Introduction

Grapevine (Vitis vinifera L.) phenology is mainly temperature-driven (Jones, 2013). Air temperature is the main factor influencing the timing of phenological phases (e.g., Gladstones, 2011; Jones, 2013) if water and radiation requirements of the plants are fulfilled (Nendel, 2010; Webb et al., 2007). Consequently, grape phenology models in viticulture are usually based on air temperature observation data as a single input parameter (e.g., Caffara & Eccel, 2010; Cola et al., 2014; Molitor et al., 2014b; Parker et al., 2011; Parker et al., 2020).

Many phenological models assume linear forcing effects of air temperature by accumulating temperatures above a defined threshold to temperature sums or cumulative degree days (e.g., Amerine & Winkler, 1944; Duchêne et al., 2010; Nendel, 2010; Parker et al., 2011). However, in such unlimited temperature sum approaches the physiological evidence that above plant-specific threshold temperatures the forcing effect of the temperature will not increase further or even tends to decrease is neglected (Bonhomme, 2000; Molitor et al., 2014b; Wang & Engel, 1998; Yan & Hunt, 1999). Consequently, Molitor et al. (2014b) incorporated (i) an upper threshold temperature and (ii) a heat threshold temperature in their temperature sum-based model approach and demonstrated a significantly improved model accuracy compared to the uncapped cumulative degree day approaches previously published. Meanwhile, this approach has been proven to be of high accuracy in a broad range of cultivars using a unified threshold triplet of 10, 20 and 30 °C as lower, upper and heat temperature thresholds (Molitor et al., 2020). This “UniPhen” approach is covering all phenological stages according to the BBCH scale and is open for the incorporation of any other grape cultivar based on high-resolution observation data (Molitor et al., 2020).

Climate change is threatening viticultural production (van Leeuwen et al., 2024). Global near-surface mean air temperature has increased by over 1 °C over the past century, and all major climate projections are projecting a further increase by up to 3 °C by 2100 (IPCC, 2021). Moreover, global warming is expected to lead to a general increase in the frequency and intensity of extreme weather events (IPCC, 2021) such as droughts and heat waves (Fraga et al., 2020), threatening grape yield as well as wine quality and typicity.

In viticulture, weather and climate control grapevine growth, physiology, yield, and berry composition (Santos et al., 2020). Hence, climate change is a direct threat to the socio-economic sustainability of many viticultural regions (e.g., Lereboullet et al., 2013; Mosedale et al., 2016). Furthermore, a large portion of the total active ingredient volume of pesticides used in agriculture is applied in vineyards, and the use of fungicides in viticulture alone accounts for more than 70 % of the total fungicide use in Europe (Wingerter et al., 2021). Via the application of fungicides, viticulture contributes to climate change by releasing greenhouse gases during pesticide synthesis, transport, and application in the field via viticultural machinery. Addressing climate change in viticulture requires significant efforts to (i) adapt to changing environmental conditions as well as (ii) mitigate greenhouse gas emissions by promoting more sustainable forms of vineyard management.

In this context, the cultivation of new fungus-tolerant or fungus-resistant grape cultivars (Töpfer & Trapp, 2022), the so-called “PIWIs - pilzwiderstandsfähige Rebsorten”, has received increasing attention in recent years as they bear the potential to address climate change adaptation and mitigation simultaneously. PIWIs are new, cross-bred cultivars bearing resistance genes stemming from American or Asian Vitis species (Pedneault & Provost, 2016) and display reduced susceptibility towards major fungal diseases in viticulture, namely downy and powdery mildew. Consequently, PIWI cultivation allows for a substantial reduction in the use of fungicides (minus 50–80 %) (Töpfer & Trapp, 2022), which represent 96 % of pesticides used in viticulture (Marinho et al., 2020).

