Movement of Plasmodiophora brassicae resting spores in windblown dust

Abstract

Clubroot (Plasmodiophora brassicae Wor.) of canola (Brassica napus L.) was first detected in Alberta, Canada, near Edmonton in 2003. Since that time it has become a severe problem in canola production, spreading throughout central Alberta, with isolated cases of the disease also identified in southern Alberta, Saskatchewan and Manitoba. Clubroot is believed to be spreading mostly via the movement of P. brassicae-infested soil on farm machinery. This study examined the potential for the spread of the pathogen in windblown dust from infested fields. Dust samplers were placed at four locations, three in central Alberta and one in southern Alberta, in 2011 and 2012. Soil particles in suspension and moving through saltation were collected in samplers at heights ranging from 0 to 1 m. Total genomic DNA was extracted from the collected soil samples and subjected to conventional polymerase chain reaction (PCR) and quantitative PCR analysis for the detection and quantification of P. brassicae in the samples. The DNA of P. brassicae was detected in samplers at all locations in both years. There was no clear association between sampling height and the frequency of samples testing positive for the presence of P. brassicae DNA, or between sampling height and the amount of P. brassicae DNA. The results indicate that P. brassicae resting spores can be carried by wind-borne dust and that wind-mediated dispersal likely contributes to spread of this pathogen.

Résumé

La hernie (Plasmodiophora brassicae Wor.) du canola (Brassica napus L.) a été initialement découverte en Alberta en 2003, près d’Edmonton. Depuis, se répandant partout dans le centre de l’Alberta, c’est devenu un grave problème en ce qui a trait à la production du canola. Des cas isolés ont également été répertoriés dans le sud de l’Alberta, en Saskatchewan et au Manitoba. On croit que la hernie se répand principalement par le sol contaminé par P. brassicae qui adhère à la machinerie agricole. Cette étude a examiné la possibilité que l’agent pathogène soit disséminé par de la poussière de sol contaminé soufflée par le vent. En 2011 et 2012, des échantillonneurs de poussière ont été installés à quatre endroits, soit trois dans le centre de l’Alberta et un dans le sud. Des particules de sol en suspension et se déplaçant par saltation ont été collectées dans les échantillonneurs à des hauteurs variant de 0 à 1 m. L’ADN génomique total a été extrait des échantillons de sol collectés et soumis à l’analyse par réaction en chaîne de la polymérase (PCR) classique et par PCR quantitative pour y détecter et y quantifier P. brassicae. Au cours des deux années, l’ADN de P. brassicae a été décelé dans les échantillonneurs aux quatre endroits. Il n’y avait pas de lien précis entre la hauteur d’échantillonnage et la fréquence d’occurrence positive d’ADN de P. brassicae dans les échantillons, ou entre la hauteur d’échantillonnage et la quantité d’ADN de P. brassicae collectée. Les résultats indiquent que les spores de conservation peuvent être transportées par des poussières soufflées par le vent et que la dispersion éolienne contribue vraisemblablement à la dissémination de cet agent pathogène.

Keywords:

• Brassica napus
• canola
• clubroot
• dispersal
• dust
• Plasmodiophora brassicae

Mots clés:

• Brassica napus
• canola
• dissémination
• hernie
• Plasmodiophora brassicae
• poussière

Introduction

Clubroot, caused by the obligate parasite Plasmodiophora brassicae Woronin, is an important disease of canola (Brassica napus L.) and other Brassica crops. The disease has rapidly become a major concern in the production of canola in Alberta, Canada. Clubroot was first observed on canola in 2003 in 12 fields in the Edmonton, Alberta area (Tewari et al. 2005). In the following decade, P. brassicae has spread and clubroot disease has now been confirmed in over 1400 canola fields (Strelkov et al. 2014). While the majority of the infested fields are still found in central Alberta, some P. brassicae infestations have been identified in the southern parts of the province (Strelkov et al. 2011), and isolated infestations also have been reported from Saskatchewan and Manitoba (Strelkov & Hwang 2014).

