In order to develop an effective HWT protocol, experiments were conducted both in vitro and in artificially inoculated autoclaved wood under sterile environmental conditions to determine the heat tolerance of the target fungal isolates (Figure 1). Thereafter, a nursery experiment was carried out under practical conditions to test the long-term effectiveness of a treatment alone and in combination with BCA application on artificially inoculated scion cuttings. Finally, the HWT protocol was implemented on naturally infected scion cuttings, where the pathogen presence was assessed via real-time polymerase chain reaction (PCR).

Figure 1. Diagram showing the workflow of the protocol development for hot water treatment (HWT) applied to eliminate GTD pathogens during the grapevine propagation process (CFU = colony forming unit, MEA = malt extract agar, Ta SC1 = Trichoderma atroviride SC1).

The effect of HWT was tested on P. chlamydospora, P. minimum and D. seriata. Three isolates per species were used in this study, except for D. seriata for which one isolate was tested. Commercially available isolates, obtained from the DSMZ-German Collection of Microorganisms and Cell Cultures GmbH, and isolates from the Institute of Plant Protection, Dienstleistungszentrum Ländlicher Raum (DLR) Rheinpfalz were included (Table 1). All the fungal isolates were cultivated on 2 % malt extract agar (MEA; 2 % agar, 2 % malt extract) plates, except for D. seriata, which was grown on 4 % oatmeal agar (2 % agar, 4 % oatmeal). Fungal colonies were sub-cultured monthly on agar plates and incubated in the dark at 21 °C in a cultivation chamber.

Liquid spore suspensions of P. chlamydospora and P. minimum were prepared by flooding 2-3-week-old colonies grown on agar plates with sterile distilled water. In the case of D. seriata, a spore suspension was prepared according to Úrbez-Torres et al. (2010). The suspensions were filtered through a 30 µm pore diameter sterile gauze and the densities were adjusted to 10⁴ spores/mL using a hemocytometer (Fuchs Rosenthal Counting Chamber, NanoEnTek Inc., Waltham, USA).

The sensitivity to HWT of the ungerminated spores and mycelium development of P. chlamydospora and P. minimum was determined in a laboratory experiment (Figure 1). For each fungal pathogen, three isolates listed in Table 1 were tested individually. Therefore, a 50 mL spore suspension of the isolates was prepared and suspended in distilled water (spore suspension) or in 2 % liquid malt extract (ME; mycelium suspension), obtaining a final volume of 100 mL in a 200 mL Erlenmeyer flask. The spore suspension was directly treated with HW in a laboratory hot water bath for 30 min at 50 °C and 55 °C, and for 45 min at 40 °C, 45 °C, 50 °C, and 55 °C. The mycelium suspension was prepared by incubating the spore suspension on a shaker at 100 rpm for three days. The control suspension was untreated but remained on the laboratory bench at room temperature (~21 °C) for the duration of the treatment. Following treatment, 10 µL of spore or mycelial suspensions were plated onto 2 % MEA, and the plates were incubated for seven days in the dark at ~21 °C in a cultivation chamber. The number of colony-forming unit (CFU) of the isolates was counted and the treatment efficacy was calculated relative to the control. The experiment was repeated three times with ten replicates/treatments with time-temperature combinations and isolates.

The sensitivity of P. chlamydospora and P. minimum to HWT at several temperature-time combinations was tested on artificially inoculated sterile wood (Figure 1). Therefore, one-year-old ‘Müller-Thurgau’ dormant cane segments were autoclaved at 121 °C for 30 min. The most virulent isolates (Pch117179, Pmi63194) of each pathogen were selected for testing based on internal laboratory results. The sterile cuttings were vacuum inoculated for 15 min under 2 × 10^-3 mbar pressure in a table vacuum apparatus with 700 mL spore suspension/isolate of each species. Following inoculation, the surface of the canes was rinsed with distilled water and the cuttings were treated in a laboratory hot water bath for 30 min at 50 °C and 55 °C, and for 45 min at 40 °C, 45 °C, 50 °C and 55 °C. Treatment was carried out on the day of inoculation or four days after incubation in the dark at ~21 °C to enable colonisation and mycelium development in the cane segments. Control canes were inoculated but not exposed to HWT. Following the incubation period, the cane segments were HWT at the same time-temperature combinations as previously described. The treated cuttings were surface sterilised by flaming, subdivided longitudinally, and plated on 2 % MEA plates. The presence or absence of the fungal pathogens was assessed after seven days of incubation in the dark at ~ 21°C in a cultivation chamber, and the growth rate was calculated relative to the untreated control. The experiment was conducted three times with five replicates per treatment for each tested isolate.

