− sclerotinia treatments tovisualise the response to other treatments.

Blackleg. The major response in 2002 was to blacklegfungicide treatment, with significant interactions at bothsites between blackleg fungicide treatment and sowing date,and a significant Blackleg × Variety interaction at Galong(Table 4, Fig. 3a). At Galong, Hyola 60 out-yielded Rainbowat all sowing dates. In general the yield response to blacklegfungicide was greater in Rainbow than in Hyola 60 (Fig. 3a).This was consistent with the higher levels of blackleg lodgingmeasured in that variety (Fig. 3b) and was also reflected ina significant Variety × Blackleg interaction when analysedacross all 3 sowing dates at Galong (Table 4), although theinteraction was significant only at Sowing 2 when analysedseparately (Fig. 3a). The largest yield response to the blacklegfungicides at Galong (1.06 t/ha or 58.5%) was observed forRainbow at Sowing 1 (Fig. 3a), whereas a much smallerresponse was observed at Sowing 3 despite a similar levelof blackleg lodging at these 2 sowing times (Fig. 3b).

At Wallendbeen, there was no significant blacklegtreatment effect on yield for Sowing 1 in either variety,but significant interactions between variety and fungicidetreatments in the later sowings (Table 4, Fig. 3a). At Sowing 2,both varieties responded to the blackleg fungicide treatments,and the response tended to be greater in Hyola 60 (0.60 t/ha)than in Rainbow (0.45 t/ha) despite very low levels of lodging

Galong

0

100

200

300

400Wallendbeen

29/4 13/5 27/5 29/4 13/5 27/50

2

4

6

8

10

29/4 13/5 27/5 29/4 13/5 27/5

nsnsnsnsns

ns

ns

nsns

ns

(a)

(b)

Yie

ld (

g/m

2 )B

lack

leg

(% lo

dgin

g)

Hyola-60 Rainbow

Sowing date

VarietyHyola-60 Rainbow

Fig. 3. Effect of blackleg fungicide treatment on (a) canola yieldand (b) incidence of blackleg lodging at Galong and Wallendbeen in2002. Data are means of sclerotinia-treated and untreated plots. ns, Nosignificant difference between the untreated (solid) and treated (hatched)plots within each sowing date.

208 Australian Journal of Agricultural Research J. A. Kirkegaard et al.

in Hyola 60 (Fig. 3b). In addition to finding isolated lodgedplants in Hyola 60, L. maculans was also isolated from cutstems with symptoms of internal necrosis, confirming that theresistance to blackleg in Hyola 60 had begun to break downat that site. At Sowing 3, there was a significant response(0.32 t/ha) to the blackleg fungicide treatment in Rainbow,but not in Hyola 60, which is more consistent with the levelsof blackleg lodging observed in the varieties at that sowing(Fig. 3a, b). In the moderately resistant variety Rainbow,the magnitude of the yield response to blackleg across thesites and sowing dates was not closely related to the levelof lodging observed (Fig. 3a, b). For example, there was noresponse in yield to the large reduction in blackleg lodging inthe first sowing at Wallendbeen, whereas a smaller reductionat the same sowing time at Galong increased yield by over50%. Overall, at Wallendbeen, there was no difference in theyield of the varieties (Table 4), but in the analysis of separatesowing dates, var. Rainbow yielded better than Hyola 60 atSowing 3, particularly in the blackleg-treated plots (Fig. 3).The higher yield of Rainbow compared with Hyola 60 in theblackleg-treated plots at Sowing 3 suggests that there wasa yield advantage for Rainbow unrelated to blackleg. It islikely that the smaller plants and earlier flowering time forvar. Rainbow reduced the level of water stress at the latersowing in this dry season. This is supported by the lowerharvest index for Hyola 60 (0.22) than for Rainbow (0.25)at Sowing 3, compared with a similar harvest index for bothvarieties (0.28) at the earlier sowings.

Yield declined with later sowing at Galong, althoughthe effect of blackleg tended to mask this response inthe untreated plots (Table 4, Fig. 3). At Wallendbeen, theunexpected response to blackleg fungicides in Hyola 60at Sowing 2, contrasted with a more general trendtowards reduced yield with later sowing for both varietiesat the site.

