Scion × Rootstock Response on Production, Mineral Composition and Fruit Quality under Heavy-Calcareous Soil and Hot Climate

: Since there are many di ﬀ erent factors which a ﬀ ect peach orchard proﬁtability, it is necessary to conduct long-term studies to ﬁnd the best scion / rootstock combination within the constraints imposed by local climate and economic conditions. The inﬂuence of six peach-almond hybrids (“Adafuel”, “Adarcias”, Garnem”, “GF 677”, “PADAC 9902-01” and “Rootpac 70”); one P. persica × P. davidiana hybrid: “Cadaman”; ﬁve hexaploid plums (“Adesoto 101”, “PM 44 AD”, “PM 105 AD”, “PM 150 AD”, and “Penta”); three plum × peach-almond hybrids (“PADAC 04-01”, “PADAC 04-03” and “PADAC 99-05”), and one plum x almond hybrid (“Rootpac R”) on agronomic, basic fruit quality parameters, and ﬂower and leaf mineral nutrients of the “Big Top” nectarine was tested. The highest vigor was induced by “Adafuel”, while the lowest vigor was exhibited by “PM 105 AD”, although the latter did not di ﬀ er from the rest of the rootstocks. Yield e ﬃ ciency was higher for “Adesoto 101” and lower for “Adafuel”. Regarding the soluble solids content (SSC), “PADAC 04-01” rootstock induced the highest average content, although it did not di ﬀ er from “Adarcias”, “Adesoto 101”, “PADAC 99-05”, “PM 105 AD”, and “Rootpac R”. In addition, signiﬁcant e ﬀ ect of rootstock was found on the ﬂower and leaf mineral analysis traits evaluated. The medium (“Adarcias”, “Adesoto 101”, “PM 105 AD”) to high vigorous plum based hybrid rootstocks (“PADAC 04-01”, “PADAC 99-05” and “Rootpac R”) showed the tendency to induce a better fruit quality, based on higher concentration of SSC, and other yield attributes, demonstrating their commercial interest as new rootstocks for peaches. Considering its overall performance, “PADAC 04-01” appears as a new promising rootstock alternative to peach-almond hybrids on heavy and calcareous textured soils.


Introduction
In fruit production worldwide, cultivar is a key factor determining productivity, fruit quality and economic value [1]. The continuously release of new peach (Prunus persica) cultivars by breeders from around the world, together with the consistent improvement on orchard management strategies by growers and the commercial interest of peach production for fruit fresh consumption, allowed Spain to be the first world peach exporter and currently the largest producer in the European Union [2,3]. This commodity is mainly produced in the Ebro Valley (Mediterranean basin), where most of the  [29] "GF 677" P. amygdalus × P. persica Vigorous Open-pollinated INRA, France HT S Bernhard and Grasselly [30] "PADAC 04-01" P. cerasifera × (P. amygdalus × P. persica) Vigorous Controlled cross CSIC, Spain HT HT Moreno [9] "PADAC 04-03" P. cerasifera × (P. amygdalus × P. persica) Vigorous Controlled cross CSIC, Spain HT HT Moreno [9] "PADAC 99-05" P. cerasifera × (P. amygdalus × P. persica) Vigorous Controlled cross CSIC, Spain HT HT Moreno [9] "PADAC 9902-01" (P. amygdalus × P. persica) × (P. persica × P. davidiana) Vigorous Controlled cross CSIC, Spain HT HT Moreno [9] "Penta" P. domestica Semi-vigorous Open-pollinated CRF, Italy T T Nicotra and Moser [31] "PM 44 AD" P. insititia Semi-vigorous Open-pollinated CSIC, Spain T T Moreno [9] "PM 105 AD" P. insititia Semi-vigorous Open-pollinated CSIC, Spain T T Moreno [9] "PM 150 AD" P. insititia Semi-vigorous Open-pollinated CSIC, Spain T T Moreno [9] "Rootpac ® 70" P. persica × (P. amygdalus × P. persica) Vigorous Controlled cross AI, Spain HT MT Jiménez et al. [20] "Rootpac ® R" P. cerasifera × P. amygdalus Vigorous Controlled cross AI, Spain HT HT Pinochet [32] a AI = Agromillora Iberia S.L. private nursery, Monistrol d'Anoia; CITA = Centro de Investigación y Tecnología Agroalimentaria de Aragón, Zaragoza; CRF = Centro di Ricerca per la Frutticoltura, Rome; CSIC = Consejo Superior de Investigaciones Científicas, Zaragoza; INRA = Institut National de la Recherche Agronomique, Villenave d'Ornon Cedex. b Information obtained from different authors [9,16,17,33,34]. Abbreviations: HT, highly tolerant; T, tolerant; S, susceptible; MT, moderately tolerant.  [35]. Similarly, the field plot was classified as having a silty-clay-sandy texture ( Table 2) following the USDA soil textural classification system. In this area (cold-semiarid Mediterranean climate), the average annual precipitation amounts to 345 mm, the diurnal temperature variation amounts to 13.3 • C, and the reference evapotranspiration ranges from 997 mm to 1140 mm [35]. Cultural management practices such as fertilization, winter pruning, and spring thinning, were conducted as in a commercial orchard. Hand-thinning was carried out at 40 days after full bloom (DAFB) to maintain a minimum spacing between fruits (approximately 20 cm). Foliar pesticides and insecticides were applied as necessary, following industry standards. The experiment was established in a randomized block design with five replicates for each rootstock. Guard rows were used to avoid edge effects. Trial was established in the field during the winter of 2008-2009. Trees were trained to a low density open-vase system (5 m × 4 m), a very common planting system 10 years ago in peach orchards, when the trial was set up. The trial was level-basin irrigated with a dose of 1000 m 3 ha −1 every twelve days during the summer, for the first three years until 2012. Later, drip irrigation was established to follow standard irrigation management of peach orchards. The criteria were following modern procedures by fruit growers and having new irrigation facilities recently stablished at the experimental field we adopted this new procedure, with a maximum dose of 250 m 3 ha −1 week −1 during the months with higher water requirements (June and July).

