Introduction

Ascochyta blight (causal agent Ascochyta rabiei (Pass.) Labrousse, teleomorph Didymella rabiei (Kovachevski) von Arx) is the most important foliar disease of chickpea (Cicer arietinum) worldwide (Nene and Reddy 1987). The disease affects all aerial parts of the plant, originating from seed, and spread by wind and rain splash (Pande et al. 2005). Complete yield losses have been reported (Kaiser and Muehlbaur 1988; Navas-Cortes et al. 1998) and seed quality is commonly affected, with seed sometimes becoming unmarketable. An epidemic of ascochyta blight across northern New South Wales during the 2010 growing season was conservatively estimated to reduce the yield of commercially grown chickpea crops by thirty percent, even when fungicide was repeatedly applied (Kevin Moore pers. comm.).

Knowledge of isolate population variation within and among chickpea growing regions is required in order to determine the risk for the pathogen to potentially overcome disease management and control strategies. In particular, it is important to determine if the pathogen is able to adapt, and if so, to estimate the rate of adaptation to cause disease on newly released and widely adopted cultivars from the national resistance breeding program, such as Genesis 090 (Pulse Australia 2009c) and PBA HatTrick (Pulse Breeding Australia 2009).

To date, relatively low genetic diversity has been detected in the Australian A. rabiei population compared to that observed elsewhere (Phan et al. 2003a; Pradhan 2006; Leo et al. 2011). The detection of a low genetic diversity was most likely due to relatively few introductions resulting in localised founder populations that have then spread through the movement of infected seed and other materials (Phan et al. 2003a). Also, although known to be heterothallic, only a single mating type (MAT1-2) has been detected in Australia using a molecular diagnosis tool among all of the Australian isolates collected to date (Phan et al. 2003b; Pradhan 2006). Each of these factors contributed to the different pathosystem observed in Australia compared to the situation in other countries where both mating types have been identified and molecular diversity was high (Udupa et al. 1998; Navas-Cortes et al. 1998; Jamil et al. 2000; Chongo et al. 2004; Vail and Banniza 2009).

Whilst informative in determining levels of neutral genetic diversity among individuals within and between populations and in tracing genetic movement and exchange; molecular tools such as microsatellite markers do not provide an indication of phenotypic diversity among isolates. In order to determine differences in the ability to infect and cause disease on host genotypes, actual pathogenic interactions must be assessed. Aggressiveness is the natural variation in the severity of disease caused by isolates within the pathogen population towards a host genotype (Taylor and Ford 2007). Several previous studies have demonstrated the variation in pathogenicity and aggressiveness among A. rabiei isolates worldwide (Navas-Cortes et al. 1998; Chen et al. 2004).

Pathogenicity groups of A. rabiei have been identified in many countries (Vir and Grewal 1974; Udupa et al. 1998; Jamil et al. 2000; Chen et al. 2004; Pradhan 2006) based on measurements of aggressiveness of infection by isolates over a relatively small number of host genotypes. Whether all of these groups are distinct pathotypes, or part of a continuum of aggressiveness within the population is unknown. Vail and Banniza (2008) were unable to categorise 99 isolates collected from the Canadian prairies into discrete pathotypes and concluded that there was a continuous distribution of pathogenicity.

In order to produce chickpea cultivars with robust Ascochyta resistance, breeders need to screen germplasm against A. rabiei isolates that represent the diversity of pathogenicity within the current pathogen population. If pathogenic differences among isolates are present in Australia, isolates representing the breadth of these differences should be selected for screening purposes. Cultivars that are selected as resistant to all of the possible pathotypes will then stand the best chance of longevity in the Australian chickpea industry.

Two newly released Australian cultivars, Genesis 090 and PBA HatTrick, have moderately resistant Ascochyta blight ratings. PBA HatTrick is a cross between the Australian cultivar Jimbour with a susceptible ascochyta blight rating, and a resistant Iranian landrace (Pulse Breeding Australia 2009). Genesis 090 is an introduction from ICARDA Syria, selected and released by the National Chickpea Breeding Program in 2005 (Pulse Australia 2009c). Although cultivars Genesis 090 and PBA HatTrick have a resistant disease rating they are not immune to A. rabiei. Therefore, it is possible that in further large scale field plantings of these varieties, the pathogen will adapt to overcome these resistance sources. To date, neither PBA HatTrick or Genesis 090 have experienced a significant yield loss due to Ascochyta blight (Pulse Breeding Australia 2009; Pulse Australia 2008), although widespread adoption of these varieties by Australian chickpea growers is likely to increase selection pressure on the pathogen.

Understanding the risk to the durability of current resistance sources used in the Australian chickpea industry will better inform and prepare breeders and farmers for the disease management strategies that will be necessary to maintain good crop yields and quality. Therefore, the aims of this study were to 1) identify if differences in aggressiveness among Australian A. rabiei isolates exist from a range of locations and host genotypes, in order to assess the current risk from the most aggressive isolates to the newly released resistant chickpea cultivars, and 2) identify a representative group of isolates that may be used as a selection tool in future Australian chickpea resistance breeding programs.

