**Yield and Nutritional Quality of Vesuvian Piennolo Tomato PDO as A**ff**ected by Farming System and Biostimulant Application**

**Gianluca Caruso 1,\*, Stefania De Pascale 1, Eugenio Cozzolino 2, Antonio Cuciniello 2, Vincenzo Cenvinzo 1, Paolo Bonini 3, Giuseppe Colla <sup>4</sup> and Youssef Rouphael <sup>1</sup>**


Received: 7 August 2019; Accepted: 29 August 2019; Published: 2 September 2019

**Abstract:** Scientific investigations are being increasingly devoted to biostimulant effects on vegetable yield and quality, with the perspective of sustainable crop management. Two farming systems (conventional or organic) in factorial combination with two biostimulant treatments (tropical plant extract (PE); legume-derived protein hydrolysate (PH)) plus a non-treated control were compared in terms of tomato fruit yield, yield components, mineral composition, functional and nutritional indicators. PE- and PH-based biostimulants resulted in higher plant biomass, PH even in higher leaf area index, compared to non-treated control. Marketable yield was not significantly affected by farming system. PH and PE gave higher yield than non-treated control. PH treatment led to higher fruit number than the control, whereas PE incurred significant increase in yield only under organic farming. The mean fruit weight attained the highest value upon PE application under conventional management. Colour component a\* (redness) was higher with the conventional system compared to the organic one, whereas an opposite trend was shown by the organic acids malate, oxalate and isocitrate. Irrespective of the farming system, the soluble solids, fruit brightness (L\*) and redness as well as the target organic acids malate, oxalate, citrate and isocitrate were significantly higher than untreated plants by 10.1%, 16.1%, 19.8%, 18.9%, 12.1%, 13.5% and 26.8%, respectively, with no significant differences between the PH- and PE-based biostimulants. Higher lipophilic activity and total ascorbic acid concentration but lower lycopene were recorded under organic management. PE and PH application resulted in higher total phenol and ascorbic acid as well as in lycopene content, and lipophilic antioxidant activity than the non-treated control. Biostimulants proved to be an effective sustainable tool for enhancing tomato fruit yield and functional quality both under conventional and organic vegetable systems.

**Keywords:** antioxidant activity; functional quality; lycopene; organic farming; protein hydrolysate; *Solanum lycopersicum* L.; tropical plant extract

#### **1. Introduction**

Organic horticulture has been increasing worldwide for the past two decades, as a result of rising demand of consumers for healthy and safer food [1], accounting for 3.5 million ha in 2014, which is almost twofold compared to 2008 [2]. Indeed, this farming management is environmentally-friendly due to food production with minimal harm to ecosystems as well as minimal use of inputs in particular fertilizers and pesticides [3]. However, the lower yield compared to conventional agriculture, i.e. −20% according to Ponisio et al. [4] and −5% to −34% as reported by Seufert et al. [5], represents a disadvantage of organic farming. The latter yield reduction is mainly associated to higher biotic pressure caused by parasites, pests and pathogens [4,6] and to nutrient limitation in particular N and P [7] which limits production in several organic-based systems [8]. In fact, the rate of major minerals such as nitrogen and phosphorus released from organic fertilizers and crop residues do not often meet the crop demand during the highest rate plant growth, leading to significant yield reduction [9].

Within both conventional and organic farming systems, the use of naturally derived plant biostimulants is a promising sustainable approach [10,11], aiming to enhance (i) plant nutrient availability/uptake/assimilation and use efficiency, (ii) abiotic stress tolerance as well as (iii) product quality [12–14]. Within biostimulants, protein hydrolysates (PHs) are mainly made of amino acids, polypeptides and oligopeptides derived from proteins of animal or plant origin upon partial hydrolysis [15] and can be applied to seeds, leaves or soil in several forms (liquid or granular) [12]. Tropical plant extract (PE) and especially legume-derived protein hydrolysates (PHs) obtained from vegetal origin proteins have been drawing interest in world agricultural areas, compared to animal-derived ones, due to both their higher agronomic value [16] and no use constraints in organic farming. Moreover, PE or PH application to leaves and/or roots reportedly elicit physiological processes, thus resulting in enhancement of growth [17,18], production and quality [18,19], tolerance to abiotic stressors, such as drought, soil and water salinity, extreme temperature, nutrient deficiency, soil acidity and alkalinity [11,20–25]. Notably, PE or PHs also encourage plant activity of key enzymes involved either in N or C metabolism [12,24,26,27]. In addition, PH treatment may boost crop performances, by eliciting auxin- and gibberellin-like activities through bioactive peptides [17,28,29]. PE and PHs also exert indirect effects on plants, as they modify the architecture of roots and increase their hair surface expansion, thus enhancing macro- and microelement uptake [26,28,30–32]. However, limited scientific literature are available with regard to the effect of foliar applications of PH or PE in interaction with either conventional or organic farming on agronomical and fruit quality responses of tomato landraces, in particular the long shelf-life cherry tomato landrace 'Pomodorino del Piennolo del Vesuvio' (PPV), a typical niche product of Campania (Italy) horticultural sector.

In the perspective of the above mentioned topics, a two-year experiment was carried out to assess the response of cherry tomato landrace PPV to foliar applications of a vegetal protein based hydrolysate or a tropical plant extract biostimulant in interaction with organic or conventional crop system, in terms of yield, mineral composition, functional and sensorial quality attributes.

#### **2. Materials and Methods**

#### *2.1. Growing Conditions and Experimental Protocol*

The experimental research was carried out on open field grown tomato (*Solanum lycopersicum* L.) 'Piennolo del Vesuvio D.O.P.' ecotype Riccia, in Portici (Naples), southern Italy characterized by a typical Mediterranean climate, in 2016 and 2017. The soil was sandy-loam having 77% sand, 14.5% silt, 8.5% clay, with soil electrical conductivity of 342 μS cm<sup>−</sup>1, 1.6% organic matter, 0.94 g kg−<sup>1</sup> N, 63.9 mg kg−<sup>1</sup> P2O5, 1.8 g kg<sup>−</sup><sup>1</sup> K2O. The monthly air temperature (day/night) and rainfall recorded at the plant level, expressed as means of the two research years, were the following: 21.6 ◦C, 7.9 ◦C and 47.2 mm in April; 24.5 ◦C, 11.3 ◦C and 56.3 mm in May; 29.5 ◦C, 15.8 ◦C and 23.7 mm in June; 32.2 ◦C, 17.1 ◦C and 17 mm in July.

A factorial combination of biostimulant application (B) and farming system (F) was applied, based on two biostimulant treatments (PH or PE) plus a non-treated control and two farming systems (organic or conventional). The experimental design was a randomized complete-block design with three replications, yielding 18 experimental units (3 B × 2 F × 3 replications). Each experimental unit consisted of an 8 square meter plot. Tomato seedlings were transplanted on 25 and 24 April in the first and second growing season respectively, at a plant density of 4 plants m<sup>−</sup>2.

The two commercial PH and PE-based biostimulants 'Trainer'®and 'Auxym'®were kindly provided by Italpollina S.p.A., Rivoli Veronese, Italy. The legume-derived PH biostimulant obtained through enzymatic hydrolysis contains 75% of free amino acids and peptides, 22% of carbohydrates and 3% of mineral nutrients. The detailed aminogram of the product along with the phenolics, flavonoids and elemental composition were reported by Rouphael et al. [31] and Paul et al. [25]. The PE biostimulant obtained by fermentation of tropical plants contains 54% of free amino acids and peptide, 17% carbohydrate, 23% mineral nutrients, 6% vitamins and 0.22% phytohormones as reported in detail by Rouphael et al. [33] and Caruso et al. [32].

Cherry tomato plants were sprayed with a solution containing 3 and 2 ml L−<sup>1</sup> of PH- or PE-based biostimulant, or with water (non-treated control), four times during the growing season at 7-day intervals, starting in coincidence with the early growth of the first fruit truss.

Organic farming practices were performed in compliance with the EC Regulation 834/2007 and related subsequent updates. Both in conventional and organic systems, the fertilization was carried out with 153 kg ha−<sup>1</sup> of N, 39 kg ha−<sup>1</sup> of P2O5 and 223 kg ha−<sup>1</sup> of K2O. Phosphorus was completely supplied at planting, whereas nitrogen and potassium were given both prior to crop establishment (31% and 55% for N and K2O respectively) and the remainder on dressing. Under the organic management a 6-5-13 Bioilsa organic-mineral fertilizer (based on hydrolyzed collagen and meat flour), N (11%) and N-K (7%–21%) hydrolyzed protein manure were used; ammonium sulphate, potassium sulphate, potassium nitrate and ammonium nitrate were supplied to the conventionally grown crops. Drip irrigation started when the soil available water capacity decreased to 80%. Crop protection was performed against downy mildew, tomato leaf miner, aphids, whitefly, and red spider.

#### *2.2. Yield, Biometric Assessments and Leaf Color Measurements*

Harvests of fully ripe fruits were performed from 14 July to 2 August, as an average of the two research years, and the marketable yield, number of fruits per plant and the mean fruit mass were determined on a sub-plot of 4 m2. Fruits that were deformed or misshaped were considered unmarketable. The final leaf area was measured on 10 plants in each experimental plot using a Licor 3000 electronic area meter (Licor, Lincoln, NE, USA) and then the leaf area index was calculated. A sample of the fresh material was dried at 70 ◦C for about 3 days until reaching constant weight, to determine dry aboveground biomass.

Cherry tomato color was measured on the two sides of 10 fruits per experimental unit using Minolta CR-300 Chroma Meter (Minolta Camera Co. Ltd., Osaka, Japan) in order to obtain the color space parameters, in particular L\* (brightness), a\* (redness) and b\* (yellowness).

#### *2.3. Juice Total Soluble Solids and Fruit Dry Matter Content*

The cherry tomato PPV fruits were homogenized in a blender for 2 min and the homogenate was filtered, then the total soluble solids content was measured using the Bellingham and Stanley digital refractometer (model RFM 81). The tomato fruit dry matter percentage was also determined after drying the fresh material at 70 ◦C for about 3 days until reaching constant weight. The dried tomato fruit samples were collected for further mineral analysis.

#### *2.4. Mineral and Organic Acids Analysis*

The desiccated cherry tomato fruit tissues were ground in a Wiley Mill to pass through an 841 μm screen and used for macro-mineral profile analysis, sodium content and organic acids as described in detail by Rouphael et al. [31] and Kyriacou et al. [34]. Phosphorus, potassium, calcium, magnesium, sulfur, sodium, malate, oxalate, citrate and isocitrate were separated and quantified by ICS-3000 ion chromatography (Dionex, Sunnyvale, CA, USA) coupled to a conductivity detector. Macronutrients, sodium and organic acids concentrations were expressed on a dry weight basis (g kg−<sup>1</sup> dw).

#### *2.5. Antioxidant Activity Analysis*

The lipophylic and hydrophilic antioxidant activities were assessed on extract from freeze-dried cherry tomato PPV fruits (200 mg) added with methanol and distilled water, respectively. The antioxidant activity of the lipophilic and hydrophilic extract fractions were measured with the 2,20-azinobis 3-ethylbenzothiazoline-6-sulfonic acid ABTS [35] and with the N,N-dimethyl-pphenylenediamine (DMPD) methods [36], respectively. The absorbance of the solutions for LAA and HAA were measured at 734 and 505 nm, respectively. Lipophylic and hydrophilic antioxidant activities were expressed as mmol of Trolox (6-hydroxy-2,5,7,8-tetramethylchro man-2-carboxylic acid) and mmol ascorbic acid per 100 g of dw [36].

#### *2.6. Antioxidant Molecules Analysis*

The total ascorbic acid and polyphenols were assessed spectrophotometrically based on the protocol by Kampfenkel et al. [37] and the Folin–Ciocalteau procedure [38], respectively, after slight modifications [34]. For quantification, ascorbate and gallic acid were used as external standards to build calibration curves both for total ascorbic acid and total polyphenols content. The absorbance of the solutions for total ascorbic acid and total polyphenols were measured at 525 and 765 nm, respectively, and the results were expressed as mg ascorbic acid on 100 g fw and mg gallic acid per 100 g dw. Lycopene content was also assessed spectrophotometrically, based on the protocol by Sadler et al. [39], and for the quantification pure lycopene (Sigma, St. Louis, MO) was used to build the calibration curves. The absorbance of the lycopene hexane solution was measured at 472 nm. Lycopene content was expressed in mg 100 g−<sup>1</sup> fw.

#### *2.7. Statistical Processing*

All agronomical and qualitative data were subjected to three-way analysis of variance using the software package SPSS. The means were separated by DMRT test at 0.05 significance level. All the agronomical and quality variables were not significantly affected by the growing season (i.e., year) or its interactions with the two experimental factors applied, and therefore the mean data of the two years were reported.

#### **3. Results and Discussion**

#### *3.1. Yield and Morphometric Measurements*

As reported in Table 1, the legume-derived protein hydrolysate (PH) resulted in the highest leaf area index and biomass of the vegetative plant parts, though the latter variable was not significantly different from that recorded under the effect of the tropical plant extract (PE).

Marketable yield and its components, fruit number and mean weight, were significantly affected by biostimulant treatment, whereas no differences were recorded between conventional and organic systems. Moreover, fruit number and mean weight were also significantly influenced by the interaction between the two experimental factors (Table 1). The application of PH-based biostimulant resulted in the highest yield but PE biostimulant also gave a significantly higher yield compared to non-treated control (+18.7% and +11.2%, respectively); these outcomes stemmed from the combined effects of fruit number and mean weight (Table 1).

The two latter variables were significantly affected by the interaction between the two studied factors (Table 1). For instance, the fruit number per plant was just connected to the effect of PE-based biostimulant, leading to higher fruit number than the control only under organic system, whereas PH-based biostimulant always showed the best effect. Moreover, the mean fruit weight attained the highest value upon PE application under conventional management, the latter being also higher than that obtained with the organic system, whereas no differences were recorded between the remaining comparisons.


**Table 1.** Plant growth parameters and yield indicators of 'Piennolo del Vesuvio' cherry tomato as affected by farming system and biostimulant application. dw, dry weight. ns, \* nonsignificant or significant at *p* ≤ 0.05, respectively. Different letters within each column indicate significant differences according to Duncan's multiple range test(*<sup>p</sup>* ≤ 0.05).

In contrast with the present research findings, in previous investigation [40] conventional management of different vegetable species led to higher yield than organic one. Consistently with our results, Colla et al. [30] detected growth and yield increase of tomato in greenhouse upon PHs application, which is a whole crop cycle extension of the short-time stimulation effect observed on tomato treated with PH extracts [17,41]. Notably, the effects shown by the applied biostimulant on plants is different from the nutritional input elicited by fertilizers [42]. Indeed, in our research tomato plants showed different patterns of yield components response to the applied substances in interaction with the crop system (Figure 1).

Foliar applications of PE and PHs may have triggered in tomato plants a physiological mechanism linked to the enhanced content of signaling molecules which are the prevailing PE and PH components [12]. In this respect, low-sized molecules such as peptides and free amino acids can regulate plant phenological progress upon their easy absorption through leaves and roots by promoting endogenous biosynthesis of phyto-hormones [43]. Consistently, other authors [17,18,31,33,44] reported that plant growth, fruit setting and yield were enhanced by the auxin- and sometimes gibberellin-like activity of the mentioned biostimulants.

PE- and PH-based biostimulants are likely to boost plant development and yield through: (i) stimulating cell proliferation by signaling molecules such as specific amino acids connected to nitrogen metabolism (i.e., glutamic and aspartic acids) and soluble peptides; (ii) vitamin provision targeted to cell protection from oxidation; (iii) encouraging plant metabolism with micronutrients supply ([26] and references cited therein). Moreover, an important increase in cytokinins content was promoted by biostimulant application in *Spinacea oleracea* [45]. An additional action pattern of PE and legume-derived PH consists of enhancing macronutrient uptake and assimilation through modulation of root biomass, density and lateral root number, as well as microbial activity with the consequent higher availability of soil nutrients [13,30].

**Figure 1.** *Cont*.

**Figure 1.** Interaction between farming system and biostimulant application on 'Piennolo del Vesuvio' cherry tomato fruit number per plant (**a**) and mean weight (**b**). Different letters mean significant difference according to Duncan's multiple range test at *p* ≤ 0.05. Lowercase letters refer to the comparison between biostimulants, whereas capital letters to the comparison between farming systems within each biostimulant application.

In other experiments, *Lactuca sativa* L. sprayed with PE or PH showed a 11% higher biomass than non-treated control, which may be as a consequence of both the stimulation exerted by the most represented substances such as amino acids and key peptides and of enhancement of cultivable epiphytic bacteria as well as their species richness and diversity [46]. Overall, the direct and/or indirect mode of actions of the applied biostimulants may have boosted both growth and crop productivity of treated cherry tomato plants compared to the non-treated control treatment.

#### *3.2. Fruit Colorimetry, Nutritional Quality and Mineral Profile*

Farming system significantly affected some target indicators of tomato fruit colorimetry and nutritional quality as well as mineral composition (Tables 2 and 3). Two out of the three variables characterizing the colour (a\* and b\*) were higher under conventional management compared to the organic one; conversely, the organic acids malate, oxalate and isocitrate attained higher concentrations in the organically grown berries (Table 2). In the present research, both tomato fruit dry matter percentage and soluble solids were not significantly affected by farming management, whereas in previous investigations asparagus spears [47] and leek pseudo-stems [48] organically grown in southern or northern Europe respectively showed higher dry matter and sugar content than those managed conventionally.


**Table 2.** Flavor compounds and fruit colorimetry of 'Piennolo del Vesuvio' cherry tomato as affected by farming system and biostimulant application. TSS, total soluble solids. ns, nonsignificant or significant at *p* ≤ 0.05, respectively. Different letters within each column indicate significant differences according to Duncan's multiplerange test (*<sup>p</sup>* ≤ 0.05).

**Table 3.** Fruit mineral composition of 'Piennolo del Vesuvio' cherry tomato as affected by farming system and biostimulant application.


ns, \* nonsignificant or significant at *p* ≤ 0.05, respectively. Different letters within each column indicate significant differences according to Duncan's multiple range test (*<sup>p</sup>* ≤ 0.05).

Regardless of the farming system, the soluble solids, fruit brightness and redness as well as the target organic acids malate, oxalate, citrate and isocitrate were significantly higher than untreated plants by 10.1%, 16.1%, 19.8%, 18.9%, 12.1%, 13.5% and 26.8%, respectively, with no significant differences between the PH- and PE-based biostimulants (Table 2). The highest fruit juice soluble solids and organic acids obtained in biostimulant-treated plants independently on the formulate could be considered important key quality attributes for consumer satisfaction [49]. Consistently with our findings, Rouphael et al. [31,33], Colla et al. [18] and Ertani et al. [19] reported the increased content of soluble solids, glucose and fructose in greenhouse grown *Solanum lycopersicum* and *Capsicum chinensis* fruits upon the treatment with biostimulants, derived from tropical plant extract fermentation, enzymatic hydrolysis of legume and alfalfa plants or by extraction of red grapes.

Minerals content is essential for the quality of fruit vegetables including tomato. Based on two surveys carried out in Finland and the USA, Levander [50] demonstrated that the contribution of vegetables to dietary intake of phosphorus, potassium, calcium, magnesium and sodium is 7–11%, 31–35%, 5–7%, 18–24% and 11%, respectively. The present work has generated important information regarding the relative abundance of minerals in cherry tomato landrace and its variation range across farming system and biostimulant application. In this respect, K was found by far the most abundant mineral, followed by Ca, P, Mg, S and Na (Table 2).

For all measured minerals no significant interaction between farming system and biostimulant application was observed (Table 3). Neither farming system nor biostimulant application had significant effect of Ca content in fruit (average 5.7 g kg<sup>−</sup>1). The effect of biostimulant application on tomato fruit mineral profile was much more pronounced than the farming system. K and Mg were positively affected by both biostimulants compared to non-treated control, with no significant difference between them. PH-based biostimulant exhibited a higher content of P; in addition, both commercial biostimulants had a better effect on S content compared to the untreated control, with PH showing the highest values (Table 3). In other investigations, compared to non-treated plants the application of a PH-based biostimulant resulted in better nutritional status: higher K and Mg content in tomato [18,31] and in spinach [33] grown under protected cultivation.

In the present research, the increased concentration of cherry tomato fruit K and Mg induced by the application PH-based biostimulant might have been mediated through several direct/indirect mechanisms involving: (i) enhanced mineral uptake promoted by root growth stimulation encouraging absorption, translocation and accumulation of nutrients [17,51]; (ii) higher nutrient transporter expression in cell membranes [24,52]; (iii) the action of PH biostimulant bioactive compounds (soluble peptides, carbohydrates and free amino acids) in strengthening the sink effect and therefore the movement of nutrients within the plant [42].

#### *3.3. Antioxidant Activity and Bioactive Content*

Fruit vegetables in particular tomato are considered good sources of lipophilic and hydrophilic antioxidant molecules such as lycopene, total ascorbic acid and polyphenols. The influence of farming system and biostimulant application on antioxidant activities and bioactive compounds are reported in Table 4. Neither farming system nor biostimulant application had a significant effect on hydrophilic antioxidant activity (average 10.9 mmol ascorbate eq. 100 g−<sup>1</sup> dw). When averaged over biostimulant application, higher lipophilic activity and total ascorbic acid concentration but lower lycopene were recorded under organic management compared to the conventional one. Moreover, no significant differences between the two farming systems arose with regard to hydrophilic antioxidant activity and phenols content (Table 4). Consistently with our results, in previous research carried out on strawberry in southern Italy [53], organic farming resulted in higher fruit ascorbic acid than the conventional management. As for the biostimulant application, both the PH and PE biostimulants resulted in higher lipophilic antioxidant activity as well as phenols, ascorbic acid and lycopene concentration than non-treated control, with no significant differences between the two commercial biostimulants used (Table 4).


*Agronomy* **2019** , *9*, 505

The phytochemical homeostasis requires enzymatic activities leading to a stabilization of the concentration of antioxidants which show an increase both in response to free radical production [19,31] and when K and Mg in the tissues are high [31]. In this respect, the protection against oxidative stresses in maize plants was primed by both protein hydrolysate and plant extract based biostimulant through the expression of superoxide dismutases activity-regulating genes [54], which catalyze the enzymatic dismutation of superoxide to H2O2 [55]. The application of protein hydrolysates in greenhouse conditions encouraged the synthesis of ascorbate, p-coumaric, chlorogenic acid, capsaicin and antioxidant activity in *Capsicum chinensis* L. fruits [19], as well vitamin C in tomato fruits [18,31]. Similarly, *Spinacia oleracea* phenolic acids production was enhanced by biostimulant application [45], through the phenylalanine ammonia lyase pathway [56]. Therefore, the foliar application of plant biostimulants such as PH or PE can be instrumental in satisfying increasing consumer standards for the functional quality aspects of fresh cherry tomato PPV landrace [57,58].

#### **4. Conclusions**

From research carried out in southern Italy on tomato landrace 'Piennolo del Vesuvio D.O.P.' the effective application of plant biostimulants based on tropical plant extract or legume-derived protein hydrolysate on fruit yield, nutritional and functional attributes arose. Indeed, both formulates overall enhanced production, quality, mineral and antioxidant indicators either under organic or conventional farming systems. Controversial outcomes stemmed from the comparison between the two crop managements, as conventional farming resulted in better colored and lycopene richer fruits, but higher organic acids, ascorbic acid content and lipophilic antioxidant activity was recorded when organic procedures were applied. The present study allows us to draw important conclusions relevant to the significant contribution of biostimulant application in making sustainable even a conventional tomato farming system.

**Author Contributions:** Conceptualization, G.C. (Gianluca Caruso) and Y.R.; data curation, G.C. (Gianluca Caruso), E.C. and Y.R.; formal analysis, G.C. (Gianluca Caruso); funding acquisition, G.C.( Giuseppe Colla) and P.B.; investigation, G.C. (Gianluca Caruso), E.C., A.C. and V.C.; methodology, G.C. (Gianluca Caruso), S.D.P., E.C. and Y.R.; supervision, Gia.Ca.; writing—original draft, G.C. (Gianluca Caruso) and Y.R.; writing—review and editing, G.C. (Gianluca Caruso), S.D.P., G.C.( Giuseppe Colla) and Y.R.

**Funding:** This work was partially supported by Italpollina Company (Rivoli Veronese, Italy).

**Acknowledgments:** The authors wish to thank Roberto Maiello for his assistance with laboratory equipments.

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

#### **References**


© 2019 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/).

*Review*

## **Melatonin as a Chemical Substance or as Phytomelatonin Rich-Extracts for Use as Plant Protector and**/**or Biostimulant in Accordance with EC Legislation**

#### **Marino B. Arnao \* and Josefa Hernández-Ruiz**

Department of Plant Physiology, University of Murcia, 30100-Murcia, Spain; jhruiz@um.es **\*** Correspondence: marino@um.es; Tel.: +34 868887001

Received: 3 September 2019; Accepted: 19 September 2019; Published: 21 September 2019

**Abstract:** Melatonin (*N*-acetyl-5-methoxytryptamine) is a ubiquitous molecule present in animals and plants, and also in bacteria and fungi. In plants, it has an important regulatory and protective role in the face of different stress situations in which it can be involved, mainly due to its immobility. Both in the presence of biotic and abiotic stressors, melatonin exerts protective action in which, through significant changes in gene expression, it activates a stress tolerance response. Its anti-stress role, along with other outstanding functions, suggests its possible use in active agricultural management. This review establishes considerations that are necessary for its possible authorization. The particular characteristics of this substance and its categorization as plant biostimulant are discussed, and also the different legal aspects within the framework of the European Community. The advantages and disadvantages are also described of two of its possible applications, as a plant protector or biostimulant, in accordance with legal provisions.

**Keywords:** biostimulant; fertilizer; melatonin; phytomelatonin; plant protector; plant stress

#### **1. Introduction**

Melatonin (*N*-acetyl-5-methoxytryptamine) is a biogenic amine derived from the amino acid tryptophan, which was discovered in 1958 in the cow pineal gland by Lerner and cols. [1]. Two years later, it was detected in humans and its chemical structure was elucidated. This molecule, which was initially only related to changes in the structure of melanocytes in amphibians, fish and reptiles, was soon found to act as a neurohormone in mammals [2,3]. Since its discovery it has become one of the most researched molecules. In animals, it presents a multitude of physiological actions such as a role in the circadian rhythms of several molecules, and its influence on sleep–wake cycles, mood, motor activity and body temperature changes [4–7]. Its influence on food intake and its relationship with metabolic syndrome has also been demonstrated [8–10]. In other more specific situations such as the physiology of the retina, the immune system, sexual behavior and as an anti-cancer effector, melatonin also has a relevant role [11–15]. In addition, interesting and extensive reviews on the role of melatonin in animals and humans can be consulted [16–23].

In 1995, the presence of melatonin in plants was discovered [24–27]. During the following years there was much reluctance on the part of researchers to accept this, since some refused to believe that a neurohormone could be present in plants, and much less that it had any role in their physiology. A key piece was the elucidation of the melatonin biosynthesis route in plants, localized between the mitochondria, chloroplasts and cytoplasm of cells, and which has been studied with great accuracy by K. Back and J. Kong in rice and *Arabidopsis* plants [28–30]. However, it is now fully accepted that melatonin is present in all plant species and that it presents a panoply of interesting actions. Indeed,

several studies have demonstrated its role in processes such as seed germination, growth and the development of seedlings, leaves and roots. It takes part in organogenesis processes such as rooting and fruiting, and in processes of leaf and fruit senescence. It acts as a protector of the photosynthetic and stomatic system, and as a regulator of various enzymes of the metabolism of carbohydrates, lipids, amino acids, nitrogen, sulfur and phosphorus. It also has a role in the secondary metabolism, enhancing the synthesis of flavonoids, anthocyanins, and carotenoids, among others. It regulates its own biosynthesis and that of several plant hormones such as auxin, abscisic acid, gibberellins, cytokinins, ethylene, polyamines, jasmonic acid and salicylic acid [31–39].

Of all the aspects investigated, its protective action against stress situations has been the most researched and about which most is known. Melatonin exerts a protective action, mediated by major changes in gene expression, both against abiotic (cold, heat, drought, waterlogging, salinity, alkalinity, acid rain, chemical contamination by heavy metals, UV radiation) and biotic (bacteria, fungi, virus) stressors. As a result, plants are more tolerant and/or resistant to the negative action of such stressors [31,36,40–43] (see below). The term "biostimulant" was first proposed to denote "materials that, in minute quantities, promote plant growth" by Zhang and Schmidt (1997) [44]. Later, the definition was modified by Kaufman et al. (2007) as: "Biostimulants are materials, other than fertilisers, that promote plant growth when applied in low quantities" [45]. According Du Jardin (2015), the following definition is proposed: "A plant biostimulant is any substance or microorganism applied to plants with the aim to enhance nutrition efficiency, abiotic stress tolerance and/or crop quality traits, regardless of its nutrients content", and extended as "plant biostimulants also designate commercial products containing mixtures of such substances and/or microorganisms" [46]. Under the EC (European Community) regulation: "Plant biostimulants will be EC marked as fertilizing products stimulating plant nutrition processes independently of the products' nutrient content with the sole aim of improving one or more of the following characteristics of the plant and the plant rhizosphere or phyllosphere: Nutrient use efficiency, tolerance to abiotic stress, crop quality, availability of confined nutrients in the soil and rhizosphere, humification and degradation of organic compounds in the soil". Extensive revision works on this topic can be consulted [47,48]. The objective of this work is to provide sufficient data to establish the clear protective role of melatonin against adverse environmental situations, and to discuss the possible global use of melatonin as a biostimulant and/or bioprotective agent. Current legislation of the EC, is taken into account and the advantages and disadvantages of its use in plant crops destined for animal and human consumption are analyzed.

#### **2. Melatonin as a Regulator of Plant Stress Physiology**

Although there was much evidence in the 1990s that melatonin could exert some role as an antioxidant agent in animal cells and tissues, it was not until 2004 and 2006, in carrot cells and Chinese licorice (*Glycyrrhiza uralensis* Fisch.), that the possible protective role of melatonin in plants was suggested [49–51], although some curious and previous data existed [52]. The initial idea that melatonin in plants, as in animals, could play an important role as an antioxidant was taking shape and results in this regard became ever more plentiful [53–57]. In addition, studies on melatonin as a possible plant regulator were also progressing, especially since the initial studies of Arnao and cols. on the role of melatonin in plant growth and development, and the so-called auxin-like activity [58–64].