Nursery experiments were carried out to assess the long-term effect of HWT and its influence on the efficacy of Trichoderma (Figure 1). For this purpose, a 600 mL spore suspension of each of the following species was prepared: two combined isolates of P. chlamydospora (Pch117179 and Pch101571) and P. minimum (Pmi117607 and Pmi63194), and one isolate of D. seriata (Dse41). Similarly, a suspension of Vintec^® containing T. atroviride SC1 (Ta SC1) was obtained by suspending the formulated granulate in 600 mL tap water (2 g/L; 10¹⁰ conidia per gram formulated product) according to the manufacturer’s recommendations. One-year-old ‘Dornfelder’ shoots were collected from the experimental vineyards of DLR Rheinpfalz. One-internode cuttings underwent vacuum inoculation as described above using spore suspensions of the individual GTD pathogens, and in combination with Ta SC1 at a ratio of 1:1. The inoculated cuttings were stored in a cooler at ~2 °C for seven days. HWT and the handling of propagation material was carried out under practical conditions seven days post-inoculation (dpi) at 50 °C for 45 min in a 6,000 L professional hot water tank in March of each experimental year. The material was stored at ~2 °C for four weeks until the grafting process took place. The cuttings were grafted with an Omega-cut on healthy ‘SO4’ rootstock and sealed with paraffin. The grafted plant material was incubated in callusing boxes for two weeks, then planted in the nursery in a randomised block design in May in both experimental years.

Plant foliage protection against powdery and downy mildew infestation and trimming was carried out based on common practice. The planting material was uprooted in November in each experimental year and stored in cold storage wrapped in plastic bags until the last sampling time point. Wood samples were taken directly after HWT (7 dpi), six months post inoculation (mpi) from plants in the nursery, and 12 mpi during the cold storage phase. The bark of the scion cuttings was removed, and samples were taken from three different parts of each cutting. A 0.5 cm wood chip was removed every centimetre starting 1 cm below the top of the scion. The wood chips were surface sterilised by flaming, then cut into four equal-sized fragments. Twelve wood fragments originating from the same cane were plated on the same agar plate containing 2 % MEA supplemented with 0.1 % tetracycline hydrochloride (MEA+TE). The presence of the pathogens was visually assessed after three weeks of incubation in the dark at ~21 °C in a cultivation chamber. The incidence of pathogen infections was calculated from the mean of three sections for each cane. The experiment was carried out in each of two independent experimental years, with a total of 60 replicates per treatment per year, and 20 replicates per sampling time point.

DNA was extracted from ~100 mg wood material using the method described by Pouzoulet et al. (2013). DNA quality was verified via gel electrophoresis and final DNA concentration was measured via NanoDrop 2000 (Thermo Fisher Scientific GmbH, Osterode, Germany). Real-time PCR was performed to detect the fungal pathogens and Ta SC1 in triplicate using reaction mixtures at a total volume of 25 µL and containing 12.5 µL 2 × iTag Universal SYBR Green Supermix Buffer (Bio-Rad Laboratories, Inc., Hercules, USA), and forward and reverse primers at a concentration of 0.3 µM, 10 µL nuclease-free H₂O with 1 µL DNA template at a dilution of 1:10 in a Rotor-Gene Q cycler (Qiagen, Hilden, Germany). PCR products were purified using GenepHlow™ Gel/PCR Kit (Geneaid Biotech Ltd., New Taipei City, Taiwan) according to the user manual. Sequencing of PCR products was conducted by Eurofins Genomics GmbH (Ebersberg, Germany). Sequences were edited using the BioEdit Sequence Alignment Editor (version 7.0.5.3) and alignments were processed via Clone Manager 9.2 software.

Identification of Ta SC1 was carried out at the end of each experimental year. Three samples were randomly selected from each Ta SC1 treatment group for molecular analysis. A 0.5 cm segment was removed from the scion 2 cm below the top cut and chopped into thin slices. Amplification of the endochitinase 42 gene was carried out by the forward (Ech42 Fw; 5´- GTTCTGAGGCTGGAAGTTGC -3´) and reverse primer (Ech42 Rv; 5´- ACGCCGTCTACTTCACCAAC -3´) pair designed by Savazzini et al. (2008) and applying the cycling programme described by the authors. Sequence analysis of the amplicons was carried out to identify the presence of the probe (Ech42 P; TACCCCTTCAATCACCAATTGTTAG) specific to the target isolate.

