Plant Pathology (2011) 60, 566–574 Doi: 10.1111/j.1365-3059.2010.02400.x

Potential biological control of clubroot on canola andcrucifer vegetable crops

G. Penga*, L. McGregora, R. Lahlalia, B. D. Gossena, S. F. Hwangb, K. K. Adhikaric,

S. E. Strelkovd and M. R. McDonaldc

aAgriculture and Agri-Food Canada, 107 Science Place, Saskatoon, SK S7N 0X2; bCrop Diversification Centre North, Alberta

Agriculture and Rural Development, 17507 Fort Road, Edmonton, AB T5Y 6H3; cDepartment of Plant Agriculture, University of

Guelph, 50 Stone Road East, Guelph, ON N1G 2W1; and dDepartment of Agricultural, Food and Nutritional Science, University of

Alberta, 410 Agriculture ⁄ Forestry Centre, Edmonton, AB T6G 2P5, Canada

Clubroot caused by Plasmodiophora brassicae is an emerging threat to canola (Brassica napus) production in western Can-

ada, and a serious disease on crucifer vegetable crops in eastern Canada. In this study, seven biological control agents and

two fungicides were evaluated as soil drenches or seed treatments for control of clubroot. Under growth cabinet conditions,

a soil-drench application of formulated biocontrol agents Bacillus subtilis and Gliocladium catenulatum reduced clubroot

severity by more than 80% relative to pathogen-inoculated controls on a highly susceptible canola cultivar. This efficacy

was similar to that of the fungicides fluazinam and cyazofamid. Under high disease pressure in greenhouse conditions, the

biocontrol agents were less effective than the fungicides. Additionally, all of the treatments delivered as a seed coating were

less effective than the soil drench. In field trials conducted in 2009, different treatments consisting of a commercial formula-

tion of B. subtilis, G. catenulatum, fluazinam or cyazofamid were applied as an in-furrow drench at 500 L ha)1 water vol-

ume to one susceptible and one resistant cultivar at two sites seeded to canola in Alberta and one site of Chinese cabbage in

Ontario. There was no substantial impact on the susceptible canola cultivar, but all of the treatments reduced clubroot on

the susceptible cultivar of Chinese cabbage, lowering disease severity by 54–84%. There was a period of 4 weeks without

rain after the canola was seeded, which likely contributed to the low treatment efficacy on canola. Under growth cabinet

conditions, fluazinam and B. subtilis products became substantially less effective after 2 weeks in a dry soil, but cyazofamid

retained its efficacy for at least 4 weeks.

Keywords: Brassica napus, Brassica rapa subsp. chinensis var. utilis, Plasmodiophora brassicae, seed treatment,

soil drench

Introduction

Clubroot, caused by the plasmodiophorid pathogen Plas-modiophora brassicae, is a serious disease of brassicacrops worldwide. In western Canada, clubroot wasreported on canola in central Alberta for the first time in2003 (Tewari et al., 2005) and the disease has since beenfound in more than 456 fields in the province of Alberta(Strelkov et al., 2010). There are about 4Æ7 million hect-ares of spring canola crops in western Canada, and thecontinuing spread of clubroot is becoming a threat tocanola production across the region. In eastern Canada,clubroot causes severe damage on a range of crucifer veg-etable crops. Soil liming and crop rotation are recom-mended for managing clubroot, but success has beensporadic or limited (McDonald et al., 2004).

*E-mail: Gary.Peng@agr.gc.ca

Published online 8 December 2010

566

Until recently, there was a lack of effective options forcontrol of clubroot on canola, largely because all of thecommercial cultivars were highly susceptible (Strelkovet al., 2006). In addition, the impact of agronomicapproaches to clubroot management such as early seed-ing and long cropping rotations out of canola were inade-quate when used alone, especially in a longer and warmergrowing season (Gossen et al., 2010) or for highly suscep-tible canola cultivars (Wallenhammar et al., 2000).A resistant canola cultivar became available in 2009.While important to clubroot management, genetic resis-tance has generally been race specific (Diederichsen et al.,2006) and can break down when virulent races increasein the pathogen population. Four pathotypes of P. brassi-cae have been identified in the population in Alberta(Strelkov et al., 2007; Xue et al., 2008), and there arepotentially more that are present at low frequency. There-fore, it is prudent to develop an integrated strategy forsustainable management of clubroot that includes, but isnot entirely dependent on, cultivar resistance (Dixon,2003; Donald & Porter, 2009). Additional measures

ª 2010 Agriculture and Agri-Food Canada

Plant Pathology ª 2010 BSPP

Biocontrol of clubroot 567

including soil nutrient management (Webster & Dixon,1991a,b; Dixon & Page, 1998), cultural practices basedon better understanding of pathogen biology (McDonald& Westerveld, 2008; Gossen et al., 2009), crop rotation(Wallenhammar, 1996), and fungicides (Suzuki et al.,1995; Cheah et al., 1998; Takeshi et al., 2004) or biofun-gicides (Cheah et al., 2000; Peng et al., 2009) will proba-bly help the performance and longevity of geneticresistance.

Several studies have illustrated the potential for reduc-ing clubroot using naturally occurring microorganisms(Narisawa et al., 1998; Arie et al., 1999; Joo et al., 2004).Some of these organisms produce anti-microbial metabo-lites against P. brassicae (Arie et al., 1998; Kim et al.,2004), while others colonize roots (Hashiba et al., 2003;Usuki & Narisawa, 2007) and induce resistance to thedisease (Morita et al., 2003). Microbial control of club-root is attractive because certain soil microbes can colo-nize root and ⁄ or the rhizosphere and so potentiallyprovide durable protection. So far, no biocontrol agenthas been available for clubroot control, but several,including Bacillus subtilis (Serenade), Gliocladium cate-nulatum (syn. Clonostachys rosea f. catenulate) (Pre-stop), Streptomyces griseoviridis (Mycostop), S. lydicus(Actinovate) and Trichoderma harzianum (Root Shield),have been registered in Canada for control of othersoil-borne diseases. If effective, these biocontrol agentsmay be integrated for clubroot control in Canada.The objectives of this study were to: (i) assess thepotential of commercially available biofungicides forclubroot control, (ii) investigate potential deliveryoptions for field application and (iii) evaluate the efficacyof selected biocontrol agents against clubroot under fieldconditions.