Weather Conditions
Meteorological data from 2013-2017: the average daily temperature (Tavr, • C), the average minimum daily temperature (Tmin, • C), and the average maximum daily temperature (Tmax, • C) as well as rainfall (mm) and solar radiation (Sr, W m −2 ) from blooming (first fortnight of March) until harvest date (first fortnight of July), which covers each fruit growing period, were downloaded from the meteorological automated station located very close (≈100 m) to the field trial.

Tree Mortality and Field Determinations
Tree health and survival were monitored throughout the trial. Dead trees were recorded each year, from 2010 until 2017, when growth measurements were taken in December and tree mortality was expressed as a percentage. It should be noted that for rootstock "PM 150 AD", less than three replicates were available in the last year of study (2017). Consequently, it was excluded of the statistical analysis.
The trunk cross-sectional area (TCSA, cm 2 ) and yield (kg/tree) were determined from 2010 until 2017. Trunk circumferences were measured during the dormant season (December) at 20 cm above the graft union. At harvest, all fruits from each tree were weighed to determine total yield per tree (kg/tree). Cumulative yield and cumulative yield efficiency (cumulative yield in kilograms per tree per final TCSA) of each scion/rootstock combination were computed in 2017.

Fruit Quality Determinations
During harvest, 20 mature fruits (flesh firmness in the range of 40-50 N and fruit color ≥80% of fruit surface) of each tree were randomly selected each season for the period of 2013 to 2017. Sampled fruits were randomly selected at commercial harvest by a single person to keep consistency of maturity grade. Fruits were individually weighted and average fruit weight (AFW) was determined. Soluble solids content (SSC) of fruit juice was measured with a digital refractometer (Atago PR-101, Tokyo, Japan) and was expressed as ºBrix. To analyze titratable acidity (TA), 5 g of homogenized samples were diluted with 50 g of distilled water and microtitrated with 0.1 N NaOH. The results were expressed as g malic acid/100 g FW. Ripening index (RI) was calculated based on the SSC/TA ratio. Flesh firmness (Newtons) was measured using a penetrometer (Model FT-327) on two paired sides of each fruit, by removing one mm thick disk of skin. From 2013 to 2016, color parameters were determined and values of L* (brightness or lightness), a* (−a*, greenness, +a*, redness), b* (−b*, blueness, +b*, yellowness), C* (chroma) and h (hue angle) were measured using a colorimeter (Chroma Meter, CR-400 Konica Minolta) [39].

