Quantifying Yield Losses in Canola (Brassica napus) Caused by Verticillium longisporum

Abstract

Verticillium stripe, a soilborne disease of canola (Brassica napus) caused by Verticillium longisporum, was first identified on the Canadian Prairies in 2014. Despite its increasing incidence, the impact of this disease on canola yields has not been quantified. To address this gap, the relationship between Verticillium stripe severity and yield was investigated in two canola hybrids, ‘45H31’ and ‘CS2000’, at two infested field sites near St. Albert, Alberta, in 2020 and 2021. In 2020, a year with above-average rainfall, both hybrids developed moderate levels of the disease, whereas in 2021, a drought year, symptoms and signs of infection were milder. Regression analysis indicated that seed yield per plant declined with increasing Verticillium stripe severity in both years of the study. In both hybrids, the relationship between disease severity and yield was best explained by second-degree quadratic equations. Although single-plant seed yield declined by up to 80% with increasing Verticillium stripe severity, these reductions did not translate into significant yield losses at the plot level, suggesting that losses experienced by individual plants were offset by reduced competition among the surviving plants. These results underscore the complexity of assessing disease impacts solely based on symptom severity.

1. Introduction

Verticillium stripe is a vascular disease caused by the soilborne fungus Verticillium longisporum (C. Stark) Karapapa, Bainbr. & Heale, which mainly affects cruciferous hosts including canola (oilseed rape; Brassica napus L.) [1,2]. In contrast to the wilt symptoms caused by the closely related fungus Verticillium dahliae Kleb [3], infection by V. longisporum does not typically result in wilting in B. napus [4,5]. Instead, dark unilateral striping develops on the main stem around the pod-filling stage, approximately 3–4 weeks prior to harvest; this is followed by the formation of fungal microsclerotia below the epidermis and in the stem pith [4,6]. The infected tissues take on a shredded, whitish-gray appearance, with a peeling back of the stem epidermis at maturity. The microsclerotia can survive on infected residues and in the soil following decomposition of the host tissues and may remain viable for more than 10 years, serving as an inoculum for future infections [1,5].

Serious losses in oilseed rape attributed to V. dahliae were first reported in Sweden in the 1960s [7]. Several other studies found significant variation in yield losses resulting from infection by Verticillium species, ranging from 10% to 50% [8,9,10]. Losses on a single plant were reported to be as high as 80% for V. longisporum-infected winter oilseed rape in Germany, but only when disease incidence and severity were high [9]. In contrast, a more recent study found insignificant or inconsistent yield losses despite increased disease incidence, suggesting the need for further study of the effect of V. longisporum on yields [4].

In Canada, Verticillium stripe is a new disease, having first been identified on canola in the Prairies in 2014 [10]. No fungicides are registered for control of the disease, and little is known regarding the Verticillium stripe resistance of Canadian canola cultivars. While research on the development of disease management strategies has been initiated, there is no information on the relationship between Verticillium stripe severity and yields in hybrid canola. Such information is needed to determine the extent of the threat posed by this disease, particularly given the importance of canola to the Canadian economy. Canola is Canada’s most valuable crop, contributing over $29.9 billion annually to the national economy and supporting 207,000 jobs across the supply chain [11]. The objective of this study was to evaluate the impact of V. longisporum infection on the yield of hybrid canola, including the relationship between pathogen virulence and inoculum concentration on disease severity and yield under field conditions in western Canada.

2. Materials and Methods

2.1. Grain Inoculum Preparation

Wheat grain inoculum was prepared following [12]. In brief, V. longisporum isolate VL43, originally collected from diseased canola plants sampled around Edmonton, Alberta, was grown in 10 cm diameter Petri dishes filled with potato dextrose agar (PDA). Fresh cultures were incubated in darkness at 22 °C for 14 days, at which time the colonies were cut into about 100 small pieces and mixed with water-soaked, sterilized wheat grain (950 mL grain per culture) in an autoclaved Hi-Patch mushroom grow bag (Western Biologicals, Aldergrove, BC, Canada). The inoculated wheat grain was incubated in the dark at 22 °C for 4 weeks and then placed in a drying oven (Fisherbrand™ Isotemp™ General Purpose Heating and Drying Ovens. Thermo Fisher Scientific Inc, Ottawa, Canada) at 30 °C for 2 days. The grain was then ground and separated through a 2 mm mesh sieve using a grain mill.

