Introduction
Grey mould, caused by Botrytis cinerea Pers., is one of the most common vineyard diseases and can cause severe damage. Bioclimatic conditions in Northwestern Spain, with warm temperatures during the day and high humidity during the night, favour the development of this fungal infection, which reduces crop productivity markedly. To prevent this disease and to diminish its impacts, the most common strategy employed by winegrowers is the systematic application of chemical fungicides, generally following preset calendars based on the grapevine phenological growth stages (Bugiani et al., Reference Bugiani, Govoni, Bottazzi, Giannico, Montini and Pozza1995). However, these fungicides should only be applied when a real risk of unacceptable economic damage in the vineyard is detected, in order to avoid the consequences associated with their excessive use, such as the appearance of resistant fungus or the alteration of beneficial mycological flora.
Knowledge of pathogen biology and disease cycles, including interactions between pathogen, environment and host, is essential to avoid or to reduce the consequences of any plant disease (De Wolf and Isard, Reference De Wolf and Isard2007). Disease appearance is a consequence of the occurrence and interaction of three causal agents, which are (i) a susceptible plant host, (ii) the presence a virulent pathogen, (iii) suitable environmental conditions for the pathogen development. The absence of any of these components prevents plant disease (Stevens, Reference Stevens1960; Agrios, Reference Agrios2005). This natural condition is one of the paradigms of plant pathology known as disease triangle, and is the basis for an Integrated Pest Management strategy guided to chemical products reduction in crop managing (Stevens, Reference Stevens1960; Agrios, Reference Agrios2005).
Optimal meteorological conditions for B. cinerea development are high relative humidity and warm temperatures during the grapevine reproductive cycle (Broome et al., Reference Broome, English, Marois, Latorre and Aviles1995; Rodríguez-Rajo et al., Reference Rodríguez-Rajo, Jato, Fernández-González and Aira2010). Furthermore, microclimatic conditions in the grapevine canopy markedly influence disease development because leaf removal near clusters significantly reduces the incidence and severity of Botrytis bunch rot (English et al., Reference English, Thomas, Marois and Gubler1989). From the phenological point of view, the most widely accepted grapevine phenological stages to be susceptible to B. cinerea colonization and infection are flowering (stage 6 of BBCH-Biologische Bundesanstalt für Land und Forstwirtschaft, Bundessortenamt und CHemische Industrie scale) and ripening of berries (stage 8 of BBCH) (McClellan and Hewitt, Reference McClellan and Hewitt1973; Esterio et al., Reference Esterio, Auger, Droguett and Arroyo1996). In addition, aerobiological studies can be used to quantify possible pathogen presence in the vineyard using biosensors that measure daily and hourly airborne spore concentrations (Fernández-González et al., Reference Fernández-González, Rodríguez-Rajo, Jato, Aira, Ribeiro, Oliveira and Abreu2012). Since many authors have related fungal disease levels at a given time to airborne spore concentrations during previous periods, airborne spore concentrations can be used as bio-indicators of pathogen development (Jeger, Reference Jeger1984; Carisse et al., Reference Carisse, Savary and Willocquet2008).
The main objective of the current study was to reduce Botrytis fungicide treatment in vineyards of the Ribeiro Designation of Origin area (one of the protected winemaking areas of Spain due to the quality and history of its wines) by detecting the main risk periods of Botrytis development and infection. To achieve this, a comprehensive study was conducted of the airborne fungal conidia in the atmosphere of the vineyard, vines phenology and optimal weather conditions for infection. The combination of these factors describes the behaviour of the fungus in relation to the environmental conditions of the vineyard agro-system, and potentially identifies the optimal time for fungicide application.
Materials and methods
The studied vineyard is located in Cenlle, at 42°18′55.88″N–8°6′3.28″W (Datum WGS84), and 199 m above sea level (SIXPAC, 2018). It belongs to the Ribeiro Designation of Origin region (D.O.), in Northwestern Spain (Fig. 1). Steep valleys and hillsides characterize this area. The particular Oceanic-Mediterranean transitional eco-climate of this region is favoured by its southern situation in Galicia, as well as by natural barriers that protect this territory from sub-Atlantic storms. According to the Multicriteria Climatic Classification System (MCC), most winemaking areas in this region, watered by the Miño River, would be defined as temperate and warm, sub-humid, with very cold nights (Blanco-Ward et al., Reference Blanco-Ward, Garcia and Jones2007).