Clubroot development is associated with the formation of large galls or clubs on the roots of susceptible hosts, which impede water and nutrient uptake from the soil and can result in stunting, wilting and even plant death when symptoms are severe. Infection of canola is associated with significant reductions in yield and quality of the crop (Pageau et al. 2006; Howard et al. 2010). Plasmodiophora brassicae forms long-lived resting spores that can survive in the soil for up to 20 years (Wallenhammar 1996) and serve as inoculum for subsequent crops. The resting spores germinate to produce primary zoospores which begin the infection process. Although they are biflagellated and motile in water, zoospore motility in the rhizosphere is limited (Hwang et al. 2012).

The movement of the pathogen is thought to occur most frequently via the movement of P. brassicae-infested soil, typically on farm equipment. Other methods of pathogen spread have been examined or suggested, including its dispersal in dust occurring as an external contaminant of seeds and tubers (Rennie et al. 2011), or via water or water-mediated soil erosion (Datnoff et al. 1984; Howard et al. 2010). One of the main clubroot management strategies in Alberta has focused on limiting the spread of P. brassicae from field to field through the sanitization of field equipment (Alberta Clubroot Management Committee 2008). Frequent equipment sanitation may not be practical, however, on a commercial scale of farm production. Plasmodiophora brassicae has continued to spread in the province, and anecdotal reports have emerged of the likely movement of clubroot through soil erosion from one field to the next. However, the importance of soil erosion as a mechanism of dispersal for P. brassicae is not clear and has not been documented scientifically.

The detection of P. brassicae resting spores in soil can be difficult, as they are microscopic and are suspended in the soil matrix. Traditionally, the presence of P. brassicae in soils has been determined by bait plants, where susceptible plants are grown in potentially infested soil and later examined for clubroot symptoms. Unfortunately, this method is time consuming, requiring partial development of symptoms for detection, and requires spore concentrations exceeding 1000 spores g⁻¹ soil in order to detect the pathogen (Faggian & Strelkov 2009). The development of PCR-based diagnostic techniques has allowed for the more rapid and sensitive detection of P. brassicae spores in soil samples. Rapid and sensitive conventional PCR assays are available that enable the detection of the pathogen at levels as low as 100 spores g⁻¹ soil (Faggian et al. 1999; Cao et al. 2007). Quantitative PCR (qPCR) protocols now allow the detection of P. brassicae in soil samples and determination of the spore concentration in infested soil (Rennie et al. 2011; Wallenhammar et al. 2012). Furthermore, PCR-based techniques allow the detection of P. brassicae in soil samples too small for traditional bioassays.

Given the increasing prevalence of clubroot in Alberta and the anecdotal reports of field to field movement of the disease with soil erosion, the objective of the current study was to evaluate whether or not P. brassicae can be routinely detected in soil moved by wind from clubroot-infested fields and the quantity of pathogen inoculum carried. This information will help to understand the epidemiology of P. brassicae, particularly in the canola cropping systems in western Canada, and help refine clubroot containment strategies.

Materials and methods

Study sites

Four experimental sites were selected for the study. Three of the sites were located in central Alberta, and consisted of a commercial field in Parkland County (‘Parkland’ site, 14 ha), a clubroot nursery located at the Crop Diversification Centre North, Alberta Agriculture and Rural Development, Edmonton (‘CDCN’ site, 5 ha), and research plots located in a commercial field in a rural part of northeast Edmonton (‘50th ST’ site, 2 ha). The fourth site was located in a commercial field in Newell County, near the town of Bassano, in southern Alberta (‘Bassano’ site, 193 ha). Unlike the other sites, the Bassano site was regularly irrigated by a centre pivot irrigation system during the study years.

The CDCN and 50th ST sites consist of typically miscellaneous undifferentiated mineral soils with variable landscape characteristics and soil series distribution (Alberta Soil Information Viewer [Internet] c2001-2014). The Parkland site is an orthic dark grey chernozem with an undulating, high relief landform of limited slope (4%) (WTB4/U1h) (Alberta Soil Information Viewer [Internet] c2001-2014). The Bassano site is a brown solog on medium-textured till with an undulating, high relief landform of limiting slope (4%) (HDY16/U1h) (Alberta Soil Information Viewer [Internet] c2001-2014). All four sites were actively utilized for agricultural purposes for the duration of the study (generally, a canola-cereal rotation). In the Edmonton area, average monthly precipitation between May–August of 2011 and 2012 was approximately 44–342 mm, while the average monthly precipitation in the Bassano area was 40–184 mm for the same period (AgroClimatic Information System. [Internet] 2014).