2.2. Field Experiments

Field experiments were conducted at the St. Albert Research Station (53.6951° N, 113.6327° W), University of Alberta, in 2020 and 2021. Each year, two trials were established at separate locations on the station grounds, designated as Site 1 and Site 2. The soil at all sites was a black chernozemic sandy loam with a pH of 7.5. The two sites were adjacent to one another in both years. The crop rotation history for both field sites was obtained from research station records. In the 10 years prior, rotations included barley, wheat, and field peas. No Brassica crops had been grown during this period, and the fields were assumed to be free of V. longisporum prior to experimental inoculation. The canola hybrids ‘45H31’ and ‘CS2000’ were used for the experiments and evaluated at four different inoculum levels (see below); control treatments did not receive any inoculum. The experiments were arranged in a split-plot design with four replicates per treatment for each canola hybrid. Individual plots were 9 m² (6 m × 1.5 m) in 2020 and 4.5 m² (3 m × 1.5 m) in 2021 and consisted of four rows spaced 0.3 m apart with individual plots spaced at 0.6 m. There was a 2 m tilled buffer between replicate treatments. The various inoculum levels were generated by manually applying grain inoculum at specific rates: 25 mL per 3 m row or 50 mL per 6 m row (low inoculum), 50 mL per 3 m row or 100 mL per 6 m row (medium-low), 75 mL per 3 m row or 150 mL per 6 m row (medium), and 100 mL per 3 m row or 200 mL per 6 m row (high inoculum). Plots were seeded at the time of inoculation using a push seeder at a rate of 0.35 g seeds per 3 m row or 0.7 g per 6 m row. The trials were planted on 17 May 2020 and 26 May 2021. All plots received uniform fertilization before seeding, including 90 kg/ha of N, 40 kg/ha of P₂O₅, and 20 kg/ha of K₂O. Rainfall in 2020 was higher than average from May to August (Figure A1), resulting in flooding in one replicate at Site 2. Consequently, this replicate was excluded from the analysis. Precipitation in 2021 was below average, resulting in drought-like conditions (Figure A1).

2.3. Assessments of Disease Severity and Seed Yield

Ten plants were selected randomly from each plot at maturity, uprooted, and placed in paper bags. They were then transported to the laboratory, where the plants were dried for 5 days at 30 °C in a drying oven. Each plant was rated for Verticillium stripe severity on a 0–4 scale according to [13], where 0 = no symptoms or signs of disease; and 4 = stem entirely necrotic and covered with microsclerotia, shredding of the epidermis with most pods lost. Seed yield per plant was also recorded for each sampled plant. The remaining plants in each plot were harvested using a small plot combine on 13 October 2020 and 30 September 2021 to assess the total seed yield per plot.

2.4. Statistical Analysis

To evaluate the effects of hybrid type and inoculum level on Verticillium stripe severity and single plant seed yield, an analysis of variance (ANOVA) was conducted. The data sets from the two years of the study were examined separately, without considering the site effect. An adjusted R² was utilized to estimate the fit of the regression model. Regression equations evaluated seed yield with increases in disease severity. Differences were considered significant at p ≤ 0.05 unless otherwise noted. The percentage yield loss at each disease severity rating (0–4) was calculated based on yield loss, with no disease = 0% yield loss. The regression equations estimated the percentage of yield loss for each unit increase in disease severity. All analyses were conducted with R v. 4.3.3 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria, 2013).

3. Results

3.1. Field Experiments in 2020

An early symptom of Verticillium stripe, half-sided yellowing of the leaves, was observed by early July in most of the plots treated with V. longisporum at both sites in 2020. Symptoms and signs of infection in the mature canola plants included discoloration and shredding of the stem, necrosis, and the presence of microsclerotia (Figure A2). No symptoms were noted on plants in the non-inoculated plots at any time, resulting in disease severity ratings of zero (

Table 1. Mean Verticillium stripe severity, mean single plant yield and mean plot yield of the canola hybrids ‘45H31’ and ‘CS2000’ in field experiments with different quantities of Verticillium longisporum inoculum in 2020 and 2021.