It was not until the publication of results on the action of melatonin on changes in gene expression that the extent and potential of melatonin as a regulatory agent of multiple physiological processes in plants became widely known [64–70]. Exceeding previous expectations, melatonin is capable of activating all known molecular stress mechanisms in plants. Thus, gene regulatory factors involved in the response to cold, high temperatures, salinity, drought, chemical toxicity, etc., and also biotic stress, are up-regulated by melatonin [31,38,40,41,43]. Melatonin also regulates the expression of multiple enzymes related to hormonal homeostasis, up-regulating or down-regulating the expression of genes that encode enzymes of the biosynthetic or catabolic pathways of plant hormones including indole-3-acetic acid (auxin), gibberellins such as gibberellin-4 (GA4), cytokinins, abscisic

acid (ABA) and ethylene. It also others regulators such as salicylic acid (SA), jasmonic acid (JA) and polyamines [31–33,35,38,69,71–75]. In general, subjecting plants to a stressful situation—which leads to an increase in endogenous levels of melatonin—or treatment with exogenous melatonin, results in a stress tolerance response mediated by specific stress response factors and changes in the endogenous levels of plant hormones involved in the response [31–36,38,40–43,75–85]. In addition, the recent identification of a melatonin receptor in *Arabidopsis thaliana* has opened new expectations related to its role as a new plant hormone [86]. Figure 1 shows these aspects in a condensed form.

**Figure 1.** Model of redox network/melatonin action on abiotic and biotic stress responses.

#### **3. Beneficial Responses to Melatonin Treatments in Di**ff**erent Crops in Stress Situations**

Studies conducted with melatonin in plants under both abiotic and biotic stress are numerous. Table 1 compiles many of the studies with an agronomic interest since they deal primarily with crop species for human consumption. Table 1 presents studies classified by plant species, where there are many physiological aspects that are investigated in which melatonin exerts some generally beneficial action. These include seed germination, the growth and vegetative development of plants; photosynthesis, its pigments, photorespiration, stomatic conductance and water economy; the yields of seeds and fruits in adverse conditions; osmoregulation, ion exchange and adjustments in osmotic and hydric potentials, and the regulation of the different metabolisms of carbohydrates, lipids, nitrogen compounds, sulfur and phosphorus cycles. In regards to the secondary metabolism, melatonin induces the biosynthesis of flavonoids, anthocyanins and carotenoids, among others; in hormonal homeostasis, it intervenes in the regulation of all plant hormones and its own biosynthesis. It promotes the rooting process of primary, secondary and adventitious roots while during foliar senescence, melatonin regulates the expression of chlorophyll degradation-related and senescence-induced genes. In the postharvest control of fruits, melatonin increases the ethylene and lycopene content, and regulates many enzymes of the cell wall, ethylene biosynthesis, and primary and secondary metabolisms. It also helps preserve cut flowers; in fruiting it induces parthenocarpy. Finally, its role in bacterial, fungal and viral pathogenic infection should be emphasized, slowing damage and stimulating systemic acquired resistance (SAR) to favor crop health.

Obviously, all the above plant physiology aspects are of interest for application in plant production. Indeed, while many of the above studies were at a laboratory level, others have already been put into practice in crops with excellent results.

In general, exogenous melatonin applications are made through the root system, in irrigation water, or by spraying leaves. In the last case, no adjuvant is needed since melatonin is an amphipathic molecule that crosses biological membranes and the waxy cuticles. Melatonin is transported via the xylem from the roots to the rest of the organs of the plant quite effectively [87,88].
















 Increased content or increased action. ↓, decreased content or decreased action; ABA, abscisic acid; AGR, absolute growth rate; ASA, ascorbic acid; CGR, crop growth rate; Chls, chlorophylls; CMC, component materials categories of fertilizers; EC, European Community; ECHA, European Chemical Agency; EU, European Union; GA4, gibberellin-4; GABA, γ-aminobutyric acid; GSH, glutathione; JA, Jasmonic acid; LAI, leaf area index; MDA, malondialdehyde; MAPKK, mitogen-activated protein kinase cascade; NAR, net assimilation rate; OXI1, oxidative signal-inducible1 kinases; PFC, product function categories of fertilizers; ROS, reactive oxygen species; RWC, relative water content; SA, salicylic acid; SAR, systemic acquired resistance; TA, total valuable acidity; TCA, Krebs cycle; TSS, total solid soluble.

↑,

#### **4. Melatonin in the Health and Environment of EC**

In accordance with the Classification, Labelling and Packaging (CLP, EC-No 1272/2008) regulation, which is based on the United Nations' Globally Harmonized System, which has a purpose to ensure a high level of protection of health and the environment, as well as the free movement of substances, mixtures and articles, the European Chemical Agency (ECHA) classified melatonin (EC No. 200-797-7 (CAS 73-31-4), *N*-(2-(5-methoxyindol-3-yl)-ethyl)-acetamide), as a non-hazardous substance in terms of physical and chemical hazards. With respect to human health, it is classified as a non-hazardous substance in the oral, dermal, inhalation and irritation categories, and in regards to mutagenicity and carcinogenicity. However, melatonin is classified as a health hazard substance (code H-361) in terms of reproductive toxicity because it is suspected of damaging fertility or an unborn child. This classification reflects one of its multiple functions as an animal hormone, in which its participation in the modulation of sexual behavior in mammals has been demonstrated, and also, it is believed, the same of fertility [233,234]. In fact, it is usually applied to sheep as a hormonal regulator of sexual zeal to homogenize the reproductive process in ovine, with demonstrated higher conception and pregnancy rates when applied [235]. Nevertheless, melatonin is classified as non-hazardous in terms of its possible damage to the environment and atmosphere.

#### **5. Melatonin as an Active Substance or as a Plant Biostimulator**/**Protector in Crops: Concepts and Legal Considerations in EC**

After many changes and adaptations, the EC finally seems to have established its policy regarding the authorization, classification, use, distribution, importation, management, etc., of plant protectors and fertilizers, in an attempt to improve agricultural production, while minimizing risks and hazards for humans, animals and the environment. In order to establish the minimum basis for the possible use of melatonin in plant production and post-harvest application, several requirements regarding its human consumption must be taken into account:

(i) Melatonin is a highly studied substance that has given rise to abundant physicochemical and biological data; (ii) there are numerous studies in animals and humans regarding its beneficial effects on health, in aspects as diverse as neurodegenerative, immunological, liver, renal, heart, skin and gastrointestinal diseases, in addition to osteopathy, retinopathy, etc. It also helps in the treatment of various cancers, particularly, chemical and radiological therapies; (iii) in regards to melatonin for human consumption, although it is classified as a drug in the EC, there are some cases in which it does not need a medical prescription, such as those where the amount of melatonin is less than 1 mg. Generally, these are used for jet-lag and sleep disorders. In many other countries (e.g., USA, Canada) melatonin is not treated as a drug, but as a food supplement; (iv) in no case has melatonin been declared as toxic, even at the intake of 1 g/day. Only some slight side effects such as migraine and headache have been described.

The possible use of melatonin in plant production involves particular aspects such as: (i) Melatonin is a molecule that exists in all living things, from bacteria to humans, but also in plants, algae, fungi, etc.; (ii) its action in animals and humans is well known since it has been investigated for many years. In plants, although many physiological effects of melatonin are known, new data are being acquired every day; (iii) in all cases, only positive effects have been described, all beneficial for the development of plants (the same can be said for animals); (iv) little information is available on its effect on bacteria and fungi, especially those that are part of the soil microbiota (rhizosphere); (v) there are also few or no data on its effect on the environment, in particular on agricultural and aquatic fauna; (vi) the levels of melatonin described in plants, and which appear to be effective in pharmacological treatments known to date, are much higher than those described in animals or humans, which may be a cause for caution.

Council Directive 91/414/EEC of 15 July, 1991 concerning the marketing of plant protection products provides rules governing plant protection products and the active substances contained in those products. This old directive has been replaced by two more current ones that are as follows:


If we review the actions confirmed so far for melatonin in plants, we find that melatonin exerts a clear action as a plant protector in situations of biotic stress against bacterial, fungal and viral diseases (Regulation #2), but it can also be used as an agent against situations of abiotic stress (Regulation #1). Thus, Regulation #2 says in point 22:

"Certain substances, mixtures and micro-organisms, referred to as plant biostimulants, are not as such inputs of nutrients, but nevertheless stimulate plants' natural nutrition processes. Where such products aim solely at improving the plants' nutrient use efficiency, tolerance to abiotic stress, quality traits or increasing the availability of confined nutrients in the soil or rhizosphere, they are by nature more similar to fertilising products than to most categories of plant protection products. They act in addition to fertilisers, with the aim of optimising the efficiency of those fertilisers and reducing the nutrient application rates. Such products should therefore be eligible for CE marking under this Regulation and excluded from the scope of Regulation (EC) No 1107/2009".

These two regulations attempt to classify the substances and products applicable to crops into two large groups: Those that are plant protectors (phyto-sanitary) (Regulation #1) and those that can be used as fertilizers (Regulation #2). As we have seen in the previous section, melatonin is classified as a health hazard substance (code H-361) for its reproductive toxicity in ECHA, so its possible authorization as an active substance by regulation EC 1907/2006 of Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) could be difficult.

Although Regulation #1 on plant protection products extends the concept of an active substance, since it includes microorganisms and preparations (art. 1 point 2): This Regulation shall apply to substances, including micro-organisms having general or specific action against harmful organisms or on plants, parts of plants or plant products, referred to as 'active substances', some interesting restrictions appeared in:


Thus, taking into account all this legal information, and ruling out the possibility of using melatonin as an active substance (pure chemical substance) for agronomic application, the possibility of using plant, bacterial, algae, or fungi extracts rich in melatonin would remain. Thus, a good plan

might be to use plant (or other) extracts rich in melatonin as a fertilizer, in the category of biostimulants. A biostimulant could also be defined as a formulated product of biological origin that improves plant productivity as a consequence of the emergent properties of its constituents. Thus, biostimulants could be defined by their demonstrated mode of action and origin, or solely by their demonstrated beneficial impact on plant productivity. The challenges in developing a definition are also complicated by the multi-component and largely undefined composition of many biostimulant products and the possibility that the activity of a biostimulant may not be explained by the presence of any individual constituent, but is a result of the interaction of many constituents in the product. Indeed, most biostimulants in use today are complex mixtures of chemicals derived from a biological process or the extraction of biological materials [236].

According Regulation #2 (EU 2019/1009) on fertilizing products, in Annex I, Product Function Categories (PFCs) of EU fertilizing products, in Category 6, two types of plant biostimulant can be developed: Microbial plant biostimulants (subtype A) and non-microbial plant biostimulants (subtype B). In Annex II, it says: "An EU fertilizing product shall consist solely of component materials complying with the requirements for one or more of the CMCs listed in this Annex", where the different component materials categories (CMC) were defined. Of interest are the following:


The strategies to obtain melatonin-rich extracts may involve microorganisms (PFC6A) or plants (PFC 6B). At present, there seem to be no data on the production of melatonin by bacteria or fungal cultures. The objective to obtain melatonin-rich plants (CMC2) is ambitious since phytomelatonin levels in plants are usually very low, and less than 5–10 ng per gram of plant. An exhaustive classification of many plants according to their phytomelatonin content can be consulted [37,237,238]. Generally, medicinal plants have high phytomelatonin content, but this tends to vary widely due to the varied origin of plants, technical conditions of growth, variety, post-harvest treatment, etc. Several strategies can be followed: (i) Selecting plant species with high levels of phytomelatonin which can be extracted and concentrated, and (ii) inducing the biosynthesis of phytomelatonin in in vitro cultured pre-selected plant tissues. A discussion on this aspect can be consulted [239]. Our group is developing a formulation where only aromatic/medicinal plants are used to obtain a botanical mixture rich in phytomelatonin through the application of a simple process. A rigorous plant selection protocol and careful management will ensure high phytomelatonin content in the plant extracts generated. The formulation and its protocol are being patented before being made available to interested companies for commercial exploitation. We are currently characterizing it and conducting the appropriate studies and bioassays in plants to confirm its beneficial biological activity related with its high phytomelatonin content.

Figure 2 shows, according to the legislation analyzed, the pros and cons of melatonin (as a chemical substance) and phytomelatonin-rich extracts and its possible regularization as a plant protector or fertilizer (biostimulant).

**Figure 2.** Pros and cons of the possible use of chemical melatonin and rich-phytomelatonin extracts according to EC legislation.

#### **6. Future Prospects**

Numerous studies with melatonin have resulted in a set of data that indicate the excellent beneficial effects that this compound has on plants, especially in stress situations. It should not be forgotten that melatonin is a natural compound, endogenous to plants and other organisms including humans. It is this last aspect that makes it more interesting and also more delicate or sensitive, when using it as a plant protective agent or as a biostimulator. However, demonstrating through trials that its use is possible in crops and does not entail risks to human and wildlife health will be the only way forward in this field. The alternative of using phytomelatonin-rich extracts seems more interesting, but also more laborious. The search and selection of plants with high endogenous levels of phytomelatonin is a first requirement for subsequent extraction and preparation. The analysis and study of its potential as a protector against plant stress will throw light on the true effect on crops. However, although many aspects of the mechanism of action of phytomelatonin are already known, there are other relevant aspects to study as: (i) The optimal mode of application, time and rate; (ii) the phenological state; (iii) the effect on rhizosphere; (iv) the persistence in soil or in foliar applications; (v) the synergic or antagonic effects with other plant treatments (pesticides, fertilizers, etc.), among others. Obviously, companies in the phytochemical sector (manufacturers) will need to start field studies and deal with possible legal regularization.

#### **Abbreviations**



**Author Contributions:** The manuscript was conceived by M.B.A. and written and revised by M.B.A. and J.H.-R. **Funding:** Not financial support available for this review.

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

#### **References**


© 2019 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*

### **E**ff**ect of Vegetal- and Seaweed Extract-Based Biostimulants on Agronomical and Leaf Quality Traits of Plastic Tunnel-Grown Baby Lettuce under Four Regimes of Nitrogen Fertilization**

**Ida Di Mola 1,\*, Eugenio Cozzolino 2,\*, Lucia Ottaiano 1, Maria Giordano 1, Youssef Rouphael 1, Giuseppe Colla <sup>3</sup> and Mauro Mori <sup>1</sup>**


Received: 3 September 2019; Accepted: 20 September 2019; Published: 22 September 2019

**Abstract:** Nitrogen (N) fertilizers play a crucial role in agriculture, representing a powerful tool for farmers for increasing yields throughout the seasons under both optimal and suboptimal conditions. At the same time, their synthetic/chemical nature could have several influences on ecosystems and human health. For this reason, there is an urgent need to find new and more sustainable means of production to increase plant productivity and optimize nitrogen use. An experiment was conducted in a plastic tunnel to assess the response of baby lettuce crop to the foliar application of three plant biostimulants (PBs): Legume-derived protein hydrolysate (LDPH) 'Trainer®', tropical plant extract (TPE) 'Auxym®' and seaweed extract (SwE) from *Ecklonia maxima* 'Kelpak®' under different N rates of 0, 10, 20 and 30 kg N·ha−1. The responses of baby lettuce plants were assessed in terms of yield, growth parameters and physicochemical composition of the leaves. The fresh yield of baby lettuce in both biostimulant-treated and untreated plants was positively affected by increasing N rates from 0 to 20 kg N·ha<sup>−</sup>1, reaching a plateau thereafter indicating luxury N conditions at 30 kg N·ha<sup>−</sup>1. However, high N fertilizer application (20 and especially 30 kg N·ha−1) resulted in undesirable decreases in antioxidant activities and total ascorbic acid (TAA). Under non-fertilized regimens, foliar PBs application boost growth and yield of baby lettuce in comparison to non-treated plants. Foliar spray with LDPH and especially SwE elicited significant increases in marketable fresh yield (averaging 14%, 6% and 7% at 10, 20 and 30 kg N·ha<sup>−</sup>1, respectively) compared to TPE and untreated plants. Improved agronomical performance of baby lettuce under optimal (10 kg N ha−1) and especially suboptimal N regimens (0 kg N ha−1) was associated with increasing photochemical efficiency and a better activity of photosystem II (higher Soil Plant Analysis Development-SPAD index and chlorophyllous pigments biosynthesis). The application of LDPH enhanced antioxidant capacity and TAA in baby lettuce leaf and did not increased nitrate content as recorded in SwE and TPE treatments. Overall, plant biostimulants may be considered as a sustainable tool of production to increase leafy vegetable productivity in low fertility soils.

**Keywords:** *Lactuca sativa* L.; legume-derived protein hydrolysate; nitrate; tropical plant extract; seaweed extract

#### **1. Introduction**

In recent years, the consumption of fresh-cut leafy vegetables has increased and among them, baby leaf lettuce is very widespread. Baby leaf lettuce is widely cultivated in Italy under both open field and greenhouse conditions [1]. Baby leaf vegetables are characterized by a short cycle but it requires a correct agronomic management to avoid high levels of nitrate accumulation and pesticide residues in the final product [2]. Therefore, there is a paramount interest in enhancing its production and quality, and at the same time reducing the nitrate levels of leafy vegetables, in order to overcome the legal limit for the marketing imposed by the European Community (Reg. n◦ 1258/2011).

With the aim to boost yield and to contain the risk of nitrate accumulation in the leaves, the research community is focusing on the use of sustainable production technologies, including application of beneficial microorganisms (Plant growth promoting Rhizobacteria, mycorrhiza and *Trichoderma*) and plant biostimulants [3–6]. In function of their origin, non-microbial plant biostimulants can be classified into five categories: (i) Seaweed extracts and microalgae, (ii) protein hydrolysate (PH) and amino acid containing products, (iii) plant extracts, (iv) humic substances and (v) silicon, with the first three groups commanding 75% of the market share [7–10]. Protein hydrolysate and amino acids containing products are normally obtained by enzymatic and/or chemical hydrolysis depending on the organic matrix (animal or vegetal) and are characterized by high percentages of amino acids and peptides, followed by carbohydrates and small amounts of micronutrients [7,8]. Moreover, plant extracts are normally produced through the fermentation of tropical plants and contain amino acids and peptides, carbohydrates but also vitamins and micronutrients with small quantities of phytohormones [8], while seaweed extracts particularly the brown macroalgae are obtained through a process called 'cold cell burst' and contain polysaccharides, osmolytes (proline and betaines), macro- and micro-nutrients, brassinosteroids and phytohormones (auxins, cytokinins and gibberellins; [11,12]).

Recent studies carried out on vegetable crops including leafy greens have demonstrated that foliar and/or root applications of plant or seaweed-based biostimulants elicit several physiological and molecular processes, thus resulting in improvements in growth, productivity, nutritional quality and nutrient use efficiency (NUE) and tolerance to abiotic stressors such as drought, soil and water salinity, nutrient deprivation and extreme temperatures [12–24]. The beneficial effects of vegetal- and seaweed-based biostimulants have been attributed to direct and indirect stimulation mechanisms [9]. The direct stimulation action of biostimulants include: (i) Activity enhancement of key enzymes involved in carbon and nitrogen (N) metabolism [13,20,25], (ii) eliciting hormone-activity in particular auxinand gibberellin-like activities through bioactive peptides [26–28], (iii) physiological, biochemical and anatomical changes such as the production of antioxidant enzymes, pigments, secondary metabolites and smallest cell guard length and width [7,8,12,29]. In addition to direct mechanisms, indirect modes of actions on agronomical performance and nutrient uptake and assimilation have been also reported when vegetal- or seaweed extract-based biostimulants were applied as substrate drench and/or foliar spray [8,29]. The better nutritional status in biostimulant-treated plants in comparison to untreated plants has been mostly associated with root system modulation (increases in root biomass, root length and diameter and lateral root branching [8,12,29]).

Among the different agronomical claims of plant biostimulants, the capacity to improve NUE in particular, N is one of the most important claims supporting their placement in the market for both economic and environmental reasons [9]. However, limited scientific literature are available regarding the effects of plant biostimulants on vegetable crops under sub-optimal N regimens [20,22,30–32]. For instance, Sestili and co-workers [20] demonstrated that the application of a PH at optimal and sub-optimal N regimens enhanced hydroponically grown tomato performance, especially substrate drench. Interestingly, the same authors observed that protein hydrolysate at low N conditions upregulated gene expression for amino acid transporter and glutamine synthetase, leading to a higher assimilation of N with a positive impact on plant growth. Similarly, Carillo et al. [22] reported that foliar application of PH, especially under suboptimal N fertilization regimes (0 or 15 kg N ha<sup>−</sup>1) boost marketable yield of greenhouse spinach due to an enhancement of nutrient acquisition and to an

increase in total amino acids in plants as well as to an improvement of photochemical efficiency, thus boosting yield.

Since there is ample evidence of species-specific response to plant biostimulants, especially that of leaf biostimulant permeability (through leaf cuticule and stomatal aperture) and thus the efficacy of the biostimulant product is species-dependent [9]; there is an urgent need among researchers to assess the effect of vegetal- and seaweed based-biostimulants on baby lettuce performance at different N fertilization regimes.

Taking into account all the previous considerations, an experiment was conducted in a plastic tunnel to assess the response of baby lettuce crop to the foliar application of the legume-derived PH 'Trainer®', tropical plant extract 'Auxym®' and seaweed extract from *Ecklonia maxima* 'Kelpak®' under different N rates of 0, 10, 20 and 30 kg N·ha<sup>−</sup>1. The responses of baby lettuce plants were assessed in terms of yield, leaf morphometric parameters and leaf quality traits.

#### **2. Materials and Methods**

#### *2.1. Experimental Setting, Plant Material and Design*

The experiment was carried out in a unheated plastic tunnel covered by polyethylene during the winter 2018 growing season (January 16—March 12) at Gussone Park, experimental site of the Department of Agricultural Sciences (40◦48.870 N; 14◦20.821 E; 70 m above sea level) located in Portici, southern Italy. The trend of daily maximum and minimum air temperature inside the plastic tunnel is reported in Figure 1. The baby leaf lettuce (*Lactuca sativa* L.) cv. 'Zarina' (ISI Sementi SpA, Parma, Italy) was used as the selected crop.

**Figure 1.** Trend of the maximum and minimum air temperature inside the plastic tunnel during the growing period of baby lettuce.

A factorial combination of N fertilization (N) and biostimulant application (B) was applied based on four increasing N fertilization levels (0, 10, 20 or 30 kg N·ha<sup>−</sup>1; N0, N10, N20 and N30, respectively) and three plant biostimulants (seaweed extract—SwE, legume-derived protein hydrolysate—LDPH and tropical plant extract—TPE) and a non-treated control. The experimental design was a randomized complete-block design with three replications, yielding 48 experimental units (4N × 4B × 3 replications) established in large lysimeters of reinforced fiber glass with a diameter of 0.70 m and a depth of 0.60 m. Each experimental unit consisted of one large lysimeters. The lysimeters were filled with a soil having the following chemical and physical characteristics reported in Table 1.


**Table 1.** Physical and chemical properties of the soil used in this work.

#### *2.2. Nitrogen Fertilization Levels, Cultural Practices and Biostimulants Application*

The baby leaf lettuce was hand seeded on January 16 at a plant density of 2500 seeds·m<sup>−</sup>2. The N was applied as calcium nitrate (26%) in a single operation 14 days after sowing. The calcium nitrate was used based on standard commercial practices used in Italy.

The three commercial SwE, LDPH and TPE-based biostimulants were made by 'Kelpak®' (Kelp Products (Pty) Ltd., Cape Town, South Africa), 'Trainer®' and 'Auxym®' (Italpollina S.p.A., Rivoli Veronese, Italy), respectively.

The SwE obtained through 'cold cell burst' mainly contained phytohormones (auxins and cytokinins with a very high auxin-to-cytokinin ratio), carbohydrates, amino acids, vitamins (B1, B2, C and E) and macro- and micro-nutrients [18,19]. The LDPH-based biostimulant contained free amino acids and peptides (75%), carbohydrates (22%) and mineral nutrients (3%). The detailed aminogram was reported by Paul et al. [33,34]. The TPE biostimulant obtained by fermentation of tropical plants contained 54% of free amino acids and peptide, 17% carbohydrate, 23% mineral nutrients, 6% vitamins and 0.22% phytohormones as reported in detail by Caruso et al. [23,24]. Baby lettuce leaf plants were sprayed with a biostimulant solution containing 3 mL·L−<sup>1</sup> of SwE and LDPH and 2·mL L−<sup>1</sup> for TPE-based biostimulant, or with water (non-treated control), five times during the growing season at 7-day intervals, starting three weeks after sowing. The relative doses of the three commercial plant biostimulants were used based on manufacturer recommendations. The volume of the solution used during the five foliar applications was 100 mL per square meter.

#### *2.3. Plant Growth Parameters, Marketable Yield, Leaf Colorimetry and Sampling*

On March 12, the baby leaf lettuce was harvested in all experimental units. The leaf area was measured using an electronic leaf area meter (Li-Cor3000, Li-Cor, Lincoln, NE, USA) in order to calculate the leaf area index (LAI). The marketable fresh yield was also measured and expressed in tons per ha, and a sub-sample was oven dried at 70 ◦C for 3 days in order to determine the leaf dry matter percentage, and the dry samples were consequently used for the mineral analysis. Furthermore, the specific leaf weight (leaf dry weight per unit area; mg·cm<sup>−</sup>2) as well as leaf succulence (leaf water content per unit area; mg·cm<sup>−</sup>2) were also recorded.

Leaf colorimetry was measured on the upper side of 10 leaves per experimental unit using Minolta CR-300 Chroma Meter (Minolta Camera Co. Ltd., Osaka, Japan) in order to obtain the color space parameters (*L*\*, *a*\* and *b*\*) and a portable chlorophyll meter SPAD-502 (Konica Minolta, Tokyo, Japan) was also used to measure the SPAD (Soil Plant Analysis Development) index.

Batch samples of fresh leaves from each experimental unit were frozen in liquid nitrogen immediately after harvest, lyophilized Christ, Alpha 1–4 (Osterode, Germany) and stored at −80 ◦C until further analysis.

#### *2.4. Antioxidant Capacity Analysis*

The lipophylic and hydrophilic antioxidant capacities were assessed on extract from freeze-dried baby lettuce leaves (200 mg) added with methanol and distilled water, respectively. The antioxidant activity of the lipophilic and hydrophilic extract fractions was measured spectrophotometrically based on the methods of Re et al. [35] and Fogliano et al. [36], respectively. The absorbance of the solutions for lipophilic and hydrophilic extract fractions were measured at 734 and 505 nm, respectively. Lipophylic and hydrophilic antioxidant activities were expressed as mmol of Trolox and mmol ascorbic acid per 100 g of dry weight (dw) [36].

#### *2.5. Chlorophyllous Pigments and Nitrate Analysis*

Chlorophyll and carotenoids content of the baby lettuce leaves were also assayed spectrophotometrically after the extraction of the fresh material (500 mg) using pure acetone as described in detail by Lichtentahler and Buschmann [37], whereas the nitrate content was determined based on the method of Sah [38]. The absorbance of the solutions for chlorophyll a and b, carotenoids and nitrate were measured at 662, 645, 470 and 550 nm. The chlorophyllous pigments were expressed as mg g−<sup>1</sup> fresh weight (fw), while the nitrate content was expressed as mg kg−<sup>1</sup> fw.

#### *2.6. Total Ascorbic Acid Analysis*

The total ascorbic acid (expressed as mg ascorbic acid on 100 g fw) was also assessed spectrophotometrically based on the protocol by Kampfenkel et al. [39]. The absorbance of the solution for total ascorbic acid was measured at 525 nm.

#### *2.7. Statistical Processing*

Morphological and qualitative data were statistically analyzed by a 2-way ANOVA using the SPSS 21 software package for Windows 2007. The means were separated by a Duncan's test (significance level 0.05).

#### **3. Results and Discussion**

#### *3.1. E*ff*ect of N Fertilization Levels and Biostimulant Application on Yield and Growth*

The results regarding morphological parameters and marketable yield of baby lettuce are reported in Figure 2; Figure 3 and Table 2. For marketable fresh yield and leaf area index (LAI) significant interaction between fertilization (F) and biostimulant application (B) was observed, whereas leaf succulence and specific leaf weight (SLW) were only influenced by the two tested factors with no significant F × B interaction (Figures 2 and 3, and Table 2). The fresh yield of baby lettuce in both biostimulant-treated and untreated plants was positively affected by increasing N fertilization levels from 0 to 20 kg N·ha−1, reaching a plateau thereafter indicating a luxury N conditions at 30 kg N·ha−<sup>1</sup> (Figure 2). The marketable fresh yield of baby lettuce at N0 was clearly higher by 19% in biostimulant-treated plants compared untreated plants, with no significant differences between the three plant biostimulants tested (Figure 2). Interestingly, foliar spray with LDPH and especially SwE elicited significant increases (average 14%, 6% and 7% at 10, 20 and 30 kg N·ha<sup>−</sup>1, respectively) compared to TPE and untreated plants (Figure 2). Similarly to the effects on marketable fresh yield, the leaf area index (LAI) in SwE and LDPH-treated plants at 10, 20 and 30 kg N·ha−<sup>1</sup> was significantly higher compared to baby lettuce treated with TPE or untreated plants, whereas under non-fertilized

conditions LAI was significantly higher in biostimulant compared to untreated plants, irrespective of the commercial biostimulants used (Figure 3).

**Figure 2.** Effects of nitrogen (N) fertilization levels (0, 10, 20 and 30 kg N·ha<sup>−</sup>1; N0, N10, N20 and N30, respectively) and biostimulant applications (untreated control, SwE: Extract of seaweed *Ecklonia maxima*, LDPH: Legume-derived protein hydrolysate and TPE: Tropical plant extract) on the marketable fresh yield of baby lettuce plants. Different letters indicate significant differences according to the Duncan's test (significance level 0.05).

**Figure 3.** Effects of nitrogen (N) fertilization levels (0, 10, 20 and 30 kg N·ha<sup>−</sup>1; N0, N10, N20 and N30, respectively) and biostimulant applications (untreated control, SwE: Extract of seaweed *Ecklonia maxima*, LDPH: Legume-derived protein hydrolysate and TPE: Tropical plant extract) on the leaf area index (LAI) of baby lettuce plants. Different letters indicate significant differences according to the Duncan's test (significance level 0.05).


**Table 2.** <sup>E</sup>ffects of nitrogen (N) fertilization levels (0, 10, 20 and 30 kg N·ha<sup>−</sup>1; N0, N10, N20 and N30, respectively) and biostimulant applications (untreated control, SwE: Extract of seaweed *Ecklonia maxima*, LDPH: Legume-derived protein hydrolysate and TPE: Tropical plant extract) on leaf succulence and specific leaf weight (SLW) of baby lettuce plants.

NS, \*, and \*\* indicate non-significant, significant at *p* < 0.05, significant at *p* < 0.01, respectively. Different letters indicate significant differences according to the Duncan's test (significance level 0.05).