2003 season

The relative importance of the 2 diseases differed at the2 sites in 2003. At Galong, the low levels of sclerotiniastem rot infection (Table 3) were reflected in the absenceof any significant response to fungicide treatment for anysowing date, whereas there were significant responses toblackleg treatment and variety (Table 4). At Wallendbeen,both diseases had relatively low incidence (Table 3) but therewas an overall response to sclerotinia treatment, resultingfrom a significant response in Sowing 1, but no responseto blackleg treatment (Table 4). As a result, the data forGalong have been pooled for +/− sclerotinia treatment tohighlight the significant responses to blackleg treatment,whereas the data for Wallendbeen have been pooled for+/− blackleg treatment to highlight responses to sclerotiniatreatment (Fig. 4).

At Galong, there was a significant response to blacklegfungicide treatment at all sowing dates in Rainbow, but no

Galong

0

100

200

300

400Wallendbeen

0

100

200

300

400

1/5 13/5 11/6 1/5 13/5 11/60

2

4

6

8

10

12

1/5 13/5 30/5 1/5 13/5 30/5

Scl

erot

inia

(% in

fect

ed s

tem

s)

0.0

0.5

1.0

1.5

2.0

2.5

ns

ns

nsns

nsns

ns

nsns ns

nsns ns

(a) (b)

Yie

ld (

g/m

2 )B

lack

leg

(% lo

dgin

g)Hyola-60 Rainbow

Sowing date

VarietyHyola-60 Rainbow

Fig. 4. Effect of disease treatments on canola yield and diseaseincidence in 2003. Data for Galong (a) are meaned for sclerotinia-treated and untreated plots, and data for Wallendbeen (b) are meanedfor blackleg-treated and untreated plots. ns, No significant differencebetween untreated (solid) and treated (hatched) plots within eachsowing date.

response in Hyola 60 (Fig. 4a). The average yield response toblackleg treatment in Rainbow was 0.38 t/ha, and was lowestin Sowing 2 (0.14 t/ha) despite the high level of lodgingobserved at that sowing time. There was no lodging observedat Galong in Hyola 60. In the blackleg-treated plots, therewas no significant effect of sowing date and no obvious yieldadvantage in Hyola 60.

At Wallendbeen, the yield response to sclerotiniatreatment was significant only in Sowing 1, consistent withthe higher levels of stem infection observed for that sowingtime (Fig. 4b). The magnitude of the yield response tosclerotinia treatment in Sowing 1 (0.39 t/ha for Hyola 60)is surprisingly high given the relatively low levels of steminfection observed at the site (1–2%) (Fig. 4a, b). In thesclerotinia-treated plots, Hyola 60 significantly out-yieldedRainbow at all sowing dates (Table 4, Fig. 4a), althoughthe yield advantage was lower in Sowing 1 (0.3 t/ha) than inthe later sowings (0.5–0.6 t/ha). This apparent yield penaltyfor Hyola 60 with early sowing appears to be related toexcessive vegetative growth (total biomass of 10 t/ha cf.8.2 t/ha for Rainbow) leading to the exhaustion of soil waterand significant plant water stress during grain filling. This wasreflected in a much lower harvest index of 0.18 for Hyola 60compared with 0.23 for Rainbow in Sowing 1, which wasconsiderably less than the overall average of 0.32 for bothvarieties at the later sowing times.

Effect of disease on canola yield in a high-rainfall zone Australian Journal of Agricultural Research 209

Comparison of potential yield estimates usingAPSIM-Canola with measured yield

The yield predicted for the variety Rainbow using APSIM-Canola, and the measured yield for sclerotinia- and blackleg-treated plots of Rainbow are compared for each site, year, andsowing date in Fig. 5. In general, the potential yields predictedby the model are similar to, or slightly lower than thosemeasured yields in the disease-treated plots at both sites.This indicates that when diseases are controlled or absent,canola at these sites yielded to its potential as predicted bythe model. The exception to this was at Wallendbeen in 2002where the simulated potential yield was much higher thanthe yield measured at the site. At that site, high temperatureson 22 September at mid-flowering are likely to have causedpremature cessation of flowering and hence limited grain set(Morrison and Stewart 2002). The effect of high temperatureon flowering and grain set is currently not captured byAPSIM-Canola.

