Comparing Selection Criteria to Select Grapevine Clones by Water Use Efficiency

Citation: Mairata, A.; Tortosa, I.; Douthe, C.; Escalona, J.M.; Pou, A.; Medrano, H. Comparing Selection Criteria to Select Grapevine Clones by Water Use Efficiency. Agronomy 2022,12, 1963. https://doi.org/ 10.3390/agronomy12081963 Academic Editor: Saseendran S. Anapalli Received: 15 July 2022 Accepted: 13 August 2022 Published: 19 August 2022 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). agronomy Article Comparing Selection Criteria to Select Grapevine Clones by Water Use Efficiency Andreu Mairata 1,* , Ignacio Tortosa 2, Cyril Douthe 2, JoséMariano Escalona 2, Alicia Pou 1 and Hipólito Medrano 2 1Departamento de Viticultura, Instituto de Ciencias de la Vid y del Vino (Gobierno de La Rioja, Universidad de La Rioja, CSIC), Finca La Grajera, Ctra. De Burgos Km 6, 26007 Logroño, Spain 2INAGEA, Department of Biology, Universitat de les Illes Balears, cta. de Valldemossa Km 7.5, 07122 Palma de Mallorca, Spain *Correspondence: andr[email protected] Abstract: The current climate change is forcing growth-adapted genotypes with a higher water use efficiency (WUE). However, the evaluation of WUE is being made by different direct and indirect parameters such as the instantaneous leaf WUE (WUE i ) and isotopic discrimination of carbon ( δ13 C) content of fruits. In the present work, WUE has been evaluated in these two ways in a wide collection of grapevine genotypes, including Tempranillo and Garnacha clones, and Tempranillo on different rootstocks (T-rootstocks). A total of 70 genotypes have been analysed in four experimental fields over two years. The parameters used to measure WUE were the bunch biomass isotopic discrimination ( δ13 C) and the intrinsic WUE (WUE i ), defined as the ratio between net CO 2 assimilation and stomatal conductance. The genotypes with the highest and lowest WUE were identified, differences between them being found to be of more than 10%. Generally, the two parameters showed coincidences in the clones with the highest and lowest WUE, suggesting that both are valuable tools to classify genotypes by their WUE in grapevine breeding programs. However, δ13 C seemed to be a better indicator for determining WUE because it represents the integration over the synthesis time of the sample analysed (mainly sugars from ripening grapes), which coincides with the driest period for the crop. Moreover, the WUE i is a variable parameter in the plant and it is more dependent on the environmental conditions. The present work suggests that carbon isotopic discrimination could be an interesting parameter for the clonal selection criteria in grapevines by WUE. The main reasons were its better discrimination between clones, the fact that sampling is less time-consuming and easier to do than WUE i , and that the samples can be stored for late determinations, increasing the number of samples that can be analysed. Keywords: vitis; WUE; clonal selection; carbon isotope discrimination; photosynthesis; stomatal conductance; vid; breeding 1. Introduction Grapevine is a traditional Mediterranean crop with a long history that completes its biological cycle during the driest and warmest months of the year. Vine cultivation is mainly located in semi-arid areas with an irrigation water contribution that implies the over-exploitation of available water [ 1 , 2 ]. Furthermore, climate change is causing more frequent and longer droughts and heatwaves combined with increasingly unpredictable torrential rainfall that reduces the actual soil available for the vines [ 3 , 4 ]. These grounds lead to troubling situations of economic and environmental conflict. Spain