Blackleg disease of canola mitigated by resistant cultivars and four-year crop rotations in western Canada

Abstract

In western Canada, canola is traditionally grown in rotation once every 4 years to restrict losses due to pests. Recently, growers have begun to produce canola more intensively due to market opportunities and cultivar improvements. This study was initiated to investigate the consequences of more intensive production of canola in rotations, integrated with currently available disease management practices, including blackleg resistant cultivars and fungicides. A 4 replicate split-plot experiment was established at Scott and Melfort, Saskatchewan with canola rotation treatments ranging from yearly to every second, third or fourth season in combination with pea, wheat and flax using a blackleg resistant and a blackleg susceptible cultivar. Sub-plots were fungicide treatments for blackleg and sclerotinia stem rot diseases. Blackleg incidence and severity was increased in rotations comprising more than one canola crop every four years, regardless of cultivar, although the blackleg resistant cultivar was much less affected than the blackleg susceptible cultivar. The yield of the blackleg resistant canola cultivar was similar among rotations that included canola every 2, 3 and 4 years. Fungicide application was of limited value to maintain canola yield. The results suggest that canola cultivars with strong blackleg resistance can be grown more intensively than once every 4 years with limited yield reduction. However, the increased severity of infection and amount of infested residue produced as canola rotations are intensified, which occurs even with resistant cultivars, increases the risk of inoculum carry-over, resistance breakdown and yield loss. Therefore, it would be prudent for western Canadian canola growers to adhere to less intensive inclusion of canola in rotations, such as one canola crop in 4 years, as an effective blackleg management strategy.

Résumé

Dans l'Ouest canadien, le canola est traditionnellement cultivé en rotation, tous les quatre ans, afin de réduire les pertes causées par les maladies et les insectes. Récemment, les agriculteurs ont commencé à produire de plus en plus de canola en raison des débouchés et des cultivars améliorés. Cette étude a été lancée pour examiner les conséquences d'une production de canola en rotations plus intensives, intégrée aux pratiques courantes de gestion des maladies, y compris l'utilisation de cultivars résistants à la jambe noire et de fongicides. Une expérience a été conduite avec un dispositif split-plot à 4 répétitions à Scott et Melfort, en Saskatchewan, basée sur des rotations effectuées annuellement ainsi que tous les deux, trois et quatre ans avec des pois, du blé et du lin, en utilisant un cultivar de canola résistant et un cultivar de canola sensible à la jambe noire. Des sous-parcelles ont été utilisées pour le traitement de la jambe noire et de la pourriture sclérotique par l'application de fongicides. La fréquence et la gravité de la jambe noire ont augmenté lorsque les rotations étaient de moins de quatre ans, indépendamment du cultivar de canola utilisé, bien que le cultivar résistant fût moins touché par la maladie que le cultivar sensible. Le rendement du cultivar résistant à la jambe noire était semblable dans les rotations bisannuelles, trisannuelles et quadriennales. L'application de fongicides a été peu efficace pour ce qui a été de maintenir le rendement. Les résultats suggèrent que les cultivars qui affichent une forte résistance à la jambe noire peuvent être semés plus d'une fois tous les quatre ans tout en subissant des pertes de rendement limitées. Toutefois, l'accroissement de la gravité de l'infection et du volume de résidus infectés résultant de l'intensification des rotations, ce qui survient également avec les cultivars résistants, augmentent les risques d'arrière-effets de l'inoculum, de perte d'efficacité de la résistance variétale et de pertes de rendement. Par conséquent, les producteurs de l'Ouest canadien seraient avisés de ne pas utiliser le canola dans un programme de rotations aussi intensif : une culture de canola tous les quatre ans semble être efficace comme stratégie de gestion de la jambe noire.

Keywords:

• Brassica napus
• crop rotation
• disease management
• foliar fungicide
• genetic resistance
• Phoma lingam
• phoma stem canker

Mots clés:

• Brassica napus
• fongicide foliaire
• gestion des maladies
• jambe noire
• Phoma lingam
• résistance génétique
• rotation des cultures

Introduction

Crop rotation is defined as the growing of a planned sequence of different crops in recurring succession on the same area of land (Campbell et al., 1990), and was recognized to have benefits to crop production more than 2000 years ago (Lawes & Gilbert, 1894). Diverse crop rotations have been shown to be more sustainable for many reasons, such as improved control of crop pests, management of moisture and fertility, and reduced marketing risk. Crop rotation is a fundamental control strategy for residue-borne plant diseases (Curl, 1963). Its major benefit is the detrimental impact on the growth, survival and reproduction of the causal organism, thereby reducing the pathogen population to a level low enough that significant crop damage does not occur. Historically, productivity and sustainability of cropping systems have been based on complexity or diversity with adherence to crop rotations of 3 years or longer sequences of various crop species (Cook, 2006). The use of less diverse crop rotations generally results in increased pest problems, which include diseases, weeds and sometimes insects, with the result that yields decline (Bullock, 1992; Cook, 2006). Recently, however, the trend in crop production worldwide is to specialize in the production of a limited number of crops to maximize net returns despite evidence that crop monocultures are detrimental in the long term due primarily to the negative impact on crop health, which typically leads to reduced yields, even with the greater use of pesticides and possibly tillage (Cook, 2006). In western Canada, crop sequence research indicates that more diverse rotations tend to have fewer pest problems and lower production risk than rotations based heavily on either cereal or broadleaf crops (Bailey et al., 2000; Johnston et al., 2005; Kutcher et al., 2011).