Materials and methods

Plant, pathogen and controlled environmenttreatments

A highly susceptible canola cultivar, Fortune RR, wasused in the initial growth cabinet tests. The seed wassown in a soil-less mix in tall narrow plastic pots (4 cmdiameter, 20 cm depth, Steuwe and Sons) called cone-tainers. The soil-less mix consisted of one part fine sandto 12 parts of a 1:2 peat moss:vermiculite (v ⁄ v) mixtureamended with 1% (w ⁄ v) of 16-8-12 (N:P:K) control-released fertilizer and CaCO3, pH 8Æ0. In the greenhousetrials, cultivar 34-6S RR (also highly susceptible) wasused. Seed was sown in Sunshine #3 potting mix (pH5Æ8–6Æ2, SunGro Horticulture) in 10 cm diameter plasticpots. The plants were kept at 18–23�C with 14 h photo-period and light intensity of 512 lmol m)2 s)1 in thegrowth cabinet and 230 lmol m)2 s)1 in the green-house.

Clubroot galls were collected from infested canolafields in central Alberta. The pathogen population fromthese collections may represent a mixture of pathotypes,but pathotype 3 would certainly predominate (Strelkov

Plant Pathology (2011) 60, 566–574

et al., 2006, 2007; Xue et al., 2008). Galls were air driedand stored at )15�C until required. To extract restingspores, about 3 g of dried galls were soaked in 150 mLdistilled water for 2 h to soften the tissue and then macer-ated in a blender at high speed for 2 min. The resultingslurry was filtered through four layers of 0Æ3 mm nyloncloth and the spore concentration estimated using ahaemocytometer. In the greenhouse trials, naturallyinfested field soils were also used as a less artificial meansof infestation.

The commercial formulation of the biocontrol agentsB. subtilis, G. catenulatum, S. griseoviridis, S. lydicus,T. harzianum, G. virens and B. subtilis var. amylolique-faciens were tested under growth cabinet conditions.The latter two agents were not included in greenhousetrials due to their poor efficacy in initial tests and non-registration status in Canada. The fungicide fluazinam[3-chloro-N-[3-chloro-2,6-dinitro-4-(trifluoromethyl)phenyl]-5-(trifluoromethyl)-2-pyridinamine] was effec-tive against clubroot in crucifer vegetable crops (Cheahet al., 1998; Donald et al., 2002) and has been registeredrecently in Canada for clubroot control on vegetablecrops. Cyazofamid [4-chloro-2-cyano-N,N-dimethyl-5-p-tolylimidazole-1-sulfonamid] is a relatively new fungi-cide with specific activity against Oomycetes (Takeshiet al., 2004) and high efficacy against clubroot (Mitaniet al., 2003). Further information on these products andtheir rates used in soil-drench treatments is detailed inTable 1.

Treatment application, plant inoculation andassessment of clubroot severity

In controlled environment conditions, all of the treat-ments were applied as a soil drench at 50 mL per plantunless stated otherwise. Each plant was inoculated with5 · 107 P. brassicae resting spores (about 2Æ5 · 106 g)1

growth medium) in the growth cabinet trials, and 1 · 107

or 2 · 107 resting spores (about 1Æ5 or 3 · 105 g)1

growth medium) in the greenhouse trials by pipettingresting-spore suspensions into growth media around theplant. In initial growth cabinet trials, the products wereapplied prior to the pathogen inoculum to allow earlierestablishment of biocontrol agents. In later trials, growthmedia were generally infested with the pathogen before atreatment was applied to better mimic infestation of fieldsoils. The timing of treatment application and plant inoc-ulation is described for specific trials below. The inocu-lated plants were watered with acidified water (pH 6Æ3),and the soil in each pot was kept highly moist by dailywatering. At 4–8 weeks after inoculation, each root wasrinsed with tap water and assessed for clubroot severityusing a 0–3 scale (Strelkov et al., 2006), where 0 = nogalling; 1 = small galls only, on <1 ⁄ 3 of roots; 2 = smallor medium-sized galls on 1 ⁄ 3 to 2 ⁄ 3 of roots; and3 = severe galling, medium- to large-sized galls with morethan 2 ⁄ 3 of roots affected. A disease severity index (DSI)was calculated over all of the plants of a treatment usingthe following formula:

Table 1 Biocontrol agents and fungicides selected for assessment of clubroot control, caused by Plasmodiophora brassicae, under controlled environment

conditions

Microbial species or chemical

common name Trade name Formulationa Rate (a.i. L)1)b Registrant information

Bacillus subtilis Serenade ASO SL 5Æ0 · 1010 cfu AgraQuest Inc., Davis, CA, USA

Bacillus subtilis var. amyloliquefaciens Taegro WP 5Æ0 · 1010 cfu Novozymes Biologicals Inc., Salem, VA USA

Gliocladium catenulatum Prestop WP 1Æ5 · 109 cfu Verdera Oy, Espoo, Finland

Streptomyces griseoviridis Mycostop WP 2Æ5 · 108 cfu Verdera Oy, Espoo, Finland

Gliocladium virens SoilGard 12G G 1Æ5 · 107 cfu Certis USA L.L.C., Columbia, MD, USA

Streptomyces lydicus Actinovate AG G 2Æ5 · 107 cfu Natural Industries Inc., Houston, TX, USA

Trichoderma harzianum Root Shield WP 1Æ5 · 108 cfu BioWorks Inc., Victor, NY USA

Fluazinam Allegro 500 F SL 0Æ25 g ISK Biosciences Corp., Concord, OH, USA

Cyazofamid Ranman 400SC SC 0Æ22 g ISK Biosciences Corp., Concord, OH, USA

aProduct formulation: SL: soluble (liquid) concentrate; WP: wettable powder; G: granules; SC: suspension concentrate.bRates were based on the active ingredient (a.i.) recommended by manufacturers: cfu for biocontrol agents and g for the fungicide.

