Recent changes in fungicide use and the fungicide insensitivity of plant pathogens in Canada

Abstract

This review examines how changing fungicide usage in Canada is increasing the risk of fungicide insensitivity in pathogen populations, using examples from field, vegetable and horticultural crops. The current generation of systemic fungicides is selective, effective and has a reduced impact on the environment compared with older contact fungicides. However, many of these new fungicides also have an increased risk of fungicide insensitivity. On the Canadian prairies, a trend towards specialization of production within regions has resulted in limited cropping rotations, or even no rotation, in many fields, and yield targets are increasing. As a result, the frequency and extent of fungicide usage is increasing rapidly and insensitivity to certain fungicides has been identified in several crop pathogens. Case studies of fungicide insensitivity on Ascochyta rabiei on chickpea, Mycosphaerella pinodes on field pea, Fusarium sambucinum and Phytophthora infestans on potato, Pythium violae, P. sulcatum and other species on carrot, and Botrytis cinerea on grapes and berry crops are examined. Fungicide manufacturers have taken action to protect the efficacy of high-risk fungicides by formulating them with active compounds with a different mode of action. Also, producers have been advised to rotate fungicides with different modes of action to minimize the risk of insensitivity. Monitoring for fungicide insensitivity is central to early warning of an impending issue with insensitivity, for disease risk prediction and for avoiding disease control failures. In addition, information on insensitivity at the population level could be used to develop and implement insensitivity management programmes.

Résumé

À partir d’exemples provenant des grandes cultures, des cultures légumières et des cultures horticoles, ce compte rendu examine à quel point les changements dans l’utilisation des fongicides au Canada engendrent un risque accru d’insensibilité aux fongicides chez les populations d’agents pathogènes. La génération actuelle de fongicides systémiques est sélective ainsi qu’efficace et a réduit les effets néfastes des plus anciens fongicides de contacts sur l’environnement. Toutefois, plusieurs de ces nouveaux fongicides ont également accru le risque d’insensibilité à leur égard. Sur les Prairies canadiennes, une tendance à la spécialisation des productions dans certaines régions a concouru à limiter la rotation des cultures dans de nombreux champs, voire à l’éliminer, et les cibles de rendement augmentent. Par conséquent, la fréquence et l’ampleur de l’utilisation des fongicides s’accroissent rapidement et l’on a déjà signalé des cas d’insensibilité à certains fongicides chez plusieurs agents pathogènes. On examine donc des études de cas d’insensibilité aux fongicides chez Ascochyta rabiei sur le pois chiche, chez Mycosphaerella pinodes sur le petit pois, chez Fusarium sambucinum et Phytophthora infestans sur la pomme de terre, chez Pythium violae, P. sulcatum et d’autres espèces sur la carotte, et chez Botrytis cinerea sur le raisin et une variété de petits fruits. Les fabricants ont pris des mesures pour protéger l’efficacité des fongicides à haut risque en les préparant avec des composés actifs qui possèdent des modes d’action différents. De plus, on a conseillé aux producteurs d’alterner l’utilisation des fongicides possédant différents modes d’action afin de minimiser les risques d’insensibilité. La surveillance de l’insensibilité aux fongicides est essentielle à une alerte précoce relative à un problème imminent d’insensibilité, permettant de prédire un risque de maladies et d’éviter les échecs de lutte contre les maladies. En outre, on pourrait utiliser l’information sur l’insensibilité à l’échelle de la population pour élaborer et mettre en place des programmes de gestion de l’insensibilité.

Keywords:

• contact fungicides
• field crops
• fungicides
• horticultural crops
• single-site mode of action
• systemic fungicides
• vegetable crops

Mots clés:

• cultures horticoles
• cultures légumières
• Fongicides
• fongicides de contact
• fongicides systémiques
• plantes de grande culture
• site d’action précis

Introduction

The purpose of this review is to examine how recent changes in fungicide usage in Canada are affecting the development of fungicide insensitivity in pathogen populations. Changes in usage pattern in the extensive production systems of field crops on the Canadian prairies have been particularly rapid in recent years, but examples are drawn from field, vegetable and horticultural crops.

Crop producers have been managing plant diseases since plants were domesticated. Fungicides were added to the arsenal of management strategies in the 1800s with the introduction of Bordeaux mixture. Insensitivity to fungicides was first detected in the 1970s (benzimidazol). In recent years, an increasing emphasis on food quality has made disease reduction via in-crop application of fungicides an integral component of the production system for many crops. At the same time, expansion of international trade and travel has increased the frequency of introduction of new plant pathogens/strains into Canada and production areas around the world. Climate change may also be contributing to changes in disease distribution and in the frequency of fungicide applications needed to protect a crop in a competitive marketplace. Finally, the importance of plant pathogens and pesticide residues in crops as a barrier to trade is increasing. As a major exporter of crops, changes that can potentially affect exports are of particular concern to the agricultural industry in Canada.

Contact fungicides such as mancozeb and chlorothalonil, which remain on the exterior of the plant, are generally older chemistries that have activity against a broad spectrum of plant pathogens. Systemic fungicides are taken up by the crop and transported within the plant to some degree. Systemic fungicides are, with a few exceptions, newer chemistries that are active against a narrower range of pathogens. Contact fungicides interfere with multiple metabolic pathways in the pathogen, while systemics generally affect a specific site in a specific pathway and so are less toxic to non-target organisms (many are classified as ‘reduced risk’ pesticides). Contact, broad-spectrum fungicides are generally applied at higher rates but are less expensive per application than the newer reduced-risk products. Also, contact fungicides are often less effective because they need to be in place before the pathogen arrives, so any plant tissue that is not thoroughly covered (e.g. a newly emerged leaf) is not protected (Brent & Hollomon 2007; Mueller et al. 2013).