Fig. 1. Location of Ribeiro wine Designation of Origin in Galicia, at Northwestern Spain.
Detection of the pathogen presence in the vineyard
Airborne fungal propagule Botrytis concentrations were determined using a Lanzoni VPPS-2000 (Lanzoni s.r.l.) volumetric pollen-spore trap (Hirst, Reference Hirst1952) located in the central part of the vineyard and situated 2 m above ground level in order to avoid confounding spore trap measurements with plant growth. Aerobiological sampling period took place during the active Vitis vinifera L. season over the studied years, from 2008 to 2015. The vegetative period started on 20 March and ended at harvest date (usually in the second fortnight of September). A Lanzoni trap was calibrated to sample a constant volume of 10 litres of air per minute; air particles passed through a cylindrical drum covered by a Melinex film impregnated with a 2% silicone solution, as a spore-trapping surface. This drum was changed weekly (daily from June to August), and the exposed tape was cut into seven pieces, which were mounted on separate glass slides. Botrytis cinerea spores were counted following the protocol proposed by the Spanish Aerobiological Network (REA) (Galán et al., Reference Galán, Cariñanos, Alcázar and Domínguez2007). For spores identification and count, we analysed two lines along the slides at 400× magnification using a light optical microscope. Results were expressed as spores when referring to total values, or spores/m3 of air when referring to daily mean values (Galán et al., Reference Galán, Ariatti, Bonini, Clot, Crouzy, Dahl, Fernandez-González, Frenguelli, Gehrig, Isard, Levetin, Li, Mandrioli, Rogers, Thibaudon, Sauliene, Skjoth, Smith and Sofiev2017).
Phenological observations of the main susceptible stages to infection
A phenological study was carried out in order to relate plant developmental stages to the detected spore levels. Field observations were conducted during the active grapevine season for each study year. Among the multiple cultivars grown in the study area, we considered the Treixadura cultivar, which is the preferential autochthonous cultivar of the Ribeiro D.O. area. We randomly selected 22 plants that were observed weekly, except during the flowering stage, when the number of observations was increased to twice a week. Phenological stages were monitored using the scale recommended by Lorenz et al. (Reference Lorenz, Eichhorn, Blei-holder, Klose, Meier and Weber1994), adopted by the BBCH as a standardized scale for phenological grapevine observations (Meier, Reference Meier2001). Five main grapevine growth stages were considered: stage 1 – leaf development, stage 5 – inflorescence emergence, stage 6 – flowering, stage 7 – development of fruits and stage 8 – ripening of berries. For grapevine phenological calendar development, we considered the start date of each stage to be the date when 50% of the studied plants had reached that stage.
Suitable meteorological conditions for Botrytis spore germination
Meteorological data were obtained from an HOBO (ONSET HOBO® USB Micro Station Data Logger – H21-USB), located in a row of the vineyard at 1.5 m above ground level next to the spore sampler. The monitored parameters were maximum, mean and minimum temperatures, relative humidity and dew point. Information about rainfall and wind speed was obtained from the MeteoGalicia Meteorological Station located in the Viticulture and Enology Station of Galicia (EVEGA) in Leiro, at 5 km from the study vineyard (MeteoGalicia, 2018).
The Magarey generic model was adapted to forecast the suitable meteorological conditions for fungal plant pathogen infection. This model estimates the wetness duration requirement needed to accomplish a critical disease intensity at a given temperature. Botrytis cinerea infection is defined as having a 20% disease incidence on an infected part of the plant (Magarey et al., Reference Magarey, Sutton and Thayer2005). This model was applied to the study vineyard in order to determine the potential risk of disease development related to weather conditions. It is based on a temperature response function scaled to the minimum and optimum values of the surface wetness duration requirement.
First, possible disease development periods were identified as wet periods, with relative humidity RH ⩾95% within the hourly values of each day. If various wet periods are separated by a dry period of RH <95% in the same day, they can be summed if the dry period (D) is lower than the D 50 value. This parameter is defined as the duration of a dry period that will result in a 50% reduction in disease compared with a continuous wet period (Magarey et al., Reference Magarey, Sutton and Thayer2005). If a dry period (D) is higher than D 50, which is 13 h for B. cinerea (Bregaglio et al., Reference Bregaglio, Donatelli and Confalonieri2013), the wet periods cannot be summed and two different infection risk periods are considered for the same date.
After that, a temperature response function (Equation 1) for each wet period was applied:
where T: average temperature (°C) during the wetness period, T min: minimum temperature for infection, T max: maximum temperature for infection and T opt: optimum temperature for infection.