568 G. Peng et al.

DSIð%Þ ¼P½ðrating classÞðno. plants in the rating classÞ�

ðno. plants in treatmentÞð3Þ � 100

Efficacy trials under growth cabinet conditions

In these initial trials, all of the biocontrol and fungicideproducts were assessed against the pathogen control.Because clubroot had not been reported on canola in theprovince of Saskatchewan, these trials were carried out ina level-2 containment facility at the Saskatoon ResearchCenter to ensure that the pathogen did not escape. Spacein the containment facility was very limited, so eachexperimental unit consisted of seven plants. The studywas laid out in randomized complete block design, withthe blocks being three separate time periods. Each prod-uct was applied as a soil drench at 3 days after seeding,and suspensions of P. brassicae resting spores were addedat 7 days after seeding. The plants were inoculated afterthe treatments were applied to provide an opportunityfor the biocontrol agents to become established beforebeing challenged. The plants were maintained in a growthcabinet for 4–5 weeks and then rated for clubroot inci-dence and severity.

Efficacy trials under greenhouse conditions

Separate experiments were carried out to assess theefficacy of selected products when delivered as asoil-drench or seed treatment. In each experiment, twosources of pathogen inoculum were used: (i) naturallyinfested field soils were mixed with non-infested soil-less mix at 1:1 and 1:2 (v ⁄ v) for trials 1 and 2, respec-tively and (ii) applying resting spores to non-infestedsoil-less mix at 7 days after seeding (1 h prior to treat-ment), at 2 · 107 spores per plant in trial 1 and 1 · 107

spores per plant in trial 2. The field soil was diluted withsoil-less mix because of high clubroot DSI (almost100%) observed in canola plots at the field location. Alltreatments were applied 7 days after seeding for both

naturally infested soil and artificially inoculated soil-lessmix (1 h after inoculation).

For the soil-drench application, products were appliedat 50 mL per pot. For the seed-treatment application,canola seeds were immersed in a product preparation at10· the soil-drench concentrations for 5 min and then airdried for 1 h prior to seeding. For each of the greenhousetrials, the study was laid out in a randomized completeblock design with four replicates and 10 plants per repli-cate. The plants were maintained in a greenhouse for8 weeks and then assessed for clubroot incidence andseverity. Each greenhouse trial was repeated.

Effect of water volume on efficacy

The soil-drench treatment at 50 mL per plant representsa large volume of water when translated to a field scale(equivalent to 42 000 L ha)1 in an ‘in-furrow’ field appli-cation), and would be impractical for product delivery incanola crops. To assess the possibility of using a reducedcarrier volume, a commercial formulation of B. subtilis(13 L ha)1), G. catenulatum (1Æ4 kg ha)1), fluazinum(2Æ9 L ha)1) and cyazofamid (0Æ54 L ha)1) were appliedas an in-furrow drench at about 500, 2500 and12 500 L ha)1 of water volume shortly after seeds weredropped. Soil-less mix in each conetainer was infestedwith 5 · 107 P. brassicae resting spores 2 days prior toseeding and watered immediately after sowing with 15–30 mL of acidified water (pH = 6Æ3), depending on thewater volume applied, to provide a similar level of soilmoisture to all treatments. Thereafter plants of eachtreatment were watered as described earlier. The trialwas laid out in a randomized complete block design withtwo blocks (replicates in time) and seven plants per exper-imental unit. Each plant was assessed for clubroot sever-ity 4–5 weeks after seeding.

Field trials

In 2009, two trials of canola and one of Chinese cabbage(B. rapa subsp. chinensis var. utilis) were conducted toassess the efficacy of four products, based on the efficacy

Plant Pathology (2011) 60, 566–574

Biocontrol of clubroot 569

observed in previous trials, against clubroot under fieldconditions. These treatments were applied as a liquid intothe seed furrow at 5Æ2 · 1013 colony forming units(cfu) ha)1 of B. subtilis, 1Æ4 · 1011 cfu ha)1 of G. cate-nulatum, 1Æ45 kg active ingredient (a.i.) ha)1of fluazi-num, or 0Æ22 kg a.i. ha)1 of cyazofamid formulationdescribed in Table 1. The liquid volume delivered wasabout 500 L ha)1. The canola trials were conducted intwo commercial fields with heavy clubroot infestationnear Edmonton, Alberta. Pathotype 3 of P. brassicae pre-dominates at these sites (Strelkov et al., 2007). The Chi-nese cabbage trial was conducted at the Muck CropResearch Station, University of Guelph near Bradford,Ontario, where the soil is infested with pathotype 6 ofclubroot pathogen (Cao et al., 2009).

The trials were arranged in a randomized completeblock design with four replicates. Each plot consisted offour 6 m-long rows with 21 cm row spacing for canolaand 44 cm for Chinese cabbage, sown at 6Æ5 kg ofseed ha)1 at about 2Æ5 cm depth. Susceptible (S) andresistant (R) lines were used for both canola and Chinesecabbage trials; cvs. 45H26 (S) and 45H29 (R) for canola(source: Pioneer Hi-Bred Canada), and Mirako (S) andYuki (R) for Chinese cabbage (source: Mirako – BejoSeeds Inc.; Yuki – Stokes Seeds Ltd.). All of the productswere applied in-furrow at 500 L ha)1 using a calibratedbackpack sprayer. Water was applied to the control.Clubroot severity was assessed at full bloom in canolaand at 8 weeks in Chinese cabbage after sowing, by dig-ging about 25 plants from the central 4 m area of eachplot and rating each plant as described previously.

