Recurring Challenges from a Necrotrophic Fungal Plant Pathogen: a Case Study with Leptosphaeria maculans (Causal Agent of Blackleg Disease in Brassicas) in Western Australia

Abstract

• Background Blackleg disease of Brassica napus, caused by the necrotrophic fungus Leptosphaeria maculans, causes severe yield losses in Australia, Europe and Canada. In Western Australia, it nearly destroyed the oilseed rape industry in 1972 when host genotypes and conducive environmental conditions favoured severe epidemics. The introduction of cultivars with polygenic resistance and the adoption of sound cultural practices two decades later helped to manage the disease. These were abandoned by many farmers in recent years in favour of the effective but ephemeral resistance conferred by the single dominant gene-based resistance derived from B. rapa ssp. sylvestris. Recently, several cultivars carrying this gene have collapsed widely within a period of 3 years after their commercial release. An environment conducive to the disease and the association of the pathogen with susceptible hosts in Western Australia for over 80 years together have led to the proliferation of L. maculans races, amounting to half of all races delineated to date from Europe, including the United Kingdom, Canada and Australia.

• Scope This review demonstrates the problems that emerge when traditional cultural practices employed, along with cultivars containing polygenic resistance to a serious necrotrophic pathogen, are discarded in preference to the exclusive deployment of effective but ephemeral single dominant gene-based resistance to the disease across Southern Australia.

• Conclusions Single dominant gene-based resistance currently available, on its own, will not confer durable resistance to blackleg disease in oilseed rape. Return to earlier management practices, including reliance upon polygenic resistance and induced resistance, may be the best currently available options to maintain production in regions across Southern Australia predisposed to severe epidemics.

INTRODUCTION

Blackleg [causal agent Leptosphaeria maculans (Desm.) Ces. et de Not.; anamorph Phoma lingam (Tode:Fr.) Desm.] is a serious disease of oilseed rape (Brassica napus L. and Brassica rapa L.) worldwide (West et al., 2001). While this necrotrophic pathogen causes cotyledon and leaf lesions during the vegetative phase of the host plant, it is the crown canker phase in the adult plant that leads to dramatic yield losses of oilseed rape (Barbetti et al., 1975a, b; Bokor et al., 1975). Blackleg causes major economic losses in the main oilseed rape growing regions in Europe (West et al., 2001), Canada (Gugel and Petrie, 1992) and in Australia (Barbetti, 1975a; McGee and Emmett, 1977; Salisbury et al., 1995), despite the fact that each of these countries has different growing seasons, cultivars, agricultural practices and climates (West et al., 2001).

There have been two distinct periods in Western Australia when oilseed rape has faced collapse from blackleg. Firstly, in the early 1970s when there was widespread total collapse of the susceptible Canadian cultivars used at that time and, more recently, in 2003 onwards when the industry was again threatened, this time with collapse of the widely utilized cultivars containing single dominant gene-based resistance derived from B. rapa ssp. sylvestris (hereafter referred to as Brs R, but the gene is as yet unnamed (Li and Cowling, 2003). These recurring and serious challenges from L. maculans in Western Australia over the past 30 or more years make the situation there an ideal case study for better managing this and similar necrotrophic fungal pathogens. This review provides a biological basis and historical account of what has occurred in Western Australia. It outlines the particular characteristics of the pathogen and its environment in Western Australia that make this disease so severe, and highlights the key issues needing to be addressed to better manage this disease in the future.

With such a long period of continuous presence of L. maculans in Western Australia, it is not surprising that this region was the first anywhere to record the existence of L. maculans isolates in a field nursery that were able to cause characteristic disease lesions on cotyledons and crowns of cv. Surpass 400, and overcome the Brs R in this cultivar (Li et al., 2003). This occurred during the second season after commercial release of this cultivar in Australia in 2000 (Easton, 2001), but on a farm where some experimental materials containing the Brs R gene had been grown previously during the 1998 and 1999 seasons (Hua Li et al., 2003). However, the time frame within which resistance is overcome in different regions can vary due to the influence of factors such as the degree of selection pressure and the fitness of the resistance-breaking strains (Pang and Halloran, 1995). These findings clearly support the suggestion by Parlevliet (1993), that narrow sources of resistance in plants are unlikely to be durable in the long term. In many cases, single dominant gene-based resistance similar to that conferred by the Brs R gene, always showing resistance in the form of a hypersensitive reaction, have been considered to be non-durable (Lindhout, 2002; Parlevliet, 2002).

It is not the intention of this review to repeat coverage given in previous reviews relating to various aspects of blackleg disease in Australia, such as those by Salisbury et al. (1995), West et al. (2001) and Howlett (2004). Instead, the outcomes of studies from Western Australia over the past 30 years will be presented together with biological aspects to explain the basis of the recurrence of severe epidemics and the proliferation of unusually large numbers of races to date in this region.

PROLONGED ASSOCIATION OF HOST AND PATHOGEN

Leptosphaeria maculans has been known to be present on brassicas in Western Australia for at least 80 years. It was first reported in Western Australia in 1925 by Carne (1925) on both cabbage [B. oleracea L. var. capitata (L.) Alef.] and cauliflower [B. oleracea L. var. botrytis (L.) Alef.]; and subsequently reported on kale (B. oleracea L. var. acephala Bailey) in 1937 (Anon., 1937) and on swede [B. napus L. var. napobrassica (L.) Reichenb.] in 1959 (Chambers, 1959). The first confirmed isolates on oilseed rape (B. napus) were collected in 1971 (Bokor, 1972; Barbetti, 1975a). In 1973, L. maculans was confirmed to be in the wild cruciferous weed population, first on wild Raphanus raphanistrum L. (Anon., 1973; Barbetti, 1978b) and, by 1974, it was also confirmed to be in wild turnip (B. tournefortii Gouan) (Barbetti, 1978b). Both R. raphanistrum and B. tournefortii are widely distributed throughout Western Australia (Gioia, 2004a, b) and are thought to have been introduced during European colonization, for example, since the middle of the 19th century for R. raphanistrum (Anon., 2004).

Commercial cropping of oilseed rape was originally introduced into Western Australia in the early 1960s (Colton and Potter, 1999) with cultivars predominantly from Europe as well as from Canada (Roy and Reeves, 1975). Following the industry collapse in 1972 due to blackleg, it was clear that if an oilseed rape industry was to survive in Western Australia then better adapted cultivars that were resistant to blackleg were essential. Because of this, there effectively was no significant oilseed rape industry in Western Australia from 1974 until 1991.