By increasing PIWI cultivation, the wine industry could thus actively contribute to reducing greenhouse gas emissions and ensure the sustainability of the sector under changing climatic conditions, while reducing its overall environmental footprint. Accordingly, increased PIWI cultivation may not only contribute to achieving pesticide reduction targets, but also to the EU Biodiversity strategy for 2030. Furthermore, the cultivation of PIWI cultivars represents an economical risk reduction strategy under increasingly frequent extreme weather conditions such as long-lasting rain periods in summer, causing severe calamites of fungal diseases and hampering access to the vineyards with machines.

Despite their potential benefits, PIWIs are presently cultivated only in 0.5 % of the viticultural area of Luxembourg (Anonymous, 2023) while in countries like Switzerland or Germany, the PIWI cultivars exceed 3 % of the nationwide planted area (Destatis, 2023; OFAG, 2023). In the relatively new and rapidly growing grape-growing country, Belgium, meanwhile, PIWIs represent 29.6 % of the viticultural area (Anonymous, 2024). Anyhow, there is still a lack of detailed information about the viticultural traits of these new cultivars, including the response of grape phenology to heat consumption that is key to consider the suitability of PIWI cultivars across different climatic conditions. The growing interest in PIWI cultivars demonstrated the need to integrate them into existing models. Hence, there is a need for more detailed and transregional studies on the differences in phenological development of different PIWI cultivars, especially in the context of climate change and fungal disease pressure.

The aim of the present investigation was (i) to obtain datasets of high-resolution phenological observations of PIWI cultivars (and traditional reference cultivars) at a broad range of European locations, (ii) to apply the temperature sum based UniPhen model for all cultivars of observation allowing for a precise simulation of the phenological development at all BCCH stages between beginning of bud swelling and berries ripe for harvest and (iii) to discuss the potential implementations of UniPhen “PIWI” in applied viticulture as well as in viticultural climate change impact research.

Materials and methods

1. Experimental vineyards

The phenological monitoring was carried out at 15 experimental vineyards across Europe (Figure 1). All sites are located along a topographical gradient with Ath (Belgium) and Haidegg (Austria) as the two extremes with elevations of 60 and 600 m a.s.l., respectively. All sites provide a fair representation of the climatic conditions and soil types that characterise the vineyards cultivated in Europe outside the Mediterranean region. According to the Köppen-Geiger climate distribution in Europe (Beck et al., 2018), the sites Wädenswil and Haidegg characterised the continental cold climate with dry season and warm summer (Dfb), Laimburg characterised the temperate climate with hot summer and no dry season (Cfa), and the rest of the sites the temperate climate with warm summer and no dry season (Cfb). The main soil types are represented by a Cambisol (CM) soil unit that is present in eight of the experimental vineyards (Table 1). The Luvisol (LV) is found in Ath (Belgium) and Kreuznach (Germany), and the soil of the remaining four sites are Fluvisols (FV), Phaeozems (PH), Planosols (PL), and Regosols (RG).

Systematic phenological assessments took place in the years 2021 to 2024 in the following locations (Table 1):

Only those locations providing four years of complete data sets were selected for model calibration.

Phenology of the following cultivars was recorded: Cabernet blanc, Cabernet Cortis, Cabertin, Calardis blanc, Johanniter, Monarch, Muscaris, Pinotin, Regent, Sauvignac, Solaris, Souvignier gris, Villaris, Divico, Helios, Bronner, Cabaret noir, Cabernet Carbon, Divona, Rondo, Hibernal, Prior, Baron, Floreal, Satin noir, Vidoc, Artaban, Laurot, VB 1-22, Donauriesling, Müller-Thurgau, Pinot noir, Riesling, Chardonnay, Chasselas and Dornfelder. Only those location × cultivar combinations with complete data sets for all four years of observation, leading to at least 8 complete data sets (covering all BBCH stages between 01 and 89) per cultivar, were used as input data for model calibration (Table 2).

In total, calibration data sets consisted of the following complete data sets (covering all BBCH stages between 01 and 89 in all four years of observation) originating from the following locations (Table 3).

All cultivars were trained in a cane-pruned vertical shoot positioning system (VSP). Six plants were monitored per cultivar; the plants of observation were the same in all four years.

