Downy mildew (DM, caused by Plasmopara viticola (Berk. & M.A. Curtis) Berl. & De Toni), powdery mildew (PM, Erysiphe necator (Schwein.) Burrill), grey mold (GM, Botrytis cinerea Pers.), and black-rot (BR, Phyllosticta ampelicida (Engleman) Van der Aa) are the main foliar and bunches diseases affecting grape production worldwide. Under favourable weather conditions, they cause substantial yield and quality losses (Scribner and Viala, 1888; Ferrin and Ramsdell, 1978; Pallotta et al., 1998; Calonnec et al., 2004; Hoffman et al., 2004; Jermini et al., 2010).

Disease control in vineyards is generally achieved by using fungicides, copper, and sulphur, all of which have negative effects on human health, the vineyard environment, and biodiversity (Rusch et al., 2015; La Torre et al., 2018; Lamichhane et al., 2018; Reiff et al., 2021). As a consequence, safer disease control methods are needed. The use of mathematical models and Decision Support Systems (DSSs) to improve the scheduling of plant protection product application, as well as the use of biocontrol agents, botanicals, and natural resistance inducers (Cohen et al., 1999; Hamiduzzaman et al., 2005; Caffi et al., 2007; Rossi et al., 2014; Pertot et al., 2017; Rossi et al., 2019) are sustainable alternatives that can be combined for effective Integrated Pest Management (IPM) in vineyards. Resistant grapevine varieties are also a promising component of IPM (Lu, 1997). Since the early 2000s, resistance to grape diseases has been a major concern of breeding programs (Delame et al., 2019; Miclot et al., 2019; Zini et al., 2019), and a number of resistant varieties have been developed through hybridization of Vitis vinifera with wild American and Asian Vitis species and have also been identified among varieties of V. vinifera from Central Asia (Gessler et al., 2011; Foria et al., 2019). At present, more than 30 quantitative trait loci (QTL) associated with resistance to downy and powdery mildews, as well as to other harmful organisms (Phyllosticta ampelicida, Diaporthe ampelina, Agrobacterium spp., Pierce’s disease, Daktulosphaira vitifoliae, and Xiphinema index) have been identified (www.vivc.de accessed on 19 May 2022) and used in breeding programs (Di Gaspero and Cattonaro, 2010), with emphasis on DM and PM (Zini et al., 2015; Guimier et al., 2019; Schneider et al., 2019). Partial resistance to DM and PM modifies some resistance components such as infection frequency, latency period, lesion size, spore production, infectious period, and infectivity (Bove and Rossi, 2020). Therefore, partial resistance does not stop infections but reduces disease progress in the vineyard (Bove et al., 2021).

The development of resistant varieties requires screening in both the laboratory and the field. Although leaf disc bioassays have been developed for screening resistant genotypes under laboratory conditions (Vezzulli et al., 2018; Eisenmann et al., 2019; Bove and Rossi, 2020; Possamai et al., 2020), less research has been focused on the assessment of resistance on leaves and especially on bunches under field conditions. As a consequence, the information on the resistance level of bunches is quite limited.

Varieties expressing partial resistance to DM and PM require some disease control via fungicide application (Zambon et al., 2019) in order to avoid loss and to maintain the pathogen at low population densities so as to reduce the development of resistance-breaking genotypes (Schwander et al., 2012; Foria et al., 2019). In addition, some varieties with resistance to DM and PM are highly susceptible to other diseases like GM and BR (Foria et al., 2019), which may need control interventions. However, when a vineyard manager decides that DM- and PM-resistant varieties require fungicide applications, the manager has no or few tools for determining the intensity and the scheduling of these applications.