The resurgence of the oilseed rape industry in Western Australia was rapid, with the total areas cropped rising from approx. 15 500 ha in 1991 to 420 000 ha sown in 2004. The cultivar Surpass 400, released Australia-wide in 2000 (Easton, 2001; Burton, 2003) as the most resistant cultivar to L. maculans, initially showed excellent resistance to L. maculans in the field. Only hypersensitive resistant reactions, characterized by small dark brown necrotic local lesions without pycnidia, were evident on the cotyledons, leaves or stems. This resistance to L. maculans in cv. Surpass 400 originated from B. rapa ssp. sylvestris (Crouch et al., 1994; Easton, 2001; Buzza and Easton, 2002) and was considered to be controlled by a single dominant gene (Li and Cowling, 2003). This led to a significant change in cultivar usage in Western Australia since 2001, from cultivars containing polygenic resistance towards those cultivars containing the Brs R gene, of which cv. Surpass 501TT in particular but also, to a lesser extent, Surpass 603CL together constituted approximately two-thirds of the total area sown in Western Australia by 2003 (Littlewood, 2003). Most sources of host resistance in Australian spring-type oilseed rape are from winter-type oilseed rape material from France or via Japanese spring-type cultivars such as Norin 20, which is a parent of many Australian cultivars (Salisbury and Wratten, 1999). Like Australia, Canada initially sourced most of its blackleg resistance either from France (e.g. cultivars Jet Neuf and Cresor) or Asia (e.g. cultivars Norin 20 and Mutu), but more recently has utilized spring-type materials from Australia, such as cv. Maluka (Rimmer et al., 1995). It is clear that there are different virulence alleles in the L. maculans populations between France, Australia and Canada. For example, there are fewer races in France and Canada than in Australia (Balesdent et al., 2005). It is intriguing that most virulent L. maculans races to date in Western Australia (Hua Li et al., 2005a; Hua Li, K. Sivasithamparam and M. J. Barbetti, unpubl. res.) were found on an experimental farm at which blackleg disease screening has been undertaken annually for the majority of the past 30 years.

THE INOCULUM OF THE PATHOGEN

The blackleg pathogen population in Australia is almost exclusively the ‘A’ type (West et al., 2001) and hence is truly L. maculans rather than the ‘B’ types, now established as the separate species L. biglobosa by Shoemaker and Brun (2001), which forms a significant proportion of the Leptosphaeria pathogen population on oilseed rape in most other countries. The main source of inoculum is the windborne ascospores (sexual spores) that are produced in pseudothecia on infested oilseed rape stubbles in autumn and winter (Bokor, 1972). The main periods of release of ascospores vary in the different regions, but occur predominantly during late autumn to the end of winter (Salam et al., 2003). The ascospores can travel up to 8 km (Bokor et al., 1975). As a consequence of the Mediterranean environment in the cropping region of Western Australia, particularly in the high rainfall areas, seedling emergence often coincides with the early ascospore showers (Salam et al., 2003), as the timing of both are primarily determined by rainfall (Barbetti and Khangura, 1999). This may contribute to the more frequent occurrence of severe epidemics there than in other countries.

Wherever blackleg disease occurs in the world on oilseed rape, it is the airborne L. maculans ascospores that are considered the main source of inoculum (Bokor et al., 1975; McGee and Emmett, 1977; Gladders and Musa, 1980; Hershman and Perkins, 1995; West et al., 2001). Ascospore discharge patterns vary over time, with both location (Rempel and Hall, 1993; Khangura and Barbetti, 2001) and season (Petrie, 1995; Khangura and Barbetti, 2001) due to variation in the timing of pseudothecial maturity (West et al., 1999). Determining the developmental stages of the pseudothecia prior to ascospore discharge and manipulating these stages opens up possibilities for interfering with the development of the teleomorph of L. maculans on the residues, an issue that will be dealt with in more detail later in this review.

The role of pycnidiospores (asexual spores) is not well defined for the different regions where oilseed rape is grown. While their role in the epidemiology is considered to be minor in Europe (West et al., 2001), in Australia they have been shown to be able to play a significant role in field epidemics (Barbetti, 1976) and the additional impact on the disease epidemic from pycnidiospores may be one reason for such severe disease epidemics in Australia. Studies undertaken by Hua Li et al. (2004b) on the infection processes of L. maculans ascospores and pycnidiospores were the first to report key differences in the infection processes by the two spore types, with ascospore germination, penetration and development of symptoms on cotyledons occurring up to 2 d earlier than that with pycnidiospores. This finding, along with those of Wood and Barbetti (1977a), that ascospore inoculation gave consistently higher numbers of infected plants compared with pycnidiospore inoculations, may explain why ascospore infection is considered to be of overriding importance in generating severe, early seedling disease epidemics (Salisbury et al., 1995; West et al., 2001). This may also explain why epidemics in Australia initiated solely by pycnidiospores develop relatively slowly (Barbetti, 1976). Despite this, it is clear that pycnidiospores can play a very significant role in epidemics in Australia; perhaps more so than other major oilseed rape-growing regions.

Mahuku et al. (1996) showed that co-infection by pycnidiospores of weakly and highly virulent isolates of L. maculans in oilseed rape resulted in a reduction of lesion size on leaves as compared with a highly virulent isolate alone. They also reported that it resulted in an induction of systemic acquired resistance. In a recent controlled environment study, Hua Li et al. (2005b) showed that the overcoming of the Brs R gene could be prevented by co-inoculation of plants with a strain of the pathogen avirulent on the host. This is an exciting discovery that could be utilized in the management of the disease in the field and is discussed further in the section below on ‘Cultural control: novel cross-protection’.

THE DISEASE

In Western Australia, in the early 1970s, some of the most severe blackleg epidemics occurred (Bokor et al., 1975) on the susceptible introduced cultivars grown at that time. For example, in Western Australia in 1972, some 49 200 ha of oilseed rape produced only 7500 tonnes of seed (average yield of 159 kg ha⁻¹) (Barbetti, 1975a). By 1973 the industry had already declined markedly and in that year only 3200 ha were grown (Barbetti, 1975a). Although the development of highly resistant cultivars in recent years has seen the re-establishment of the oilseed rape industry in Western Australia since the early 1990s, significant losses still occurred, such as the A$18·6 and A$49·4 million reported for the 1998 and 1999 crops, respectively (Khangura and Barbetti, 2001).

The continuing relatively poor performance against blackleg disease of cultivars in Western Australia is probably due to increased disease pressure resulting from increased areas of infested residues (Barbetti and Khangura, 1999), to the appearance of more aggressive isolates of L. maculans (Hua Li et al., 2004a, 2005a), and due to the recent overcoming of single dominant gene-based resistance (Li et al., 2003). Commercial Australian B. napus cultivars containing the Brs R gene were grown on increasingly large acreages across Australia in the early 2000s, particularly in Western Australia, where the percentages of the total oilseed rape area sown to cultivars containing the Brs R gene were 0, 8·6, 56·1 and 66·4 % for years 2000, 2001, 2002 and 2003, respectively (Littelwood, 2003).

An important implication from the study of Hua Li et al. (2005a) was that the nature of background polygenic resistance, also termed ‘quantative resistance’ by Lindhout (2002), present in Australian cultivars with single dominant gene-based resistance may be inadequate to prevent eventual collapse from severe canker damage, as they had observed with cv. Hyola 60 and has now been confirmed in commercial plantings in the 2004 cropping season in Western Australia (Hua Li, M. J. Barbetti and K. Sivasithamparam, unpubl. res.). Blackleg disease remains the most severe disease threat to the sustainability of this ‘recovered’ oilseed rape industry in Western Australia. Simple cultural measures that may provide successful management of blackleg disease in eastern Australia, such as maintaining crop isolation from only the residues of the immediate previous year, as recommended by Marcroft et al. (2003, 2004a), are clearly inadequate in Western Australia, where there is strong evidence that residues of 2–3 years of age or more are still capable of causing severe blackleg epidemics in some situations (Barbetti and Khangura, 1999).