The site-specific mode of action of systemic fungicides has a strong up-side, because their potential for a negative impact on health, safety and the environment is reduced. However, single-site-specific activity can also be an important drawback, because the pathogen may rapidly develop insensitivity to a specific active ingredient, or an entire class of actives. How quickly this occurs depends on the pattern of usage and the ability of the pathogen to develop insensitivity without sacrificing virulence and aggressiveness (FRAC 2011). There is no evidence of insensitivity to contact fungicides from field studies (Gisi & Sierotzki 2008), and so these fungicides have maintained their efficacy over time (Brent & Hollomon 2007).

One example of a class of active compounds with a high risk of pathogen insensitivity is strobilurin fungicides. Strobilurins bind at the Qo site of cytochrome b in the cytochrome bc1 enzyme complex (and so are known as Qo inhibitors, or QoIs) found in the mitochondrial membranes of fungi (Bartlett et al. 2002). The site-specific mode of action of these active compounds increases the potential for insensitivity to develop in pathogen populations (Koller 1991). Many fungal species are reported to have dramatically reduced levels of sensitivity to strobilurin fungicides associated with substitution of a single amino acid in the cytochrome b site (Chin et al. 2001; Pasche et al. 2005; Grasso et al. 2006). Insensitive isolates generally tolerate very high levels of fungicide (Ypema & Gold 1999; Avila-Adame & Koller 2003; Mondal et al. 2005), resulting in complete loss of disease control when these fungicides are used as solo products (Gisi et al. 2000).

The risk of a pathogen population developing insensitivity to a high-risk fungicide is a cumulative function of several key indicators: (i) genetic diversity of the pathogen; (ii) potential for rapid production and dispersal of pathogen propagules; and (iii) need for repeated application of the fungicide to provide disease management each year (Brent & Holloman 2007). Polycyclic pathogens generally produce propagules rapidly, and windborne spores ensure rapid dispersal. Production systems with large fields and low diversity (within crop and over time) generally require more frequent application of fungicide than systems with high diversity. Insensitivity in closely related species is another strong indicator of risk of insensitivity to that class of compounds. Additional information on these relationships is available at the FRAC website (FRAC 2011).

Fungicide efficacy can be substantially reduced by the development of insensitivity. Since the fungicide still effectively controls sensitive isolates, the frequency of insensitive isolates increases under the selection pressure imposed by each additional application, until the fungicide no longer controls the disease (Ma & Michailides 2005). There are two types of insensitivity: quantitative and qualitative. In quantitative insensitivity, the population becomes less sensitive to each subsequent application of the fungicide, but high rates of fungicide are still effective. In qualitative insensitivity, isolates are almost completely insensitive and control is no longer possible at label rates.

There can be a temptation to blame all instances of disease control failure on loss of fungicide efficacy caused by insensitivity. Factors that could contribute to disease control failure include inappropriate application timing or rate, poor crop coverage and wrong or deteriorated product. Misidentification of the pathogen/problem can also result in disease control failures that might resemble pathogen insensitivity. For example, repeated use of strobilurin fungicides to prevent early blight of potato (Solanum tuberosum L.) caused by Alternaria solani Sorauer has contributed to an increase in the prevalence of brown spot disease. Brown spot is caused by A. alternata (Fr.:Fr.) Keissl., which produces similar foliar symptoms to early blight, but is insensitive to strobilurins (Pasche et al. 2003).

There are also instances where disease control failures resulted from increased biological degradation of the active ingredient in soil, rather than a change in pathogen sensitivity. For example, enhanced microbial degradation following application of the carboximide fungicides vinclozolin and iprodione resulted in loss of control of soilborne pathogens such as Sclerotium cepivorum (Walker et al. 1986; Mitchell & Cain 1996). Enhanced degradation of metalaxyl has also been identified and related to failure to control cavity spot of carrot (Farrar et al. 2002; Hiltunen & White 2002).

Loss of sensitivity to a fungicide has often been assessed based on reduction in mycelium growth (Figs. 1 and 2) or spore germination of the target fungus in the presence of a discriminant dose of the fungicide on agar media. However, these assessments are slow and time-consuming, and often represent an obstacle to processing large numbers of samples. Insensitivity to fungicide can also be monitored using conventional or real-time PCR to detect single nucleotide polymorphisms (SNPs) that indicate insertion, replacement or deletion of specific nucleic acids that are associated with fungicide insensitivity (Van Der Heyden et al. 2014). Use of these molecular tools has increased sampling capacity and effectiveness, and allowed research to be conducted on populations rather than isolates. Moreover, molecular assays can be used to identify the molecular basis for insensitivity at the SNP level to obtain genotypic patterns of insensitivity. This facilitates study of sub-population dynamics (Ma & Michailides 2005; Van Der Heyden et al. 2014). Information on the frequency of isolates carrying different SNPs in a fungal population is important because the SNPs may differ their impact on pathogen fitness and level of insensitivity (Veloukas et al. 2011; Billard et al. 2012).