When averaged over biostimulant application (F × B = not significant), the leaf succulence increased quadratically by increasing N fertilization levels from 10 to 30 kg N·ha−<sup>1</sup> with no significant difference among the three N fertilization rates, whereas the SLW declined at 20 and 30 kg N·ha−<sup>1</sup> (Table 2). Averaged over N fertilization levels, significant differentiation in terms of leaf succulence and SLW was recorded in response to biostimulants application with the higher values of succulence observed with LDPH and TPE followed by SwE as opposed to untreated plants, whereas the lowest values of SLW were recorded in baby lettuce treated with *Ecklonia maxima* extract (Table 2).

The stimulation effect of commercial biostimulants (6%–19%) recorded in the current research is in line with previous studies carried out on greenhouse fresh tomato treated with seaweed extracts of *E. maxima* or *Ascophyllum nodosum*, LDPE and TPE (7%–25%; [12,40]) but far lower than those recorded on greenhouse spinach [19]. The different stimulation effect among tested species indicates a crop-specific differential response to plant biostimulant applications and thus requires additional crop-specific studies to optimize plant biostimulants application, taking into consideration the following factors: environment, management practice and plant morphological traits (e.g., leaf permeability and cuticle thickness [9,28]).

Interestingly, LDPH (at 0 kg N·ha−1) and SwE (at 10, 20 and 30 kg N·ha−1) are likely to boost growth response and crop productivity as a consequence of the presence of bioactive molecules such as amino acids (tryptophan, glutamic and aspartic acids), soluble peptides (in LDPH) and polysaccharides (laminarans, fucoidans and alginates), phenolic compounds, osmolytes (proline, betaine and manitol) and phyohormones (abscisic acid, auxins, brassinosteroids, cytokinins and gibberellins) (in SwE) [8,29]. These former molecules present in seaweed and PH-based biostimulants may have triggered a signal transduction pathway through elicitation of endogenous phytohormone synthesis, thus leading to a higher crop productivity compared to untreated-baby lettuce plants [19,20]. Another possible mechanism of action (indirect mechanism) behind the stimulation of LAI and marketable fresh yield could be the modulation of the root system architecture in terms of root biomass, root volume and length and higher root branching triggered by tryptophan in LDPH and auxins in SwE, which improved nutrient uptake/translocation/assimilation, leading to a higher agronomical performance [12,19,41,42]. Our results are in agreement with those of Carillo and co-workers [22] who reported that foliar application of LDPH at a rate of 4 mL·L−<sup>1</sup> under suboptimal N fertilization

conditions (0 and 15 kg N·ha<sup>−</sup>1) increased the fresh yield of greenhouse spinach through an increase of the nutritional status (higher macronutrient accumulation), better photosynthetic activity and improving the total acid content.

#### *3.2. E*ff*ect of N Fertilization Levels and Biostimulant Application on Leaf Colorimetry and SPAD Index*

Among the physical properties that may affect the purchasing decisions of vegetable consumers is product appearance, in particular, the color of the vegetable [43]. In the present study, no significant interaction between N fertilization and biostimulants application was recorded for the three leaf colorimetric parameters lightness (*L*\*), green color intensity (negative values of *a*\*) and yellow color intensity (positive values of *b*\*) (Table 3). The colorimetric CIELAB components *L*\* and *b*\* were significantly influenced by the two tested factors, whereas *a*\* was only affected by N fertilization levels (Table 3). Increasing the N fertilization levels from 0 to 30 kg N·ha−<sup>1</sup> yielded lighter baby lettuce leaf expression by increasing *L*\* values, but with a decrease in *a*\* values (Table 3). Moreover, when averaged over N fertilization levels, the foliar application of SwE and TPE-based biostimulants elicited an increase in *L*\* values compared to the untreated control, whereas LDPH treatment exhibited intermediate values (Table 3).

**Table 3.** Effects of nitrogen (N) fertilization levels (0, 10, 20 and 30 kg N·ha<sup>−</sup>1; N0, N10, N20 and N30, respectively) and biostimulant applications (untreated control, SwE: Extract of seaweed *Ecklonia maxima*, LDPH: Legume-derived protein hydrolysate and TPE: Tropical plant extract) on leaf hunter color parameters of baby lettuce plants.


NS, \*, and \*\* indicate non-significant, significant at *p* < 0.05, or significant at *p* < 0.01, respectively. Different letters indicate significant differences according to the Duncan's test (significance level 0.05).

Interestingly, the foliar application of commercial plant biostimulants improved the SPAD index significantly; this is an important physiological parameter having a crucial role on the primary metabolism of plants. With the exception of under N20, where no significant difference in the SPAD index was observed, between biostimulants-treated and untreated plants, the foliar application with SwE (at 10 and 30 kg N·ha<sup>−</sup>1) and with the three commercial biostimulants (at 0 kg N·ha<sup>−</sup>1) incurred a significant increase in the SPAD index (Figure 4). Our findings have been also demonstrated in many leafy vegetable species such as jute, lettuce, and spinach [16,19,22]. The highest SPAD values observed after the application of plant biostimulants in particular extracts from brown macroalgae could be attributed to several putative mechanisms like the following: (i) better translocation of soluble sugars via the phloem, (ii) increases in the biogenesis of chloroplast, as well as (iii) limited chlorophyll degradation, and thus, delayed senescence [29,44,45].

**Figure 4.** Effects of nitrogen (N) fertilization levels (0, 10, 20 and 30 kg N·ha<sup>−</sup>1; N0, N10, N20 and N30, respectively) and biostimulant applications (untreated control, SwE: Extract of seaweed *Ecklonia maxima*, LDPH: Legume-derived protein hydrolysate and TPE: Tropical plant extract) on the SPAD index of baby lettuce plants. Different letters indicate significant differences according to the Duncan's test (significance level 0.05).

#### *3.3. E*ff*ect of N Fertilization Levels and Biostimulant Application on Nitrate Accumulation and Biochemical Parameters*

Nitrate was affected by both N fertilization levels and biostimulant application, without significant F × B interaction (Table 4). The nitrate concentration in baby lettuce leaf was negatively affected by N fertilization levels. Increasing the N fertilization from 0 to 30 kg N·ha−<sup>1</sup> increased the nitrate accumulation in leaves, especially at 20 and 30 kg N·ha−1, where the content of nitrate was above the upper limits set by the European Union (EU) for safe lettuce marketing (Commission Regulation No. 1258/2011; 3000 to 5000 mg NO3 −·kg−<sup>1</sup> of lettuce depending on growing season and cultivation conditions). On the other hand, when averaged over N fertilization levels, the nitrate concentration in LDPH treated plants was significantly lower on average by 21.2% compared to baby lettuce treated with SwE or TPE and it was not significantly different than untreated-baby lettuce plants (Table 4). The capacity of LDPH, which is mainly composed of soluble solids and especially amino acids, to accumulate less nitrate in the leaf tissue, could be attributed to a molecular mechanism such as the up-regulation of genes involved in N metabolism such as nitrate reducatse, and consequently, to an augmenting assimilation of nitrates into amino acids [46,47]. Furthermore, other studies conducted by Sady and Smole ´n [31] and Smole ´n and Sady [32] on carrots and spinach, respectively, reported that after the foliar application of 'Pentakeep V' containing 5-aminolevulinic acid was able to reduce nitrate accumulation in combination with a 50% N dose, whereas an opposite trend was observed in combination with 100% N. The authors concluded that nitrate accumulation in response to biostimulant application may change in relation to several interacting variables including species, variety and N application rates.



NS, \*, and \*\* indicate non-significant, significant at *p* < 0.05, or significant at *p* < 0.01, respectively. Different letters indicate significant differences according to the Duncan's test (significance level 0.05).

One of the beneficial responses of plant biostimulants application is an increase in chlorophyllous pigments such as chlorophyll a, b and total, as well as carotenoids. This was the case in the current research study, since the foliar application of LDPH and TPE incurred a significant increase in chlorophyll a and b and consequently the total chlorophyll compared to SwE and untreated-baby lettuce plants (Table 4). Furthermore, the content of carotenoids was positively affected by the foliar application of SwE and LDPH compared to the control treatment (Table 4). This beneficial effect of vegetal and seaweed extract-based biostimulants on carotenoids and especially chlorophyll content has been recorded also in corn, jute and eggplant [21,22,48,49]. The increase in chlorophyll a and total content in both LDPH and TPE (characterized by the high percentage of free amino acids (75% and 54%, respectively [23,24]) could be attributed to the higher content of primary amino acids in the vegetal-based treated plants as amino acids (e.g., alanine, aspartate, asparagines and glutamate) which help to boost chlorophyll content, and consequently, increase photosynthetic activity as well as the quantum yield of O2 evolution [22].

#### *3.4. E*ff*ect of N Fertilization Levels and Biostimulant Application on Antioxidant Capacity and Bioactive Content*

Lipophilic (LAA) and hydrophilic (HAA) antioxidant activities as well as total ascorbic acid (TAA) were significantly affected by both factors with a significant F × B interaction (Table 5). Antioxidant scavenging activity was an important functional quality parameter in assessing the nutritional properties of foods including leafy vegetables, since lipophilic (e.g., β-carotene, lutein, α-tocopherol, etc.) and hydrophilic (e.g., vitamin C, caffeic acid, ferulic acid, quercitin, etc.) antioxidant molecules impart beneficial effects to human health, as these bioactive molecules are known to play a primary role in delaying oxidative damage, thus, preventing a wide range of diseases [50–54]. In the current study, LAA, HAA and TAA of the baby lettuce ranged from 19.9 to 32.3 mmol trolox 100·g−<sup>1</sup> dw, from 3.0 to 8.2 mmol ascorbic acid 100 g−<sup>1</sup> dw and from 6.8 to 33.4 mg g<sup>−</sup>1, respectively (Table 5).



\*, \*\* significant at *p* < 0.05 and 0.01, respectively. Different letters indicate significant differences according to the Duncan's test (significance level 0.05).

High N fertilizer application (20 and especially 30 kg·N ha<sup>−</sup>1) resulted in undesirable decreases in HAA and TAA of baby lettuce leaves (Table 5), which is in agreement with the results of Wang et al. [55] who reported that high N fertilization levels can result in undesirable changes in the quality attributes of fruit and leafy vegetables such as soluble solids and ascorbic acid leading to a decrease in commercial, nutritional and functional quality.

The vegetal- and seaweed extract-based biostimulants applied to baby lettuce resulted in higher antioxidant capacity and bioactive content depending on the N fertilization levels compared to untreated control treatment. For instance, at N0 the highest antioxidant activities and TAA compared to the untreated control were recorded in baby lettuce treated with LDPH and SwE-based biostimulant plants, respectively, whereas at N10 and N30 the highest LAA and HAA contents were observed in TPE and LDPH treated plants, respectively (Table 5). Our findings on the effect of plant biostimulants on nutritional and functional quality of the product were in line with previous research on vegetal-based biostimulants (protein hydrolysate and plant extract) conducted by Caruso et al. [23], in which foliar application at weekly interval increases the LAA, HAA and TAA contents of perennial wall rocket compared to the non-treated control. Similarly, Vasantharaja et al. [56] demonstrated that the application of seaweed extract-based biostimulant (*Sargassum swartzii*) boosted the antioxidant activity and the bioactive content (e.g., phenols and vitamin C) of cowpea. A mechanistic explanation of the beneficial effect of plant biostimulants, in particular LDPH, on the biosynthesis of antioxidant molecules could be due to: i) the activity stimulation of key enzymes involved in antioxidant homeostasis in cells, and ii) the higher macro- and micro-nutrient assimilation of biostimulant-treated plants which could contribute to the synthesis of amino acids, phenylalanine and tyrosine [7,40].

#### **4. Conclusions**

The idea of working with plant biostimulants to increase yield under both optimal and suboptimal conditions is gaining interest among leafy vegetable growers, as well as private companies and researchers for both economic and environmental reasons. The foliar application of vegetal and seaweed extract-based biostimulants, in particular SwE and LDPH enhanced plant growth, and productivity especially under sub-optimal N regimens, and to a lesser extent; at 20 and 30 kg <sup>N</sup>·ha<sup>−</sup>1. The foliar application of SwE and LDPH was effective in supporting better physiological and biochemical status in terms of the SPAD index, chlorophyll and carotenoids content leading to a higher agronomical performance. Interestingly, the leaf quality traits of baby lettuce leaf can be improved by biostimulation action, especially with LDPH which delivered leaves with high antioxidant activity and total ascorbic acid as well as low nitrate content. The results of the current experiment highlight the benefit of using vegetal and seaweed extract-based biostimulants in baby lettuce to improve productivity under both optimal and especially suboptimal N regimens, bringing benefits to farmers and to the environment.

**Author Contributions:** Conceptualization, M.M.; methodology, I.D.M., E.C.; software, L.O.; validation, M.M., I.D.M., E.C. and L.O.; formal analysis, L.O.; investigation, I.D.M.; resources, M.M.; data curation, I.D.M. and L.O.; writing—original draft preparation, I.D.M.; writing—review and editing, Y.R.; visualization, M.G. and G.C.; supervision, G.C. and M.M.; project administration, M.M.; funding acquisition, M.M.

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

#### **References**


© 2019 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*

### **Seed Coating with Thyme Essential Oil or** *Paraburkholderia phytofirmans* **PsJN Strain: Conferring Septoria Leaf Blotch Resistance and Promotion of Yield and Grain Isotopic Composition in Wheat**

**Maissa Ben-Jabeur 1,\*, Zayneb Kthiri 1, Kalthoum Harbaoui 2, Karima Belguesmi 2, Maria Dolores Serret 3, Jose Luis Araus 3,\* and Walid Hamada 1,\***


Received: 13 July 2019; Accepted: 18 September 2019; Published: 26 September 2019

**Abstract:** Septoria leaf blotch (SLB) is considered one of the most devastating diseases affecting global wheat production. Biostimulant application is among the modern approaches in plant protection to overcome the impact of SLB's fungicide resistance. In this manner, the effect of coating seeds with thyme essential oil or *Paraburkholderia phytofirmans* PsJN strain on SLB severity and yield components (spikes/m2, straw yield (SY), grain yield (GY) and thousand kernel weight (TKW)) were assessed under field conditions for 3 years. The effect on physiological traits and nitrogen and carbon isotope composition (δ15Ngrain, δ13Cgrain) and nitrogen and carbon content (Ngrain, Cgrain) of grains was assessed in one year of study. The increasing SLB severity decreased all yield components, increased δ15Ngrain and Cgrain content and slightly decreased δ13Cgrain as the resulting effect of *Zymoseptoria tritici* inducing stomatal opening and leaf necrosis. Across the years, both treatments alleviated the SLB adverse impact by reducing SLB severity, increasing spikes/m2, SY, GY and TKW. Both treatments ameliorated grain quality by increasing Cgrain content and decreasing δ13Cgrain and δ15Ngrain. The difference between the performance of thyme oil or PsJN strain in terms of intensity and stability is discussed and considered to be linked to the different triggered systemic resistance and the associated amount of costs deriving from resource allocation towards defense processes.

**Keywords:** Septoria; wheat; *Paraburkholderia phytofirmans*; thyme essential oil; isotope

#### **1. Introduction**

Globally, wheat leads all crops in terms of cultivated area and continues to be the most important food grain source for humans [1]. The high consumption of hard (or durum) wheat in some countries is associated with a decrease in wheat production resulting from ongoing climate change causing a rise of drought stress and the emergence of more aggressive pathogens [2], which leads to above-average imports to meet needs for consumption. Septoria leaf blotch (SLB), caused by the hemibiotroph *Zymoseptoria tritici*, constitutes one of the major constraints affecting durum wheat global production resulting in yield losses [3] and shriveled grains, which is undesirable for industries as they result in low flour extraction rates in milling and provide poor quality for feeding livestock [4]. Since the introduction of fungicides in the 1980s, chemical control is currently one of the main approaches used to manage SLB [3,4]. However, fungicide resistance and its associated environmental impact is now a widespread problem [5].

Biostimulants are considered as products modifying biochemical and physiological processes in plants, neutralizing the adverse impact of weather conditions and reducing the occurrence of diseases by stimulating plant growth, strengthening plant defenses and improving nutrition efficiency leading to sustainable crop yield [6]. In this context, this study's interest focused towards assessing the effect of the biostimulants thyme oil and *Paraburkholderia phytofirmans* PsJN strain against SLB severity via the seed coating technique. Our previous experiments revealed that seed coating with both agents induced seed priming associated with increased germination, the emergence of seedlings, shoot and root development, and a decreased root to shoot ratio [7]. Moreover, coating seeds with either thyme oil or *P. phytofirmans* revealed great potential in controlling SLB under controlled conditions [8]. Thyme oil and PsJN strain differed in their mode of action. Thyme oil induced systemic programmed cell death (PCD) with higher frequency of formed papillae, high peroxidases activity and H2O2 amount, and low catalase and phenolic compounds, indicating systemic acquired resistance (SAR), and the necrotic area was reduced to 30% with reduced pycnidial density to 1.8%. While PsJN strain encountered hyphae and condensate for biofilm formation, the induced local PCD with less frequency of formed papillae, low peroxidases activity and H2O2 amount, and low catalase and phenolic compounds, indicated induced systemic resistance (ISR), and the necrotic area was reduced to 10% with reduced pycnidial density to 9.4%. Despite the potential of biostimulants in achieving disease control under controlled conditions, their performance under field conditions could be less imposing. Hence, the effect of thyme essential oil and PsJN strain under field conditions on SLB severity, yield components and carbon and nitrogen stable isotope composition in durum wheat grains are examined.

#### **2. Materials and Methods**

#### *2.1. Plant Material*

A Tunisian variety of durum wheat (*Triticum turgidum* L. subsp. *Durum* (Desf) Husn.); 'Karim', known for its sensitivity to SLB, was used.

#### *2.2. Seed Coating Treatment*

Just before sowing, the seeds were coated with either thyme essential oil or *Paraburkholderia phytofirmans* PsJN strain. Thyme essential oil was extracted by hydro distillation from dried aerial parts of *Thymbra capitata* (L.) Cav. (chemotype carvacrol, voucher specimen D 1186-3), and harvested during the flowering stage from the plain of Kef (Tunisia, 36◦23 N, 8◦79 E). The obtained essential oil was distributed into 1 mL-amber-glass vials and stored at 4 ◦C for subsequent use. The chemical composition of the oil was investigated and carvacrol was identified as the major compound according to Ben Jabeur et al. [9]. The concentration of thyme oil was adjusted to 5 ppm before use with adding 0.5% of dimethyl sulfoxide (DMSO) as a solubilizing agent to assure the homogenous application of the essential oil. The bacterial inoculum of *P.phytofirmans* PsJN strain (provided by Pr. Ait Barka, University of Reims, France) was produced by transferring one colony to 20 mL of King's B liquid medium, incubated at 27 ◦C at 150 rpm for 48 h. The bacteria were collected by centrifugation at 8000 rpm for 5 min and washed and the concentration was adjusted to 108 CFU.mL−<sup>1</sup> before use with phosphate-buffered saline (PBS) (10 mM, pH 6.5). The coating product Agicote Rouge T17 (AEGILOPS Applications, Val de Reuil, France), specific for cereal seeds, containing propane-1,2-diol (5–10%), polyethylene glycol mono(tristyrylphenyl)ether (5–10%), and 1,2-benzisothiasol3(2H)-one (0.0357%), was used [10]. The coating technique consists of preparing the appropriate volume of the coating solution mixture based on the quantity of seeds required for each experimental plot. Each 10 g of wheat seeds required 40 μL of the coating product Agicote Rouge T17 and 400 μL of either thyme oil (5 ppm) or PsJN inoculum (108 CFU.mL<sup>−</sup>1), (400 μL of water was used as a control). Then, the coating

mixture was applied progressively to wheat seeds in a continuous rotation, using a portable rotating drum apparatus (SUNCOO, Atlanta, GA, USA) with a speed of 2800 rpm, at an ambient temperature (20 ± 2 ◦C) until complete adhesion and absorption, to assure the homogeneous distribution of the coating mixture among the seeds. The final concentration of products per seed was 10−<sup>5</sup> μL of coated thyme oil/seed and 210<sup>4</sup> CFU of coated PsJN strain/seed. Prior to the evaluation of the effect of coating seeds with thyme oil, the effect of the coating product was evaluated in the laboratory. The positive or negative effects of the coating product on seed germination and seedling growth were not detected and its inertness was assured.

#### *2.3. Experimental Design for Field Trials*

The experiments were conducted at the experimental station in Oued-Beja (CRGC), located in the sub-humid bioclimatic zone of Tunisia, for three years; 2015–2016 and 2017 under rainfed conditions (Table 1). The soil type of the experimental area is mostly clay loam with pH 7.2 (Table 2). A complete random block design with three replicates was used. The plots size was 1 × 3 m spaced by 1.5 m. Each plot consisted of 6 rows; with a row spacing of 0.15 m. The sowing was carried out in the first week of December at a sowing density of 350 seeds /m2. The plants were inoculated with 107 spores/ml of *Z. tritici* twice. After full emergence of the third leaf and at stem elongation, a CO2-pressurized knapsack sprayer was used. Nitrogen (ammonium nitrate) was applied at 25 kg N/ha at sowing and at the stem elongation stage.

**Table 1.** The climatic conditions (temperature, precipitation, humidity) of the three years in the experimental station of Oued Beja.


**\*** Sums for precipitation; average values for the rest.

**Table 2.** Soil's physicochemical characteristics of Oued Beja station.


#### *2.4. E*ff*ect of Seed Coating with PsJN Strain and Thyme Oil on Plant Physiology, Disease Control and Yield Components*

At anthesis, five leaves within each plot were selected for nondestructive measurements of leaf chlorophyll content, using a portable meter (SPAD 502 plus, Minolta, UK), and stomatal conductance of the flag leaf with a leaf porometer (Decagon, Pullman, Washington, USA). In addition, the following measurements were performed for each plot at the canopy level: The canopy normalized difference vegetation index (NDVI), with a spectroradiometer (GreenSeeker@Trimble, Westminster, Colorado, USA), canopy temperature using an infrared thermometer (Fluke, Everett, Washington, USA). For disease scoring, 15 plants were sampled from each plot, all leaves were taken for assessing the vertical disease progress and estimated for severity according to Eyal et al. [11]. Since the difference in vertical

disease progress upon the samples was not observed, the diseases assessment was conducted at the leaf numbered flag leaf-3, the highest leaf showing symptoms. The leaves were scanned, and the images were analyzed using ImageJ software (the National Institute of Mental Health, Bethesda, MD, USA). The extent of the necrotic area was determined, according to Stewart and McDonald [12]. Briefly, the background was removed from each image and the total leaf area and green leaf area in the pixel was calculated using color thresholding in the red-green-blue (RGB) color space as formulated: Septoria severity (%) = (total leaf area-green leaf area)/total leaf area <sup>×</sup> 100. At harvest, 1 m2 of each plot was hand harvested, and then straw yield (SY, Mg ha<sup>−</sup>1), number of spikes/m2, thousand kernels weight (TKW, g) and grain yield (GY, Mg ha<sup>−</sup>1) were measured.

#### *2.5. E*ff*ect of Seed Coating with PsJN Strain and Thyme Oil on Total Nitrogen and Carbon Content and Stable Carbon and Nitrogen Isotope Composition*

The total N and C content and the stable nitrogen isotope signature in the dry matter of the mature grains sampled from each plot of the third field trial (2017) were analyzed at the Scientific Facilities of the University of Barcelona. Approximately 1mg of each sample and reference materials were weighed into tin capsules and measured with an elemental analyzer (Flash1112EA; Thermo Finnigan, Bremen, Germany) coupled with an isotope ratio mass spectrometer (Delta CIRMS, Thermo Finnigan, Bremen, Germany) operating in continuous flow mode in order to determine the total C and N content and the stable carbon (13C/ 12C) and nitrogen (15N/ 14N) isotopes' ratios. The ratios were expressed in δ notation [13], as δ13C = ( 13C/ 12C) sample/ ( 13C/ 12C) standard <sup>−</sup>1, where sample refers to the plant material and standard to Pee Dee Belemnite (PDB) calcium carbonate, and as δ15N = ( 15N/ 14N) sample/ ( 15N/ 14N) standard <sup>−</sup>1, where sample refers to plant material and standard refers to N2 in air.

#### *2.6. Statistical Analysis*

The effects of the treatments and years and their interaction on SLB severity and yield components were determined through a two-factor (treatment × year) analysis of variance (ANOVA) with RStudio 1.1.463 (R Foundation for Statistical Computing, Vienna, Austria). The effects of the treatments on physiological traits, yield components and grain stable isotopic compositions were determined through a one-factor ANOVA (treatment). The least significant difference (LSD) test was used to assess the differences between the treatment means. The clustered Pearson correlation matrices were generated in the RStudio environment using the mean values of all traits to study the relationships between all parameters analyzed within each treatment. The data of the non-inoculated control and inoculated control were correlated (Figure 1, IC) assessing for relationship between traits in wheat-*Z. tritici* interaction. The data of the inoculated control and plants treated with PsJN strain were correlated (Figure 1, CB), and the data of the inoculated control and plants treated with thyme oil were correlated (Figure 1, CT) for extracting the potential mode of action of each treatment in conferring disease resistance and yield improvement.

**Figure 1.** A correlation matrix for physiological traits, yield components and grain stable isotope composition (2017 year of study). Treatments; IC: inoculated control, CB: coated with PsJN strain,

CT: coated with thyme oil. Traits; 13 C: δ13Cgrain, 15N: δ15Ngrain, C: Cgrain, N: Ngrain, GY: Grain yield, SY: Straw yield, TKW: Thousand kernel weight, CT: Canopy temperature, SLB: SLB severity. The darker, bigger blue squares indicate a stronger positive correlation. The darker, bigger brown squares indicate stronger negative correlation.

#### **3. Results**

#### *3.1. Climatic Features and Sources of Variances of 3 Years of Study*

The data in Table 1 show that the experimental season 2016 is the season most favoring SLB compared to the other seasons tested. It was characterized by a higher amount of annual precipitations, lower maximal temperatures and high humidity. By contrast, the experimental season 2017, was characterized by drier weather due to a lower amount of precipitation, a higher maximal temperature and lower humidity. In fact, the analysis of variance revealed a highly significant (*p* < 0.001) effect for SLB severity (%), straw yield (SY) and grain yield (GY), and thousand kernel weight (TKW) was also significantly (*p* < 0.01) affected between the years. The effect of treatment (T) and the interaction year x treatment (Y × T) was highly significant (*p* < 0.001) for all four traits (Table 3).

#### *3.2. E*ff*ect of SLB Severity on Wheat Yield Components in Control Plants*

SLB was spotted in the non-inoculated control. Nevertheless, SLB severity was less compared to the inoculated control (Table 3). Therefore, a comparison between the inoculated control and non-inoculated control revealed that field artificial inoculation of wheat with *Z.tritici* increased SLB severity over naturally occurring levels, facilitating the study of the effect of treatment on wheat yield under infested conditions. Furthermore, SLB severity varied according to the variability in climatic conditions between the years. The highest severity occurred at the driest season 2017. SLB decreased significantly all yield components of the cultivar 'Karim' specifically and compared with the control. The grain yield reduced by 0.2, 0.3, and 0.5 Mg ha−<sup>1</sup> in 2015, 2016, and 2017 respectively.

#### *3.3. E*ff*ect of Seed Coating Treatment on SLB Severity and Yield Components*

Both treatments showed a great potential in controlling SLB under field conditions (Table 3). The plants coated with thyme oil reduced SLB severity by 22%, 25.5%, and 53.2% in 2015, 2016, and 2017 respectively compared to the inoculated control. The plants coated with PsJN strain reduced SLB severity by 30%, 24%, and 48.3% in 2015, 2016, and 2017 respectively compared to the inoculated control. In season 2015, when water availability was high, PsJN strain was more efficient than thyme oil in reducing SLB severity. In seasons 2016 and 2017, when water availability decreased, thyme oil was more efficient than PsJN strain in reducing Septoria severity. In fact, a significant treatment by year interaction was observed for SLB.

The treatment with PsJN strain increased all yield components in the 3 seasons, not only with regard to the inoculated control but also compared with the non-inoculate control (Table 3), and the increased intensity varied among the years, most likely due to environmental factors. Contrastingly, thyme oil increased TKW compared to the inoculated control and decreased it compared to the non-inoculated control in all seasons. Furthermore, thyme oil had different effects on GY and SY among the 3 years. In 2015, in which the rainfall was more abundant in the vegetative growth stage (December–February) than the grain filling stage (April), thyme oil increased SY and decreased GY. In 2016, in which rainfall was limited in the vegetative growth stage (December-February) and abundant at the heading and anthesis (March), thyme oil decreased SY and increased GY. In 2017, in which rainfall was abundant in both vegetative growth stage (December–February) and grain filling stage (April), thyme oil increased both SY and GY.


**Table 3.**Effect of treatments on SLB severity and yield components of durum wheat evaluated in three-year-study. The F values are shown, and the symbols indicate statistical significance (\*\*, *p* < 0.01; \*\*\*, *p* < 0.001), values with different superscript letters are significantly different classes accordingthe LSD test (*<sup>p</sup>* ≤ 0.05).LSD: least significant difference; SLB: Septoria leaf blotch; SY: Straw yield; GY: Grain yield; TKW: Thousand Kernels weigh; NIC: non-inoculated control;inoculatedcontrol;CB:coatedwithPsJNstrain;CT:coatedwiththymeoil.

 IC:

#### *3.4. E*ff*ect of Seed CoatingTreatment and SLB Severity on Physiological Traits, Yield Components and Grain Isotopic Composition*

#### 3.4.1. Effect of SLB in Control Plants

On the control plants inoculated with *Z.tritici*, during vegetative growth, the green leaf area was reduced compared with the other treatments (Figure 2), as shown by the reduction in the canopy vegetation index NDVI, and the decrease in leaf chlorophyll content (SPAD), while stomatal conductance increased and the carbon isotope composition (δ13C) of the grains slightly decreased. At harvest, SLB severity caused a reduction in GY and biomass as well as in the yield components spikes/m<sup>2</sup> and TKW and altered the grain composition by increasing Cgrain content and δ15Ngrain (Table 4). SLB had no effect on Ngrain content. The behavior of *Z.tritici*, the effect of SLB on the wheat physiological state, and the impact on yield components and grain composition was confirmed by the negative correlation between traits in cluster 1: SPAD, NDVI, spikes/m2, GY, SY, TKW, canopy temperature, δ13Cgrain and the traits in cluster 2: SLB severity, stomatal conductance, Cgrain content, δ15Ngrain (Figure 1, IC).