Importance of early seedling infection

In Australia, maximum yield loss results from infections that occur early in the seedling stage (Barbetti and Khangura, 1999; Khangura and Barbetti, 2001), when the host is most vulnerable (Bokor et al., 1975; Barbetti and Khangura, 1999). Cotyledon infection in particular is the major source of infection for subsequent canker development (Barbetti, 1975a, 1978a), with the highest correlation between cotyledon infection and subsequent crown canker development resulting from inoculation on very young seedlings (Cargeeg and Thurling, 1979). While later infections are considered to be less significant in terms of yield loss, in some situations significant yield loss can still occur (Barbetti, 1975c). In genotypes containing effective single dominant gene-based resistance, all growth stages normally show high resistance (Crouch et al., 1994) and blackleg inoculum pressure during seedling emergence and establishment is not critical for subsequent yield. For all other cultivars, it is the early seedling infection, and in particular the infection of cotyledons, that is of overriding importance in causing the most severe disease epidemics that occur in Australia (Bokor et al., 1975; Barbetti and Khangura, 1999). Although inoculation on older plants can still lead to crown canker development, it is often too late to cause severe yield loss. In contrast, in the other major oilseed rape growing regions of the world outside Australia, such as North America and Europe, cotyledon infection is generally of little importance in the majority of situations (West et al., 2001), probably because of the lack of close synchronization of timing of ascospore release with seedling establishment that occurs in southern Australia. For example, on winter-type B. napus in the United Kingdom, Hammond and Lewis (1986) found that the most damaging cankers were generally formed from leaf infections occurring before the onset of rapid stem extension.

Environment conducive to the pathogen

The south-west cropping belt in Western Australia is a region with a Mediterranean climate, with a cool moist winter growing season with rainfall peaking in mid-winter and with up to 75 % of rainfall in the period May–October (Potter et al., 1999). The late spring period is often characterized by severe moisture deficiency and this is then followed by a hot summer period frequently with little or no significant rainfall. The unfavourably hot and dry Mediterranean summer favours the conservation of inoculum of most residue-borne necrotrophic fungal pathogens in this region (Sivasithamparam, 1993). The soils in this region are old, severely weathered, heavily leached, inherently nutritionally deficient (particularly in phosphorous, potassium, nitrogen and some trace elements), have very low biological activity and frequently poor structure. Plants growing in these nutrient-impoverished soils are severely predisposed to necrotrophic pathogens (Sivasithamparam, 1993).

OVERCOMING SINGLE DOMINANT GENE-BASED RESISTANCE

Hua Li et al. (2003) initially determined the severity of cotyledon lesions and crown cankers in cv. Surpass 400 with L. maculans strains that were capable of overcoming the Brs R gene, such as UWA S4, UWA S7 and UWA S12. The disease caused by these strains was not as severe as that shown by the highly susceptible cultivars, showing that polygenic resistance, termed ‘residual resistance’ by Zadoks (1961), is also present in cv. Surpass 400. Surpass 400 and the highly susceptible cultivars, which Hua Li et al. (2003) used, have different parental origins (Crouch et al., 1994; Buzza and Easton, 2002; Easton, 2001; Li and Cowling, 2003). Such polygenic resistance could be important, even in cultivars considered to be protected by single dominant gene-based resistance. However, in 2003 and 2004, there were reports of overcoming of resistance in Australian commercial cultivars containing the Brs R gene, initially in Western Australia (Hua Li et al., 2003; Li et al., 2003) and later in other parts of southern Australia (Canola Association of Australia, 2004a). Hua Li et al. (2004a) subsequently demonstrated, for the first time, that the Brs R gene contained in other cultivars, such as Surpass 402 CL, Surpass 404 CL, Surpass 501 TT, Surpass 603 CL and Hyola 60, all commercially released in 2001 (Burton, 2003), could also be overcome by some Western Australian isolates of L. maculans.

Studies by Li et al. (2003) and Hua Li et al. (2003, 2004a, 2005a) clearly demonstrate the occurrence of a high degree of physiological specialization amongst L. maculans populations, as evident in the specific pathogenicity of Western Australian isolates on specific cruciferous taxa. This confirms studies by others in Australia (Purwantara et al., 2000; Sosnowski et al., 2001; Barrins et al., 2002). It is this heterogeneity in L. maculans populations that probably allows this pathogen to overcome host resistance, especially where it is single dominant gene-based resistance. Hua Li et al. (2005a) found the virulence of L. maculans isolates in Western Australia was related to their origin and highlighted the increased virulence of some particular isolates on cultivars containing the Brs R gene, compared with isolates collected 1 or 2 years earlier (Hua Li et al., 2003; Li et al., 2003). Hua Li et al. (2005a) not only confirmed the existence of highly virulent resistance-breaking isolates in Western Australia but that these isolates were present over the cropping seasons of 2001 and 2002. With increased selection pressure resulting from widespread deployment in Western Australia of cultivars containing the Brs R gene, it is not surprising that the more aggressive isolates had a significant adverse impact on these particular types of cultivars in commercial situations in Western Australia in the 2004 season (Hua Li, M. J. Barbetti and K. Sivasithamparam, unpubl. res.).

As polygenically inherited resistance is more stable, utilization of this type of resistance is likely to provide much more durable control of blackleg (Parlevliet, 1993; Chèvre et al., 1996; Pilet et al., 1998). It is vital to include very strong and durable polygenic resistance in addition to any single dominant gene-based resistance that might be deployed in oilseed rape cultivars. Durable disease resistance is widely accepted as a desirable objective (Stuthman, 2002). One possible way to achieve durability is to maximize high levels of polygenic resistance. This needs to be a key strategy for Australian breeders, even in cultivars containing single dominant gene-based resistance (Hua Li et al., 2003, 2004a). This will at least ensure that any future overcoming of resistance does not result in the complete collapse of the cultivar from blackleg disease as now occurs in some field situations in Australia.

One of the most significant changes in disease management in canola (oilseed rape) occurred during the period 2000–2003 when single dominant gene-based resistance was effective in Australian cultivars. Growers quickly recognized that simply utilizing this resistance alone could provide them with effective management of blackleg and that there was no longer any requirement to assess disease risks or to apply any cultural or chemical management options. The failure of single dominant gene-based resistance in Western Australia has highlighted to growers the risks and penalties associated with abandoning traditional cultural management strategies, including use of polygenic resistance as a major strategy (Canola Association of Australia, 2004). There is now strong interest in integrated management strategies rather than sole use of host resistance, including improved risk assessment for individual growers.

Leptosphaeria maculans race diversity and source of diversity of genes

The races of L. maculans recently designated by Balesdent et al. (2005) from isolates collected in Europe, Australasia and Canada were dominated by Western Australian isolates that constituted 13 of the 26 races delineated to date. As these 13 races were delineated from a random batch of 18 isolates, it is highly likely that several more races will be delineated as the screening is extended to include more isolates from across the Southern Australian grain belt. The very high degree of diversity of isolates from Western Australia is not surprising, and explains both the ability of this pathogen recently to overcome the Brs R gene (Li et al., 2003) along with observations of apparent reductions in the level of field resistance of one or more Australian cultivars that effectively only contain polygenic resistance (Ballinger and Salisbury, 1996; Gororo et al., 2004; M. J. Barbetti, Hua Li and K. Sivasithamparam, unpubl. res.). It is clear from the work of Rouxel et al. (2003b) that there have been one or more major shifts in populations of L. maculans in Australia previous to the introduction of the Brs R gene. This shift could be due to the use of cultivars developed in Western Australia, carrying the Rlm4 single dominant gene-based resistance (Roy and Reeves, 1975; Roy, 1978). Similar population shifts from avirulence to virulence have also been observed in France for the single dominant resistance gene Rlm1 between 1990 and 2000 (Rouxel et al., 2003a), with associated overcoming of this host resistance by L. maculans in commercial crops in France (Brun et al., 2000; Balesdent et al., 2002; Rouxel et al., 2003a).