Methodology

Isolate and host material

A representation of the population present during the 2009 season was collected and consisted of 24 A. rabiei isolates from six locations in Victoria, New South Wales and South Australia (Table 1). Two methods were used to collect isolates depending on the presence or absence of pycnidia. When pycnidia were present in lesions, they were picked off using a needle and placed into 1 mL of sterile water. The solution was then vortexed for 2 min before one loop of the solution was streaked over a V8 agar plate [containing 20% V8 vegetable juice, (Campbell’s Grocery Ltd), 2% technical agar (DifcoTM), 0.375% calcium carbonate (Merck) in distilled water]. Three days later a single colony was taken off the plate using a sterile needle under a dissecting microscope and placed onto V8 agar. When a lesion was present without visible pycnidia, the leaf tissue was surface sterilised using a 2.6% sodium hypochlorite solution for two minutes and plated onto V8 agar. Growth from the sterilised leaf material which was identified as A. rabiei was then subcultured onto a V8 agar plate. All isolates were single spore derived and maintained at 20°C with a 12-h photoperiod until required.

Table 1 Chickpea genotypes from which isolates were recovered, location of collection and corresponding isolate label of the 24 Ascochyta rabiei isolates for Fig. 1
Full size table

Twelve chickpea genotypes were selected as representative of a potential differential range in disease reaction to A. rabiei in consultation with Australian chickpea breeders and pathologists. They included current commercial cultivars as well as lines used as parents in the Australian breeding program, and susceptible commercial cultivars (Table 2). Chickpea plants were grown in 20 cm diameter pots containing commercial potting mix (BioGro), four plants per pot. Three replicates were sown for each of the genotype x isolate combinations, as well as three control pots containing the cultivar Howzat. All plants were grown in a shade house facility at the Grains Innovation Park, Department of Primary Industries, Horsham, Victoria.

Table 2 Chickpea genotypes used to screen Ascochyta rabiei isolates and their assigned ascochyta blight resistance ratings
Full size table

Inoculation and bioassay

Four weeks prior to inoculation, the 24 A. rabiei isolates were plated onto water agar containing autoclaved chickpea tissue. Inoculum was produced four weeks later by flooding the surface of the plate containing pycnidia with sterile water for 10 min. A glass rod was then used to gently scrape the surface and dislodge spores. A haemocytometer was used to adjust the concentration of spores to 1 × 105 spores/ml and Tween 20 (BDH) was added (0.05%) as a surfactant. The inoculum was sprayed onto four-week-old chickpea plants using a handheld pressurised sprayer (Preval® Power Unit). At the same time, the control plants of Howzat were sprayed with sterile water and Tween 20 only. Each plant was then covered with an individual minidome (MaxValu cup 6) (Chen et al. 2005) for 24 h to maintain humidity.

Each plant was scored for disease symptoms using the following rating scale, adapted from Singh et al. (1981), at 7, 14, 21 and 28 days post inoculation: Leaf lesions; 1 = No leaf symptoms, 3 = Pin prick lesions, 5 = Small, inconspicuous necrotic lesions without pycnidia, 7 = Individual lesions with darker margins and some pycnidia and 9 = Lesions with dark margin coalesced containing pycnidia.

The area under the disease progress curve (AUDPC) was then calculated using the following equation:

$$ {\text{AUDPC}} = \sum {\left[ {{(}{{\text{X}}_{\text{i}}} + {{\text{X}}_{{{\text{i}} + {1}}}})/2} \right]({{\text{t}}_{{{\text{i}} + 1}}} - {{\text{t}}_{\text{i}}})} $$

where, Xi is the blight score of the ith evaluation, Xi+1 is the blight score of the i + 1th evaluation and (ti+1 – ti) is the number of days between the evaluations.

Minitab® Statistical Software (release 15.1.1.0) was used for Analysis of Variance and NTSYSpc (version 2.21) was used for Principal Component Analysis (PCA) of means.

Results

Analysis of variance showed highly significant differences in mean disease scores (P < 0.0001) for both hosts and isolates as was the isolate by host interaction (P < 0.0001). There was no isolate by genotype combination that had a mean disease score value of one, where no leaf symptoms were present. The variation in the ability of individual isolates to cause disease across the host genotypes assessed was substantial and mean comparison of AUDP of isolates on each host genotype represented continua rather than identifying distinct groups (Table 3).

Table 3 Area under the disease progress curve for 24 Ascochyta rabiei isolates screened over 12 chickpea genotypes. LSD represents Fisher’s least significant difference at 5%
Full size table

Similarly, principal components analysis of area under the disease progress curve measures resulted in a continuum rather than discrete isolate groupings (Fig. 1). The first axis (Dim-1) accounted for 72.5% of the variation, the second (Dim-2) accounted for 9.8%, and the third (Dim-3) accounted for 7.1%. A cumulative total of 89.41% of the variation was explained by these first three axes.