#### 3.4.2. Effect of Seed Coating with PsJN Strain

Disease resistance was observed and characterized by a higher green leaf area (Figure 2) and SPAD values, and lower SLB severity and stomatal conductance compared to the inoculated control (Table 4). The plant growth promoting effect of coating seeds with PsJN strain was remarkably observed from (i) an increase in SPAD, and NDVI in the vegetative growth phase and increase in SY, GY, TKW at harvest (Table 4), and (ii) the positive correlation among SY, GY, SPAD, spikes/m2, TKW, NDVI (Figure 1, CB, cluster 1). Concerning grain composition, the coating with PsJN strain and Cgrain content was positively correlated to SY, GY, SPAD, spikes/m2, TKW, NDVI (Figure 1, CB, cluster 1), and decreased δ15Ngrain and δ13Cgrain, which is most likely related to a lower canopy temperature, stomatal conductance, and SLB severity compared to the inoculated control (Figure 1, CB, cluster 2). No effect was observed on Ngrain content.

#### 3.4.3. Effect of Seed Coating with Thyme Oil

Disease resistance was observed and characterized by a higher green leaf area (Figure 2) and SPAD values, lower SLB severity, and a lower stomatal conductance, resulting in a higher canopy temperature compared to the inoculated control. Coating seeds with thyme oil increased GY, SY, spikes/m<sup>2</sup> and TKW compared to the inoculated control (Table 4). Concerning grain composition, thyme oil increased Cgrain content which was positively correlated to GY, SY, spikes/m2, TKW, canopy temperature and SPAD. The effect of thyme oil on decreasing δ13Cgrain and δ15Ngrain content is most likely related to an increase in stomatal conductance mediated by a lower SLB severity (Figure 1, CT, cluster 2) and NDVI was the less correlated trait. No effect was observed on Ngrain content.


#### **4. Discussion**

#### *4.1. E*ff*ect of Climate on Variability of SLB Severity, Yield Components among the 3 Years*

Despite the less favoring conditions for disease development in the dry season 2017, SLB severity was the highest. In fact, one of the fundamental concepts in plant pathology illustrates that plant disease occurrence requires a three-way interaction of a susceptible host, a virulent pathogen and an environment suitable for disease development, which is referred to as the disease triangle [14].The drought and temperature stresses, associated with climatic change as well as anthropogenic air pollutants as is the case of elevated O3 levels, have the potential to: (i) Accelerate tissue necrosis favoring infection by necrotrophic pathogens, drawing nutrients from dead host tissues; (ii) reduce the major plant defense processes against pathogens due to reduced photosynthate production and the activation of the ABA-responsive signaling pathway [15,16]. SLB significantly decreased straw yield, grain yield and the yield components of the cultivar 'Karim' specifically and compared with the control in the three years of study, which agrees with the sensitive attitude of this cultivar reported [17].

#### *4.2. E*ff*ect of SLB on PhysiologicalTraits, Yield Components, and Stable Isotopic Composition*

SLB was spotted in the non-inoculated control due to the natural aerial epidemics in the experimental station zone considered as a hot spot for SLB [17]. In season 2017, the green status of plants (SPAD and NDVI) decreased with the increasing SLB severity as expected since symptoms of SLB involve chlorotic and necrotic lesions in leaves, thus reducing the green leaf area. Furthermore, SLB caused a decrease in canopy temperature (CT) and an increase in stomatal conductance (SC). This constitutes a part of *Z.tritici* hemibiotrophic behavior causing early malfunction of stomatal regulation through the stimulation of a stomatal opening leading to an increase in the transpiration rate and energy dissipation, and the subsequent decline of canopy temperature [18]. All these metabolic modifications provoked by SLB are thought to contribute to the decreasing grain yield, straw yield, number of spikes/m2, and the decreasing grain quality through the modification of TKW, δ15Ngrain, C grain and δ13C grain.

Carbon content in grains is derived from photosynthetic fixation occurring during grain filling, from diffusion of CO2 from the air into the leaves (and the non-laminar parts) through stomata and carboxylation by Rubisco, and from earlier-assimilated carbon remobilized from vegetative organs [19]. Through these enzymatic and physical processes, C3 plants discriminate against 13C in favor of 12C leading to lower δ13C/δ12C ratio [20]. The values of the δ13C/δ12C ratio in C3 plants have been shown to vary depending on the balance between CO2 diffusive supply (stomatal conductance) and the enzymatic demand for CO2 (net photosynthetic assimilation), which defines the intercellular versus atmospheric ratio of CO2 (Ci/Ca) in the photosynthetic organ [19–21]. In this context, multiple mechanisms could be involved in the alteration of carbon metabolism by SLB, decreasing δ13Cgrain content and increasing Cgrain content: (i) The induced stomatal opening by SLB results in an increase of CO2 supply to carboxylation sites; (ii) during the long latent biotrophic period, and referred as the symptomless growth phase, the pathogen suppresses the plant defense response which consumes the carbon skeleton components resulting in an increase in the carbon reserve [22]; (iii) during the necrotrophic phase, the pathogen releases the early suppressed plant defense resulting in the accumulation of ABA responsible for increasing the carbohydrate content in leaves and for enhancing their remobilization to grains [22,23]; (iv) in the necrotrophic phase, the pathogen causes a decrease in the photosynthetic capacity associated with less chlorophyll resulting in an increase in the Ci/Ca ratio, therefore decreasing the δ13C [24]. On the other hand, the nitrogen content in grains is derived from direct nitrogen assimilation from roots during grain filling and from remobilization of earlier-assimilated nitrogen from vegetative organs to developing grains [25]. The natural variation of the stable nitrogen isotopes 15N/ 14N assessed through the nitrogen isotope composition (δ15N) is linked to nitrogen sources used by the plant (NH4 <sup>+</sup> uptake will induce 15N enrichment compared to NO3 −), to the activity of enzymes involved in the assimilation of ammonium (glutamine synthetase, GS) or nitrate (nitrate reductase, NR),

to the nature of compounds resulting from nitrogen fractionation. Proteins are generally 15N enriched compared to chlorophyll, lipids, amino sugars and alkaloids [26], and to volatilization, translocation, or nitrogen recycling in the plant [25]. SLB, decreasing δ15Ngrain ( 15N/ 14N) and not influencing total Ngrain content at the same time, suggests that SLB both increased the isotopic fraction 15N and decreased the isotopic fraction 14N. In this context, multiple mechanisms could be involved in the decrease of the isotopic fraction 14N by SLB: (i) During the long latent biotrophic period, pathogens successfully acquire primary and secondary nitrogen sources available in the living tissues by enzymatic digestion of host cell walls, by invading neighboring cells, or by inducing nutrient leakage from the surrounding tissues [27] resulting in decreased 14N leaf storage in the vegetative growth stage; (ii) at the metabolic level, *Z.tritici* causes a decrease in N assimilation and remobilization via reducing the activity of the enzymes NR, GS and GDH starting from the first phase of infection leading to decreased 14N leaf and a resulting decrease in 14Ngrain [28]; (iii) SLB causing chlorotic and necrotic lesions induce N retention in the diseased plant parts, thus decreasing N remobilization to grain resulting in decreased 14Ngrain [29]; (iv) stomatal-opening induced by *Z.tritici* can cause an increase in N compounds volatilization resulting in decreased 14N leaf storage, thus a decrease in later 14Ngrain content [26]. Moreover, the mechanism involved in the increase of the isotopic fraction 15Ngrain tends to be the effect of SLB on increasing grain protein (15N enriched) content as a consequence of the loss of photosynthetic leaf area and, therefore, of carbohydrate availability to the developing grain [26,30].

#### *4.3. E*ff*ect of PsJN Strain*

Coating seeds with PsJN strain showed a great potential for controlling SLB under field conditions in the three years of study and tends to be the most stable treatment by increasing all yield components (GY, SY, spikes/m2and TKW) despite the different climatic conditions. Disease resistance was associated to the alleviation of the plant damage induced by *Z.tritici* behavior characterized by less stomata openings and enhanced chlorophyll pigmentation observed in the 2017 year of study. This could be referred to the bacterial direct effect in altering the fungal development, and the indirect effect in triggering induced systemic resistance (ISR) within the plant tissues and promoting shoot and root growth [8]. The increase in photosynthesis (SPAD) and yield components is thought to be related to the effect of PsJN strain on: (i) Inducing seed priming resulting in metabolic changes that involve phenolic compound accumulation and growth promotion of root and shoot parts starting from the seedling emergence stage [7]; (ii) decreasing the plant ethylene level by decreasing ACC levels in plants via the bacterial 1-aminocyclopropane-1-carboxylate (ACC) deaminase activity resulting in a delay of senescence and prolonged photosynthetic activity of green tissue [31]; (iii) producing the growth regulator indole 3-acetic acid (IAA) that stimulates the development of the root system, thereby increasing nutrient absorption [32].

More specifically, in a way to understand the effect of the interaction PsJN strain-*Z.tritici* on carbon and nitrogen metabolism, the total carbon content (Cgrain) and fractionation (δ13C, δ15N) were analyzed in the grains. The effect of PsJN strain on decreasing δ13Cgrain and increasing Cgrain content compared with the inoculated and the non-inoculate controls suggests that this effect is mostly related to its potential in improving the plant water status due to the enhanced root development conferring a higher amount of captured water [7,24]. The effect of PsJN strain on decreasing δ15Ngrain compared to the inoculated control and simultaneously not influencing total Ngrain content suggests that PsJN strain both increases the isotopic fraction 14Ngrain and decrease the isotopic fraction 15Ngrain and could be interpreted as: (i) The enhanced N uptake and assimilation during vegetative growth and remobilization during grain filling leading to increased 14Ngrain [25]; (ii) the enhanced photosynthesis and water status leading to nitrogen fractionation into chlorophyll, lipids, amino sugars rather than proteins in the vegetative growth resulting in decreased δ15Ngrain [25]; (iii) and/or as the consequence of the alleviation of SLB's adverse effects.

#### *4.4. E*ff*ect of Thyme Oil*

Coating seeds with thyme oil showed a great potential in controlling SLB under field conditions in the three years of study and seems to be more efficient in controlling SLB compared to PsJN strain according to SLB severity values. The thyme oil effect on yield components tends to be dependent on climatic conditions since the latter had different effects on GY and SY among the 3 years. In 2015, in which rainfall was more abundant in the vegetative growth stage (December–February) than the grain filling stage (April), thyme oil increased SY and decreased GY. In 2016, in which rainfall was limited in the vegetative growth stage (December-February) and abundant at the heading and anthesis (March), thyme oil decreased SY and increased GY. In 2017, in which rainfall was abundant in both vegetative growth stage (December–February) and grain filling stage (April), thyme oil increased both SY and GY. This suggests that thyme oil increases the growth rate of the assimilatory organ dependent on water availability. Thyme oil seems to be ineffective in promoting grain yield when there is an interaction disease x water deficit at the grain filling stage. This is thought to be the side effect of the activation of the systemic acquired resistance SAR [8], which induces the energy allocation towards defense related mechanisms and limits energy availability towards drought-tolerance mechanisms when water deficit occurs at the grain filling stage. According to the 2017 one year of study, disease resistance was branded by the absence of the plant damage induced by *Z.tritici* behavior, resistance was characterized by less stomata opening and the absence of chlorophyll deterioration which is most likely due to thyme oil's direct effect via hampering the fungal development and indirect effect via inducing SAR within plant tissues [8].The thyme oil effect behind enhanced GY, SY, spikes/m2 and TKW of wheat is thought to be related to both: (i) The elicitor effect inducing seed priming resulting in the metabolic changes that involve peroxidase, phenolic compounds accumulation and the growth promotion of root and shoot parts starting from seedling emergence stage [7]; (ii) the alleviation of SLB's adverse effect. Concerning grain composition, the effect of thyme oil on increasing Cgrain content and decreasing δ13Cgrain suggests that this effect is mostly related to thyme oil's potential in improving the plant water status due to the enhanced root elongation conferring a higher water uptake [7,24]. The thyme oil effect on decreasing δ15Ngrain compared to the inoculated control and simultaneously, not influencing the total Ngrain content suggests that thyme oil both increases the isotopic fraction 14Ngrain and decreases the isotopic fraction 15Ngrain and could be explained by: (i) The enhanced N uptake during vegetative growth as a consequence of the thyme oil priming effect on inducing intracellular acidification of plant cells [7] was found to increase N uptake [33], leading to increased 14N [25]; (ii)the enhanced water status leading to nitrogen fractionation into lipids, amino sugars rather than proteins in the vegetative growth resulting in decreased δ15Ngrain [20]; (iii) and/or as the consequence of the alleviation of SLB's adverse effect.

#### *4.5. Comparison between Treatments and Insight to Cost*/*Gain Balance*

The effect of PsJN strain and thyme oil differed among the three years of study. Concerning their effect on crop protection against SLB, in season 2015, when water availability was high, PsJN strain was more efficient than thyme oil in reducing SLB severity. Contrastingly, in seasons 2016 and 2017, when water availability decreased, thyme oil was more efficient than PsJN strain in reducing SLB severity. It is suggested that this difference is most likely due to their different induced type of resistance. Thyme oil triggers systemic acquired resistance causing the systemic stomatal closure [8], thus preserving water content and, by the way, decreasing the drought side effects. However, PsJN strain triggers induced systemic resistance (ISR) causing local stomatal closure only in the presence of a pathogen [8], thus maintaining the normal water dissipation rate. By this way, the energy needed for SLB resistance is expected to decrease due to energy allocation towards drought-tolerance mechanisms when a water deficit occurs, as in the years 2016 and 2017.

The better impact of PsJN strain on yield components and grain composition compared to thyme oil is suggested to be related also to the distinct defence mechanisms and can be explained by the selective cost–benefit scenario of inducible defences [22]. Thyme oil is considered to trigger constitutive defence and PsJN strain is considered to trigger induced defence [8]. The plant defence is a costly business, requiring energy and resources that would otherwise be used for growth and development [22,34]. In this context, the constitutive resistance triggered by thyme oil, where the activation occurs before the onset of the disease, is considered to be a costly advantage causing higher allocation of resources. While, the induced resistance triggered by PsJN strain, where defences are only activated following pathogen attack and only at the site of infection, is considered a less pricey advantage compared to constitutive resistance [22,34].

#### **5. Conclusions**

This study revealed that economic losses in durum wheat due to increased SLB severity can result from losses in straw yield, grain yield and grain quality. Coating seeds with either thyme oil or PsJN strain showed potential in counteracting the deleterious effects of SLB and the promotion of straw yield, and grain yield and quality. The data showed that the impact of thyme oil and PsJN differed in terms of intensity and stability. Further, it is considered to be linked to the different growth promoting effects and the different triggered systemic resistance and associated amount of costs deriving from resource allocation towards defense processes. This cost-benefit of induced resistance in the variety 'Karim' of durum wheat gives insight into the worth of studying the effects of PsJN strain or thyme oil in other varieties of wheat in order to seek better interaction which minimizes the costly effect of biostimulants.

**Author Contributions:** K.H. and K.B. managed and directed the field trials at the Regional Field Crops Research Center (CRRGC) at Béja, Tunisia. M.B.-J. and Z.K. conducted collection of samples and nondestructive field measurements. M.B.-J. analyzed the samples and wrote the paper under the supervision of J.L.A., M.D.S. and W.H.

**Acknowledgments:** This work was financially supported by the Ministry of Research and Higher Education of Tunisia [grant number LR02AGR02], ARIMNet, and the Spanish Ministry of Economy and Competitiveness (MEC) [grant number AGL2016-76527- R]. J.L.A. acknowledges the support of ICREA Academia, Government of Catalonia, Spain.

**Conflicts of Interest:** The authors declare that they have no conflicts of interest.

#### **References**


© 2019 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* **Biostimulants for Plant Growth Promotion and Sustainable Management of Phytoparasitic Nematodes in Vegetable Crops**

### **Trifone D'Addabbo 1,\*, Sebastiano Laquale 2, Michele Perniola <sup>3</sup> and Vincenzo Candido <sup>2</sup>**


Received: 28 August 2019; Accepted: 4 October 2019; Published: 7 October 2019

**Abstract:** The parasitism of root-knot nematodes, *Meloidogyne* spp., can cause heavy yield losses to vegetable crops. Plant biostimulants are often reported for a side-suppressive effect on these pests and many commercial products are increasingly included in sustainable nematode control strategies. Source materials of most biostimulants derived from plant or seaweed raw materials were documented for a reliable suppression of root-knot nematode species, whereas the suppressiveness of microbial biostimulants was found largely variable, as related to the crop and to environmental factors. Chitosan-based biostimulants were also stated for a variable phytonematode suppression, though clearly demonstrated only by a few number of studies. In a preliminary experimental case study, four commercial biostimulants based on quillay extract (QE), sesame oil (SO), seaweeds (SE), or neem seed cake (NC) were comparatively investigated for their effects against the root-knot nematode *M. incognita* on potted tomato. Soil treatments with all the four biostimulants resulted in a significant reduction of nematode eggs and galls on tomato roots, though NC and SO were significantly more suppressive than QE or SE. In addition, almost all biostimulant treatments also resulted in a significant improvement of tomato growth compared to the non-treated control. These preliminary results seem to confirm the literature data and clearly indicate the potential role of biostimulants for a safe nematode management both in organic and integrated crop systems.

**Keywords:** biostimulants; phytoparasitic nematodes; suppressiveness; sustainable management

#### **1. Introduction**

Phytoparasitic nematodes are among the most harmful pests of vegetable crops, responsible for an annual yield loss amounting to 9–15% of the world crop yield [1]. Most of these losses are due to root-knot nematode species, *Meloidogyne* spp., causing poor plant growth and reduced crop yield and quality and reducing plant resistance to other biotic and abiotic stresses [2]. Traditionally, control of these pests relied on soil treatments with synthetic nematicides, but the increasing demand for a higher crop safety to the environment and humans has led to a progressive dismission of these products, giving a strong impulse to the search and the implementation of control strategies based on natural mechanisms, such as the use of plant biostimulants [3].

Plant biostimulants derived from natural materials have been receiving a growing interest by researchers, farmers, and industrial companies, as considered an effective tool for improving crop productivity [4]. The previous unclear and misunderstanding legislation frame led to include among the biostimulants a large variety of products with different activities, such as growth enhancers, plant

strengtheners or conditioners, resistance elicitors, as well to registration procedures variable among countries or even within the same country [5]. The uncertain legislative frame resulted in the immission in the market of a large variety of biostimulants stated for a suppressiveness on phytoparasitic nematodes, because of their content of raw materials (plants, seaweeds, microorganisms, and more) widely demonstrated for an activity against phytonematode species [6–8]. However, the recent EU Regulation 2019/1009 [9] has restricted the definition of fertilizing products and biostimulants and, therefore, many of these borderline products are destined to be classified as phytochemicals, dealing with more complex and expensive registration procedures.

Because of the increasing technical and economic relevance of these products, the aim of this study is to provide a review of the main groups of nematode-suppressive plant biostimulants actually available in the market and to indicate their potential for an effective but safe nematode management by a preliminary experimental case study on the root knot nematode *M. incognita* Kofoid et White (Chitw.) on tomato.

#### **2. The State-of-the-Art**

#### *2.1. The Market Supply*

A survey of the Italian market in 2018 revealed the presence of almost 40 different commercial plant biostimulants/strengtheners declaring a side activity on phytoparasitic nematodes on their labels (Table 1). More than 50% of these commercial products were based on plant raw materials, such as extracts, seed oils or green and seed biomasses, whereas another 25% was represented by seaweed derivatives. There was only one chitosan-based formulate, whereas the remaining others were microbial formulations. Only four products were clearly described as nematotoxic and the activity of other nine formulates was related to nematode repellence, disorientation, or antifeeding effects, whereas the remaining products were generically described as enhancers of plant resistance or of unfavorable soil conditions.


**Table 1.** Commercial biostimulants reporting an activity against phytoparasitic nematodes available in the Italian market at December 2018.


**Table 1.** *Cont.*

<sup>1</sup> L = liquid; D = dry meals, P = pellets, G = granules; <sup>2</sup> 1 = biostimulant; 2 = rooting; 3 = fertilizing; 4 = plant defense enhancement; 5 = increase of soil beneficial microflora; 6 = creation of a nematode-unfavorable environment; 7 = repellence, antifeeding, disorientation; 8 = toxicity. Products applied in the case study are reported in bold.

#### *2.2. The Literature Review*

Plant-derived biostimulants previously documented for an activity on phytonematodes were mostly liquid formulations of extracts and oils or, at a less instance, granular or powder seed meal or cake derivatives. A large number of plant biostimulants based on sesame seed oil [10], quillay water extract [11,12], or meals from biomasses or seeds of *Brassicaceae* plants and neem [13–15] were previously demonstrated for a suppressive activity on root-knot nematode populations on field and greenhouse tomato.

Seaweed extracts were found to cause an almost complete mortality of root-knot nematode juveniles and eggs in in vitro studies [16,17], as well as formulations of the extracts from seaweed species *Ascophyllum nodosum* L. and *Ecklonia maxima* Osbeck were reported for an effective control of root-knot nematodes also in soil experiments on tomato [18–20]. In addition to extract derivatives, a strong suppression of *Meloidogyne* spp. infestations on fruit or vegetable crops was described also for soil amendments with biomasses of seaweeds *Uva lactuca* L. and *Spatoglossus schroederi* Agardh (Kützing), may be due to their high content of phenolics and other bioactive compounds [21,22]. In addition to Meloidogyne species, suppressive activity of seaweed products was also detected on nematode parasites economically relevant to tropical or subtropical vegetable crops, such as *Helicotylenchus indicus* Siddiqui, *Belonolaimus longicaudatus* Rau, or *Radopholus similis* Cobb (Thorne) [23–26].

Literature studies are available also on the suppressive activity of chitosan and/or its derivatives, both alone or combined with other suppressive materials (agricultural wastes, plant compounds, biocontrol agents), either on root-knot nematodes [27–30] and other phytoparasitic species i.e., the soybean cyst nematode *Heterodera glycines* Ichinoe and the pinewood parasite *Bursaphelenchus xylophilus* (Steiner et Buhrer) Nickle [31–33].

Most of the microbial biostimulants reported as active on phytoparasitic nematodes were formulations of arbuscular mycorrhizal fungi [34,35]. Suppressiveness to root-knot nematodes of these products, either alone or combined with other microorganisms or plant extracts, was documented both in field and greenhouse [36–39]. Moreover, their activity was demonstrated also on other phytonematode parasites, such as *Nacobbus aberrans* Thorne et Allen or *Helicotylenchus multicinctus* (Cobb) Golden on field banana and greenhouse tomato, respectively [40,41]. In addition to mycorrhizal fungi, formulations of other fungal or bacterial biocontrol agents (*Trichoderma* spp., *Bacillus* spp.) or nitrogen fixers (*Azospirillum* spp., *Azotobacter* spp.) were also reported for controlling *M. incognita* in glasshouse tomato and field sunflower [42–44], or improving crop tolerance to the cyst nematode *Heterodera schachtii* Schmidt and more generically to soil phytoparasitic nematophauna [45,46].

#### **3. An Experimental Case Study**

#### *3.1. Materials and Methods*

A sandy soil (64.4% sand,18.7% silt, 16.9% clay, 0.8% organic matter, pH 7.5; 18.2% soil average humidity, 23.5% field capacity, 12.9% wilting point), artificially infested with the root-knot nematode *M. incognita* (8 eggs and juveniles mL−<sup>1</sup> soil) was poured into 2.5 L clay pots. Soil was then treated with three commercial liquid biostimulants derived from quillay (*Quillaja saponaria* Molina) extract (Tequil Multi®, Fertenia) (QE), sesame (*Sesamum indicum* L.) oil (NeMax®, Sumitomo Chemical) (SO) or brown algae (*Laminaria* spp.) extract (AgriPrime Nematec®, BioAtlantis) (SE), and a granular formulation of neem (*Azadirachta indica* Juss) cake (Neem Soil®, Serbios) (NC). QE, SO and SE were applied at transplant and 15 and 30 days later at amounts corresponding to 60, 10, and 2 L ha−1, respectively, whereas NC was incorporated to the soil at a 1000 kg ha−<sup>1</sup> rate two weeks before transplanting. The same treatments were also provided to pots containing non-infested soil. Soil treated with the nematicide Oxamyl (OX), applied at a 10 L ha−<sup>1</sup> field rate 3 days before tomato transplant and 15 days later, and non-treated soil, both infested (NT) and non-infested (NI) by *M. incognita*, were used as controls. One-month-old tomato seedlings (cv. Harvester) were transplanted in each pot, providing five replicates for each treatment in comparison.

The pots were arranged in a randomized block design in a plastic greenhouse at 25 ◦C, where they were maintained for 75 days, receiving a regular irrigation but no additional pesticide or fertilizer treatment. At the end of their permanence in the greenhouse, plants were uprooted and weight of green and root biomass was recorded for each plant. Root gall formation was estimated according to a 0–10 scale [47] and nematode multiplication on tomato roots was determined by extracting eggs and juveniles by the Hussey and Barker's method [48]. Data were statistically analyzed by ANOVA and treatment means were compared by the Fisher's Least Significant Difference Test at *P* ≤ 0.05, using PlotIT 3.2 (Scientific Programming Enterprises, Haslett, MI) software.

#### *3.2. Results*

The number of nematode eggs and juveniles on tomato roots were always significantly lower in the soil treated with the four biostimulants or OX than in NT soil (Figure 1A). Moreover, the multiplication of *M. incognita* in pots treated with NC or SO was not statistically different from OX and significantly lower than the treatments with QE and SE. Finally, QE resulted to be significantly more suppressive than SE.

Treatments with the four biostimulants and OX also resulted in a significantly lower number of root galls in comparison with NT (Figures 1B and 2). As for nematode eggs and juveniles, the formation of galls in soil treated with NC and SO was statistically lower than QE and SE, though only NC was significantly not different from OX. No statistical difference occurred between the number of galls from QE and SE.

Tomato plant biomass in soil infested by *M. incognita*, either non-treated and treated with the biostimulants or OX, was always significantly lower than NI (Figure 3A). Green biomass from plants in soil treated with QE was significantly larger compared to all the other treatments and NT. Adversely, weight of green biomass from pots treated with the other three formulates was not significantly different from NT and statistically lower than OX.

Weight of the tomato roots from all the treatments but NC was significantly higher than the NT (Figure 3B). Moreover, QE resulted in a root biomass significantly heavier than the other three biostimulants and OX and not different from NI. Finally, SE resulted in a root growth statistically not different from OX but higher compared to NC and SO.

**Figure 1.** Multiplication of the root-knot nematode *Meloidogyne incognita* (**A**) and gall formation (**B**) on the roots of tomato cv. Harvester in soil non-treated (NT) or treated with commercial biostimulants based on neem cake (NC), sesame oil (SO), seaweed extract (SE), and quillay extract (QE) or with nematicide Oxamyl (OX). Bars tagged with the same letters are not statistically different (*P* ≤ 0.05) according to the Least Significant Difference's Test.

**Figure 2.** Roots of tomato plants cv. Harvester from soil treated with commercial biostimulants based on neem cake (NC), sesame oil (SE), seaweed extract (SE) and quillay extract (QE) or with nematicide Oxamyl (OX) and from non-treated soil (NT).

**Figure 3.** Weight of green biomass (**A**) and roots (**B**) of tomato plants cv. Harvester in soil non-treated (NT) or treated with commercial biostimulants based on neem cake (NC), sesame oil (SO), seaweed extract (SE), and quillay extract (QE) or with nematicide Oxamyl (OX). Bars tagged with the same letters are not statistically different (*P* ≤ 0.05) according to the Least Significant Difference's Test.

#### **4. Discussion**

The experimental case study indicated that biostimulants can also provide a satisfactory nematode suppression, as confirming previous findings from literature studies. However, these results aim to be only indicative of the potential use of biostimulants in nematode management and need to be validated by future trials in field conditions, as well as different combinations of biostimulants should be also tested to verify a potential synergism among different products.

The mechanisms of biostimulants suppressiveness to nematodes are only partially known or simply hypothesized. Seaweed activity on phytoparasitic nematodes was generally attributed to their content of secondary metabolites, such as steroids, triterpenoids, alkaloids, and phenols, known for a nematicidal activity or as plant resistance elicitors [49,50]. Analogously, the suppressiveness to phytonematode populations of plant-based biostimulants is mainly related to nematotoxic metabolites both preformed in raw plant material (saponins, fatty acids, alkaloids and more) or released during the plant materials degradation in soil [51,52]. Induction of a systemic plant resistance to nematode penetration has been also documented for some active compounds of plant-derived biostimulants, such as neem azadiractin or chestnut (*Castanea sativa* Mill.) tannins [53,54]. Nematode suppression by microbial biostimulants was generally attributed to the induction of crop defense responses to nematode invasion [55,56]. Additional or alternative mechanisms, such as a competition for nutrients and space or the synthesis of nematicidal microbial metabolites have been also suggested [57–59]. The nematicidal effectiveness of chitosan products was generally attributed to the induction of a local or systemic plant resistance [60], though an enhancement of nematode-suppressive rhizospheric bacteria and fungi has been also hypothesized [36,40].

In our study, only QE was confirmed for a biostimulant effect on tomato growth, as limited only to the root biomass for SO and SE or nil for NC. The growth effect of QE can be attributed to the

high content of triterpenic saponins, widely acknowledged for significant plant growth regulating properties [61], in *Q. saponaria* extracts.

Chemical composition of plant-based biostimulants can change according to a range of environmental and agronomic factors [62], as well as the nematode suppressiveness of microbial formulations may vary according to microbial strains, crop species/varieties, and environmental conditions [63]. Variable effects on soil phytonematode populations were also documented for chitosan products, as strictly dependent on the molecular weight of raw materials [32,64]. The unstable composition is a serious constraint to the full implementation of biostimulants in nematode management strategies, as leading to a fluctuating activity in field and, consequently, to a difficult certification of nematicidal performances and registration of commercial products [51]. A preliminary standardization of source raw materials and manufacturing processes should ensure constant suppressive performances and a successful market presence to the future commercial plant biostimulants addressed to nematode management. Moreover, preliminary toxicological screenings should be provided for any new biostimulant, as to exclude the presence of compounds with an unknown toxicological profile or the persistence of human pathogens in materials of animal origin [51].

In conclusion, plant biostimulants can also play a relevant role in the future nematode management strategies, as providing an acceptable nematode suppression in addition to their main activity of plant growth and ensuring a full safety to the other biotic soil components. It may be reasonably expected that the Regulation 2019/1009 [9] will lead to the disappearance of products with a direct toxicity to nematodes activity, because of the high costs of their registration as pesticides, as limiting the market to the products working through plant resistance improvement. A stand-alone application of these products can be reasonable only in organic crop systems, where few nematode control tools are available, or in short-cycle crops where the short pre-harvest intervals do not allow the use of most synthetic nematicides. However, a combination with other chemical or nonchemical control tools can justify the application of these products also in conventional crop systems. Benefit–cost ratio of treatments with the kind of products analyzed in this work should be always evaluated before their application as nematode suppressants, because of the high market price of these products which limit their use preferably to high value crops.