In this work, we characterised the “resistance patterns” of 16 grape varieties in an experimental vineyard with respect to the four main pathogens of grape: P. viticola, E. necator, B. cinerea, and P. ampelicida. Some of these varieties carry one or more Rpv (acronym for resistance to Plasmopara viticola) loci and/or Ren (acronym for resistance to Erysiphe necator) loci that confer partial resistance to DM and PM. We define the resistance pattern as the resistance level of a variety against all of the diseases (i.e., DM, PM, GM, and BR in this study) in a vineyard. The resistance pattern of a variety may affect disease management because it indicates which pathogens may develop and require light or intensive control.

The present research characterised the resistance pattern of 16 grape varieties to four main grape pathogens, i.e., P. viticola, E. necator, B. cinerea, and P. ampelicida. Most of 16 varieties carry one or more Rpv loci conferring partial resistance to DM and/or Ren loci conferring partial resistance to PM. Resistance was assessed in the field and relative to a susceptible control, the variety ‘Merlot’, over four grape-growing seasons (2017, 2018, 2019, and 2021). The conduciveness of environmental conditions for the diseases differed among the four growing seasons such that epidemics of DM, PM, and BR developed on both leaves and bunches with varying severities, while GM affected only bunches and only at very low levels. Therefore, incomplete information was collected with respect to GM. Low levels of GM detected in this study are likely due to unfavourable weather conditions. Further studies are needed to investigate the resistance to GM of the varieties showing resistance to DM and/or PM.

In previous research, resistance to mildews was largely evaluated by performing monocycle, in vitro bioassays involving the artificial inoculation of leaf discs (Stein et al., 1985; Eibach et al., 1989; Staudt and Kassemeyer, 1995; Brown et al., 1999; Boso et al., 2006; Boso and Kassemeyer, 2008; Bellin et al., 2009); leaf disc inoculation has also been used to study the components of partial resistance to P. viticola in some resistant varieties (Bove and Rossi, 2020). Before the current study, little information was available about the behaviour of these resistant varieties under field conditions (Fischer et al., 2004; Wingerter et al., 2021), in which infection cycles follow each other during the season so that the resistance components related to sporulation (i.e., length of latency and infectious periods, sporulation rate, and infectiousness of zoosporangia; Bove and Rossi, 2020) may have effects that are undetected in monocycle leaf disc bioassays. For this reason, the whole epidemic course of each disease was considered and summarized by the AUDPC values in the current study. To our knowledge, this is the first research report of the field evaluation of resistance to DM, PM, and BR in resistant grapevine varieties expressed both on leaves and bunches under natural conditions of inoculum and without application of fungicides.

In this research, we introduce the term “resistance pattern” to describe the resistance level of a variety to all of the diseases occurring in a field experiment. This concept is similar to the term “resistance pattern” that is commonly used in bacteriology with respect to antibiotic resistance; based on the CDCs (Centers for Disease Control and Prevention) glossary of terms related to antibiotic resistance, the term describes the antibiotic resistance of a single isolate to multiple antibiotics.

Overall, our research divided the 16 grape varieties into four main groups (CLU1 to CLU4) based on the resistance pattern to the three diseases. The first group (CLU1) included only the control, ‘Merlot’, which had low resistance to mildews and intermediate resistance to BR. The second group (CLU2), made up of ‘Merlot Khorus’ and ‘Merlot Kanthus’, had high resistance to DM conferred by Rpv12 and Rpv3, respectively. ‘Merlot Khorus’ and ‘Merlot Kanthus’ share the same pedigree (Merlot x 20/3) and have been genetically studied only for Rpv loci, while the presence of Ren and Rgb loci is still uncertain. Given the results of the current study, it is unlikely that these varieties carry Rgbs; both varieties were severely affected by BR over the 4 years of the study, with AUDPC values on both leaves and bunches greater than those of the sensitive control, ‘Merlot’. CLU2 had intermediate resistance against PM, with E. necator colonies observed mainly on leaves rather than on bunches. The third cluster (CLU3) included only ‘Rkatsitelii’, a native Georgian variety that showed intermediate resistance to DM despite the absence of Rpv loci in its germplasm. The intermediate resistance of ‘Rkatsitelii’ to DM reflected the low level of resistance of its leaves and the high level of resistance of its bunches during the four growing seasons ( Figure 3A ). ‘Rkatsitelii’ also had intermediate resistance to BR, with its leaves less resistant to the disease than its bunches ( Figure 3C ). ‘Rkatsitelii’ had high levels of PM on both its leaves and bunches, with AUDPC values on its leaves much greater than those on the leaves of the control, ‘Merlot’. The fourth cluster (CLU4), which included the majority of the investigated varieties, showed high resistance to the three diseases. On these varieties, DM, PM, and BR symptoms were rarely observed on leaves, and bunches remained generally healthy over the four growing seasons.