The study by Balesdent et al. (2005) clearly indicates that Australian isolates of L. maculans, in general, have a wider range of virulences compared with isolates from France and Canada. Perhaps this is due, at least in part, to widespread commercial deployment of cultivars containing the Brs R gene in Australia, that have not only been sources of new resistance-breaking isolates (Hua Li et al., 2003; Li et al., 2003), but have also been sources of isolates with increased levels of virulence against polygenic resistance (Hua Li et al., 2004a, 2005a). Hua Li et al. (2005a) found isolates that could overcome resistance in B. juncea, B. nigra and B. carinata, similar to earlier findings for B. juncea in eastern Australia (Ballinger and Salisbury, 1996; Chen et al., 1996; Purwantara et al., 1998; Marcroft et al., 2002) and for B. juncea, B. nigra and B. napus–B. juncea recombinant lines in France (Brun et al., 2000, 2001; Somda et al., 1996, 1998, 1999). The presence of a wider range of virulences in Australian L. maculans populations may also be a consequence of repeated exposures to alternative sources of resistance to L. maculans populations at experimental sites, including blackleg disease nurseries, and at sites where there have been small plantings of alternative cruciferous host crops. The diversity of virulences in Western Australia may have existed previous to the introduction of the Brs R gene. This is evident in the ability of strains WAC 2242 and WAC 2244, isolated in 1972 from B. napus in Western Australia, that are capable of overcoming the Brs R gene (Hua Li et al., 2005a). In the absence of widespread canola cropping between 1973 and 1991, alternate hosts, including cruciferous weeds, may not only have maintained populations but may also have contributed to the development of new virulent races.

Boom–bust cycle from using single dominant gene-based resistance

Studies from Western Australia on cultivars with the Brs R gene (Hua Li et al., 2003) and those from France for B. napus–B. juncea lines (Brun et al., 2000) and commercial cultivars containing Rlm1 resistance (Rouxel et al., 2003a), demonstrated that single dominant gene-based resistance can rapidly be overcome following exposure of this type of Brassica germplasm to field populations of the pathogen. In relation to the management of the disease, this raises several issues concerning the reliability of single dominant gene-based resistance and the need for alternative strategies. Pyramiding of resistance genes has been successful in the wheat-rust pathosystem (Pederson and Leath, 1988) and if achieved in the L. maculans–B. napus pathosystem with other single dominant and/or effective polygenic resistance genes, may be a useful way to produce highly promising cultivars with more durable resistance (Lindhout, 2002). Unfortunately, the nature and number of polygenic resistance genes in B. napus is currently undefined. However, such a strategy could still be evaluated for L. maculans in oilseed rape. The field efficiency of any race-specific resistance in B. napus, even in the short-term, will depend upon the population structure of L. maculans (Ansan-Melayah et al., 1997) (e.g. do strains with the ability to overcome race-specific resistance already exist in the population?), the migration of virulent pathotypes (Ansan-Melayah et al., 1997) (e.g. movement of blackleg-infested seed), and how such resistance is deployed and the interaction of population dynamics and genetics (van den Bosch and Gilligan, 2003). While single dominant gene-based resistance can be efficient for a period, under field conditions, provided the corresponding avirulent strains of the pathogen remain prevalent (Ansan-Melayah et al., 1995; Balesdent et al., 2001), future commercial use of existing single dominant gene-based resistance will have to be carefully managed to maximize the period of its durability. How this can best be achieved is still not yet clear. As outlined by van den Bosch and Gilligan (2003), measuring durability as ‘the time from introduction of the cultivar to the time when the frequency of the virulence gene reaches a preset threshold’ is at best questionable. They suggest that it is more appropriate to consider three measures of durability, namely (1) the expected time until invasion of the virulent genotype; (2) the virulence frequency related to the time required for the virulent genotype to take over the pathogen population; and (3) the additional yield due to the additional number of uninfected host growth days. Unfortunately, there was no attempt by breeders or growers, at the time of release of cultivars containing Brs R, to implement management strategies in Australia to maximize the period of effective use of the Brs R gene.

It is clear from the study of Balesdent et al. (2005) that the pathogen, especially in Australia, does not discriminate between hosts with single dominant gene-based resistance. Widespread deployment of single dominant gene-based resistance potentially provides the ideal situation both for rapid overcoming of resistance and for loss of effectiveness of host resistance genes, duplicating the boom–bust cycle for this pathogen in France (Balesdent et al., 2005). While the speed of adaptation of L. maculans to resistant host genotypes may depend on many factors, such as the frequency of genetic recombination and the degree of selection pressure imposed for virulence (Pang and Halloran, 1995), it is clear that environmental conditions in Australia are ideal for extensive pseudothecial development on residues over the autumn and/or early winter period (Bokor et al., 1975; West et al., 2001; Salam et al., 2003; Wherrett et al., 2004), which could further assist the rapid selection of L. maculans populations that overcome host resistance genes. Adaptation within L. maculans populations to overcome host resistance genes is potentially more rapid where there is a lack of diversity in host resistance genes (Parlevliet, 1993).

OPTIONS TO MANAGE BLACKLEG

Host resistance: seedling vs. adult host resistance

The use of resistant cultivars is the preferred way of controlling blackleg and has been used with varying degrees of effectiveness worldwide (Newman and Bailey, 1987) and is the major means for managing the disease in Australia (Barbetti and Khangura, 1999; Salisbury and Wratten, 1999) and elsewhere (West et al., 2001). Resistance to L. maculans is often categorized as either adult stage resistance or seedling resistance. Adult plant resistance is generally considered to be polygenic and strongly influenced by environmental conditions, while seedling resistance is often described as a single-gene trait (Ferreira et al., 1995; Ansan-Melayah et al., 1998; Pilet et al., 1998). Different scenarios have been reported by different researchers in relation to seedling vs. adult host resistance. For example, field studies in Western Australia demonstrated a clear link between cotyledon and adult plant susceptibility/resistance, both in B. napus cultivars with and without the Brs R gene (Hua Li et al., 2004a). Similarly, Brassica species with a B-genome (genome from B. nigra) have been reported to have a strong relationship between seedling and adult plant responses. They possess a resistance to blackleg that is ‘absolute and stable’ (Sacristán and Gerdemann, 1985, 1986) and is effective throughout the life of the plant (Roy, 1984; Rimmer and van den Berg, 1992; Zhu et al., 1993; Struss et al., 1996), such that cotyledon tests can provide an efficient method to select for blackleg resistance in the B-genome species such as B. juncea (Chèvre et al., 1997).

For B. napus cultivars (without the Brs R gene), a similar relationship between seedling and adult plant responses has also been reported in Australia (Hua Li et al., 2003, 2004a) and elsewhere (Bansal et al., 1994). However, examples of poor correlation between seedling and adult plant field reactions in some studies (Gugel et al., 1990; Ballinger and Salisbury, 1996; Pilet et al., 1998) could have resulted from the inoculation method used, the seedling stage selected for inoculation, or the level of infection (Newman and Bailey, 1987), or due to the virulence of the L. maculans isolates (Purwantara et al., 1998). A clearer definition of the relationship between seedling and adult host resistance, using precise levels of inoculum on seedlings under controlled conditions, is needed if available resistance sources and types are to be fully exploited by breeders.

