Improving the understanding of fungicide resistance in Australian viticulture
Abstract
The grapevine diseases powdery mildew (Erysiphe necator), Botrytis bunch rot (Botrytis cinerea) and downy mildew (Plasmopara viticola) are managed with the use of fungicides. The development of fungicide resistance leads to reduced efficacy and field failure causing significant economic impacts. This project aimed to improve the understanding of fungicide resistance in Australian viticulture. In general, all three pathogens showed varying levels of reduced sensitivity to most of the fungicide groups tested, with resistance confirmed for fungicides from gp 13 for E. necator, gps 9, 12 and 17 for B. cinerea and gps 4 and 11 for P. viticola. It is recommended that fungicides be withdrawn from the spray program where resistance has been detected and that continued monitoring is carried out where reduced sensitivity has occurred. Techniques have been refined for improved monitoring of fungicide resistance with the development of a mini-greenhouse system to maintain and test isolates, rotorod spore traps to monitor for resistance, an in-field qPCR detection pipeline, Next Generation Sequencing to detect resistance and a heterologous yeast expression system to test resistance-associated mutations. For E. necator, there were strong relationships between phenotyping and genotyping for gp 11 (QoIs), but not for gp 3 (DMIs), and the resistance mutants G143A (QoI) and Y136F (DMI) were fit and stable for at least six months. The mutants H242R/Y (gp 7, SDHI, E. necator) and G1105S (gp 40, CAA, P. viticola) were not detected, but reduced sensitivity was recorded for both. Continued phenotypic and genetic monitoring will be crucial to manage and minimise fungicide resistance in Australian viticulture. Future research will develop in-field and high-throughput techniques in combination with rotorod spore trapping to achieve shorter turnaround and a high capacity for screening of known resistance-associated mutations.
Summary
Grapevine disease management can be compromised by resistance to fungicides used to control the most economically important foliar diseases in Australian viticulture; powdery mildew (Erysiphe necator), Botrytis bunch rot (Botrytis cinerea) and downy mildew (Plasmopara viticola). Fungicide resistant populations add to production costs by reducing fungicide efficacy, which can lead to the failure of spray programs for disease management. While laboratory testing has shown occurrences of fungicide resistance with these diseases in Australia, there is a need to evaluate the sensitivity of populations and hence the efficacy of the various fungicide products available to grape growers. The relationship between laboratory results obtained through bioassay (phenotyping), genetic analysis (genotyping) and field failure is still not well understood. In addition, the relative fitness of resistant populations compared with sensitive wild types (WT) in the field is still unclear. Several methods for monitoring fungicide resistance were investigated, such as the use of spore trapping, Next Generation Sequencing (NGS) and in-field real-time detection to increase the number of samples that can be processed. This project seeks to address these issues through the combined expertise of the research team in plant pathology and molecular diagnostics. The information generated will contribute to maintaining effective use of the available fungicide groups into the future, while reducing the economic and environmental impact of fungal diseases in Australian viticulture.
To determine the incidence and severity of fungicide resistance in Australian vineyards, isolates of E. necator, B. cinerea and P. viticola were collected between 2016 and 2021. A range of different collection methods were adopted and improved, together with a range of laboratory assays and genetic techniques. Analysis of sensitivity included the following fungicide groups: Quinone outside Inhibitors (QoIs, group 11), Demethylation Inhibitors (DMIs, gp 3), Amines (gp 5; spiroxamine), Aza-naphthalenes (gp 13; proquinazid and quinoxyfen), Succinate dehydrogenase inhibitors (SDHI, gp 7; boscalid and pydiflumetofen), Aryl-phenyl- ketones (gp 50; pyriofenone), Phenyl-acetamide (gp U6; cyflufenamid), Anilino-Pyrimidines (AP, gp 9), PhenylPyrroles (PP, gp 12), KetoReductase Inhibitors (KRI, gp 17), Carboxylic Acid Amides (CAA, gp 40) and Quinone outside Inhibitor, stigmatellin binding type (QoSI, gp 45).
A new mini-greenhouse system was developed to maintain E. necator populations and effectively conduct experiments to obtain consistent results. In addition, field label rate concentrations were introduced to the fungal growth testing regime, leading to the new ‘reduced sensitivity’ category. In general, E. necator isolates showed varying levels of reduced sensitivity to most of the fungicides tested; QoIs (20-62% of samples collected), DMIs (9-50%), spiroxamine (17%), proquinazid (58%), boscalid (42%), pydiflumetofen (30%) and pyriofenone (41%), but resistance was only confirmed for quinoxyfen in 60% of samples collected from regions in Vic, SA and WA. However, all tested isolates were sensitive to cyflufenamid. Continued monitoring for E. necator resistance is required to detect any shifts in resistance, especially where reduced sensitivity was detected.