Recent changes in fungicide usage

Research into crop diversification in the 1970s and 1980s identified several crops with potential for western Canada. The enthusiastic uptake of these new crops by growers resulted in canola and pulses (field pea, lentil and chickpea) becoming important components of the production system on the Canadian prairies, which had previously been dominated by cereals, forages and summer fallow. Diverse cropping rotations, supplemented with genetic disease resistance, became the main mechanism of disease management (Krupinsky et al. 2002), supplemented with occasional application of foliar fungicides (Table 1). The longer intervals between a particular crop that are possible in diverse cropping rotations have been shown to reduce the impact of many important field crop diseases in Canada (Bailey et al. 2000, 2001; Holm et al. 2006; Kutcher et al. 2013; Peng et al. 2014). However, a trend towards specialization of production within regions has resulted in short (or even no) rotation in many areas (Kutcher et al. 2013). Also, yield targets have increased following a number of years of high commodity prices, and fungicide usage has increased rapidly in response. From 1996 to 2006, the mean proportion of conventional crop land on the Canadian prairies that received at least one application of foliar fungicide increased from 7% to 11%. However, the proportion of the area treated with a fungicide increased to 23% in 2011, and this proportion is continuing to increase rapidly (T.K. Turkington, personal communication). In contrast, the mean proportion of agricultural land that received an application of foliar fungicide in Ontario increased from 8% in 1996 to 11% in 2006, which reflected a similar pattern to the prairies, but was only 17% in 2011 (Table 1). This indicates that the rapid increase in fungicide usage in Saskatchewan and Manitoba may not be occurring as quickly in Ontario.

Table 1. Changes in the proportion of the area under conventional field crop management¹ that was treated with fungicide each year, based on Canadian Census of Agriculture figures.

Province	Area (M ha)	Area treated with fungicide (%)				Change over time (%)
		1996	2001	2006	2011	1996–2006	2006–2011
Alberta	7.0	7.7%	8.1%	9.4%	14.9%	21%	59%
Saskatchewan	10.3	5.3%	9.2%	9.0%	21.3%	70%	136%
Manitoba	3.6	10.2%	21.0%	22.5%	47.1%	121%	110%
Total (Prairies)	20.8	7.0%	10.9%	11.3%	23.4%	61%	107%
Ontario	2.3	7.6%	8.7%	11.1%	16.8%	47%	51%

¹Estimate based on the area treated with inorganic fertilizer, which excludes organic production and many forage stands (Statistics Canada 2011).

In vegetable crops, disease pressure has steadily increased as intensive crop production has expanded. Increasing emphasis on quality has made disease management through application of fungicides critical in sustaining growth and delivering a competitive product. As a result, fungicides are applied frequently. Contact fungicides are often preferred in vegetable crops because they are (usually) less expensive and have a broad spectrum of activity. For some vegetable crops, contact fungicides are the only materials that were registered for use. However, many vegetable crops have a dense crop canopy or other growth habit that reduces access by the contact fungicides, so use of a systemic fungicide can provide more effective disease reduction.

In grape and berry crops, flowers are important sites for infection by Botrytis cinerea Pers.:Fr. (Jarvis 1977). After infection, the pathogen can remain quiescent for several weeks. Near harvest, when the sugar content of berries increases, B. cinerea rapidly invades mature fruits, causing complete loss of infected berries. As a result, fungicide applications are required during the flowering period and, to a lesser extent, near harvest. An application of fungicide is generally recommended for grapes at the beginning of flowering, 80% cap fall, and at bunch closure. Application for berry crops is recommended at the beginning of the flowering period (10% bloom), at full bloom (first green berries) and at the end the flowering period. During warm wet seasons, a fourth fungicide application could be required a few days before harvest. Growers are moving from reliance on multi-site fungicides to single-site fungicides. However, this increases the risk that fungicide insensitivity will result in failure of disease control.

Case Study 1. Fungicide insensitivity in pathogens of pulse crops

Ascochyta blight on chickpea

Chickpea (Cicer arietinum L.) is a high-value, drought-tolerant, nitrogen-fixing legume crop adapted to the driest region of the Canadian prairies (Chongo & Gossen 2001). Ascochyta blight, caused by Ascochyta rabiei (Pass.) Labrousse [teleomorph Didymella rabiei (Kovachevski) Arx], is the most important constraint to chickpea production in the region (Gan & McDonald 2002) and is important around the world (Nene & Reddy 1987). Management of ascochyta blight relies on partial crop resistance (Ahmed et al. 2006), crop rotation and tillage (Gossen & Miller 2004) and application of fungicides (Reddy & Singh 1990; Gan et al. 2006).

Strobilurin fungicides are highly effective against ascochyta blight and so were widely and repeatedly used when they first became available for use on chickpea in Canada in 2003. However, the pathogen was at high risk for developing insensitivity to strobilurins because it reproduces sexually by means of wind-borne ascospores (Armstrong et al. 2001), is genetically diverse (Chongo et al. 2004) and repeated applications of foliar fungicides are required to provide acceptable disease management each year (Reddy & Singh 1990; Gan et al. 2006). Also, insensitivity to a strobilurin (QoI) fungicide had already been reported in a related pathogen, Didymella bryoniae (Fuckel) Rehm (Olaya & Holm 2001; Stevenson et al. 2002). Fungicide insensitivity can be dispersed rapidly across large areas via airborne ascospores (Torianni et al. 2008), so the presence of airborne spores made the risk of rapid breakdown of disease management a regional issue rather than a local issue.