The trend over the past 30 years in western Canada has been to grow a greater diversity of crops rather than primarily cereals and fallow, while reducing or eliminating tillage (Gan et al., 2010). Over the last 10–12 years, however, growers have again specialized, with more and more farmers producing canola in rotation more frequently than once every 4 years, which has been the standard recommendation (Rimmer et al., 2003). Instead, 3-year rotations of canola with cereals and pulses have been suggested to be sustainable (Cathcart et al., 2006). Two-year rotations of cereals with canola, and even canola monoculture, are becoming common across western Canada (M. Hartman, 2012; A. Kubinec, Manitoba Agriculture, Food and Rural Initiatives, personal communication; V. Vakulabharanam, Saskatchewan Ministry of Agriculture, personal communication). A 2011 survey of canola growers in Alberta showed that 2% of canola was seeded into fields that had been canola in 2010, 34% seeded into fields that produced canola in 2009, 27% in 2008, and 36% was canola in 2007 or prior to 2007 (Blacksheep Strategy Inc., 2012).

More intensive production of canola was facilitated in the 1990s by the introduction of cultivars with good blackleg [Leptosphaeria maculans (Desmaz.) Ces. & De Not.] resistance, tolerance of the herbicides glyphosate, glufosinate and imidazolinones, and large yield improvements due to the development of hybrid cultivars. Since 1985, both yield and value of canola have increased steadily. By contrast, yield and price of wheat, which is the most commonly grown crop in the region, has remained fairly constant over the same time period. Canola producers have responded to market opportunities and canola industry promotion by producing greater amounts of canola, in large part through more frequent production on a given field. Improved blackleg resistance and weed control, compared with the technology available in the 1970s and 1980s, were important factors that have made this possible. However, the adoption of the 2-year, wheat-canola rotation by canola growers is relatively recent (beginning in the mid- to late 1990s) and the sustainability of this practice is questionable because it is highly dependent on both blackleg resistance and herbicide tolerance. Producers as well as industry need to understand the consequences of intensive canola rotations to prepare for unwanted outcomes such as increased pest problems. In Europe, Christen & Sieling (1995) found that diseases, particularly blackleg, were a cause of yield reduction of intensive oilseed rape rotations. In western Canada, blackleg continues to be one of the most common diseases of canola (Dokken-Bouchard et al., 2011; Lange et al., 2011; McLaren et al., 2011).

The objective of this study was to determine the implications of intensive production of canola, while considering the cultivar and pesticide improvements that have been made since the 4-year rotation was recommended. Related to this, the study evaluated the impact that the frequency of canola in rotation would have on diseases other than blackleg and weeds, and the estimated yield and quality losses, to provide a basis for developing revised crop rotation recommendations for canola.

Materials and methods

Locations, rotations and cultivars

The study was conducted at two locations in Saskatchewan. At Scott, the experiment was initiated in 1998 on a Dark Brown Chernozem (Typic Boroll) loam soil on the transition between the semi-arid and sub-humid prairies, and continued until 2007. At Melfort, the experiment was initiated in 1999 on a Black Chernozem (Udic Boroll) silty clay soil in the sub-humid prairies, and continued until 2006. Both experimental sites had been continuously cropped prior to establishment of the study, although canola had never been grown before 1998 at the Scott site. Field experiments were designed as 4 replicate split-plots of 7 rotations with all phases of each rotation present every year. The rotations were: continuous canola (C), continuous pea (P), wheat (W)-P, W-C, W-P-C, W-P-W-C, and W-flax (F)-W-C. The continuous P and the W-P rotations are not discussed in this paper. This scheme resulted in the frequency of canola occurring every year or every second, third or fourth year. As an additional factor, two cultivars of canola were included: one a herbicide (glufosinate) tolerant, blackleg resistant, hybrid (HYB) typical of the most advanced cultivars available at the time, and the other a herbicide non-tolerant, blackleg susceptible, open pollinated (OP) cultivar typical of those in use when the 4-year rotation recommendation for canola was developed. Rotation phases were designated as main plots with fungicide treatments as sub-plots. Because the first year of study at each site was an establishment year, data from that year were excluded.

Cultivars of crops used as rotational crops were adapted to the production sites. The HYB canola cultivar was always an Invigor™ brand cultivar (2663 from 1999–2003 at Scott and 2000–2003 at Melfort, 5030 at both locations in 2004, and 5020 thereafter to reflect the latest genetic improvements). The same OP cultivar, Westar, was grown at both sites every year. The field pea, wheat and flax cultivars also changed periodically to reflect the most suitable ones available at the time.

Seeding and fertility details

Crops were always direct-seeded, with a hoe drill with 20 cm row spacing (Scott, plot size 6.4 × 13.7 m) or with an air drill with 23 cm row spacing (Melfort, plot size 7.6 × 15.2 m). At Scott, canola was generally sown first (median date 10 May, range of 5–17 May), followed by pea (median date 13 May, range of 17–21 May), wheat (median date of 13 May, range of 9–20 May) and flax (median date of 15 May, range of 8–27 May). At Melfort, all crops were sown the same day during the first 10 days of May in 2000, 2001 and 2004, between 11–20 May in 2002, 2003, 2005 and 2006 and on 28 May in 1999. In 2004, pea was sown 1 day earlier than other crops at Melfort.

Both cultivars of canola were sown at rates between 6.7 to 8.2 kg ha⁻¹ each year based on 1000 seed weight to achieve a seed density of 150 seeds per m⁻². At Melfort, terbufos insecticide at 865 g ai ha⁻¹ was applied as a seed row granular treatment each year for added control of fleabeetle (Coleoptera: Chrysomelidae). Seeding rates for other crops were as recommended for western Canada (Anonymous, 2007).