Fig. 1
figure 1

Principal Components Analysis (PCA) of Ascochyta rabiei isolates based on their ability to cause disease (measured as area under the disease progress curve ) on twelve chickpea differentials. Axis Dim-1 accounts for 72.48% of the variation, Dim-2 9.81%, and Dim-3 7.12%. See Table 1 for isolate information. Isolates labelled with a ‘V’ were collected in Victoria, an ‘S’ were collected in South Australia and an ‘N’ were collected in New South Wales

Full size image

Isolates did not form any clear groupings but were spread over the PCA analysis. The only observable division in the isolates occurred in the middle of the graph. The isolates on the right hand side of the PCA were the most aggressive with decreasing aggressiveness towards the left of the graph. The four most aggressive isolates were from Victoria (09HOR04, 09KAL09 and 09KAN19) and Tamworth New South Wales (09TAM05). The two least aggressive isolates (09KIN11 and 09MEL04) were both collected from South Australian locations.

Discussion

A continuum of aggressiveness within the Australian A. rabiei population with no clear groupings was evident. In pathosystems where resistance are quantitative (Flandez-Galvez et al. 2003), it is unlikely that isolates will group in discrete pathotypes (Caten 1987). The A. rabiei – C. arietinum pathosystem was thought to be based on a quantitative interaction (Gowen et al. 1989; Vail and Banniza 2008), which suggested that a continuum of pathogenicity was more likely than true pathotypes. Although many other investigations have grouped isolates into pathotypes, in most cases few host genotypes have been used to differentiate them (Udupa et al. 1998; Jamil et al. 2000), and further investigation may find that the isolates tested did not represent the pathogen population. The statistical interaction between host genotypes and A. rabiei isolates also indicated that the pathosystem was based on quantitative resistance (Parlevliet 1979).

Of the four most aggressive isolates identified, three originated from locations in Victoria and one from a New South Wales location, over a thousand kilometres apart. It is likely that seed movement was responsible for the spread of isolates, the same method which allowed A. rabiei to enter Australia (Cother 1977). The location from which each isolate was collected was random in relation to aggressiveness among the isolates (Fig. 1), which also indicates that related A. rabiei isolates had been dispersed throughout the chickpea growing regions, most likely through infected seed. Molecular evidence has also shown that the origin of isolates did not explain any variation seen in genetic diversity (A. Leo pers. comm.). As Western Australia has a more strict process for regulating the importation of seed from other states (Quarantine 2009), it was possible that the isolates collected from this state may have been distinct from the eastern Australian isolates but this is yet to be determined.

A range of pathogenicity within the A. rabiei population was identified, and breeders can now select from these isolates to ensure that germplasm is screened against a representation of field populations. The use of these isolates in breeding programs will only be applicable for germplasm that is expected to be commercialised in the same areas that the isolates were collected from. For this reason the isolate selection used in screening breeding material should represent the future growing region of any specific cultivar. Therefore, it will be important to screen isolates from chickpea growing regions that were not covered in this study, such as Western Australia, and increase the number of isolates screened from regions such as Tamworth, NSW, where only two isolates were examined.

If recombination of the pathogen was occurring in Australia it would be possible for a wider range of aggressiveness to occur in the future. In order to ensure the reliability of this selection tool in breeding programs, continued sampling of isolates from all Australian chickpea growing regions should occur to identify any new isolates. Also, as this research is based on a single biological experiment and environmental replication, multiple environments testing will be required to validate the findings prior to implementation. As well as screening with single spored isolates, disease nurseries should also be utilised to assess resistance over a wide range of locations and against a wide range of isolates. This will help to ensure that breeding material is screened against the widest range of isolate aggressiveness even if the most aggressive isolates have yet to be collected.

Mean AUDPC values of each cultivar used in this study (Table 3) indicated that the published resistance ratings are mostly accurate with respect to the 2009 A. rabiei population (Table 2). The only exception is the cultivar ‘Almaz’ which had a much higher mean AUDPC than the moderately resistant and moderately susceptible cultivars.

As isolates have been found that can cause substantial disease on the most resistant genotypes it will be very important to emphasise both cultural and chemical control measures, and that growers do not rely on host resistance alone. The resistant genotypes such as Genesis 090 and PBA HatTrick are recommended to receive a foliar fungicide spray at early podding, as well as monitoring before this point, although further fungicide sprays are unlikely to be required (Pulse Australia 2009c; Pulse Breeding Australia 2009). It is important that this early monitoring occurs, so that if pathogenicity or aggressiveness has increased within the pathogen population, early disease symptoms can be observed and disease management procedures put into place.