**4. Discussion**

Nowadays, both a high yield performance and fruit quality attributes (intrinsic and extrinsic) are very important due to the functional role of vegetables in the human diet [29].

Along with approaches such as genetic improvement and optimization of the cultivation conditions, herbaceous grafting technique represents a toolbox for securing or increasing yield stability and fruit quality with or without stress conditions [26,34–37]. In this paper, we investigated the effects of *S. aethiopicum* gr. *gilo* (accessions 1 and 2) versus the interspecific hybrid of *S. melongena* × *S. aethiopicum* gr. *gilo* on plant performance and fruit quality attributes of 'Scarlatti' eggplant F1 hybrid (black cylindrical shape). All the results are compared with those obtained from non-grafted, self-grafted, and *S. torvum* rootstock grafted plants.

Grafting success is an important concern when using new scion–rootstock combinations [37]. Generally, Solanaceous crops are grafted by cleft or tube grafting methods [4,20,38]. In our experiment using the tube grafting technique, grafting success rates of 89.1% for *S. aethiopicum* gr. *gilo* (accession 1), 93.6% for *S. aethiopicum* gr. *gilo* (accession 2), and over 99% for grafting onto 'Scarlatti', *S. torvum*, and *S. melongena* × *S. aethiopicum* gr. *gilo* hybrid, were observed. Our findings are in accord with those obtained by Bletsos et al. [4] and Rahman et al. [39], who reported a high grafting success rate of eggplant cultivars grafted onto *S. torvum*. Although, *Solanum aethiopicum* gr. *gilo* is a close relative species of *S. melongena* [16], from our results it appears that graft incompatibility might exist especially when 'Scarlatti' is grafted onto *S. aethiopicum* gr. *gilo* accession 2. Thus, we may assume that the genotype is a crucial aspect for the identification of a suitable rootstock as significant differences may exist between different accessions of a species. Regarding the *S. melongena* × *S. aethiopicum* gr. *gilo* hybrid rootstock, our outcomes are in accord with those obtained by Gisbert et al. [9], who reported a high graft success rate when interspecific hybrids of eggplants were used as rootstocks. Our results are also consistent with those of Sabatino et al. [10], who reported a high graft success rate of eggplant 'Birgah' F1 hybrid (violet globose shape) grafted onto *S. melongena* × *S. aethiopicum* rootstock, confirming the capacity of this interspecific hybrid to give good results with two different types of common eggplant. In fact, Birgah is one of the violet F1 hybrid belonging to a local Sicilian fruit typology that has been subjected to lower breeding improvement efforts when compared to elongated black-purple typelike 'Scarlatti', which is world-wide diffused especially in the most industrialized countries. Our experiment confirms

that rootstock might play a major role on plant biometric parameters. Plant height, which may be considered an indicator of vigor was highest in plants with *S. aethiopicum* gr. *gilo* (accession 1), *S. torvum*, and interspecific hybrid *S. melongena* × *S. aethiopicum* gr. *gilo* rootstocks and lowest in those plants with *S. aethiopicum* gr. *gilo* (accession 2) rootstock grafts. Therefore, vigor of the rootstock is important in conferring scion vigor. In the absence of grafting incompatibility signals, grafted plants may also develop more rapidly, hence contributing to earliness. In our work, plants grafted onto *S. torvum* and *S. melongena* × *S. aethiopicum* gr. *gilo* hybrid rootstocks flowered earlier than those grafted on the other rootstocks tested. Our results are consistent with those obtained by Gisbert et al. [9] and Sabatino et al. [10], who reported that plant vigor is positively related to fruit earliness. Increased earliness has also been reported for other fruiting vegetables such as melon grafted onto *Cucurbita* rootstock [40,41]. In our experiment, grafting has proved a useful technique to enhance eggplant 'Scarlatti' F1 hybrid yield traits. We also found that plants grafted onto *S. melongena* × *S. aethiopicum* gr. *gilo* hybrid rootstock had higher total and marketable yield plant−<sup>1</sup> than plants grafted onto *S. aethiopicum* gr. *gilo* (accessions 1 and 2), confirming that this latter rootstock has little value for enhancing eggplant yield. Conversely, interspecific hybrid of *S. melongena* × *S. aethiopicum* gr. *gilo* rootstock demonstrated great improvements in agronomic performance due to grafting. However, in our study grafting did not increase the number of marketable fruits. These findings are different from those of Gisbert et al. [9] and Sabatino et al. [10], who reported that eggplants grafted onto interspecific hybrids produced consistently more fruits per plant than ungrafted ones. However, despite in the present work grafting did not significantly improve the number of fruits per plant, the data recorded in our experiment supported the trend established in the experiments by Gisbert et al. [9] and Sabatino et al. [10]. On the other hand, these findings seem to be consistent with those obtained by Maršiˇc et al. [42], who reported that, in the second trial, grafting did not increase the number of fruit per plant in eggplant 'Galine' F1 (eggplant cylindrical shape). Therefore, it seems that the number of fruit per plant is a yield component influenced by the genetic background of the rootstock scion combination.