Flower and Leaf Mineral Analysis
Flower and leaf mineral element concentrations were determined in 2017, i.e., in year nine after planting, using healthy trees for sampling. One hundred whole fully opened flowers per tree (including petals, sepals, reproductive parts, bracts and peduncles) were taken at full bloom (in early March). Flowers were sampled from the central part of the shoots around the tree crown. These fresh flowers were then put into an air forced stove (60 • C) and 24 h later ground to a fine powder (in a conventional coffee mill). Leaf sampling was carried out at 120 days after blooming (DAFB), as described in other Prunus rootstocks studies [33,40]. It was approximately the harvest time period of "Big Top" nectarine in our climate conditions.
The mineral element composition of the dried tissue was determined using the method of C.I.I. et al. [41], as previously reported by Mestre et al. [33]. Total N was determined by Dumas method (LECO Inc. FP-528, Madrid, Spain); the macroelement P was analyzed spectrophotometrically by the phospho-vanadate colorimetric method (ThermoSpectronic Helios β); and the macroelements Ca, K, Mg, and the microelements Cu, Fe, Mn and Zn by atomic emission spectroscopy (ICP, Horiba-Jobin Yvon, Active-M, Longjumeau, France).
The DOP index (deviation from optimum percentage) on leaves was estimated for the diagnosis of the nutritive status of the trees [42]. The DOP index was calculated from the leaf analysis by the following mathematical expression: where C is the nutrient concentration in the sample to be studied and C ref is the nutrient concentration considered as optimum, both values given on a dry matter basis. The C ref has been taken from optimum values proposed by different authors [43,44]. The ΣDOP is obtained by adding the values of DOP indices irrespective of sign. The larger was the ΣDOP the greater was the intensity of imbalances among nutrients.

Data Analysis
All variables evaluated followed normal distribution based on Shapiro-Wilk W Test. Data were subjected to a two-way ANOVA analysis in order to examine year (Y), rootstock (R), and Y × R interaction. The total variability of each parameter was estimated using the total sum squares of two-way ANOVA results. Additionally, the variability expressed as percentage of the total sum of squares for year, rootstock, and the interaction between both was calculated. When the F test was significant, means were separated by the HSD Tukey test (p ≤ 0.05). Pearson correlation analyses were Agronomy 2020, 10, 1159 6 of 22 performed to study correlations among traits evaluated. The hierarchical agglomerative cluster was done to represent similarity between rootstocks and variables evaluated. All statistics analysis was designed using JMP (Version 14.2; SAS Institute Inc., Cary, NC, USA).

Weather Conditions
The average daily temperature (Tavr) and the average maximum daily temperature (Tmax) were higher in the fruit growing season of 2017, especially when compared to 2013 and 2016 seasons (Table 3). Moreover, year 2017 had less rainfall accumulated compared to the rest of the four years. Years 2014 and 2015 also stood out for having high Tavr and Tmax, whereas 2013 for being the year with more rain accumulated during the fruit growing period. Over the five years of weather data presented in this study, a general trend of increasing temperatures is observed, except for 2016. Nowadays, rising temperatures and extreme weather events are increasingly affecting the Mediterranean basin, an area classified as extremely vulnerable to global warming [5]. High temperatures together with excessive radiation negatively affect peach production by compromising key developmental stages of the fruit as well as affecting the final fruit quality and storability.

Tree Mortality
At the ninth year after budding, growing conditions generated varying levels of tree mortality, the highest (100%) with the peach-almond hybrid "Garnem", and the hexaploid plums "PM 44 AD" and "Penta" (Figure 1). Consequently, they were excluded of the trial due to their high tree mortality experienced during the first years after planting. In contrast, the peach-almond hybrids "Adarcias" and "Rootpac 70", as well as the interspecific plum based rootstocks "PADAC 04-01" and "Rootpac R" survived well. Mestre et al. [33] also reported the poor and good adaptation of "Garnem" and "Rootpac R" respectively, growing under root asphyxia and heavy soil conditions. The peach-almond hybrids "GF 677" and "PADAC 9902-01", the hexaploid plums "PM 105 AD" and "PM 150 AD", and the peach-plum hybrid "PADAC 99-05", experienced also a high tree mortality mostly during the first growing years (40% of dead trees), in contrast with results previously reported for "GF 677" [40,45]. Nevertheless, "Adesoto 101" experienced a 20% mortality rate, in agreement with results obtained by other authors [20,33]. In this trial, tree mortality could be attributed to the sensitivity to root asphyxia caused by waterlogging conditions [33] or susceptibility to pathogens as Phytophtora spp. [40,46] or other pathogens [47]. Short-waterlogging conditions are very common in clay-loam textured soils [48], the most frequent soils type in the Ebro Valley and the Murcia region, the largest producer areas in Spain. Since this country is the first peach producer in Europe, that type of soils is very common in the most important peach growing areas. Our field plot was classified as having a silty-clay-sandy texture. This texture can induce waterlogging problems and root asphyxia to sensitive rootstocks as peach seedlings and peach-almond hybrids [49]. Moreover, heavy textured soils facilitate flooding with the occurrence of intense rains, commonly concentrated in very short periods of the summer in the Mediterranean area [48]. Typical symptoms observed included an initial leaf wilting and defoliation followed by dead of twigs, and further of larger branches, and finally tree declined and died. Ziegler et al. [49] suggested that in fruit trees (especially in peach, as one of the most susceptible Prunus species to root anoxia in flooded soils) flooding stress can affect significantly plant survival and growth. Tolerance to root anoxia is determined by rootstock characteristics, and the differences in the responses to flooding should be considered in production settings where it is not possible to properly control irrigation to prevent short periods of flooding or in soils that do not drain irrigation or rain water quickly [48,49].