is the country with the largest viticulture area in the world and is the thirdbiggest wine producer [ 5 ], predominantly in a Mediterranean climate where the irrigated vineyard area was 41.5% in 2021, 0.3% higher than in the previous year [6]. Consequently, current data and future predictions point to the important need to optimise irrigation water use to improve environmental sustainability and the economic balance of the crop. Agronomy 2022,12, 1963. https://doi.org/10.3390/agronomy12081963 https://www.mdpi.com/journal/agronomy Agronomy 2022,12, 1963 2 of 15 Water use efficiency (WUE) in grapevine is a major topic in applied and fundamental research [ 7 ]. The research for drought-adapted cultivars and clones will become an indispensable requirement in semi-arid conditions. Previous work demonstrated the variability of WUE between cultivars and clones [8–13]. The favourable results in classic genetic selection, the existence of a very wide diversity of cultivated grapevine varieties [ 14 , 15 ] and the continuous progress in genomics [ 16 ] offer the genus vitis a wider genetic range to adapt grapevines to situations of increased water stress [ 17 ]. This background, coupled with continuous technological progress, offers the necessary conditions to find more drought-adapted grapevine genotypes. Nowadays, the application of genomic and genetic engineering tools makes it very attractive for grape breeding due to the long time needed with traditional methods [ 18 ]. The utilization of molecular markers can easily identify quantitative trait loci (QTL) that affect traits of interest to accelerate the introduction in host plants using the backcrossing method [ 19 ]. The breeding method is assisted by molecular markers. Genetic engineering could make it possible to obtain new varieties/clones with, for example, high yield, disease resistance, different sugar content, early maturity, or drought tolerance [ 20 ]. Until now, very little progress has been seen in new commercial varieties. In recent decades, the main selection programs developed were focused on clonal selection inside the more commercial varieties because of the legal frameworks of wine protection. Breeding new varieties would require a long administrative process and acceptance by regulatory boards and consumers. In contrast, the selection of clones within an authorised variety was immediately accepted [14]. One of the problems in the selection of genotypes by WUE is how to estimate this parameter. Conceptually, WUE reflects the balance between carbon gains and water loss. This balance can be measured at different levels from leaf instantaneous gas fluxes to plant production [ 21 ]. At the leaf level, the ratios between CO 2 assimilated (A N ) and transpiration (E) or stomatal conductance (g s ) determine the WUE of the plant. “Intrinsic water use efficiency” is determined by factors that the plant can control (A N /g s , WUE i ), less influenced by environmental conditions than the “instantaneous water use efficiency” (AN/E, WUEinst) [22]. These leaf determinations should be taken as representative of the water efficiency over the plant cycle. It is a selection criterion with a clear physiological basis, even though the daily and seasonal measurements of WUE i are “instantaneous”. To overcome these limitations, biomass determination of stable carbon isotope abundance, in particular the 13C ratio, was proposed as a reliable indicator of WUE [23–25]. Photosynthetic processes discriminate between the 12 C and 13 C isotopes due to their different diffusion between the atmosphere and chloroplasts. This discrimination against 13 C ( δ13 C) also occurs in the