BSNE sampler set up

Big Spring Number Eight (BSNE) samplers (Custom Products, Big Spring, TX) (Fryrear 1986) were set up at the four research locations in 2011 and 2012. These are passive samplers that consist of an enclosed metal pan with a 10 cm² front inlet. Each sampler has a wind vane allowing the sampler to rotate around the supporting pole and orient the opening into the wind. Airborne dust is deposited into the pan as the wind decelerates through the sampler. In 2011, at the Parkland, CDCN and 50th ST sites, the samplers were placed at 0.1, 0.35, 0.6, 0.8 and 1.05 m above ground level with five samplers per post. Since the prevailing wind direction in the area is from west to east, the majority of the posts were placed on the eastern side of the fields, to allow for more frequent dust collection from the study sites. Four posts were located on the east side of the field at the CDCN and 50th ST sites, and three posts were located on the east side of the field at the Parkland site. One post was placed west of each field at all of the sites. At the Bassano site, there were five posts positioned along the perimeter of a natural gas lease (Fig. 1), and each post had four samplers positioned at 0, 0.25, 0.5 and 0.75 m above ground level. The samplers were kept in the fields from July to October in 2011.

Fig. 1 (Colour online) Location of the Big Spring Number Eight dust sampler (Custom Products, Big Spring, TX) posts and weather stations in: (a) a field research site located in northeast Edmonton (‘50th ST site’), (b) a clubroot nursery in northeast Edmonton (‘CDCN site’), (c) a commercial field in Parkland County (‘Parkland site’) and (d) a commercial field in Newell County, in Southern Alberta (‘Bassano site’). The distribution of P. brassicae resting spores, as assessed by quantitative PCR analysis of soil samples collected in each field, also was analysed for the Parkland and Bassano sites, and is indicated in panels (c) and (d). Green circles indicate no resting spores detected. Beige circles indicate locations infested with P. brassicae, with the diameter of the circle corresponding to the resting spore concentration (spores g⁻¹ soil).

In 2012, the arrangement of the BSNE samplers was modified slightly at the three sites in central Alberta, to get better coverage over the entire lengths of the fields. At CDCN, 50th ST and Parkland, the number of BSNE samplers per post was reduced to three at heights of 0.1, 0.5 and 1 m above ground level. The number of posts was increased to eight at CDCN and 50th ST and to seven at Parkland. Where possible, parity was maintained with respect to the number of posts on the east and west sides of the fields, and were arranged as illustrated in Fig. 1a–c. The BSNE samplers were deployed on 11 May 2012 and were kept in the fields until 26 October 2012. Vegetation was mechanically trimmed throughout the season to prevent obstruction of the samplers.

At Bassano, the number of posts and samplers as well as the sampler heights along the natural gas lease were kept the same as in 2011 (Fig. 1d). Four additional posts were installed in the south-west portion of the field, alongside an irrigation canal (Fig. 1d).

BSNE dust collection and analysis

Dust from the BSNE samplers deployed at the CDCN, 50th ST and Parkland fields was collected approximately every 2 weeks. Dust from the Bassano BSNE samplers was collected twice during the sampling period. The dust was removed by rinsing the interior of the sampler with deionized water and transferring the resulting slurry to a 50 mL conical tube. The tubes were centrifuged at 4500 rpm (Heraeus Megafuge 40R centrifuge, rotor 3608, Thermo Scientific, Ottawa, ON) for 15 min to pellet all of the debris. The samples collected in 2011 were allowed to air dry at room temperature in their collection vessels. The samples collected in 2012 were processed in the same manner, except that they were dried in a laboratory oven (37°C) until all moisture was evaporated. The pellets were weighed after collection and pooled with the previous collection. The gross weight of all the dust was measured at the end of the season. Samples were stored at 4°C until the end of the sampling season.