Insensitivity to QoI fungicides in A. rabiei was first reported from baseline studies (isolates that were never exposed to QoI fungicides) in western Canada in 2004 (Gossen & Anderson 2004) and a rapid shift to insensitivity to QoIs occurred across large areas of the Northern Great Plains in 2007 (Chang et al. 2007; Gossen, Chang et al. 2008; Wise et al. 2008, 2009; Thaher 2011). The life science industry in western Canada was informed immediately that populations of A. rabiei in the region were no longer sensitive to QoIs (Gossen, Chang et al. 2008, 2009). As a result, one company removed this use from their QoI label, and replaced their strobilurin fungicide with a dual-action fungicide for ascochyta blight management across North America. Extension and industry specialists passed the information to retailers and crop advisors in Canada and the northern USA. Most producers across the region changed their fungicide selections in response to this timely warning, and levels of ascochyta blight remained low to moderate in 2007–2012 (Bassendowski & Gossen 2011, 2012).

Mycosphaerella blight on field pea

Mycosphaerella blight, caused by Mycosphaerella pinodes (Berk. & Blox.) Vestergr (anamorph Ascochyta pinodes (Berk. & Blox.) Jones), is an important disease of field pea (Pisum sativum L.) in western Canada and around the world (Davidson & Ramsay 2000; Banniza & Vandenberg 2003; Bretag et al. 2006). Yield losses in Canada can be as high as 50% (Conner et al. 2007). Management of mycosphaerella blight requires repeated fungicide application each year (Beasse et al. 2000; Warkentin et al. 2000). QoI fungicides have excellent efficacy against this pathogen and so are widely used. Mycosphaerella pinodes was identified as being at a moderate to high risk of developing insensitivity to QoI fungicides because it is genetically diverse (Xue et al. 1998), is a polycyclic pathogen with windborne spores and populations of several closely related pathogens are insensitive to QoIs (Grasso et al. 2006; Keinath 2009).

Based on this risk assessment, isolates of M. pinodes were collected from across the Northern Great Plains and compared with a baseline group with no previous contact with QoIs. All of the baseline isolates were sensitive to QoIs. Of the 324 new isolates assessed, all of the isolates collected from the USA were sensitive but 19 isolates from Canada were insensitive. The mean EC₅₀ of the insensitive isolates was 180 mg L⁻¹, corresponding to 65× the label rate (Bowness 2013). Extension specialists and the life science industry in Canada were advised of the potential problem (Bowness et al. 2011, 2012). As a result, recommendations have generally changed from single-action QoI fungicides to dual-action fungicides, which are less prone to loss of sensitivity. Fortunately, QoIs are not used as intensively on field pea as on chickpea, so a rapid shift to insensitivity is not anticipated.

Other pathogens of concern

Almost all of the research on fungicide insensitivity on field crops on the Canadian prairies has focused on pathogens of pulse crops, but there is an old example of fungicide insensitivity on alfalfa (Medicago sativum L.) grown for seed (Gossen et al. 2001). Pulse crops represent a relatively small component of the production area in the region, even though they are grown on millions of hectares each year. Canola (Brassica napus L.) has become the largest crop in the region, surpassing wheat. The level of fungicide insensitivity in important pathogens such as Leptosphaeria maculans (Desmaz.) Ces. & De Not. (blackleg) on canola and Fusarium graminearum Schwabe (teleomorph Gibberella zeae (Schw.) Petch) (fusarium head blight) of wheat is not known. The rapid rate of increase in fungicide usage on field crops in the region indicates that insensitivity may quickly become an issue in these pathogens, if indeed it is not already resulting in losses in disease control.

Case Study 2. Fungicide insensitivity in pathogens of potato

Fusarium dry rot

Fungicide insensitivity is not a new issue in disease management in potato crops. For example, treatment of seed pieces with the benzimidazole fungicide thiophanate-methyl (TPM) and pre-storage applications of thiabendazole (TBZ) fungicide initially provided effective reduction in silver scurf caused by Helminthosporium solani Durieu & Mont. However, these fungicides were quickly rendered ineffective by fungicide insensitivity (Kawchuk et al. 1994). Benzimidazoles bind to the tubulin molecule to inhibit microtubule assembly and cell division. Although TPM does not belong structurally to the benzimidazoles, it acts via the alkyl benzimidazole carbamate, carbendazim. Insensitivity often occurs between actives, such as TBZ and carbendazim, which have related structures and a similar mechanism of action.

Dry rot of potato tubers associated with infection by Fusarium spp. can cause severe losses in tuber quality and yield in Canada. One of the pathogens that causes dry rot, F. sambucinum Fuckel (teleomorph Gibberella pulicaris), also produces trichothecene mycotoxins (Kawchuk et al. 2002). Benzimidazole fungicides were initially used to manage dry rot, with TBZ as a post-harvest treatment and TPM as a seed treatment (Holley & Kawchuk 1996), which is the same pattern of use that led to development of insensitivity to silver scurf.

Pathogen insensitivity to benzimidazoles was initially observed as a result of substantial increases in the incidence of dry rot and the loss of entire potato storages. Insensitivity to TBZ was found in F. sambucinum (Fig. 1), which is the dominant pathogen causing potato dry rot in Canada, but was rare in other Fusarium spp. on potato. The EC₅₀ values for benzimidazole-insensitive isolates of F. sambucinum ranged between 34 and 71 mg L⁻¹ for TBZ and > 2500 mg L⁻¹ for TPM (Kawchuk et al. 1994; Holley & Kawchuk 1996). Production practices were quickly changed to avoid the use of TPM seed treatments. Also, the use of TBZ as a post-harvest treatment was limited to badly bruised tubers that were to be stored for more than several months and post-harvest TBZ was recommended only for farms experiencing dry rot caused by sensitive F. sambucinum or other Fusarium spp.