For B. cinerea, total resistance frequency for fungicides from gps 9, 12 and 17 was close to 16%. New genotypes associated with gp 9 and 17 resistance, not previously described in grape isolates in Australia, were characterised. These include mutations in Pos5 (P293S), Mdl1 (E407K, G408R) and Erg27 (F412I/V). These new mutations are ideal targets for future development of qPCR assays for in vitro or in-field testing. Further resistance monitoring of these three key chemical groups is still required in all wine regions, and especially in regions where frequencies are elevated, such as Pemberton, WA and King Valley, Vic. Of particular interest is gp 12, where larger shifts in sensitivity are possible.
Resistance of P. viticola at field rate was detected for QoI fungicides (gp 11; 8% of isolates tested) and PhenylAmide (PA) fungicide (metalaxyl, gp 4; 19%), with the majority of resistant populations detected from regions in NSW and Victoria. In addition, reduced sensitivity was detected in 2-47% of isolates tested for fungicides from group 4, 11, 40 and 45. PA and QoI fungicides should be withdrawn from spray programs where resistance to metalaxyl and pyraclostrobin has been detected. It is recommended that monitoring for resistance be continued, especially where reduced sensitivity has been detected.
There were strong relationships between phenotyping and genotyping for QoIs. Bioassay results matched with presence of the mutant G143A (responsible for resistance to QoIs) with 92% and 90% agreement in powdery and downy mildew, respectively. This mutant will be an ideal candidate for high-throughput and in-field testing.
The relationship between phenotyping and genotyping for DMI fungicides in E. necator was weak and unclear as other factors may contribute to the resistance, such as gene copy number or other mutations. The Y136F mutant was detected in most of the isolates, but the phenotyping results showed only limited reduction in sensitivity. More research will be necessary before reliable genotyping can be performed for DMI resistance.
G143A and Y136F mutants were fit and stable for at least six months under laboratory and greenhouse conditions. This suggests that withdrawal of QoIs and DMIs from the spray program for six months may not be long enough to reduce mutant populations of E. necator.
The H242R/Y mutation linked to SDHI resistance was not detected in any E. necator populations, however reduced sensitivity was detected. Similarly, the G1105S mutant, linked to CAA resistance, was not detected in any isolates of P. viticola, however reduced sensitivity was again detected. Therefore, it will be important to continue monitoring for SDHI and CAA resistance in Australian vineyards.
The efficacy of DMI fungicides was evaluated when exposed to populations of E. necator, with varying sensitivity. The heterologous yeast expression system (HYES) revealed that there were significant shifts in sensitivity associated with the Y136F mutation to the DMI fungicides myclobutanil and triadimenol. These results support the findings that triadimenol is less effective at controlling Y136F isolates than WT isolates. These results have shown that the HYES is suitable for testing additional Cyp51 mutations that may be identified in the future.
DMI fungicides differed in their efficacy regardless of the genetic profile (mutant or WT) of E. necator. Mefentrifluconazole and difenoconazole were more effective than myclobutanil and triadimenol in controlling powdery mildew in greenhouse conditions. Triadimenol was less effective when the Y136F mutant was present.
The relationship between copy number of the Cyp51 gene and phenotypic resistance is not clear, however there was a trend indicating that high copy number may be associated with resistance to DMI fungicides. Further research is required to investigate this relationship, which may lead to the development of a marker for resistance to DMI fungicides.
Mixing sulfur with the DMI fungicides penconazole, myclobutanil and difenoconazole had no effect on the efficacy of these fungicides for controlling powdery mildew in grapevine under greenhouse conditions. This supports label recommendations for use of sulfur, and dispels any confusion raised by previous research of mixing sulfur with DMI fungicides.
Rotorod spore traps, mini-vacuum cyclone separators, washing infected leaves and cotton bud swabs were equally successful tools to detect genetic resistance of E. necator. However, the rotorod spore trap did not require the movement of plant material, and proved to be more sensitive and practical, especially in the early stages of infection when there were no visible symptoms to be collected using the other methods.
A rapid in-field qPCR detection pipeline was successful in estimating B. cinerea group 9 and 17 resistance frequencies in commercial vineyards. This pipeline can now be used as a rapid and cost effective ‘first look’ step to estimate frequencies at vineyards with known control issues. The pipeline can also be used for the monitoring of frequencies where there are no known control issues, or to estimate frequencies within a field trial setting. Additional assays for alternate group 9 and group 17 genotypes could now be developed. The in-field pipeline was also successfully developed to target the group 11 associated mutation G143A in E. necator.
Reduced sensitivity and resistance have been detected in Australian populations of E. necator, B. cinerea and P. viticola for some of the major groups of fungicides used to control grapevine diseases. Therefore, resistance management strategies are required to minimise and delay the onset of resistance. These strategies include regular monitoring (using bioassays and/or genetic analysis), rotating fungicide chemistries, applying chemicals only when necessary, applying multi-site fungicides and other alternatives such as inorganic fungicides (sulfur and copper) and biologicals (Aureobasidium pullulans). Future research will develop in-field and high-throughput techniques to achieve shorter turnaround and a high level of screening for known resistance associated mutations.