Apparent visual quality is an important criterion when consumers decide to purchase eggplant fruits. In our study, rootstocks affect fruit length and fruit calyx length. Our results are in accord with those obtained by Gisbert et al. [9], who hypothesized that fruit shape changes are probably due to changes in the concentration of growth regulators induced by the rootstock. Our findings on apparent visual quality are partially different from those of Sabatino et al. [10], who reported that rootstock significantly influences fruit width but has no significant effect on calyx length. Thus, it seems that, although fruit shape in eggplant is highly heritable and under genetic control [43,44] and even though rootstocks may affect cultivar fruit shape parameters due to changes in the concentration of growth regulators [9], rootstock may differently affect specific fruit shape characters in relation to the scion genotype.

Our results demonstrated that rootstock may enhance H◦ color parameter of skin eggplant fruit. Our outcomes are in contrast with those of Moncada et al. [45], who revealed that *S. torvum* rootstock decreased H◦ parameter in 'Black Bell', 'Black Moon', and 'Longo' eggplants. Thus, it seems that fruit skin color parameters are differently affected by scion–rootstock combination. Moreover, in our study, fruits from plants grafted onto *S. torvum* and *S. melongena* × *S. aethiopicum* gr. *gilo* hybrid had the highest H◦ color parameter value, which is an important commercial and qualitative prerequisite in eggplant fruit. Eggplant ranks among the top 10 vegetables in terms of antioxidant fruit activity. Accordingly to these data, a high fruit pulp browning potential could be expected [46,47]. Mishra et al. [46], King et al. [48], and Prohens et al. [49] revealed that eggplant cultivars differed in their extents of post-cut browning, which could be due to variations in the PPO activity or level of soluble phenolics. Our results on pulp browning are consistent with the findings of Sabatino et al. [10], who reported little or no effect of grafting by using *S. torvum* rootstock. However, we also found that fruits from plant grafted onto *S. aethiopicum* gr. *gilo* (accessions 1 and 2) revealed the highest value in terms of ΔL30.