The obtained results allow calculation of wetness duration requirement value (W (T)), in hours, by the following expression (Equation 2):
where W (T): wetness duration requirement for the critical disease threshold at temperature T, W min: minimum value of wetness duration requirement for the critical disease threshold at any temperature and f (T) is the temperature response function. W max: parameter that indicates an upper boundary on the value of W (T).
To develop the original model, Magarey et al. used experimental data from 53 published studies on the temperature and moisture responses of different plant pathogens (Magarey et al., Reference Magarey, Sutton and Thayer2005). Information on the temperature–wetness combination effect on B. cinerea infection of grape flowers and berries comes from Nair and Allen (Reference Nair and Allen1993). Based on model parameter values identified in this study (Table 1), we calculated the W (T) values for our study area in two different phenological stages: flowering (stage 6) and ripening of berries (stage 8). The calculated values for this fungal disease indicator offer an indirect measurement of infection risk. To express the potential risk identified in each period, we graphically represented (Fig. 2) the risk periods identified with the Magarey model as the difference between the W max value for flowering (12 h) and the W max value for ripening of berries (10 h). The resulting W (T) values show a direct measure of infection risk during these episodes. We expressed this measurement as Magarey units rather than hours of required wetness duration.
Fig. 2. Airborne B. cinerea spore concentrations (light grey area), 100 spores/m3 threshold (represented by discontinuous line), grapevine phenological stages (upper lines of S-1, S-5, S-6, S-7 and S-8), anti-Botrytis treatments (arrows), Magarey suitable meteorological periods (values expressed as Magarey units = W max–W (T) for each phenological stage, in black bars), and the evaluated risk periods. The represented risk periods are the result of the combination of a susceptible phenological stage (flowering-S6 or ripening of berries-S8), one identified Magarey suitable meteorological period and the airborne spore concentrations above 100 spores/m3 (high-risk period (HR) in red triangle), between 10 and 100 spores/m3 (moderate-risk period (MR) in orange diamond) or below 10 spores/m3 (low-risk period (LR) in blue circle). (a) W max (Equation 2) for B. cinerea in flowering stage risk periods: 12 h (dark grey line). (b) W max (Equation 2) for B. cinerea in ripening of berries stage risk periods: 10 h (light grey line).
Table 1. Model parameters for infection model developed by Magarey et al. (Reference Magarey, Sutton and Thayer2005) and D50 value for Botrytis cinerea
a Model parameters obtained from the Nair and Allen (Reference Nair and Allen1993) temperature–wetness combination study.
b Reference for D 50 value in Bregaglio et al. (Reference Bregaglio, Donatelli and Confalonieri2013).
Furthermore, in order to determine the statistical relationship between airborne spore concentrations and the main weather parameters altogether, we applied a Principal Component Analysis (PCA) for the 2008–2015 data set. This statistical procedure reduced the dimensionality of the set of weather predictor variables to determine the highest influence on airborne spore concentrations. The considered variables were the Botrytis airborne spore concentrations (Botrytis), mean temperature (T mean), maximum temperature (T max), minimum temperature (T min), relative humidity (RH), dew point (Dew P), rainfall (Rain) and wind speed (Wind S). The STATGRAPHICS Centurion XVI version 16.1.11 was used for the statistical analysis.
Identification and evaluation of risk infection periods
For the identification of the real infection risk periods caused by B. cinerea, we followed the disease triangle principle taking into account the three causal agents for plant disease occurrence. First, we developed a phenological calendar for each study year based on field observations to determine the timing of the susceptible stages for B. cinerea development: flowering (stage 6 BBCH) and ripening of berries (stage 8 BBCH). From this, we considered the influence of suitable weather conditions within the susceptible phenological stages by applying the Magarey model to identify the main meteorological suitable periods of disease development. As previously stated, this model uses the hourly relative humidity and temperature to obtain a requirement of wetness duration for a critical disease threshold, taking into account meteorological conditions and its proximity to optimal fungal development conditions. Finally, in the suitable meteorological periods detected by the Magarey model during the susceptible phenological stages 6 and 8, we checked the recorded Botrytis airborne spore concentrations in order to identify the real infection risk periods during the subsequent days.
Moreover, we propose the classification of the real infection risk periods detected into three categories: high risk (HR) at spore concentrations ⩾100 spores/m3, moderate risk (MR) if spore concentrations are between 10 and 100 spores/m3 and low risk (LR) with spore concentrations <10 spores/m3.