Phosphate fertilizer was applied in the seed row at a rate of 34 kg P₂O₅ ha⁻¹ for all crops every year at Scott. At Melfort, P₂O₅ was side-banded at 20 kg ha⁻¹ in 2004 to 2006; 30 kg ha⁻¹ in 2000, 2002 and 2003 and 33 kg ha⁻¹ in 2001. No K or S fertilizer was applied at Scott, but at Melfort, both were side-banded at rates of 10–15 kg ha⁻¹ for all crops. Nitrogen (N) fertilizer was side-banded at rates that varied among crops and years depending on results of annual soil tests. In canola, applied N varied among years, from 12–78 kg ha⁻¹ at Scott and 65–99 kg ha⁻¹ at Melfort. Within a site, the same rate of fertilizer was used for each crop regardless of the rotation. In general, N application rates were higher at Melfort than at Scott, and tended to be highest for canola, lower for wheat and flax, and very low for pea. In the midst of, and shortly after, the very dry conditions of 2001–2003, N application rates were lower for some crops, because N use by preceding crops declined, resulting in higher soil test N.

Herbicides

All plots seeded to wheat or flax at Scott received 2,4-D at 55 g ai ha⁻¹ in autumn prior to the crops being grown in 1999, 2000, 2001, 2004 and 2007. Prior to seeding, all plots were treated with glyphosate at rates of 440–880 g ai ha⁻¹. The higher rate was used when growing conditions were considered unfavourable for glyphosate to provide complete weed control. Within a site year, the same rate was applied to all plots. In 2002 at Melfort, dry conditions meant that no weeds emerged prior to seeding so pre-seeding herbicide was not required.

At Scott, HYB canola was treated with glufosinate at 500 g ai ha⁻¹ every year with clethodim added at 15 g ai ha⁻¹ during 5 of 9 years when additional grassy weed control was deemed necessary. At Melfort, HYB canola received sethoxydim plus ethamethsulfuron-methyl (20 g ai ha⁻¹) in 2000; thereafter glufosinate (703 g ai ha⁻¹) was applied every year, in addition to ethafluralin (1396 g ai ha⁻¹) in 2001 and 2004, and clethodim (15 g ai ha⁻¹) in 2005 and 2006. For OP canola, ethafluralin (1235–1396 g ai ha⁻¹) was applied each year at Scott and during 2000 at Melfort. In addition, sethoxydim (210–490 g ai ha⁻¹) was applied in 14 of 16 site years, ethametsulfuron-methyl (15–22.5 g ai ha⁻¹) at 11 site years, clopyralid (151 g ai ha⁻¹⁾ at two and quialofop-p-ethyl (35 g ai ha⁻¹) at one site year. Weeds in wheat, flax and field pea were controlled with recommended herbicides, but did not include herbicide group 2 products, such as imazethapyr, due to concerns with herbicide residue.

Fungicides

Two fungicides were applied to sub-plots of each crop, the first application was with azoxystrobin (Quadris, Syngenta) at 125 g ai ha⁻¹ at the 2–6 leaf stage to control blackleg for the first years of the study (up to and including 2003). After 2003, fungicide was not applied at this early stage. From 2000–2003, except in 2002, vinclozolin (Ronilan, BASF) at 500 g ai ha⁻¹ was applied at the 20–30% flower bloom stage to control sclerotinia stem rot. Fungicide at the flowering stage was not applied due to the drought in 2002. From 2004 to 2006, boscalid (Lance, BASF) was applied at 246 g. ai ha⁻¹ for sclerotinia stem rot control. All fungicides were applied in 100 L ha⁻¹ of water. Canola seed of both cultivars was treated with thiamethoxam, difenconazole, metalaxyl and fludioxonil (Helix™, Syngenta). At Scott in 2005, an application of deltamethrin (Decis®, Bayer CropScience) at 6.2 g. ai ha⁻¹ was applied to canola during flowering to control diamondback moth larvae.

Blackleg evaluation

Blackleg severity was evaluated by visual assessment of the basal stem of each plant for the circumference of the stem exhibiting blackleg disease symptoms (extent of the lesion) at swathing (30% seed colour change). Two hundred plants per plot were assessed in the years 2000–2003 and 100 in 2004–2006 using the scale of Newman (1984) (0 = no disease symptoms to 5 = plant dead) as well as recording disease incidence (% of plants infected) for both blackleg and sclerotinia stem rot.

Harvest details

Pea, flax and wheat were direct combined when seed had dried sufficiently to thresh completely. Canola was swathed prior to harvest every year except in 2002 when the crop was highly drought-stressed. In that year, there was a high risk that swaths would be moved by wind so the canola plots were direct-harvested. After harvest, seed was air-dried before cleaning and weighed to determine yield, thousand seed weight (TSW), per cent green seed and analysed for per cent oil content.

Statistical analyses

Data were analysed with the PROC MIXED procedure of SAS (Littell et al., 2006). The effect of replicate and site (location by year combination) were considered random, and the effects of applied treatments were considered fixed. Two models were used in the analysis. The first model was for all except the first year of the study and included main effects of rotation sequence and cultivar, and their interaction. The fungicide split (to control blackleg) was not in this model because the treatment was not applied after 2003. The second model examined the main effects of rotation and cultivar plus the split plot effect of fungicide application for blackleg control, and their interactions, from the second year of the study to 2003. This model was also applied to analyse sclerotinia stem rot for Melfort 2000, the only year the disease was present. The random component for the sclerotina stem rot model included only the replicate. Fixed treatment effects were considered significant at P < 0.05.