Among the four cluster, only CLU1 had low resistance to DM, meaning that all of the other varieties were resistant to DM to some extent under field conditions; these other DM-resistant varieties included ‘Rkatsitelii’ (CLU3), ‘Reberger (CLU4), and ‘Palava’ (CLU4), which lack Rpv loci in their germplasm. The expression of resistance to DM was generally highest in the varieties carrying Rpv3, Rpv10, or Rpv12; these observations were consistent with previous studies (Basler and Pfenninger, 2003; Pezet et al., 2004; Vezzulli et al., 2018; Bove and Rossi, 2020). In a recent study, the 16 varieties considered in the current research were characterized for six components of resistance (infection frequency, latent period, lesion size, production of sporangia, infectious period, and infectivity of sporangia) to DM by leaf disc inoculation under laboratory conditions (Bove and Rossi, 2020). The findings of the latter study were confirmed by our observations. For example, ‘Merlot Kanthus’, ‘Johanniter’, ‘Bronner’, ‘Solaris’, ‘Calardis blanc’, ‘Merlot Khorus’, ‘Cabernet volos’, ‘Villaris’, and ‘Fleurtai’ showed high resistance to DM in both the laboratory assays (Bove and Rossi, 2020) and in the field experiment of the current study. In contrast, ‘Felicia’ and ‘Regent’ had low levels of resistance to DM in the laboratory (Bove and Rossi, 2020) but high levels of resistance in the field (current study). Although their AUDPC values were much lower, especially on bunches, than those of ‘Merlot’, the varieties ‘Calandro’, ‘Palava’, ‘Rkatsitelii’, and ‘Reberger’ were less resistant to DM in the current study than the previously mentioned resistant varieties, which was consistent with other studies (Boso and Kassemeyer, 2008; Bitsadze et al., 2015; Bove and Rossi, 2020).

Discrepancies between laboratory studies (conducted with high disease pressure and optimal environmental conditions for P. viticola) and field observations have been already reported (Calonnec et al., 2013; Vezzulli et al., 2018), and may be caused by differences in inoculum dose and environmental conditions, as well as by the co-occurrence of multiple diseases and the consequent competition among pathogens for the infection sites.

In studies that evaluate the resistance against a specific disease, all other diseases are usually controlled by the application of fungicides that do not affect the disease being studied. In the current research, however, no fungicides were applied during the growing season, so that the different pathogens were competing for the same infection sites. If we consider the crop as a finite collection of infection sites with two mutually exclusive states, healthy or infected (Savary et al., 2018), and if we assume that a site can be occupied by only one pathogen, then the occurrence of one disease limits the occurrence of the others. This could explain the intermediate level of resistance to DM of ‘Rkatsitelii’, which differs from the previous findings of Bove and Rossi (2020) and Bitsadze et al. (2015), who rated this variety as having low resistance to DM. In the current study, PM severely affected ‘Rkatsitelii’ and may have reduced the available sites for infection by P. viticola, resulting in an overestimation of the resistance to DM for this variety. This “competition-for-infection sites” explanation may also be applied to ‘Merlot Khorus’ and ‘Merlot Kanthus’, which were severely affected by BR over the four growing seasons. This may have reduced infections by PM and DM, leading to an overestimation of the effective level of resistance against these two diseases. Nevertheless, these two varieties showed high resistance to DM in leaf disc inoculation tests (Bove and Rossi, 2020). Pathogen-specific in vitro studies with ‘Merlot Khorus’ and ‘Merlot Kanthus’ are needed to better characterize their level of resistance to PM; because the presence of Ren loci in these two varieties is still uncertain, genetic profiling is also needed.