To characterize the benzimidazole insensitivity in F. sambucinum, homokaryons were derived from single ascospores from crosses of TBZ-insensitive with sensitive isolates. The β-tubulin gene was amplified from these homokaryons using degenerate oligonucleotides synthesized to conserved sequences of the β-tubulin gene derived from related fungi. An open reading frame containing three putative introns and a deduced amino acid sequence of 446 amino acids exhibited a high level of homology to the β-tubulin gene of other fungi. Although nucleotide and amino acid differences were observed in the β-tubulin gene, none could be linked to TBZ insensitivity (Kawchuk et al. 2002). Linkage analysis confirmed that, unlike many other plant pathogens, the isolated β-tubulin gene was single copy and was not linked to TBZ insensitivity in F. sambucinum. The mechanism of benzimidazole insensitivity in F. sambucinum remains unknown.

Insensitivity to TBZ was expected to persist in the F. sambucinum population even after application of TBZ was discontinued, because insensitive isolates have similar growth rates and virulence/aggressiveness as sensitive isolates. Current management recommendations are that fungicides with different modes of action should be applied in succession to minimize exposure to a single active ingredient. Also, these fungicides should be used in combination with a contact fungicide whenever possible. In addition, management of dry rot should include production and storage practices that minimize the incidence of dry rot, e.g. preventing tuber damage, which is required for infection by Fusarium spp. (Lynch et al. 2003).

Late blight

Late blight caused by Phytophthora infestans Mont. (de Bary) is an important disease of potato worldwide, and can also infect tomato and other members of the Solanaceae. Repeated fungicide application over the growing season is required to reduce losses, but worldwide annual losses are still estimated to be in the billions of dollars (Kawchuk et al. 2014).

Genotypes of P. infestans have shown a propensity for developing insensitivity to certain site-specific fungicides (Hu et al. 2012; Kalischuk et al. 2012; Wijekoon et al. 2013). Prior to 1993, the phenylamide fungicide mefenoxam (the R-enantiomer of metalaxyl) was effective against late blight. Mefenoxam inhibits sporulation and mycelium growth inside host tissues by inhibiting RNA polymerase-1, so mutations that change the affinity of this enzyme can result in insensitivity to this fungicide (Davidse et al. 1983). In 1993, the mefenoxam-insensitive genotypes US-7 and US-8 displaced sensitive isolates (Goodwin et al. 1998) and US-8 has continued to predominate in many production areas.

The genotypes US-8 and US-20 are consistently insensitive to mefenoxam, while genotypes US-22, US-23 and US-24 were initially highly sensitive to mefenoxam (Fig. 2). However, recent studies have demonstrated mefenoxam insensitivity among these previously sensitive genotypes (Hu et al. 2012; Kalischuk et al. 2012; Wijekoon et al. 2013). Recent epidemics of late blight in Canada and the USA have occurred in large part as the result of insensitivity to mefenoxam and long-distance movement of P. infestans on infected plant material (Hu et al. 2012; Kalischuk et al. 2012; Fry et al. 2013; Wijekoon et al. 2013).

Many new systemic fungicides with activity against late blight have been identified. Fungicides with different modes of action are being used in rotation with, or co-applied with, contact fungicides such as chlorothalonil or mancozeb. Use of mefonoxam has been reduced, and co-application of mefenoxam with a contact fungicide has become a standard practice. This change in usage pattern has lowered the selection pressure favouring insensitive strains and slowed the development of insensitivity, and has been associated with the recurrence of mefenoxam-sensitive strains (Grünwald & Flier 2005).

The latest generation of fungicides represent promising components in an integrated management strategy for late blight (Lobato et al. 2008). For example, phosphites stimulate plant defence responses, are active against oomycetes in vivo, and have a low potential impact on the environment. These fungicides have important implications for late blight management at a time when fungicide insensitivity in P. infestans is rapidly increasing (Peters et al. 2014).

Case Study 3. Fungicide insensitivity in cavity spot on carrot

Cavity spot is an important root disease of carrot in many parts of the world. The disease is characterized by superficial, elliptical brown lesions on the root surface, which may also be associated with vertical cracks. The presence of more than two or three small lesions makes the carrot crop visually unacceptable for the fresh market. The lesions are also a problem on processing carrots because they are not easily removed in the peeling process.

Cavity spot is a disease complex caused by several species of Pythium. Worldwide, P. violae Chesters & Hickman and P. sulcatum Pratt & Mitchell are the most common species, but P. sylvaticum Campbell & Hendrix and several others are also commonly isolated (Hiltunen & White 2002). Identification of the pathogens associated with cavity spot in a region is important because species differ in host range and also vary in sensitivity to metalaxyl fungicide (Hiltunen & White 2002). For example, P. sulcatum was less susceptible to metalaxyl than P. violae (White 1988). In Ontario, P. sulcatum, P. irregulare, P. violae and P. ultimum are the species most commonly isolated from cavity spot lesions (McDonald 1994; M. Tesfaendrias, personal communication). In Quebec, P. sulcatum and P. sylvaticum are most common, but P. macrosporum Vaartaja & van der Plaats-Niterink and several other species are also present (Allain-Boule et al. 2004).

The recommended cultural practices for management of cavity spot include avoiding highly infested fields and selecting carrot cultivars that are less susceptible to the disease. Most commercial cultivars are susceptible (Lu et al. 2012; Saude et al. 2014) and crop rotation is generally not effective (Hiltunen & White 2002; Lu et al. 2012). Application of the fungicide metalaxyl or its R-enantiomer, mefanoxam, is the standard practice for managing cavity spot. The fungicide is applied at seeding or early in the growing season in Canada, but timing of application varies in different regions of the world (Hiltunen & White 2002).