Mennella et al. [19] revealed that chlorogenic acid is the major monomeric phenolic compound in eggplant fruits. Although few differences were found in fruit composition traits, higher fruit chlorogenic acid content was found in fruits harvested from plants with 'Scarlatti', *S. torvum*, *S. aethiopicum* gr. *gilo* (accessions 1 and 2) rootstocks, and ungrafted plants in comparison to fruits harvested from 'Scarlatti' scions grafted onto *S. melongena* × *S. aethiopicum* gr. *gilo* hybrid. This higher chlorogenic acid concentration may be an additional indication of stress in this rootstock–scion combination, as stress conditions induce accumulation of phenolics [50,51]. Our results are dissimilar from those of Sabatino et al. [10], who reported no differences in terms of fruit chlorogenic acid content among rootstocks tested, but are consistent with those of Sabatino et al. [26] who reported that grafting significantly increased phenolic concentration in fruits from grafted eggplant landraces only in three out of four genotypes. Our outcomes on phenolic content are also consistent with those of Maršiˇc et al. [42] who demonstrated that grafting significantly increased phenolic concentration in fruits from a grafted eggplant landrace as opposed to commercial varieties. Moreover, Toppino et al. [52] discovered environment-specific QTLs associated to the amount of chlorogenic acid. Accordingly, it seems that fruit phenolic content in eggplant is a quality trait also affected by scion–rootstock interactions. Mennella et al. [19], Friedman and McDonald [53], and Friedman [54] observed that genetic, environment, and growing conditions may affect glycoalkaloids concentration in potato, tomato, and eggplant. In addition, Jones and Fenwick [55] and Krits et al. [56] assessed that the level of total glycoalkaloids in potato tubers should not exceed 200 mg·kg−<sup>1</sup> of fw (or 200 mg·100 g−<sup>1</sup> of dw). Glycoalkaloids concentration ranged from 0.98 to 8.64 mg·100 g−<sup>1</sup> of fresh weight. Rootstock affected glycoalkaloids concentration in eggplant; however, it is important to remark that glycoalkaloids amount in fruits from grafted plant remained below the recommended threshold value. Our findings are in contrast with those of Sabatino et al. [10], who observed no significant differences in terms of glycoalkaloids concentration in 'Birgah' F1 eggplant fruits from plants grafted onto several potential rootstocks including hybrids and allied species.

Generally, changes in overall fruit composition between grafted and ungrafted plants were reported in tomato [57] and pepper [58]. Our results are in accord with those of Sabatino et al. [10], who found some differences in fruit quality attributes among fruits from ungrafted or self-grafted plants.

#### **5. Conclusions**

After the ban of methyl bromide use for soil fumigation, the soilborne diseases arising from continuous cropping are major issues that negatively affect productivity in eggplant. In this scenario, grafting is an effective tool to control vegetable crop stresses. Nowadays, *S. torvum* and tomato hybrids are the main eggplant rootstocks commercially used. However, investigations on the influence of new rootstocks on biometric parameters, yield performance, and fruit quality characteristics of different eggplant groups can provide further advances in this specific grafting field. In the present study, the interspecific hybrid of *S. melongena* × *S. aethiopicum* gr. *gilo* exhibited high grafting compatibility and high yield performance. These findings, together with the absence of negative effects on apparent 'Scarlatti' F1 hybrid fruit quality and composition, demonstrate that this interspecific hybrid may be a viable rootstock alternative to *S. torvum*.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4395/9/5/223/s1, Figure S1: Monthly meteorological data from May to August of 2014 and 2015 from the meteorological station of the experimental farm of the Department of Agricultural, Alimentary and Forest Sciences, University of Palermo, Italy, Figure S2: Plant height evolution of 'Scarlatti' eggplant plant ungrafted, self-grafted and grafted onto S. torvum, S. aethiopicum gr. gilo (accession 1), S. aethiopicum gr. gilo (accession 2) and S. melongena × S. aethiopicum gr. gilo rootstocks. Bars indicate the standard error of the mean. ANOVA analysis for plant height at 50, 80, and 110 DAT showed the following significance: \*\*\* (statistically significant differences at *p*-value below 0.001), \*\*\* and \* (statistically significant differences at *p*-value below 0.05), respectively, Figure S3: Root collar evolution of 'Scarlatti' eggplant plant ungrafted, self-grafted and grafted onto S. torvum, S. aethiopicum gr. gilo (accession 1), S. aethiopicum gr. gilo (accession 2) and S. melongena × S. aethiopicum gr. gilo rootstocks. Bars indicate the standard error of the mean. ANOVA analysis for root collar at 50, 80, and 110 DAT showed the

following significance: \*\*\* (statistically significant differences at *p*-value below 0.001), \*\*\* and \*\*\*, respectively. Table S1: Rootstock effects on apparent fruit quality in 'Scarlatti' F1 scion.

**Author Contributions:** L.S., G.I., G.L.R., and F.D., conceived and designed the research. L.S. also carried out field work, analyzed the data, and wrote the manuscript. E.P. and G.M. performed laboratory analytical determination. G.I. and G.L.R. helped draft the manuscript. All authors read and approved the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** We thank Carlo Prinzivalli for his assistance in establishing the field trial.

**Conflicts of Interest:** The authors declare no conflict of interest.