DNA extraction

At the end of the season, 150 mg from the pooled dust samples collected from each BSNE sampler was transferred to a bead mill tube (MoBio Laboratories Inc, Carlsbad, CA) for DNA extraction. Where the amount of dust collected from a sampler was <150 mg, the entire amount of dust collected was processed. Total genomic DNA was extracted from all samples with a Powersoil DNA isolation kit (MoBio Laboratories Inc, Carlsbad, CA) as per the manufacturer’s instructions. The concentration and purity of the DNA extracts were quantified on a Nanodrop 2000c spectrophotometer (Thermo Fisher Scientific Inc., Waltham, MA), stored at 4°C and promptly prepared for PCR analysis. Samples were diluted in purified water immediately prior to each respective analysis to a concentration of 2 ng μL⁻¹ for conventional PCR analysis, or diluted 10-fold for qPCR analysis.

PCR analysis

All dust samples were subjected to conventional PCR and qPCR analysis to detect and quantify P. brassicae DNA, respectively. Conventional PCR was performed as per Cao et al. (2007) with the TC1R and TC1F primer set. Quantitative PCR was performed as described by Rennie et al. (2011) with the DR1R and DR1F primer set. The number of resting spores in the dust samples was estimated through qPCR by comparison with a standard curve generated with DNA samples extracted from known amounts (assessed with a haemocytometer) of P. brassicae resting spores (Rennie et al. 2011).

Resting spore distribution analysis

The distribution of resting spores in the commercial fields (Parkland and Bassano sites) was assessed by quantitative PCR analysis of soil samples collected in each field. Briefly, 100 sampling points per field were selected using a random point generator in ArcGIS (Esri Canada, Toronto, ON), and approximately 100 g of soil was collected from each point using a soil auger. The samples were air-dried and homogenized individually in a mortar with a pestle. The homogenized samples were then subjected to DNA extraction and quantitative PCR analysis as described above. Maps showing the resting spore distributions at the two sites were generated with ArcGIS (Esri Canada) using the georeferenced sampling points (Fig. 1).

Data analysis

Statistical analysis was performed using SAS version 9.3 (SAS Institute Inc., Cary, NC, USA). Spearman correlation coefficients were calculated to determine the relation between sampler height and the weight of dust collected, concentration of resting spores detected in the collected dust, and the total number of resting spores collected in the samplers. Spearman correlation was used due to the non-normal distribution of dust weights and resting spore numbers.

Results

Dust collection

In 2011, 4–3160 mg (mean = 387 mg) of dust was recovered from the samplers over the course of the growing season. In 2012, 1–1781 mg (mean = 251 mg) of dust was recovered from the samplers over the course of the growing season. In 2012, four samplers in Bassano did not collect any dust. These samplers were all mounted on the same post which was located on the south-west extent of the field (Fig. 1d). Rainfall accumulation led to overflow from some of the samplers during some intervals of the growing season. Despite these occasional losses, however, sufficient amounts of dust were collected for DNA extraction purposes from all samplers except the aforementioned ones in Bassano. In 60% of the samples, more than 150 mg of dust was obtained.

In both 2011 and 2012, the greatest amount of dust was collected at the lowest sampler heights (0 m in 2011, 0.1 m in 2012). The amount of dust collected decreased significantly above the lowest height and the differences between the other sampler heights were not as pronounced. The correlation between sampler height and dust collection was weak, but significant in both 2011 (r = −0.35, P = 0.0013) and 2012 (r = −0.47, P < 0.0001).

PCR analysis

Plasmodiophora brassicae DNA was detected by conventional PCR analysis in 14 of the 74 dust collections in 2011 and 34 of the 81 dust collections in 2012. The presence of P. brassicae DNA also was detected and quantified by qPCR in 15 of the 74 dust collections in 2011 and 41 of the 81 dust collections in 2012. The locations that tested positive for P. brassicae DNA in 2011 and 2012, as well as the concentrations of resting spores estimated by qPCR analysis, are given in Tables 1–4. Plasmodiophora brassicae resting spores were detected and quantified at both of the commercial field sites, Bassano (<1.0 × 10³–2.2 × 10⁵ spores g⁻¹) and Parkland (<1.0 × 10³–1.3 × 10⁵ spores g⁻¹), in 2011 and 2012, as well as at 50th Street (<1.0 × 10³–2.7 × 10⁵ spores g⁻¹) and CDCN (6.3 × 10³–1.9 × 10⁵ spores g⁻¹). The highest resting spore concentrations were found at the Bassano site in 2011 (2.2 × 10⁵ spores g⁻¹) and at the 50th ST site in 2012 (2.7 × 10⁵ spores g⁻¹).