Table 2 Effect of biocontrol agents and fungicides applied as a soil drench

on the disease severity index (DSI) of clubroot, caused by Plasmodiophora

brassicae, on canola grown under growth cabinet conditions (n = 3)

Treatment DSI (%)

Clubroot

suppression (%)

Inoculated control 37 abca 0

Biocontrol agent

Trichoderma harzianum 47 a )27

Streptomyces lydicus 40 ab )8

Bacillus subtilis var.

amyloliquefaciens

29 bcd 22

Gliocladium virens 24 bcde 35

Streptomyces griseoviridis 14 defg 61

Gliocladium catenulatum 7 efg 81

Bacillus subtilis 3 fg 91

Fungicide

Effect of soil dryness duration on efficacy

In a growth cabinet trial to assess the impact of early sea-son drought on treatment efficacy, B. subtilis, G. catenul-atum, fluazinum and cyazofamid formulations at therates described above were applied ‘in furrow’ in500 L ha)1 water to canola at seeding. Resting spores ofP. brassicae had been applied to the growth medium2 days prior to sowing. The duration of dryness treat-ments were applied by delaying additional watering afterseeding for 0, 1, 2, 3 and 4 weeks to simulate the effect ofdelayed rain events under field conditions. The averagewater content of the growth medium was 11Æ42% (wateractivity: 0Æ678 ± 0Æ011 at 20�C) prior to regular water-ing. Inoculated but nontreated plants served as a controlat each interval of delayed watering. Plants were grownin conetainers arranged in a randomized complete blockdesign with two blocks (replicates in time) and sevenplants per experimental unit. Clubroot severity wasassessed 4 weeks after watering was initiated for eachtreatment. Watering provided the high soil moisture con-ditions for seed germination and root infection.

Fluazinam 3 fg 91

Cyazofamid 3 fg 91

aMeans followed by the same letter(s) do not differ based on a

protected LSD test at P £ 0Æ05).

Data analysis

For each trial, homogeneity of variance was assessedusing Bartlett’s Test. PROC GLM in SAS (Statistical Analysis

Plant Pathology (2011) 60, 566–574

System (version 8.2), SAS Institute Inc.) was used foranalysis of variance, and means were separated based onFisher’s Protected LSD at P £ 0Æ05. PROC UNIVARIATE wasused to assess the normality of the data, and percentagedata (DSI) were often transferred using arcsine squareroot to improve the efficiency of the analysis (but meansare presented in pre-transformed forms for ease of com-parison). DSI values were calculated for each replicate(seven plants for growth cabinet trials, 10 plants forgreenhouse trials), and used as the response variable forsubsequent analysis. Data for each pathogen inoculumsource (inoculation versus infested field soil) were analy-sed separately due to a lack of homogeneity of varianceamong the trials.

Results

Efficacy assessment and water-volume effect undercontrolled conditions

There was no evidence of phytotoxicity on canola plantstreated with the biocontrol agent or fungicide formula-tions in any of the trials. Clubroot severity was low tomoderate (mean DSI < 50%) in the inoculated control inall three replications of the growth cabinet trial. The syn-thetic fungicides were highly effective, reducing DSI by91% relative to the inoculated control (Table 2). The for-mulation of B. subtilis, G. catenulatum and S. griseoviri-dis reduced DSI by 91–61%, but the other biocontrolagents had little or no effect.

In trial 1 of the soil-drench study in the greenhouse,clubroot levels in the control were high (DSI = 100%) forboth naturally infested field soil and inoculated soil-lessmix (Table 3). Fluazinam and cyazofamid reduced club-root severity, but the biofungicides were much less effec-tive than in the growth cabinet trials. In trial 2, pathogenpressure was reduced by increasing the dilution of the

Table 3 Effect of biocontrol agents and fungicides applied as a soil drench

on the disease severity index of clubroot, caused by Plasmodiophora

brassicae, on canola grown under greenhouse conditions (n = 4)

Treatment

Infested field

soil

Inoculated

soil-less mix

Trial 1 Trial 2 Trial 1 Trial 2

Inoculated control 100 aa 98 a 100 a 76 a

Biocontrol agent

Streptomyces griseoviridis 93 ab 77 b 93 ab 33 b

Bacillus subtilis 90 abc 68 b 88 bc 3 e

Streptomyces lydicus 90 abc 49 c 86 bc 8 de

Trichoderma harzianum 86 bc 45 cd 91 abc 23 c

Gliocladium catenulatum 85 c 36 de 88 bc 13 cd

Fungicide

Fluazinam 28 d 23 f 0 d 0 e

Cyazofamid 23 d 11 g 0 d 0 e

aMeans in a column followed by the same letter(s) do not differ

based on a protected LSD test at P £ 0Æ05.

Table 4 Effect of biocontrol agents and fungicides applied as a seed

treatment on the disease severity index of clubroot on canola grown under

greenhouse conditions (n = 4)

Treatment

Infested field soil

Inoculated

soil-less mix

Trial 1 Trial 2 Trial 1 Trial 2

Inoculated control 100 aa 80 a 100 a 75 a

Biocontrol agent

Streptomyces griseoviridis 96 ab 56 bcd 97 ab 34 bc

Trichoderma harzianum 99 a 68 ab 95 abc 28 c

Bacillus subtilis 94 abc 50 cd 93 abcd 43 b

Gliocladium catenulatum 92 abc 61 bc 88 bcd 33 bc

Streptomyces lydicus 91 abc 59 bc 100 a 35 bc

Fungicide

Fluazinam 89 bc 40 de 84 cd 7 d

Cyazofamid 75 d 34 e 83 d 7 d

aMeans in a column followed by the same letter(s) do not differ

based on a protected LSD test at P £ 0Æ05.

Table 5 Effect of carrier volume of biocontrol agents and fungicides on the

disease severity index (DSI) of clubroot, caused by Plasmodiophora

brassicae, on canola (n = 2)

Product and rate (formulation ha)1)

Water

volume

(L ha)1)a DSI (%)

Inoculated control 2500 100 ab

Biocontrol agent

Bacillus subtilis (13 L) 12 500 7 b

2500 2 b

500 14 b

Gliocladium catenulatum (1Æ4 kg) 12 500 7 b

2500 0 b

500 16 b

Fungicide

Fluazinam (2Æ9 L) 12 500 0

2500 0

500 0

Cyazofamid (0Æ54 L) 12 500 0

2500 0

500 0

aThe water volume was based on the surface area of each pot, but

the amount of product delivered was the same for each treatment.bMeans are calculated across the two replicates (in time) of the trial.