Galong

0.0

1.0

2.0

3.0

4.0

5.0

6.0

2001

2002

2003

Wallendbeen

0.0

1.0

2.0

3.0

4.0

5.0

6.0

19 Apr. 9 May 29 May 18 June

Sowing date

Yie

ld (

t/ha)

2001

2002

2003

Fig. 5. Simulated yield (lines) and measured yield (symbols) for thedisease-treated canola (var. Rainbow) plots at both sites in 2001 (solidline), 2002 (dashed line), and 2003 (dotted line).

Discussion

Overall effect of disease on canola yield

The results of these experiments support the perceptionamong leading growers and consultants that the diseasesblackleg and sclerotinia stem rot are causing significant yieldreductions in otherwise well-managed crops in the high-rainfall areas of southern NSW. A summary of the resultsindicates that either individually, or in combination, thediseases caused significant yield losses at both sites in everyyear of the study, ranging from 0.39 t/ha to 1.54 t/ha (Table 6).The estimates of yield loss associated with blackleg are basedon response to fungicide alone, and could be increased furtherif one assumes that at least some of the yield advantage ofHyola 60 over Rainbow (which was evident at both sites in allyears with the exception of 2002 at Wallendbeen), resultedfrom a reduction in blackleg infection to levels lower thanthat in the treated plots of Rainbow (see Figs 3b and 4bGalong). The ability of the crops to perform to their water-and N-limited potential when the diseases were either absentor controlled, provides clear evidence that these diseases arelikely to be one of the key causes of the low or decliningyields in the area, especially given that the use of fungicidesin the area was negligible before 2001. Although the surveyby Mead et al. (2004) did not include disease monitoring,the general reduction in yield across the entire central andsouthern NSW area, including the irrigated areas, suggeststhat the causes of yield decline are widespread and notlimited to particular soil types, management practices, orenvironments. The analysis of regional ABS data togetherwith wheat yield trends and predicted yield potential (Fig. 1)indicates that it is the failure of canola to perform in wetseasons that appears to be the most obvious reason for yielddecline. Both of the diseases considered in this study arelikely to be more prevalent in wet seasons.

Effect of blackleg

Blackleg is considered the most significant disease of canolain Australia, and the results of these experiments confirmthe prevalence of the disease in southern NSW. Blackleglodging in the variety Rainbow was observed at both sitesin all seasons (with the exception of 2003 Sowing 1 atWallendbeen), and significant responses to the fungicidetreatment were evident in all years. The lower overallresponse of the blackleg-resistant variety Hyola 60 to theblackleg fungicide treatment provides further evidence ofthe yield limitation caused by blackleg in this study. Directcomparisons of yield performance of the 2 cultivars aresomewhat confounded by their differences in flowering dateand biomass (Hyola 60 flowers later and has higher biomass),particularly in the drier 2002 and 2003 seasons. However, theyield advantage of Hyola 60 over Rainbow is clearly reducedin the blackleg-treated plots (see Figs 3a, 4a for Galong),supporting the evidence that blackleg continues to limit the

210 Australian Journal of Agricultural Research J. A. Kirkegaard et al.

Table 6. Range in yield response (t/ha) to treatment for blackleg and sclerotinia stemrot at Galong and Wallendbeen in southern NSW from 2001 to 2003

Data for both varieties are included

Galong Wallendbeen2001 2002 2003 2001 2002 2003

Blackleg 0.0 0.0–1.06 0.0–0.52 0.0–0.85 0–0.60 0.0Sclerotinia 0.17–0.76 0.0 0.0 0.85–1.54 0.0 0–0.39

yield of canola cultivars with Blackleg Resistance Ratingsof 6–6.5. As previously reported by Marcroft et al. (2004),paddock history (short canola–wheat rotations) had littleeffect on the overall levels of blackleg reported at the sites,but isolation from the previous year’s canola stubble of up to1600 m at Wallendbeen in 2002 coincided with the highestdisease severity at that site (7–10% lodging). It is possible,in years highly conducive to blackleg development, thatisolation at this scale does not prevent disease development.The different paddock histories and isolation distances inthese experiments represent typical ranges likely to beencountered in this intensive canola-growing area, so theblackleg levels can be considered representative for theseasonal conditions encountered.