Production and Fruit Quality
As depicted in Table 4, the analysis of variance (ANOVA, p ≤ 0.05) showed that year and rootstock, but not their interaction, significantly affected tree vigor (TCSA) and yield. However, fruit weight (FW) was only significantly affected by year, although the main source of variability was the interaction year × rootstock.  Short-waterlogging conditions are very common in clay-loam textured soils [48], the most frequent soils type in the Ebro Valley and the Murcia region, the largest producer areas in Spain. Since this country is the first peach producer in Europe, that type of soils is very common in the most important peach growing areas. Our field plot was classified as having a silty-clay-sandy texture. This texture can induce waterlogging problems and root asphyxia to sensitive rootstocks as peach seedlings and peach-almond hybrids [49]. Moreover, heavy textured soils facilitate flooding with the occurrence of intense rains, commonly concentrated in very short periods of the summer in the Mediterranean area [48]. Typical symptoms observed included an initial leaf wilting and defoliation followed by dead of twigs, and further of larger branches, and finally tree declined and died. Ziegler et al. [49] suggested that in fruit trees (especially in peach, as one of the most susceptible Prunus species to root anoxia in flooded soils) flooding stress can affect significantly plant survival and growth. Tolerance to root anoxia is determined by rootstock characteristics, and the differences in the responses to flooding should be considered in production settings where it is not possible to properly control irrigation to prevent short periods of flooding or in soils that do not drain irrigation or rain water quickly [48,49].