Results

Climatic conditions

Generally, precipitation amounts at both locations during the course of this study were near normal in 2000 and from 2004–2007, but drier than normal from 2001–2003 (Table 1). At Scott, the latter period received <70% of the long-term normal growing season (May–August) precipitation. Precipitation was very limited during the growing season of 2001, while 2002 started extremely dry, and crops were drought-damaged so badly that they were unable to respond to moisture received in June and thereafter. Moisture was well distributed during 2003, but below normal throughout the year. Precipitation returned to near normal during 2004, 2006 and 2007, and was >140% of the long-term growing season normal in 2005.

Table 1 . Precipitation (mm) from 1 September to 30 April and during the months of May, June, July and August of 1999 to 2007 at Scott and Melfort, Saskatchewan and long-term averages (1971–2000)

Period	1999	2000	2001	2002	2003	2004	2005	2006	2007	Long Term
Scott
Previous Sept–Apr	136	116	106	65	131	120	147	196	202	147
May	66	24	18	4	20	35	41	63	79	36
June	43	41	59	62	35	52	100	46	103	63
July	81	91	37	39	57	69	77	35	14	71
August	48	57	4	46	36	44	88	47	36	43
Growing season total	238	213	118	151	148	200	306	191	232	213
Annual total	374	329	224	216	279	320	519	387	434	359
Melfort
Previous Sept–Apr	n/a	102	102	95	176	147	150	240	n/a	169
May	n/a	15	9	4	45	18	43	63	n/a	48
June	n/a	74	23	63	64	71	177	74	n/a	73
July	n/a	106	46	5	36	56	70	112	n/a	77
August	n/a	47	11	129	24	55	97	46	n/a	58
Growing season total	n/a	242	89	201	169	200	387	295	n/a	256
Annual total	n/a	344	191	296	345	347	537	535	n/a	425

At Melfort, growing season precipitation was <80% of the long term from 2001–2004, and considerably below this in 2001 and 2003 (Table 1). During 2000, precipitation conditions were low early in the season, but improved from June through August. Precipitation was below normal for all intervals during 2001 and 2002 except August 2002. Moisture shortages became progressively worse from June through August during 2003, while 2004 was characterized by moisture shortages early in the growing season. During 2005 and 2006, annual precipitation was >120% of normal and growing season precipitation >115% of normal, with no periods of moisture shortage.

Mean temperatures during the growing season were below normal at both Scott and Melfort in 2000, 2004 and 2005 (Table 2). Growing season temperatures were near normal at Scott in 2002, 2006 and 2007, and at Melfort in 2002. Growing season temperatures were above the long-term average in 2001 and 2003 at both locations, as well as at Melfort in 2006. Frost in August 2004 at Scott had an impact on the experiments, as well as hail damage at Scott in both 2005 and 2006. A hailstorm on 12 July 2005 at Scott caused extensive damage during flowering of most crops. However, with good moisture, crops recovered reasonably well and at harvest it was difficult to determine the extent of yield loss. Hail occurred on 4 August 2006 when most of the crops were nearing maturity. Yield loss was extensive, and yield was estimated to account for the hail damage. For wheat and canola, heads and pods at several locations in the trial were collected by hand to estimate the extent of yield loss per square metre. Results were compared with estimates on similar crops in adjacent commercial fields by hail adjusters to ensure that estimates were reasonable. On that basis, we estimated losses in canola at 80% for 2006 at Scott.

Table 2 . Mean temperatures (°C) during the months of May, June, July and August of 1999 to 2007 at Scott and Melfort, Saskatchewan and long-term averages (1971–2000)

Period	1999	2000	2001	2002	2003	2004	2005	2006	2007	Long term
Scott
May	9.4	9.4	11.0	8.0	10.1	8.0	9.0	10.9	11.0	10.9
June	13.6	13.5	13.9	16.4	14.9	12.7	13.5	15.3	14.0	15.2
July	15.1	17.8	17.7	19.3	17.7	16.7	18.0	18.8	21.0	17.0
August	16.8	15.6	19.0	15.6	19.6	14.0	13.0	16.3	14.0	16.3
Mean	13.7	14.1	15.4	14.8	15.6	12.9	13.4	15.3	15.0	14.9
Melfort
May	n/a	9.1	11.6	6.8	12.5	6.7	8.6	13.3	n/a	10.8
June	n/a	13.0	14.0	17.1	15.9	12.5	13.9	18.2	n/a	15.7
July	n/a	17.6	18.4	19.8	18.0	16.5	16.9	17.8	n/a	17.4
August	n/a	16.6	19.0	16.0	19.9	13.6	14.9	18.3	n/a	16.4
Mean	n/a	14.1	15.8	14.9	16.6	12.3	13.6	16.9	n/a	15.1

Plant population

Rotation treatment affected canola plant populations (P = 0.0238, Table 3). Averaged over both cultivars and all site-years, plant populations were 98 plants m⁻² (data not shown). This exceeded the 40–60 plants m⁻² considered essential (Angadi et al., 2003; Brandt et al., 2007) to support optimum yield for both cultivars. Plant populations varied among rotations, with similar populations under monoculture canola, the 3-year rotation (W-P-C) and in the 4-year rotation (W-F-W-C). The canola monoculture likely had some volunteers from the previous year's crop. Plant populations were lower in the 2-year rotation (W-C) and the 4-year rotation (W-P-W-C) compared with other rotations. However, the magnitude of the differences was small.