There have been many reports of partial or complete failure of metalaxyl to control cavity spot. One source of failure in disease management may be that metalaxyl/mefenoxam is not effective against the Pythium spp. that predominates at the site. For example, P. sulcatum is less sensitive to metalaxyl than P. violae in the UK. A pre-plant soil test was developed to identify the Pythium spp. present in a field, so that growers could assess their risk of cavity spot (Hiltunen & White 2002). Also, Pythium isolates from diseased carrots in Michigan and California were evaluated for sensitivity to mefanoxam and two new fungicides (zoxamide and fluopicolide) with activity against Oomycetes. These new fungicides were not registered for use on carrots, so pathogen populations had not been exposed to these actives. Lack of activity against an isolate would therefore be a characteristic of the pathogen species, rather than the result of development of insensitivity from selection pressure. All of the species were sensitive to zoxamide and most were sensitive to mefanoxam. Each isolate of P. ultimum, P. sylvaticum and P. violae was sensitive to fluopicolide, but P. dissocticum, P. irregulare and P. sulcatum were not sensitive (Lu et al. 2012). Similarly, isolates of P. sulcatum, P. sylvaticum and P. macrosporum from Quebec were sensitive to metalaxyl and zoxamide, but the isolates of P. macrosporum showed a wide variation in sensitivity to zoxamide (Martinez et al. 2005). This indicates that a fungicide might be highly effective at one production site, but much less effective at another site, based on differences in the species that are dominant at the sites. Also, species that are not sensitive to a fungicide would be expected to proliferate in a particular field or region with repeated application of that fungicide. It is important to note, however, that the complex interactions that affect resource utilization and microbial ecology in soil may act as a counter-balance to this selection pressure, ensuring that the proportions of pathogenic Pythium spp. at a site change slowly in many situations.

Some failures of metalaxyl/mefanoxam to control cavity spot of carrot result from reduced persistence of the fungicide caused by enhanced microbial degradation of the active ingredient. For example, enhanced degradation of metalaxyl was identified in a soil that had been treated with metalaxyl six times over 14 years (Hiltunen & White 2002). Rapid degradation of mefenoxam was also observed in fields in California that had received repeated applications of the fungicide compared with fields where no fungicide had been applied (Farrar et al. 2002). In both instances, the pathogen population remained sensitive to metalaxyl, so the breakdown in cavity spot management was not related to insensitivity to the fungicide. Instead, the active compound was degraded so rapidly in soil that it was no longer effective at normal rates of application (Fararr et al. 2002; Hiltunen & White 2002). Enhanced degradation may be a major cause of disease control failures for these products.

Finally, fungicide insensitivity may also cause failures in control of cavity spot. In the study of isolates from Michigan and California, a few isolates of P. irregulare were highly insensitive to mefanoxam and some isolates of P. intermedium and P. sylvaticum had intermediate insensitivity. One of 116 isolates of P. violae had intermediate insensitivity to mefanoxam (Lu et al. 2012). However, an extensive study in England found no P. violae isolates that were insensitive to metalaxyl (Hiltunen & White 2002).

Several authors have noted the need for alternatives to metalaxyl to protect carrots from cavity spot (Hiltunen & White 2002; Martinez et al. 2005; Lu et al. 2012). The report that several Pythium spp. isolated from carrots in the USA were insensitive to fluopicolide, which is not registered for this use, represents a concern because it indicates that the efficacy of this active ingredient may not be durable. There is also evidence that repeated exposure to zoxamide (also not registered on carrots) can select for insensitivity in some Pythium spp. When isolates were grown on media amended with zoxamide and transferred 11 times, three isolates of P. sulatum developed some insensitivity and one isolate of P. sylvaticum developed a high level of insensitivity (> 900×) (Martinez et al. 2005). Interestingly, zoxamide effectively reduced cavity spot in an inoculated greenhouse trial with carrots grown in soil-less mix, but metalaxyl had no effect (Martinez et al. 2005). This lack of activity of metalaxyl was not related to insensitivity or enhanced degradation, and raised questions about other factors that affect efficacy, such as the fungicide binding to the organic matter in the peat-based growing medium (Martinez et al. 2005).

Currently, none of the genes associated with fungicide sensitivity in Pythium spp. have been identified and there is no information about how insensitivity affects pathogenic fitness. Thus, cultural methods for management of cavity spot, such as use of resistant cultivars, should be practiced where available. Also, banding of the fungicide over the row, to treat only a portion of the entire field, may reduce the rate of development of both fungicide insensitivity and enhanced degradation. Biological controls or products that enhance host resistance could also play a role in protecting carrots from cavity spot, but there has been no progress in these approaches to date.

Case Study 4. Fungicide insensitivity in Botrytis cinerea on grapes and berry crops

Fungicides that specifically target Botrytis spp. make up 10% of the world market in fungicides, with application to wine and table grapes representing 50% of this market segment (Dean et al. 2012). In Canada, both multi-site inhibitors and single-site fungicides are registered for management of botrytis diseases in stone fruits, grapes and berry crops. Multi-site inhibitors, such as chlorothalonil and dithiocarbamate, generally act by blocking enzymes related to respiration (Leroux et al. 2010). They are only effective if applied prior to infection (Van der Heyden et al. 2012). The single-site fungicides have a broader window of efficacy but have a moderate to high risk of development of insensitivity. Strobilurins and carboxamides affect mitochondrial respiration by inhibiting the activity of mitochondrial complex ІІІ and ІІ, respectively (Leroux et al. 2010). Dicarboximides affect osmoregulation by inducing lipid peroxidation, membrane destruction and over-production of reactive oxygen species. Sterol biosynthesis inhibitors affect C-4 demethylation during ergosterol biosynthesis (Leroux et al. 2010).

