Weather conditions and phenology
Crop forcing technique succeeded in delaying the berry ripening period to a cooler environment as it had been reported in previous works (Gu et al., 2012; Martínez-Moreno et al., 2019; Martinez De Toda et al., 2019; Oliver-Manera et al., 2022). The accumulated GDD from budburst to veraison was a robust tool to predict veraison date for forced treatments, observation reported previously in fully irrigated vines (Prats-Llinàs et al., 2020; Oliver-Manera et al., 2022) but contrary to Martínez-Moreno et al. (2019) which reported significantly more GDD required to reach veraison in forced treatments. This extra GDD requirement was attributed to a competition between the attempt to refill carbohydrate reserves after forced budburst and vine development for resources. Those contradictory results may come from the vine water status. In Martínez-Moreno et al. (2019), they observed pre-veraison Ψs lower than − 1 MPa in forced vines. Therefore, a good vine water status soon after forcing (lower than − 0.65 MPa) may benefit both, vine development and the carbohydrate reserves refilling, minimizing the effect of forcing on the required GDD for vine development.
Water consumption and plant water status
The crop coefficients (Kc) used on these study, and therefore the calculated potential ETc, matched perfectly with those reported using weighing lysimeters with Tempranillo cultivar for a FIPAR of 30% (Picón-Toro et al., 2012). Although, for the whole growing season (from budburst to leaf fall) the potential ETc was higher and rainfall lower for unforced vines than for forced vines (Table 2), after three years of experiment, the irrigation strategies we applied in forced treatments did not save water compared to the unforced treatment (Table 2). In C-RDI treatment, water deficit strategy was applied from veraison (DOY 213 on average) to harvest (DOY 256 on average) and was based on the criterion of maintaining Ψs above the threshold of -1.2 MPa proposed by Girona et al. (2006). Although from veraison to harvest, C-RDI vines are under highly water demanding environment (30% of the total ETc) and practically without rainfall (Fig. 1A, C and E and Table 2), to accomplish with the Ψs threshold criteria hardly required water supplied by irrigation (Table 2 and Fig. 2). In addition, before veraison, 32% of the ETc was compensated by rainfall in C-RDI (Table 2), which, together with the water available in the soil from previous autumn and winter rainfall, allowed to the C-RDI treatment to compensate for only 49% of the ETc by irrigation without dropping Ψs below − 0.6 MPa (Fig. 2) threshold assigned for non-stressed vines (Girona et al., 2006). As a result, although some post-harvest irrigation was required to recover vine water status to Ψs above the threshold of − 0.85 MPa for non-stressed vines (Girona et al., 2006), C-RDI supplied only 38.2% of the seasonal potential ETc by irrigation (Table 2).
On the other hand, for forced vines, the period from forced buds budburst (DOY 182 on average) to veraison (DOY 263 on average) in which no deficit irrigation was applied, coincided with the most water demanding period (representing 50% of the whole year ETc), with low compensation by rainfall (only 10% of the ETc) (Table 2). Therefore, in forced treatments, irrigation water was required to compensate almost 90% of the ETc from forced budburst to veraison which represented the 50% of the total ETc. In addition, in forced treatments, the pre-forcing water stress strategy could have saved a maximum of 35% of the total ETc (Table 2) but spring rainfall compensated 55% of the ETc, reducing the required water provided by irrigation. Only in 2019, after a dry spring (only 30% of Etc was compensated by rainfall), treatments DI and DI + RDI saved water compared to C-RDI (Fig. 1D) at the end of the season. In fact, 2019 was the only season in which Ψs, which is a very robust indicator to discriminate different vine water status independently of the environment (Santesteban et al., 2019), was clearly affected by the irrigation regime in DI and DI + RDI compared to C-RDI and RDI treatments (Fig. 2). However, when analysing pre-forcing IntΨ, which is an appropriate tool to indicate the cumulative stress (Myers, 1988), it was evident that DI and DI + RDI treatments were more stressed than RDI (Table 3).