Table 3 . Interaction effects and treatment means of rotation (R) sequence (C – canola, W – wheat, P – pea, F – flax) and cultivar (CV) (OP – open pollinated, HYB – hybrid) on plant population and seed quality (TSW – thousand seed weight and green seed) of canola over 16 site-years (Melfort, 2000–2006 and Scott, 1999–2007) Saskatchewan

Treatments	Plant population (plants m^−2)	TSW (g)	Green seed (%)
R	0.0238	<0.0001	0.1559
CV	0.4196	0.5134	<0.0001
R*CV	0.9915	0.0777	0.0696
Cont C	107.8 a	3.22 c	3.0
W-C	93.4 c	3.40 b	3.0
W-P-C	105.0 ab	3.46 ab	2.7
W-P-W-C	96.6 bc	3.48 a	2.6
W-F-W-C	99.1 abc	3.46 a	2.4
SEM^a	13.27	0.135	0.85
OP	99.1	3.40	3.3 a
HYB	101.6	3.41	2.1 b
SEM^a	12.99	0.134	0.83
^aSEM – Standard error of the mean.

Weeds

Herbicides used for weed control were highly effective at controlling unwanted vegetation in canola crops in all rotations with few exceptions, and for that reason, we have not included weed number or weed biomass data. One exception occurred at Scott in 2005, when biomass from volunteer wheat in OP canola in the W-C rotation was more than one-third of that of the canola biomass. Most of this arose from wheat heads that had fallen to the ground the preceding year as a result of sawfly damage. It also reflected the poor competitive ability of OP canola that was heavily infected with blackleg disease. It was likely that the volunteer wheat aggravated yield loss associated with disease in this instance. A second exception occurred with HYB canola in the W-C rotation at Melfort in 2006, when weed biomass was about 20% of crop biomass. Overall, there was no apparent trend for weeds to increase or decrease in severity as canola frequency in the rotation increased with either cultivar.

Diseases

Sclerotinia infections were either not observed, or only observed at trace levels at Scott during the years of this study. At Melfort, sclerotinia stem rot was present at a moderate to high incidence in one year (2000), but recorded at a disease incidence of <3% in all other years. Sclerotinia stem rot incidence of this magnitude (<3%) would have had an insignificant effect on canola yield. Other diseases noted were: alternaria black spot (Alternaria spp.), foot rot (Rhizoctonia solani Kuhn and Fusarium spp.) and aster yellows (phytoplasma), which occurred at most site-years at low or trace levels and were not believed to have an impact on yield or quality.

Blackleg was the major disease of canola observed during the course of this study at both Scott and Melfort. Blackleg incidence was considered to be moderate to high at Scott in all years of the study, ranging from 39–91% of plants infected in the fungicide untreated, continuous OP rotation (data not shown). At Melfort, blackleg incidence during the study ranged from 7–89%. Incidence was slightly lower at Melfort than Scott every year, except 2004. A similar trend was observed for the severity of blackleg on the OP cultivar, with a minimum score of 0.2 (of a possible 5.0), maximum of 3.8 at Melfort and a minimum score of 1.6 and maximum of 3.5 at Scott.

Cultivar and rotation had the greatest impact on blackleg disease symptoms and the interaction was significant for both blackleg incidence and severity. Less frequent canola production in the rotation: from monoculture to 2-year, 3-year and 4-year, resulted in significantly lower blackleg incidence and severity for both cultivars (Fig. 1). Blackleg incidence of the OP canola cultivar was doubled in canola monoculture (62%) compared with the 4-year (W-P-W-C) rotation (31%), and blackleg severity was 2.6 times as great (score of 2.1 on a scale of 5) in canola monoculture compared with the 4-year (W-P-W-C) rotation (0.8). The results were similar for the HYB cultivar, although the magnitude was much reduced: blackleg incidence was 3.3 times greater in the canola monoculture (24%) and >2 times in the 2-year canola (W-C) rotation (15%) compared with the 4-year (W-P-W-C) rotation (7%); severity was greater by a factor of 5 (0.5) in canola monoculture, and by a factor of 3 (0.3) in the 2-year (W-C) rotation, compared with the 4-year (W-P-W-C) rotation (0.1). For both cultivars, blackleg incidence and severity were not statistically different between the 4-year rotation with field pea and the 4-year rotation with flax (Fig. 1).

Fig. 1. a, Blackleg disease incidence (%) and b, blackleg disease severity (0–5 scale) for each rotation treatment and cultivar. Data are the means of 15 site-years (Melfort 2000–2006, except 2005, and Scott 1999–2007). Standard error of the mean for disease incidence = 3.40, severity = 0.10. Bars with the same letter are not significantly different at P ≤ 0.05 according to Fisher's LSD test.

Fungicide applications aimed at blackleg disease control were applied from 2000–2003 at Melfort and 1999–2003 at Scott. Fungicides were effective at reducing blackleg incidence for both cultivars; however, a significant cultivar by fungicide treatment interaction for blackleg severity indicated that the magnitude of the reduction in blackleg (incidence and severity) was greater for the OP than the HYB (Table 4; Fig. 2).

Table 4 . Interaction effects of rotation (R) sequence (C – canola, W – wheat, P – field pea, F – flax), cultivar (CV) (OP – open pollinated, HYB – hybrid) and azoxystrobin (Quadris®) fungicide (F) applications on blackleg incidence and seed yield of canola at Melfort (2000–2003) and Scott (1999–2003) Saskatchewan

Treatments	Blackleg incidence (%)	Blackleg severity (0–5)	Seed yield (kg ha^−1)	Green seed (%)	Oil content (%)
R	<0.0001	<0.0001	<0.0001	0.0940	0.0269
CV	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
R*CV	<0.0001	<0.0001	0.3445	0.7868	<0.0001
F	<0.0001	<0.0001	0.0106	0.0238	0.1035
R*F	0.3825	0.4862	0.8241	0.6230	0.0100
CV*F	<0.0001	<0.0001	0.0292	0.1399	0.1030
R*CV*F	0.0655	0.7446	0.6349	0.3973	0.0181

Fig. 2. a, Blackleg disease incidence (%) and b, blackleg disease severity (0–5 scale) for each cultivar and fungicide treatment. Data are the means of 9 site-years (Melfort 2000–2003 and Scott 1999–2003. Standard error of the mean for disease incidence = 4.5, severity = 0.12. Bars with the same letter are not significantly different at P ≤ 0.05 according to Fisher's LSD test.