Host resistance: basis of host resistance in Australian cultivars

Mechanisms of resistance to infection processes of cultivars with polygenic and single dominant gene-based resistance have been studied by Hua Li et al. (2004b) with Australian spring-type B. napus oilseed rape cultivars challenged with isolates that were incapable of overcoming single dominant gene-based resistance. They found no differences in relation to the processes of ascospore and pycnidiospore attachment, germination and penetration on Australian cultivars with either polygenic or single dominant gene-based resistance. In contrast, Sosnowski et al. (2004) reported inhibition of pycnidiospore germination and mycelial growth on cv. Hyola 60, a cultivar with the Brs R gene. However, Hua Li et al. (2004b) observed major differences post-penetration on cotyledons between resistant and susceptible cultivars. They described in detail the hypersensitive response and the host–pathogen interactions between spring-type B. napus carrying single dominant gene-based resistance and L. maculans. They showed that disease symptoms on cv. Surpass 400 were evident as early as 3–5 d post-inoculation (dpi) with ascospores and 5–7 dpi with pycnidiospores. However, on other cultivars (either susceptible or with polygenic resistance) symptoms were not evident until 10 dpi with ascospores, 12 dpi with pycnidiospores. In Australian cultivars effectively containing only polygenic resistance, Hua Li et al. (2004b) showed that fungal growth was much weaker, and host cells were relatively less damaged than in a susceptible cultivar. Cultivars with only polygenic resistance can offer useful resistance in spring-type B. napus to L. maculans, based upon the additive effects of several genes (Parlevleit, 2002). Although these ‘minor’ genes are inadequate for a hypersensitive response, they can successfully contain the disease to a significant degree (Hua Li et al., 2004b).

Host resistance: the search for new sources outside B. napus

New sources of improved resistance to L. maculans, outside B. napus, have been sought amongst Brassica species with B-genomes including B. juncea with AABB, B. nigra with BB, B. carinata with BBCC (Roy and Reeves, 1975; Roy 1978; Mengistu et al., 1991; Johnson and Lewis, 1994) and in other cruciferous hosts (Barbetti, 1978b; Salisbury, 1987, 1991; Mengistu et al., 1991; Johnson and Lewis, 1994; Tewari et al., 1996; Snowdon et al., 2000). Brassica species with the B-genome and single dominant gene-based resistance are known to have a hypersensitive type of resistance response to L. maculans that can be effective throughout the life of the plant (Roy, 1984; Rimmer and van den Berg, 1992). This response has variously been reported to be monogenic or oligogenic, both in Australia (Roy, 1984) and in other locations (Hill, 1991; Rimmer and van den Berg, 1992). Crosses between B. napus and species with the B-genome have given rise to several resistant oilseed rape lines in Western Australia (Roy, 1984) and elsewhere (Sacristán and Gerdemann, 1986; Sjödin and Glimelius, 1988, Struss et al., 1991; Zhu et al., 1993; Chèvre et al., 1996; Plieske et al., 1998; Eber et al., 1999; Brun et al., 2001). Brassica juncea, B. nigra and B. carinata are now being targeted as sources of resistance for Australian breeding programmes (Purwantara et al., 1998; Marcroft et al., 2002). Wild crucifers, both closely and distantly related to B. napus, are being targeted as potential resistance donors in other breeding programmes (Séguin-Swartz et al., 2000).

Hua Li et al. (2005a) showed that one or more Western Australian L. maculans isolates could overcome resistance in 12 of the 13 cruciferous taxa they tested. The findings of their study demonstrated the wide host range of these isolates within the Brassicaceae. This has major implications for the widespread deployment of cruciferous species, such as B. juncea, Crambe abyssinicia, Eruca vesicaria or Raphanus sativus as biofumigation green manure crops, B. juncea, B. carinata, Sinapis alba or C. abyssinicia as sources of high erucic acid industrial oils, or, in particular for all the above species, as sources of single dominant gene-based resistance to blackleg for introgression into B. napus. In addition, some strains were found that attacked cruciferous hosts other than B. napus that were equally or more virulent on B. napus. It is clear that some Western Australian field isolates of L. maculans present in the oilseed rape-growing areas of Western Australia are capable of overcoming the blackleg resistance sourced from the B-genome, such as the resistance of B. juncea, B. nigra and B. carinata (Hua Li et al., 2005a). While other cruciferous taxa such as R. raphanistrum, C. abyssinicia, E. vesicaria and R. sativus are all reported to have good levels of resistance to most isolates of L. maculans tested elsewhere, a few field isolates in the study by Hua Li et al. (2005a) were found to overcome these resistances and cause severe blackleg disease. Western Australia may be the only region with L. maculans isolates highly virulent on E. vesicaria and R. raphanistrum. This finding is confirmed by field observations at two sites in Western Australia over the past three seasons of severe natural blackleg epidemics on R. raphanistrum seedlings, in one instance causing significant seedling death (M. J. Barbetti, unpubl. res.). Camelina sativa is an annual species that has attracted increasing attention in recent years as an alternative crop for oil production in Western Australia (M. Campbell, personal communication) and elsewhere (Schuster and Friedt, 1995) and appears resistant to the L. maculans isolates tested to date in Australia (Salisbury, 1987; Hua Li et al., 2005a).

The ability to overcome resistance in a variety of cruciferous species was not only found in isolates of L. maculans collected by Hua Li et al. (2005a) in recent years, but also with isolates (e.g. WAC 2242 and WAC 2244), collected in 1972 and one or both of which overcame the resistance in B. napus cv. Surpass 400, a B. napus × B. juncea line, B. juncea, C. abyssinicia or E. vesicaria. Hua Li et al. (2005a) also found that another isolate (WAC 4048), collected in 1984, overcame the resistance in B. napus cv. Surpass 400, a B. napus × B. juncea line, C. abyssinicia and S. alba. A further isolate (WAC 2281), collected in the 1973 from R. raphanistrum, overcame the resistance in a B. napus × B. juncea line, B. juncea, R. raphanistrum, B. carinata, C. abyssinicia and S. alba. This is despite there being no evidence of host specificity in a previous study (Barbetti, 1978b) where isolates from B. napus and R. raphanistrum were equally pathogenic on B. napus, B. rapa, R. raphanistrum and B. tornefortii. Findings by Hua Li et al. (2005a) clearly illustrate a potential spectrum of virulence within L. maculans populations in Western Australia that enables them to overcome single dominant gene-based resistance in these particular cruciferous hosts, many of which have yet to be introduced as commercial crops. The broad spectrum of variation in virulence within Australian L. maculans populations clearly poses a major question in relation to future use of new sources of single dominant gene-based resistances to blackleg for introgression into B. napus (Salisbury, 1991; Crouch et al., 1994), including use of wild germplasm (Lenné and Wood, 1991) such as wild Brassica species. These findings, along with other reports describing L. maculans isolates virulent on B. juncea and B. nigra (Purwantara et al., 1998; Somda et al., 1999; Brun et al., 2001), clearly show that it is imperative to determine whether Australian L. maculans populations have a pre-existing ability to overcome alternative sources of single dominant gene-based resistance and/or a capacity to quickly acquire such an ability.