Table 1. Recovery of Plasmodiophora brassicae DNA from soil collected in Big Spring Number Eight dust samplers (Custom Products, Big Spring, TX) placed adjacent to an agricultural field (‘50th ST’ site) in a rural part of northeast Edmonton, Alberta.

Year	Sampler post	Field Side	Sampler height (m)	Total soil collected^a (g)	Conventional PCR result^b	Concentration of resting spores^c (g^−1)
2011	E	‒	0.8	2.6 × 10^–1	‒	1.6 × 10^4
2011	C	‒	0.1	2.3	+	1.7 × 10^3
2011	D	‒	0.1	3.2	+	<1000
2011	D	‒	1.05	3.7 × 10^–1	+	8.8 × 10^3
2012	A	East	0.1	5.4 × 10^–1	+	2.7 × 10^5
2012	A	East	0.5	1.7 × 10^–1	‒	7.8 × 10^4
2012	A	East	1	1.6 × 10^–1	+	6.9 × 10^4
2012	A	West	0.1	1.7	+	2.2 × 10^5
2012	B	East	0.1	1.6	+	2.9 × 10^3
2012	B	East	0.5	1.6 × 10^–1	+	1.7 × 10^3
2012	B	East	1	1.4 × 10^–1	+	1.2 × 10^3
2012	B	West	0.5	1.2 × 10^–1	+	<1000
2012	B	West	1	5.2 × 10^–2	‒	<1000
2012	C	East	0.1	1.3	+	2.3 × 10^3
2012	C	East	0.5	1.4 × 10^–1	+	1.2 × 10^3
2012	C	East	1	9.6 × 10^–2	+	<1000
2012	C	West	1	1.8 × 10^–1	+	1.2 × 10^3
2012	D	East	0.1	1.1	‒	<1000
2012	D	East	0.5	9.1 × 10^–2	+	1.3 × 10^3
2012	D	East	1	5.5 × 10^–1	+	1.1 × 10^3
2012	D	West	0.1	7.7 × 10^–1	+	1.3 × 10^3
2012	D	West	0.5	1.8 × 10^–1	‒	2.7 × 10^3
2012	D	West	1	7.1 × 10^–2	‒	3.4 × 10^3

^aSoil (dust) was collected from each sampler every 2 weeks from July to October 2011 and from May to October 2012.

^bConventional PCR analysis was conducted as per the protocol of Cao et al. (2007); + = presence of P. brassicae DNA band on agarose gel; – = absence of P. brassicae DNA band.

^cResting spore concentrations were estimated by quantitative PCR analysis based on the method of Rennie et al. (2011) as summarized in the Materials and Methods.

Table 2. Recovery of Plasmodiophora brassicae DNA from soil collected in Big Spring Number Eight dust samplers (Custom Products, Big Spring, TX) placed adjacent to a field at the Crop Diversification Centre North, Alberta Agriculture and Rural Development (‘CDCN’ site), Edmonton, Alberta.

Year	Sampler post	Field side	Sampler height (m)	Soil collected^a (g)	PCR result^b	Concentration of resting spores^c (spores g^−1)
2011	E	‒	1.05	7.0 × 10^–2	+	6.3 × 10^3
2012	A	East	1	1.2 × 10^–1	+	6.4 × 10^4
2012	A	West	1	5.5 × 10^–2	‒	4.9 × 10^4
2012	B	East	1	1.3 × 10^–1	+	5.3 × 10^4
2012	C	East	1	5.5 × 10^–2	‒	1.3 × 10^5
2012	C	West	0.5	3.8 × 10^–2	‒	8.0 × 10^4
2012	C	West	1	3.9 × 10^–2	+	1.4 × 10^5
2012	D	East	0.1	2.2 × 10^–1	+	1.9 × 10^5
2012	D	East	0.5	6.2 × 10^–2	+	8.8 × 10^4
2012	D	East	1	4.8 × 10^–2	+	1.2 × 10^5
2012	D	West	0.1	4.0 × 10^–1	+	1.9 × 10^4

^aSoil (dust) was collected from each sampler every 2 weeks from July to October 2011 and from May to October 2012.