Production and Fruit Quality
As depicted in Table 4, the analysis of variance (ANOVA, p ≤ 0.05) showed that year and rootstock, but not their interaction, significantly affected tree vigor (TCSA) and yield. However, fruit weight (FW) was only significantly affected by year, although the main source of variability was the interaction year × rootstock.
Rootstock was the main source of variability for tree vigor (TCSA) ( Table 4). The highest vigor was induced by the peach-almond hybrid "Adafuel" followed by "GF 677" (Table 5). However, the last one did not differ statistically from the rest of the rootstocks evaluated, in agreement with previous studies [11,33]. Vigorous rootstocks could appear suitable for peach production under replanting conditions or in poor and calcareous soils that might otherwise not be favorable for growing peach [26]. Massai and Loreti [50] reported also that the most vigorous rootstocks appear to offer the best guarantees for the realization of efficient and uniform plantings in replant soils. Therefore, "Adafuel" and "GF 677" could be great rootstocks candidate for this type of soils. In addition, it is worth mentioning that "PADAC 04-03" and "Rootpac 70" had similar vigor in two soil conditions, our silty-clay-sandy and calcareous soil and the loamy and calcareous soil, as reported by Reig et al. [11]. The lowest TCSA value was induced by the plums "PM 105 AD" and "Adesoto 101", although they were no statistically Agronomy 2020, 10, 1159 8 of 22 different from the rest of the rootstocks but "Adafuel". Reig et al. [11] also reported a medium vigor of these two plum rootstocks. Therefore, plums may be suitable for reducing excessive growth of peach cultivars growing in fertile soils or to increase planting density [33].  Regarding yield (kg tree −1 ), the highest variability was attributed to the year, as shown in Table 4. Yields were reduced by spring frosts in 2016. All rootstocks produced fruits from the third year after planting (2011) (Figure 2), showing more clear differences among rootstocks at the sixth year after planting (2014). Thus, at the end of the trial (2017), the average yield was significantly affected by rootstock. The highest average yield was induced by the vigorous peach-almond hybrid "Adafuel", as shown in other studies [45], although it did not differ statistically from the rest of the rootstocks except for plums "PM 105 AD" and "Adesoto 101" (Table 5). In this study, because spacing was the same for all rootstocks, higher yields per tree or per hectare are related with rootstock vigor and greater canopy volume. Recalculating the planting distance of each rootstock based on its induced vigor is desirable to compare yields. Also, "GF 677" will be considered as reference rootstock [11]. Hence, comparing them in terms of cumulative yield efficiency (kg cm −2 ), the highest value was shown for "Adesoto 101", although differences were not significant with the other rootstocks evaluated, except for "Adafuel" which induced the lowest value (Table 5). Higher vigorous rootstocks, in general, induce lowest yield efficiency (YE) [24,51]. In our study, the tendency of the plum "Adesoto 101" to show higher YE is probably due to its lower vigor and the resulting good level of yield, as previously reported with other peach cultivars [17,22,25]. On the contrary, the tendency of the peach-almond hybrid "Adafuel" to show low YE is probably due to its higher vigor [45].
Agronomy 2020, 10, x; doi: www.mdpi.com/journal/agronomy evaluated, except for "Adafuel" which induced the lowest value (Table 5). Higher vigorous rootstocks, in general, induce lowest yield efficiency (YE) [24,51]. In our study, the tendency of the plum "Adesoto 101" to show higher YE is probably due to its lower vigor and the resulting good level of yield, as previously reported with other peach cultivars [17,22,25]. On the contrary, the tendency of the peach-almond hybrid "Adafuel" to show low YE is probably due to its higher vigor [45]. The significance is designated by asterisks as follows: *, statistically significant differences at p-value < 0.05; **, statistically significant differences at p value < 0.01; ***, statistically significant differences at p-value < 0.001; ns, not significant.

Fruit Quality Determinations
The ANOVA analysis reported a significant effect of the year (Y) and rootstock (R) for all traits studied, except for firmness (FF) and fruit weight (FW) which were not affected by rootstock ( Table  4). The year x rootstock interaction was not significant for most of the traits, hence highlighting that most rootstocks responded in a similar way to changes in the weather conditions [15], except for titratable acidity (TA) and a* parameters, indicating a different rootstock performance in relation to the climatic conditions. However, except for TA, which had similar variability caused by year and the interaction year × rootstock, the variability caused by the rest of the traits evaluated was mostly caused by the year.

Effect of Year on Fruit Quality Traits
When selecting the best cultivar/rootstock combination, it is crucial to evaluate the overall variability for most of fruit quality traits. In addition, it is important to understand which fraction of this variability is explained by the year, the rootstock, and/or their interaction. It is also important to The significance is designated by asterisks as follows: *, statistically significant differences at p-value < 0.05; **, statistically significant differences at p value < 0.01; ***, statistically significant differences at p-value < 0.001; ns, not significant.

Fruit Quality Determinations
The ANOVA analysis reported a significant effect of the year (Y) and rootstock (R) for all traits studied, except for firmness (FF) and fruit weight (FW) which were not affected by rootstock ( Table 4).
The year x rootstock interaction was not significant for most of the traits, hence highlighting that most rootstocks responded in a similar way to changes in the weather conditions [15], except for titratable acidity (TA) and a* parameters, indicating a different rootstock performance in relation to the climatic conditions. However, except for TA, which had similar variability caused by year and the interaction year × rootstock, the variability caused by the rest of the traits evaluated was mostly caused by the year.

