*3.4. Mineral Content and Fruit Quality*

In the second experiment, all leaf mineral concentrations significantly differed by rootstock type, while Ca and Mg were affected also with irrigation method (Table 6). Plants grafted on commercial rootstocks had highest values for N, P and K (*p* ≤ 0.001). Ca was lowest in self-grafted plants grown with one-stem, while Mg was highest in same plants. Plants grown under both reduced irrigation techniques had significantly (*p* ≤ 0.01) highest Ca and Mg leaf concentrations. Fruit mineral concentration in same year was significantly affected by rootstock. Plants grafted on commercial rootstocks had higher values than both types of self-grafted plants (Table 7).

The effects of rootstock and irrigation on fruit quality traits are presented in Table 6. Both TSS and TA differed by rootstock type (*p* < 0.05). There were no differences among irrigation techniques in these traits, but the effect of the interaction of the rootstock × irrigation was found for TSS (*p* < 0.05) (Figure 3). Highest TSS was recorded in one-stem Attiya grown under DI and differ from values same plants under FI and PRD. Plants grafted on Emperador had higher TSS under PRD than other irrigation treatments. TA was higher in fruits grown on commercial rootstocks.


**Table 6.** Leaf mineral concentrations of self-grafted and grafted tomatoes grown under three irrigation techniques in the second year.

Significant differences between treatments (LSD test at *p* ≤ 0.05) are indicated with different letters within columns. ns—non-significant.

**Table 7.** Effect of rootstock type and irrigation treatment on tomato fruits mineral concentrations and quality parameters in the second year of the experiment with plants grown on one and two stems.


\* Significant differences between treatments (LSD test at *p* ≤ 0.05) are indicated with different letters within columns. ns—non-significant.

**Figure 3.** Total soluble solids (TSS) of tomato fruits sampled from different grafted plants grown with three irrigation techniques. Vertical bars represent SE values (*n* = 4). Different letters above column indicate a significant difference by the LSD test at *p* ≤ 0.05. ATT—one-stem Attiya; AT—two-stem Attiya; EM—Emperador; MX—Maxifort.

#### **4. Discussion**

In general, the use of commercial rootstocks resulted in highly improved plant vigor in terms of excessive vegetative growth. In our studies, the plant height and leaf number measured at 60 or 70 DAT were highest with plants grafted on Emperador and Maxifort rootstocks; the difference was even more pronounced in the second trial. Comparing height with leaf number, it can be assumed that interleaf nodes interval was not influenced by grafting or irrigation. These findings partially differed from other studies [14,20]. Similarly, vegetative biomass production (as leaf area, leaf and shoot DM) was bigger in grafted plants. For example, shoot DM in year one was 40% and 60% higher in Maxifort and Emperador grafted plants compared to self-grafted ones. Similar values were found in the second year on stem basis, but when we include all plant production (both stems) vegetative biomass is at least two or three times higher. Vigorous rootstocks had enough capacity to provide satisfactory vegetative growth by roots that can supply needed water and nutrients for assimilates production. On average, leaf area was not influenced by irrigation, as others found that tomato under DI and PRD had smaller leaf area than control and explained as soil drying affected roots reaction and production of chemical signals, i.e., changed ABA concentration or xylem sap pH that leads to stomatal closure and decreases leaf expansion growth [21].

Interactive effect between rootstock type and irrigation treatment showed that plants grafted on commercial rootstocks did not differ in photosynthetic rate (A), while both types of self-grafted ones differed depending on the applied irrigation technique. It can be concluded that grafted plant had better assimilative processes. The optimization of A under water stress could be modified by the rootstock through action on biochemical and biophysical processes [22]. Stomatal conductance and intercellular CO2 measured 20 days after starting irrigation treatments was lower under DI and PRD. These effects of reduced irrigation in some cases were noted later after initiation of irrigation [23], while in another study differences between treatments disappeared with time [24]. Valerio et al. [23] showed that lower stomatal conductance was related with leaf ABA accumulation, more ABA reduced stomatal conductance. Reducing stomatal conductance is a typical response to soil drying as stomatal closure is primary response to water deficit so plants could better contol water loss due to transpiration [7]. Stomatal closure reduced transpiration rate which was more pronounced in DI. Although, stomatal conductance was similar in both DI and PRD, it was expected that transpiration will be similar in both of them suggesting the response is mostly to the overall amount of water supplied to roots [21]. In contrast, in our study DI received 5% more water than PRD, so it seems that hydraulic signal is an important factor because plants under PRD on the wet side of the root can absorb enough water to keep higher level of transpiration [23]. Photosynthetic WUE had highest values under DI as result of lowest transpiration rate and similar photosynthetic rate to other irrigation. This can lead to lower biomass production as was noted in our study as reduced shoot biomass (although not significant) under DI and others found similar [21].