The study was conducted in collaboration with the main vine company of D.O. Ribeiro, ‘Viña Costeira’ S.R.L. Our experimental data were used by the company to regulate the treatments in their vineyards. Decisions concerning spray administration depended on the combination of phenological observation of a susceptible phenophase, the possibility that predicted meteorological conditions would allow spore germination, the consequent infection of the plants during the next 4–6 days and the detection of spore thresholds in the atmosphere of the vineyard. The fungicides applied in the vineyard were Ciprodinil 37.5%+Fludioxonil 25% or Fenhexamida 50% (WG) P/P by means of fogging and farm tractor mechanical application.
Results
Detection of the pathogen presence in the vineyard
The highest spore amount was registered in 2008, with 39 299 spores, followed by 2014 with 32 073 spores. The lowest spore amount was observed in 2011 with 5747 spores. The maximum values of daily spore concentrations were registered on 28 May 2008 and 7 May 2008 with a value of 1669 and 1495 spores/m3, respectively. In general, the main peak values were observed every year between the end of May and the beginning of July (except for the two latest seasons, where the highest values were detected on 15 September in 2014 and during 2015, on 7 May).
Phenological observations related to pathogen presence in the vineyard
The airborne spore concentrations in the vineyard were analysed in relation to the main phenological growth stages of grapevine. We determined the phenological stage with the highest airborne B. cinerea spore concentrations and the highest total annual spore amount per season (Table 2). The maximum values corresponded with a period between stage 5 (inflorescence emergence) and stage 7 (development of fruits). In addition, our study showed that the average spore concentrations per phenological stage were mostly higher during the flowering stage (years 2009, 2012 and 2013) and the previous (stage 5: inflorescence emergence) and later (stage 7: development of fruits) phenological stages (Table 2).
Table 2. Start date and length (days) of the principal BBCH growth stages (S-1 Leaf development, S-5 Inflorescence emergence, S-6 Flowering, S-7 Development of fruits, S-8 Ripening of berries), with the corresponding average and maximum airborne Botrytis spore concentrations for each stage
Suitable meteorological conditions for Botrytis spore germination
The Magarey generic fungal prediction model was applied to determine suitable meteorological disease development periods during the phenological stages of flowering (stage 6) and ripening of berries (stage 8), and it detected several episodes of disease development (Fig. 2). The model predicted, for all considered years, the lowest requirements for possible infections during flowering, with a range of values between 0.559 and 2.387 h. This indicates that flowering is the grapevine phenological stage most susceptible to Botrytis infection according to the meteorological conditions included in the Magarey model for our bioclimatic region. Considering all years during the flowering stage, the highest number of infection risk periods was identified in 2013, with eight suitable meteorological periods, while the year with the lowest number of suitable meteorological periods was 2011 with only one (Fig. 2).
During ripening of berries stage, the Magarey model also identified possible suitable meteorological periods. Generally, the total number of disease periods detected was higher during this stage, probably because the length of this growth stage is generally longer than flowering (Table 2). During this stage, the maximum number of suitable periods was detected in 2008 (21 suitable meteorological periods) whereas the minimum in 2013 (seven suitable meteorological periods). Nevertheless, the derived wetness requirement values (W (T)) were higher than the obtained during flowering, with a range of 4.073–9.772 h.
A PCA analysis was performed to ascertain the meteorological variables that most influenced spore presence in the atmosphere. The PCA analysis for the 2008–2015 data set extracted two principal components that accounted for 66.3% of the data variance. The first component (Cp1) explained 46.5% of the variance, and it grouped the mean, maximum and minimum temperature with dew point. This component clearly clustered temperature-related variables. The second component (Cp2) explained the 19.8% of variance and grouped the airborne spore concentrations with the humidity-related variables of relative humidity and rain with positive correlation coefficients, and wind speed with a negative correlation coefficient (Table 3). This indicates that wind speed has a negative influence on spore concentrations (Table 3).
Table 3. Factor loadings of the considered meteorological and aerobiological variables for the 2008–2015 data set
In bold the loadings with the largest value for each variable.