Effect of Year on Fruit Quality Traits
When selecting the best cultivar/rootstock combination, it is crucial to evaluate the overall variability for most of fruit quality traits. In addition, it is important to understand which fraction of this variability is explained by the year, the rootstock, and/or their interaction. It is also important to evaluate which climate factor (temperature, precipitation, and solar radiation) is dominant in explaining the year-to-year variability [15]. In this work, all traits were not correlated with any of the weather variables evaluated, except for SSC and TA, leading to the thought that other factors differing between years play a critical role in them, such as tree vigor or crop differences observed during the five years of study, as previously reported [52].
The Tavr was highly and positively correlated with SSC (r = 0.99, p ≤ 0.01) and the Tmax was highly and negatively correlated with TA (r = -0.91, p ≤ 0.05), thereby partly explaining the influence of the year on these two traits. Font i Forcada et al. [14] indicated that sugars profile in fruits was positively correlated with temperature factors, showing that environmental variables such as temperature and/or solar radiation influence the tree-growing environment. In contrast, Iglesias et al. [15] reported a negative correlation between SSC and growing degree days (the difference between the daily mean temperature and a base temperature of 10 • C from full bloom date to harvest date). The low air temperature could play a role in an overall reduction in fruit size and sugar concentration [53]. Font i Forcada et al. [14] reported that organic acids profile in fruits was negatively correlated with temperature factors for three years of evaluation, as happens in the present study for TA and five years. In contrast to our results, Iglesias et al. [15] reported that malic acid was not affected by weather variables on "Big Top" budded on 20 Prunus rootstocks. However, they reflected only two years of evaluation.

Effect of Rootstock on Fruit Quality Traits
Considering differences between rootstocks for fruit quality traits, the average of the five years of study showed that the greatest SSC was recorded on the three-way plum based hybrid "PADAC 04-01", although differences were not significant compared to the peach-almond "Adarcias", the plums "Adesoto 101" and "PM 105 AD", as well as the plum based hybrids "PADAC 99-05" and "Rootpac R" ( Table 6). The lowest SSC values was recorded on "Adafuel", "PADAC 9902-01", and "Rootpac 70", although differences were not significant compared to "Cadaman", "GF 677", "PADAC 04-03", and "PM 105 AD". The SSC values obtained were similar to those reported for "Big Top" in different growing conditions [11,54]. The plum "Adesoto 101" induced higher five-year average TA values than "Adafuel", "GF 677", "PADAC 04-01" and "PADAC 04-03" ( Table 6). Orazem et al. [22] evaluated the performance of "Redhaven" peach cultivar grafted on eleven Prunus rootstocks. They observed that "Adesoto" rootstock induced higher values on FW and SSC to "Redhaven" fruits when compared to five other different plums and five peach-based rootstocks. Other different studies with "Catherina" [17,25] and "Big Top" cultivars [2,14] also showed the tendency of the "Adesoto 101" rootstock to induce higher TA and specific organic acids. Despite higher acidity of fruits on this rootstock, their SSC were not affected. High sugar contents and, to a lower extent, high acid contents seem to be favourable to fruit quality as evaluated by consumers [55] and could increase the sweetness perception of the low-acid "Big Top" nectarine [14].
Fruit color is one of the most important criteria of ripening in stone fruits [56] and an important factor for marketability and consumer acceptances [57]. Indeed, consumers associate intense red color with better appearance and quality [54]. Slightly significant differences were found among rootstocks in color parameters (Table 7). Regarding L*, the lightness fruits (higher L* values) were harvested on the plum based hybrid "PADAC 04-01" and the lower L* values (brightness fruits) on "PM 105 AD", although differences were not significant with the other rootstocks evaluated. For a* values, the peach-almond "Adafuel" induced fruits with the most red colored skin (higher a* value) although not significant differences were found when compared with all rootstocks except for "Rootpac R".
It is interesting to note the tendency of "Adafuel", "Adesoto" and "PADAC 9902-01" to show more red colored fruits in 2015, when the difference between Tmax and Tmin was higher of all the years evaluated (2013 to 2016). "PADAC 04-01" had the highest b* and hue (h) values, despite not being statistically different from the rest of the rootstocks, except for "Adarcias", "PM 105 AD" and "Rootpac 70". Hue (h) showed a significant low and positive correlation with TA (r = 0.30, p ≤ 0.01) ( Table 8), meaning that less mature fruits will have more lightness hue angle (h color parameter). However, in other studies, a negative correlation was found between hue angle and TA [45]. In general, high hue angle values could indicate low acidity, in agreement with Génard et al. [58] and Ruíz and Egea [59]. Color measurements, in general, are good predictors for fruit quality parameters except for the fruit firmness because fruits with the same hue angle may have different firmness [60]. The significance is designated by asterisks as follows: *, statistically significant differences at p-value < 0.05; **, statistically significant differences at p value < 0.01.
A hierarchical agglomerative cluster analysis was used to separate the rootstocks into three groups of increasing dissimilarity (Figure 3). This cluster showed that the rootstocks were not grouped according to their genetic origin, although results confirm that the plum based hybrids "PADAC 04-01" and "PADAC 99-05" induced fruits with the higher values on SSC and several color parameters (L*, b*, C*, and h). In a rootstock trial established in a peach production area of Tunisia, the hybrid "PADAC 04-01" induced also a good agronomical performance and higher fruit quality (SSC) compared to other nine rootstocks budded with the flat peach "Subirana" [10], confirming its suitability for the Mediterranean area.
Agronomy 2020, 10, x; doi: www.mdpi.com/journal/agronomy grouped according to their genetic origin, although results confirm that the plum based hybrids "PADAC 04-01" and "PADAC 99-05" induced fruits with the higher values on SSC and several color parameters (L*, b*, C*, and h). In a rootstock trial established in a peach production area of Tunisia, the hybrid "PADAC 04-01" induced also a good agronomical performance and higher fruit quality (SSC) compared to other nine rootstocks budded with the flat peach "Subirana" [10], confirming its suitability for the Mediterranean area. In general, the plums "Adesoto 101" and "PM 105 AD", the peach-almond "Adarcias" and the interspecific plum based hybrids "PADAC 04-01", "PADAC 99-05" and "Rootpac R", showed the tendency to induce a better fruit quality, based on higher concentration of SSC, and other yield traits (Tables 5 and 6), demonstrating their commercial interest as rootstocks for peaches. In general, the plums "Adesoto 101" and "PM 105 AD", the peach-almond "Adarcias" and the interspecific plum based hybrids "PADAC 04-01", "PADAC 99-05" and "Rootpac R", showed the tendency to induce a better fruit quality, based on higher concentration of SSC, and other yield traits (Tables 5 and 6), demonstrating their commercial interest as rootstocks for peaches.