Detection of P. minimum and P. chlamydospora was carried out by the forward and reverse primers according to Pouzoulet et al. (2013), with reaction mixtures containing the same ingredients as listed above. The primer set was designed to target the ß-tubulin gene of D. seriata. Sequence alignment was carried out with Diplodia spp. isolates KARE1632 (NCBI accession no. MN318125.1), BoF99.7 (NCBI accession no. KY701766.1), AKBA8 (NCBI accession no. KX259170.1), and MRHf12 (NCBI accession no. MK388682.1) obtained from the NCBI database. Primer specificity tests were carried out on species that commonly occur in grapevine scion. Therefore, fungal colonies from the strain collection of the Institute of Plant Protection at DLR Rheinpfalz were grown on 2 % MEA; namely, Trichoderma koningiopsis, Trichoderma gamsii, Phomopsis ampelina, Penicillium expansum, Fomitiporia mediterranea, Trichothecium roseum, Botrytis cinerea, Cladosporium herbarum, and Cylindrocarpon destructans (Table S1). DNA extraction from fresh fungal colonies was carried out by scraping off fungal mycelium from two-week-old colonies grown on agar plates using a sterile scalpel. The samples were freeze-dried in liquid nitrogen and homogenised as described previously. DNA extraction was performed by DNeasy plant mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Real-time PCR reactions were prepared as mentioned above using the target-specific primers (Table 2).

The PCR programme was run as follows: initial denaturation at 95 °C for 5 min, 35 cycles of 10 s at 95 °C for denaturation, and the extension step at 61 °C for 30 s, followed by the melt curve analysis from 65 °C to 95 °C with 0.5 °C increments at 5 seconds/step. A standard curve was prepared by plotting the average threshold cycle value (Ct values) against the log concentration of the target DNA to estimate the primer efficiency calculated by the software. A dilution series of 5 ng, 500 pg, 50 pg, 5 pg, and 500 fg total fungal DNA extracted from fresh mycelium was thus amplified. Melting curve analysis was carried out by the Rotor-Gene Q Series Software (version 2.0.2.4) with threshold settings at 0.16 dF/dT. Real-time PCR product clean-up and sequencing were carried out as described earlier. NCBI Nucleotide BLAST was used to identify the species of the obtained sequences.

In order to select naturally-infected scion material for the laboratory experiment, an experimental field in DLR Rheinpfalz containing 15 years old ‘Riesling’ plants was tested for the presence of P. minimum, P. chlamydospora, and Diplodia spp. via molecular analysis and re-isolation. Twenty plants were selected for the experiment, of which 10 plants had shown symptoms during the previous vegetative phase. Wood powder samples were taken from the trunk heads by drilling a hole with a hand drill equipped with a 5 mm drill bit in December 2023. DNA extraction was conducted as described above. Additionally, 10 mg of the powder was plated on 2 % MEA + TE for re-isolation in parallel to the molecular analysis. Powder samples plated on agar were incubated for 14 days in the dark at ~21 °C, and the presence of the target pathogen was visually assessed based on morphological traits using a stereomicroscope.

The effect of HWT on Diplodia spp. on naturally infected scion cuttings was assessed (Figure 1). Therefore, four dormant one-year-old shoots were collected from each of the previously tested plants and analysed for pathogen presence. Fifteen positive and 15 negative shoots were used in the experiment, and samples were taken from the first four internodes at the basal and top sections. Approximately 100 mg of drill powder samples were taken from each section. In each trial, five sections per treatment were randomly selected as replications. The experimental setup included both healthy and diseased groups of the non-HWT control and the HWT material. Assessment was carried out via real-time PCR and via re-isolation on 2 % MEA + TE prior to treatment. HWT was carried out in a laboratory water bath at 50 °C for 45 min. Drill samples were taken 24 hours post-HWT directly next to the previous sampling location, and the presence of the pathogen was assessed via real-time PCR and re-isolation.