On the other hand, the post-veraison period in forced treatments represented only 13% of the total ETc and is likely to be compensated by rainfall in autumn (Table 2). Delayed veraison by an average of 53 days (Table 1), moved the ripening period to environmental conditions and daylength that drives to a low daily evaporative demand. As a result, we did not observe any sign of water stress after veraison in any forced treatment neither by analysing the Ψs (Fig. 2) nor the IntΨ (Table 3). The difficulty of applying post-veraison water stress in forced vines was previously reported by Martínez-Moreno et al. (2019) but neither the amount of water nor the exact irrigation strategy were reported.
It should be considered that annual rainfall in 2019 (331 mm) was close to the regional mean (340 mm) but not in 2018 (485 mm) and 2020 (501 mm). Therefore, in terms of saving water, DI and DI + RDI strategies showed promising results but not the RDI strategy.
Vine performance and grape and wine quality responses to forcing and irrigation strategy
To study plant responses to irrigation treatments, multi-year studies are recommended (Intrigliolo and Castel, 2010). As a perennial crop, grapevine responses to water stress may be extended to the subsequent year. The same idea is valid for defoliation experiments. Reducing photosynthetic capacity one year may affect the next season in form of reduced carbohydrate reserves (Bennett et al., 2005). When using the crop forcing technique, the physiological response to irrigation is even more complex than in a classical irrigation or defoliation experiments. The effects on forced vines performance may come from the period before forcing or even from the previous year (Oliver-Manera et al., 2022). The vine response to the crop forcing technique is a reduction of yield due to reduced bunch weight, the number of berries per bunch and the individual berry weight (Gu et al., 2012; Martínez-Moreno et al., 2019; Martinez De Toda et al., 2019; Oliver-Manera et al., 2022) observations which are in accordance to our results (Table 4). The yield reduction has been attributed to the low carbon availability when vines are forced, the date of forcing, an impaired seasonal carbon balance and to the temperature distribution of forced vines (very elevated at first stages of development and low after veraison) altering the duration of the developmental stages (Gu et al., 2012; Martínez-Moreno et al., 2019; Martinez De Toda et al., 2019; Oliver-Manera et al., 2022). Our results of reduced FIPAR, which is positively correlated to the whole vine photosynthesis (Poni et al., 2003), in forced treatments compared to C-RDI (Fig. 3) confirms the limitation on the photosynthetic capacity in forced vines. However, it is remarkable that in our experiment, yield in forced vines increased year after year (Table 5) suggesting a vine acclimatation to be forced. This observation is consistent with the capacity to restore carbohydrate reserves after vines are forced attributed to high source:sink ratio since forced vines give priority to vegetative growth to yield (Oliver-Manera et al., 2022) and to the tendency of the yield reduction to remain constant after two years of forcing (Martínez-Moreno et al., 2019).
When comparing different irrigation strategies of forced treatment, we observed a reduction in yield, but only caused by a reduction in the number of bunches per vine in those treatments, in which irrigation was withheld before forcing (Table 5). We did not observed a reduction of berry fresh weight, which is very sensitive to post-veraison water stress (Girona et al., 2009) in any treatment, which is consistent with the lack of post-veraison water stress (Fig. 2 and Table 3). However, the main limitation for berry post-veraison growth may be low temperature (Martinez De Toda et al., 2019) rather than water stress.