2. Assessment of phenological stages

All phenological growth stages according to the BBCH (Biologische Bundesanstalt, Bundessortenamt und Chemische Industrie) scale as defined by Lorenz et al. (1995) (Table S1) between BBCH 01 (beginning of bud swelling) and BBCH 85 (softening of berries) were recorded when 50 % of the vines or shoots exhibited the respective stage. The assessment intervals ranged from two to three days. Usually, records were taken at the same location by the same person in all years and cultivars. The date of BBCH 89 was defined as the DOY (day of the year) when 60° Oechsle (= 14.17 °Brix) was reached or exceeded for the first time in the respective year following the definition of Molitor et al. (2020). Maturity control took place at weekly intervals between veraison (BBCH 81) and harvest by collecting approximately 50 randomly selected berries (clusters from different positions of the canopy; berries from different positions in the cluster) per cultivar (avoiding berries with visible bunch rot symptoms). The date of reaching a specific phenological stage was noted as day of the year (DOY).

3. Meteorological measurements

Daily average air temperatures at 2 m height were recorded during the period of examination by standard automatic weather stations in proximity (less than 200 m) of the experimental vineyards. Average growing season (April to October) temperatures per location and year are given in Table S2.

The highest growing season temperatures were observed in the year 2022 (average 17.3 °C), which was on average 2.5 °C warmer than in 2021 (average 14.8 °C). On average, Ebringen was the location with the highest average growing season temperatures (17.5 °C), while the lowest average growing season temperatures were recorded in Ath (15.4 °C).

4. Application of the unified high-resolution phenological model approach “UniPhen”

In the present study, the approach of the high-resolution cumulative degree day-based phenological model, as previously developed for the Rivaner cultivar (Molitor et al., 2014b), parameterised for Riesling (Molitor et al., 2016) and Pinot noir (Molitor & Junk, 2019), as well as unified by using a unified global optimised temperature threshold triplet (lower threshold: 10 °C, heat threshold: 20 °C, heat threshold: 30 °C) (Molitor et al., 2020) was applied to all selected cultivars.

In this “UniPhen” approach, degree days (DD) for a specific day are calculated based on the daily average air temperatures, applying the following conditions:

DD_10,20,30 = 0, if t ≤ 10 °C or if t ≥ 40 °C

DD_10,20,30 = (t – 10), if 10 °C ≤ t ≤ 20 °C

DD_10,20,30 = 10, if 20 °C ≤ t ≤ 30 °C

DD_10,20,30 = 10 – (t – 30), if 30 °C ≤ t ≤ 40 °C

where DD_10,20,30 is the degree day for a specific day and t is the daily average temperature.

For every growth stage, degree days (DD) were summed up to cumulative degree days (CDDs) using the Equation (1):

${\text{CDD}}_{\text{10,20,30}}={\sum }_{\text{i=m+}1}^n{\text{DD}}_{\text{10,20,30}}\left(i\right)$ (1)

where CDD_10,20,30 are the cumulative degree days relative to the date of BBCH 09 in Riesling; DD_10,20,30 the value of the degree day for the day i; m the DOY when BBCH 09 was observed in Riesling; and n the DOY when the target phenological stage was reached in the respective cultivar.

The reference date was the day of the year (DOY) at which BBCH 09 was reached in Riesling (following Molitor et al. (2020)).

If no observation data for BBCH 09 for Riesling were available (Marcelin, Freiburg, Ebringen), its virtual DOY was calculated based on the observed DOY of BBCH 09 for Pinot noir. According to the data presented by Molitor et al. (2020), Pinot noir reaches BBCH 09 6.1 CDD_10,20,30 prior to Riesling. Hence, the date of BBCH 09 in Riesling was fixed at the DOY at which a temperature sum of 6.1 CDD_10,20,30 after to the DOY of observation of BBCH 09 in Pinot noir was reached or exceeded.