Identification and evaluation of risk infection periods
The results of the risk infection period evaluation were graphically represented in Fig. 2 through the combination of phenological growth stages, Magarey suitable periods and airborne spores concentrations. Additionally, we noted fungicide treatment dates due to their importance in the presence of spores in the atmosphere. Table 4 shows the chemical fungicides used, the dates of application, the corresponding daily airborne spore concentrations and the reason for each treatment decision. The decision to spray fungicide depended on the combination of the phenological observation of a susceptible phenophase, the spore threshold and the potential for suitable meteorological conditions in the vineyard that would allow spore germination and the consequent infection of the plant during the subsequent 4–6 days. According to our data and the field observations of the agronomic technicians of the company owning the experimental vineyard, we established a general spore threshold of 100 spores/m3 for a high risk of infection (HR).
Table 4. Date of the fungicide treatments, type of fungicide (C: Ciprodinil, FL: Fludioxinil, Fh: Fenhexamida), spore concentration in the atmosphere, the day of the spray and reason of spray decision (Avoidable: the Company decided to administrate the spray unilaterally; Preventive treatment: spray administration due to suitable meteorological conditions for spore germination; High spore concentration in the vineyard: spray administration due to high presence of pathogen in the atmosphere of the vineyard; Infection risk period during the next 4–6 days: spray administration due to the detection of high or moderate infection risk periods during the next 4–6 days because of the combination of a propitious phenological stage, exceedance of spore concentration thresholds and suitable Magarey meteorological periods)
Considering the interaction of the three causal agents required for infection, the year with the highest number of high-risk periods (HR) was 2008 with 15 periods during flowering and ripening stages, while the lowest number of HR was obtained for 2010 and 2015, both with two periods during flowering and ripening. The highest number of moderate-risk periods (MR) was obtained for 2008 and 2010, with 11 periods considering flowering and ripening together, while the lowest number was detected in 2014 with just three periods. We obtained HR and MR periods for all the considered seasons, but low-risk periods (LR) were only detected for 2010 and 2011, with four and one periods, respectively, during both flowering and ripening.
Analysing each susceptible phenological stage separately, we observed that the highest number of HR and MR during the stage 8 ripening of berries was detected in 2008, with 10 HR-S8 and 11 MR-S8 periods (Fig. 3). In the case of stage 6 flowering, the highest number of HR was obtained for 2013 with seven HR-S6, and the highest number of MR was registered for 2010 with five MR-S6. An average of five risk periods was found for the studied years during flowering, considering both high- and moderate-risk periods. For the ripening stage, double the number of risk periods was found, with an average of ten high- and moderate-risk periods for the studied years. Within this average, the moderate-risk periods for ripening stand out, with seven MR-S8 as the mean value (Fig. 3).
Fig. 3. Number of high-risk (HR) and moderate-risk (MR) periods for B. cinerea infection from 2008 to 2015 during the phenological stages of flowering (S6) and ripening (S8).
Discussion
Prediction models based on the main risk factors that lead crops to epidemic diseases are of great importance for the integrated management strategies. Traditionally, the main risk infection factors were associated with agricultural practices (such as crop rotation, planting dates, tillage practices, etc.), environmental conditions (propitious meteorological situations) or different host susceptibility according to the phenological stage of the plant (Rosa et al., Reference Rosa, Gozzini, Orlandini and Seghi1995; Twengström et al., Reference Twengström, Sigvald, Svensson and Yuen1998; Rossi and Giosuè, Reference Rossi and Giosuè2003; Rossi et al., Reference Rossi, Giosuè, Pattori, Spanna and Del Vecchio2003; Manter et al., Reference Manter, Reeser and Stone2005; Paul and Munkvold, Reference Paul and Munkvold2005; De Wolf and Isard, Reference De Wolf and Isard2007; Ciliberti et al., Reference Ciliberti, Fermaud, Languasco and Rossi2015). The disease triangle is one of the paradigms in plant pathology (Stevens, Reference Stevens1960) that claims the existence of a plant disease absolutely requires the interaction of a susceptible host, a virulent pathogen and environment-favourable conditions for disease development (Stevens, Reference Stevens1960; Agrios, Reference Agrios2005). Therefore, plant disease is prevented with the absence of any one of these three causal components. Based on this, the combination in our study of (i) the identification of the grapevine phenological stage vulnerable to infection, (ii) the observation of spore levels considered as a pathogen biosensor and (iii) the identification of suitable environmental conditions via the meteorological Magarey model led us to ascertain the optimal moments for phytosanitary treatment, taking into account the environmental characteristics and biological conditions of this agroecological system. The proposed method deals with the need for an effective monitoring system and the establishment of disease thresholds aimed to an appropriate decision-support system for crop management guidance on pest control based on the third principle of ‘Decision based on monitoring and thresholds’ included in the eight principles of the Integrated Pest Management (Barzman et al., Reference Barzman, Bàrberi, Birch, Boonekamp, Dachbrodt-Saaydeh, Graf, Hommel, Jensen, Kiss, Kudsk, Lamichhane, Messéan, Moonen, Ratnadass, Ricci, Sarah and Sattin2015).
