Inﬂuence of Harvest Stage and Rootstock Genotype on Compositional and Sensory Proﬁle of the Elongated Tomato cv. “Sir Elyan”

: The present experiment addressed the e ﬀ ects of two harvest stages, namely breaker (S 1 ) and turning (S 2 ), on compositional and sensory proﬁle of grafted, greenhouse elongated tomatoes cv. “Sir Elyan”. The rootstocks “He-Man”, “Interpro”, and “Armstrong” were used. The S 1 fruits showed a higher dry matter content, ﬁrmness and titratable acidity when compared to the S 2 ones (by 6%, 3%, and 15%, respectively). They showed, also, the highest L-ascorbic acid concentration and antioxidant activity. Di ﬀ erently, soluble solid content, lycopene and β -carotene increased in the S 2 fruits (by 4%, 92%, and 26%, respectively). Excepting methyl salicylate, all the volatiles detected peaked in the S 2 fruits, which were scored by panelists as sweeter and more ﬂavorful than the S 1 ones. Among the rootstock genotypes, “He-Man” promoted fruits ﬁrmness, carotenoids concentration and antioxidant activity, irrespective of the harvest stage. “Interpro” enhanced the sensory attributes of the S 1 fruits in terms of bitterness, sourness, sweetness and ﬂavor. Di ﬀ erently, sweetness, sourness, and tomato ﬂavor of the S 2 fruits were promoted by “Armstrong”. The results highlight the inﬂuence of rootstock genotype on the nutraceutical and sensorial proﬁle of “Sir Elyan” fruits harvested at early ripening stages.


Introduction
In the Mediterranean Basin, tomato (Solanum lycopersicum L.) represents a resource of primary economic and dietary importance [1]. Where quality is concerned, organoleptic traits and health-promoting compounds of fresh vegetables are becoming increasingly important [2,3], and for tomato these characteristics nowadays play a pivotal role in influencing consumers' preferences. From a nutraceutical viewpoint, tomato is a primary source of antioxidant in the human diet, with ascorbic acid, lycopene, and β-carotene being among the most effective in protecting human health from several degenerative diseases [4].
Tomato flavor results from the interaction among taste (deriving from sugars and organic acids) and aroma [5]. Tomato aroma flows from over 400 volatile organic compounds (VOCs), although only few of them are reputed of primary importance [5,6]. It has been demonstrated that these traits involve the developmentally-regulated expression of many polygenic systems, making them strongly influenced by environmental and agronomic factors [6]. Nowadays, greenhouse tomato cultivations are strongly dependent on the adoption of grafting, in order to increase their yield and resistance to biotic and/or abiotic stressors [7][8][9]. It has been reported that tomato quality variables such as acidity, sugar, flavor, aroma, color, carotenoid content, and texture can be differently influenced by the rootstock-scion combinations, also because of their interaction with external factors such as climate conditions and cultural practices [10]. Tomato fruits are harvested at different ripening stages, according to its destination. Fruit destined for the fresh market can be harvested from the mature-green to fully ripe stage [11], according to the consumers' demand and fruit typology.
In the paste decades, breeding has prioritized yield, shelf-life, and diseases resistance, which may have contributed to compromise flavor characteristics in fresh market tomatoes. However, the lack of flavor of retail tomatoes is partly due also to harvesting fruits before their full ripening is achieved. In fact, the concentration of individual volatiles in tomato fruits depends by the ripening stage at harvest, and fruit harvested before full ripening usually do not produce the characteristic volatiles associated to high quality tomatoes [11]. However, the threshold concentrations detected by humans for various aroma compounds range over many orders of magnitude. Sensory observations could be used to confirm the contribution of these compounds to odor and aroma [12].
The elongated-type tomato, that represents a significant commercial niche in Italy, could be consumed at breaker or turning stage. In a survey in Italy, the first ripening stages are preferred by the 20% of fresh tomato consumers [13].
The objective of this work was to investigate the composition and sensorial properties of tomato harvested at breaker and turning stages of an elongated tomato cultivar, grafted onto three common rootstocks in Mediterranean greenhouse cultivation.