For the most part, information about fungicide insensitivity is acquired through surveys conducted at various spatial scales (farm, region, province or country). As with any survey, the precision of fungicide insensitivity estimation increases with increasing number of samples assessed (Van Der Heyden et al. 2014).

Inventory of phenotypic insensitivity

In 2010 and 2011, a total of 230 samples were collected in 18 vineyards in the province of Quebec. Spores from 10 infected berries were collected from each vineyard. Each berry was selected because it carried a single sporulating colony of B. cinerea, and the spores were collected using individual dry BD-BBL culture swabs (Fisher Scientific, Withby, ON). For each fungicide (azoxystrobin, iprodione, pyrimethanil, boscalid, fenhexamid), a discriminant dose was added to the media following Leroux et al. (2010). The assays were conducted on 1.3% agar media, except for boscalid, where the assays were conducted on PS media (agar plus 4 g of sodium succinate, 2 g of K₂HPO₄ and 2 g of KH₂PO₄). For each sample, 0.3 mL of a spore suspension was spread over the surface of the agar medium and incubated at 19 °C in the dark for 24 h. The proportion of germinated spores was calculated based on assessment of 100–200 spores per treatment. Only spores with a germ tube at least twice the average length of non-germinated spores were considered to have germinated.

The incidence of fungicide insensitivity was relatively constant over the 2 years (Fig. 3). When calculated across years, 91% of isolates were insensitive to azoxystrobin, 61% to iprodione, 60% to pyrimethanil, 49% to boscalid and 5% to fenhexamid. The most important result was that the level of fungicide insensitivity in B. cinerea isolates in Québec vineyards was >40% for most of the fungicides tested. Despite the high level of insensitivity identified in these studies, the relationship between insensitivity in isolates and breakdown of disease management in the field is not clear. Research to identify thresholds for disease control failures is needed.

Fig. 1 (Colour online) Isolates of Fusarium sambucinum recovered from diseased potatoes and grown on potato dextrose agar plates supplemented with 10 mg L⁻¹ thiabendazole to assess fungicide sensitivity. Isolates sensitive (A) or insensitive (B) to benzimidazole fungicides were recovered in similar ratios over a 20-year period.

Fig. 2 (Colour online) Isolates of Phytophthora infestans genotype US-23 grown on V8-potato dextrose medium supplemented with 5 mg L⁻¹metalaxyl. Almost all of the Canadian isolates of P. infestans genotype US-23 were sensitive to metalaxyl (right) in 2010, but had been replaced by insensitive isolates (left) by 2013v.

Fig. 3 Proportion of isolates insensitive to selected fungicides in a conidial germination assay of 232 Botrytis cinerea samples collected from vineyards across Quebec in 2010 and 2011. Capped lines indicate standard error.

Inventory of genotypic insensitivity

Isolates of B. cinerea were collected from vineyards and berry fields in Nova Scotia, Québec, Ontario and British Columbia in 2012 as described previously. Each swab was used to inoculate solid media amended with novobiocin (100 μg mL⁻¹) and tetracycline (50 μg mL⁻¹) for purification and conservation. DNA extraction was performed using a commercial isolation kit (MP Biomedicals, Solon, OH). Isolates that were insensitive to the fungicides assessed in the phenotypic screening were identified with RFLP-PCR and PIRA-PCR assays using published primers (Veloukas et al. 2011). The detection of mutation F412wt-I-S-V related to fenhexamid insensitivity was performed using an ASPPAA assay (Billard et al. 2012).

The result of the genotypic survey was similar to that from the phenotypic assessments. The frequency of isolates carrying mutations related to fungicide insensitivity was 82% for insensitivity to azoxystrobin (mutation G143A), 83% to iprodione (I86S), 80% to boscalid (H272R, H272Y, H272L, N230I, P225F), and 3% to fenhexamid (F412I, F412S, F412V) (Fig. 4).

Fig. 4 Proportion of single nucleotide polymorphisms (SNPs) associated with specific fungicide insensitivity in Botrytis cinerea isolates collected from vineyards across Quebec in 2012. Capped lines indicate standard error.

Overall, the proportion of B. cinerea isolates carrying mutations related to fungicide insensitivity was 71% to azoxystrobin, 57% to iprodione, 44% to boscalid mutation H272R, 20% to boscalid H272Y, 1% to boscalid N230I, 3% to boscalid H272L, 0.1% to boscalid P225F and 4% to fenhexamid (Fig. 5). The pattern of insensitivity was generally similar among the provinces, except that the proportion of isolates carrying mutation G143A for azoxystrobin was higher in British Columbia and Ontario (94%) than in Quebec and Nova Scotia (66%) and the proportion of isolates carrying mutation I86S for iprodione was higher in Ontario (80%) than in the other provinces (49–64%). This pattern was similar across the four crops surveyed (cherry, grape, strawberry and raspberry), with a higher proportion of insensitivity in cherry and grapes (Fig. 5).

Fig. 5 Proportion of single nucleotide polymorphisms (SNPs) associated with specific fungicide insensitivity in Botrytis cinerea isolates collected from cherry, grape, strawberry and raspberry fields in Nova Scotia, Quebec, Ontario and British Columbia in 2012. Capped lines indicate standard error.