Despite both irrigation management and defoliation being cited as interesting techniques to be studied in grapevine management to face the global warming (Gutiérrez-Gamboa et al., 2021; Palliotti et al., 2014) there is still a huge field for research when combining both techniques. Any reduction of bunches per vine was observed when combining water deficit with leaf removal with Tempranillo (Buesa et al., 2019) neither Merlot cultivar (Yu et al., 2016). On the other hand, after five years of experiment, Williams (2012) observed lower bunches per vine in both, vines under water stress and defoliated vines but no interaction was observed. Ferlito et al. (2014) observed a reduction in yield when combined defoliation and water stress but it was dependant on the cultivar. However, early water stress applied about full bloom (Guilpart et al., 2014; Matthews and Anderson, 1989; Santesteban et al., 2011) as well as early reduction of the whole canopy functional leaf area (about full bloom and flowering) (Gatti et al., 2016; Risco et al., 2014) can result in a reduction of the number of bunches per vine the subsequent year or even the abortion of inflorescences to the same year in late pruning technique (Gatti et al., 2016). In a crop forcing study, the number of bunches per vine increased when applied irrigation increased from 60% of the pre-veraison ETc to 80% of the seasonal ETc the next year but only one year of each irrigation regime was reported (Gu et al., 2012). The inflorescence induction in grapevines it is widely accepted that occurs from budburst until before the bud enters dormancy the year before berry formation (Boss et al., 2003; Mullins et al., 1992). Therefore, in forced vines, inflorescence primordia which would become the forced crop, are formed from unforced budburst (between DOY 85 and 94 in this study) to few days after forcing pruning (at least before the time when forced inflorescences appeared). This observation is supported by the fact that, with Tempranillo cultivar, the later the vines are forced (always before the bud enters dormancy) the higher the number of forced bunches (Martínez-Moreno et al., 2019; Martinez De Toda et al., 2019; Oliver-Manera et al., 2022). In our experiment, the irrigation treatment in DI and DI + RDI (Fig. 1B, D and G) coincides with the induction of the inflorescence primordia that will become the forced crop after the forcing operation. The good negative correlation observed between the integral of stress prior to forcing and the number of bunches per vine (Fig. 4b) suggests the high sensitivity to inflorescence formation to water stress which is higher as the soil is drier (Guilpart et al., 2014). This is the case of DI + RDI treatment since less proportion of ETc was compensated through irrigation the previous year (Table 2). In fact, the first Ψs measured after forced buds were unlocked (DOY 207 and 212 for 2018 and 2019) tended to be lower (12% averaging 2018 and 2019) and the IntΨ is more negative between forcing and veraison in those treatments with pre-forcing water stress (Fig. 2) unless winter and spring rainfall compensate the pre-forcing vine water demand as in the case of 2020 (Fig. 1E and Table 2). Since Ψs is a good indicator of plant water status and is well correlated with soil water status (Intrigliolo and Castel, 2006), this suggests some recovery time required to refill the soil in DI and DI + RDI treatments likely caused by the impact of forcing pruning in the hydraulic vine system. It is accepted that cutting shoots when are under tension causes xylem embolism (Wheeler et al., 2013): Because the forcing pruning were performed under high evaporative demand and low soil water content for DI and DI + RDI treatments it is likely that the hydraulic system for both treatments needed more time to recover from the pruning. Dysfunctionality of the hydraulic system can reduce vine growth and photosynthesis (Lovisolo et al., 2010) which is consistent with our observation of slightly higher FIPAR in RDI treatment (Fig. 3).
On the other hand, it is well known that the number of bunches per vine is correlated to the carbohydrate reserves at budburst (Bennett et al., 2005; Lebon et al., 2008) which are dependent on the canopy photosynthetic capacity and the carbon utilization the year before (Holzapfel et al., 2010; Williams, 1996). Before the time vines were forced, we only observed a significant reduction in FIPAR in DI and DI-RDI, which is consistent with the lower Ψs for the same treatments (Fig. 2) compared to RDI in 2019 (Fig. 3). In addition, the averaged Ψs we observed in treatments DI and DI + RDI before forcing was about − 0.54 MPa whereas the averaged minimum Ψs was − 0.66 MPa for both treatments, which is not considered to drastically reduce leaf net photosynthesis in Tempranillo (Intrigliolo and Castel, 2011). Moreover, although soil water content is a main factor affecting the use of carbohydrate reserves (Smith and Holzapfel, 2009), vegetative growth is more sensitive to water stress than photosynthesis (Hsiao, 1973), we that mild water stress may would have resulted in an increase of carbohydrate reserves at forced budburst in DI and DI + RDI treatments, which is contrary to our results. Therefore, our results suggests that inflorescence formation is more sensitive to mild water stress than photosynthesis as previously reported (Guilpart et al., 2014). However, our results do not allow to rule out reduced carbon availability in the reduction of bunches per vine we observed in DI + RDI treatment.