^bConventional PCR analysis was conducted as per the protocol of Cao et al. (2007); + = presence of P. brassicae DNA band on agarose gel; – = absence of P. brassicae DNA band.

^cResting spore concentrations were estimated by quantitative PCR analysis based on the method of Rennie et al. (2011) as summarized in the Materials and Methods.

Table 3. Recovery of Plasmodiophora brassicae DNA from soil collected in Big Spring Number Eight dust samplers (Custom Products, Big Spring, TX) placed adjacent to a commercial agricultural field at Parkland County, Alberta (‘Parkland site’).

Year	Sampler post	Sampler height (m)	Soil collected^a (g)	PCR result^b	Concentration of resting spores^c (spores g^−1)
2011	C	0.35	2.1 × 10^–1	+	6.6 × 10^3
2011	C	0.6	2.4 × 10^–1	+	3.1 × 10^3
2011	C	0.8	1.8 × 10^–1	+	<1000
2012	A	0.1	2.2 × 10^–1	+	2.0 × 10^4
2012	C	0.5	1.9 × 10^–1	+	1.2 × 10^4
2012	C	1	1.5 × 10^–1	+	1.3 × 10^5
2012	D	0.5	1.0 × 10^–1	‒	3.1 × 10^4
2012	E	1	1.1 × 10^–1	+	1.7 × 10^4

^aSoil (dust) was collected from each sampler every 2 weeks from July to October 2011 and from May to October 2012.

^bConventional PCR analysis was conducted as per the protocol of Cao et al. (2007); + = presence of P. brassicae DNA band on agarose gel; – = absence of P. brassicae DNA band.

^cResting spore concentrations were estimated by quantitative PCR analysis based on the method of Rennie et al. (2011) as summarized in the Materials and Methods.

Table 4. Recovery of Plasmodiophora brassicae DNA from soil collected in Big Spring Number Eight dust samplers (Custom Products, Big Spring, TX) adjacent to an agricultural field near Bassano, Alberta (‘Bassano site’).

Year	Sampler post	Sampler number	Sampler height (m)	Soil collected^a(g)	PCR result^b	Concentration of resting spores^c (spores g^−1)
2011	A	2	0	4.0 × 10^–1	+	2.3 × 10^4
2011	A	4	0.25	2.1 × 10^–1	+	2.2 × 10^5
2011	A	1	0.5	2.8 × 10^–1	+	1.7 × 10^4
2011	A	3	0.75	2.5 × 10^–1	+	2.1 × 10^4
2011	A	6	0.5	2.5 × 10^–1	+	8.8 × 10^3
2011	A	8	0.75	1.6 × 10^–1	+	1.0 × 10^4
2011	D	4	0	2.7 × 10^–1	+	1.8 × 10^4
2012	A	‒	0.25	9.0 × 10^–3	+	undetermined
2012	B	‒	0	2.1 × 10^–1	+	<1000
2012	B	‒	0.25	2.5 × 10^–1	+	5.1 × 10^4
2012	B	‒	0.5	1.2 × 10^–1	+	1.3 × 10^5
2012	B	‒	0.75	5.0 × 10^–2	+	5.1 × 10^4
2012	C	‒	0	1.0 × 10^–1	+	6.3 × 10^3
2012	C	‒	0.25	5.8 × 10^–2	+	1.2 × 10^4
2012	C	‒	0.5	1.0 × 10^–1	+	3.7 × 10^3
2012	C	‒	0.75	1.0 × 10^–4	+	undetermined

^aSoil (dust) was collected twice from each sampler from July to October 2011 and twice from May to October 2012.

^bConventional PCR analysis was conducted as per the protocol of Cao et al. (2007); + = presence of P. brassicae DNA band on agarose gel; – = absence of P. brassicae DNA band.

^cResting spore concentrations were estimated by quantitative PCR analysis based on the method of Rennie et al. (2011) as summarized in the Materials and Methods.