Rootstocks may affect tomato productivity positively or negatively, although in most cases yield increased both under non and stress conditions and depended on rootstock/scion combination [24]. In our experiment, plants grafted on commercial rootstocks had highest yield as a result of more and bigger fruits per plant, as was found in other studies [11,17,25]. Enhanced fruit production could be clearly related with higher plant biomass [15]. Early yield was different between years, in the first year highest early yield was noted under DI which can be related to a more pronounced water stress that hastened fruit ripening in this treatment. Topcu et al. [26] found higher tomato early yield in PRD than DI plants in experiment with more water reduction (50%) comparing our 40%. In experiment with two stems (second year), early yield was highest in one-stem-grafted Attiya what is result of longer period of growth for two stems plants because they were trained as side-shoots from cotyledons. In both experiments, cultivars grafted on commercial rootstocks had highest total yield under all irrigation treatments. In the second year comparing yield of these plants, it was found that under FI yield did not differ from DI but differ from PRD (Figure 1). It seems that rootstocks due to its vigor have enough capacity for water uptake to sustain yield under DI. It is important to notice that

growing on two stems (2nd experiment) did not reduce yield markedly when comparing with one stem plants (1st experiment), although different cultivars were used what should be taken into account. Rahmatian et al. [16] found dry matter allocation was not influenced by grafting or stem numbers and that good balance between vegetative and generative growth can depend on rootstocks. Other studies done with ungrafted greenhouse and processing tomato mostly obtained higher yield using PRD than DI [8] or similar to DI and FI [24].

WUE is the main indicator of plant water relations and is regulated by physiological mechanisms. In both years WUEy calculated as ratio between yield and water applied per treatment was higher in rootstock-grafted plants and as expected under PRD and DI. In these treatments higher fruit yield and lower water use resulted in improved WUE. It was not shown that PRD improved WUE better than DI, which means that irrigation volume is more important than used irrigation technique in determining yield or all crop growth as was suggested before [21,23,24], although other found irrigation technique can be more important [9]. Comparing two experiments it can be seen that in double stemmed plants (2nd experiment) WUE was higher leading to conclusion water use was optimized. In addition, in 1st Experiment WUEy was much lower in FI than in the second year what can be related to use of different soil moisture meters: tensiometers and soil sensors. The tensiometers was used for hand-operated irrigation, which could have led to overirrigation in the first year. It was shown that automatic operated tensiometers was more effective, which can be compared with sensors with automatic valves in our study [27].

The leaf mineral concentrations of P and K were under range of sufficiency while others were in range (N and Mg) or above (Ca) as proposed for greenhouse tomato. Grafting is considered as an effective tool for improving nutrient uptake and use efficiency in vegetables, although those were observed under optimal nutrient status in the root zone. N, P and K had higher concentrations in the plants grafted on commercial rootstocks what was expected and already confirmed in other studies that showed that nutrient uptake depends on rootstock–scion combinations [28]. Higher leaf P in grafted plants were reported for grafted eggplant and watermelon [29,30]. Grafted plants had more vigorous root system, which could be reason for increase in active uptake of P that has low mobility in soils. Self-grafted and grafted plants had low leaf K (under sufficiency range) because fertilization was not intensive as in commercial production. Potassium is nutrient normally required in the largest amount in tomato production. Grafting promote better growth and K uptake even under low K supply as was shown by Schwarz et al. [31]. These nutrients (N, P, K) concentrations were not affected by water supply rate, although opposite was shown for N in other studies for PRD or DI in non-grafted tomatoes [32]. Increase in K concentrations under water stress was found in some non-grafted and grafted tomatoes explaining that K accumulation improves stomatal resistance which improve drought tolerance [33]. In other case, decrease in grafted tomato leaf K was noted with increase in water stress level [12].