Experimental Materials and Growth Conditions
The experiment was conducted in a 2400 m 2 greenhouse located in Southwest Sicily (36 • 50' N, 14 • 28' E; 18 m a.s.l.). Tomato plants cv. "Sir Elyan" F 1 , belonging to the medium-sized, elongated type, were grafted onto 3 rootstocks, characterized by different ability to imprint vegetative vigor to the scion: "He-Man" F 1 (low-vigor) "Interpro" F 1 (medium-vigor), and "Armstrong" F 1 (high-vigor). Transplanting was effected at the end of January adopting the following distances: 1.20 m (between double rows), 0.80 m (between paired rows) and 0.70 m (within single rows) (1.43 plant m −2 , 2 stems plant −1 ). A typical fertilization program was applied, whereas drip irrigation was provided when accumulated daily evaporation reached 25 mm. The crop was grown up to the half of July. A two-way randomized blocks design with four replications was adopted, using 6.8 × 3.6 m experimental plots, each containing 16 plants (net of borders).

Carpometric Determinations
On 10 May, 32 commercial tomato fruits from the third trusses per replicate were hand-harvested at 2 different ripening stages. These were breaker stage (16 fruits per replicate), i.e., when the berries started turning to red by their stylar end (rank 3 of the OECD Tomato Colour Gauge, hereafter S 1 ); turning stage (16 fruits per replicate), i.e., when colour change regarded~30 of the esocarp (rank 4 of OECD Tomato Colour Gauge, hereafter S 2 ). Soon after harvest, fruits were transported in the laboratory and processed for further analysis. Fruits fresh weight was determined, whereas the fruit shape index was calculated as the ratio among longitudinal and transversal diameters. Fruit firmness was determined through a Digital Texture Analyser mod. TA-XT2 (Stable Micro Systems, Godalming, UK) and defined as the force (N) needed to impress a 2 mm fruit deformation along its equatorial axis. Subsamples of collected fruits were kept in a thermo-ventilated oven at 70 • C (Binder, Milan, Italy) until constant weight was reached, in order to determine their dry matter content.

Fruit Quality Determinations
Subsamples of harvested fruit were washed with demineralized water, dried with paper and blended with a domestic food processor at room temperature. The resulting puree was centrifuged and an aliquot of the supernatant was used to determine the soluble solids content (SSC) by using a digital refractometer DBX-55A (Atago Co., Ltd., Tokyo, Japan) provided with an automatic temperature compensation system. Titratable acidity (TA) was determined using 10 g aliquots of tomato fruits poured in 50 mL of distilled water and titrated with 0.1N NaOH to an end-point of pH 8.1. TA was expressed as g L −1 citric acid (CA). The SSC/TA ratio was also calculated. Lycopene and β-carotene were extracted using the method described by Sharma and Le Maguer [14] and quantified by HPLC (equipped with a C30 Acclaim column) according to Gregory et al. [15] and Subagio et al. [16]. Ascorbic acid was extracted and quantified by HPLC (with an Ultra AQ C18 column) according to Nisperos-Carriedo et al. [17]. The antioxidant activity was determined using the free radical 2,2-diphenyl-1-picrylhydrazyl (DPPH) assay, according to Brand-Williams et al. [18].

Volatile Extraction and Analysis
Soon after harvest, fresh tomato sample were chopped and mixed. The aroma compounds identification was performed using SPME coupled with GC/MS. The fiber was chosen according to Beltran et al. [19], i.e., a 75 µm Carboxen/PDMS (CAR/PDMS) fiber (Supelco, Bellefonte, PA, USA). Before use, the fiber was preconditioned in the GC injection port at 300 • C for 1 h, then exposed for 1 h to the headspace of a 25 mL septum-sealed glass vial containing 20 g aliquot of homogenized fresh tomato. Each vial was previously immersed in a bath water at 60 • C for 15 min. GC-MS analyses were performed using an Agilent (Palo Alto, CA, USA) 6890 N GC equipped with a 30 cm length, 0.20 mm i.d., 0.20 µm film thickness, fused silica capillary column (SUPELCOWAX™ 10, Supelco). During the analysis, the GC injection port temperature was 250 • C, with a split ratio of 5:1. Helium was used as the carrier gas at a flow rate of 1.1 mL min −1 . The column temperature was held at 40 • C for 5 min, then programmed to increase by 5 • C min −1 to 220 • C, which was held for 10 min. Mass spectrometry conditions were as follows: ion source, 230 • C; electron energy, 70 eV; multiplier voltage, 1247 V; GC/MS interface zone, 280 • C; and a scan range of 35-350 mass units. Duplicate analyses were performed for each sample. Identification of the compounds was carried out by comparison of the analytes fragmentation patterns with the spectra libraries (NIST 98, US).