**Author Contributions:** Conceptualization, T.D., M.P., V.C.; data curation, S.L.; formal analysis, T.D., V.C.; investigation, S.L.; methodology, T.D., S.L.; software, S.L.; supervision, M.P.; validation, V.C.; visualization, S.L.; writing—original draft, T.D.; writing—review and editing, T.D., V.C.

**Acknowledgments:** The authors acknowledge the technical assistance of Fabio Catalano for the arrangement of greenhouse experiment and lab work.

**Conflicts of Interest:** The authors declare that the submitted work was carried out in the absence of any personal, professional, or financial relationships that could potentially be construed as a conflict of interest.

#### **References**


© 2019 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* **Chemical Composition of Winter Rape Seeds Depending on the Biostimulators Used**

### **Marek Gugała 1, Anna Sikorska 2,\*, Krystyna Zarzecka 1, Pavol Findura 3,4 and Urszula Malaga-Toboła <sup>4</sup>**


Received: 21 October 2019; Accepted: 4 November 2019; Published: 5 November 2019

**Abstract:** Plant growth regulators may reduce the negative effect of environmental stress factors and can contribute to increasing the quality and quantity of the yield. The aim of the research was to determine the effect of biostimulators on the quality of seeds of three winter rape morphotypes. Three varieties of winter rape were used: Poznaniak (population variety), PX104 (hybrid variety restored with a semi-dwarf growth type) and Konkret (hybrid variety restored with a traditional growth type). The varieties were exposed to three treatments: the biostimulator Tytanit®, the biostimulator Asahi®SL and the biostimulator Silvit®, and the control with no biostimulators. Seeds were analysed for content of crude fat, total fat and crude fibres. The biostimulators reduced total protein content (on average from 0.8 to 1.75 g·kg−<sup>1</sup> of d.m.) and increased the concentration of crude fat (on average from 0.71 to 1.93 g·kg−<sup>1</sup> of d.m.) and crude fibre (on average from 0.15 to 0.84 g·kg−<sup>1</sup> of d.m.) compared to the control. PX104 had the highest content of crude fat and total fat protein, and the lowest in crude fibre. The smallest protein content was found in seeds of the long-stem hybrid Konkret, while crude fat was lowest in the population form (Poznaniak), and crude fibre was lowest in long-stem hybrid (Konkret).

**Keywords:** anti-nutritional substances; fat; fibre; morphotype; protein

#### **1. Introduction**

Rapeseed *(Brassica napus* L. var. oleifera) is one of the most important oil-protein crops grown in the world. One of the many factors with a negative effect on the quantity and quality of rapeseed crops include unfavourable soil conditions and drought-related stress. Strong stress leads to damaged cell structures and disturbances in metabolism and as a result, can lead to photosynthesis and plant and metabolism disruption [1]. Rouphael and Colla [2] reports that plant biostimulators are products obtained from various organic or inorganic substances or microorganisms that improve plant growth, productivity and reduce the negative effects of environmental stress. Many authors [3–8] have shown that regulators of plant growth and development reduce the negative impact of abiotic stress factors. Petrozza et al. [9] showed that when a plant experiences stress, the biostimulator strengthens its stress tolerance mechanism.

Colla and Rouphael [10] and Rouphael et al. [11] emphasize that the use of biostimulators is increasingly becoming one of the basic elements of agricultural technology in many crop species around the world. Calvo et al. [12] forecast that the global market for biostimulants in consumption will increase by 14% per year.

According to El-Boray et al. [13], Przybysz et al. [14], Kocira et al. [15] and Zulfiqar et al. [16], preparations stimulating plant growth can be based on extracts of marine algae, free amino acids, humic compounds, effective microorganisms or phenolic compounds. Their use in plant cultivation has a positive effect on photosynthesis, regulation of water management and increasing the content of organic and inorganic compounds, which in turn, has a positive effect on the size and quality of the crop.

Grabowska et al. [17] and Kolomaznik et al. [18] stated that the effectiveness of biostimulators depends on many factors, including the correct selection of preparations, their dose, concentration and methods of application, as well as plant species and varieties and environmental factors.

The study assumes the hypothesis that the use of biostimulators may have a positive effect on the chemical composition of winter rapeseeds.

Due to few studies being available on the beneficial effects of growth bioregulators on the quality characteristics of winter rapeseed, and the wide interest in agricultural practice, research was undertaken to determine the effect of biostimulators on the chemical composition of three winter rapeseed varieties.

#### **2. Materials and Methods**

#### *2.1. Arrangement of the Experiment and Research Location*

The field experiment was carried out in 2013–2016 in three different fields at the Agricultural Experimental Station—Zawady (52◦03 N; 22◦33 E), belonging to the University of Natural Sciences and Humanities in Siedlce. The experiment was established in a random split-plot system in three repetitions (total number of plots 3 × 4 = 12, repeated in 3 successive crop rotations from 2013–2016). The surface of one plot was 21 m<sup>−</sup>2. The examined factors were:

I—three varieties of winter rape: Poznaniak (population variety), PX104 (hybrid variety restored with a semi-dwarf growth type) and Konkret (hybrid variety restored with a traditional growth type).

II—four types of biostimulators:

1. control—no biostimulators;

2. biostimulator Tytanit® (active substance—titanium), applied in three doses of 0.20 dm3 ha−<sup>1</sup> in the autumn (2 October 2013, 6 October 2014, 5 October 2015) at the 4–8 leaf stage (BBCH 14–18) according to the rating of Biologische Bundesantalt, Bundessortenamt and Chemische Industrie (BBCH), in the spring (26 March 2014, 23 March 2015, 21 March 2016) after the onset of growth (BBCH 21–36), and at the budding stage—early flowering (30 April 2014, 29 April 2015,4 May 2016) (BBCH 50–61) [19];

3. biostimulator Asahi®SL (active substances: sodium orto nitrophenol, sodium para nitrophenol, sodium 5-nitroguaiacolate), applied in three doses of 0.60 dm3·ha−<sup>1</sup> in the autumn (25 September 2013, 29 September 2014, 27 September 2015) at the stage of 3−5 leaves (BBCH 13−15), in the spring (26 March 2014, 23 March 2015, 21 March 2016) after the plants resumed growth (BBCH 28−30), and two weeks following the second application (10 April 2014, 7 April 2015, 4 April 2016);

4. biostimulator Silvit® (active substances: active silicon, potassium oxide, boron, zinc), applied in three doses of 0.20 dm3·ha−1, three weeks after emergence (2 October 2013, 6 October 2014, 5 October 2015) (BBCH 12−14), in spring (26 March 2014, 23 March 2015, 21 March 2016) after plants resumed growth (BBCH 28–30), and two weeks after the second application (10 April 2014, 7 April 2015, 4 April 2016).

The studies were carried out on soil classified according to WBR FAO (2014) [20] as the Haplic Luvisols group—sandy, belonging to a very good rye soil complex of the IVb botanical class. In the years of the experiment, the soil reaction (pH) ranged from 5.68 to 5.75. The soil was characterised by a low total nitrogen content (average from 0.80 to 0.90 g·kg<sup>−</sup>1), phosphorus content (average from 0.33 to 0.55 g·kg−1), potassium content (average from 0.61 to 0.67 g·kg−1) calcium content (average from 0.82 to 0.85 g·kg<sup>−</sup>1), magnesium content (average from 0.38 to 0.46 g·kg<sup>−</sup>1) and sulphur content (average from 0.11 to 0.15 g·kg<sup>−</sup>1). It has a low abundance in assimilable forms of phosphorus (average from 75.0 to 80. g·kg−1) and an average assimilability of potassium (from 200.0 to 205.0 g·kg−1) and magnesium (average from 59.0 to 61.0 g·kg<sup>−</sup>1).

The phosphorus and potassium fertilization at the dose of 40.0 kg P·ha−<sup>1</sup> and 110.0 kg K·ha−<sup>1</sup> with the first dose of 40.0 kg N·ha−<sup>1</sup> was used before sowing. Fertilization was used in the form of Lubofos for Rape at the dose of 600.0 kg, i.e., 21.0 kg N·ha<sup>−</sup>1, 26.4 kg P·ha<sup>−</sup>1, 92.1 kg K·ha<sup>−</sup>1, 34.8 kg S·ha<sup>−</sup>1, 1.2 kg <sup>B</sup>·ha<sup>−</sup>1. Fertilization rates were supplemented by 55.9 kg·ha−<sup>1</sup> of ammonium nitrate (19.0 kg N·ha<sup>−</sup>1), 29.6 kg·ha−<sup>1</sup> of triple superphosphate (13.6 kg P·ha<sup>−</sup>1) and 29 kg·ha−<sup>1</sup> of potassium salt (17.9 kg K·ha<sup>−</sup>1). The second nitrogen dose of 100.0 kg·ha−<sup>1</sup> was applied in spring before vegetation using ammonium nitrate at the dose of 255.5 kg·ha−<sup>1</sup> and ammonium sulphate at the dose of 62.5 kg·ha<sup>−</sup>1. The third dose of nitrogen 60.0 kg·ha−<sup>1</sup> was applied at the beginning of budding using ammonium nitrate at the dose of 176.5 kg·ha<sup>−</sup>1.

The three crops of rapeseed were harvested on 11 July 2014, 17 July 2015 and 14 July 2016, respectively.

#### *2.2. Chemical Analysis of Seeds*

The tests samples of winter rape seeds were analyzed for:

Crude fat (g·kg<sup>−</sup>1of d.m.)—with the Soxhlet method, which extracted the fat with petroleum ether in a Soxhlet apparatus and determides its quantity by weight, total protein (g·kg−<sup>1</sup> of d.m.) [21].

Total protein (g·kg<sup>−</sup>1of d.m.)—with the Kjeldahl method where protein nitrogen was converted to ammonium sulphate with concentrated sulphuric acid in the presence of a catalyst, the solution was alkalised, distilled and titrated with hydrochloric acid-ammonia bound with boric acid, the conversion factor Nx6.25 was used, crude fibre (g·kg−<sup>1</sup> of d.m.) [22].

Crude fibres (g·kg−1of d.m.)—with the Wenden method consisting of the quantitative determination of organic substances insoluble during cooking in an acid solution.

#### *2.3. Statistical Analysis*

Research results were statistically analysed by ANOVA. The results of the study were statistically analysed using the analysis of variance. The significance of the sources of variation was tested by the Fischer-Snedecor "F" test, and the assessment of significance of differences at the significance level *p* < 0.05 between the compared averages used Tukey's multiple intervals. Statistical calculations were made based on our own algorithm written in Excel [23].

#### *2.4. Weather Conditions*

Climatic data from 2013–2016 was obtained from the Hydrological and Meteorological Station in Siedlce. During the years of conducting the experiment, varied weather conditions prevailed (Table 1). In the second growing season, the largest annual rainfall was recorded (average of 599.2 mm) and the smallest mean annual air temperature (average of 8.8 ◦C). In this period, the annual amount of rainfall was 171.7 mm higher compared to the long-term period. The last year of tests was the warmest and most dry. The annual rainfall was 43.8 mm lower than the average for the long-term period, and the average air temperature was higher by 1.3 ◦C compared to the average from 1996–2010. Based on the calculated Sielianinov hydrothermal coefficient, the first and last year of the study were optimal, while the growing season 2014–2015 was rather wet (*K* = 1.71).



\* humid 2.5 < k ≤ 3.0, extremely humid k > 3.0.

#### **3. Results and Discussion**

#### *3.1. The Content of Total Protein Depending on the Types of Biostimulators Used*

Our own research showed that biostimulators significantly affected the reduction of total protein content in rapeseeds (Table 2). The smallest concentration was recorded on object 4, sprayed with the Silvit biostimulator. This value was lower on average by 1.75 g·kg−<sup>1</sup> of d.m. compared to the control variant. Different results were obtained by Gugała et al. [25], who did not find a significant effect of the biostimulators Tytanit, Asahi SL or Silvit for the value of this feature. Similarly, Jarecki and Bobrecka-Jamro [26,27], Kozak et al. [28] and Matysiak et al. [29,30] did not prove the effect of bioregulators and foliar fertilizers containing micro- and macro-elements for the value of this feature. While Jankowski et al. [31], after a double foliar application with boron, increased the protein content in seeds by an average of 8.8 g·kg−<sup>1</sup> of d.m. compared to the control object. In regards to seed protein, the present study's research showed the interaction of the types of biostimulators used in relation to the protein content in the seeds of the rapeseed morphotype varieties studied, which indicated the individual response of the rapeseed varieties to the biopreparations used (Table 2). The lowest protein was in the treatment of Konkret with Silvit and PX104 with Silvit, and in Pozniak with Tytanit. In all morphotypes, the highest protein content was recorded on the object where no natural growth stimulants were used. In the cultivar with the traditional growth type, the lowest protein content was found after the application of the Tytanit biostimulator, while in the other varieities it was after the use of the Silvit biostimulator. Equal protein content was found in the restored hybrids of the semi-dwarf type (PX104) after the application of the Asahi SL and Tytanit preparations. A similar tendency was observed in the restored hybrid with the traditional growth type.

The content of total seed protein was dependant on the genetic factor (Table 2). In our own research, the content of protein in the seeds of the studied winter rape varieties averaged from 361.37 to 373.42 g·kg−<sup>1</sup> of d.m. The highest concentration was found in the semi-dwarf hybrid PX104, while in the long-stem hybrid (Konkret), it was lower on average by 12.05 g·kg−<sup>1</sup> of d.m. Different results were obtained by Gugała et al. [25], who received the highest value of this feature in a hybrid with a traditional type of growth and the lowest in a semi-dwarf hybrid. Ratajczak et al. [32] did not show significant differences between heterosis morphotypes with a traditional and semi-dwarf type of growth or in the population Califorium variety.

#### *3.2. The Content of Crude Fat Depending on the Types of Biostimulators Used*

The bioregulators used in the experiment significantly influenced the increase of crude fat in winter rapeseeds (Table 2). The greatest value of this feature was noted after the use of the Asahi SL biostimulator, it was significantly smaller on the objects where Tytanit and Silvit were applied. The beneficial effect of the Asahi SL biostimulator on the fat content in seeds was also confirmed by Spychaj-Fabisiak et al. [33] and Gugała et al. [25]. Similarly, Kovácik et al. [ ˇ 34] confirmed that a two-fold application of the Tytanit biostimulator affected the increase of the fat content in rapeseeds compared to the control object. The lack of effect of biostimulators on the fat content in seeds has been demonstrated by Matysiak et al. [29,30]. The authors observed only a slight tendency to increase the value of this feature even by 3.9% in relation to the control object. Similarly, Szczepanek et al. [35] noted a small effect of stimulating plant preparations on this feature. Jankowski et al. [31] after using a boron-containing foliar preparation, found a significant increase in the content of crude fat only after its two applications in the BBCH50 and BBCH55 phases. Jarecki and Bobrecka-Jamro [25,26] did not prove the effect of foliar preparations containing micro- and macro-elements on the value of this feature.

The impact of the types of biostimulators used on the crude fat content in rapeseed depended on the genetic factor (as shown in Table 2). The lowest fat content in all tested cultivars was recorded on the control object. The population cultivar had the highest fat content after using the Asahi SL biostimulator, but after the application of all biopreparations in this cultivar, the differences in protein crude fat content were not statistically significant. The seeds of the restored hybrid with the traditional growth type were characterized by the highest content of crude fat after the application of Asahi SL, and under the influence of the other biostimulators, they were the same as on the control object. A similar tendency was observed in the semi-dwarf hybrid, with differences in the value of this trait on the objects with the Tytanit and Silvit biostimulator were not statistically significant.

The content of crude fat depending on the genetic factor is shown in Table 2. Our own research proved that the highest content of fat was a characteristic of the PX104 (restored hybrid with a semi-dwarf type of growth), it was significantly smaller by 17.66 g·kg−<sup>1</sup> of d.m. in the long-stem hybrid (Konkret), while the smallest on average by 20.57 g·kg−<sup>1</sup> was in the population form (Poznaniak). Different results of studies were obtained by Gugała et al. [25] who showed that the highest value of this feature was characteristic for a restored hybrid with a traditional type of growth, while the smallest the population (Monolit).

#### *3.3. The Content of Crude Fibre Depending on the Types of Biostimulators Used*

Natural plant preparations influenced the increase of the crude fibre content in winter oilseed rapeseeds on average from 0.15 to 0.84 g·kg−<sup>1</sup> of d.m. (Table 2). The highest value of this feature was noted on object 3 with the Asahi SL biostimulator. Different results were obtained by Gugała et al. [25]. In this study, the biostimulants did not significantly alter the crude fibre content in seeds of the rapeseed cultivars (Table 2).

The content of crude fibre depending on the genetic factor is shown in Table 2. Our own studies indicate that the highest content of crude fibre was observed in the seeds of the PX104 variety, while the smallest was in the long-stem morphotype (Konkret). Different results were obtained by Gugała et al. [25], who did not find any statistical differences in the value of this feature between the studied morphotypes.

#### *3.4. Chemical Composition Depending on Weather Conditions*

The chemical composition of seeds depending on climatic conditions in the study years is shown in Table 2. In our own research, the highest content of total protein, fat and crude fibre was obtained in seeds collected in the second year of research, in which the total precipitation was 41.9 mm higher in May, and the average monthly temperature was smaller by 0.5 °C from the average multi-year. Similar results were obtained by Chmura et al. [36] and Gugała et al. [25]. According to the authors, during the period from the end of flowering to the technical maturity stage of high protein content, temperatures of 16.2 °C were maintained on average, regardless of the sum of rainfall. M ˛aczy ´nska et al. [37] recorded a higher concentration of fat in colder years with a greater sum of precipitation, while it was lower in warm years. In our own studies, differences in the content of total protein and crude fat in the growing season of 2013–2014 and 2015–2016 were statistically insignificant, while the lowest content of crude fibre was found in seeds collected in the first year of research.



#### **4. Conclusions**

In summary, the applied biostimulators had an effect on reducing the total protein content (on average from 0.8 to 1.75 g·kg−<sup>1</sup> of d.m.) and increasing the concentration of crude fat (on average from 0.71 to 1.93 g·kg−<sup>1</sup> of d.m.) and crude fibre (on average from 0.15 to 0.84 g·kg−<sup>1</sup> of d.m.) compared to the control object. The best quality of seeds was characteristic for the semi-dwarf PX104 variety. The smallest protein content was found in seeds of the long-stem hybrid Konkret, while crude fat was lowest in the population form (Poznaniak), and crude fibre was lowest in long-stem hybrid (Konkret). Diverse climatic conditions prevailing in the years of conducting the experiment influenced the chemical composition of rapeseeds. The highest content of total protein, crude fat and fibre were obtained in the second year of studies.

**Author Contributions:** Methodology, K.Z. and M.G.; software, M.G.; writing—original draft preparation, A.S.; review and editing—P.F. and U.M.-T.

**Funding:** The results of the research carried out under the research theme No. 363/S/13 were financed from the science grant granted by the Ministry of Science and Higher Education".

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

#### **References**


© 2019 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*

### **Investigating the Impact of Biostimulants on the Row Crops Corn and Soybean Using High-E**ffi**ciency Phenotyping and Next Generation Sequencing**

**Nunzio Briglia 1, Angelo Petrozza 2, Frank A. Hoeberichts 3,**†**, Nathalie Verhoef <sup>3</sup> and Giovanni Povero 4,\***


Received: 18 September 2019; Accepted: 12 November 2019; Published: 16 November 2019

**Abstract:** Row crops represent the most important crops in terms of global cultivated area. Such crops include soybean, corn, wheat, rice, rapeseed, sunflower, and cotton. Row crops agriculture is generally an intensive system of farming used to obtain high yields by employing elevated quantities of organic and mineral fertilizers. Considering this, and the decrease in area of arable land, it becomes crucial to ensure high yield and quality using alternative strategies, such as the use of plant biostimulants. These compounds are increasingly recognized as sustainable solution to optimize nutrient uptake, crop yield, quality, and tolerance to abiotic stresses. In this work, by means of high-throughput plant phenotyping, we evaluated the effectiveness of a set of three new foliar biostimulant prototypes (coded as 52096, 52097, 52113) applied on corn and soybean at application rates 2.5 and 5 mL/L (corresponding to 1 and 2 L/ha respectively). This allowed us to select the most effective prototype (52097, commercial name "YieldOn®") in increasing digital biovolume (DB) and greener area (GGA) either in soybean (both application rates) or corn (rate 5 mL/L) and decreasing Stress Index (SI) in soybean (both application rates). Molecular mechanism of action of selected prototype 52097 was subsequently characterized through Next Generation Sequencing (NGS). In corn, genes involved in hormone (cytokinin and auxin) metabolism/catabolism, maltose biosynthesis, sugar transport and phloem loading were upregulated after application of prototype 52097. In soybean, genes involved in nitrogen metabolism, metal ion transport (mainly zinc and iron), sulfate reduction, and amino acid biosynthesis were induced. The proposed approach supports the integration of multiple omics to open new perspectives in the discovery, evaluation, and development of innovative and sustainable solutions to meet the increasing needs of row-crops agriculture.

**Keywords:** biostimulants; corn; imaging; industrial crops; maize; next generation sequencing; phenomics; plant phenotyping; row crops; soybean

#### **1. Introduction**

The increase in global population and the uncertainty produced by climate change represent big challenges for current and future agriculture [1,2]. Agricultural activity should ensure crop production systems that can tolerate increasingly adverse environmental conditions, such as drought, flooding,

and other stressful events. At the same time, it should provide adequate yields to guarantee an economic return for farmers, and high-quality produce to satisfy the demands of consumers [3]. With a decreasing acreage of arable land and the limits of genetic potential of primary crops, to reach such objective it becomes necessary to increase crop yield, producing "more with less" [4–6] and to avoid overexploitation of natural resources, such as soil and water [7]. According to this, many research projects are supporting to design energy-efficient and eco-friendly cultivation systems, which are less dependent on the use of external inputs (e.g., fertilizers) [8,9].

To achieve these goals, the use of plant biostimulants (PBS) appears to be one of the most promising strategies [10]. According to the European Biostimulant Industry Council (EBIC, 2019) [11], plant biostimulants "contain substance(s) and/or microorganisms, whose function when applied to plants or the rhizosphere is to stimulate natural processes to enhance/benefit nutrient uptake, nutrient efficiency, tolerance to abiotic stress, and crop quality". The PBS formulations are generally proprietary compositions based on micro and macro-algae, plant extracts, hormone-like compounds, complex organic materials, amino acids or humic acids. Extensive reviews have recently discussed the discovery and the characterization of the activity of PBS derived from seaweeds, especially *Ascophyllum nodosum* [12–17]. In addition, several studies on the beneficial effects of natural PBS on plant growth, production and fruit quality in various crops have been recently published [18–21]. Physiological aspects in relation to the supply of PBS, like increased root and shoot growth, tolerance to abiotic stress, plant water uptake, and reduction of transplant shock, have also been reported [22–26]. Moreover, application of specific PBS may reduce fertilizer use and nutrient solution concentrations in hydroponic systems [27]. The development of PBS can therefore be used for the modulation of some plant physiological processes such as growth stimulation, stress mitigation, leading to increase yield and nutritional value of edible organs [16–18,28].

Considering the row crops sector, effective PBS are needed. Row/industrial crops such as soybean, corn, wheat, rice, rapeseed, sunflower, and cotton represent the most important crops in terms of global cultivated area [29]. It should be pointed out that row crops agriculture is generally based on an intensive farming system aimed at obtaining high yields by the use of high external inputs including organic and mineral fertilizers [9]. This is inconsistent with a vision of sustainable eco-compatible agricultural activity. Consequently, the use of PBS represents a sustainable strategy to contribute to ensure high yield and quality of product in this sector.

Recently, it was proposed to use transcriptomics together with plant phenomics to screen PBS and characterize their influence on plant physiology including the mechanisms activated by specific formulations [16,24]. Through transcriptomics, it is possible to identify possible modes of action of different substances and in turn predicting their role as biostimulants [30]. In addition to the transcriptomic profiling via microarrays, the novel technology Next Generation Sequencing (NGS) has been recently proposed as a tool to monitor the impact of PBS on the transcriptome of non-model plants, making it feasible to perform genomics in agricultural crops [31,32].

Using phenomics, it is possible to study the effect of PBS on plant biomass accumulation and the performances of the photosynthetic apparatus based on multi-spectrum, high-throughput image analysis to detect morphometric and specific physiological parameters (e.g., "Digital" Biovolume) [33,34]. This represents a step forward compared to "classical" in vitro and in vivo bioassays based on manual determination of simple physiological and morphological traits, evaluating nutrient uptake and growth stimulation through destructive quantification of root and shoot biomass. Such measurements result in a partial evaluation of PBS effects, without giving a real explanation of the mechanisms by which certain PBS exerts their effect(s). Among the different bioassays, the root growth inhibition of cress and the chicory hypocotyl growth are the most frequently used tests [35].

On the other hand, plant phenomics, based on multi-spectrum analysis of reflected or re-emitted light from the plant crown, stem and leaves provides a series of information related to plant structure and function, for example, plant water and nutritional status, pathogen infection, as well as on the plant's ability to intercept light. The use of high-throughput imaging analysis system allows to successfully integrate the experiments involving many variables, a large number of samples, and multiple comparisons [36]. Moreover, the high-throughput image analysis system is a non-invasive method that has the potential to determine the plant phenotypic response to experimental variable(s) (e.g., abiotic stress conditions), throughout the growing (or part of it) of experimental crops [24,25,37].

Hence, very recent papers showed that the use of a "multi-omics" approach, in particular metabolomics and plant phenotyping, represents an effective tool to examine plant performances under different experimental conditions [24,34]. The application of such integrated approach could offer a better explanation of the mechanism(s) of action of different PBS molecules or compounds on crops. This can be obtained by the identification of several biomarkers of PBS action as reported in Paul et al. (2019b) and Ugena et al. (2018) [34,36].

The aim of this study was the selection and characterization of a novel biostimulant formulation conceived to increase the yield of row crops. To achieve such objective, using a phenomic approach we investigated—on corn and soybean—the effect and physiological mechanism(s) of action of three different foliar biostimulant formulations/prototypes. This allowed the selection of the most effective one, subsequently characterized at transcriptomic level to understand its molecular action. This study, based on the integration of phenomic and genomic tools, opens new perspectives to release effective formulations for row-crops agriculture.

#### **2. Materials and Methods**

#### *2.1. Plant Material and Growing Conditions*

Experiments were performed on corn (*Zea mays* L., hybrid P0423, Pioneer) from November to December 2016, and soybean (*Glycine max* L. Merr.), from May to June 2017.

Plants were grown in a greenhouse under natural light conditions at the Plant Phenomics Platform, ALSIA-Metapontum Agrobios Research Center, Italy (N 40◦23 E 16◦47 ). Temperatures, humidity (RH%), and radiation (PAR μmol m−<sup>2</sup> s<sup>−</sup>1) are reported in Supplemental Table S1.

Both species were sowed directly into white pots (16 cm diameter, 20 cm height), containing 3.5 L of substrate consisting of a 50:50 mixture of peat and river sand. The day before sowing, 20 units of nitrogen, 40 units of phosphate, and 20 units of potassium oxide were added to the substrate mixture. Both soybean and corn plants were irrigated daily with 100 mL water, for 12 days. Afterwards, water was increased to 200 mL until the end of experiment to ensure an adequate water supply.

Three different biostimulant prototypes based on different combinations of seaweed and plant extracts formulated with selected micronutrients such as Mn, Zn, Mo (prototypes coded as 52096, 52097, 52113; proprietary composition Valagro SpA) were sprayed (using a portable atomizer sprayer) on both species at the third true leaf stage. Two rates were applied: 2.5 (lower dosage) and 5 mL/L (higher dosage) (corresponding to 1 and 2 L/ha respectively) during the experiment on soybean, while 5 mL/L was the unique rate applied during the experiment on corn. Untreated control plants were sprayed with distilled water. The experimental setup was composed by 5 biological replicates (plants) for each experimental condition using a completely randomized experimental design.

For NGS and qRT-PCR analysis, samples were collected just before (*t* = 0) and after the foliar application of PBS 52097 (5 ml/L) at 8, 24, 48, and 168 h. Both corn and soybean 3 leaves were removed from the plants and immediately submerged in liquid nitrogen. For each experimental condition (treatment and time-point), three biological replicates were collected from different plants at the third fully expanded leaf stage. Each biological replicate consisted of three entire leaves (central position) collected from three individual plants and pooled.

Plants used for sampling leaves were excluded from subsequent imaging acquisition or additional leaf sampling. All samples were collected at around 9:00 a.m.

#### *2.2. Non-Destructive Measurements*

The morphological and physiological characterization of the plants was carried out, non-destructively, by plant imaging. Images of plants were acquired, throughout the experiment, with the plant phenotyping platform Scanalyzer 3D (LemnaTec GmbH). Detailed information on the platform is reported in Briglia et al. (2019) [38]. Briefly, it is composed of 2 imaging chambers visible light (RGB) and fluorescence (FLUO), respectively. FLUO images, recorded into the fluorescence imaging chamber, were used to evaluate the photosynthetic performance through the "Stress Index" (see below). For each imaging chamber three images per plant were taken, one from top view of the plant and two from side view (0◦ and 90◦).

#### 2.2.1. Digital Biovolume Assessment

Plant growth was assessed through the digital biovolume (DB) [39] as follow:

$$
\sum \text{pixel sideview } 0^\circ + \sum \text{pixel sideview } 90^\circ + \log\_{10} \left( \sum \text{pixel} \frac{topview}{3} \right) \tag{1}
$$

where pixel sideview 0◦, 90◦ and top view are the plant pixel areas from all sides and top view images.

#### 2.2.2. Color Classification

During the experiment the resulting RGB images were then analyzed by categorizing the pixels according to their color.

After the color segmentation process, that allow to separate the plant from the background, the RGB images were converted to HSI color space (Hue, Saturation, Intensity) and then the hue histogram was calculated. According to Casadesús et al. (2007) [40] the relative greener area (GGA) of each image was calculated as the sum of frequencies of the histogram classes included in the hue angle ranging from 80◦ to 180◦. The GGA were used to evaluate the health status of the plant via colour classification (e.g., green: healthy and active leaf surface; yellow: degree of the plant senescence) [40].

#### 2.2.3. Stress Index

The performance of photosynthetic system is not constant and depends mainly on the health and stress condition of a plant. When a plant is placed under stress, more fluorescent light of higher energy is released and this change in the pixel distribution can be measured using the fluorescence imaging chamber.