Detection of the pathogen presence in the vineyard
The Botrytis spore presence was constant in the atmosphere of the vineyard during the study period, as previously noted by other authors in the Iberian Peninsula (Oliveira et al., Reference Oliveira, Guerner-Moreira, Mesquita and Abreu2009; Rodríguez-Rajo et al., Reference Rodríguez-Rajo, Jato, Fernández-González and Aira2010). We proposed 100 spores/m3 as a threshold for high risk of infection caused by B. cinerea, based on our field observations and the data of the agronomic technicians of the company owning the experimental vineyard. We observed that lesions in vines started when airborne spore values were above the 100-spore level, indicating that this represents an important disease risk indicator in the vineyard, and a reasonable threshold to justify a fungicide treatment. Furthermore, Carisse and Van der Heyden (Reference Carisse and Van der Heyden2015) found similar results on their study about the influence of airborne conidia concentration on flower and stem-wound infections at three different temperatures of 15, 20 and 25°C. They observed that no infection of stem-wounds occurred under 100 conidia/m3, and the proportion of infected flowers remained low under 10 conidia/m3 rising with the increase of the spore concentrations. Based on this, we proposed a second threshold for moderate risk of infection between 10 and 100 spores/m3.
Phenological observations related to pathogen presence in the vineyard
Several authors have noticed the synchronism between the most vulnerable grapevine growth stages and B. cinerea presence in many geographical areas (Esterio et al., Reference Esterio, Muñoz, Ramos, Cofré, Estévez, Salinas and Auger2011; Ciliberti et al., Reference Ciliberti, Fermaud, Roudet, Languasco and Rossi2016; Carmichael et al., Reference Carmichael, Siyoum, Jongman and Korsten2018; Hatmi et al., Reference Hatmi, Villaume, Trotel-Aziz, Barka, Clément and Aziz2018; Martínez-Bracero et al., Reference Martínez-Bracero, Alcázar, Velasco-Jiménez and Galán2018). The most widely accepted critical stages for grey mould infection are flowering (stage 6) and ripening of berries (stage 8). During the flowering stage, pollen and sugar exudation favour the colonization of tissues by the pathogen (Esterio et al., Reference Esterio, Auger, Droguett and Arroyo1996). This nutritional effect is also evident during ripening, when the presence of sugars, which increases over the stage, has a synergetic effect with the increasing ontogenic susceptibility of grapes as they mature, and propitious meteorological conditions. This effect increases disease intensity in the stage closest to the grape harvest (Latorre, Reference Latorre1986; Bulit and Dubos, Reference Bulit, Dubos, Pearson and Goheen1988; Kretschmer et al., Reference Kretschmer, Kassemeyer and Hahn1994). Frenguelli (Reference Frenguelli and Moscato2001) noted a Botrytis airborne spore concentration peak in September, when the fungus develops on senescence leaves besides on ripening fruits. Furthermore, infection events during flowering have special interest due to fungal colonization of floral debris. This process represents an important inoculum source for late infection episodes, affecting during maturation and before harvest when the berries susceptibility to B. cinerea infections increases (Holz et al., Reference Holz, Coertze and Basson1997; Wolf et al., Reference Wolf, Baudin and Martínez-Ochoa1997). Our results support these affirmations as we found that the highest spore levels were registered near flowering stage. These findings agree with the values obtained by the Magarey fungal disease model, as we detected risk periods for critical disease incidence at this phenological stage for the study area. Furthermore, we also found high spore concentrations during the ripening of berries stage, corroborating the results noted by other authors (Fernández-González et al., Reference Fernández-González, Rodríguez-Rajo, Jato and Aira2009; Rodríguez-Rajo et al., Reference Rodríguez-Rajo, Jato, Fernández-González and Aira2010).