A statistical analysis was performed using RStudio software (version 4.3.2, R. Posit Software, PBC, Boston, MA). CFU count data of the in vitro HWT assay was analysed by fitting the data to a negative binomial generalised linear model (GLM). Post hoc pairwise comparisons were conducted to assess the differences between groups using the Šídák method (P ≤ 0.05). Binomial GLM was used to analyse the results of the wood sensitivity test and incidence in the nursery study and field study. To examine the differences in mean values between groups, the ‘emmeans’ package version 1.10.0 was applied using the Šídák or Tukey methods (P ≤ 0.05). To estimate the marginal means of non-zero groups, a linear model was fitted to the data using the ‘lm’ function in RStudio, and the resulting model was then analysed using the ´emmeans´ function. Contrasts were calculated to assess whether the estimated means significantly differed from zero. Model validation was conducted using the DHARMa package (version 0.4.6).

The results of HWT on conidial germination of three isolates of P. minimum and P. chlamydospora showed that, in vitro, the tested pathogens survived higher treatment temperatures in their mycelium stage compared to the ungerminated conidia stage (Figure 2). The treatment duration of 30 min did not influence the treatment efficacy at temperatures of 50 °C and 55 °C; all treatments effectively inhibited spore germination of both tested fungal pathogens (Figures 2A and 2B). P. minimum revealed a higher tolerance to HWT compared to P. chlamydospora when subjected to lower temperatures (40 °C and 45 °C). Treatments on P. minimum at the lowest temperature (45 min, 40 °C) did not significantly decrease the number of CFUs of the tested isolates compared to the non-treated control (Figure 2A). For isolates Pmi117607 and Pmi63194, this treatment reached an efficacy of only 5.6 % and 9.0 %, respectively. While this temperature/duration combination had no effect on isolate Pmi9896 (Table 3), the treatment at 45 °C for 45 min reduced the number of CFUs of Pmi117607 significantly (by 60.0 %) when compared to the non-treated control. This treatment showed an efficacy of 32.6 % against Pmi63194 and 53.8 % against Pmi9896. Spore germination of Pmi63194 and Pmi9896 was inhibited at 50 °C. Pmi117607 tolerated 50 °C, but conidial survival at this temperature was negligible (Figure 2A). In the case of P. chlamydospora, treatment at 40 °C of ungerminated spores decreased the conidial survival of Pch101571 by 93.8 %, by 89.5 % of the wild isolate Pch9886 (Figure 2B) and by 62.6 % of Pch117179 compared relative to the non-treated control. The temperatures of 45 °C completely inhibited the spore germination of all the tested P. chlamydospora isolates (Table 3).

Regarding the mycelium developmental stage of P. minimum, Pmi117607 was observed to survive treatment for 30 min at 50 °C, with an efficacy of 98.2 % against the isolate (Figure 2C). The results from the longer treatment durations showed that P. minimum tolerated treatments at 40 °C with a treatment efficacy ranging between 4.9 % and 54.5 % (Table 3). Pmi117607 showed the highest tolerance to HWT at the mycelium stage, since it survived treatments of 50 °C regardless of the duration of the treatment (Figure 2C). A high HWT treatment efficacy was observed against P. chlamydospora mycelium, the lowest being 93.4 % at the lowest tested temperature against Pch101571. With further temperature increase, the number of CFUs of the tested isolates was found to decrease (Figure 2D). Efficacy at 50 °C reached more than 99 % in case of Pch101571 and Pch9886. Temperatures at 55 °C inhibited the growth of all the isolates.

Figure 2. Effects of hot water treatments (HWT) comprising different combinations of temperature and exposure time on the survival of conidia (A, B) and mycelium (C, D) of three isolates of P. minimum and P. chlamydospora (CFU = colony forming unit).

Groups sharing the same letter are considered statistically indistinguishable. Significant differences between treatments were analysed by fitting a negative binomial GLM followed by a Sidak’s post hoc test. Different letters indicate significant differences at P ≤ 0.05 (n = 30).