Hybrid	Treatment	Disease Severity		Mean Single Plant Yield (g)		Mean Plot Yield (kg/ha)
		2020	2021	2020	2021	2020	2021
‘45H31’	Control	0.0 A	0.1 a	4.2 AB	6.0 a	83.5 A	86.0 a
	Low	0.3 AB	0.3 a	2.8 ABC	7.5 a	70.8 A	71.1 a
	Medium-Low	0.7 ABC	0.4 a	2.1 BC	8.9 a	64.0 A	75.3 a
	Medium	1.1 BCD	0.1 a	1.7 C	9.1 ab	66.2 A	90.7 a
	High	1.4 CD	0.1 a	1.8 C	9.3 ab	71.8 A	88.4 a
‘CS2000’	Control	0.0 A	0.0 a	4.7 A	6.2 a	62.7 A	90.2 a
	Low	0.8 ABCD	0.0 a	2.4 ABC	8.0 a	88.5 A	80.0 a
	Medium-Low	1.3 BCD	0.3 a	2.2 BC	8.7 a	86.9 A	100.9 a
	Medium	1.9 D	0.1 a	1.6 C	9.6 ab	73.5 A	78.4 a
	High	1.5 CD	0.1 a	2.3 BC	13.2 b	76.7 A	107.8 a

Note: Field plots were located at two sites in the St. Albert Research Station, University of Alberta. Treatments refer to the relative amount of V. longisporum grain inoculum applied to the plots. Verticillium stripe severity was assessed on a 0–4 scale. Means in a column followed by the same letter are not significantly different at p ≤ 0.05. See Section 2.2 Field Experiments for full details.

). In general, the mean Verticillium stripe severity increased with higher inoculum levels in 2020 (Table 1). On ‘45H31’, the most severe disease was observed in the medium (mean rating = 1.1) and high inoculum (1.4) treatments, which were significantly greater than observed in the non-inoculated control (Table 1). On ‘CS2000’, the severity of Verticillium stripe was significantly greater in the medium-low (mean rating = 1.3), medium (1.9), and high inoculum treatments (1.5) relative to the control (0.0) (Table 1).

A trend of decreasing single-plant yield with increasing inoculum was observed in 2020. The mean single plant yield for ‘45H31’ in the non-inoculated control treatment was 4.2 g, significantly greater than observed in the medium (1.7 g) and high inoculum (1.8 g) treatments (Table 1). Values for the low (2.8 g) and medium-low (2.1 g) inoculum treatments were intermediate within this range. Single plant yield for ‘CS2000’ was 4.7 g in the non-inoculated control, significantly greater than in the medium-low (2.2 g), medium (1.6 g), and high (2.3 g) inoculum treatments (Table 1). No significant differences in total plot yields were detected for any of the treatments in either hybrid, with these values ranging from 576 g to 752 g and from 564 g to 797 g for ‘45H31’ and ‘CS2000’, respectively (Table 1).

3.2. Field Experiments in 2021

No early symptoms of Verticillium stripe were observed in 2021, and disease severity at maturity on both hybrids remained mild. Additionally, weak symptoms of Verticillium stripe (mean rating = 0.1) were noted in the non-inoculated plots of ‘45H31’. No significant differences in disease severity were detected in either hybrid, although the numerically highest mean disease severities were found in the medium-low inoculum treatment for both ‘45H31’ (0.4) and ‘CS2000’ (0.3) (Table 1).

In 2021, no significant differences in mean seed yield per plant were noted among inoculum treatments except for the non-inoculated control (6.2 g) vs. the high inoculum treatment (13.2 g) for ‘CS2000’ (Table 1). The mean seed yield per plant was 6.0 g and 6.2 g for ‘45H31’ and ‘CS2000’, respectively, in the non-inoculated controls. This was not significantly (p ≤ 0.05) different from the seed yields per plant for these hybrids in the low (‘45H31’ = 7.5 g, ‘CS2000’ = 8.0 g), medium-low (8.9 g, 8.7 g), medium (9.1 g, 9.6 g), or high inoculum treatments (9.3 g, 13.2 g) (Table 1). No trends were observed for the total plot yield for either of the hybrids, and any differences between treatments were not significant (Table 1).