ribulose biphosphate carboxylase/oxidase (RuBisCo) reaction catalysed by the Rubisco enzyme and is attenuated when the CO 2 concentration in chloroplasts decreases due to stomatal closure. In consequence, the differential proportion of carbon isotopes in plant dry matter results in an integrative estimate of the relationship between photosynthetic rate and stomatal aperture (WUE i ) throughout the synthesis period of the analysed biomass [25]. Tempranillo and Garnacha are among the most widely cultivated varieties in Spain [ 14 ]. In addition, the use of drought-tolerant rootstocks in grapevine helps minimise the effect of water stress through improved water uptake and transport [ 26 ], also controlling plant transpiration through chemical response [27] and hydraulic signalling [28]. Measurements of δ13 C and WUE i have been used in previous work as indicators of WUE in the grapevine [ 12 , 13 , 23 , 29 ]. In this context, the objectives of this work were: (i) Analyse the variability of WUE between clones of the Garnacha and Tempranillo cultivar and Tempranillo on different rootstocks (T-rootstocks), (ii) Evaluate the discrimination capacity of WUE i and 13 C isotopic ratio in two years of field-growing vines data and, (iii) Compare both parameters as operative selection criteria by their interest in grapevine clone breeding. Agronomy 2022,12, 1963 3 of 15 2. Materials and Methods 2.1. Experimental Sites and Plant Material The experiment was carried out in four experimental plots: two located in Logroño (La Rioja, Spain), one in Haro (La Rioja, Spain) and the last one in Miranda de Arga (Navarra, Spain). In total, 58 clones of two cultivars (Garnacha and Tempranillo) and 12 genotypes of rootstocks on Tempranillo (T-rootstocks) were measured over two years (2015 and 2018). Clone groups were randomly distributed in each experimental plot. In each field, leaf gas exchange measurements (WUE i ) were realised in August and berry samples ( δ13 C) were collected at maturity in September and October. The environmental conditions of the climatic stations closest to the experimental fields were described in the two years of study (Table 1). Data were collected and averaged by month from 1 April to 31 October. Growing degree days (in ◦ C day −1 ) were calculated as daily Tmean – Tbase (only positive values, T base = 10 ◦ C), and reference evapotranspiration (ET0) was calculated using the Penman–Monteith method [30,31]. 2.2. Leaf Gas Exchange Measurements Instantaneous leaf gas exchange measurements were done using an open infrared gas analyser system (Li-6400xt; Li-Cor, Inc. Lincoln, NE, USA). Leaf net photosynthesis (A N ) and stomatal conductance (g s ) were measured in a fully exposed mature leaf (one measure per plant and 4–6 plants per clone). The CO 2 concentration reference was 400 µ mol CO2mol−1 air with a flow rate of 350 µ mol (air) min −1 . All the measurements were always taken above the 1500 µ mol m 2 s −1 active photosynthetic radiation (PAR) between 10:00 and 13:00 (local time) using a 6 cm 2 chamber [ 13 ]. Intrinsic water use efficiency (WUE i ) was calculated as the ANand gsratio. 2.3. Carbon Isotope Ratios The carbon isotope ratio ( δ13 C) was determined from samples of 30 berries/plants collected at harvest, at the same plants measured for WUE i (4–6 plants per clone). Berry samples were oven-dried (taking the seed out) and δ13 C was analysed in 2 ± 0.1 mg aliquots of berry powder samples (Thermo Flash EA 1112 Series) [ 23 ]. Determinations of δ13 C were carried out using an Elemental analyser (NC2500, Carlo Erba Reagents) coupled to an isotope ratio mass spectrometer (Thermoquest Delta Plus, ThermoFinnigan). The carbon isotope ratio was expressed as δ13 C = [(R s− R b )/R b ] × 1000 [ 25 ], where R s is the ratio 13 C/ 12 C of the sample. R b is the 13 C/ 12 C of the PDB (PeeDee Belemnite) standard (0.0112372) and was measured every seven samples. 