Sensory Analysis
The UNI EN ISO 13299:2016 [20] sensory profile method was used to measure any difference in sensory characteristics of tomatoes. Twelve trained (ISO 8586:2012) [21] panelists (six females and six males, 28-40 years old) with a broad expertise in vegetables were trained in 3 sessions, using both commercial and experimental samples to familiarize with scales and procedures. The panelists, using a discontinuous scale between 1 (absence of sensation) and 9 (extremely intense), evaluated the intensity of the sixteen attributes selected on the basis of frequency (≥60%): 1 for appearance (freshness); 1 for tactile hand feel (firmness); 3 for odor (herbaceous, tomato and off odors), 3 for flavor (herbaceous, tomato and off-flavors); 4 for taste (salt, sour, sweet and bitter); 4 for rheological properties (crunchy, juicy, mealy, peel thick) ( Table 1). The evaluation sessions were conducted in the sensory laboratory (UNI EN ISO 8589:2014) [22] of Di3A (University of Catania) from 11:00 a.m. to 12:00 a.m. in individual booths illuminated with a white light. Tomato samples were served on plates, coded with three-digit numbers and water was provided to panelists for rinsing between samples. The order presentation was randomized among panelists and sessions. All data were acquired by a direct computerized registration system (FIZZ Byosistemes. ver. 2.00 M, Couternon, France).

Statistical Procedures
Collected and calculated data attributable to ratio scales were firstly subjected to Shapiro-Wilk and Levene's test, in order to check for normal distribution and homoscedasticity, respectively, then to a factorial "rootstock × ripening stage" (R × S) analysis of variance (ANOVA), according to the experimental layout adopted in the greenhouse. Percentage data were Bliss' transformed before the ANOVA (untransformed data are reported and discussed), whereas multiple mean comparisons were performed through Fisher's protected LSD test (p = 0.05). Sensory data were subjected to a two-way non-parametric ANOVA using Friedman's test followed by the calculation of Kendall's coefficient of concordance, in order to check the independence of observations. Means separation was performed in all pairwise comparisons by using the Mann-Whitney's U-test, with an associated P-level calculated according to the Bonferroni's correction. A correlation analysis was also performed, in order to define possible relationships among volatiles concentration and sensory scores. All calculations were performed using Excel version 2016 (Microsoft Corporation, Redmond, WA, USA) and Minitab version 16.1.1 (Minitab Inc., State College, PA, USA).

Carpometric Traits
Average fruit weight and fruit shape index were both affected by R × S interaction. Passing from S 1 to S 2 , the former variable significantly increased only in "Sir Elyan" grafted onto "Interpro" (+13.6%), whereas shape index decreased only in "Sir Elyan" grafted onto "Armstrong" (−5.3%) ( Table 2). Fruit dry matter showed a similar trend in all the grafting combinations, decreasing from 7.4 (S 1 ) to 7.0% (S 2 ), whereas fruit firmness proved to be higher in "Sir Elyan" grafted onto "He-Man" than onto the other rootstocks, and in S 1 than S 2 stage (Table 2).

Taste Variables
Soluble solid content (SSC) proved to be significantly higher in S 2 than in S 1 fruits, and when "Sir Elyan" was grafted onto "Interpro" (Table 3). Differently, a significant R × S interaction was recorded for titratable acidity (TA) since, passing from S 1 to S 2 , it significantly decreased in "Sir Elyan" grafted onto "Interpro" and "Armstrong" (−19.7% and −12.8%, respectively) and did not change using "He-Man" as rootstock (Table 3). Both main factors significantly affected the SSC/TA ratio, as in S 2 fruits it was 18.7% higher than the S 1 ones, showing also a higher value in "Sir Elyan" grafted onto "He-Man" (Table 3). Table 3. Soluble solid content (SSC), titratable acidity (TA) and their ratio in tomatoes "Sir Elyan" as affected by rootstock and ripening stage (mean ± standard error). Different letters among factor means (bold numbers) indicate significance at Fisher's LSD test (p = 0.05). NS: not significant.