The results of the current study increase our understanding on the resistance of grape bunches to diseases. When all the varieties were considered, the positive correlation between the resistance of leaves vs. bunches for all three diseases seems to support the results of Calonnec et al. (2013) on the reliability of bioassays on leaves for predicting disease severity on bunches. However, when the correlations were assessed for groups of varieties showing a similar degree of resistance, the correlation between the AUDPC on leaves and bunches was significant only when the resistance level was low, with the exception of PM. We therefore conclude that the bioassays on leaves can indicate the level of resistance of bunches in the field for susceptible varieties but not for resistant ones.

More than 70% of the fungicide used in Europe is applied to vineyards (Muthmann and Nadin, 2007). According to the Directive 2009/128/EC, the use of plant genotypes carrying resistance genes to major grape diseases may help to reduce the application of plant protection products and thus the negative effects of those products on the environment and human health and may also reduce vineyard production costs. Although some studies have shown that resistant grapevine varieties can reduce the application of fungicides by 100% (Wingerter et al., 2021), other studies indicate that resistant varieties still require the application of biocides to protect against certain oomycete and fungal pathogens (Töpfer et al., 2011; Delmas et al., 2016). By defining the resistance patterns of 15 resistant varieties to three grapevine diseases (DM, PM, and BR), our study provides a more nuanced and more useful perspective than that provided by single-pathogen laboratory studies on the resistance of these varieties under field conditions. Information on the resistance patterns could support both the strategic and tactical decisions of grape disease management (Rossi et al., 2012; Pertot et al., 2017; Rossi et al., 2019).

At the strategic level, an understanding of resistance patterns may help growers select the variety to be planted in a specific environment. For example, in a location with frequent spring rains and high moisture, i.e., a location where PM rather than DM and BR prevails, a variety showing high or intermediate resistance to E. necator would be preferred such as varieties grouped in CLU2 and CLU4 in the current research. On the other hand, in locations where DM and BR are favoured, varieties of CLU4 should be preferred. The planting of varieties grouped in CLU1 and CLU3 should be carefully considered in locations with a history of severe grapevine diseases, especially in organic farms in which the use of synthetic fungicides is banned.

At the tactical level, if a variety expresses high resistance to DM and PM but high susceptibility to BR, intensive control of P. ampelicida but light control of P. viticola and E. necator would be required. This could be the case of ‘Merlot Khorus’ and ‘Merlot Kanthus’ in CLU2. For the varieties grouped in CLU4, tactical disease management can require few fungicide applications, which can be scheduled using an infection risk-based approach. For this purpose, information on resistance patterns could be used to modify mathematical models so that they provide variety-specific predictions of disease risk. These models usually predict the risk of infection based on the weather conditions and, in some cases, the growth stage of the host (Rossi et al., 2008; Caffi et al., 2011; González-Domínguez et al., 2015; Rossi et al., 2015; Brischetto et al., 2021), but they do not presently use information concerning the resistance level of the variety. The usefulness of such models to growers would be increased if they include resistance patterns and thereby provide more precise information on whether and when a fungicide treatment would be needed for any grape variety.

The raw data supporting the conclusions of this article will be made available by the authors upon request, without undue reservation.

VR mainly contributed to the conception and the design of the study. IS and FB carried out the data collection in field. All authors contributed to the analysis of results and collaborated in writing the manuscript. All authors contributed to the article and approved the submitted version.

This research was funded by research fundings of Università Cattolica del Sacro Cuore.

The authors thank Gultekin Hasanaliyeva and Othmane Taibi (Università Cattolica del Sacro Cuore) for their valuable assistance in data collection in field.

Author FB was employed by the company Horta srl.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.