In both 2011 and 2012, no correlation was found between sampler height and resting spore concentration (2011: r = −0.12 P = 0.66; 2012: r = 0.17, P = 0.17). There also was no correlation between sampler height and the total number of resting spores detected in the samplers (2011: r = −0.35 P = 0.20; 2012: r = −0.03, P = 0.80). In 2012, the highest total number of resting spores occurred at the lowest sampler height, where the dust collections were frequently larger. The management of weeds surrounding the samplers was improved in 2012 by increasing the frequency of mechanical trimming.

Resting spore distribution analysis

The distribution of P. brassicae resting spores, as assessed by PCR analysis of soil samples collected in each field, was determined for the Parkland and Bassano sites (Fig. 1). At the Parkland site, eight of 100 samples tested positive for the presence of P. brassicae DNA. These samples corresponded to the eastern side of the field, near the entrance (Fig. 1). At the Bassano site, P. brassicae DNA was detected in 58 of the 100 soil samples tested.

Air movement

Air movement across the Parkland and 50th ST sites was punctuated with wind gusts throughout 2012 as high as 51 and 20 km h⁻¹, respectively. While there were portions of exposed soil between the research plots located at the 50th ST site, the Parkland site was sown to a commercial crop and there was little exposed soil following canopy closure. Storm events were observed in 50th ST throughout the summer of 2012, while Parkland had a reduced impact of storm events in mid July, ostensibly the point where the crop canopy fully developed. Due to limitations on available weather monitoring stations, only the 50th and Parkland weather data were recorded.

Discussion

This study was conducted to evaluate whether or not P. brassicae can be carried with windborne dust collected at the edges of clubroot-infested fields. The identification of P. brassicae DNA by conventional PCR and qPCR analysis in dust samples collected from all four field sites, examined in both years of the study, indicates that the movement of pathogen resting spores in windborne dust is possible and perhaps common. There has been little work examining the potential spread of clubroot by wind since this possibility was originally suggested in the late 19th century (Carruthers 1893; cited in Chupp 1917). Nonetheless, the detection of P. brassicae resting spores in eroded soil is not surprising. In Chinese cabbage (Brassica rapa L. ssp. pekinensis) crops, much of the pathogen population has been found to occur at the soil surface and immediate sub-surface (Kim et al. 2000). It is likely that a similar situation would occur in canola crops. With so many resting spores close to the surface, any process that moves soil or dust also would be expected to transport the spores as well. These findings, therefore, confirm the earlier suggestions and provide the first experimental proof of the movement of spores in wind-borne dust.

Ten more positive samples were detected by qPCR than by conventional PCR. This may be due in part to the increased sensitivity of the qPCR assay. Additionally, the two PCR methods were performed on different scales of DNA dilution (PCR: 2 ng μL⁻¹, qPCR: 1/10).

Plasmodiophora brassicae DNA was not identified in all of the individual samplers. The clubroot pathogen, like other soil-borne microorganisms, often has a patchy distribution within fields (Wallenhammar et al. 2012). Hence, differences in the concentration of resting spores in soil collected at the various sampling points likely reflect this uneven distribution in the fields. In the Parkland field, for example, inoculum was found to be localized near the field entrance. Soil collected from the samplers closest to this location accounted for the majority of positive detections of P. brassicae DNA in both 2011 and 2012. Similarly, in Bassano, almost all positive detections were in the samplers that were closest to the clubroot-infested north-east corner of the field. Unfortunately, soil sampling was not performed across the entire Bassano field as it was in the Parkland field.

There was no strong correlation between the height of the sampler and the concentration or total number of resting spores recovered. This suggests that there is no clear association between resting spore numbers and soil particle sizes, as larger particles are expected to accumulate in the lower samplers. No attempt was made, however, to determine the size of the particles recovered in this experiment. There was a large increase in the number of positive samples recovered in 2012 compared with 2011, particularly at the CDCN and 50th ST sites. Given the fact that these were research sites, greater field activity associated with field plot experiments is expected. Agricultural machinery and vehicles were regularly observed driving around and near the samplers at both these sites; increases in this activity could contribute to increased soil erosion. Moreover, there was some exposed soil between the research plots located at the 50th ST and CDCN sites, whilst little exposed soil was observed at the commercial crop sites following canopy closure.