Fruit Nutraceutical Profile
As regards the L-ascorbic acid content, the S 2 fruits showed a significant decrease when "Sir Elyan" was grafted onto "He-Man" (−21.4%) while for the other rootstocks no statistical variations among ripening stages were recorded (Table 4). On the average of grafting combinations, lycopene content increased by 91.5% passing from S 1 to S 2 , whereas showed significantly lower values in "Sir Elyan" grafted onto "Armstrong", and the highest content in "Sir Elyan" grafted onto "He-Man" ( Table 4). The β-carotene content was affected by R × S interaction since, passing from S 1 to S 2 , the higher increase was noticed in the grafting combination "Sir Elyan"/"Armstrong" than in the other ones (Table 4). DPPH significantly decreased passing from S 1 to S 2 using "Interpro" and "Armstrong" rootstocks (Table 4).

Sensory Analysis
Tomato samples significantly differed for 11 out the 16 sensory attributes (Table 6). In particular, the bitter perception decreased in the S 2 fruits only in the grafting combinations "Sir Elyan"/"Armstrong" (−21.3%); the crunchy decreased in the S 2 fruits of "Sir Elyan" grafted onto "He-Man" and "Interpro" with a trend particularly evident in the latter (−32.6%). Both firmness and freshness in general dropped in the S 2 fruits, particularly in "Sir Elyan" grafted onto "Interpro" (−19.2% and −6.0%, respectively). Similarly, herbaceous flavor and odor showed a general decrease in the S 2 fruits with the strongest reduction in tomato grafted onto "He-Man" (−15.6% and −18.2%, respectively). Overall, salt and sour perceptions were higher in the S 1 fruits as compared to the S 2 ones, with the highest differences recorded in "Sir Elyan" grafted onto "Interpro" (+34.4%) and "Armstrong" (+13.8%), respectively. Differently, the sweet perception showed a different trend passing from S 1 to S 2 stage in the rootstock treatments. This sensory attribute increased using "He-Man" (+13.1%) and 'Armstrong (+12.3%) rootstocks and decreased in "Interpro". Tomato flavor and odor generally peaked in the S 2 fruits, with a gradient recorded in the grafting combinations "Sir Elyan"/"Armstrong" (+18.8%) and "Sir Elyan"/"Interpro" (+24.5%). Table 6. Mean scores of 16 sensory attributes of tomatoes "Sir Elyan" differing for rootstock and ripening stage. Different letters within each row indicate significance at Mann-Whitney's U-test (p = 0.04885).