The only site-year where sclerotinia stem rot was deemed moderate to high was at Melfort, 2000. This was the second year of the study at that site, so there were no established rotation effects on sclerotinia stem rot, but rather effects possibly resulting from differences attributed to the previous crop stubble: canola, pea and wheat. There was an interaction of stubble type (rotation) and canola cultivar (P = 0.0032), and of canola cultivar and fungicide (P = 0.0464) on sclerotinia stem rot incidence. The incidence of the disease on the OP was 21% for canola-on-canola, 44% for canola-on-pea and 62% for canola-on-wheat, but for the HYB: 35%, 35% and 38%, respectively. The cultivar × fungicide interaction occurred because fungicide application reduced sclerotinia incidence for the HYB (39% for the unsprayed treatment to 35% for the fungicide treatment) but increased it for the OP (47% for the unsprayed treatment to 50% for the fungicide treatment).

Seed yield

Canola yield was lowest in the monoculture canola rotation, regardless of cultivar (Fig. 3) although the rotation by cultivar interaction was significant (P = 0.015). For the blackleg susceptible OP cultivar, yield was increased as frequency of canola in the rotation decreased. Yield was 835 kg ha⁻¹ in monoculture of the OP cultivar and increased to 1063 kg ha⁻¹ (27%) in the 2-year rotation (wheat-canola), 1254 kg ha⁻¹ (50%) in the 3-year rotation (W-P-C) and 1308 kg ha⁻¹ (57%) for the two 4-year rotations. Yield was 23% greater in the two 4-year rotations than the 2-year rotation, but there was no difference among the 3-year and 4-year rotations.

Fig. 3. Yield of each canola cultivar (OP and HYB) for each rotation treatment. Data are the means of 16 site-years (Melfort 2000–2006 and Scott 1999–2007). Standard error of mean = 184. Bars with the same letter are not significantly different at P ≤ 0.05 according to Fisher's LSD test.

The yield of the blackleg resistant HYB canola was always greater than the yield of the blackleg susceptible OP canola in the same rotation treatment (Fig. 3). The yield advantage varied among site-years, but over the years and under the conditions of this study, the HYB (1659 kg ha⁻¹) averaged 60% higher than the OP (1153 kg ha⁻¹). Similar to the OP, yield of HYB canola was lower in monoculture than in rotations of 2 or more years. Averaged over all site-years, the lower yield associated with growing HYB canola continuously (1333 kg ha⁻¹) compared with every second year (1687 kg ha⁻¹) was 21%, or every third year (W-P-C) was 27%. Unlike the OP cultivar, 2- and 4-year rotations of the HYB cultivar were similar in yield. The 3-year rotation, which included pea prior to canola, had greater yield than the 2-year rotation (9%) or the 4-year rotations (5% greater than the 4-year rotation including pea and 8% greater than the 4-year rotation including flax). The impact of blackleg disease severity on yield was estimated using analysis of covariance with site-year as a fixed effect; a one unit increase in disease severity (0–5 scale) reduced yield of the OP by 187 kg ha⁻¹ (P < 0.0001) and yield of the HYB by 194 kg ha⁻¹ (P < 0.0003) (data not shown).

Despite reduced blackleg incidence and severity of both the OP and HYB cultivars when fungicide was applied, there was limited yield benefit for the OP (8%) and no benefit for the HYB (Fig. 4, cultivar by fungicide interaction P = 0.0290).

Fig. 4. Yield of canola cultivars (OP and HYB) with and without fungicide treatment. Data are the means of 9 site-years (Melfort 2000–2003 and Scott 1999–2003. Standard error of mean = 276. Bars with the same letter are not significantly different at P ≤ 0.05 according to Fisher's LSD test.

In the single site-year when sclerotinia was observed (Melfort, 2000), the benefit of fungicide on yield was confounded between the fungicide application for blackleg and that for sclerotinia stem rot. However, averaged over the 16 site-years, there was no yield benefit to applying fungicide to control sclerotinia stem rot to the HYB (Fig. 4); for the OP, the benefit of this application contributed in only a small way (1 site-year of 9) to the 8% yield increase observed.

Seed quality

Rotation had an effect on TSW but cultivar did not (Table 3). Monoculture canola had the lowest TSW, followed by the 2-year rotation. Thousand seed weight was greatest for the 3- and 4-year rotations. The amount of green seed in the harvest sample differed only between cultivars, 3.3% for the OP and 2.1% for the HYB, but there was no difference among rotations (Table 3). Fungicide application targeting blackleg reduced the amount of green seed in the harvest sample (Table 4), from 2.0% in the untreated plots to 1.6% in the treated plots. There was a rotation × cultivar interaction for seed oil content (Fig. 5), and in the years fungicide was applied for blackleg disease control, a rotation × cultivar × fungicide interaction occurred (data not shown). The HYB cultivar had higher oil content than the OP regardless of rotation, and little variation among rotations and fungicide application. For the OP, oil content was lower in monoculture canola, increased as the frequency of canola was reduced, and increased with fungicide application to control blackleg, but only for monoculture canola.