Means followed by the same letter do not differ based on a

protected LSD test at P £ 0Æ05.

570 G. Peng et al.

field soil and reducing the number of resting sporesapplied to the soil-less mix. This resulted in a lower DSIon the control, especially in the soil-less mix. All ofthe treatments were substantially more effective than intrial 1 (Table 3).

The same pattern of response was observed in the seed-treatment experiment. In the inoculated control in thetrial 1, DSI was 100% and none of the biocontrol agentswere effective, and the fungicides were only marginallysuppressive (Table 4). When the inoculum pressure wasreduced in trial 2, all of the treatments were effective,although the efficacy was substantially lower than in therespective soil-drench treatment. Overall, the syntheticfungicides were more effective than the biocontrol agentsin both the soil-drench and seed-treatment applicationsin these greenhouse trials.

Under growth cabinet conditions, water volumes from500 to 12 500 L ha)1 had no effect on product efficacy.Low levels of clubroot developed in the biocontrol agenttreatments, but no clubroot developed in the treatmentwith synthetic fungicide (Table 5).

Field trials

A canola trial seeded on May 30, 2009 was abandoneddue to extremely poor emergence caused by droughtconditions before and after seeding. The trial seeded at asecond site on June 1, 2009 also experienced 4-weekpost-seeding drought, but showers in July helped manyseedlings to eventually emerge. Severe clubroot haddeveloped by the end of the season, despite the dry condi-tions, with 78% DSI on the nontreated susceptible culti-var. None of the fungicide ⁄ biocontrol agent treatmentshad substantial impact on clubroot severity, with DSIranging from 60 to 65Æ4%. The average DSI was muchlower on the resistant canola cultivar (0)11Æ3%) than onthe susceptible cultivar.

In the trial on Chinese cabbage, precipitation that fell2 days after seeding allowed seedlings to emerge quickly.Clubroot severity on the susceptible cultivar at harvestwas moderate (DSI = 44Æ8%). The fungicide ⁄ biofungi-cide treatments reduced the DSI to 6Æ9–20Æ36% (Fig. 1),which was about 85–55% disease suppression relativeto the control. There was no substantial difference inefficacy among the four products tested. DSI on resistantcv. Yuki was 99% lower than on the susceptible cv.Mirako, with only a few plants displaying any clubrootsymptoms.

Plant Pathology (2011) 60, 566–574

Treatment

Ave

rage

dis

ease

sev

erity

inde

x (%

)

0

10

20

30

40

50

60

Cv. Mirako (S)Cv. Yukli (R)

Control G. catenulatum B. subtilis Fluazinum Cyazofamid

Figure 1 Mean clubroot severity, caused by Plasmodiophora

brassicae, on Chinese cabbage as affected by

biofungicide ⁄ fungicide treatments and cultivar resistance near

Bradford, Ontario in 2009 (n = 4).

Duration of dryness (week)0 1 2 3 4

Ave

rage

dis

ease

sev

erity

inde

x (%

)

0

20

40

60

80

100

Inoculated controlG. catenulatumB. subtilisFluazinumCyazofamid

Figure 2 Effect of post-seeding dryness duration on efficacy of

biofungicides ⁄ fungicides against clubroot, caused by

Plasmodiophora brassicae, on canola under growth cabinet

conditions (n = 2).

Biocontrol of clubroot 571

Soil dryness

The impact of post-treatment dryness duration differedamong the products assessed. All of the products main-tained their efficacy for 1 week in dry soil, but efficacy ofB. subtilis and fluazinum products declined substantiallyafter 2 weeks. The fungicide cyazofamid was still effec-tive after 4 weeks in dry soil (Fig. 2).

Discussion

In the controlled environment trials on canola whereinoculum pressure was moderate, several of the biocon-trol agents reduced clubroot severity when applied as asoil drench. They were not effective when inoculum pres-sure was high. For example, in trial 1 of the greenhousetrial (DSI = 100% in the control), none of them reducedclubroot severity. In contrast, when pathogen inoculum

Plant Pathology (2011) 60, 566–574

was reduced by about 50% in trial 2, the B. subtilis andS. lydicus formulation provided a similar reduction inclubroot severity to that of the fungicides fluazinam andcyazofamid in soil-less mix. Similarly, in growth cabinettrials where DSI in the control was generally below 50%,the B. subtilis and G. catenulatum treatments were aseffective as the fungicides.

Narisawa et al. (2005) found that control of clubrooton Chinese cabbage using the endophytic fungus Hetero-conium chaetospira was more consistently effective whenpathogen inoculum was at or below 1 · 105 restingspores per g of soil, especially under high soil moistureconditions. The low efficacy of the biocontrol agentsunder high disease pressure indicates that it will likely benecessary to use them as components of an integrated dis-ease management system (Donald et al., 2006; ratherthan using them on their own. Used in combination withcultural ⁄ agronomic measures such as crop rotation (Wal-lenhammar, 1996; Cheah et al., 2006) and soil amend-ment (Webster & Dixon, 1991a; Dixon & Page, 1998)that help reduce pathogen load in the soil or infectionpotential, may allow the natural products to reach theirefficacy potential. By comparison, inoculum load had asmaller impact on the efficacy of the fungicides fluazinamand cyazofamid.

In initial growth cabinet trials, biocontrol agents wereapplied prior to the pathogen and this timing probablyallowed them to achieve maximum efficacy potentialbecause they could establish in the rhizosphere andoccupy the surface of roots or root hairs before the patho-gen. In the greenhouse trials, the treatments were appliedto naturally infested soils 7 days after seeding to targetthe peak of zoospore production and infection (Asanoet al., 2000)., This treatment timing, however, wasnumerically less effective than applications prior to path-ogen inoculation for all of the products, including the fun-gicides. It is likely that in the infested soils some earlyinfection might have already occurred by the time oftreatment (7 days after seeding), thus lowering the effi-cacy. It may also be possible that the treatments are moreeffective against zoospore infection than inactivatingpathogen plasmodium established in roots. Therefore,timing biocontrol agents prior to root infection by theclubroot pathogen increases the efficacy of treatment.When in subsequent trials, these were applied to infestedgrowth media at sowing more effective control wasachieved.