2.4. Statistical Analysis Every cultivar and plot was analysed independently due to their differences in climate, soil, crop management, vine characteristics, etc. Two-way analysis of variance (ANOVA) was used to evaluate the effects of the factors and their interactions on all the variables measured and calculated (Table 2). Then, the WUE i – δ13 C regressions obtained in each group were demonstrated. Seven separated (one per group) one-way ANOVAs were performed to check where the parameters were significant. Distribution and homoscedasticity were analysed using the Shapiro–Wilk test and Levene’s statistic. When differences were found, a post-hoc test (Duncan) was applied to determine which genotypes were different and estimate a ranking [ 32 ]. Data analyses were performed with SPSS 22.0 (IBM Corp., Armonk, NY, USA). Any differences were accepted with a p-value > 0.05. Agronomy 2022,12, 1963 4 of 15 Table 1. Climatic conditions of the three experimental sites. The values are the average of the maximum (Tmax) and minimum temperature (Tmin) and the sum of the cumulative precipitation (P), the reference evapotranspiration (ET0) and the growing degree days (GDD) accumulative by months from April to October in 2015 and 2018 [30,31]. Year T max (◦C) T min (◦C) P (L m−2) ET0 (mm Month−1) GDD (◦C Month−1) Roda April 2015 18.1 6.9 16.2 113.1 72.7 May 21.8 10.0 7.2 144.6 169.7 June 26.6 12.3 63.8 158.6 268.9 July 30.6 15.2 10.7 190.1 370.6 August 28.5 14.1 37.1 167.8 339.3 September 22.1 10.5 26.8 98.8 170.1 October 18.1 8.6 58.8 65.9 96.9 23.7 11.1 220.6 938.9 1488.2 April 2018 17.1 7.6 97.7 97.8 76.0 May 19.9 8.5 51.3 111.4 120.2 June 25.0 12.8 33.3 134.6 249.6 July 28.4 15.6 78.9 156.3 347.7 August 29.6 14.7 0.0 157.2 356.4 September 27.4 13.6 41.4 113.8 287.3 October 19.0 8.3 66.0 65.0 115.3 23.8 11.6 368.6 836.1 1552.5 La Grajera–Vitis Provedo April 2015 18.7 7.3 21.0 105.0 81.7 May 22.6 10.8 2.6 142.5 196.0 June 27.7 14.1 42.8 170.7 307.5 July 31.5 17.0 34.9 197.8 410.1 August 29.0 15.3 19.1 163.9 364.0 September 23.0 11.6 13.3 100.7 204.6 October 18.5 9.1 33.2 62.9 111.4 24.4 12.2 166.9 943.5 1675.3 April 2018 17.3 7.5 86.1 90.6 76.7 May 19.9 9.8 65.6 112.7 138.4 June 25.2 13.9 39.3 137.4 272.0 July 28.9 16.7 117.6 168.7 371.3 August 30.0 16.4 0.0 167.2 386.1 September 27.3 14.5 45.4 113.1 302.0 October 19.5 9.3 28.3 70.7 132.4 24.0 12.6 382.3 860.4 1678.9 Vitis Navarra April 2015 19.8 6.0 11.9 101.1 85.3 May 23.5 10.2 2.4 148.9 206.7 June 28.3 13.4 70.7 165.9 317.7 July 30.9 16.0 0.3 184.2 395.9 August 28.8 14.2 10.3 142.7 350.4 September 23.5 10.4 12.9 95.2 202.6 October 19.9 8.5 24.1 57.0 116.8 24.9 11.3 132.6 895.0 1675.4 April 2018 18.3 6.5 61.3 93.1 80.3 May 21.0 9.0 29.2 119.4 149.3 June 25.8 12.9 48.6 145.8 275.9 July 30.2 15.7 56.3 174.8 386.9 August 29.8 14.9 6.3 159.4 372.4 September 28.0 12.9 16.2 113.3 300.4 October 20.7 8.4 16.6 72.6 141.1 24.8 11.5 234.5 878.4 1706.3 Agronomy 2022,12, 1963 5 of 15 Table 2. g s and intrinsic water use efficiency (WUE i ) average and their standard deviations in Tempranillo, Garnacha and rootstock cultivars in the different fields and years analysed. 