Recent developments in molecular detection of fungicide insensitivity have enhanced the surveillance and monitoring capacities of researchers and crop advisors. Estimates of insensitivity levels provided by classical phenotypic assays were generally similar to those obtained with PCR-based assays of specific SNPs in B. cinerea. The benefit of using molecular tools to test for fungicide insensitivity is not only to provide insight on the specificity of a polymorphism towards a particular insensitivity phenotype, but also to provide a more precise characterization in terms of fitness and the practical implications of insensitivity. For example, each of the five known SNPs for insensitivity to boscalid were found in each province. Mutation H272R was by far the most frequent (44% of isolates), followed by H272Y (20%). The three other mutations occurred at a low frequency. The value of the insensitivity factor associated with the two dominant SNPs is low to intermediate, while the other SNPs confer a much higher degree of insensitivity (Veloukas et al. 2011; Yin et al. 2011). Field insensitivity is influenced by the proportion of insensitive individuals carrying SNPs associated with a low, intermediate or high insensitivity factor, so the proportion of specific SNPs in the population has a very important practical significance for field insensitivity. Interpreting the results of insensitivity assays based on the frequency of each SNP may lead to a more accurate estimation of the potential impact of this insensitivity on practical disease management.

Summary and conclusions

In this review, insensitivity to fungicides was shown to be an important issue in crop production across Canada. Also, the need for management of insensitivity via rotating or tank mixing products with different modes of action, monitoring pathogen populations for insensitivity, and identifying alternatives to fungicide application, was a consistent factor across crop types and regions.

Case Study 1 demonstrated that fungicide usage has increased rapidly on the Canadian prairies in response to widespread use of short rotations, in combination with high yield targets for field crops. This represents a very substantial change in the production system. Also, many of the newer single-site-of-action fungicides have a moderate to high risk of fungicide insensitivity, unlike the older contact active compounds, which have a very low risk of insensitivity. As a result of increased usage and increased risk, insensitivity to certain fungicides has already been identified in several pathogen populations and other fungicides are at risk. Manufacturers have moved to protect the efficacy of several high-risk fungicides by formulating them together with active compounds that have a different mode of action. Another approach is for producers to rotate fungicides with different modes of action to minimize the risk of development of insensitivity.

Minimizing the need for fungicide application through crop rotation, utilization of genetic resistance and use of agronomic packages that optimize crop growth will also reduce the risk of fungicide insensitivity. However, monitoring for insensitivity is crucial to identifying the onset of insensitivity, its distribution and evolution, and to identify factors that influence insensitivity. As such, monitoring is an essential first step in the design and implementation of management strategies to minimize insensitivity.

Case Study 2 described situations that represent the status quo for fungicide sensitivity of vegetable crops in Canada. Fungicide usage is high in many vegetable crops, where visual quality is a critical component in a competitive market. These are crops where producers are aware that the normal patterns of fungicide usage can result in development of fungicide insensitivity, and new fungicides are routinely being sought to replace classes of fungicides when their efficacy is lost to insensitivity. For most vegetable crops, rotation of fungicides together with crop rotation generally provides effective disease management, without the need for rapid response to changes in pathogen insensitivity within a growing season.

In Case Study 3, many of the reported instances of breakdown in management of cavity spot on carrots are likely caused by incorrect identification of the causal agent and/or enhanced degradation of the fungicide. However, insensitivity also contributed to management failures in some situations, and rotation of fungicides with different modes of action is currently not an option. Cavity spot on carrot appears to be a system where use of molecular tools for rapid identification of the pathogen(s) at a site, combined with assessment of the presence of mutations for insensitivity, could be used to reduce the risk of management failures.

Case Study 4 of fungicide insensitivity in vineyards and berry crops represents a situation where insensitivity has developed against many of the most common fungicides. Molecular assessments are being used to make disease management using fungicides possible. A very high proportion of the production area receives repeated application of fungicide throughout the growing season, and there is no opportunity to use crop rotation to reduce disease pressure on the crop. Information about the status of fungicide insensitivity is used for fungicide selection and to assess the need to implement insensitivity management strategies. Researchers and crop advisors are focusing on techniques to screen for fungicide insensitivity in real time, to provide producers with recommendations for selecting effective fungicides for their situation and on how best to manage insensitivity.

Management of fungicide insensitivity is important for growers and agribusiness because it affects yield, stability of production and economic sustainability. Information on the level of insensitivity is essential for making informed disease-management decisions. Management of fungicide insensitivity is also indirectly important to consumers because it can affect quality and product availability. Fungicide insensitivity has been an important issue for agriculture in Canada for many years, but its importance is increasing as fungicide usage becomes the norm in field crops, as it is already for vegetable and horticultural crops. Monitoring high-risk pathogens of field crops to identify the development of insensitivity is clearly required, before this change in production practice results in large-scale breakdown of disease control.

Acknowledgements

The authors thank the many colleagues who provided isolates and diseased specimens for our studies on fungicide insensitivity. We also thank the Agriculture Development Fund of Saskatchewan, Saskatchewan Pulse Growers, Cavendish Farms, Conagra Limited Lamb-Weston Division, Potato Growers of Alberta, the Conseil pour le développement de l’agriculture du Québec, Canadian Agricultural Adaptation Program and AAFC for financial support of this research. Finally, thanks to the Canadian Phytopathological Society for supporting the symposium that formed the basis of this invited review.