The must acidity and malic acid concentration were increased, and the pH reduced by the forcing operation with level of TSS within the range (22.5–23.5°Brix) previously determined although lower than the C-RDI (Table 6). Therefore, the TSS:TA ratio was enhanced due to the forcing. All of them were goals of the crop forcing technique and are in accordance with results previously reported with Tempranillo (Lavado et al., 2019; Martínez-Moreno et al., 2019; Martínez De Toda, 2021). The increased must acidity and the lower pH in forced treatments has been attributed to the lower temperature in which berries ripen because malic acid degradation as a respiratory substrate is dependant of the temperature (Martinez De Toda et al., 2019). The irrigation treatment did not alter must quality which allow to go on researching new irrigation strategies based on our results. However, although wine acidity was increased, alcohol content decreased, the effect on colour attributes and TPI differed depending on the year (Table 7) and clearly failed in 2020 which could be related to the mildew affection in C-RDI which could increase the source sink relationship increasing the anthocyanin biosynthesis (Mirás-Avalos et al., 2017), closely related to wine colour in red varieties. However, DI treatment succeeded in increasing TPI in season 2019 compared to forced vines under water stress and unforced vines. The latter observation is consistent with previous works which observed higher polyphenol content in well irrigated forced than unforced vines attributed to a better coupling of sugar and anthocyanins accumulation due to more suitable temperature regime after veraison in forced vines (Gu et al., 2012; Lavado et al., 2019). Regarding the effect of water stress, Buesa et al. (2019) observed that rainfed and defoliated vines soon before veraison reduced TPI compared to undefoliated and irrigated vines, which was attributed to a reduced carbohydrate availability after veraison. In forced vines, carbohydrate uptake at a whole vine level is dramatically reduced even for fully irrigated vines (Oliver-Manera et al., 2022), therefore, since 2019 was the only year in which post-veraison Ψs for RDI and DI + RDI treatments dropped below Ψs -1.0 MPa which can limit leaf photosynthesis (Intrigliolo and Castel, 2011), reduced carbohydrate availability may have affected phenolic biosynthesis.
In a climate change scenario in which water demand by plants and aridity are expected to increase, those techniques which increases yield reducing water consumption becomes a priority (Flexas et al., 2010). This is not the case of the crop forcing technique and the irrigation strategies studied in our experiment. Due to more water consumption and reduced yield in forced than unforced vines, WUE was dramatically reduced in forced vines regarding the irrigation strategy (Table 4 and Table 5). Mainly, there are two ways to increase WUE in forced vines: i) yield must be increased maintaining or reducing water supplied or, ii) water consumption should be reduced maintaining or increasing yield. Early deficit irrigation (from full bloom to veraison) reduces berry size and vine vigour (Intrigliolo and Castel, 2010) which are also responses to the crop forcing technique. Therefore, in our opinion, early water deficit strategy should be avoided in forced vines. However, based on our results and the apparent positive yield response to the increase of water supply suggested by Gu et al. (2012), we suggest that sustained deficit irrigation (a controlled reduction of the ETc) would be a good irrigation strategy to increase WUE. In addition, in this study we used previously defined Ψs thresholds determined under environmental in which vines are adapted to grow. However, it is well known that Ψs thresholds are affected by phenology but also by the vapour deficit pressure (Olivo et al., 2009). Therefore, sustained deficit irrigation strategies would be interesting matter of research to define physiological irrigation management thresholds in forced vines. Interestingly, a recent variation of the crop forcing called double cropping in which primary crop and the leaves from the first six nodes were not removed, overcame the yield reduction associated to forced vines (Poni et al., 2021). Therefore, research in irrigation strategies and water stress responses using this new technique would be of interest to enhance WUE.