2015 2018 gs (mol H2O m−2s−1) WUEi (mmol CO2mol−1H2O) gs (mol H2O m−2s−1) WUEi (mmol CO2mol−1H2O) Tempranillo Roda 0.311 ±0.108 a 57.8 ±15.5 c La Grajera 0.097 ±0.045 c 118.9 ±23.6 a 0.098 ±0.054 c 90.5 ±14.1 a Vitis Provedo 0.198 ±0.072 a 76.7 ±17 b Vitis Navarra 0.164 ±0.063 b 76.5 ±16.5 b Garnacha Vitis Navarra 0.236 ±0.048 b 70.6 ±11.3 b T-Rootstock Vitis Navarra 0.233 ±0.095 b 64.6 ±13.9 bc General 0.126 ±0.069 B 103.2 ±29.1 A 0.236 ±0.104 A 68 ±16.8 B Gs: Two-way ANOVA: Year **, Field ***, Year ×Field ** EUAi: Two-way ANOVA: Year ***, Field ***, Year ×Field *** Different lower case letters indicate a difference between groups of the same year with Duncan’s test (p< 0.05). *** p-value < 0.001; ** p-value < 0.01. 3. Results 3.1. Comparison of Fields and Year Effect The climatic data of the areas of the experimental plots were analysed: three fields located in the region of La Rioja and the other in the region of Navarra, both located in the north of Spain (Table 1). La Grajera and Vitis Provedo fields were located in Logroño (La Rioja), and the third one in Haro (West of La Rioja), near Roda’s winery. Vitis Navarra was located near Larraga, Navarra. All fields were characterised by a Mediterranean climate, with a warm and low rainfall in summer. Roda’s plot is characterised by a less warm summer, with a 10% lower accumulation of growing degrees. Generally, Vitis Navarra had less precipitation and drier climate conditions. In 2018, there was a precipitation increase of 129% in La Grajera and Vitis Provedo, 67% in Roda and 77% in Navarra fields (compared with 2015). Nevertheless, GDD, ET0 and temperatures remained very stable between the two years. Stomatal conductance is determined by plant water status and, at the time, determines WUE [ 33 ]. For this reason, the water status was estimated as g s for all the plots and cultivars (Table 2). The La Grajera farm showed higher water stress (g s < 0.1 mol H 2 O m −2 s −1 ) in both years. Interestingly, the difference in rainfall does not determine the water status of the plants between plots in the same year. Roda’s field (2018) showed a higher stomatal conductance, reaching values close to 0.5 mol H2O m−2s−1. Significant differences in g s and WUE i were observed between fields, cultivars and years. Comparing the average water status between the two years, there was an increase (2018 reached 2015) in stomatal conductance (+87%) and, therefore, a significant decrease in WUEi(−59%). Even though there was a large range of g s in the experimental fields, a good correlation (R 2 : 0.7686) was founded between ln WUE i and g s values in all groups, plots and years analysed (Figure 1). Agronomy 2022,12, 1963 6 of 15 Agronomy 2022, 12, 1963 6 of 15 Figure 1. Linear regression between the natural logarithm intrinsic water use efficiency (WUEi, AN/gs) and stomatal conductance (gs) representing the clonal groupings analysed (LG: La Grajera; VN: Vitis Navarra; VP: Vitis Provedo; R: Roda). 3.2. Genotypic Characterisation of WUE Due to the high variability in water status between plots and years (Table 2), an independent analysis was carried out for each field, year and cultivar. Table 3 shows the average WUEi data of the clones analysed by year, cultivar and plot. Letters represent significant differences between genotypes (p-value < 0.05). A total of seven independent analyses were carried out. Only Tempranillo’s cultivar in Vitis Provedo (2015) showed non-significant differences between clones in WUEi. For the WUEi, a maximum value of 143.1 mmol CO2 mol−1 H2O and a minimum of 40.8 mmol CO2 mol−1 H2O were obtained at 1048 clone (Tempranillo, La Grajera) in 2015 and 137 clone (Tempranillo, Roda) in 2018, respectively. Great variability was observed between clones of the same group, for example, between the genotype 140RU and RG9 (T-rootstocks, 2018), where the percentage increase was 91.55%. The same statistical analysis was performed with the δ13C data (Table 4). Significant differences