Despite considerable research on the management ofblackleg in canola, there are no reliable data available toquantify the yield loss caused by the disease under currentmanagement guidelines in southern NSW. Previous studiesin Victoria suggested that for each 1% increase in the numberof plants with severe blackleg stem cankers, yield is reducedby 0.7% (McGee and Emmett 1977), although these datawere for older varieties and are possibly now outdated.Compilation of the data collected in these experimentssuggests that the effect may be much greater (Table 7).Where significant reductions in blackleg lodging coincidewith significant yield increases, the data show quite a rangein the yield response to reduced lodging from 1.1% to 8.6%increase in yield at Galong, and from 1.8% to 9.1% atWallendbeen for each 1% reduction in lodging. The meanyield increase associated with a 1% reduction in lodging forthe 2 sites was very similar (5.1% and 5.4% for Galong and

Table 7. Summary of the % yield increase associated with each1% reduction in blackleg lodging for variety Rainbow

Data for 2002 and 2003 are summarised from Figs 3 and 4. Data forWallendbeen 2001 are described in the text. Only data where asignificant yield response to blackleg treatment coincided with

significant reductions in lodging are included

Galong Wallendbeen2001 2002 2003 2001 2002 2003

Sowing 1 – 8.4 8.6 – 1.8 –Sowing 2 – 6.1 1.1 6.8 6.9 –Sowing 3 – 3.2 3.4 9.1 2.7 –Site mean 5.1 5.4

Wallendbeen, respectively). This is considerably higher thanthat previously reported. The variation is not surprising, andis likely to have many causes including: (1) the yield effects ofblackleg, which are associated with internal or root rot and notwith the degree of lodging alone; (2) effects of the fungicideson diseases other than blackleg, such as white leaf spotor downy mildew; (3) the capacity of crops to compensatefor lodged plants under certain seasonal conditions; and(4) the confounding effects of other factors on yield, suchas sclerotinia stem rot or drought. Evidence for all of thesepossibilities can be found within the data where relativelylarge yield responses to blackleg treatment occur in theabsence of lodging (e.g. Hyola 60 Sowing 2, Wallendbeen2002 Fig. 3), or where large reductions in the level of lodginghave no impact on yield (Rainbow sowing 1, Wallendbeen2002, Fig. 3). Marcroft et al. (2004) showed that individualplants with internal stem necrosis suffered yield loss of 7.4%when 20–40% of the stem cross-section was necrotic and upto 29.4% when 80–100% was necrotic. Thus, significant yieldloss can result from infection in unlodged plants. In spite ofthe variation apparent in this rule of thumb, the average 5%yield reduction associated with each 1% increase in lodging inthese experiments (relevant for lodging up to 10%) indicatesthe critical importance of maintaining adequate control overthis disease.

The observation of isolated lodging and internal necrosisin Hyola 60 in 2002 provided an early indication that themajor gene resistance to blackleg present in that cultivarhad begun to break down. This occurred subsequently indevastating circumstances on the Lower Eyre Peninsula inSouth Australia in 2003, where the resistance breakdownoccurred across a large area, with yield losses of up to90% (Sprague et al. 2006). That experience underscoresthe importance of continued development of broad sourcesof blackleg resistance as well as the need to finely tuneother management approaches to disease control, includingfungicide use in high-risk situations.