Additionally, there are major implications for any future widespread deployment, particularly in Australia, of non-B. napus cruciferous species as crops in their own right. Two canola-quality B. juncea mustards will be sown as large strip trials across Australia in 2005 to allow commercial evaluation of oil and meal quality and for seed multiplication for commercial release in 2006 (Salisbury et al., 2004). This is despite the fact that B. juncea-attacking L. maculans isolates are already confirmed to be present in eastern (Purwantara et al., 1998) and Western Australia (e.g. WAC2242) since 1972 (Hua Li et al., 2005a). This impending commercial release of B. juncea without any scheme to effectively manage the deployment of resistance suggests the need for caution by Australian oilseed rape breeders, especially in light of the overcoming of the Brs R gene. Clearly, many of these potential alternative sources of single dominant gene-based resistance outside of B. napus could be expected to fail after 3–4 or more years of commercial field exposure to Western Australian L. maculans populations, particularly in the absence of protocols to strategically manage the deployment of such resistances. Such strategies must be developed and implemented in association with any future deployment of single dominant gene-based resistance, such as strategies considered for France (Pinochet et al., 2003).

Host resistance: need to return to primary reliance on polygenic resistance for more durable management of L. maculans virulence genes

There is clear and strong evidence to indicate the potential to rely upon cultivars with polygenic resistance. First, there is the common occurrence of relatively low levels of the blackleg on cultivars with polygenic resistance in South Australia and Western Australia in areas where the Brs R gene has been overcome (Canola Association of Australia, 2004; Hua Li, M. J. Barbetti and K. Sivasithamparam, unpubl. res). Secondly, polygenic resistance has generally remained effective against blackleg in the field, despite the presence of strains capable of overcoming this resistance under controlled conditions (Hua Li et al., 2004a) and despite indications of declining effectiveness of polygenic resistance over time in one or more cultivars (Gororo et al., 2004). Thirdly, the ability of certain avirulent strains to prevent the Brs R gene from being overcome by certain strains (Hua Li et al., 2005b) may also have potential for further protection of cultivars (see section on ‘Cultural control: novel cross protection’). Despite these limitations and concerns, targeting polygenic resistance should remain a future focus of Australian breeding programmes until a way is found to utilize currently known single dominant gene-based resistance sources in a more durable way.

Only in France, has there been any real attempt to promote a strategy for better management through durable resistances based on three practices: develop methodologies to better characterize host genotypes, promote appropriate agronomic practices and closely monitor L. maculans population behaviour (Pinochet et al., 2003). For Australia, it has been Western Australia that has led the way in developing a long-standing and effective approach to promote the best cultural and management practices (Barbetti and Carter, 1986; Barbetti, 1994; Barbetti and Carmody, 1998; Barbetti and Khangura, 1999, 2000; Barbetti et al., 2000a, b, 2003). It was Western Australia that initiated studies to define the nature of host resistances available in Australian breeding programmes (Hua Li et al., 2004b).

Host resistance: managing increasing L. maculans virulence

There is an increasing frequency of isolation of strains with increased virulence against cultivars containing the Brs R gene (Hua Li et al., 2005a). For example, of the isolates collected from Western Australia in 2001, when the area sown to cultivars with the Brs R gene constituted 8·6 % of the total area sown (Littlewood, 2003), 22 % of isolates collected overcame the Brs R gene present in cv. Surpass 400 (Hua Li et al., 2003). However, of the isolates taken in the following year (2002), when cultivars containing the Brs R gene constituted 56 % of the area sown to oilseed rape in Western Australia (Littlewood, 2003), 78 % of isolates collected overcame resistance in cv. Surpass 400 (Hua Li et al., 2005a). This situation is similar to that observed in France by Chèvre et al. (1996) who proposed that ‘the presence of the B. nigra blackleg resistance gene in oilseed rape cultivars may induce a change in the structure of the L. maculans populations which might induce breakdown of the introduced resistance’. It is probable that it may not be an induction but rather a selection for specific populations by the specific hosts. The climatic conditions in Western Australia favour the retention of infested residues for long periods The masses of pseudothecia produced on them over the autumn and/or early winter period, may assist L. maculans populations to evolve and/or be selected rapidly to overcome host resistance. From 2000 to 2003, B. napus cultivars containing the Brs R gene were grown on increasingly large acreages across Western Australia (Littlewood, 2003). The increased proportion of infested residues resulting from this widespread deployment of cultivars containing the Brs R gene contributed to the development of L. maculans strains, such as UWA 192 (Hua Li et al., 2004a), that could have a significantly adverse impact on the yield in future seasons.

While Hua Li et al. (2005a) found that B. napus cv. Dunkeld was relatively susceptible to most L. maculans isolates from B. napus cultivars (without the Brs R gene), it was more resistant to most isolates from either B. juncea or B. napus cultivars with the Brs R gene. This may help explain the relatively good performance of certain commercial crops of cultivars effectively containing only polygenic resistance at sites where cultivars containing the Brs R gene from blackleg have failed (Canola Association of Australia, 2004; Hua Li, M. J. Barbetti and K. Sivasithamparam, unpubl. res.). However, it must be noted that there are now in existence in the Western Australian field environment, pathogen strains that currently are compatible with both the monogenic and polygenic forms of resistance.

Host resistance: need to monitor virulence alleles in L. maculans populations

There is clearly a need to monitor L. maculans virulence alleles in L. maculans populations in order to predict changes in the spectrum of host resistance needing to be deployed in specific regions. While there are molecular markers for some virulence alleles, there is a need to develop further markers for the rest of the virulence alleles, such that single dominant resistance genes can be defined in new cultivars in relation to the specific virulence alleles present in areas where these cultivars will be grown. It appears that there has been, and continues to be, a particularly rapid selection by host R genes for specific strains of L. maculans in Australia. The segregation of strains based on the AvrLm1-9 alleles (Balesdent et al., 2005) of Western Australian strains of L. maculans is based on host differentials (Hua Li et al., 2005a; Hua Li, K. Sivasithamparam and M. J. Barbetti, unpubl. res.), and is particularly timely. Studies using host differentials have identified several (Hua Li et al., 2004a, 2005a) potential races there, in addition to the seven ‘adult races’ identified earlier by Ballinger and Salisbury (1996) in eastern Australia.