in 13C content between clones were observed in all groups. The mean values of 13C have a range of 8 ‰, including values between −21.5 ‰ (clone RG8, T-Rootstocks, 2018) and −29.4 ‰ (clone 807, Tempranillo, La Grajera, 2015). Integrating both parameters, the high efficiency of Tempranillo clones 807 was clearly shown (La Grajera, 2015), as well as that of VN32 (Vitis Navarra, 2015), 1048 (La Grajera, 2018) and six (Roda, 2018), that of the Garnacha clone ENTAV 136 (Vitis Navarra, 2018) and the T-rootstock clone RG2 (Vitis Navarra, 2018). In addition, some genotypes stand out for their low WUE, including Tempranillo clones 1084 (La Grajera, 2015) and 137 (Roda, 2018), the Garnacha clone EV15 (Vitis Navarra, 2018) and the T-rootstock clone RG8 (Vitis Navarra, 2018). Figure 1. Linear regression between the natural logarithm intrinsic water use efficiency (WUE i , A N /g s ) and stomatal conductance (g s ) representing the clonal groupings analysed (LG: La Grajera; VN: Vitis Navarra; VP: Vitis Provedo; R: Roda). 3.2. Genotypic Characterisation of WUE Due to the high variability in water status between plots and years (Table 2), an independent analysis was carried out for each field, year and cultivar. Table 3shows the average WUE i data of the clones analysed by year, cultivar and plot. Letters represent significant differences between genotypes (p-value < 0.05). A total of seven independent analyses were carried out. Only Tempranillo’s cultivar in Vitis Provedo (2015) showed non-significant differences between clones in WUE i . For the WUE i , a maximum value of 143.1 mmol CO 2 mol −1 H 2 O and a minimum of 40.8 mmol CO 2 mol −1 H 2 O were obtained at 1048 clone (Tempranillo, La Grajera) in 2015 and 137 clone (Tempranillo, Roda) in 2018, respectively. Great variability was observed between clones of the same group, for example, between the genotype 140RU and RG9 (T-rootstocks, 2018), where the percentage increase was 91.55%. The same statistical analysis was performed with the δ13 C data (Table 4). Significant differences in 13 C content between clones were observed in all groups. The mean values of 13 C have a range of 8 ‰ , including values between − 21.5 ‰ (clone RG8, T-Rootstocks, 2018) and −29.4 ‰ (clone 807, Tempranillo, La Grajera, 2015). Integrating both parameters, the high efficiency of Tempranillo clones 807 was clearly shown (La Grajera, 2015), as well as that of VN32 (Vitis Navarra, 2015), 1048 (La Grajera, 2018) and six (Roda, 2018), that of the Garnacha clone ENTAV 136 (Vitis Navarra, 2018) and the T-rootstock clone RG2 (Vitis Navarra, 2018). In addition, some genotypes stand out for their low WUE, including Tempranillo clones 1084 (La Grajera, 2015) and 137 (Roda, 2018), the Garnacha clone EV15 (Vitis Navarra, 2018) and the T-rootstock clone RG8 (Vitis Navarra, 2018). Agronomy 2022,12, 1963 7 of 15 Table 3. Mean values and standard deviations of WUE i (A N /g s ) in the genotypes and clones studied. 2015 Tempranillo La Grajera Vitis Navarra Vitis Provedo 86 110.7 ±19.8 bcdef VN1 74.3 ±9.5 bRJ43 69.4 ±6.8 n.s. 232 123.7 ±18.8 abcd VN31 77.5 ±8.5 bRJ78 83.3 ±27.4 n.s. 260 131 ±2.6 ab VN32 97 ±6.5 aVP11 65.6 ±9.3 n.s. 280 121.5 ±15.2 abcd VN33 64.9 ±12.7 bVP24 83.9 ±9.8 n.s. 518 132.4 ±9.3 ab VN42 79.6 ±22.2 ab VP25 87.3 ±13.6 n.s. 560 124.7 ±34 abcd VN69 65.6 ±16.3 bVP28 66.7 ±10.2 n.s. 807 129.4 ±26.8 ab VP8 80.9 ±23.6 n.s. 814 136.3 ±10.6 ab 843 129 ±5.5 abc 1041 90.1 ±24.8 ef 1048 143.1 ±8.9 a 1084 81.7 ±18.8 f 1089 96.6 ±25 def RJ26 132.9 ±22.8 ab RJ43 113.1 ±15.4 abcde RJ51 99.3 ±18.5 cdef RJ78 126.5 ±14.5 abcd RJ79 123.1 ±18 abcd 2018 Garnacha T -Rootstock Tempranillo Vitis Navarra Vitis Navarra La Grajera