5. Data analysis

To determine general temperature sum thresholds to simulate all phenological stages between BBCH 01 and BBCH 89 in all cultivars, average CDD_10,20,30 were calculated for each cultivar, relative to the date of reaching BBCH 09 in Riesling. Standard deviations of the cumulative degree days were calculated for each phenological stage in each cultivar and each location. To normalise the relative value of the standard deviations caused by the daily DD values, standard deviations were divided by the theoretical DD at 15 °C (approximate growing season temperature in Remich in the period of observation, selected following Molitor et al. (2020)). These normalised standard deviations, SD_15° are given for all phenological stages, all cultivars and all locations.

The R script used to calculate the CDD per cultivar is available in Jiménez-Rodríguez (2024) as UNIPHEN-Tool v1.0.0.

To assess the goodness of fit of the model, mean bias errors (MBE) and mean absolute errors (MAE) were calculated for all years, locations and cultivars as an average of all BBCH stages. A positive MBE value indicates that, on average, the model overestimates actual observations; a negative value indicates that the model underestimates actual observations under specific annual or local conditions (Caffara & Eccel, 2010).

Results

1. Phenological observation data

Days of the year (DOY) on which the phenological stages BBCH 01 to 89 were reached in the years 2021 to 2024 at the selected locations (complete data sets over four years) in the 16 cultivars of observation are shown in Table S3.

2. Cultivar specific temperature sum thresholds for the different phenological stages

Figure 2 shows the average CDD_10,20,30 threshold values relative to BBCH 09 in Riesling for the different cultivars. On average, the lowest CDD_10,20,30 value for BBCH 09 was recorded for Muscaris (–13) and the highest for Pinotin (3). The lowest CDD_10,20,30 value for BBCH 61 was observed for Solaris (195). Divico showed the lowest CDD_10,20,30 value for BBCH 81 (601), Solaris for BBCH 89 (797). For all those three stages highest values were observed for Riesling (BBCH 61: 251, BBCH 81: 859, BBCH 89: 1012) (Figure 2). Figure 3 displays graphically the differences between the different cultivars in CDD_10,20,30 values for the respective stages.

Lowest SD_{15 °C} values (except BBCH 09 for Riesling as reference) were observed in case of all stage × cultivar combinations for Muscaris at BBCH 11 (0.6 = 0.6 days at 15 °C) and highest for Riesling at BBCH 79 (19.2). On average of all cultivars, lowest SD_{15 °C} values were recorded for BBCH 09 (1.2) and highest for BBCH 79 (13.4). On average of all stages, the lowest SD_{15 °C} values were observed in Sauvignac (5.0) and the highest in Divico (6.7), with 5.5 being the average SD_{15 °C} value over all 16 cultivars (Figure 4).

Highest MBE values (recording later than simulated) were recorded in Cabernet Cortis in Geilweilerhof (20.1 CDD_10,20,30) and lowest (recording earlier than simulated) in Pinot noir in Marcelin (–21.4 CDD_10,20,30). On average of all cultivars highest MBEs were observed in case of Geilweilerhof (7.6 CDD_10,20,30) and lowest in case of Freiburg (–12.7 CDD_10,20,30).

Highest MAE values were observed for Pinot noir in Marcelin (23.1 CDD_10,20,30) and lowest for Sauvignac in Neustadt (7.0 CDD_10,20,30). Likewise, on average of all cultivars highest MAE values were reached in Marcelin (17.6 CDD_10,20,30) and lowest in Neustadt (10.2 CDD_10,20,30) (Table 4).

Concerning the vintage effect, highest MBE values were recorded in Muscaris in 2024 (12.9 CDD_10,20,30) and lowest in Regent in 2021 (–23.0 CDD_10,20,30). On average of all cultivars, highest MBE were observed in 2023 (7.3 CDD_10,20,30) and lowest in 2021 (–11.9 CDD_10,20,30). In 2021, the MBE values of all cultivars were negative.