The lower clubroot severity in the initial growth cabi-net trials (compared to the greenhouse trials) was proba-bly caused by the high pH (8Æ0) of the growth mediumused. Application of acidified water (pH = 6Æ3) aftertreatment did not change the soil pH substantially (datanot shown). This issue was resolved in later trials byswitching to a commercial growth medium (Sunshine #3,pH = 5Æ8–6Æ2), which generally allowed clubroot severityto reach a level >90% in the pathogen controls. This expe-rience also underlines the effectiveness of enhancing soilpH to alleviate clubroot impact as described previously(Webster & Dixon, 1991a,b; Dixon & Page, 1998), even

572 G. Peng et al.

under high disease pressure conditions. For many canola-growing areas in western Canada where soil pH levels areclose to or above 8Æ0, there may be substantial antago-nism in the soil against clubroot, which may complementbiocontrol agents in clubroot control. Although the directimpact of soil pH on efficacy of B. subtilis and G. catenul-atum is not well understood, other studies have foundthat metabolites synthesized by biocontrol agents can beinfluenced by pH (Slininger & Shea-Wibur, 1995;Moreno-Mateos et al., 2007). This aspect requiresfurther investigation for B. subtilis and G. catenulatum.

In the field trials, all of the products reduced clubrootseverity on the susceptible cultivar of Chinese cabbage,but were less effective on the susceptible cultivar ofcanola. It is also noteworthy that the biocontrol agentsshowed similar efficacy to that of fungicides in field trials,which underscores their potential. Clubroot severity wasmoderate in the nontreated control of susceptible Chinesecabbage, but severe in the control of susceptible canola.The greater efficacy on Chinese cabbage than on canolasupports the observation in the controlled environmenttrials that biocontrol agents are more effective underlower inoculum ⁄ disease pressure conditions. Anotherpossible cause of the relatively low efficacy of all of theproducts in the canola trial may be the prolonged soil dry-ness experienced after seeding. It is likely that when rainfinally fell at 4 weeks after sowing, little or no activeingredient of the products remained in the soil. Thishypothesis was supported by the impact of dry conditionsafter sowing on product efficacy observed under con-trolled environment conditions. The efficacy of most ofthe products dropped substantially after being held in drysoil for more than 1 week. The exception was the fungi-cide cyazofamid, which was still highly effective after4 weeks. Caution is required when extrapolating from acontrolled environment study to field conditions.Although the water activity of the growth medium(0Æ678) represents very dry soil conditions in the field, noextremes or fluctuation of temperature were used underthe controlled environment condition. Additionalresearch is required to better understand product behav-iour in the soil, but detailed assessments of the persistenceof these agents under dry conditions were beyond thescope of the current study.

Seed treatment with biocontrol agents or fungicidesreduced clubroot severity when inoculum pressure waslow, but soil-drench treatments were consistently moreeffective. In most systems involving biocontrol agents,the extent of rhizosphere colonization by the biocontrolagent is important in their efficacy against soil-bornepathogens (Yang et al., 2009; Begum et al., 2010). Seedtreatments deliver only a small fraction of the number ofliving propagules per seedling compared to a soil drenchapplication, but would be more practical than a soildrench for use in the extensive agricultural systememployed for canola production. Also, formulations ofthese products have not been optimized for seed treat-ment application, so the observation that seed treatmentsshowed some activity is encouraging. It is possible that

products delivered as a seed treatment can be formulatedto help them colonize the emerging roots ⁄ root hairs andmove efficiently with the growing points of seedlingroots.

The mechanism(s) that underlie the activity of thesebiocontrol agents for clubroot control has not been deter-mined, but antibiosis is a likely candidate. Hyphaldegrading enzymes have been identified in the rhizo-sphere treated with the commercial formulation ofG. catenulatum (Chatterton & Punja, 2009, 2010) andthe antibiotics surfactin and iturin have been reported tooccur in the rhizosphere treated with the formulation ofB. subtilis (Kinsella et al., 2009). In a preliminary study,cell-free filtrates of the G. catenulatum and B. subtilis for-mulations, as well as suspensions of pure fungal or bacte-rial cultures, all reduced clubroot severity on canolaunder controlled environment conditions (Peng et al.,2010). This strongly indicates that both biocontrol agentsproduce antimicrobial metabolites against clubroot, atleast under cultural conditions. These metabolites did notinhibit germination of resting spores (Lahlali et al.,2010), but more likely interact with pathogen zoospores.It is not clear if these agents are able to penetrate roots ofcrucifer plants and stimulate them against clubroot orother soil-borne diseases as reported with other endo-phytes (Morita et al., 2003; Bacon & Hinton, 2007).These modes of action are being investigated, and a betterunderstanding of them will provide useful informationfor designing efficient product-delivery strategies againstclubroot.

The resistant lines of both canola and Chinese cabbageprovided effective reductions in clubroot severity underfield conditions. For example, the resistant cultivar ofcanola had about 95% less disease than the susceptiblecultivar under high disease pressure. These observationsprovide strong support for using genetic resistance as acornerstone in a clubroot management system. It is recog-nized that sources of genetic resistance are very limited(Hirai, 2006; Diederichsen et al., 2009) and resistancecan also be eroded quickly (Oxley, 2007). Therefore, it isbelieved that cultivar resistance should be employed judi-ciously to prolong the effectiveness and enhance efficacy(Donald & Porter, 2009). An integrated approach shouldbe considered for sustainable management of clubroot(Dixon, 2003; Donald et al., 2006). In Canada, cultivarresistance is likely to be the mainstay for this managementpackage, but additional measures including soil nutrientmanagement (Dixon & Page, 1998; Webster and Dixon,1991a, b), seeding date selection (McDonald & Wester-veld, 2008; Gossen et al., 2009), crop rotation (Wallen-hammar, 1996; Cheah et al., 2006), and chemical orbiological control (Suzuki et al., 1995; Narisawa et al.,1998; Cheah et al., 2000; Takeshi et al., 2004) should allbe included as options in the management tool box.