Effect of sclerotinia stem rot

Sclerotinia stem rot is generally described as a sporadicdisease in Canada and Australia, and it proved to be so duringthe 3-year period of this study. The disease was present athigh levels at both sites in 2001, absent at both sites in 2002,and present at low levels at both sites in 2003. The levelsof sclerotinia stem rot measured at the sites in 2001 (up to

Effect of disease on canola yield in a high-rainfall zone Australian Journal of Agricultural Research 211

30–40% infection) were at the upper end of levels reportedby Hind et al. (2003) for the 1998–2000 period across theregion, but these high levels accord with the increasing trendin the disease incidence that those authors observed overthat period. The drier spring conditions of 2002 and 2003reduced disease incidence, although early sown crops atWallendbeen were infected at low levels (2–3%) in 2003.Although the disease is considered sporadic on a nationalscale, it is likely to be more prevalent and cause greateryield loss in areas such as south-eastern NSW where goodspring rainfall and vigorous crops facilitate disease infectionand development. The relationships derived between theincrease in yield and the reduction in stem infection resultingfrom sclerotinia treatment at the 2 sites in 2001 suggest thatfor every 1% increase in stem infection, yield is reducedby an average of 1.3%. This is considerably higher thanthe 0.5% yield loss reported for each 1% increase in steminfection in Canada (Morrall and Thomson 1991), but mayreflect the generally larger plants and lower populationsunder Australian conditions, which increase the potentialyield loss associated with main-stem infection of individualplants compared with Canadian crops. Up to 3 sprays werenot enough to eliminate the stem infection altogether in2001, and had little effect on the level of infection in higherbranches. Guidelines to maximise the effectiveness ofexpensive fungicide sprays to reduce the disease are urgentlyrequired. Given that the average infection levels reportedfrom a much wider zone of southern NSW (including lowerrainfall areas) from 1998 to 2000 were 10% (Hind et al.2003), it is clear that the potential effects on yield(13% assuming a multiplier of 1.3) are significant andguidelines for the control of the disease are urgently needed.It is interesting to note that the period from 1998 to 2001, inwhich the yield decline in canola was most evident, coincidedwith a period of increasing sclerotinia stem rot incidenceacross the region.

Effect of sowing date

Sowing date was not included as an important treatment inits own right in these experiments, but was used to maximisethe chances of capturing a greater range of disease severityin each season. The results for the 3 seasons of this studysuggest that, in general, sclerotinia stem rot infection wasmore severe with earlier sowing, whereas blackleg wasmore severe with later sowing. Thus, irrespective of sowingdate, there is a high likelihood that at least one of thesediseases will affect crops. Paddock history also had littleeffect on disease development as both diseases occurredin 2001 in paddocks that had not grown canola for 4 yearsbecause of their relatively close proximity to stubble. Thisis consistent with the mode of infection of both diseasesthrough wind-blown ascospores and indicates that isolation,rather than rotation, is a more effective management tool(Marcroft et al. 2004).

The results of these experiments have demonstrated thesignificant effect that the diseases blackleg and sclerotiniastem rot are having on canola yields in the higher rainfall areasof south-eastern NSW. The level of yield loss associated withboth diseases provides a clear indication that cost-effectivecontrol measures are urgently required in order to maintaincanola as a viable and important break crop in the farmingsystems of the area.

Acknowledgments

We thank Grain Growers Australia and Grains Research andDevelopment Corporation for funding this work through theBest Bet Canola Project administered by Harden and DistrictRural Advisory Service with advice from Best Bet CanolaManagement Committee. We also thank Mr Charlie Baldry,Mr Steve Woodhead, and Mr Neil McColl for use of their landfor the experiments, and Mr Geoff Howe, Mr John Graham,and Mr Peter Hines for expert technical assistance.

References

Australian Bureau of Statistics (2005) ‘Agriculture commoditycollection (data available on request).’ (Australian Bureau ofStatistics: Canberra, ACT)

Hammond KE, Lewis BG, Musa TM (1985) A systemic pathway in theinfection of oilseed rape plants by Leptosphaeria maculans. PlantPathology 34, 557–565.

Hind TL, Ash GJ, Murray GM (2003) Prevalence of sclerotinia stem rotof canola in New South Wales. Australian Journal of ExperimentalAgriculture 43, 163–168. doi: 10.1071/EA02027

Holland JF, Robertson MJ, Kirkegaard JA, Bambach R, Cawley S(1999) Yield of canola relative to wheat and some reasons forvariability in the relationship. In ‘Proceedings 10th InternationalRapeseed Congress’. Canberra, ACT. (CD-ROM, GCIRC:Paris, France)

Isbell RF (1996) ‘The Australian soil classification.’ (CSIROPublishing: Collingwood, Vic.)