Fig. 5. Oil content (%) of each canola cultivar (OP or HYB) in each rotation. Data are the means of 16 site-years (Melfort 2000–2006 and Scott 1999–2007). Standard error of mean = 1.3. Bars with the same letter are not significantly different at P ≤ 0.05 according to Fisher's LSD test.

Discussion

Previous research in Europe, USA and western Canada has indicated that greater crop diversification results in increased overall productivity (Christen & Sieling, 1995; Krupinsky et al., 2002; Johnston et al., 2005; Kutcher et al., 2011) and our study supports these findings. As the frequency of canola in the rotation increased, the incidence and severity of blackleg disease was observed to increase, regardless of cultivar. Differences in the incidence and severity of the disease were observed between cultivars, indicating the effectiveness of genetic resistance in combating this disease. At all site-years, the HYB cultivar was always much less affected by the disease. Equally important for the control of this pathogen was the frequency of canola in the rotation. The incidence and severity of blackleg was greater in rotations that included canola more frequently than once every 4 years. This is in agreement with previous research in western Canada (Guo et al., 2005; Johnston et al., 2005), Europe (Fargue-Lelièvre et al., 2011) and Australia (Salisbury et al., 1995; West et al., 2001), although in Australia, Marcroft et al. (2004) reported that isolation of canola crops from each other is more important than the frequency of canola in the rotation to mitigate blackleg.

Leptosphaeria maculans is a residue-borne pathogen harboured on infected stems of canola (Rimmer et al., 2003). The benefit of crop rotation to control blackleg, similar to other residue-borne diseases, depends on the length of time the infected residue requires to degrade. Under continuous monoculture or rotations too short to allow infected canola residue to fully decompose, L. maculans can complete and repeat its lifecycle, resulting in its perpetuation and multiplication. In Manitoba, the number of ascospores and pycnidiospores of L. maculans trapped in 4-year rotations was lower than in more intensive canola rotations (Guo et al., 2008). In Saskatchewan, L. maculans-infected canola residue with the ability to harbour spores of the pathogen was detected up to 5 years after the previous canola crop (Petrie, 1995). This indicates that blackleg disease management might be improved further with canola frequency in the rotation reduced to only once in 5 or more years. However, the rate of residue breakdown depends on environmental conditions, i.e. under dry conditions residue breakdown is slower than under wet conditions (Sosnowski et al., 2006). Recommendations for the frequency of canola in rotation to mitigate blackleg differ from once every 2 years in the State of Victoria, Australia, to once every 2–3 years in the State of South Australia, because the cool, wet conditions in Victoria promote faster residue decomposition than the warm, dry conditions in South Australia. In the State of Western Australia, only one canola crop every 4 years is recommended because hot, dry summers preserve residues for longer periods, similar to conditions in western Canada.

Increased amounts of infected canola residue will contribute to more inoculum and a larger pathogen population in future years, and therefore to greater potential infections of future crops, in rotations of less than 4 years. In addition, larger pathogen populations result in increased selection pressure on the pathogen to overcome the genetic resistance in canola cultivars to L. maculans (Rouxel et al., 2003; Li et al., 2005), as well as for selection of pathogen strains insensitive to fungicides, should fungicide application be adopted as a control strategy. Variability for virulence of L. maculans has been documented in western Canada (Kutcher et al., 2007, 2010) and fungicide insensitivity has been reported for other pathogens in western Canada (Gossen & Rimmer, 2001; Gossen & Anderson, 2004). For these reasons, it is of concern that the amount of blackleg-infested residue in this study was noted to increase dramatically as measured by the incidence (proportion of infected plants) and severity (a measure of the amount of infected tissue) as the frequency of canola in rotation increased. Despite cultivation of a blackleg resistant cultivar, the number of infested plants more than doubled in the 2-year canola rotation, and more than tripled in continuous canola compared with the 4-year rotation (Fig. 1a). Similarly, the severity of infection was 5 times greater in the continuous rotation and 3 times greater in the 2-year rotation than in the 4-year rotation (Fig. 1b).

Guo et al. (2005) found that a diverse rotation (canola rotated with wheat or wheat and flax) reduced the incidence and severity of blackleg in canola regardless of tillage system (conventional or zero). In the continuous canola treatment, tillage reduced blackleg, but the differences between tillage systems, in terms of blackleg, were small when a diverse rotation was followed. These results support the conclusions of Johnston et al. (2005), who found that when a crop (barley, wheat, canola, flax and field pea) was seeded on its own stubble, the yield and quality were usually lower than when produced on the stubble of another crop, and that the difference was related to major pathogens affecting productivity. Similar conclusions regarding the reduction of crop diseases with increasing crop diversity in the rotation were reached by Bailey et al. (1992, 2000), Brandt & Zentner (1995), Zentner et al. (2002) and others as reviewed by Krupinsky et al. (2002).

Sclerotinia stem rot had a significant impact on canola production only at 1 site-year. The incidence of sclerotinia stem rot of the OP was higher in canola seeded on pea stubble than canola on canola stubble, possibly due to susceptibility of pea to sclerotinia, which may have occurred in 1999, although this was not detected. Most likely, levels were greater in canola on pea than canola on canola due to more rapid canopy closure and a thicker, denser canopy as a result of improved fertility and soil health. The greatest incidence of sclerotinia stem rot occurred for canola grown on wheat stubble, possibly due to greater soil moisture retention of the wheat stubble than canola or pea stubbles, which may have facilitated earlier apothecia development, and greater numbers of sclerotia that germinated to form apothecia. The benefit of fungicide appeared to be greater for the HYB cultivar than the OP, possibly because the canopy of the HYB was denser than that of the OP (not measured), which may have provided a more favourable canopy environment for the pathogen in the HYB compared with the OP.