The Stress Index was calculated according to Petrozza et al. (2014) [25] as (*Fx* − *Fy*)/(*Fx* + *Fy*) where *F* is the number of pixels in the *x*, *y* color classes, under the assumption that any impairments of the photosynthesis result in a change of pixel number at the x and y color class [25].

The *x*, *y* color classes were determined experimentally, by examining the hue histogram. Values of photosynthetic Stress Index vary from +1, poor photosynthetic efficiency, to −1, greater photosynthetic efficiency and should be considered only as a relative level when compared to other plants in the same experiment.

#### *2.3. RNA-Seq Analysis*

Single samples from leaves from untreated control plants (UTC) and from those treated with PBS 52097 (24 h after application) were used for RNA-sequencing. For each sample, total RNA was isolated using a CTAB-based protocol as described by Chang et al. (1993) [41]. RNA-seq libraries were prepared according to the so-called "dUTP method" to generate mRNA-seq libraries [42,43]. In short, mRNA was purified from 4 μg total RNA using oligo-dT beads, fragmented, and converted to cDNA. Libraries were subsequently made using the Illumina mRNA-Seq Sample Preparation Kit according to the manufacturer's instructions. An amount of 4 pmol of each library was sequenced by BaseClear B.V. (The Netherlands) using the Illumina HiSeq2500 system, with a read length of 50 nucleotides.

Single-end sequencing reads were filtered using the Illumina Casava pipeline version 1.8.3 and Illumina Chastity filtering. Additional filtering on the remaining reads was performed using the FASTQC quality control tool version 0.10.0. For RNA-seq analysis, sequence reads were mapped (per sample) to the reference using CLC Bio Genomics Workbench software (version 5.1.5). As a reference for corn, the publicly available B73 reference sequence (AGPv3.22, downloaded from the ZmGDB genome browser) consisting of 63,241 sequences was used. For soybean, the Glycine max\_275\_Wm82.a2.v1 primary transcripts [44], consisting of 88,647 sequences were used as a reference. To determine gene expression levels and differential gene expression, RPKM values (read counts corrected for library size and transcript length) were calculated using the CLC Bio software. Differentially expressed genes (DEGs) were selected by calculating the ratio of the RPKM value of treated samples over the RPKM value of untreated samples. Only genes with at least 50 reads and an RPKM value of over 5 in at least one sample were considered.

To functionally categorize the corn and soybean gene sequences, gene ontology (GO) terms were assigned to each assembled contig using Blast2GO software (version 3.1). GO terms provide a controlled vocabulary to describe the functions of genes across species. Blast2GO is an automated tool for the assignment of GO terms based on sequence similarity [45]. Statistical assessment of GO term enrichment in groups of DEGs were done using Fisher s Exact Test in combination with false discovery rate (FDR) correction for multiple testing.

#### *2.4. qRT-PCR Analysis*

qRT-PCR analysis was performed on single samples collected 8, 24, 48, and 168 h after PBS(s) application. Total RNA was isolated as described above. First-strand cDNA was prepared using 80 ng total RNA and qScript™ cDNA Supermix (Quanta Biosciences). Two and a half μL of 1:5 diluted first-strand cDNA was used as a template in the subsequent PCR, performed on a Bio-Rad CFX using 5 pmol of both primers (sequence of primers reported in Supplemental Table S2) and PerfeCta SYBR Green SuperMix UNG (Quanta Biosciences) in a final volume of 12.5 μL per reaction, according to the manufacturer s instructions. All transcript levels were normalized using a eukaryotic translation initiation factor gene (corn) or actin gene (soybean) as a control.

#### *2.5. Statistical Analysis of Data*

The statistical analysis was performed using R software (3.3.2 version; R foundation for Statistical Computing, Vienna, Austria). Phenotyping results were analyzed using one-way analysis of variance (ANOVA), and the means were compared with Duncan's New Multiple Range Test (*p* < 0.05).

#### **3. Results and Discussion**

#### *3.1. Phenomic Parameters*

#### 3.1.1. Digital Biovolume

Plant development was assessed through the DB which is a morphometric measurement previously employed in high-throughput (HTP) studies to monitor the influence of abiotic stresses, mainly drought, on plant growth [34,38].

The DB of soybean plants was significantly improved after 10 days from the application of each of the three prototypes tested both at lower (2.5 mL/L) and higher (5 ml/L) concentration, in comparison with UTC (Figure 1A,B). However, the most consistent results (higher DB increase) were obtained using formulation 52097 at the lower dosage (2.5 mL/L; Figure 1A) reaching—at the end of experiment—the mean DB value of 49.58 k units (+72% compared to UTC plants). Such improvement in DB was clearly observed after 10 days from prototype application and maintained during time until the end of DB measurements in soybean, which was 26 days after treatment (Figure 1A).

Parallel measurements on corn confirmed that the 52097 PBS prototype exerted the higher increasing effect on DB in comparison with the other treatments and UTC plants (Figure 1C). In this case, beside an early plant response to the treatment observed already 7 days after treatment, the plants treated with 52097 maintained a greater DB throughout the experiment. No statistically significant differences between the 52113 and control plants were noted. At day 10 after treatment the plants treated with the 52097 reached a DB level of 27.92 k units, 81% higher than that of the UTC plants.

For both soybean and corn, plants treated with prototype 52097 showed a constant and consistent increase in DB accumulation compared to plants treated with 52096 or 52113.

**Figure 1.** Mean values ± SE (*n* = 5) of digital biovolume (DB) measured on (**A**) soybean plants treated at lower dosage, (**B**) soybean plants treated at higher dosage, and (**C**) corn plants treated at higher dosage. Dashed lines and filled black circles (-) identify the untreated control plants (UTC).

#### 3.1.2. Greener Area

The value of greener area (GGA) in soybean showed a general decrease during time (Figure 2A,B), as expected, due to the progression of phenological phases, which lead to senescence and yellowing. This was observed in untreated soybean plants, but also in treated with prototypes 52096 and 52113. Interestingly, soybean plants treated with prototype 52097 at both rates showed a consistent persistence of optimal (ranging from 0.7 to 0.8) GGA values during time (Figure 2A,B), even during the latest phenomic measurements (19–26 DAT).

For the first 12 days after treatment (DAT), no statistically significant differences were observed between the treatments, since all plants showed the same mean GGA value around 0.8 (Figure 2A,B). Starting from 14 days after treatment the first yellowing/sign of senescence were recorded on the untreated control plants and the plants treated with 52113 and 52096 prototypes. At the end of the experiment it was possible to see how, at both rates tested, the application of 52097 on soybean allowed a higher level of GGA, in particular around 0.77 (2.5 ml/L dosage) and 0.75 (5 mL/L dosage). On the other hand, control plants and plants treated with 52113 and 52096 prototypes reached values between 0.40 and 0.28 respectively (Figure 2A,B). It can be concluded that only prototype 52097 was able to preserve and improve GGA in comparison with the other experimental conditions. This can be attributed to a positive effect of prototype 52097 on the "stay-green" condition, that is known to allow plants to maintain high photosynthetic activity [46], and gain benefits on biomass accumulation, as confirmed by the data previously shown on DB (Figure 1).

The positive results on GGA observed after application of prototype 52097 were visibly clear by looking at the set of pictures taken during the cycle, by mean of the visible camera (Figure 2C).

Considering the same test on corn, all the formulations exerted a slight increase in GGA in comparison with untreated control (UTC) plants (Figure 2D), statistically significant at 7, 10, and 13 days after treatment with prototypes 52096 and 52097.

**Figure 2.** Mean values ± SE (*n* = 5) of greener area (GGA) measured on (**A**) soybean plants treated with biostimulants at lower dosage, (**B**) soybean plants treated with biostimulants at higher dosage at 12, 14, 19, and 26 days after treatment (DAT). (**C**) Acquired RGB images of representative UTC (top) and 52097-treated (bottom) soybean plants, showing the effect of PBS application on color and growth during the trial (from T0 to 26 days after treatment). (**D**) GGA measurements taken on corn plants at 7, 10, 13, and 15 days after treatment (DAT). Solid black bars identify the untreated control plants (UTC).

#### 3.1.3. Stress Index

The measurement of Stress Index of treated and untreated soybean plants did not show statistically significant differences during the first four data acquisitions time (Figure 3). As expected, due to the plant cycle progression and senescence, during the last three data acquisitions a higher Stress Index—ranging from 0.4 to 0.7—was observed for UTC. Treating plants with 52096 or 52113 did not affect the Stress Index. Interestingly, soybean plants treated with 52097 showed a lower level of Stress Index than the other treatments, with values stable around 0.2 (Figure 3).

**Figure 3.** Mean values ± SE (*n* = 5) of Stress Index (SI) measured on (**A**) soybean plants treated at lower dosage, (**B**) soybean plants treated at higher dosage. Dashed lines and filled black circles (-) identify the untreated control plants (UTC).

This was not observed on corn plants, where both UTC and treated plants showed a Stress Index value ranging from 0.2 to 0.4 throughout the experiment (Supplemental Table S3).

#### *3.2. Molecular Analyses*

Based on the obtained results, although some differences were observed between soybean and corn, we selected compound 52097 as the best candidate for further NGS analyses. Leaf tissue from untreated soybean and corn plants was compared to its 52097-treated counterpart by RNA-seq analysis. Per sample, over 25 million single reads were generated and mapped to the relevant reference transcriptome (see Material and Methods). For corn, around 77% of the available sequencing reads could be mapped to this assembly, for soybean this was 89%. By comparing 52097-treated samples to untreated controls, differentially expressed genes (DEGs) were identified. Naturally, the number of DEGs depended on the fold-change threshold applied (Table 1). In general, the number of DEGs was higher in corn than in soybean. Lists of the 20 most upregulated genes for both crops are provided in Supplemental Table S4 (soybean) and Table S5 (corn).

**Table 1.** Number of differentially expressed genes (up and down-regulated) in corn and soybean 24 h after application of formulation 52097 when compared to mock-treated control plants.


#### *3.3. Functional Annotation Using Gene Ontology*

Of the 63,241 sequences present in the reference transcriptome from corn, 49,426 sequences (78%) could be functionally annotated using GO, meaning that one or more biological processes, molecular functions, or cellular localizations could be linked to these sequences based on sequence homology. For soybean, from the 88,647 genes present in the soybean transcriptome, 70,776 sequences (80%) could be functionally annotated using GO.

For corn, all DEGs up- or downregulated more than 3-fold were used for enrichment analysis. Several biological processes, including nitrogen assimilation, maltose biosynthesis, and cytokinin metabolism were enriched among the 331 upregulated genes from corn (Figure 4A). Analysis of the 593 downregulated corn transcripts resulted in 55 enriched GO-terms (biological process). By filtering out the most reduced GO-terms, that is, removing parent terms of already present statistically significant child GO terms, a list of 13 significantly enriched biological processes remained (Figure 4B). These terms included divalent metal ion transport, response to carbohydrate, phenylalanine degradation, and flavonoid biosynthesis.

For soybean, first, all DEGs up- or downregulated more than 3-fold were used for enrichment analysis. Analysis of the 65 upregulated soybean transcripts resulted in 25 enriched GO-terms (biological process), including metal ion transport, sulfate reduction, asparagine biosynthesis, and serine metabolism (not shown). However, there were no significantly enriched GO terms among the 59 downregulated soybean transcripts. For this reason, it was decided to include all soybean contigs with a FC greater than 2. This resulted in 16 significantly enriched (reduced) GO terms for the 278 upregulated contigs (Figure 4C), and 13 for the 321 downregulated contigs, including auxin-activated signaling, sulfur amino acid metabolism, iron transport, and sulfur compound biosynthesis (Figure 4D).

From both crops, two individual DEGs were selected using the results from the GO enrichment analysis. For corn, these were a cytokinin dehydrogenase (CKX; Figure 5A) and a glutamine synthetase (GS; Figure 5B). CKX catabolizes the plant hormone cytokinin and plays an important role in cytokinin regulated processes [47]. GS is required for nitrogen assimilation and allocation within the plant, and for nitrogen remobilization in both source and sink tissues [48]. GS is important for ammonium assimilation in roots, during senescence, and during photorespiration. Several studies have indicated that GS plays an essential role in plant development and yield formation in cereals. For example, in corn, leaf-localized GS are of specific importance for the development of the cob with respect to kernel number and kernel size [49]. A putative GS gene is induced by treatment with product 52097 (Figure 5B).

In addition, 3 more genes from corn were selected on the basis of their functionality: An SPX domain-containing protein (named after SYG1/Pho81/XPR1 proteins; Figure 5C), a NRT1/PTR FAMILY (NPF) protein (Figure 5D, and a polyol/monosaccharide transporter (PMT; Figure 5E). It is well described that plant growth and development are highly dependent on the availability of inorganic phosphate (Pi). Among the many proteins involved in the plant response to Pi starvation, proteins containing the SPX domain are key players involved in the maintenance of internal levels of Pi. Indeed, SPX genes have been reported to be induced upon Pi starvation in roots and shoots, and proteins harboring the SPX domain have been shown to be involved in P use efficiency [50]. Members of the plant NITRATE TRANSPORTER 1/PEPTIDE TRANSPORTER (NRT1/PTR) family display protein sequence homology with peptide transporters in animals. In comparison to their animal and bacterial counterparts, the plant NRT1/PTR family proteins transport a wide variety of substrates: nitrate, peptides, amino acids, dicarboxylates, glucosinolates, IAA, and ABA [51]. The transcript identified here shows the highest similarity to the first identified member of the NRT1/PTR family: NRT1.1. NRT1.1 is an Arabidopsis nitrate transporter that also functions as a nitrate sensor and can transport auxins. As such, it links nutrient and hormone signaling [51]. PMTs are proteins capable of transporting a range of sugar alcohols and monosaccharides including glucose, fructose, sorbitol, mannitol, xylitol,

xylose, and galactose [52]. PMTs are believed to be involved in phloem loading [52]. Hence, the induction of a PMT gene identified in this study could point towards increased phloem loading.

**Figure 4.** Gene ontology (GO) term enrichment analysis of the differentially expressed contigs from prototype 52097-treated plants (24 h after application; false discovery rate (FDR) < 0.05). (**A**) Upregulated GO terms in corn. (**B**) Down-regulated GO terms in corn. (**C**) Upregulated GO terms in soybean. (**D**) Down-regulated GO terms in soybean. The absolute number of contigs in the test set is represented by the bars in the graphs on the right (numbers of contigs in the test/reference sets are reproduced next to each bar).

**Figure 5.** Gene expression of selected genes as determined by qRT-PCR. Samples were collected before (T0), and 8, 24, 48, and 168 h after treatment with formulation 52097. Expression of genes (**A**) *CKX (cytokinin dehydrogenase)*, (**B**) *GS (glutamine synthase)*, (**C**) *SPX domain-containing protein*, (**D**) *NPF* family protein (*NRT1*/*PTR*), and (**E**) *PMT* (*polyol*/*monosaccharide transporter*) was analysed in corn, while (**F**) *ZIP* (ZRT, IRT-like transporter) and (**G**) *AS* (*asparagine synthetase*) expression was assessed in soybean samples. Relative expression levels are in ddCt.

For soybean, we selected a ZIP (ZRT, IRT-like protein; Figure 5F) transporter and an asparagine synthetase (AS; Figure 5G). ZIP transporters are important during uptake and transport of zinc and iron and other divalent metal cations [53]. AS, like GS, functions in nitrogen metabolism [54].

For all these genes, expression was determined in leaf samples collected on several timepoints after application of prototype 52097. It was observed that data obtained by qRT-PCR corroborated the NGS results. Twenty-four hours after application, the differences in gene expression between treated and untreated leaves were very comparable. The additional timepoints showed that GS and SPX expression already peaked 8 h after application (Figure 5B,C), whereas the other 5 genes reached their maximum expression after 24 h (Figure 5A,D–G).

#### **4. Conclusions**

This study highlights the use of high-throughput/efficiency plant phenotyping (phenomics) together with Next Generation Sequencing to investigate the effectiveness and mechanism of action of

new biostimulant formulations. Such formulations were specifically conceived as foliar applications to increase yield of different row/industrial crops, such as corn and soybean.

Phenomic-based measurements of digital biovolume, Greener Area, and Stress index allowed us to select 52,097 (commercial name "YieldOn®") as the most effective prototype among the ones tested. Subsequently, through NGS a deep characterization of the molecular mechanisms by which the biostimulant under investigation exerts its positive effect was performed. This analysis explained the mechanism of action of the biostimulant under investigation, which in corn upregulated specific processes like nitrogen and phosphate assimilation and metabolism, maltose biosynthesis, sugar transport and phloem loading, hormone (cytokinin) metabolism. In soybean nitrogen metabolism, metal ion transport (mainly zinc and iron), sulfate reduction, and amino acid biosynthesis were upregulated.

In conclusion, the results showed in this work support the integration of multiple "omics" as robust and objective tools in the discovery, evaluation, and development of innovative, sustainable, and targeted solutions to meet the emerging needs of row-crops agriculture.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4395/9/11/761/s1, Table S1: Temperatures, humidity (RH%), and radiation (PAR μmol m−<sup>2</sup> s<sup>−</sup>1) parameters measured during the trial. Table S2: Sequence of primers used for qRT-PCR analysis; Table S3: Stress Index measurements in corn; Table S4: lists of the 20 most upregulated genes for soybean; Table S5: lists of the 20 most upregulated genes for corn.

**Author Contributions:** N.B. Conceptualization, main writing and structuring of the paper, High efficiency phenotyping Data Curation and Formal analysis; A.P. Methodology, High efficiency phenotyping Data Curation; F.A.H. Writing, Methodology, Next Generation Sequencing Data Curation and Formal Analysis; N.V. Methodology, Next Generation Sequencing Data Curation and G.P. Conceptualization, writing, review, editing and supervision.

**Funding:** The research conducted was funded by Valagro SpA, which provided biostimulant prototypes that were blindly identified by a numbered code to avoid any conflict of interest in the data production. The prototypes were named only at the end of the experiments.

**Acknowledgments:** The authors wish to thank Prem Warrior and Giuseppe Montanaro for their critical review of the manuscript. YieldOn® is a trademark registered by Valagro in Italy and other countries.

**Conflicts of Interest:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

#### **References**


© 2019 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*

### *Bacillus subtilis* **CBR05 for Tomato (***Solanum lycopersicum***) Fruits in South Korea as a Novel Plant Probiotic Bacterium (PPB): Implications from Total Phenolics, Flavonoids, and Carotenoids Content for Fruit Quality**

**Murugesan Chandrasekaran 1, Se Chul Chun 2, Jae Wook Oh 3,\*, Manivannan Paramasivan 4, Ramesh Kumar Saini <sup>5</sup> and Jesudoss Joseph Sahayarayan <sup>6</sup>**


Received: 7 September 2019; Accepted: 29 November 2019; Published: 3 December 2019

**Abstract:** Plant growth-promoting rhizobacteria (PGPR) are naturally occurring soil biota which benefit plants by improving plant productivity and immunity. The aim of the present work was to evaluate the effect of the inoculation of PGPR strain, *Bacillus subtilis* CBR05 on the quality of tomato fruits produced under greenhouse conditions. Results were compared with mock-inoculated control and market sample. We found a significant increase in total phenol and flavonoid contents of tomato fruits in PGPR strain *B. subtilis* CBR05 inoculated plants compared to those of mock-inoculated control and market sample. Moreover, *B. subtilis* CBR05 inoculation stimulated antioxidant activities and levels of carotenoid (β carotene and lycopene) content in plants. In addition, the inoculation of the strain *B. subtilis* CBR05 produced the highest content of lycopene (21.08 μg/g FW) in tomato fruits as compared to mock-inoculated plants. Our results show that the PGPR strain *B. subtilis* CBR05 is a versatile soil bacterium that enhances tomato production by elevating antioxidant activities and carotenoid (β carotene and lycopene) levels in fruit.

**Keywords:** *Bacillus subtilis*; tomato; antioxidant activity; carotenoids; probiotics; PGPR

#### **1. Introduction**

Tomato (*Solanum lycopersicum*) is regarded as the second most vegetable crop next to potato in the agricultural implications of human consumption. According to agricultural statistics, tomatoes along with sweet corn and snap beans constitute 93% of crop production and processing strategies (Agricultural Statistics, United States Department of Agriculture (USDA), 2016). The positive benefits of tomato consumption have been rigorously proved against a variety of diseases like chronic degenerative diseases, owing to the escalated content of significant phytochemicals with potent health benefits, like the carotenoids (β-carotene and lycopene), the glycoalkaloids (dehydrotomatine and α-tomatine),

ascorbic acid, tocopherols, and many phenolic and flavonoid compounds [1–3]. Tomato also contributes as a major dietary ingredient for Vitamin A and C which implies increased per individual consumption in the United States and many Western countries [2,4]. Fruit ripening in tomato comprises a cascade of events on biochemical, physiological, and structural perspectives involving the influence of secondary metabolites that confer flavor, aroma, texture, and appearance of the tomato [5,6]. Accumulation of large quantities of pigments, especially lycopene and β carotene, inside the plastoglobules of chromoplast provides a visual indication that the fruit is mature and suitable for consumption [4].

Agronomic practices are recognized as a vital factor in determining the nutritional quality of tomato crops [7,8]. The nutrient contents in tomato fruits depend on the environment in which they grow [9,10]. Nowadays, the use of crop modeling to identify effective farmer strategies to counteract adverse future climatic conditions has become a standard in climate change impact assessments [11–13]. Over the past few years, a variety of methods have been proposed to comprehensively assess fruit quality and its relationship with water, including principal component analysis (PCA), analytic hierarchy process (AHP), gray relational analysis (GRA), and technique for order preference by similarity to ideal solution (TOPSIS) [14–17]. The challenge of producing fresh fruits and vegetables is increasing both yield and quality to satisfy consumers as the environment changes in ways that are deleterious to crop species [18]. The quality of agricultural products is affected by many pre- and postharvest factors [9]. The utilization of biofertilizers that mitigate these adverse environmental effects has become a feasible and beneficial production practice. Plant growth-promoting rhizobacteria (PGPR) may be considered as preharvest biotic factors that mitigate adverse environmental effects and promote improved crop yield and quality [19,20].

Among various PGPR approaches, *Bacillus* species are considered as likely candidates due to their broad-spectrum antagonistic activity against phytopathogens, production of long-lived and stress-tolerant spores, secondary metabolites, lytic enzymes, resistance to adverse environments, and plant growth promotion [19,21–23]. *Bacillus subtilis* plays a significant role in improving plant growth and tolerance to both biotic and abiotic stresses. PGPR strains also act as bio-stimulants of phytohormones and peptide synthesis [20,24], but studies of the PGPR strain, *B. subtilis* CBR05 on tomato have not yet appeared. Preharvest factors that directly affect crop yield and quality can be summarized into biological factors comprising pathological, entomological, and animal issues, which was found to be nullified upon increased usage of PGPR. The dire need for assessment of tomato and tomato-based products is given significant attention concerning nutrition and quality relying on the nature of the variety, maturity at harvest, effective transport, and storage [25]. Characterization of the carotenoids, mainly β-carotene and lycopene during storage and various ripening stages, shows drastic developments in sustainable yield and quality parameters of tomato [26]. This information adds to our understanding of temporal differentiation of nutritionally significant phytochemicals during ripening of tomato fruits. The objective of this study was to evaluate the effects of *B. subtilis CBR05* on the quality of tomato fruits under greenhouse conditions.

#### **2. Material and Methods**

#### *2.1. Chemicals and Reagents*

Authentic standards of carotenoid, all-*E*-lutein, was purchased from Cayman Chemical Company, Michigan, USA. All-*E*-β-carotene were purchased from Sigma-Aldrich, St. Louis, MO, USA. Only the HPLC grade of organic solvents was employed in carotenoid extraction (Daejung, Siheung*-si*, Korea).

#### *2.2. Bacterial Strain and Culture Conditions*

*B. subtilis* CBR05 isolated from our lab maintained on tryptic soy agar (TSA) plates. For long-time storage, bacterial cultures were maintained in tryptic soy broth at −80 ◦C. For experimental purposes, the cultures were transferred to TSA (MBCell, Seoul, Korea) and incubated at 30 ◦C for 24 h. The inoculum mixture of the strain was prepared by culturing in nutrient broth and incubating

at 28 ◦C with constant shaking at 130 rpm. The bacterial cells were centrifuged at 10,000 g for 10 min at 4 ◦C. The cell precipitate was resuspended in 10 mM MgCl2 and the cell concentration of bacterial suspensions of *B. subtilis* CBR05 was adjusted to 10<sup>8</sup> colony-forming unit (CFU)/mL (OD600 = 1.0) for further studies.

#### *2.3. Plant and Growth Conditions*

Tomato seeds (Korean cultivar, Kwangbok) utilized for this study were obtained from a Korean seed resource center, Seoul, South Korea. They were surface sterilized in sodium hypochlorite, rinsed several times with distilled water, and planted onto pots containing sterilized growth media (Peat moss with perlite in a ratio of 3:1). Two sets of plants (three plants per set) were maintained, one without PGPR (mock-inoculated control) inoculation and the other with PGPR strain, *B. subtilis* CBR05 inoculum. All treatments were placed randomly in the greenhouse and replicated 3 times. Plant growth-promoting *B. subtilis* CBR05 were applied under sterile conditions to the base of the plants close to the roots to ensure better colonization. Plants were maintained under greenhouse conditions at a temperature of 25 ◦C with watering carried out every alternate day, to make a better availability for nutrition and plant growth promotion.

#### *2.4. Antioxidant Assays*

#### 2.4.1. DPPH Assay

1,1-diphenyl-2-picrylhydrazyl (DPPH) assay was performed to assess the fractions exhibiting scavenging property of free radicals in vitro [27]. Then, 0.2 mM solution of DPPH in ethanol was added to the fraction of aliquot at concentrations (100 μg/mL). The mixture was allowed to stand for 30 min and the absorbance was measured at 517 nm using a UV-Visible spectrophotometer. The percent scavenging activity was determined and Trolox was used as the standard.

#### 2.4.2. ABTS (2,2'-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) Assay

A Trolox equivalent antioxidant capacity (TEAC)/ABTS assay was conducted based on the method of Ramos et al. [28]. The ABTS solution (7 mM) was oxidized with potassium peroxodisulfate (2.45 mM) for 16–18h at room temperature. The ABTS solution was diluted with solvents. An aliquot (100 μg/mL) was mixed with diluted ABTS solution and the absorbance was read at 734 nm. Trolox and ascorbic acid were used as reference standards.

#### *2.5. Determination of Total Phenolic Contents*

The Folin–Ciocalteu method was used to estimate various concentrations concerned with the total phenolic content. The extracts were dissolved in absolute methanol and later 200 μL of the extract was mixed with 800 μL of 1 N Folin–Ciocalteu reagent (1: 10). After 5 min at room temperature, 3 mL of sodium carbonate (15%) was added to the extracts. Following incubation for 30 min at room temperature, the absorbance was read at 765 nm using a UV-spectrophotometer. A standard curve for gallic acid equivalents (GAE) (milligrams per gram of extract (mg GAE/ge)) was utilized to evaluate the concentration of total phenolic compounds. Analyses were performed in triplicate per each extract.

#### *2.6. Determination of Total Flavonoid Contents*

Screening the content of flavonoids was done by a modified protocol as reported previously [29]. The extracts were dissolved in absolute methanol. In a 15 mL conical tube, 1 mL of a sample was mixed with 0.3 mL of 5% sodium nitrite, followed by incubation for 5 min. After incubation, 0.3 mL of aluminum chloride (10%) and 2 mL of sodium hydroxide (1 mol/L) were added to the reaction mixture, and the absorbance was read at 496 nm with a UV-spectrophotometer, using catechin as the standard. Quercetin equivalents (QE) present per g of extract (mg QE/ge) was used to quantify the expression levels.

#### *2.7. Extraction and Quantification of Carotenoids*

Carotenoids were extracted in triplicates and quantified according to previously established protocol with minor modifications [30,31]. All the preparations were performed in low light conditions to avoid the degradation of carotenoids. Three independent biological samples were extracted separately. Briefly, one whole tomato fruit was finely chopped and mixed thoroughly. Five grams of chopped fruits (exact to 0.001 g) from each treatment were separately transferred into test tubes containing 20 mL of acetone and 0.1% butylated hydroxytoluene (BHT: *w*/*v*). The samples were homogenized with a mechanical homogenizer and centrifuged at 5000× *g* (5 min at 4 ◦C temperature). The supernatant was recovered and pelleted samples were repeatedly extracted until the pallets became colorless. Supernatants from all extractions were pooled and vacuum-dried in a rotary evaporator (Temperature < 35 ◦C) (Büchi RE 111, Switzerland).

#### *2.8. HPLC Analysis*

The extract was recovered with 10 mL of methylene chloride (CH2Cl2) containing 0.1% BHT and transferred to an amber color HPLC vial for HPLC analysis. The chromatographic separation was achieved using an Agilent Model 1100 HPLC instrument (Agilent Technologies Canada Inc., Mississauga, ON, Canada) equipped with a degasser, autosampler, dual pump, and diode array detector (DAD). Samples were scanned (200–800 nm) with 0.05 min (1 s) response time, 8.0 mm slit width, and a detection wavelength of 450 (for most of the carotenoids) and 470 nm (for lycopene). The bandwidth was ±16 nm for all detection wavelengths. Similarly, 600 nm was used as a reference wavelength with ±50 nm bandwidth in all detections. The column used was a YMC, C30 carotenoid column, 250 × 4.6 mm, 5 μm (YMC, Wilmington, NC, USA), and the chromatographic data were recorded with ChemStation LC 3D software. The column thermostat was maintained at 25 ◦C temperature. Then, 20 μL of standards and samples were injected with an autosampler. The solvent system consisted of Methanol: methyl tertiary butyl ether (MTBE): water (81:15:4) (Mobile phase A) and MTBE: Methanol (91:9) (Mobile phase B). The gradient elution was 0%–100% B in 90 min, and 5-min post-run at a flow rate of 1 mL/min.

#### *2.9. Statistics*

All of the experiments were conducted in triplicate and results were tabulated as the Mean ± standard deviation (SD). Statistical significance of the data was determined using one-way analysis of variance (one-way ANOVA) followed by Fisher LSD (Least Significant Difference) test. Data analyses were performed using Sigmastat v8.02 (Systat Software Inc., San Jose, CA, USA). A *p* value of ≤ 0.05 was considered significant.