Kirkegaard JA, Christen O, Krupinsky J, Layzell D (2004)‘Break crop benefits in temperate wheat production.’ http://www.cropscience.org.au/icsc2004/symposia/2/1/191 kirkegaardja.htm

Li H, Sivasithamparam K, Barbetti MJ (2003) Breakdown of a Brassicarapa subsp. sylvestris single dominant blackleg resistance genein B. napus rapeseed by Leptosphaeria maculans field isolates inAustralia. Plant Disease 87, 752.

Marcroft SJ, Sprague SJ, Pymer SJ, Salisbury PA, Howlett BJ (2004)Crop isolation, not extended rotation length, reduces blackleg(Leptosphaeria maculans) severity of canola (Brassica napus)in south-eastern Australia. Australian Journal of ExperimentalAgriculture 44, 601–606. doi: 10.1071/EA03087

McGee DC, Emmett RW (1977) Blackleg (Leptosphaeria maculans(Desm.) Ces. et de Not.) of rapeseed in Victoria: crop lossesand factors which affect disease severity. Australian Journal ofAgricultural Research 28, 47–51. doi: 10.1071/AR9770047

Mead A, Nicol H, Cole C (2004) Identifying constraints to increasedcanola yield in central and southern NSW. Report to Grains Researchand Development Corporation (GRDC) on Project DAN00050.GRDC, Canberra, ACT.

Morrall RAA, Thomson JR (1991) ‘Petal test manual for sclerotinia incanola.’ (University of Saskatchewan: Canada)

Morrison MJ, Stewart DW (2002) Heat stress during flowering insummer Brassica. Crop Science 42, 797–803.

212 Australian Journal of Agricultural Research J. A. Kirkegaard et al.

Norton R, Kirkegaard JA, Angus JF, Potter T (1999) Canola in rotations.In ‘Canola in Australia: the first thirty years’. (Eds PA Salisbury,TD Potter, G McDonald, AG Green) pp. 23–28. (OrganisingCommittee 10th International Rapeseed Congress, ISBN 0-646-38131-8)

Rimmer SR, Buchwaldt L (1995) Diseases. In ‘Brassica oilseeds’.(Eds DS Kimber, DI McGregor, G McDonald, AG Green)pp. 111–140. (CAB International: Wallingford, UK)

Robertson MJ, Kirkegaard JA (2005) Water-use efficiency of drylandcanola in an equi-seasonal rainfall environment. Australian Journalof Agricultural Research 56, 1373–1386.

Salisbury PA, Ballinger DJ, Wratten N, Plummer KM, Howlett BJ(1995) Blackleg disease on oilseed Brassica in Australia: a review.Australian Journal of Experimental Agriculture 35, 665–672.doi: 10.1071/EA9950665

Sosnowski MR, Ramsay M, Murray G, Scott E, Wilmshurst C (2001)Symptoms of blackleg (Leptosphaeria maculans) on the roots ofcanola in Australia. Plant Pathology 50, 808. doi: 10.1046/j.1365-3059.2001.00631.x

Sosnowski MR, Scott ES, Ramsey MD (2004) Infection of Australiancanola cultivars (Brassica napus) by Leptosphaeria maculans isinfluences by cultivar and environmental conditions. AustralasianPlant Pathology 33, 401–411. doi: 10.1071/AP04045

Sprague SJ, Marcroft SJ, Hayden HL, Howlett BJ (2006) Major generesistance in Brassica napus overcome by Leptosphaeria maculanswithin three years of commercial production in southeasternAustralia. Plant Disease 90, 190–198.

Turkington TK, Morrall RAA, Gugel RK (1991) Use of petal infestationto forecast sclerotinia stem rot of canola: evaluation of early bloomsampling 1985–90. Canadian Journal of Plant Pathology 13, 50–59.

West JS, Kharbanda PD, Barbetti MJ, Fitt BDL (2001) Epidemiologyand management of Leptosphaeria maculans (phoma stem canker)on oilseed rape in Australia, Canada and Europe. Plant Pathology50, 10–27. doi: 10.1046/j.1365-3059.2001.00546.x

Manuscript received 24 May 2005, accepted 28 October 2005

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