Although this study demonstrates the presence of P. brassicae in airborne dust, additional research is required to gain a full understanding of erosion as a mechanism of pathogen movement. The detection of resting spores by conventional PCR or qPCR analysis does not conclusively indicate that the spores are viable (Vincelli & Tisserat 2008), and wind movement of soil may desiccate any microorganisms within (Dowding 1969). Nonetheless, resting spores have been shown to be quite capable of surviving dry soil conditions (MacFarlane 1952) and harsh environments, including the digestive tracks of animals and composting systems (Karling 1968; Ylimaki et al. 1983).

The potential distance that wind could transport P. brassicae also was not addressed by this study. It is unlikely the amounts of windborne resting spores moved over long distances are sufficient to initiate new infestations in most situations. The heterogeneous distribution of resting spores in farm fields is evidence that locally limited mechanisms are of prime importance in the dispersal of P. brassicae. The unlikeliness of long-distance movement of clubroot is also evidenced by the relatively slow movement of clubroot over wider regions of the prairies, and the development of genetically distinct pathogen populations between Brassica production regions (Strehlow et al. 2013). Nevertheless, the possible significance of long-range wind dispersal of the pathogen should not be dismissed. Even if infrequent, successful long-distance spore movement could allow the pathogen to be introduced into previously disease-free areas. The detection of quantifiable levels of P. brassicae DNA in several isolated fields in the provinces of Saskatchewan and Manitoba (Strelkov & Hwang 2014), combined with the identification of mild clubroot symptoms in a few canola crops in those provinces (Dokken-Bouchard et al. 2012; Kubinec et al. 2014), suggest the possible long-range introduction of pathogen inoculum via wind and pathogen establishment. In the current study, positive PCR and qPCR results were obtained from soil particles recovered from BSNE samplers set at heights of approximately 0.5–1 m. At these heights, significant numbers of soil particles in the 0.05–0.1 mm range would be expected to occur (Chepil 1957). Soil particles in the range of 0.05–0.1 mm could become airborne (Hagen 1999) and would typically be deposited within 1–2 km of the source site. Should these soil particles be contaminated with P. brassicae resting spores, dispersal of the clubroot pathogen via wind-mediated soil erosion could be expected to occur over a distance of at least 1–2 km. Fine soil particles, which may be contaminated with resting spores, have the potential to be dispersed over considerable distances of perhaps 10–100 km or more (Zingg & Chepil 1950; Fryrear et al. 1991). Given the small size of P. brassicae resting spores, i.e. 3–4 μm, individual resting spores that become suspended as a result of wind-mediated soil erosion also could be expected travel considerable distances.

Short-distance wind dispersion could potentially expand resting spore infestations within an infested field or immediately adjacent fields. The localized saltation of soil could spread P. brassicae over short distances within and between fields, further reducing crop yields in already infested fields. This has been found to occur frequently with nematode species (White 1953; Brown 1984). Concerted, short-distance wind dispersion could contribute to the expansion of clubroot infestation into adjacent areas. The focus of clubroot containment in Alberta has been on the prevention of soil movement through equipment sanitization. Sanitation efforts could be undermined by local and regional movement of infested dust. The amount of airborne dust in western Canada has decreased significantly in the past 20 years, and this is believed to be attributable to the widespread adoption of soil conservation techniques such as reduced tillage and direct seeding (Fox et al. 2012). Farming practices that minimize soil erosion may have additional benefits beyond soil conservation, helping to mitigate local or even regional spread of P. brassicae and other soil-borne pathogens.

As the first report on wind dispersal of P. brassicae in Canada, this research provides results that could be the basis of future research or progress in the management of clubroot. The transport of P. brassicae by wind-blown dust particles could enhance the long-term spread and reduce containment of clubroot of canola in Canada.

Acknowledgements

We are grateful to the farmer cooperators who owned the commercial fields in which some of the experiments were conducted.

Additional information

Funding

We thank Agriculture and Agri-Food Canada and the Canola Council of Canada for financial support of this research through the Clubroot Risk Mitigation Initiative under the Growing Forward Program.