Flower and Leaf Mineral Analysis
The mean concentrations of several macro-and micro-mineral elements from flowers and mature leaves for each rootstock are shown in Tables 9 and 10. Significant effect of rootstock was found on all mineral elements evaluated, except for leaf N concentration and flower Fe concentration. Reference values on a dry weight basis have been developed for leaf elemental concentrations [43,44]. However, there are not reference values for flower elemental concentrations and very few works are available [40,61,62]. These experiments indicated that the peach flowers, which are quite large and develop before leaves, constitute a promising way of entry into the tree for essential nutrients such as N, K, Mg, Fe and Mn. On the basis that a shortage can be also crucial at the nutrient-demanding flowering stage, early season foliar applications of fertilizers can also improve peach fruit yield and quality [62].  In this study, all rootstocks induced optimum leaf values for the assessed macro-elements according to Villar and Villar [44], except "PADAC 04-03" and "PADAC 99-05" for P, and "Adesoto 101", "PADAC 04-01", "PM 105 AD" and "Rootpac R" for Ca and Mg, which had lower values than the optimum. Nevertheless, P, Ca and Mg values were adequate for all rootstocks agreeing to Leece [43], with the exception of the plum based rootstocks "PADAC 04-01", "PADAC 99-05", "PM 105 AD" and "Rootpac R" showing marginal values for Ca. The tendency of plum based rootstocks to show lower leaf Ca values were also reported by comparing plum and peach based rootstocks in a calcareous soil of Tunisia [63].
Most rootstocks had leaf Fe, Mn and Zn deficiencies or close to deficiency (Table 10), according to reference values [43,44]. This could suggest according to the field fertilization that the roots from most of the rootstocks evaluated in this study have problems with uptake of these nutrients from the soil solution, due to the bicarbonate ions present into the typical soil solution of calcareous soils with high pH, making their concentrations in leaves lower than the optimum. The tendency of "Rootpac R" and "PADAC 99-05" to induce higher leaf Fe concentration could indicate higher tolerance to iron-chlorosis in calcareous soils. Despite the low leaf Fe values, no iron chlorosis symptoms were observed on all rootstocks evaluated. According to reference values proposed by Leece [43], all rootstocks showed Mn values lower than optimum except "Adafuel" and "Adarcias", showing the interest of these rootstocks in this type of soils. Mn deficiency has been also reported for peach rootstocks in heavy and calcareous soils [33,40], probably due to the lack of solubility of this element in soils with a high pH. Furthermore, increased Ca in soil or an excess of phosphoric acid fertilization might decrease or block Mn uptake [61,64,65].
Based on adequate values reported by Leece [43] and Villar and Villar [44], DOP index was calculated for each element separately (Tables 11 and 12) and for all elements together (Table 12). Significant differences among rootstocks were found on Ca, Mg, Cu and Mn elements for both ways of calculating the DOP index. Among rootstocks, "PADAC 04-03" was the most and the least balanced rootstock for Ca and Cu elements respectively, whereas "PM 105 AD" showed the opposite. This last one also showed wider imbalanced Mn value, whereas "Adafuel" and "Adarcias" showed the contrary. However, for the ΣDOP which had a high correlation between Leece [43] and Villar and Villar [44] (r = 0.86, p <0.0001), there were not significant differences among rootstocks.