The source of the L. maculans virulence variation is particularly intriguing and of wide relevance internationally. For Western Australia, the population shifts associated with the sexual recombination capacity of this fungus are favoured by numerous factors: the conducive environment for particularly severe epidemics; the extensive residue carryover; the exposure of L. maculans populations to a wide range of cruciferous host resistances; the importation into Western Australia of infested seed from other regions (Wood and Barbetti, 1977b); and consistently permitting the disease to exceed economic thresholds in terms of management strategies applied. All these factors are likely to have contributed to the wide range of virulences in Western Australia. However, the long association of L. maculans with cultivated and wild weed cruciferous species in Western Australia has been the most significant source of this diversity of virulences. The role of the saprophytic phase in the post-senescence saprophytic growth, colonization and survival (sensu Garrett, 1970) of the pathogen remains to be determined. During the growing season, the pathogen is considered to sporulate mainly from pycnidia (asexual) in lesions on cotyledons, leaves and stems, while the predominant production of pseudothecia (sexual) occurs on senesced stems. There is, however, no evidence to suggest that the fungus can survive parasitically or saprophytically on non-hosts or their residues between rotations of oilseed crops. The effective production of inoculum in oilseed rape residues 1–3 years after harvest in Western Australia contrasts with the observation that only stubbles of immediate preceding crops are important elsewhere (Marcroft et al., 2004a). This is likely to be a consequence of the ‘conservation of inoculum’ (sensu Sivasithamparam, 1993) that occurs in the Mediterranean environment of Western Australia and not from saprophytic spread of the fungus. This raises the potential need to manage changes in the pathogen field populations through control of cruciferous weeds. The genetic structure of the Western Australian L. maculans population needs to be analysed in order to determine its evolutionary potential (McDonald and Linde, 2002a, b). If it has a high evolutionary potential then the L. maculans population there is more likely to overcome genetic resistance. In addition, a sound knowledge of the population genetic structure of the pathogen may offer an insight into the best breeding strategy for durable resistance (McDonald and Linde, 2002a).

Perhaps the unrestricted transfer of L. maculans infested seed (e.g. into Western Australia from countries outside Australia and from eastern Australia) and the importation of infested residues from Australia (McGee and Petrie, 1978), may partially explain why different regions have some common virulences in their L. maculans populations. It is vital that the virulences present in Australia, and particularly in Western Australia, are not inadvertently imported to other major oilseed rape-producing countries of the world. It will be an impossible challenge to prevent random transmission of seedborne infections across national and international borders. If indeed such successful transmissions are rare, it is most likely due to unfitness of the strain(s) in the new environment rather than the effectiveness of current quarantine protocols.

Cultural control: back to the basics for disease management

In the presence of virulent pathogen populations, polygenic host resistance needs to be utilized within a framework of good integrated management practices as developed for this disease earlier in Australia (Barbetti and Carter, 1986; Barbetti and Carmody, 1998; Barbetti et al., 1999, 2000a, b). Such cultural practices include residue management (Barbetti and Khangura, 1999; Wherrett et al., 2003) and avoidance of early seedling disease events (Salam et al., 2003). It is essential for the management of this disease that growers ‘go back in time’, to again focus on integrated disease management, if they are to make secure their future in the oilseed rape industry. In addition to using cultivars with the best polygenic resistance suited to their particular region, growers need to follow the basic procedures outlined by Barbetti et al. (2003), namely (1) avoid proximity to residues, especially those from the immediate previous season; (2) dispose of residues where possible; (3) consider using fungicides, especially where host resistance is intermediate and disease pressure is likely to be moderate to severe; and (4) try to sow at a time to avoid synchronization of seedling emergence and major ascospore showers.

There have been many previous studies aimed at determining the efficacy of chemicals for control of blackleg. For example, foliar fungicide applications in Australia (Barbetti, 1975b; Brown et al., 1976) and in other countries (Steinbach et al., 1991; Schramm and Hoffman, 1992; Rempel and Hall, 1995; West et al., 2002); seed treatments in Australia (Ballinger et al., 1988a, b; Barbetti and Khangura, 1999; Khangura and Barbetti, 2002) and elsewhere (Kharbanda, 1989); and fungicide coating of fertilizers in Australia (Ballinger et al., 1988a, b; Khangura and Barbetti, 1999, 2002). However, contrasting results have often been obtained across these studies. Unfortunately, the earlier studies with fungicides in the 1970s and 1980s frequently gave, at best, only limited relief from damage caused by blackleg (Barbetti, 1974, 1975b, 1982a, b, c, 1988a, b) under the Western Australian environment in association with the host resistances available at those times. However, the prime focus in the application of cultural control measures is to reduce the disease inoculum to below the threshold of that needed for serious yield losses to occur.

Cultural control: an integrated approach

It cannot be overestimated the pivotal role that cultural practices play, not only in relation to their direct effect in reducing the degree of exposure to the inoculum, but in maximizing the effectiveness of available host resistance, be it polygenic or dominant gene-based resistance. Significant beneficial epidemiological synergism can be derived through combining disease management tactics at the regional scale. This presents the pathogen with multiple barriers to overcome. In addition, integrated management has the potential to increase the durability of host resistance (Mundt et al., 2002). Cultural practices such as burning of infested residues are widely practiced in some years in Western Australia. However, burial/disposal of residues through cultivation is now not popular since the widespread adoption of minimum tillage practices (Barbetti and Khangura, 1999; Barbetti et al., 2000a, b, 2003). Manipulation of sowing date (Khangura and Barbetti, 2004) specifically based upon accurate regional forecasts for the timing of ascospore showers, offers significant avoidance of the heaviest ascospore showers in some regions and seasons (Salam et al., 2003). There has been strong interest in this forecasting advice in Western Australia, both prior to the availability of cultivars with the Brs R gene, and, particularly, following their recent demise. However, there are other cultural possibilities that have not been evaluated in Australia, such as the manipulation of nitrogen availability, as recently assessed in France (Aubertot et al., 2004).

Cultural control: potential for interfering with pathogen development on residues

Conditions that are conducive for disease development may not necessarily be the same as those for pathogen reproduction. Marcroft et al. (2004b) showed that residues from B. carinata, B. nigra, S. alba and B. napus cv. Surpass 400 had lower pseudothecial densities and discharged fewer ascospores than residues of other B. napus cultivars. They suggested that if this trait of low blackleg inoculum from residues could be introgressed into commercial oilseed rape cultivars, blackleg disease severity could be substantially reduced, resulting in higher and more stable yields. However, it is possible that such restriction on the ability of L. maculans to sporulate on residues increases selection pressure on the fungus to overcome it and may contribute towards the rapid overcoming of the Australian cultivars carrying the Brs R gene.

The recent overcoming of resistance in cultivars with the Brs R gene in the oilseed rape industry in Australia highlights the importance of utilizing the full range of alternative means for blackleg control. The disease control potential for directly interfering with pathogen development on infested residues using chemicals could be significant. The direct application of chemicals to oilseed rape residues in order to decrease blackleg disease pressure on the crop has been researched (Wherrett et al., 2003, 2004; Humpherson-Jones and Burchill, 1982; Rawlinson et al., 1984; Petrie, 1995; Turkington et al., 2000). However, no commercially feasible method of chemically treating infested oilseed rape stubbles or residues yet exists. Wherrett et al. (2003), in particular, demonstrated significant disruption of L. maculans pseudothecial development on oilseed rape residues from the application of the demethylation-inhibiting fungicides, fluquinconazole and flutriafol, and also with a herbicide, glufosinate-ammonium, with almost total control of ascospore emissions. Wherrett et al. (2003) described two potentially useful scenarios in regard to utilizing chemicals to manipulate disease pressure on the crop. First, a number of chemicals, such as fluquinconazole, flutriafol and gluphosinate-ammonium, were able to delay pseudothecial development and subsequent ascospore discharge was decreased, giving growers the potential to minimize synchronization of ascospore discharge with early crop establishment. Secondly, a situation where pseudothecial development was not delayed, but the number of ascospores released was reduced (e.g. due to ziram application). This latter scenario would only be effective if the reduction was sufficient to bring disease inoculum to below that required for a severe disease epidemic.