Roda ARA-2 79.2 ±8.2 ab 1103P 74.7 ±5ab 232 93.7 ±14.2 ab 6 79.6 ±5a ARA-24 57 ±4.7 e110R 66.8 ±6.2 abcd 1048 105.1 ±6.4 a108 49.4 ±9c ARA-4 81.6 ±7.9 ab 140RU 41.4 ±2.8 e1052 101.1 ±3.3 ab 137 40.8 ±3.4 c ENTAV 136 88.6 ±5.7 a420A 64.1 ±8.2 bcd 1078 88 ±10 abc 156 47.9 ±4.8 c ENTAV 141 73.5 ±10.9 bcd RG2 72.6 ±8ab 1084 72.8 ±6.4 c166 48.2 ±1.1 c EV11 67.3 ±1.8 cde RG3 55 ±11.8 d1371 86.1 ±14.2 bc 178 77.2 ±16.3 ab EV13 65.5 ±9.4 de RG4 58.1 ±6.1 cd 203 66.4 ±9.3 b EV14 63.2 ±5.6 eRG6 71.6 ±5.4 ab 215 52.6 ±4.6 c EV15 63.1 ±6.7 eRG7 75.2 ±9.9 ab 243 46.2 ±5.8 c RJ21 61.1 ±1.5 eRG8 42.4 ±4.9 e336 66.9 ±7.1 b VNQ 75.7 ±5.1 bc RG9 79.3 ±11.9 a SO4 68.5 ±11.8 abc Different letters indicate statistical differences within each group by Duncan test (p< 0.05). An integrator value was obtained for the genotype in each group by adding the proportional distribution of the relative standard deviation of the values obtained in WUE i and δ13 C. With this method, it was possible to define water efficiency for each clone according to the values obtained for both parameters. The clones were defined as very efficient (residual > 15%), efficient (15 to 5%), normal (5 to − 5%), inefficient ( − 5 to − 15%) or very inefficient (<−15%), depending on the values obtained of the calculated percentages. Of the 70 clones analysed, 14 showed to be very efficient and 21 to be very inefficient. 1048 genotype (Tempranillo) was defined as a very efficient genotype in both years. In the Tempranillo cultivar (largest number of clones analysed), very efficient genotypes in water use efficiency were 814 and 1048 (La Grajera, 2015), VN32 (Vitis Navarra, 2015), VP25 (Vitis Provedo, 2015), 1048 (La Grajera, 2018) and 6, 178, 203 and 336 (Roda, 2018). In contrast, genotypes defined as very inefficient were 1041, 1084, 1089 and RJ51 (La Grajera, 2015), VN33 and VN69 (Vitis Navarra, 2015), RJ43 and VP11 (Vitis Provedo, 2015), 1084 and 1371 (La Grajera, 2018) and 108, 137,156,166 and 243 (Roda, 2018). In the Garnacha cultivar (Vitis Navarra, 2018), the genotype defined as very efficient was ENTAV 136, and the very Agronomy 2022,12, 1963 8 of 15 inefficient ones were ARA-24, EV15 and RJ21. The T-rootstock genotypes (Vitis Navarra, 2018), clones defined as very efficient were 1103P, RG2, RG7 and RG9. In contrast, the very inefficient ones were 140RU, RG3 and RG8. Table 4. Mean values and standard deviations of 13 C isotopic discrimination ( δ13 C ‰ ) in the genotypes and clones studied. 2015 Tempranillo La Grajera Vitis Navarra Vitis Provedo 86 −22 ±0.8 abc VN1 −27.1 ±0.7 bc RJ43 −26.3 ±0.4 a 232 −21.7 ±0.2 aVN31 −27.8 ±0.4 cRJ78 −25.9 ±0.5 a 260 −22.4 ±0.4 abcd VN32 −25.3 ±0.7 aVP11 −26.6 ±0.4 a 280 −22.2 ±0.2 abcd VN33 −27.4 ±0.5 bc VP24 −24.4 ±0.6 c 518 −22.9 ±0.5 bcdef VN42 −27.3 ±0.8 bc VP25 −24.3 ±0.4 c 560 −21.8 ±0.9 ab VN69 −26.6 ±0.6 bVP28 −24 ±0.3 c 807 −21.5 ±0.2 aVP8 −25.2 ±0.6 b 814 −22.6 ±0.6 abcde 843 −22.4 ±0.4 abcd 1041 −24.8 ±0.5 gh 1048 −23 ±0.5 cdef 1084 −25.8 ±1.8 h 1089 −24.1 ±0.5 fg RJ26 −23.9 ±0.7 fg RJ43 −24 ±0.3 fg RJ51 −23.9 ±0.7 fg RJ78 −23.6 ±0.4 ef RJ79 −23.2 ±0.5 def 2018 Garnacha T -Rootstock Tempranillo Vitis Navarra Vitis Navarra La Grajera Roda ARA-2 −24.8 ±0.7 bc 1103P −26.3 ±0.3 b232 −22.7 ±0.8 a6−24.9 ±0.8 ab ARA-24 −24.8 ±0.1 bc 110R −28 ±0.4 ef 1048 −23.2 ±0.1 a108 −26.6 ±0.3 de ARA-4 −24.8 ±0.4 bc 140RU −28.5 ±0.5 f1052 −23.2 ±0.8 a137 −27.1 ±0.9 e ENTAV 136 −24.6 ±0.3 abc 420A −27.7 ±0.3 de 1078 −23.9 ±0.7 a156 −26.6 ±0.4 de ENTAV 141 −27 ±0.6 eRG2 −25.4 ±0.7 a1084 −24.1 ±0.6 a166 −26.2 ±0.2 cde EV11 −24 ±0.7 aRG3 −27.4 ±0.4 cde 1371 −25.8 ±2.3 b178 −24.9 ±0.7 ab EV13 −24.2 ±0.2 ab RG4 −27.1 ±0.9 bcd 203 −24.5 ±0.5 a EV14 −24.2 ±0.6 ab RG6 −26.7 ±0.4 bc 215 −26 ±0.8 cd EV15 −25.8 ±0.6 de RG7 −26.8 ±0.4 bc 243 −25.6 ±0.6 bc RJ21 −24.9 ±0.2 bc RG8 −29.4 ±0.3 g336 −24.7 ±0.9 ab VNQ −25.2 ±0.3 cd RG9 −26.9 ±0.8 bcd SO4 −26.4 ±0.7 b Different letters indicate statistical differences within each group by Duncan test (p< 0.05). 