Suitable meteorological conditions for Botrytis spore germination
The Magarey generic fungal infection model applied in this study was one of the first to demonstrate that a single temperature-driven equation can be used to simulate infection response for several vineyard pathogens (Magarey et al., Reference Magarey, Sutton and Thayer2005). This model is a simplification of the ecosystem functioning based on meteorological variables, but actually, other parameters have marked influence on disease progress processes. A higher number of infection risk periods was detected during the ripening of berries stage, although with higher model values than those obtained for the flowering stage, which indicates that plants are less susceptible to this fungal disease because of weather conditions. These elevated values of hours of wetness duration requirement indicate that the meteorological conditions were not as favourable for fungal development during the ripening of berries stage as during flowering in the Northwestern Spain region.
The results of the PCA statistical analysis showed the influence of humidity-related variables and wind speed on Botrytis airborne spore presence in the atmosphere. These results accurately describe the development and dispersal behaviour of the pathogen as humidity is widely considered as a critical factor for grey mould spore germination and infection (English et al., Reference English, Thomas, Marois and Gubler1989; Broome et al., Reference Broome, English, Marois, Latorre and Aviles1995). The negative correlation found between airborne spore concentrations and wind speed reflects the predominant dispersion mechanism of propagules in Botrytis species, usually dry conidia scattered by wind (Holz et al., Reference Holz, Coertze, Williamson, Elad, Williamson, Tudzynski and Delen2007). Moreover, this negative association could indicate that the spores detected in the atmosphere of the vineyard are released in the study plot itself, instead of being transported from other areas through long distance transport processes. This circumstance is reflected in a positive correlation between wind and airborne spore concentrations (Moreno-Grau et al., Reference Moreno-Grau, Aira, Elvira-Rendueles, Fernández-González, Fernández-González, García-Sánchez, Martínez-García, Moreno, Negral, Vara and Rodríguez-Rajo2016).
Identification and evaluation of risk infection periods related to fungicide application
The proposed method for the identification of the infection risk periods, which account for meteorological conditions, the vines phenological stage and the pathogen's presence, represents a valuable tool for the development of fungicide application schedules. The identification of the main infection risk periods based on spore thresholds makes possible the disease detection between 4 and 6 days before the symptoms appearance, as it is verified in field observations (data not shown). Once the spores are present in the air of the vineyard in higher concentrations than the marked threshold, they still need 4–6 more days (depending on the phenological stage) to develop a new fungus and lesions under propitious meteorological conditions. Carisse et al. (Reference Carisse, Savary and Willocquet2008) also found a significant correlation between airborne spore concentration on a given date and lesion density 1 week later for unmanaged and managed sites on their study.
For the evaluation of risk periods (HR, MR or LR) for disease development, we applied the airborne spore-level thresholds for each category to the identified Magarey suitable periods during the flowering and ripening of berries stages. The obtained results of the evaluation of infection risk periods for the eight studied seasons showed, on average, the same number of three HR infection periods for ripening and flowering. Nevertheless, marked differences were found for the MR periods as we detected two for flowering and seven for ripening. This suggests that the critical grapevine phenological stage for B. cinerea infection in our bioclimatic area is the ripening stage.
During the years prior to the present study, 4–5 sprays against grey mould were applied annually in the vineyard following preset calendars based on phenology. The first treatment was conducted annually 2 weeks before flowering, and then during the June to August period, three to four additional treatments were conducted (one in June, another in July and usually one to two in August). The combination of the Magarey model, the aerobiological data and the phenological observations represents a stronger resource for disease risk prediction taking into account both environmental conditions and fungal development, and allowing to reduce the number of phytosanitary treatments on the vineyard. Using the proposed methodology, we either prevented the appearance of lesions or reduced the presence of lesions, as the sprays were applied at a time prior to the visibility of lesions in plants. In the present study, from 2008, annual chemical treatments in the vineyard were reduced to 2–3 depending on the year, achieving a 25–35% reduction in fungicide treatments. The first treatment was conducted during May, but not during all years, and the second and third treatments in July and August. The spray administration depended on the combination of phenological observation of a susceptible phenophase, the exceedance of the 100 spores/m3 threshold and that the potential suitable meteorological conditions in the vineyard would allow spore germination during the subsequent 4–6 days.
The decision for most of the applied fungicide treatments in the study vineyard from 2008 to 2015 was motivated by high airborne spore concentrations or by high/moderate infection risk period detection (because of the combination of a propitious phenological stage, the rise of the spore concentration and a suitable Magarey meteorological period). Moreover, during the studied period, two preventive treatments were administrated in the years 2009 and 2010, and three avoidable treatments were unilaterally applied by the company on 22 July 2010, 18 July 2011 and 17 July 2015. These treatments were applied during the phenological stage 7 of development of fruits (which is not considered as a susceptible stage for Botrytis infections), and under low airborne spore concentrations.