Acknowledgements

We thank Turnbull GD and Manolii VP for technicalassistance, and SaskFlax, the Alberta Canola Growers

Plant Pathology (2011) 60, 566–574

Biocontrol of clubroot 573

Commission, the Canola Council of Canada, and the PestManagement Centre of AAFC for financial support.

References

Arie T, Kobayashi Y, Okada G, Kono Y, Yamaguchi I, 1998.

Control of soilborne clubroot disease of cruciferous plants by

epoxydon from Phoma glomerata. Plant Pathology 47,

743–8.

Arie T, Kobayashi Y, Kono Y, Gen O, Yamaguchi I, 1999. Control

of clubroot of crucifers by Phoma glomerata and its product

epoxydon. Pesticide Science 55, 602–4.

Asano T, Kageyama K, Hyakumachi M, 2000. Germination of

surface-disinfected resting spores of Plasmodiophora brassicae

and their root hair infection in turnip hairy roots. Mycoscience

41, 49–54.

Bacon CW, Hinton DM, 2007. Potential for control of seedling

blight of wheat caused by Fusarium graminearum and related

species using the bacterial endophyte Bacillus mojavensis.

Biocontrol Science and Technology 17, 81–94.

Begum MM, Sariah M, Puteh AB, Zainal Abidin MA,

Rahman MA, Siddiqui Y, 2010. Field performance of bio-

primed seeds to suppress Colletotrichum truncatum causing

damping-off and seedling stand of soybean. Biological

Control 53, 18–23.

Cao T, Manolii VP, Hwang SF, Howard RJ, Strelkov SE, 2009.

Virulence and spread of Plasmodiophora brassicae [clubroot] in

Alberta, Canada. Canadian Journal of Plant Pathology 31,

321–9.

Chatterton S, Punja ZK, 2009. Chitinase and b-1,3-glucanase

enzyme production by the mycoparasite Clonostachys rosea f.

catenulata against fungal plant pathogens. Canadian Journal of

Microbiology 55, 356–67.

Chatterton S, Punja ZK, 2010. Factors influencing colonization of

cucumber roots by Clonostachys rosea f. catenulata, a biological

disease control agent. Biocontrol Science and Technology 20,

37–55.

Cheah LH, Page BBC, Koolaard JP, 1998. Soil-incorporation of

fungicides for control of clubroot of vegetable brassicas. In: Pro-

ceedings of the 51st New Zealand Plant Protection

Conference, August 11–13, 1998. Hamilton, New Zealand:

New Zealand Plant Protection Society, 130–3.

Cheah LH, Veerakone S, Kent G, 2000. Biological control of

clubroot on cauliflower with Trichoderma and Streptomyces

spp. New Zealand Plant Protection 53, 18–21.

Cheah LH, Gowers S, Marsh AT, 2006. Clubroot control using

Brassica break crops. Acta Horticulturae 706, 329–32.

Diederichsen E, Beckmann J, Schondelmeier J, Dreyer F, 2006.

Genetics of clubroot resistance in Brassica napus ‘Mendel’. Acta

Horticulturae 706, 307–11.

Diederichsen E, Frauen M, Linders EGA, Hatakeyama EK, Hirai

EM, 2009. Status and perspectives of clubroot resistance

breeding in crucifer crops. Journal of Plant Growth

Regulation 28, 265–81.

Dixon GR, 2003. Clubroot, a case for integrated control.

Plantsman 2, 179–83.

Dixon GR, Page LV, 1998. Calcium and nitrogen eliciting

alterations to growth and reproduction of Plasmodiophora

brassicae (clubroot). Acta Horticulturae 459, 343–9.

Donald C, Porter I, 2009. Integrated control of clubroot. Journal

of Plant Growth Regulation 28, 289–303.

Plant Pathology (2011) 60, 566–574

Donald EC, Porter IJ, Lancaster RA, 2002. Strategic application of

lime, fertilisers and fungicides for improved control of clubroot.

Cruciferae Newsletter 24, 81–2.

Donald EC, Porter IJ, Faggian R, Lancaster RA, 2006. An

integrated approach to the control of clubroot in vegetable

Brassica crops. Acta Horticulturae 706, 283–300.

Gossen BD, McDonald MR, Hwang SF, Kalpana KC, Peng G,

2009. Managing seeding date to minimize clubroot

(Plasmodiophora brassicae) damage in canola and vegetable

brassicas. In: Proceedings of International Clubroot Workshop

2009. Kunming, China, 34–40.

Gossen BD, McDonald MR, Kalpana KC, Hwang SF, Strelkov SE,

Peng G, 2010. Impact of climate change on clubroot of canola

on the Canadian prairies. In: Proceedings of the 81st Annual

Meeting of the Canadian Phytopathological Society with

the Pacific Division of the American Phytopathological

Society, 20–23 June 2010, Vancouver, Canada, 50.

Hashiba T, Morita S, Narisawa K, Usuki K, 2003. Mechanisms of

symbiosis and disease suppression of root endophytic fungus.

Nippon Bogeikagaku Kaishi – Japan Society for Bioscience,

Biotechnology and Agrochemistry 77, 130–3.

Hirai M, 2006. Genetic analysis of clubroot resistance in Brassica

crops. Breeding Science 56, 223–9.

Joo GJ, Kim YM, Kim JW et al., 2004. Biocontrol of cabbage

clubroot by the organic fertilizer using Streptomyces sp. AC-3.

Korean Journal of Microbiology and Biotechnology

32, 172–8.

KimS,ShinC,MoonS et al., 2004. Isolationandcharacterization

ofStreptomyces sp.KACC91027againstPlasmodiophora

brassicae. Journal of Microbiology and Biotechnology 14,

220–3.