#### **3. Results**

In the present investigation, the effects of the PGPR strain, *B. subtilis* CBR05 inoculation on the maintenance of carotenoids, total phenolics, flavonoid contents, and antioxidant properties were evaluated. The results revealed that the PGPR strain, *B. subtilis* CBR05 has the capacity to improve the plant growth and change some of the tomato fruit quality characteristics under greenhouse conditions. Tomato plants treated with PGPR strain, *B. subtilis* CBR05 showed significantly higher biomass compared with mock-inoculated controls. Significant increases in root length and dry weight, over mock-inoculated controls, were achieved in green-house conditions (data not shown). In addition, antioxidant activities of three tomato fruits (mock-inoculated, *B. subtilis* CBR05 inoculated, and market fruit) were determined using both DPPH and ABTS radical scavenging method. The extract of fruits from plants inoculated with *B. subtilis* CBR05 strain was the most active against DPPH and ABTS radical and that of fruits from the market, the differences being significant when compared to mock-inoculated control (Figure 1).

**Figure 1.** Antioxidant assays ((2,2'-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) and 1,1-diphenyl-2-picrylhydrazyl (DPPH)). Values are mean ± standard deviation from three replicates. Bars followed by the same letter(s) are not significantly different (*p* ≤ 0.05).

The total phenolic content of the fruit extracts is shown in Figure 2. Among the fruits, the PGPR strain, *B. subtilis* CBR05 inoculated plants had the highest phenolic content followed by market fruits. Moreover, the flavonoid contents were also elevated in PGPR treated tomato fruits as compared to those of both mock-inoculated control and market fruits (Figure 2). We did not find any significant differences between mock-inoculated control and market fruit. The results of this study show that *B. subtilis* CBR05 is an effective probiotic agent for the promotion of tomato fruit quality.

**Figure 2.** Total phenol and flavonoids contents. Values are mean ± standard deviation from three replicates. Bars followed by the same letter(s) are not significantly different (*p* ≤ 0.05).

The carotenoids, such as lycopene and β-carotene, were extracted and quantified (Table 1). The contents and composition of carotenoids were found to differ among control, *B. subtilis* CBR05 inoculated, and market fruits (Figure 3). Carotenoid levels (lycopene) in tomato fruits in plants treated with PGPR strain, *B. subtilis* CBR05 were higher than both the control and market fruits (Table 1; Figure 3).


**Table 1.** Contents of carotenoids in tomato fruits.

In each experiment, means followed by different superscript letters are significantly different among the treatment groups (*p* ≤ 0.05). Values are mean ± standard deviation from three replicates.

**Figure 3.** Carotenoids' contents (lycopene and β-carotene). Values are mean ± standard deviation from three replicates. Bars followed by the same letter(s) are not significantly different (*p* ≤ 0.05).

The present investigation extends their results by comparing the contents of carotenoids from the fruits using HPLC–DAD (Figure 4). Using this methodology, all-E- β -carotene and all-E- lycopene were identified as the major carotenoids in tomato fruits, based on retention time with standards and by comparing the peak spectra recorded with a DAD during the analysis. The chromatograms (470 nm) and the peak spectra of major identified peaks were shown in Figure 4. The other minor carotenoids were not quantified due to the unavailability of standard compounds. We did not find any significant differences in β-carotene. However, in the present study, we have recorded a significantly higher amount of lycopene (All-E-lycopene) in fruits during ripening in PGPR strain, *B. subtilis* CBR05 inoculated plants than those of both mock-inoculated control and market fruits. These results validate the productive roles of the *B. subtilis* CBR05 in enhancing the nutritional potential of tomato fruits.

**Figure 4.** HPLC chromatogram of carotenoids.

#### **4. Discussion**

The use of PGPR is increasing in agriculture and may offer an attractive alternative to synthetic chemicals and fertilizers. Plant growth-promoting microorganisms are efficient microbial competitors that can promote plant growth by producing phytohormones and/or by increasing available nutrients through production of secondary metabolites or act as biocontrol agents to protect plants from infection by phytopathogens [19,21–24]. There have been many reports on PGPR and their effective roles [19–24]. Insufficient experimental work has been reported to speculate on the mechanisms of PGPR effects on fruit quality. In the present study, PGPR strain, *B. subtilis* CBR05 isolated from rice were used as inoculants for tomato plants grown under greenhouse conditions. Our results showed that PGPR inoculations significantly increased the total biomass and root length compared to those in the control. *B. subtilis* CBR05 appears to impart plant growth promotion effects that are distinct from other commercial biocontrol agents. Tomato fruits are a good source of antioxidant compounds that can reduce harmful oxidation reactions in the human body, thus preventing various diseases associated with free radical oxidation, such as cardiovascular and neurological disorders and cancer [1–4].

Antioxidant activity has been widely used to test the ability of plant extracts to act as free radical scavengers [32]. In the present investigation, *B. subtilis* CBR05 had a net positive effect on the antioxidant activity measured by the DPPH and ABTS scavenging capacity, which seems to indicate that the bacteria acted as a regulator of the synthesis of antioxidant compounds in the plant (Figure 1). Strong scavenging of ion radical was exhibited by the inoculated tomato fruits, thus showing that *B. subtilis* CBR05 inoculation increased the radical scavenging capacity of tomato fruits. In a previous study, *B. licheniformis* inoculated plants had increased antioxidant profiles in tomato plants under greenhouse conditions [32]. Similar results concerned with enhanced fruit quality and marketable grade have been reported for other crops under the influence of PGPR [33] but this report is the first of its kind to specifically address PGPR strain, *B. subtilis* CBR05 in improvement of fruit quality. Moreover, PGPR enhances fruit characteristics based on the mediation of increased availability of nutrients to plants like phosphorous and iron, enhancing the nutritional status of the plants in the rhizosphere [33–35].

Phenolic contents perform an essential role in plant resistance and defense against phytopathogens, which are closely linked with reactive oxygen species (ROS). This study reveals that among the selected tomato fruits, PGPR strain, *B. subtilis* CBR05 inoculated tomato fruits had the highest amount of phenolics (Figure 2). Some phenolic compounds may prevent oxidative damage in vivo and thus protect against the development of the disease such as cardiac disease and cancer [1,2]. This might be considered as useful for health purposes. In addition, inoculation of *B. subtilis* significantly increased

flavonoid content compared to those of both control and market fruits. Similarly, increases in total flavonoids content by *B. licheniformis* have been reported for tomato fruits [32]. In our previous studies, we also found that *B. subtilis* CBR05 inoculation enhanced the accumulation of peroxidase and polyphenol oxidase enzymes, which are involved in the metabolism of phenols and flavonoids [22,23]. Hence, this shows that *B. subtilis* colonization induces resistance against biotic and abiotic stress agents. The results of antioxidant assays also revealed that tomatoes are a rich source of antioxidants, thus their habitual consumption can potentially help to combat the oxidative stress.

Regulation of carotenoid biosynthesis and high-accumulation lycopene during tomato fruit development is widely studied [36–38]. In the present study, we found a significantly higher amount of lycopene (All-E-lycopene) in *B. subtilis* CBR05 inoculated tomato fruits (Figures 3 and 4). Lycopene possesses the highest antioxidant potential among the carotenoids and several other antioxidants found in fruits and vegetables [39]. Thus, the addition of PGPR enhances lycopene content in tomato fruits and can potentially contribute to antioxidant levels of diets. This potent antioxidant activity of lycopene protects from a variety of ROS and reactive nitrogen species (RNS), thus helping in preventing chronic diseases in humans [31,36]. Similar to the lycopene contents, the DPPH and ABTS antioxidant activity of *B. subtilis* CBR05 inoculated tomato fruits was much higher than both the control and market fruits. Earlier, we also reported that defense-related enzymes in tomato after treatment with *B. subtilis* CBR05 efficiently combated *X. campestris* pv, *vesicotoria* [40] and induction of defense-related enzymes like superoxide dismutase, catalase, peroxidase, and polyphenol oxidase assessment revealed the up-regulation of glucanase and phenyl ammonia lyase indicating induced systemic resistance (ISR) in tomato. The earlier results established that the antioxidant capabilities of tomatoes are naturally present. Further, it was specifically proved that *B. subtilis* CBR05 mechanism of disease resistance against *X. campestris* pv, *vesicotoria* was confirmed for the involvement of the de novo pathway involved in Vitamin B6 biosynthesis [41]. In addition, known bacterial elicitors of ISR are microbial associated molecular patterns triggering immunity. When this fails, microbial effector-triggered immunity is induced and leads to programmed cell death. It increases the plant's systemic resistance to subsequent pathogen challenge by PGPR. Moreover, plant probiotic bacterium (PPB) could be used to reduce the use of chemicals (fertilizers, pesticides) in agriculture. This could lead to improved quality at reduced costs and could provide the basis for more sustainable agriculture [42]. Organic agriculture has been widely promoted and adopted to establish better sustainability in food production and crop protection. *Micromonospora* has been regarded as a PPB due to rhizobia helper bacteria properties in *Medicago sativa* L. [43]. Similarly, *Phyllobacterium* and *B. licheniformis* also show promising benefits for increasing vitamin C content across various functional foods and have been considered as PPB, devoid of economic loss [32,44].

In the present investigation, we extracted and quantified phenols, flavonoids, and carotenoids, in the mock-inoculated control, *B. subtilis* CBR05 inoculated plants and market fruits. Among them, PGPR strain, *B. subtilis* CBR05 inoculated tomato fruits were found to have the richest source of lycopene, total phenolics, and flavonoids contents. Additionally, the PGPR strain, *B. subtilis* CBR05 inoculated tomato fruits showed potent antioxidant activities. The significantly higher lycopene content and radical-quenching activity of PGPR strain, *B. subtilis* CBR05 inoculation confer tomato fruits with more nutritional value, and their consumption can minimize oxidative stress-mediated chronic diseases. Thus, biofertilizers based on PGPR may be a viable alternative to improve the nutraceutical quality of greenhouse-produced tomato fruits.

**Author Contributions:** Conceptualization, M.C., and S.C.C.; methodology, M.C., R.K.S., and S.C.C.; formal analysis, M.C., R.K.S., M.P., and S.C.C.; writing—Original draft preparation, M.C., M.P., J.J.S., and J.W.O.; writing and editing, M.C., J.J.S., M.P., and J.W.O.

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

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

#### **References**


© 2019 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* **E**ff**ects of Water Stress and Modern Biostimulants on Growth and Quality Characteristics of Mint**

#### **Hosam O. Elansary 1,2,\*, Eman A. Mahmoud 3, Diaa O. El-Ansary <sup>4</sup> and Mohamed A. Mattar 5,6,\***


Received: 5 November 2019; Accepted: 15 December 2019; Published: 18 December 2019

**Abstract:** Natural biostimulants combine different elicitors that may influence economic properties of herbal crops, such as mint. Mint (*Mentha longifolia* L.) plants were subjected to three water levels based on container substrate capacity (CSC; 100% CSC, 70% CSC, and 50% CSC) and/or applications of four biostimulants (CRADLE™, Mobilizer™, Nanozim De'Lite™ [ND], and Nanozim NXT™ [NN]). ND and NN exhibited higher vegetative growth and root dry weight than the control (without biostimulants) and other treatments. NN produced the highest fresh and dry mint yields under all water levels. Irrigation water-use efficiency (IWUE) of NN was highest (2.78 kg m<sup>−</sup>3) with 70% CSC, whereas the control produced the lowest IWUE (1.85 kg m<sup>−</sup>3) with 100% CSC. Biostimulants boosted physiological and metabolic responses, including gas exchange, leaf water potential, relative water content, and proline accumulation of stressed plants. NN treatment with 70% CSC had the highest essential oil (EO) ratio (3.35%). Under 70% and 50% CSC with NN treatment, the proportion of 1,8-cineol increased and that of pulegone decreased in EOs. Increased antioxidant activities, reduced H2O2 levels, and increased catalase and superoxide dismutase activities were observed. Applications of ND and NN during water stress conditions increased economic and medicinal properties of mint EOs with applications in the agricultural and pharmaceutical industries.

**Keywords:** *Mentha longifolia*; biostimulants; *Ascophyllum nodosum*; humic acid; antioxidants

#### **1. Introduction**

Mint plants have a long history as traditional medicinal plants [1]. *Mentha longifolia* L. belongs to the family Lamiaceae and naturally occurs in Egypt, Saudi Arabia, and most Arabian countries. The fresh/dried plants are mainly used as an herbal medicine for the treatment of indigestion, menstrual pain, coughs, asthma, fever, and headaches [2,3]. The fresh leaves are used in soft drinks and as garnishes for salads in some countries. The essential oil (EO) is used in the pharmaceutical, cosmetic, and food industries [4]. The EOs exhibit strong antimicrobial activity against several microorganisms [3].

Water stress is one of the major limiting factors for agriculture and food safety worldwide [5]. This stress causes reduced vegetative growth and great losses to farmers. Different studies have focused on the effects of water stress on the growth parameters and EO yield. Zade et al. [6] reported that water stress decreased peppermint plant fresh and dry weight, leaf number, plant height, and root dry weight but nonetheless increased EOs compared to that of normal irrigation in greenhouse and field experiments. Figueroa-Pérez et al. [7] showed that water stress decreased fresh and dry weights of peppermint but increased composition of plant secondary metabolites and antioxidant capacity. Ekren et al. [8] reported that plant height and yield of purple basil were negatively affected by water stress, whereas the EO content increased and irrigation water-use efficiencies were not significantly affected. Shormin et al. [9] showed that the harmful effects of water stress on Japanese mint yield could not be compensated by high nitrogen quantities. Farahani et al. [10] also reported the highest content of EO in balm occurred at 60% field capacity (FC). However, in other studies, Khorasaninejad et al. [11] showed that water stress had negative effects on some growth parameters and EO content of peppermint plants. Razmjoo et al. [12] found that this stress reduced some growth parameters and EO content of chamomile.

Several approaches have been applied to control water stress, such as the use of biostimulants. Modern biostimulants have been produced to increase the productivity and the quality of horticultural crops and help the plants tolerate stress conditions. Some of these biostimulants are mixtures of seaweed extracts, humic acid, and macro and micro elements, whereas other products contain mixtures of mycorrhiza and seaweed extracts, as well as other micro elements. Seaweed extracts work as elicitors for plant secondary metabolites, including EOs and may increase the pharmaceutical properties against microorganisms [13,14]. However, the effects of the mixtures of seaweed extracts and other elicitors, such as humic acid and specific minerals have not been investigated for mint plants. Further, water stress may cause significant changes in the EO composition and these changes might cause parallel changes in the antimicrobial properties of respective EOs.

In this investigation, our goal was to determine the effects of water stress and commercial biostimulants on the growth, physiology, secondary metabolites, and antioxidant activities of mint (*Mentha longifolia* L.). We propose that these natural biostimulants modulate growth, EO ratio, and EO constitutes, leading to enhanced bioactivity of mint plants. These effects indeed have the potential to have future agricultural industry applications.

#### **2. Materials and Methods**

#### *2.1. Plant Material*

Uniformly rooted cuttings of mint (*Mentha longifolia* L.) were brought from nurseries of the Alexandria University farm in February 2018 and 2019 (as two successive growing seasons). The species was identified and vouchered by Hosam Elansary in the Faculty of Agriculture, Alexandria University. The sandy soil (75.5%, 13.2%, and 11.3% of sand, silt and clay, respectively) samples were air dried and sieved with a 2 mm mesh. The soil had an FC of 20.5%, wilting point of 9.6%, electrical conductivity of 0.36 mS cm<sup>−</sup>1, organic matter of 1.4%, pH of 6.2, total nitrogen of 0.085%, total phosphorus of 0.05%, and total sulfur of 0.03%. After proper soil preparation, the plants were grown in 2.1 L plastic pots containing the natural sandy soil supplemented with Crystalon® (65 kg N ha−<sup>1</sup> as urea, 40 kg P2O5 ha−<sup>1</sup> as triple superphosphate, 34 kg K2O ha<sup>−</sup><sup>1</sup> as potassium sulfate, and 2 g L−<sup>1</sup> media) in the greenhouse. The temperature inside the greenhouse ranged between 15.0 ◦C (night) and 27.3 ◦C (day) and the relative humidity ranged between 67% and 72% during the growing period. The photosynthetic active radiation was approximately 1000 μmol m−<sup>2</sup> s−<sup>1</sup> at noon. Daily watering by drip irrigation was applied to reach the full pot substrate FC. Pots were irrigated equally for 30 days after transplantation (DAT). Mint plants were harvested at 90 DAT. Container substrate capacity (CSC) is the maximum amount of water that can be retained by the substrate after the discharge because of gravity [15]. Before planting, the gravimetric method was used to determine CSC or FC by watering the plants to saturate the soil then the pots were left to drain for 60 min and the volume of drained water was quantified and the difference between the supplied and drained water volumes were considered the volumetric water

retained by the soil (i.e., *CSC*). The amount of water applied (*AWA*) to compensate for the soil water deficit to reach the FC is calculated as follows:

$$AWA = \left(\mathbb{C} \text{SC} - \theta\_v\right) DA \tag{1}$$

where θ*<sup>v</sup>* is soil water content at the irrigation event, *D* is the soil depth, and *A* is the surface area of the pot.

#### *2.2. Treatments*

The plants were subjected to three watering levels of CSC (100%, 70%, and 50%) after 30 DAT and/or single biostimulant of four commercial biostimulants, namely, CRADLE™, Mobilizer™, Nanozim NXT™, and Nanozim De'Lite™ (Biostadt, Mumbai, India). CRADLE (CR) powder is a mycorrhizal biofertilizer developed by InGene Organics, India and was used at g L−<sup>1</sup> growing soil. Mobilizer (Mob) is a granular mycorrhizal biofertilizer mixed with kelp seaweed extract (*Macrocystis pyrifera*), humic acid, and amino acids and was applied at g L−<sup>1</sup> growing soil. Nanozim De'Lite (ND) is a granular formulation of 25% (*w w*<sup>−</sup>1) seaweed (*Ascophyllum nodosum*), 25% (*w*/*w*) carbohydrates, 2% (*w w*−1) amino acid, and 1% (*w*/*w*) potassium (K2O), and was used at 1 g L−<sup>1</sup> with irrigation water. Nanozim NXT (NN) is a liquid mixture of 15% (*w w*−1) seaweed (*Ascophyllum nodosum*), 5% (*w*/*w*) humic acid, 1% (*w w*−1) potassium (K2O), 0.01% (*w w*−1) phosphorus (P2O5), 0.05% (*w*/*w*) alginic acid, 0.05% (*w w*<sup>−</sup>1) hydrolyzed protein, and several micronutrients and was applied at 1.5 mL L−<sup>1</sup> of irrigation water. The doses of the biostimulants and method of applications matched the manufacturer recommendations and untreated plants with biostimulants were considered the controls. Plants were grouped into three blocks containing 10 replicates per treatment [3 water levels (100%, 70%, and 50% CSC) × (4 biostimulants +1 control "without biostimulant") = 15 treatments] and totaling 450 plants (150 plants/block × 3 blocks) in a completely randomized design.

#### *2.3. Measurements*

#### 2.3.1. Morphological and Physiological

Following 9 weeks of treatments, several morphological measurements were determined including leaf number (plant−1), leaf area (cm2 plant−1), plant heights (cm), plant fresh weight (g), plant dry weight (g), and root dry weight (g). Irrigation water-use efficiency (IWUE, kg m−3) was calculated by dividing the fresh weight of the plant (kg) by the total *AWA* (m3) to each treatment during the growing period [16]. A digital area meter was used to determine the leaf area. The dry weights were determined following drying at 35 ◦C in an oven until reaching a constant weight.

Gas exchange measurements were performed on fully expanded leaves, under clear, sunny conditions using a portable photosynthesis system analyzer (ADC BioScientific, LCi, Bioscientific, Ltd., Hoddesdon, UK) and included photosynthetic rate (A), transpiration rate (E), and stomatal conductance (gs). Leaf midday water potential and midday relative water content were calculated at the end of the experiments at noon following the methods of Elansary et al. [17]. Leaf proline composition was also determined following the methods of Elansary et al. [18].

#### 2.3.2. Essential Oil and Gas Chromatography/Mass Spectrometry (GC/MS)

The EOs were obtained by hydro-distillation of dried leaves for 1 h in Clevenger type glass equipment in the Department of Plant Production, King Saud University. The EO ratio was determined per treatment and the EOs were maintained dry by subjecting samples to anhydrous sodium sulfate, then stored at 4 ◦C. A Thermo Scientific, Trace GC Ultra was used coupled with a mass spectrometer (ISQ). A TG-1MS column (narrow bore, length 30 m × 0.32 mm ID, 0.25 μm film thickness) was used and the carrier gas was helium. The machine was programed with a starting temperature of 45 ◦C, then a gradual increase was made to 165 ◦C (4 ◦C min<sup>−</sup>1), followed by an increase to 280 ◦C (15 ◦C min<sup>−</sup>1), and ending with holding time of 15 min. A 2 μL sample of each EO was injected at 250 ◦C on a splitless mode flow (1 mL min−1) for splitless time (3 min) followed by another split flow (10 mL min−1). The FID was also accomplished in the same column and program. A homologous series of *n*-alkanes (C10–C36) was used to identify the compounds by the retention time and indices were coupled with a mass spectral search program (NIST Ver. 2.0) and WILEY libraries. Selected references from the literature were also used for comparison purposes [1,13].

#### 2.3.3. Antioxidant Potential

Lipid peroxidation levels expressed as thiobarbituric acid reactive substances (TBARS), catalase (CAT) activity, H2O2, and superoxide dismutase (SOD) activities were quantified in frozen petal tissues [19].

#### *2.4. Statistical Analyses*

Data obtained from both years (2018 and 2019) were subjected to analyses of variance (ANOVA) with a completely randomized design to determine the significance of differences among treatments in SPSS Version 22 software. Standard errors (SE) were calculated from the data presenting for the mean of 20 replicates from both years. The least significant differences (LSD) method at *p* ≤ 0.05 was used to compare all means at each watering level [20].

#### **3. Results**

#### *3.1. Morphological Responses*

Plants subjected to different water levels of *CSC* and/or biostimulants showed different morphological responses as shown in Table 1. The plants were subjected to three watering levels of *CSC* (100%, 70%, and 50%) and/or single biostimulants (CR, Mob, ND, and NN). Under 100% *CSC*, NN-treated plants showed the highest leaf number and leaf area and was followed by ND, Mob, and CR. ND and NN treatments at 70% *CSC* and 50% CSC showed the highest increase in leaf number and leaf area as compared to those of other treatments, as well as those of the control. The tallest plants and highest root dry weight were found treatments with NN. CR and Mob showed no significant differences under 50% *CSC*. On the other hand, when water stress increased (i.e., 70% *CSC* and 50% *CSC*), the values for leaf number (20.19% and 45.77%), leaf area (24% and 52.40%), plant height (17.30% and 41.35%), and root dry weight (15.70% and 49.59%) decreased significantly. Plant fresh and dry weights increased significantly in biostimulant-treated plants compared to those of the control, and the highest increase was for plants subjected to NN followed by ND under different water levels (Table 2). NN-treated plants with 70% *CSC* water level had the highest IWUE (average of 2.78 kg m<sup>−</sup>3) in both growth seasons, which was statistically different from that of all other treatments.

#### *3.2. Physiological and Metabolic Performance*

Figure 1a–c shows the effects of water stress and biostimulants on gas exchange, namely, A, E, and gs of mint plants during the growing seasons in 2018 and 2019 (shown as averages). Under 100% *CSC*, the A of mint plants showed a (*p* < 0.05) significant increase in plants subjected to NN (8.35 μmol CO2 m−<sup>2</sup> s−1) compared to that of other treatments (Figure 1a). The reduction in irrigation water to 70% *CSC* and 50% *CSC* caused significantly reduced A values in control plants. However, biostimulant treatment increased the values of A, wherein NN-treated plants under 70% *CSC* and 50% *CSC* had the highest values (7.4 and 4.9 μmol CO2 m−<sup>2</sup> s<sup>−</sup>1, respectively), followed by ND-treated plants (7.3 and 4.6 μmol CO2 m−<sup>2</sup> s<sup>−</sup>1, respectively).



Data represent the mean calculated from *n* = 100, 60, or 20 for each water level, biostimulants or their interaction, respectively, in two growth seasons. \*\*\* (*<sup>p</sup>* ≤ 0.001). Means differing in lowercase letters (a–e) within each row indicate significant differences at *p* ≤ 0.05. Means differing in uppercase letters (A–E) within each mean row and each mean column indicate significant differences according to the least significant difference (LSD) test at *p* ≤ 0.05. CSC (container substrate capacity), Control (without biostimulants), CR (CRADLE™), Mob (Mobilizer™), ND (Nanozim De'Lite™), NN (Nanozim NXT™), and Biosti. (biostimulant).

**Table 2.** Average plant fresh and dry weights and irrigation water-use efficiency (IWUE) for mint plants grown under greenhouse conditions during the growth season in 2018 and 2019, subjected to three water levels, 100%, 70%, and 50% *CSC*, and four biostimulants, CR, Mob, ND, and NN, as well as a control (without biostimulants).


lowercase letters (a–e) within each row indicate significant differences at *p* ≤ 0.05. Means differing in uppercase letters (A–E) within each mean row and each mean column indicate significant differences according to the least significant difference (LSD) test at *p* ≤ 0.05. CSC (container substrate capacity), Control (without biostimulants), CR (CRADLE™), Mob (Mobilizer™), ND (Nanozim De'Lite™), NN (Nanozim NXT™), and Biosti. (biostimulant).

(**c**)

**Figure 1.** *Cont.*

(**f**)

**Figure 1.** Photosynthetic rate (**a**), transpiration rate (**b**), stomatal conductance (**c**), leaf midday water potential (**d**), leaf relative water content (**e**), and proline content (**f**) responses of mint plants, average of two growing seasons, as affected by three water levels, 100%, 70%, and 50% *CSC* (container substrate capacity), and four biostimulants, CR (CRADLE™), Mob (Mobilizer™), ND (Nanozim De'Lite™), and NN (Nanozim NXT™), and a Control (without biostimulants). FW (fresh weight). Different capital letters on top indicate significant differences between water levels at *p* ≤ 0.05. Different letters on top of columns indicate significant differences between biostimulants across water levels at *p* ≤ 0.05. Bars indicate the means ± SE of the mean (*n* = 20).

The values of E were higher in plants subjected to biostimulants compared to that of the control (Figure 1b). Treating plants with CR, Mob, and ND showed no significant differences in E under different *CSC* treatments. However, the E in NN-treated plants was significantly (*p* < 0.05) increased by 5.6%, 21%, and 17.6%, respectively, compared with that of the control plants under 100%, 70%, and 50% *CSC*.

The values of gs increased in plants treated with NN (203, 163, and 125 mmol m−<sup>2</sup> s<sup>−</sup>1, respectively) compared to that of other biostimulant treatments under 100%, 70%, and 50% *CSC* (Figure 1c). Statistically significant differences in gs under 50% *CSC* were found only between the ND treatment and both CR and Mob treatments, but significant effects on gs to the same treatments were not observed for 100% and 70% *CSC*. Irrespective of the biostimulant treatments, there were significant (*p* < 0.05) differences between the water levels treatments, where the 100% *CSC* was superior.

Leaf midday water potential increased in plants subjected to different biostimulants under 70% *CSC* (average of −0.95 MPa) and 50% *CSC* (average of −1.26 MPa) compared to that of 100% *CSC* (average of −0.67 MPa), as shown in Figure 1d. The NN treatment had the lowest water potential under different water level treatments. The leaf relative water content increased significantly (*p* < 0.05) in plants subjected to NN under different water levels (Figure 1e), where NN-treated plants with 100% *CSC* had the highest value of 88.3%. The proline content (Figure 1f) increased in plants subjected stress conditions only (70% and 50% *CSC*) and the increases were higher in plants subjected to biostimulants than that of the control (without biostimulants) treatment. The NN-treated plants with 50% CSC had the highest proline content value of 89.1 μg g−<sup>1</sup> fresh weight.

#### *3.3. EO Ratio and Constitutes*

The EO ratio was increased in response to biostimulant treatments as shown in Figure 2. Control treatments had the lowest EO ratio of 2.7%, 2.8%, and 2.1% fresh weight, respectively, under 100%, 70%, and 50% *CSC*. In NN-treated plants, the average EO ratio was increased by 21.57%, followed by that of ND (15.81%), in relation to that of the control plants under water level treatments. There was a significant difference in the EO ratio (*p* < 0.05) of mint plants between different biostimulant treatments within each water level, except between CR and Mob (*p* > 0.05) under water stress conditions of 70% and 50% *CSC*. Irrespective of the biostimulant treatments, the EO ratio under 70% *CSC* was not significantly higher than that of 100% *CSC*.

**Figure 2.** Essential oil ratio of mint plants, average of two growing seasons, as affected by three water levels: 100, 70, and 50% *CSC* (container substrate capacity), and four biostimulants: CR (CRADLE™), Mob (Mobilizer™), ND (Nanozim De'Lite™), and NN (Nanozim NXT™), in addition to Control (without biostimulants). Different capital letters on top are significant differences between water levels at *p* ≤ 0.05. Different letter on top columns indicate significant differences between biostimulants across water levels at *p* ≤ 0.05. Bars give the means ± SE of the mean (*n* = 20).

Major EO constitutes in all treatments were 1-menthone, isopulegone, pulegone, α-pinene, 1,8-cineol, and α-terpineol ratios as shown in Tables 3 and 4. The 1-menthone, isopulegone and pulegone were significantly (*p* < 0.001) reduced in plants subjected to water stress and biostimulant treatments (Table 3). Irrespective of the biostimulant treatments, 70% and 50% *CSC* plants exhibited decreased 1-menthone by 5.3% and 9.6%, respectively, compared to that of 100% *CSC* plants. Likewise, isopulegone for these plants was decreased by 5.3% and 8.6% and pulegone by 6.9% and 8.5%, respectively. Irrespective of the water level treatments, ND and NN treatments significantly reduced 1-menthone, isopulegone, and pulegone, by 23.8% and 33.3%; 8.1% and 18.6%; 11.1% and 18% on average, respectively, compared to that of the control treatment. However, the application of biostimulants significantly (*p* < 0.001) increased the α-pinene, 1,8-cineol, and α-terpineol ratios under different water stress conditions (Table 4). These ratios were significantly (*p* < 0.001) different between 100%, 70%, and 50% *CSC* plants, where 100% CSC showed the lowest values. The NN-treated plants yielded the highest ratios for α-pinene (4.3%), 1,8-cineol (36.1%), and α-terpineol (3.4%) at 50% *CSC*.