Highest MAE values were observed for Regent in 2023 (25.1) and lowest for Calardis blanc in 2022 (7.0 CDD_10,20,30). On average of all cultivars highest MAE values were recorded for 2021 (16.3 CDD_10,20,30) and lowest for 2022 (9.7 CDD_10,20,30) (Table 5).

Discussion

1. Model application for PIWI cultivars

The application of the UniPhen for 13 PIWI cultivars based on at least eight observation data sets revealed the average CDD_10,20,30 thresholds relative to BBCH 09 in Riesling until the respective BBCH stage is reached per cultivar. It allows for a classification of the relative precocity of all 13 PIWI cultivars (plus three traditional reference cultivars) in each of the 31 stages of the BBCH scale between 01 and 89. As described before (Molitor et al., 2020), the relative precocity of the different cultivars is not stable over the different phenological stages, for example, Calardis blanc is one of the latest cultivars in leaf development, while one of the earliest in beginning of flowering and again the latest PIWI cultivar to reach harvest maturity (Figure 2). According to their precocity in budburst, Muscaris and Solaris could be classified as “early” while Pinotin and Cabertin are rather late, even slightly later than the traditional reference Riesling (consequences are discussed in the paragraph Practical applications).

The present investigations confirmed the general suitability of the cumulative degree day approach using three threshold temperatures as proposed first by Molitor et al. (2014b) and unified later by the introduction of the optimised temperature sum threshold triplet 10, 20 and 30 °C in the UniPhen approach (Molitor et al., 2020). Observed global average normalised standard deviations correspond to 5.5 days at 15 °C (or 2.75 days at 20 °C) demonstrating the strong robustness of the UniPhen model also for the PIWI cultivars.

Generally, for all cultivars, normalised standard deviations SD_{15 °C} are lowest around budburst. In this period normalised standard deviations correspond to 1 to 2 days, indicating the high model precision. While in stages of inflorescence emergence standard deviations slightly increase, during flowering (BBCH 61-69) and fruit-set (BBCH 71) the SD_{15 °C} values correspond to standard deviations of less than 6 days. As described before (Molitor et al., 2020) for traditional cultivars, higher average normalised standard deviations were observed in the bunch closure stages 77 and 79. Here, as well as in the stages of inflorescence emergence the perception of the observer is more subjective than in case of leaf emergence or flowering (Verdugo-Vasquez et al., 2017). Furthermore, the degree of fruit set, water and nutrient availability, as well as the yield potential have an influence on the cluster architecture and, in consequence, the moment of fruit set. Standard deviations corresponding to 13.3 days at 15 °C were observed for stage BBCH 89. These standard deviations can partly be explained by the observation intervals (7 days (weekly maturity control) versus 2–3 days for the other stages) as well as annual differences in sugar accumulation caused by differences in crop load (Santesteban & Royo, 2006).

The effects of the locations as well as the vintages were evaluated using the mean bias error and the mean average error. While in Geilweilerhof on average of all years and cultivars the stages were recorded 7.6 CDD_10,20,30 later than simulated, for Freiburg this was the case 12.7 CDD_10,20,30 earlier (corresponding to 2 to 3 days at 15 °C). These minor effects might be explained by the differences in the perception of the observers or potentially micro-climatic effects caused by the position of the weather station relative to the vineyard of observation. Observed average mean average errors at the different locations between 10.2 CDD_10,20,30 in Neustadt and 17.6 CDD_10,20,30 in Marcelin might be reasoned by differences in the observation interval or the consistency of the observations.

Interestingly, concerning the deviations caused by the vintage, all cultivars showed in 2021 a negative mean bias error, meaning that all stages were on average earlier observed than simulated, with deviations corresponding to an average of 2 days at 15 °C. While this deviation of 2 days might be acceptable in the application of the model, this systematicity might have its origin in the consistently lower temperatures in 2021 compared to the two following years and/or non-linear dependencies between the temperatures recorded by the weather stations and the apparent temperatures in the vineyard canopy.