Discussion
Under the specific conditions of our experiment, the S 1 fruits showed a higher dry matter content, consistent with their higher firmness, this last feature indicating less advanced metabolic processes when the reference ripening stage was achieved. Indeed, the decline in fruit firmness coincides with the up-regulation of several cell wall degrading enzymes, as well as with the dissolution of the middle lamella, leading to the reduction of the intercellular adhesion and cell wall depolymerization [26]. The grafting combination "Sir Elyan"/"He-Man" yielded the fruits with the highest firmness in both harvest stages, indicating the possibility to influence this trait by selecting the most suitable rootstock. This is an important commercial modification brought by grafting, since textural properties are implicated in fruits' shelf life and transportability, as well as on the perception of their flavor profile [10].
Soluble sugars (mainly glucose, fructose and sucrose) and organic acids (mainly citric and malic) are primary compounds of tomato fruits, whose amount are commonly measured through the soluble solid content (SSC) and titratable acidity (TA), respectively. From a sensorial viewpoint, their measure is linked to the perceived sweetness (SCC) and sourness (TA) of tomatoes, whereas the SCC/TA ratio describes the overall balance among them in the perceived taste [27]. All these variables are reputed primary contributors to the perceived flavor of tomato fruits [27]. In our experiment, the S 1 fruits were characterized by a decreased SSC and an increased TA, overall indicating their less sweet, more acidic taste. So our results confirm the lower sugar content characterizing tomato fruits harvested at earlier ripening stages [28]. Moreover, our results agree with previous reports of a TA increase up to breaker stage, and its subsequent decline with further ripening [29].
It is generally accepted that vigorous rootstocks show a higher sink strength, competing with the fruits for photosynthates accumulation [30]. Accordingly, we recorded a higher SCC/TA ratio when "Sir Elyan" was grafted onto "He-Man" (i.e., the least vigorous rootstock) than onto "Armstrong" (the most vigorous one), indicating a modified tendency of the fruits to accumulate sugars on the basis of the rootstock vigor.
When compared to other fruits, tomato shows only moderate ascorbic acid (AsA) content, but its dietary importance implies that even small variations in this micronutrient can have relevant effects for consumers [31]. The ascorbic acid concentration we recorded showed a tendency to peak in S 1 fruits. This is consistent with previous findings about the higher AsA biosynthetic capacity of younger fruits (up to 1.4-fold higher), probably to support their higher rates of cell division and expansion [31]. Indeed, it has been reported that AsA plays an important role in plants, related to cell division and cell wall synthesis [26]. On the other hand, the AsA drop in the S 2 fruits mirrored the decrease in fruit dry matter content, indicating a prevailing ripening-driven dilution effect in "Sir Elyan" tomatoes. For this nutritional trait, the rootstock-related differences proved to be significant only in the S 1 fruits, with the grafting combination "Sir Elyan"/"He-Man" (i.e., the least vigorous one) showing the highest concentration. Accordingly, the lowest fruit AsA content characterizing the most vigorous grafting combinations has been explained through their higher vegetative biomass, resulting in a redistribution or accumulation of this molecule in other plant fractions [32].
From a nutraceutical viewpoint, tomato is recognized as one of most important suppliers of carotenoids and phenolic acids, whose content contributes to the antioxidant capacity of this vegetable [33], and the shift in their concentration during ripening is one of the major changes accompanying the improvement of fruit palatability. The S 1 fruits showed compositional evidences of backwarded ripening process compared to the S 2 ones, resulting from a higher antioxidant activity and a lower concentration of the main carotenoids. These results agree with previous observations about the pivotal role of ripening stage in influencing these important nutraceutical traits of tomato [34]. Indeed, regarding the main tomato pigments, the carotenogenesis flows from the degradation of the photosynthetic membranes and metabolization of chlorophylls, leading to a progressive accumulation in tomato fruits of the C40 isoprenoids lycopene and β-carotene [35]. Regarding the rootstock effect, when grafted onto "He-Man", "Sir Elyan" fruits showed the best nutraceutical profile, both in terms of lycopene, β-carotene, and antioxidant capacity, irrespective of the ripening stage. This is an important outcome of our experiment, demonstrating the possibility of an overall functional improvement of greenhouse tomatoes using an appropriate graft combination, also in the light of their usually poorer nutraceutical composition when compared to field-grown tomatoes [36].