Despite the achieved reduction in treatments in comparison with preset calendars, several authors have demonstrated the effectiveness of plant disease control strategies based on disease forecasting. Madden et al. (Reference Madden, Ellis, Lalancette, Hughes and Wilson2000) found similar disease incidence in crops with standard treatments based on preset calendars to those with chemical application based on the identification of environmental conditions which favour sporulation and infection integrated into a warning system. A secure and effective disease control can be implemented by using this kind of models.
Furthermore, we propose to administrate a treatment in the previous period to the flowering stage, identified as highly vulnerable for our climatic region by means of the Magarey model, in order to prevent latent infection and the appearance of disease symptoms. This fact was notable in the last two studied years, 2014 and 2015, where a preventive chemical treatment was applied during the previous phenological stage to flowering (stage 5 inflorescence emergence) coinciding with very high airborne spore concentration levels. The treatment greatly contributed to the control of the latent disease, which would turn into inoculum sources for later infections by colonization of senescent floral debris and aborted berries (Wolf et al., Reference Wolf, Baudin and Martínez-Ochoa1997; Rodríguez-Rajo et al., Reference Rodríguez-Rajo, Jato, Fernández-González and Aira2010). It was possible to control later infection cycles in 2015 and to reduce spore levels in 2014. Moreover, a possible treatment reduction could be achieved in later stages in 2008, 2009 and 2013 with the application of a preventive treatment during the stage 5, since the highest airborne fungal propagule load coincided near to flowering stages in these years but no treatment was applied during this critical phase.
Finally, it is notable that the proposed method can also act as a crop protection tool against the consequences of climate change because the increase in the variability of climatic conditions may affect vegetal phenology and pathogen biology (Dalla Marta et al., Reference Dalla Marta, Grifoni, Mancini, Storchi, Zipoli G and Orlandini2010; Lamichhane et al., Reference Lamichhane, Barzman, Booij, Boonekamp, Desneux, Huber, Kudsk, Langrell, Ratnadass, Ricci, Sarah and Messéan2015). This climatic variability and changes are not considered in preset calendars for chemical product applications. The developed evaluation of infection risk can be adapted to other fungal grapevine pathogens, or even for other crops, by considering the specific pathogen requirements, the environmental conditions and the vegetal susceptibility according to the plant phenological stage (Paul and Munkvold, Reference Paul and Munkvold2005; De Wolf and Isard, Reference De Wolf and Isard2007). These conditions affect the pathogen–crop relationship at a microscale level because the canopy microclimate regulates fungal growth and development. Several factors, such as wind speed within the canopy, temperature, atmospheric humidity and leaf wetness vary markedly depending on agricultural practices that potentially change canopy architecture, such as leaf removal, plant spacing, cultivar selection or irrigation practice (English et al., Reference English, Thomas, Marois and Gubler1989).
Conclusions
The proposed method resulted in a useful tool with which to dynamically predict the main grey mould infection risk periods by means of: phenological observations to identify the susceptible stages, exceedance of the 100 spores/m3 threshold in the atmosphere of the vineyard and the possibility that propitious meteorological conditions enhance spore germination over the subsequent 4–6 days. This supposes a time window enough for winegrowers to apply the required chemical treatments before the Botrytis lesions appearance on the crop. A 25–35% reduction in the number of fungicide treatments was achieved following the proposed method in the studied vineyard. This reduction promotes the protection of the environment and human health, and the reduction of economic costs with an added improvement in the products obtained. These are the most important current challenges of winegrowers, especially those related to the wine Designation of the Origin area.
Authorship declaration
The authors have contributed significantly and they are in agreement with the data presented in the present study.
Financial support
This work was funded by Xunta de Galicia CITACA Strategic Partnership (Reference: ED431E 2018/07) and the AGL2014-60412-R Economy and Competence Ministry of Spain Government project. Fernández-González M. was supported by FCT (SFRH/BPD/125686/2016) through HCOP-Human Capital Operational Program, financed by ‘Fundo Social Europeu’ and ‘Fundos Nacionais do MCTES’. González-Fernández E. was supported by the Ministry of Sciences, Innovation and Universities (FPU grant FPU15/03343). Piña-Rey A. was supported by Xunta de Galicia Pre-doctoral Period Support Program (ED481A-2017/xxx).
Conflict of interest
The authors declare that there are no conflicts of interest.
Ethical standards
Not applicable.