3.3. Regression Models

In 2020, regression analysis indicated that the relationships between disease severity and seed yield per plant at the two sites were best described by quadratic equations (Figure 1A,B). In the case of ‘45H31’, the regression model was y = 3.5 − 0.33x − 0.12x², with the expected average seed yield ranging from 0.26 g to 3.5 g per plant at Site 1 (Figure 1A). At Site 2, the expected average seed yield ranged from 0.792 g to 1.8 g per plant with a regression model y = 1.8 + 0.052x − 0.076x² (Figure 1B). For ‘CS2000’, the regression model at Site 1 was y = 4.6 − 1.3x + 0.08x², and the expected average seed yield ranged from 0.68 g to 4.6 g per plant. At Site 2, the regression model for this hybrid was y = 2.7 − 0.023x − 0.14x², with the expected average seed yield ranging from 0.368 g to 2.7 g per plant (Figure 1A,B).

The regression models for percentage yield loss per plant vs. disease severity at Site 1 in 2020 were y = 5.7 + 8.8x + 3.3x² for ‘45H31’ and y = 6.1 + 27x − 1.6x² for ‘CS2000’ (Figure 1C). At Site 2, these models were y = 9.2 − 2.7x + 3.8x² for ‘45H31’ and y = 4.5 + 0.88x + 5.1x² for ‘CS2000’ (Figure 1D). Yield losses exceeding 60% were estimated for both hybrids at a Verticillium stripe severity ≥ 3 at Site 1. Furthermore, in the case of ‘45H31’ at both sites, plants with a disease severity rating of 1 showed a higher percentage yield loss compared with those with a severity rating of 2. However, as the severity increased further from 2 to 4, there was a subsequent decrease in yields.

As was found in 2020, the relationships between disease severity and seed yield per plant in 2021 were also best explained by quadratic equations. For hybrid ‘45H31’ at Site 1, the regression model was y = 10 − 7.1x + 1.3x², with the expected average seed yield ranging from 0.4 g to 10 g per plant. At Site 2, the expected average seed yield ranged from 0.8 g to 7.4 g per plant with a regression model of y = 7.4 − 6.7x + 1.5x² (Figure 2A,B). In the case of ‘CS2000’ at Site 1, the regression model was y = 9.6 − 5.9x + 0.93x², and the expected average seed yield ranged from 0.27 g to 9.6 g per plant. At Site 2 for this hybrid, the regression model was y = 10 − 8.6x + 1.8x², with the expected average seed yield ranging from 0.4 g to 10 g per plant (Figure 2A,B).

In 2021, the regression models for percentage yield loss per plant vs. disease severity were y = 1.2 + 68x − 13x² and y = 2.3 + 89x − 20x² for ‘45H31’ at Sites 1 and 2, respectively (Figure 2C,D). For ‘CS2000’, the models were y = −0.3 + 61x − 9.5x² at Site 1 and y = 1.5 + 85x − 18x² at Site 2 (Figure 2C,D). Both hybrids showed yield losses exceeding 50% at disease severities ≥ 1 at both sites in 2021. At Site 2, plants with a disease severity of 2 showed a greater percentage of seed yield loss than those with a severity rating of 3 for both hybrids (Figure 2D).

4. Discussion

Verticillium stripe is an emerging disease of canola in Canada, leading to concern regarding its potential impact on the production of this crop [9,14]. A systemic understanding of the impact of Verticillium stripe on canola yields is critical for assessing the necessity and efficacy of various disease management approaches. To investigate the relationship between disease severity, inoculum density, and seed yield, field trials were conducted with two canola hybrids, ‘45H31’ and ’CS2000’, over two years. To our knowledge, this is the first report on yield losses in canola caused by Verticillium stripe under Canadian conditions.

Given the absence of Brassica crops in the preceding decade, it is likely that the soils at the experimental field sites were initially free of V. longisporum microsclerotia. This supports the inference that observed infections resulted from the introduced inoculum. The results indicated varying effects of inoculum density on disease severity across the two years of the study. In 2020, significant differences were observed in Verticillium stripe severity between inoculum treatments. Plots treated with medium-low (‘CS2000’) or medium (‘45H31’) to high densities (both hybrids) of V. longisporum inoculum showed higher disease severity compared with non-inoculated controls. These findings align with an earlier greenhouse study, which indicated significant differences in Verticillium stripe severity between high inoculum and control treatments at the adult plant stage [13]. In contrast, in 2021, no significant differences were found between any of the inoculum treatments or controls for either hybrid examined. An absence of a significant effect on disease severity from artificial inoculation of the soil, even at the highest inoculum density, was also reported in a European study with winter oilseed rape [9]. The authors suggested that, in years with unfavorable environmental conditions for disease development, delayed colonization of the host plant helps to mitigate the impact of infection on yield [9].