The study by Wherrett et al. (2003) also demonstrated that measurements of pseudothecial maturity could serve as a reliable indicator within a monitoring programme to predict the timing of ascospore discharge from oilseed rape residues, providing growers with a sowing date that minimizes synchronization with ascospore showers and the highly susceptible early seedling stage. The study by Wherrett et al. (2003) supported the model of Salam et al. (2003) that pseudothecial maturity and ascospore discharge in the southern Australian Mediterranean environment are programmed to coincide with seedling emergence and early development.

Cultural control: strategies to bring inoculum below a threshold

Management strategies that reduce the population size of the pathogen can have significant effects on the epidemiology of a fungal pathogen. This aspect is considered to be separate from the potential of integrated management to increase the durability of resistance through impacts on pathogen population genetics (Mundt et al., 2002). Wherrett et al. (2004) established, for the first time, the relationship of reduced inoculum levels to subsequent reductions in seedling disease and crown cankers in field-sown oilseed rape. In particular, they showed evidence of an inoculum threshold of 25 × 10⁵ ascospores stem⁻¹, above which, severity of seedling disease and crown cankers in oilseed rape cv. Dunkeld were little further affected. At spore loads below this threshold, however, they showed that there was a highly responsive relationship between inoculum level and disease where reductions in ascospore number from 25 × 10⁵ ascospores stem⁻¹ (estimated to be 2·8 × 10¹¹ ascospore ha⁻¹) to 5 × 10⁵ ascospores stem⁻¹ (estimated to be 5·6 × 10¹⁰ ascospore ha⁻¹) gave pronounced decreases in both seedling disease and crown cankers. While this relationship could be expected to be affected by host genotype and environment, their data still provided a potential avenue for minimizing yield loss from blackleg, especially in southern Australia where conditions frequently encourage synchronization of ascospore discharge with seedling establishment (Barbetti and Khangura, 1999; Salam et al., 2003). Crucially, the Wherrett et al. (2004) study demonstrated that crown canker severity could be reduced by nearly half as a result of significant reductions in ascospores discharged during the seedling phase from chemical treatment of residues and they estimated a yield increase of up to 900 kg ha⁻¹ as a consequence. This highlights the tremendous potential benefits that could come from interfering with L. maculans development on oilseed rape residue, prior to and during the early seedling stage. If the future widespread adoption of herbicide-tolerant genetically modified oilseed rape occurs in Australia, the application of herbicides such as glufosinate-ammonium could potentially provide additional benefits by way of reducing blackleg hazard levels during the growing season. The study of Wherrett et al. (2004) now clearly points to the target of what needs to be achieved through the deployment of cultural, chemical and genetic components of integrated disease management. The nature of agriculture in the southern Australian environment precludes the total elimination of the pathogen, but allows inoculum levels to be reduced below the threshold required for severe yield penalty.

Cultural control: novel cross protection

It is known that challenging a plant with an avirulent isolate can result in a more rapid response to infection, such that this could potentially provide increased protection against subsequent challenges from virulent pathogens (Kuc, 1987). Recent studies by Hua Li et al. (2005b) established that co-inoculation of pycnidiospores of an avirulent strain of L. maculans (UWA P11) with those of a Brs R gene breaking strain (UWA 192) prevented the overcoming of the Brs R gene in cv. Surpass 400. Hua Li et al. (2005b) showed that on its own an avirulent strain induced a hypersensitive reaction on cotyledons of cv. Surpass 400 and a virulent strain induced a fully susceptible reaction. However, when applied as a mixture to cv. Surpass 400, they found that the combination of strains gave only a hypersensitive response when the avirulant strain was as low as 10 % of the volume of the inoculum and this effect was still detectable at a low level at 1 % of the volume of the inoculum. This effect of the avirulent strain was localized and there was no evidence of a systemic effect between cotyledons or between separate point inoculations on the one cotyledon. It is possible that the observed effects could have been, at least in part, due to competitive antagonism between the different strains. These findings were in contrast to a study by Mahuku et al. (1996) who were the first to claim the existence of systemic acquired resistance in B. napus by demonstrating that the co-inoculation of highly virulent and weakly virulent isolates of L. maculans on a highly susceptible B. napus cultivar Westar reduced lesion size on leaves compared with the highly virulent isolate alone. This effect could only be demonstrated where the avirulent strain had been inoculated <64 h following inoculation with the virulent strain (Mahuku et al., 1996).

Cultivars containing only polygenic resistance appear to remain unaffected, or even perform better, in some fields where the overcoming of the single dominant gene-based resistance has occurred (Canola Association of Australia, 2004; Hua Li, M. J. Barbetti and K. Sivasithamparam, unpubl. res.). In contrast, Western Australian strains of L. maculans can overcome this same polygenic resistance under glasshouse conditions (Hua Li et al., 2004a, 2005a). The reason for the relative effectiveness of polygenic resistance in field situations where the Brs R gene has been overcome and, in certain cases, the retained effectiveness of this Brs R gene in some field situations (Hua Li, K. Sivasithamparam and M. J. Barbetti, unpubl. res.), is unclear and may not be due simply to inherent resistance to individual strains or races of L. maculans. Rather, the observed resistance may be due to various interactions observed in the glasshouse between L. maculans isolates by Mahuku et al. (1996) and Hua Li et al. (2005b). In particular, the results of Hua Li et al. (2005b) may explain the occurrence of crops of cultivars containing the Brs R gene that escape major collapse from the disease in the presence of virulent strains.

In cropping areas of Southern Australia where populations of virulent L. maculans represent a significant disease risk, exposure of cultivars with single dominant gene-based resistance to a source of canola residues colonized by an avirulent L. maculans might prove an effective means of reducing the risk. In situations in Australia, where fields are still sown with cultivars containing the Brs R gene, such resistance could be challenged by the presence of one or more resistance-breaking strains of L. maculans. In such sites, applying, or sowing near to, oilseed rape residues infested with one or more avirulent strains of L. maculans opens up a potentially novel management option where cultivars containing the Brs R gene could be protected from being overcome by the virulent resistance-breaking isolates. The significant implication of this work is that it has the potential for extending the period of effectiveness of single dominant gene-based resistance by manipulating field situations such that oilseed rape crops carrying single dominant gene-based resistance could be effectively immunized from the effect of virulent strains by the presence of avirulent strains of L. maculans. This exciting area of novel control warrants further investigation on a commercial scale, both for management of current situations in Australia where cultivars containing the Brs R gene are still being grown, and for its potential for management of other single dominant gene-based sources of host resistance that may be deployed in the future. Ideally, consideration needs to be taken of using strains incapable of outcrossing with virulent strains in the field and to optimize the inoculum levels to ‘out-compete’ virulent field strains. There should be little potential hazard from avirulent strains such as UWA P11 as they do not pose a disease threat to the host carrying the Brs R gene, even at high inoculum doses. This also could be another example where even a defeated resistance gene could still be used to provide crop resistance when integrated with other control measures as proposed by Pink (2002), especially when used in combination with strong polygenic resistance and effective cultural control measures.

Resolution of this challenge is not expected to be simple. It is likely, following an understanding of the basic cell biological signalling and genomic information involved in the host–pathogen interaction between L. maculans and oilseed rape, that we will be in a position to fully exploit available polygenic and single dominant gene resistances. Hopefully, this will lead to modifications to cultural practices and the host genome that will effectively interfere and/or block the inoculum behaviour and the parasitic activity of the pathogen and allow us to manage current and future threats from new and existing strains of this pathogen.