3.3. Ability of δ13C and Leaf Gas Exchange Values to Measure WUE The relationship between values of the two estimates of the WUE was analysed in Figure 2which represents the relationship between δ13 C and WUE i values among the seven groups of genotypes analysed. A significant relationship (Pearson correlation of 0.699) was observed between both parameters (p-value < 0.05). Only in the Garnacha cultivar of Vitis Navarra (2018) was this relationship insignificant. Agronomy 2022,12, 1963 9 of 15 Agronomy 2022, 12, 1963 9 of 15 0.699) was observed between both parameters (p-value < 0.05). Only in the Garnacha cultivar of Vitis Navarra (2018) was this relationship insignificant. Figure 2. Linear regression between carbon isotopic discrimination (13C) and WUEi of the data set grouped in the clone sets analysed (LG: La Grajera; VN: Vitis Navarra; VP: Vitis Provedo; R: Roda). ** p-value < 0.01; * p-value < 0.05; n.s: not significant. A good correlation was also shown for the values of the residual percentages between the δ13C and WUEi (Figure 3). The WUEi percentages have a larger range of oscillation, reaching 37.5% compared to 11.4% for δ13C. Of the 70 genotypes analysed, 17 did not match the trend of the mean values of the residual percentages. Interestingly, 6 of the 11 clones of the Garnacha cultivar (Vitis Navarra, 2018) did not follow the general trend. However, there was a great general relationship (Pearson correlation of 0.585; p-value < 0.01) between the residual percentages of the parameters. Once the genotypes were ranked by WUEi or δ13C, their relative position was quite coincident for both parameters. As shown in (Figure 4), each genotype value was correlated according to its value in both parameters (WUEi and δ13C). Interestingly, 83% of the genotypes were positioned in close to three positions in both parameters and 71% of genotypes in less than two positions. In general, there was a good correlation in the relative position of genotypes classified as best and worst according to the WUE. Figure 2. Linear regression between carbon isotopic discrimination ( 13 C) and WUE i of the data set grouped in the clone sets analysed (LG: La Grajera; VN: Vitis Navarra; VP: Vitis Provedo; R: Roda). ** p-value < 0.01; * p-value < 0.05; n.s: not significant. A good correlation was also shown for the values of the residual percentages between the δ13 C and WUE i (Figure 3). The WUE i percentages have a larger range of oscillation, reaching 37.5% compared to 11.4% for δ13 C. Of the 70 genotypes analysed, 17 did not match the trend of the mean values of the residual percentages. Interestingly, 6 of the 11 clones of the Garnacha cultivar (Vitis Navarra, 2018) did not follow the general trend. However, there was a great general relationship (Pearson correlation of 0.585; p-value < 0.01) between the residual percentages of the parameters. Once the genotypes were ranked by WUE i or δ13 C, their relative position was quite coincident for both parameters. As shown in (Figure 4), each genotype value was correlated according to its value in both parameters (WUE i and δ13 C). Interestingly, 83% of the genotypes were positioned in close to three positions in both parameters and 71% of genotypes in less than two positions. In general, there was a good correlation in the relative position of genotypes classified as best and worst according to the WUE.