Rotation differences on plant emergence were small, and likely had only minimal impact and only at site-years where overall densities were low. Increased numbers of plants following pea may reflect less pea residue as compared with residues of other crops in other rotations. This may have led to higher soil temperatures in treatments that included pea, favouring a higher per cent emergence of canola (Nykiforuk et al. 1994) or reduced interference of pea residues with seed placement. Hwang et al. (2009) examined canola emergence under controlled conditions in soil retrieved from this study. The soilborne canola root disease pathogens Fusarium spp., Rhizoctonia spp. and Pythium spp. were isolated from the soils. In that study, populations of these pathogens decreased and emergence of canola plants generally increased with increasing diversity of rotation. However, Hwang et al. (2009) observed emergence of canola was sometimes greater in soil from the continuous canola rotation than the 4-year rotation, similar to our observations (Table 3). Increased plant emergence in monoculture canola likely reflected emergence of volunteer canola from the preceding crop. Lower plant densities in the W-C rotation suggested that volunteers may only be an important consideration in the crop grown immediately after canola.

Canola seed quality was affected by cultivar and rotation, which may have been related to the impact of blackleg. Thousand seed weight was lower in monoculture canola and in the 2-year rotation (W-C) than the 4-year rotations, but reduction in seed weight is of little importance to canola growers in terms of product quality. However, seed size can affect early seedling growth and vigour of canola grown for seed, and smaller seed size may be an indicator that seed filling has been hampered by disease or other forms of biotic or abiotic stress. The amount of green seed that was present in the seed sample varied from year to year or between Scott and Melfort as a result of climatic conditions, i.e. hail likely delayed seed maturation in 2005 and 2006 at Scott. However, only cultivar had a consistent impact on the amount of green seed in the harvest sample, which was greater for the OP than HYB, possibly due to the effect of blackleg, which may have delayed maturity for less severely infected plants. There was no effect of rotation, cultivar or fungicide on the HYB cultivar, but oil content of the OP was improved by fungicide application in the monoculture rotation treatment. However, regardless of rotation or fungicide treatment, oil content was greater for the HYB than the OP.

Canola yield reflected the impact of blackleg on the crop, i.e. the monoculture canola rotation had the greatest incidence and severity of blackleg and the lowest yield compared with rotations with reduced frequency of canola. The degree of impact was demonstrated statistically by the analysis of covariance: as blackleg severity increased, yield of both canola cultivars was reduced by a similar amount. Lower yield in monoculture canola, particularly for the HYB, may also have been a reflection of less available soil N due to the increased demand from canola compared with other crop species. Brandt et al. (2007) reported that hybrid canola cultivars have higher N use efficiency and can benefit from higher rates of N than OP cultivars, agreeing with the findings of Smith et al. (2010). The fact that the 3-year rotation of HYB canola had greater yield than the 2-year rotation, or the 4-year rotation that included flax, but not pea, reflected the benefit of growing the crop on pea stubble or with pea in rotation, which may also have been due in part to greater soil N availability. This was despite slightly greater blackleg incidence and severity in the 3-year rotation, compared with either 4-year rotation. In the 4-year rotations, canola was grown on wheat stubble and therefore did not benefit from the added N or moisture provided by the previous pea crop. However, for OP canola, blackleg incidence and severity was likely the main factor responsible for the trend to reduced yield in the 3-year rotation compared with the 4-year rotations, despite the benefit of the previous pea crop in the 3-year rotation. Although there was reduced blackleg incidence and severity in HYB canola with increasing length of rotation, there appeared to be little impact on yield. In the analyses of soil residual nutrients in samples obtained from the present study (Malhi et al., 2011 a, 2011 b), soil residual N was lowest in rotations where overall yield was highest either because the higher-yielding HYB canola was grown or because rotation effects resulted in higher yield. The differences in residual soil N were small relative to differences in N removal between rotation treatments, suggesting that soil N was a less important factor explaining canola yield differences than disease.

There were great differences in blackleg incidence and severity between the OP and the HYB cultivars, which reflected genetic improvement against this disease. This was certainly a factor, although not the sole factor, responsible for the yield difference between the cultivars over the various rotations. Many canola cultivars available during the course of this study were assessed as moderately resistant and some were even moderately susceptible to blackleg, and therefore were expected to perform somewhere within the range observed for the HYB and the OP cultivars used in this study. The increase in blackleg incidence and severity as the frequency of canola in the rotation increased, regardless of cultivar, indicated that this practice can be expected to result in an increase in the pathogen population, even on resistant cultivars. Yield data would suggest that progress was made in genetic resistance to blackleg in the HYB cultivar, but this has not overcome the need for at least 3-year, if not 4-year rotations, to achieve optimum canola yield and to reduce the risk of resistance breakdown of current canola cultivars.

Fungicide application to control blackleg was not of benefit on the HYB cultivar, reflecting the limited impact of blackleg on these R (resistant)-rated cultivars. The small yield improvement as a result of fungicide application on the blackleg susceptible OP cultivar suggests that this disease control strategy may be of limited value to western Canadian canola growers, should genetic resistance break down. However, more research is currently being conducted to determine if other disease control products, application timings and multiple applications may improve disease control.

Acknowledgements

The authors are grateful for the excellent technical support provided by Larry Sproule and Don Gerein at Scott and Colleen Kirkham, Dan Cross and Brett Mollison at Melfort. Thank you to SaskCanola, the Manitoba Canola Growers Association (MCGA) and the Alberta Canola Producers Commission (ACPC) for financial support through the Canola Agronomic Research Program of the Canola Council of Canada. Matching funding from the Matching Investment Initiative of Agriculture and Agri-Food Canada is acknowledged.