2. Practical applications

Generally, the UniPhen approach is designed to allow the simulation of phenological phases, where environmental conditions or crop cultural measures, for example, have a distinct influence on factors such as yield formation (Molitor & Keller, 2016), wine typicity (Molitor & Junk, 2019) or susceptibility towards pests or diseases (Molitor et al., 2016; Molitor et al., 2020).

Usually, PIWI cultivars are not completely resistant against fungal infections, but of (i) lower susceptibility and/or (ii) their period of susceptibility of specific organs is shorter than in traditional cultivars since ontogenetic resistance appears phenologically earlier. Frequently, a limited number of fungicide applications in PIWI cultivars is recommended in the period of highest susceptibility for berry infections (Schumacher et al., 2024). Periods of highest susceptibility, however, differ in their length between different PIWI cultivars, for example, Cabernet Cortis and Solaris are susceptible towards Plasmopara viticola until BBCH 71-73, whereas Souvignier gris is still susceptible at BBCH 75 (Schumacher et al., 2024). Hence, simulating the phenological development supports a more targeted timing of fungicide treatments in PIWIs depending on their phenology driven status of susceptibility. The knowledge about the phenological status of different PIWI cultivars potentially contributes to a further reduction of pesticide use if applications outside the period of susceptibility are abandoned, for example, some PIWI cultivars might need to be treated at a specific point in time while other cultivars passed already their cultivar specific phenological growth stage of ontogenetic resistance. Here, the combination of the knowledge of cultivar-specific phases of susceptibility with UniPhen “PIWI” might deliver valuable decision support for users.

The obtained classification of relative cultivar precocity at different stages opens new chances for cultivar selection under practical conditions, especially in regions where late frost damage might constitute a serious threat for sustainable wine production (e.g., Kartschall et al., 2015; Leolini et al., 2018; Molitor et al., 2014a; Mosedale et al., 2015), such as in many cool-climate viticulture regions including the new winegrowing regions emerging in recent years. Especially in these regions, PIWIs are frequently the cultivars of choice due to their reduced susceptibility as well as the missing customer’s habit to prefer traditional cultivars. Under high late frost risk conditions cultivars with early bud development and early bud burst might be avoided based on the information derived from the present model.

Solaris and Muscaris were identified as thermal-temporally early in budburst and, hence, more likely to be damaged by late frost events at dates where other cultivars are not yet susceptible, since green shoots and leaves have not yet emerged. Taking into consideration a long-term average (1991–2020) April to May (period where late frost damage takes place) temperature in Remich of 12.4 °C, the observed difference of 10 (Solaris) or even 13 (Muscaris) CDD_10,20,30 compared to Riesling could be translated to a precocity of four days for budburst and, hence, a four or five days longer risk period for late frost damage. This fact should be considered when thinking about cultivar selection in regions or locations where late frost damage is likely. Concerning the late frost damage risk, Pinotin or Cabertin might be interesting candidates since they are the only PIWI cultivars with a slightly later budburst than the traditional reference cultivar Riesling. Indeed, under cool climate conditions such as in the newly emerging Northern European grape growing region, the ideotype of a PIWI would be one that exhibits a late/delayed bud burst combined with an early full maturity. In fact, the observations as well as the methodology of the present paper could also be useful for breeders in selecting parental material for the breeding of new cultivars with those ideal characteristics.

Generally, PIWI cultivars could potentially pave the way for long-term adaptation strategies to achieve higher climate resilience (Terleth & Tavernar, 2022; Tscholl et al., 2024). Consequently, the percentage of PIWI cultivars in new plantations is increasing in recent years – especially, but not exclusively, in the new winegrowing regions and at higher altitudes, which have in past not been suitable for viticulture due to limited heat consumption. However, so far little was known about the cultivar specific heat demand of PIWI cultivars to produce grapes with adequate maturity for wines of high quality. The BBCH scale defined the stage BBCH 89 as “berries ripe for harvest”. Hence, the temperature sum necessary to reach BBCH 89 is linked to the annual heat demand for a specific cultivar to produce fully mature grapes. Using UniPhen “PIWI” as a bioclimatic indicator allows to describe the suitability of different locations or altitudes for the cultivation of specific cultivars under changing climatic conditions.