As regards the aroma components, excepting methyl salicylate, the S 1 fruits showed the lowest content for all the detected volatiles. The most severe reductions regarded the carotenoid-derived 6-methyl-5-hepten-2-one and β-ionone (−92% and −67%, respectively), the lignin-related 3-methyl-1-butanol (−78%) and the lipid-derived E-2-hexenal and E-2-heptenal (−70% and −66%, respectively). Consistent with the findings of Raffo et al. [28], such reductions clearly indicate a disturbance in the formation of tomato key odorants induced by early harvests. Indeed, it is known that the ability of tomatoes to form lipid-derived volatiles increases as the fruits ripen [37], so the reduced concentration we recorded in S 1 fruits in terms of C6 aldehydes and related alcohols (namely 1-hexanol and 3-hexen-1-ol) probably reflects the developmentally-induced lower expression of genes encoding for lipoxygenases and hydroperoxide lyases converting unsaturated fatty acids to 13-hydroperoxides and then to C6 aldehydes [38]. On the other hand, the increase in 6-methyl-5-hepten-2-one and β-ionone we found in the S 2 fruits, is consistent with their higher concentration of precursors lycopene and β-carotene, respectively [39]. On the contrary, methyl salicylate is typically associated to anti-herbivores defense of plant tissues [40], hence the higher concentration of this last volatile clearly indicates a less advanced ripening process in S 1 fruits, whose compositional traits still refer to the seed-protecting role of the pericarp. From a sensory viewpoint, the S 1 fruits were characterized as firmer and with the highest freshness, herbaceous odor and flavor. Differently, the S 2 fruits were generally scored as less acidic and with stronger typical tomato odor and flavor. Such modifications were clearly distinguishable, despite in the latter fruits a significant increase was recorded in "green" and "grassy" volatiles such as hexanal, E-2-hexenal, octanal, 3-hexen-1-ol, or 1-pentanol. This finding is only apparently contradictory, since in tomato no single compound has been found to be reminiscent of fresh ripe tomatoes, whose aroma, instead, springs from a multitude of several volatiles, blended in appropriate concentrations [41]. In this view, the correlation analysis revealed that 1-pentanol, β-ionone, and 6-methyl-5-hepten-2-one gave the strongest contribution to tomato flavor perception, whereas β-ionone and hexanal were the volatiles most tightly linked to the increased tomato odor perception. On the other hand, the alcohol 1-pentanol, the aldehydes hexanal and E-2-hexenal, as well as the apocarotenoid β-ionone and 6-methyl-5-hepten-2-one displayed a pool of negative correlations with some traits typically associated to less ripe tomatoes (namely crunchy, firmness, freshness, and herbaceous flavor), highlighting their primary role in accompanying the organoleptic evolution of tomatoes during ripening. It is interesting to note that, consistent with its higher concentration in unripe fruits, methyl salicylate was positively correlated to the perceived sour, salt and bitter, so suggesting a possible interactive effect of this compositional trait in generating these peculiar perceptions in less ripe "Sir Elyan" fruits. As regards the studied rootstocks, the volatile composition was somewhat erratic and unpredictable on the basis of the rootstock characteristics. Irrespective of the harvest stage, the grafting combination "Sir Elyan"/"He-Man" gave the highest concentration for 6 out of the 12 volatiles detected (including the fruity volatiles β-ionone and 6-methyl-5-hepten-2-one), whereas the combinations "Sir Elyan"/"Armstrong" gave the highest concentration for 5 volatiles (including 1-pentanol, 1-hexanol and methyl salicylate, all associated to "green" notes). Only 3-hexen-1-ol showed the highest concentration when "Sir Elyan" was grafted onto "Interpro". Such diversity gave a different ability to the rootstocks to offset the fruit sensory modifications brought by the different harvest stages. Indeed, when grafted onto "He-Man" or "Interpro", "Sir Elyan" gave S 1 fruits characterized by an attenuated bitter perception, consistent with their lower concentration of methyl salicylate. Moreover, the grafting combination "Sir Elyan"/"Interpro" gave less acidic S 1 fruits, characterized also by improved sweetness and tomato flavor. Differently, "He-Man" and "Armstrong" were able to accentuate the perception of sweet, sour and typical tomato flavor in the S 2 fruits, which, despite a lower SSC, mirrored their higher 6-methyl-5-hepten-2-one content when compared to "Interpro". This finding is consistent with the widely accepted idea that some fruity/floral volatiles improve the perception of sweetness in tomato fruits [24].

Conclusions
Our findings indicate that the harvest stage significantly affected both nutraceutical and eating quality profile of "Sir Elyan" tomatoes, but a certain dichotomy emerged among the functional profile and some sensorial traits in "Sir Elyan" fruits, in relation to their harvest stage. On the other hand, the rootstock genotype influenced both the nutraceutical and quality traits of the fruits. Despite none of the studied rootstocks was able to improve all the quality traits considered in this experiment, our results suggest the possibility to selectively improve the nutraceutical or sensorial profile of "Sir Elyan" fruits by selecting the most appropriated rootstock, provided that its interactive effects with the harvest stage are taken into account.