Based on hierarchical agglomerative cluster analysis, three clusters were identified in both flower ( Figure 4) and leaf ( Figure 5) mineral status.
With regards to flower nutrient concentrations (Figure 4), the first group, which included "Adafuel" through "Cadaman", was characterized by rootstocks with high Mg, medium to low Zn, and very low Cu (Figure 4). In addition, "Adafuel" had high P and Ca concentrations and very low Cu, whereas "Cadaman" had high Mn (Tables 9 and 10, Figure 4), in agreement with Zarrouk et al. [40] when those rootstocks were budded with the "Queen Giant" nectarine and the "Tebana" peach in a similar calcareous soil. The second cluster, which included rootstocks from "GF 677" to "Rootpac 70", had medium to high N and P values, and medium to low Mn, K, Fe, Cu and Ca values (Figure 4). The last cluster, "Adesoto 101" through "Rootpac R", showed low values of N and P, and medium to high values of K, Fe, Cu and Ca. When mineral element concentrations of flowers are compared with those obtained in leaves, only the values of Ca, Mg, and Mn appeared lower than in leaves, as reported for pear [62], peach [40], and cherry trees [61].
Concerning leaf nutrient concentrations, the first group, which included "Adafuel" through "PADAC 04-03", had medium to high Ca, Mg and Mn values, and low Fe and K values ( Figure 5). "Adafuel" had also high N and Mn values in agreement with Zarrouk et al. [40]. PADAC 04-03 had high Zn and Ca values in agreement with Reig et al. [11]. The second cluster included "Adarcias" through "Rootpac R" which had high Ca, Mg and Mn, and low N and Cu values. The last group, "Adesoto 101" through "PM 105 AD", had medium to high N, Cu, Fe and Zn, and low Ca, Mg and Mn values. Reig et al. [11] also reported high leaf K values on "PM 105 AD" and high Fe values on "Rootpac 70". Several strong and moderate correlations were found when grouping the macro-and micronutrient values for both flowers and leaves. Thus, leaf Ca was highly and positively correlated with leaf Mg (r = 0.75, p ≤ 0.01) and moderately and positively correlated with leaf Mn (r = 0.60, p ≤ 0.01) ( Table 8). It is worthy to mention that Mg had a moderate positive correlation when flower and leaf samples were compared (r = 0.49, p ≤ 0.01), suggesting that Mg flower analysis could be used to detect and treat any deficiencies at an earlier stage in peach trees in calcareous soils.       Rootstocks labeled in red belong to Cluster 1, labeled in green to Cluster 2, and labeled in blue to Cluster 3.

Conclusions
This study demonstrates that the rootstock greatly influences some important tree and fruit traits, such as tree survival, vigor, yield efficiency, leaf and flower mineral nutrition and SSC. The vigor-controlling rootstocks ("Adarcias", "Adesoto 101" and other plums), and several interspecific plum based hybrids ("PADAC 04-01", "PADAC 99-05" and "Rootpac R"), showed the tendency to induce a better fruit quality. In addition, "Adarcias", "PADAC 04-01", and "Rootpac R" were the best adapted rootstocks in heavy-calcareous soil conditions and demonstrating their commercial interest as rootstocks for peaches. Therefore, new interspecific rootstocks are available based on their good performance and tolerance to abiotic stresses and adapted to Mediterranean conditions. Funding: Financial support was provided by the Spanish Ministry of Science and Innovation (grants AGL2011-24576, RFP 2015-00019 and RTI2018-094176-R-C33) co-funded by FEDER and the Regional Government of Aragon (grants A44 and T07-17R). C. Font i Forcada was also supported by the Ministry of Science and Innovation with a "Juan de la Cierva-Incorporación" grant (IJCI-2015-23397).