Disease development caused by V. longisporum is reported to be favored by dry and increased temperature conditions [15,16]. In the current study, however, Verticillium stripe was milder under the warmer and drier conditions experienced in 2021 vs. 2020. This suggests possible variation between environmental conditions and pathogen development in the field and/or the influence of other factors. German studies have found that the relationship between higher temperatures and increased levels of disease in oilseed rape is not consistently observed [17].

In 2020, there was a clear trend of decreasing single plant seed yield with increasing Verticillium stripe severity. However, in 2021, there were generally no significant differences in single-plant seed yields, except for a significant increase detected only in the high inoculum treatment in ‘CS2000’. This unexpected outcome could reflect a decrease in the number of plants surviving to maturity in 2021. This reduction likely led to less competition for resources among the remaining plants, potentially enabling them to compensate for losses associated with the disease. In canola, there is an established inverse relationship between plant density and yield [18]. While stand establishment appeared to be poorer in some of the high inoculum treatments, it was not quantified in this study. It is possible that inoculation with V. longisporum reduced seedling emergence, although there is limited research on the impact of this fungus on this parameter. Nevertheless, other research has demonstrated that soilborne diseases, such as root rot and damping off, caused by Fusarium spp. and Rhizoctonia solani Kühn [12,19], notably reduce seedling emergence. Total plot yields were similar across treatments in both years of this study, further suggesting that losses experienced by individual plants were offset by reduced competition among the surviving plants.

In both years of the experiments, microsclerotia on infected tissues were detected in early September, near the end of the growing season in western Canada. This pattern mirrors the findings from a German study, where late colonization of plant tissue by V. longisporum led to an exponential increase in disease at the pod-filling stage in winter oilseed rape [17]. The delayed onset of symptoms in the field has also been reported in cauliflower (B. oleracea L. var. botrytis) infected by Verticillium species [3]. Infection by V. longisporum involves the germination of microsclerotia, penetration of lateral roots, and eventual spread into the vascular tissue [20]. The timing of these events can significantly affect yields, and environmental conditions may play an important role in determining this timing, as demonstrated for V. dahliae in cotton [21].

In both hybrids and across both years of the study, single plant seed yield showed a negative correlation with Verticillium stripe severity. Plants with severe infection (disease rating > 3) experienced reductions in single plant yields surpassing 60%. Similar observations were reported earlier in winter oilseed rape, where individual plant yields were negatively correlated with disease severity caused by V. longsiporum [9]. The current study suggests that the relationship between Verticillium stripe severity and yield loss is best described by quadratic equations. The expected average reductions in single plant yields ranged from 17.5% to 82.8% and from 61.8% to 92.3% across both hybrids in 2020 and 2021, respectively, surpassing previous European estimates of 10% to 50% in winter oilseed rape [9]. The differences in potential yield impact could arise from variations in the physiological and morphological traits between winter oilseed rape and spring canola, as well as the diverse environmental and soil conditions present in Europe and western Canada. Environmental factors such as soil temperature and humidity are indeed likely contributors to these yield effects [17]. This study provided an initial assessment of the relationship between Verticillium stripe severity and canola yields under conditions in western Canada. Although single-plant seed yield declined by up to 80% with increasing Verticillium stripe severity, these reductions did not translate into significant yield losses at the plot level. Across both years and hybrids, total seed yield per plot remained statistically unaffected by inoculum level, suggesting that inter-plant compensation may mitigate individual plant losses. These results underscore the complexity of assessing disease impacts solely based on symptom severity.

Given the lack of registered fungicides for the control of this fungus, there is a need for enhanced resistance in commercial canola hybrids. Improved resistance, combined with other strategies, may help to ensure the sustainable management of the Verticillium stripe of canola.