As the previous tests had indicated high HWT efficacy at 50 °C for 45 min, this temperature-treatment combination was selected for further tests in the nursery experiments. None of the tested pathogens had been isolated prior to planting in the experimental year of 2021, whereas contamination of the non-inoculated samples had reached 45 % incidence prior to planting in 2022 (Table 4). Based on morphological characterisation, Diplodia spp. was identified in the contaminated control samples. Over the growing season, the presence of Diplodia spp. in non-inoculated control samples reached 10 % in 2021 and 40 % in 2022. Still, the incidence of the artificially inoculated samples was significantly higher than in non-inoculated control samples. All tested GTD pathogens successfully colonised the wooden cuttings and these pathogens were detectable one-year post-inoculation after all the usual nursery procedures associated with grapevine propagation had been carried out. P. chlamydospora was re-isolated from non-treated cane samples at 7 dpi in more than 70 % of the samples in both experimental years. At the first sampling time point, P. minimum incidence reached 90 % in 2021 and 40 % in 2022. The highest D. seriata incidence on the non-treated control canes was observed at 7 dpi, with 100 % in 2021 and 90 % in 2022. P. chlamydospora incidence throughout the experimental years reached more than 65 %, of P. minimum it exceeded 60 % and D. seriata remained above 35 % at all sampling time points. The application of Ta SC1 resulted in long-lasting inhibition of pathogen growth – as determined by the visual assessment – after cultivation on media. However, in the case of a cane sample inoculated with P. minimum and treated with Ta SC1, an incidence of 5 % was recorded. While no Ta SC1 growth could be observed on the agar plate for this sample, the inoculated pathogen was re-isolated. Following HWT, the incidence of the individual pathogens decreased significantly at the first sampling time point. No pathogen recovery was detectable at 7 dpi in the case of P. chlamydospora and D. seriata, and P minimum incidence reached only 5 %. In general, the treatment was not capable of fully eliminating the targeted pathogens. Throughout the two experimental years, P. chlamydospora was recovered at the last sampling time points. By contrast, D. seriata was entirely eliminated in 2021, and in the following experimental year, it was detected on 10 % of the samples at 6 mpi.

It is worth noting that the incidence of D. seriata in the inoculated HWT samples might have originated from natural infection, as the incidence of non-inoculated samples also increased at this time point (6 mpi in 2022). At the second sampling time point, P. minimum incidence reached 40 %, but no pathogen was detected at 12 mpi in 2021. In 2022, P. minimum incidence after HWT increased from 20 % to 30 % at the last sampling time points of the observation period.

A complete, long-term eradication of all tested pathogens was achieved through the combined treatment of HWT and Trichoderma. Trichoderma growth was not influenced by HWT, and, at the end of the two experimental years, the recovery of the Trichoderma-based BCA product from the wood sections exceeded 95 % (Table S2). The presence of the SC1 strain was confirmed through sequencing, indicating 90 % recovery of Ta SC1 in the treated samples (Table S3).

In order to obtain naturally infected scion material, a 15-year-old Riesling vineyard was tested for the presence of P. minimum, P. chlamydospora and Diplodia spp. According to the results of the real-time PCR assessment, P. minimum was present on the trunks of 14 % of the plants, but pathogen re-isolation was only successful in one sample. In general, P. chlamydospora was present in 73 % of the samples according to the molecular analysis (Table S4), while Diplodia spp. were present in 32 % of the tested plant trunks, which was confirmed by re-isolation on agar plates. Seventeen out of 88 shoots were tested positive for possessing Diplodia spp., which partially originated from mother plants that had not shown any symptoms in the previous vegetative season. The pathogen was detected in the basal first three internodes of the shoots (Table S5). Due to the high incidence of Diplodia spp. in the shoot material, this species was selected as the model pathogen for testing the molecular analysis method for the evaluation of HWT efficacy.

According to the visual assessment of the wood fragments, the percentage of plants infected with Diplodia spp. significantly decreased following HWT at 50 °C for 45 min (Figure 4): pathogen incidence was 20 % in treated plants compared to 93 % in untreated infected plants. The sequence analysis confirmed that it was D. seriata which had been isolated and cultivated from these infected samples. No pathogen presence was detected in the non-treated negative control group, whereas an incidence of 7 % was detected in the HWT group via visual assessment of the wood fragments.

According to the results of the molecular analysis, pathogen incidence in the untreated infected material reached 80 %, with two samples showing an amplification signal, but the melt curve remained below the threshold limit. Diplodia spp. was detected in 53 % of the infected HWT samples one-day post-treatment, with no significant impact of the treatment compared to the non-treated infected group. The incidence of the non-infected control samples treated with HWT was 7 % according to the real-time PCR detection, confirming the observations made on the agar plates. This incidence was reached in one sample colonised by Diplodia mutila, which was verified by sequence analysis.

Figure 4. Incidence (%) of Diplodia spp. in one-year-old cane samples non-treated or exposed to hot water treatment (HWT) grouped as infected (positive) and non-infected (negative) according to visual and molecular (real-time PCR) analysis.

Significant differences are indicated with different letters above the bars based on binomial GLM followed by Tukey’s post hoc test (P ≤ 0.05, n = 15).