Especially the classification of relative precocity of the maturity period (BBCH 81-89) might be used for the selection of PIWI cultivars in climate change adaptation strategies. Particularly high temperatures in the maturation period have been demonstrated to negatively affect wine fruitiness and its aroma (Duchêne et al., 2010). Consequently, PIWI cultivars with a thermal-temporally early ripening period leading to hot temperatures during maturation such as Solaris should be avoided in the traditional Central European wine growing regions – especially since a twofold increase of the ripening period temperatures is expected with ongoing climate change (Molitor & Junk, 2019). On the other hand, early ripening PIWI cultivars are of highest interest for cool winegrowing regions outside the traditional regions as well for viticulture at higher altitudes. Here, it is recommended to check prior to the planting of a vineyard if the annual heat consumption of a region or location fits to the heat demand of the cultivar of interest. The information about the general suitability of cultivars for cultivation under specific climatic conditions can be derived from UniPhen “PIWI”. Based on long-term observation data, (i) virtual dates of reaching specific phenological stages as well as CDD_10,20,30 values at the end of the season can be calculated and (ii) potentially fitting cultivars might be selected, for example, if the CDD_10,20,30 value at the end of the season reaches 800, only Solaris might be a suitable candidate out of the 13 PIWI cultivars tested here. Since the cultivar selections are long-term decisions, additionally, future climate projections might be incorporated to simulate the future heat consumption for a specific region.

Generally, the observed systematic biases caused by the location were low and might be caused by differences in the perception of the observers. The observed mean bias errors between –12.7 and 7.6 CDD_10,20,30 refer to less than 3 days earlier or less than 2 days later observations than simulated, respectively. Consequently, the model proved to be valid under present Central European conditions. Under completely different climatic conditions, the model may need to be validated first with local observation data before being applied. However, the chance that cultivars with heat demands corresponding to the conditions in Central Europe (as the present ones), might be cultivated under clearly hotter conditions appears unlikely. Generally, it must be taken into consideration that extreme weather events, like late frost damage or hailstorm damage, destroying the green plant tissues, might lead to a temporary stop or even a reset of the phenological development. In such cases the model might not be applicable.

Conclusions

Based on a broad set of observation data, the UniPhen model was applied and allows now for a precise simulation of all 31 BBCH stages between the beginning of bud swell (01) and berries ripe for harvest (89) for 13 PIWI cultivars. UniPhen “PIWI” could be used (i) to simulate periods of highest susceptibility towards fungal diseases and contribute to reduce fungicide use to a minimum as well as (ii) a bioclimatic indicator for the suitability of a location/region for the cultivation of specific PIWI cultivars, considering late frost risk and minimum heat demand to produce high quality wines. Present data and the model itself might help to alleviate lowering the existing barriers for grape growers (due to a lack of experience and knowledge) to grow these alternatives to traditional, susceptible cultivars.

UniPhen “PIWI” as well as the UniPhen approach in general is open to be extended for other grape cultivars and application under different or future climatic conditions.

Acknowledgements

The authors thank M. Schultz, D. Dam, C. Simon, P. Lopes (IVV), T. Kaltenbach, C. Mertes (WBI Freiburg), A.-K. Ertel (JKI Geilweilerhof), L. Kuenzler (Agroscope Wädenswil) for supporting the phenological assessments.

This work was partly funded by the Luxembourgish Ministry of Agriculture, Food and Viticulture in the framework of the research projects “VinoManAOP”, “VinoManAOP2” and “PIWI³”, as well as by funds of the Federal Ministry of Food and Agriculture (BMEL) based on a decision of the parliament of the Federal Republic of Germany via the Federal Office for Agriculture and Food (BLE) under the Federal Programme for Ecological Farming in the framework of the research project “VITIFIT”.