Regarding Ca2<sup>+</sup> and Mg2<sup>+</sup>, a significant increase in tomato Ca leaf concentration was found due to grafting what is in line with other reports [29,31]. In addition, both DI and PRD resulted in more leaf Ca than plants under FI. It was found that tomatoes under PRD had increased Ca uptake due to higher plant water status and lower stomatal conductance [34]. Higher Ca uptake induced by grafting are important for the tomato fruits due to the possibility of blossom-end rot incidence. Different than Ca, in grafted tomato was found lower leaf Mg what is in line with previous studies and could be also influenced by rootstock and cultivars used [5,14]. It seems that grafting somehow decrease Mg uptake in grafted vegetables, but reason it is not yet clear. Possible higher Ca uptake reacts antagonistically to Mg uptake, which could be related to specific transport systems [35]. Under reduced irrigation treatments higher leaf Mg was measured and same was found for mini watermelons [30]. Mg2<sup>+</sup> ion has largest hydrated radius among cations and this property makes Mg2<sup>+</sup> bind weakly to negatively charged soil colloids and root cell walls [36], which could lead to decreased Mg uptake under FI conditions due to leching in sub-root zones. The fruit mineral concentrations was influenced

by rootstock type showing that highest values in the plants grafted on Emperador and Maxifort. Other found effect of rootstock, but also influence of water stress on fruit minerals [18].

Higher TSS was affected by plant type with highest values in one stem plants. Interactive analysis showed it is mostly result of highest values of same plant type under DI. The enhanced TSS in that treatment could be result of water stress, although it is not clear why similar was not found in self-grafted two stem plants. Self-grafted double stemmed plants possible use more assimilates for additional vegetative growth [16]. Grafted plants had lower TSS what is often found even when used different cultivars and rootstocks [25]. For grafted plants vigorous roots can be additional sinks for assimilates and also better water uptake can result in dilution effect of fruits sugars [10]. Under PRD all plant types had similar TSS so it can be concluded that self and grafted plants with this irrigation type changed mechanisms responsible for results recorded under DI. Grafting on commercial rootstocks decreased Mg leaf content, which can possible lead to latent Mg deficiency influencing carbohydrate partitioning requiring for obtaining maximum yield and ensuring sugar accumulation in fruits [37]. In our study, both rootstocks increased the TA. Their increase by grafting was also found in many other experiments under different conditions [5]. Grafting under regular and low K resulted in higher TA, independent of K in fruits [31], while in our study K in fruits grafted on both rootstocks was higher compared to self-grafted plants. It is known that K concentration in fruits can be positively related with acid content, although further investigations are needed.

#### **5. Conclusions**

In the present experiment, we evaluated growth, gas-exchange parameters, yield, WUE and leaf mineral concentrations and fruit quality of self-grafted and tomato grafted on two commercial rootstocks cultivated in greenhouses in Mediterranean climate under three irrigation techniques: FI, DI and PRD. First year plants were grown with one stem and in the second with two stems.

In conclusion, these studies for the first time demonstrates the effects of parallel usage of different reduced irrigation techniques on grafted tomato vegetative and generative traits. Grafting onto commercial rootstocks improved plants growth and yield both in cultivation with one or two stems. Grafted plants under DI had minimal yield reduction compared to FI in double stemmed plants. WUE was highly improved with grafting and application of PRD and DI. That was more pronounced in experiment with two stem plants and could be result of different biomass partitioning and irrigation scheduling based on soil moisture sensors. Leaf mineral concentrations were higher in grafted plants as possible better uptake of vigorous rootstocks while Mg was reduced what imply contrasting rootstock–scion interactions. These findings indicate that grafted plants can be grown under moderate DI before PRD and that two stem plants could be used under that irrigation regime.

**Author Contributions:** Conceptualization, B.U., G.D., K.Ž.; methodology, B.U., M.R.; validation, G.D., G.V.S.; formal analysis, B.U., G.D.; investigation, B.U., M.R., G.V.S., M.M., A.M.; resources, B.U., G.D.; data curation, M.R., M.M., A.M.; writing—original draft preparation, B.U.; writing—review and editing, K.Ž., G.D.; visualization, B.U.; supervision, B.U., G.D., M.R.; project administration, B.U.; funding acquisition, B.U. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was financed by the Ministry of Agriculture, Croatia through project "Innovative technologies for tomato production enhancement and quality improvement" with project number 2015-13-01.

**Acknowledgments:** The authors are thankful to Silvia Miliši´c, Jelena Dumani´c and Željko Bili´c for their help maintaining experiments, collecting data and providing chemical analyses. In addition, special thanks to families Hrabar and Jankovi´c for providing usage of greenhouses and help in conducting experiments.

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

#### **References**


© 2020 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 (http://creativecommons.org/licenses/by/4.0/).

*Article*