Kinsella K, Schulthess CP, Morris TF, Stuart JD, 2009. Rapid

quantification of Bacillus subtilis antibiotics in the rhizosphere.

Soil Biology and Biochemistry 41, 374–9.

Lahlali R, Peng G, Gossen BD et al., 2010. Investigation on modes

of action for the biofungicides Serenade and Prestop in clubroot

control. In: Proceedings of the 81st Annual Meeting of the

Canadian Phytopathological Society with the Pacific Division

of the American Phytopathological Society, 20–23 June

2010, Vancouver, Canada, 53–4.

McDonald MR, Westerveld SM, 2008. Temperature prior to

harvest influences the incidence and severity of clubroot on two

Asian brassica vegetables. HortScience 43, 1509–13.

McDonald MR, Kornatowska B, McKeown AW, 2004.

Management of clubroot of Asian Brassica crops grown on

organic soils. Acta Horticulturae 635, 25–30.

Mitani S, Sugimoto K, Hayashi H, Takii Y, Ohshima T, Matsuo

N, 2003. Effects of cyazofamid against Plasmodiophora

brassicae Woronin on Chinese cabbage. Pest Management

Science 59, 287–93.

Moreno-Mateos MA, Delgado-Jarana J, Codon AC, Benıtez T,

2007. pH and Pac1 control development and antifungal activity

in Trichoderma harzianum. Fungal Genetics and Biology 44,

1355–67.

Morita S, Azuma M, Aoba T, Satou H, Narisawa K, Hashiba T,

2003. Induced systemic resistance of Chinese cabbage to

bacterial leaf spot and Alternaria leaf spot by the root endophytic

fungus, Heteroconium chaetospira. Journal of General Plant

Pathology 69, 71–5.

Narisawa K, Tokumasu S, Hashiba T, 1998. Suppression of

clubroot formation in Chinese cabbage by the root endophytic

574 G. Peng et al.

fungus, Heteroconium chaetospira. Plant Pathology 47, 206–

10.

Narisawa K, Shimura M, Usuki F, Fukuhara S, Hashiba T, 2005.

Effects of pathogen density, soil moisture, and soil pH on

biological control of clubroot in Chinese cabbage by

Heteroconium chaetospira. Plant Disease 89, 285–90.

Oxley S, 2007. Clubroot Disease of Oilseed Rape and other

Brassica Crops. Edinburgh, UK: Scottish Agriculture College:

Technical Note, TN 620.

Peng G, Gossen BD, Strelkov SE, Hwang SF, McDonald MR,

2009. Effect of selected biofungicides for control of clubroot on

canola. Canadian Journal of Plant Pathology 31, 145–6.

Peng G, Hwang SF, McDonald MR et al., 2010. Control of

clubroot (Plasmodiophora brassicae) with microbial and

synthetic fungicides. Phytopathology 100, S99.

Slininger PJ, Shea-Wibur MA, 1995. Liquid-culture pH,

temperature, and carbon (not nitrogen) source regulate

phenazine productivity of the take-all biocontrol agent

Pseudomonas fluorescens 2-79. Applied Microbiology and

Biotechnology 43, 794–800.

Strelkov SE, Tewari JP, Smith-Degenhardt E, 2006.

Characterization of Plasmodiophora brassicae populations from

Alberta, Canada. Canadian Journal of Plant Pathology 28,

467–74.

Strelkov SE, Manolii VP, Cao T, Xue S, Hwang SF, 2007.

Pathotype classification of Plasmodiophora brassicae and its

occurrence in Brassica napus in Alberta, Canada. Journal of

Phytopathology 155, 706–12.

Strelkov SE, Manolii VP, Zequera M, Manolii E, Hwang SF, 2010.

Incidence of clubroot on canola in Alberta in 2009. Canadian

Plant Disease Survey 90, 123–5.

Suzuki K, Sugimoto K, Hayashi H, Komyoji T, 1995. Biological

mode of action of fluazinam, a new fungicide for Chinese

cabbage clubroot. Annals of the Phytopathological Society of

Japan 61, 395–8.

Takeshi O, Terumasa K, Shigeru M, Norifusa M, Toshio N, 2004.

Development of a novel fungicide, cyazofamid. Journal of

Pesticide Science 29, 147–52.

Tewari JP, Strelkov SE, Orchard D, Hartman M, Lange R,

Turkington TK, 2005. Identification of clubroot of crucifers on

canola (Brassica napus) in Alberta. Canadian Journal of Plant

Pathology 27, 143–4.

Usuki F, Narisawa K, 2007. A mutualistic symbiosis between a

dark septate endophytic fungus, Heteroconium chaetospira, and

a non-mycorrhizal plant, Chinese cabbage. Mycologia 99,

175–84.

Wallenhammar AC, 1996. Prevalence of Plasmodiophora brassicae

in a spring oilseed rape growing area in central Sweden and

factors influencing soil infestation levels. Plant Pathology 45,

710–9.

Wallenhammar AC, Johnsson L, Gerhardson B, 2000. Agronomic

performance of partly clubroot-resistant spring oilseed turnip

rape lines. Journal of Phytopathology 148, 495–9.

Webster MA, Dixon GR, 1991a. Boron, pH and inoculum

concentration influencing colonization by Plasmodiophora

brassicae. Mycological Research 95, 74–9.

Webster MA, Dixon GR, 1991b. Calcium, pH and inoculum

concentration influencing colonization by Plasmodiophora

brassicae. Mycological Research 95, 64–73.

Xue C, Cao T, Howard RJ, Hwang SF, Strelkov SE, 2008.

Isolation and variation in virulence of single-spore isolates of

Plasmodiophora brassicae from Canada. Plant Disease 92,

456–62.

Yang J, Kloepper JW, Ryu CM, 2009. Rhizosphere bacteria

help plants tolerate abiotic stress. Trends in Plant Science 14,

1–4.

Plant Pathology (2011) 60, 566–574