#### *3.4. Antioxidant Activities*

There was a significantly (*p* < 0.05) reduced accumulation of lipid peroxidation and H2O2 in plants subjected to different biostimulants, as shown in Figure 3. The NN-treated plants yielded the lowest values of lipid peroxidation (57, 46, and 27 μmol TBARS g−<sup>1</sup> fresh weight, respectively) and H2O2 (2.6, 4.6, and 5.9 μmol g−<sup>1</sup> fresh weight, respectively) under 100%, 70%, and 50% *CSC*. However, control plants had the highest values of lipid peroxidation (64, 54, and 37 μmol TBARS g−<sup>1</sup> fresh weight, respectively) and H2O2 (2.9, 5.2, and 7.2 μmol g−<sup>1</sup> fresh weight, respectively). It was observed that there were no significant differences between the control and CR treatments in lipid peroxidation and H2O2 under different water levels, except for the 70% *CSC* for lipid peroxidation. In 50% *CSC*, there were only significant differences between the Mob and ND treatments. On the contrary, there were significant (*p* < 0.05) increases in the activities of CAT and SOD of leaf extracts of biostimulant-treated plants under normal and water stress conditions (Figure 3). The highest CAT and SOD activity values were found in NN treatments (increasing 25.5% and 40.3%, respectively), followed by that of ND-treated plants (increasing 17.9% and 26.8%, respectively) comparing with those of the control treatment, which had the lowest values. The CAT activity was significantly (*p* < 0.05) increased by 30% and 58.2% on average, respectively, in plants subjected to water stress (70% and 50% *CSC*) compared to that of the normal (100% *CSC*) condition, whereas SOD activity was increased by 79.3% and 123.7% on average, respectively.



392


Data represent the mean calculated from *n* = 100, 60, or 20 for each water level, biostimulants or their interaction, respectively, in two growth seasons. \*\*\* (*<sup>p</sup>* ≤ 0.001). Means differing in lowercase letters (a–e) within each row indicate significant differences at *p* ≤ 0.05. Means differing in uppercase letters (A–E) within each mean row and each mean column indicate significant differences according to the least significant difference (LSD) test at *p* ≤ 0.05. CSC (container substrate capacity), Control (without biostimulants), CR (CRADLE™), Mob (Mobilizer™), ND (Nanozim De'Lite™), NN (Nanozim NXT™), and Biosti. (biostimulant).

**Figure 3.** *Cont.*

**Figure 3.** Lipid peroxidation, H2O2, catalase (CAT), and superoxide dismutase (SOD) activities of mint plants, average of two growing seasons, as affected by three water levels, 100%, 70%, and 50% *CSC* (container substrate capacity), and four biostimulants, CR (CRADLE™), Mob (Mobilizer™), ND (Nanozim De'Lite™), and NN (Nanozim NXT™), and a control (without biostimulants). FW (fresh weight). Different capital letters on top indicate significant differences between water levels at *p* ≤ 0.05. Different letters on top columns indicate significant differences between biostimulants across water levels at *p* ≤ 0.05. Bars provide the means ± SE of the mean (*n* = 20).

#### **4. Discussion**

Water stress is one of the major limiting factors of the growth and productivity of plants worldwide [21]. The amount of irrigation water applied influenced the biomass and EO yields of mint. The fresh and dry weights of mint were decreased with the irrigation water stress because of vegetative growth (i.e., leaf number and plant height), which decreased under water deficit conditions. Reduction in growth parameters as a consequence of drought has also been described in peppermint [6,7,11], Japanese mint [9], purple basil [8], balm [10], and chamomile [12]. The irrigation water level of 50% *CSC* had a negative effect on EO yield of mint. This is in agreement with earlier findings in peppermint [11] and chamomile [12], and in contrary to the results found in the previous studies from Ekren et al. [8] in purple basil and Farahani et al. [10] in balm.

The application of biostimulants under water stress conditions (70% and 50% *CSC*) showed enhanced growth by means of increased leaf number, plant height, root dry weight, fresh and dry weights, and IWUE. These morphological improvements are mainly attributed to the composition of these biostimulants. The most active biostimulant in this study was NN, which is composed of a unique

mixture of important compounds: seaweed extract (15%), humic acid (5%), macro (potassium 1% and phosphorus) and micro elements, alginic acid, and hydrolyzed protein, as described in the materials and methods. The major constitutes of the NN biostimulant were seaweed extracts (*Ascophyllum nodosum*) and humic acid. The application of *Ascophyllum nodosum* extracts as plant biostimulants has been reported in several studies [19,22]. Humic acid may increase the leaf area, stem diameter, plant dry weight in different plants [23,24] and may ameliorate stress conditions in tomatoes [25]. Potassium, phosphorus, and microelements play critical roles in the growth and morphology of most plants [26,27]. However, the mixture was superior in the ameliorating effects against water stress in mint plants compared to other commercial biostimulants. The second biostimulant showing relatively high morphological performance was ND, which is mainly composed of *Ascophyllum nodosum* extracts (25%), carbohydrates (25%), (*w*/*w*) amino acid (2%) and potassium (1%). ND showed slightly lower morphological promoting effects than that of NN. CR and Mob are mainly composed of mycorrhizal biofertilizer. However, Mob contains additional components, including seaweed extract (*Macrocystis pyrifera*), humic acid, and amino acids, which may explain the slight increased vegetative performance of Mob compared to that of CR. Furthermore, *Macrocystis pyrifera* has been reported to have stimulatory effects on plant growth [28].

Gas exchange parameters (*gs*, *E*, and *A*) are important indicators of the physiological performance of plants under stress conditions. The increase in *gs* under stress conditions in response to external factors is strongly related to enhanced gas exchange through the stomata [29]. The increased gas exchange is normally reflected as enhanced transpiration and photosynthesis rates in the leaves [30]. There were increases in the gas exchange in plants treated with different biostimulants under water stress conditions, which indicated that these biostimulants acted as effective stress ameliorants. Leaf water potential and relative water content reductions might be associated with stress conditions [31,32]. They increased in this study in plants subjected to different biostimulants, indicating enhanced metabolic performance of treated plants. Furthermore, the increased proline composition in biostimulant-treated plants reflected enhanced stress tolerance as found in previous studies using other external elicitors [22,33].

The main constitutes of the EO were pulegone and 1,8-cineol. A previous study on the same species from Egypt reported comparable composition of both compounds [1]. There were fluctuations in the main constitutes of EO, as well as specific compounds, including pulegone, isopulegone, 1-menthone, 1,8-cineol, α-pinene, and α-terpineol. Secondary metabolites, such as cineole are usually associated with terpenes [34] and this explains the parallel increase in 1,8-cineol, α-terpineol, and pinene. 1-menthone and isopulegone are metabolites of pulegone. The pulegone is not favored in the EO composition of mint plants because of its carcinogenic effects at high doses [1], whereas, cineol is a favored compound in EOs because of its medicinal applications and pharmaceutical potential [35,36]. The application of NN showed the highest increase in 1,8-cineol and related terpenes ratios and lowest compositions of pulegone compared to that of the control and other biostimulant treatments. This result suggests that NN application may have future applications in medicinal plants, such as mints. The use of the NN biostimulant is a novel approach for enhancing the chemical composition of the EOs of mint plants by reducing hazardous compounds and increasing useful ones as found in this study.

In this study, the application of seaweed extract-based biostimulants mixed with humic acid and/or macro elements represented a novel tool for the enhancement of the medicinal properties of major medicinal plants, such as mints. The achievement of enhanced antioxidant activities of the EOs of mint might be of great importance for agricultural and related pharmaceutical industries. The oil of mints is routinely used in perfume and cosmetic preparations, as well as in the food industries, such as in chocolate and soft drinks. The development of new EO compositions with increased antioxidant properties will increase the additive value of the medicinal crop and will assist in reducing dependence on synthetic antioxidants to control ROS accumulation.

#### **5. Conclusions**

This study revealed an association between the application of specific biostimulants and the increase/decrease of the main EO composition (cineol and pulegone) of mint plants. The application of this finding is related to the agricultural, medicinal, and pharmaceutical industries. There were increases in the morphological characteristics, physiological performance, and EO ratio of biostimulant-treated plants. The morphological and physiological enhancements indicated increased tolerance to water stress. Further, biostimulant-treated plants showed higher antioxidant activities, reduced accumulation of H2O2, and increased CAT and SOD activities, which indicated an antioxidant stress tolerance activation mechanism in treated plants. The application of biostimulants to mint plants increased the quantity and quality of produced EOs and enhanced the medicinal properties, as well as that of the traditional medicinal crop. ND and NN are recommended under water stress conditions in mint.

**Author Contributions:** H.O.E., conceptualization, funding acquisition, investigation, supervision, project administration, methodology, writing—original draft; M.A.M., formal analysis, investigation, data curation, methodology, project administration, writing—review and editing; E.A.M., supervision, visualization; D.O.E.-A., supervision, visualization. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was financially supported by the Deanship of Scientific Research at King Saud University through research group number RG-1440-022.

**Acknowledgments:** The authors extend their appreciation to the Deanship of Scientific Research at King Saud University for their financial support of the present study.

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

#### **References**


© 2019 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/).

*Review*

### **Arbuscular Mycorrhizal Fungi and Associated Microbiota as Plant Biostimulants: Research Strategies for the Selection of the Best Performing Inocula**

**Luca Giovannini 1,**†**, Michela Palla 1,**†**, Monica Agnolucci 1, Luciano Avio 1, Cristiana Sbrana 2, Alessandra Turrini <sup>1</sup> and Manuela Giovannetti 1,\***


Received: 18 December 2019; Accepted: 9 January 2020; Published: 11 January 2020

**Abstract:** Arbuscular mycorrhizal fungi (AMF) are beneficial soil microorganisms establishing mutualistic symbioses with the roots of the most important food crops and playing key roles in the maintenance of long-term soil fertility and health. The great inter- and intra-specific AMF diversity can be fully exploited by selecting AMF inocula on the basis of their colonization ability and efficiency, which are affected by fungal and plant genotypes and diverse environmental variables. The multiple services provided by AMF are the result of the synergistic activities of the bacterial communities living in the mycorrhizosphere, encompassing nitrogen fixation, P solubilization, and the production of phytohormones, siderophores, and antibiotics. The tripartite association among host plants, mycorrhizal symbionts, and associated bacteria show beneficial emerging properties which could be efficiently exploited in sustainable agriculture. Further in-depth studies, both in microcosms and in the field, performed on different AMF species and isolates, should evaluate their colonization ability, efficiency, and resilience. Transcriptomic studies can reveal the expression levels of nutrient transporter genes in fungal absorbing hyphae in the presence of selected bacterial strains. Eventually, newly designed multifunctional microbial consortia can be utilized as biofertilizers and biostimulants in sustainable and innovative production systems.

**Keywords:** arbuscular mycorrhizal symbiosis; mycorrhizosphere; AMF associated bacteria; plant growth-promoting bacteria; biofertilizers; phosphate-solubilizing bacteria; siderophore production

#### **1. Introduction**

In the next decades, the major challenge for agriculture will be the adoption of a new paradigm, sustainable intensification, to meet human needs for the production of enough food at a global scale while maintaining environmental quality and reducing the input of chemical fertilizers and pesticides [1]. These objectives may be pursued by giving more attention to beneficial soil microorganisms that play key roles in the maintenance of long-term soil fertility and health, the reduction of chemical inputs in agriculture, the promotion of plant nutrition, and the production of safe and high-quality food [2]. Among them, arbuscular mycorrhizal (AM) fungi (AMF) represent a key functional group, positively affecting plant growth, nutrition, and health. AMF are obligately biotrophic organisms that establish mutualistic symbioses with the roots of all major land plant taxa, including the most important food

crops such as cereals, pulses, fruit trees, vegetables, medicinal plants, and other economically relevant species such as sunflower, cotton, sugarcane, tobacco, coffee, tea, cocoa, rubber, and cassava [3]. Within food crops, the only exceptions are represented by genera and species belonging to Brassicaceae and Chenopodiaceae, which are non-mycorrhizal plants.

In exchange for plant photosynthates, AMF facilitate the uptake and transfer of mineral nutrients, such as phosphorus (P), nitrogen (N), sulfur, potassium, calcium, copper and zinc, from the soil to their host plants by means of the extraradical mycelium (ERM) extending from colonized roots into the soil [3]. Such a fungal structure represents one of the critical elements of the AM symbiosis, as the flow of nutrients translocated to the root cells of host plants is highly dependent on its structure, extent, and interconnectedness. ERM functions as an efficient absorbing system, given the high surface-to-volume ratio of the mycelium, which is able to uptake soil nutrients beyond the depletion zone around roots, and the presence of nutrient transporter genes in the hyphae [4]. Besides plant nutrition improvement, AMF facilitate the completion of biogeochemical cycles, increase plant tolerance to biotic and abiotic stresses, carbon sequestration and soil aggregation [5], and the content of health-promoting phytochemicals [6,7] (Figure 1).

**Figure 1.** Schematic drawing representing the impacts of arbuscular mycorrhizal fungi (AMF) and beneficial bacteria on plant performance and soil fertility. On the left: a visual representation of the AMF life cycle and factors affecting the different AMF developmental stages; on the right: mycorrhizal helper (MH) and plant growth promoting (PGP) bacteria synergistically interacting with AMF.

Several studies showed that the multiple services provided by AMF are the result of the synergistic activity of diverse bacterial communities living in the mycorrhizosphere, strictly associated with their spores and extraradical mycelium and playing diverse plant growth-promoting (PGP) roles, from nitrogen fixation and P solubilization and mineralization to the production of indole acetic acid (IAA), siderophores, and antibiotics [8,9]. Such microbiota was identified not only by culture-independent methods but also by culture-dependent approaches, which allowed their functional characterization, aimed at detecting the best performing bacterial strains, to be used in combination with selected AMF as biofertilizers and biostimulants in innovative and sustainable food production systems [10].

The aim of this review is to provide an overview of the recent developments which contributed to disclose the biostimulant properties of AMF and their associated bacteria and to propose the best research strategies for the selection of functional isolates and consortia to be utilized as high-quality inocula in sustainable agriculture.

#### **2. Arbuscular Mycorrhizal Fungi**

AMF belong to the phylum Glomeromycota, encompassing ten out of eleven families: Acaulosporaceae, Ambisporaceae, Archaesporacea, Claroidoglomeraceae, Diversisporaceae, Gigasporaceae, Glomeraceae, Pacisporaceae, Paraglomeraceae, and Sacculosporaceae (http://www.amfphylogeny.com/, accessed on 7 January 2020). Given their status of obligate biotrophs, the AMF life cycle cannot be completed in the absence of host plants. It starts with an asymbiotic phase, during which spores germinate in response to physical factors such as moisture, temperature and pH, producing hyphae with a limited lifespan [11]. In the presence of root exudates from host plants, a differential hyphal morphogenesis occurs, with germling hyphae reorienting the direction of elongation and initiating a differential branching pattern [12–14]: this pre-symbiotic phase is followed by physical contact between AMF hyphae and host roots, with the differentiation of appressoria, which give rise to hyphae growing intercellularly within the root cortex, eventually penetrating in root cells and producing highly branched hyphal tree-like structures similar to haustoria, the arbuscules. Arbuscules are the key structures of mycorrhizal symbioses, as at their level nutrient exchanges between the two partners take place: AMF obtain carbon (up to 20% of plant photosynthates) and lipids from the host plant and release mineral nutrients absorbed and translocated by ERM [15–17]. Two types of root colonization have been detected: *Arum*-type and *Paris*-type [3]. In the *Arum*-type, the AM symbiont spreads intercellularly between cortical root cells, forming terminal arbuscules on intracellular hyphal branches [18]. In the *Paris*-type, the fungus grows directly from cell to cell within the cortex and forms intracellular hyphal coils and intercalary arbuscules along the coils. However, most of the data available on AMF derive from studies carried out on the *Arum*-type mycorrhizal symbioses, which are widely distributed in natural and agricultural ecosystems. Beyond arbuscules, several AMF species produce intraradical vesicles, which are spore-like storage structures containing lipids. After receiving host carbon, the fungal symbiont is able to grow extraradically, colonize the surrounding soil, absorb mineral nutrients to be transferred to the host plant, interact with rhizosphere and soil microorganisms, colonize the roots of other plant living nearby (even belonging to species, genera and families different from their host), and also translocate mineral nutrients from one host to another [19,20]. The life cycle is closed by the formation of asexual spores by ERM, functional to the maintenance of a high mycorrhizal potential of the soil and, consequently, of soil biological fertility (Figure 1).

#### **3. AMF Functional Diversity: Colonization Ability and E**ffi**ciency**

So far, 323 AMF species have been described (http://www.amf-phylogeny.com/amphylo\_species. html, accessed 3 December 2019), though only a few species have been investigated for their functional diversity, in order to detect and select the best isolates to be used in agriculture. As a consequence, most of the available commercial inocula are prepared with *Rhizoglomus irregulare* (syn. *Rhizophagus irregularis*, formerly *Glomus intraradices*) and *Funneliformis mosseae* (formerly *Glomus mosseae*), that are generalist symbionts, widespread all over the world in almost all soils and climatic zones [3]. In order

to exploit the great inter- and intra-specific diversity, the general criteria to be applied when selecting the most efficient AMF isolates are outlined here.

The two fundamental fungal characteristics to be taken into account are colonization ability, which refers to fungal capacity of a rapid and extensive root colonization, and efficiency, represented by fungal symbiotic performance, in terms of plant growth and nutrition.

#### *3.1. Colonization Ability*

A high root colonization ability is the essential prerequisite for any AMF isolate to be designed for agricultural utilization, as it should be able to compete with highly competitive native AMF. AMF colonization ability does not depend only on fungal genotype, but also on soil characteristics and plant genotype, which may influence the different steps of mycorrhizal establishment, from spore germination to appressorium formation and intraradical growth.

The first variable affecting the competitive ability of an AMF strain is represented by spore dormancy, which may be relieved by storage at 5–10 ◦C for 5–6 weeks; nevertheless, it is extremely important to know which AMF isolates produce dormant spores when selecting strains for inoculation. As an example, different species of the genera *Glomus*, *Funneliformis*, and *Acaulospora* show spore dormancy, while species such as *Gigaspora gigantea* and *Gigaspora margarita* are able to germinate as early as one day after incubation [21]. It is unfortunate that only a few works have investigated this critical element, which should be further studied not only at the species but, most importantly, at the isolate level, as the producers of commercial inocula often reproduce their own strains.

A key fungal characteristic directly linked to AMF establishment and persistence in the field is represented by spore germination, which is affected by different factors such as soil pH and nutrient content, temperature, soil bacteria, and pesticides. Poor information is available on soil variables, suggesting that the different AMF strains show optimum germination when cultivated in environments with characteristics similar to those from which they were originally isolated. Thus, for example, *Acaulospora laevis*, predominant in low pH soils, germinates well at pH 4–5, while *Dentiscutata heterogama* (formerly *Gigaspora heterogama*), isolated from warm climates, germinates best at 34 ◦C [21], although nine AMF, isolated and maintained in tropical areas, showed very different germination rates, ranging from 8% to 78%, when cultured in the same environmental conditions [22]. It has long been known that spore germination can be stimulated by soil microorganisms, from Actinobacteria to Pseudomonads, although the most relevant role is played by bacteria living in intimate association with AMF, often located on and within spore wall layers (mycorrhizospheric bacteria) [9]. Actually, many bacterial taxa able to degrade biopolymers were recently detected in spore homogenates by culture-independent methods, suggesting a possible chitinolytic activity on chitin of spore walls that could enhance spore germination [23,24]. It is interesting to note that a recent molecular work reported the ability of six AMF isolates to recruit different bacterial communities on their spores, belonging to Actinomycetales, Bacillales, Burkholderiales, Pseudomonadales, and Rhizobiales, possibly exerting an activity on spore germination [25]. As to pesticides, their effects on spore germination are different depending on the target organisms. Several fungicides, like copper hydroxide and mancozeb, were able to inhibit spore germination of *F. mosseae* in vivo, while flutolanil, azoxystrobin, fenpropimorph, and fenhexamid inhibited germination of *R. irregulare* spores in vitro [26,27]. On the other hand, other fungicides, such as fosetil Al, metalaxyl and different herbicides, seem to exert no activity on spore germination even if the results obtained on the same substance in different investigations were often contradictory [28].

After germination, another important variable affecting the competitive ability of AMF towards native fungi is represented by the ability of germlings to produce an extensive and interconnected hyphal network, which is essential for increasing the chance of coming into contact with a host root. Germling growth may be affected by the same environmental variables quoted above, but depends largely on fungal genotype as it can range from 0.25 up to 104 and 544 mm of hyphal length per germling in the same experimental in vitro conditions [11]. It is important to underline that the possibility to contact host roots and to establish the symbiosis is greatly extended by the ability of germling

hyphae to become interconnected through hyphal fusions (anastomoses): this capacity represents a fundamental survival strategy for AMF germlings, which can plug into compatible extraradical networks, gaining immediate access to plant-derived carbon [29]. Anastomosis formation is highly related to the fungal genotype, as species belonging to the families of Glomeraceae and Acaulosporaceae show a high frequency of hyphal fusions, while members of the family Gigasporaceae do not form fusions interconnecting different hyphae [30,31]. The length, viability, and interconnectedness of germling hyphae are affected by various pesticides: for example, fungicides containing benomyl and fenhexamid, even at doses below the recommended field rate, inhibited hyphal growth of *F. mosseae*, affected mycelial viability, and induced abnormal hyphal branching, while the herbicide glufosinate ammonium decreased mycelial growth and viability, and also the anastomosis rate [32,33].

When AMF germlings come into contact with a host root, a differential hyphal morphogenesis is induced, characterized by an increase in hyphal branching, functional to the production of appressoria on the root surface [12,34]. Appressoria are swollen, multinucleate structures formed as early as 36 h after the contact between germlings and roots [35], and represent the signs of fungal recognition of the host plant. A prompt production of a large number of appressoria, which is requisite for a rapid root colonization, characterizes the most infective AMF, as it makes them highly competitive with native symbionts. Several works investigated this AMF functional trait: an old, but not obsolete work, reported that the number of appressoria may range from 2.6 to 21.1 and from 4.6 to 10.7 per mm of root length in field-grown strawberry and apple, respectively [36], while more recent works found 10.2–80.5 appressoria per plant in parsley and aubergine inoculated in microcosms with *F. mosseae* [37,38]. The same fungus showed variable results depending on host plants: for example, it produced 3.6 appressoria per mm of root in *Medicago truncatula*, 9.7 in *Prunus cerasifera*, and 1.26 in *Trifolium pratense* [39–41]. On the other hand, *G. margarita* produced only 0.01 appressoria per mm of root when inoculated on *Allium cepa* [42]. The dynamics of appressoria formation was monitored in a time-course experiment, showing that the first structures were produced after 36, 48, and 60 h, depending on the fungal genotype [35].

Appressoria produce intraradical hyphae able to establish the mycorrhizal symbiosis by rapidly spreading in the apoplastic space between root cortical cells, although the levels of root colonization greatly vary among AMF and plant genotypes. While such variability among different AMF species have been assessed in countless experiments aimed at evaluating fungal performance in terms of plant growth, the susceptibility of different plant genotypes to mycorrhizal colonization has been investigated only in recent works, reporting large differences among 11 sunflower cultivars (range 8.6–78.7%) and 108 durum wheat varieties (range 10–44%) [43,44].

#### *3.2. E*ffi*ciency*

The efficiency of the different AM fungal isolates is generally interpreted as their ability to increase plant growth and nutrient uptake, and evaluated by considering the relevant fungal variables such as ERM development, extent, interconnectedness, viability, and rate of nutrient uptake and translocation, that are directly linked to the occurrence of fungal transporter genes in the absorptive extraradical hyphae [4].

ERM length density, assessed after destructive extraction from the soil, showed a large variability among AMF species, ranging from 1.1–6.9 to 3–5 and 10 m/g soil in *Acaulospora laevis*, *F. caledonius* (formerly *Glomus caledonium*), and *Scutellospora calospora*, respectively [11]. Recent works have reported higher hyphal lengths (up to 22 m/g soil) produced by *R. irregulare* isolate BEG 87 [45]. It is worth mentioning the ERM growth rate, which was 738–1067 and 3.1–3.8 mm/day in bidimensional and tri-dimensional experimental systems, respectively [20,46].

ERM structure and interconnectedness have been investigated by nondestructive tests, which provided both qualitative [47,48] and quantitative data. For example, ERM produced by members of the family Glomeraceae, widely distributed in agricultural soils, is highly interconnected by means of anastomoses between contacting hyphae (67–77% in *F. mosseae*), reaching the value of

100–410 anastomoses per gram of soil [20,31,49]. On the contrary, hyphae of members of the families Gigasporaceae, Ambisporaceae, and Paraglomeraceae are not able to fuse after contact, in vivo [50]. Nevertheless, within the Glomeraceae family, self-incompatible interactions between contacting hyphae may occur, with frequencies ranging from 5% to 32% [29,50,51]. Further extensive studies addressed such a clue, revealing major differences among three glomalean AMF: in particular, when grown in symbiosis with five different plant species, *F. mosseae* and *R. irregulare* ERM showed anastomosis frequency of 26–48% and 36–54%, respectively, while *F. coronatus* never exceeded 7.7% [52]; length and density affect AMF symbiotic performance, positively correlating with plant growth responses and nutrient levels [53]. Specifically, AMF isolates showing a high anastomosing ability are able to tolerate soil disturbance, such as tillage, by producing large mycorrhizal networks capable of re-establish interconnections after disruption [54–57]. ERM length and structure may be affected by pesticides, as reported by a recent work performed using a whole-plant experimental system, i.e., in *F. mosseae*, ERM length and density decreased in the presence of the herbicides dicamba and glufosinolate and the fungicides benomyl and fenhexamid, while ERM length and density increased in the presence of two mycorrhizospheric bacteria, *Ensifer meliloti* (formerly *Sinorhizobium meliloti*) and *Enterobacter ludwigii* [58]. Such recent novel data stress the need for further studies to evaluate the impact of agrochemicals and biocontrol agents on ERM structure and activity in a large number of AMF taxa in order to detect the most resilient isolates able to maintain a high mycorrhizal inoculum potential in soil.

Beyond the mentioned phenotypic parameters, viability, which is the most important factor affecting ERM functionality in soil, has been poorly investigated. A few studies reported that metabolic activity occurred in 63–96%, 96–100%, and 100% of extraradical hyphae in *R. irregulare*, *F. mosseae,* and *Rhizoglomus clarum* (formerly *Glomus clarum*), respectively [46,59,60]. A recent study posed the interesting question of whether ERM could survive and maintain colonization ability after plant harvest, thus representing a source of inoculum for the successive crops. The authors, utilizing an in vivo whole-plant experimental system and two worldwide distributed glomalean AMF, *F. mosseae* and *R. irregulare*, revealed that ERM viability and functionality are uncoupled from the host plant lifespan, as, after shoot removal, its growth from detached roots was comparable with that from intact plants and continuous for at least 150 days [61]. Accordingly, ERM represents a long-term survival structure able to maintain mycorrhizal potential and biological fertility in agricultural soils.

AMF efficiency is highly correlated with the rate of P translocation to the host plant: alas, only scanty information is available, showing that in *F. mosseae*, P fluxes in hyphae were 3.4 <sup>×</sup> <sup>10</sup>−<sup>8</sup> mol cm−<sup>2</sup> <sup>s</sup>−<sup>1</sup> [62]. However, as the transfer of nutrients flowing in the extraradical hyphae can occur exclusively through appressoria, which are the unique structures connecting soil-based to root-based mycelium, a high number of appressoria produced on the root surface is a key factor affecting not only AMF colonization ability but also their efficiency.

Studies on the occurrence of nutrient transporter genes in AMF extraradical hyphae have mostly been performed in vitro, using transgenic root organ cultures and few species, i.e., *R. irregulare* and *R. intraradices*. The results showed that a number of nutrient transporter genes (ammonium, phosphorus, zinc) are differentially regulated, depending on the availability of various mineral or organic compounds [4,63]. However, as transformed roots show an altered hormonal balance and sugar acquisition, possibly affecting the physiology of the mycorrhizal symbiosis, diverse whole-plant experimental systems were devised, encompassing other AMF species, *F. mosseae*, *F. coronatus*, and *G. margarita* [64]. Further extensive investigations focusing on nutrient transporters gene expression in extraradical mycorrhizal mycelium produced by a large number of AMF isolates are needed in order to achieve a deeper knowledge of differences in AMF efficiency and to select the best performing symbionts to be used as inocula, if also meeting the other quality characteristics concerning colonization ability and efficiency.

#### **4. AMF E**ffi**ciency in the Enhancement of Plant Health-Promoting Compounds**

In the light of the new findings on plant secondary metabolism being modulated by AMF, the concept of efficiency should be expanded to take into consideration the production of healthpromoting compounds, a theme of the highest concern not only to scientists but also to consumers and producers as phytochemicals may reduce oxidative damages, prevent chronic and heart diseases, and decrease the risk of mortality from cancer [65–67]. The levels of such compounds, mainly represented by carotenoids, glucosinolates, polyphenols, including flavonoids, isoflavones and anthocyanins, are affected by different variables such as plant genotype, agronomic techniques, soil characteristics, and also by mycorrhizal symbioses [6].

For example, sweet basil (*Ocimum basilicum*) inoculated with *Glomus* spp. increased the production of rosmarinic and caffeic acids, and of essential oils [68,69], while *R. intraradices* affected the gene expression of key enzymes involved in basil rosmarinic acid biosynthetic pathway [70]. *Echinacea purpurea* inoculated with *R. irregulare* and *G. margarita* showed higher concentrations of caffeic acid derivatives, alkylamides, and terpenes [71], while *R. irregulare* inoculated on *Stevia rebaudiana* enhanced its content of the health-promoting compound steviol glycoside [72]. Interestingly, diverse AMF isolates differentially affected the production of specific phytochemicals; for example, the levels of thymol derivatives in the roots of *Inula ensifolia* were more enhanced by *R. clarus* than by *R. irregulare* [73], while in basil leaves the production of camphor and alfa-terpineol were enhanced by *Gigaspora rosea* but not by *G. margarita,* which decreased the total content of essential oils, in particular that of eucalyptol, linalool, and eugenol [68].

Despite the good results obtained by utilizing medicinal plants and herbs, only a few food crops have been investigated for their levels of health-promoting compounds upon mycorrhizal inoculation, i.e., lettuce, onion, tomato, maize, artichoke, strawberry, pepper, and sweet potato [7]. Most experimental works utilized either AMF inocula composed of a mixture of species, obtained from commercial producers or single species inocula, often represented by *R. irregulare* or *F. mosseae*. Also, molecular studies focused on the assessment of the levels of transcripts encoding the enzymes of the pathways relevant to the production of health-promoting secondary metabolites mainly utilized the same two species [7]. This has impaired the evaluation of the efficiency of different AMF, aimed at selecting the best performing symbionts in the production of beneficial phytochemicals. Accordingly, in the years to come, in-depth investigations should fully exploit the wide physiological and genetic diversity of AMF, testing the highest possible range of diverse species, isolates, and lineages within isolates. In addition, transcriptomic studies would allow the identification of AMF strains differentially expressing genes relevant to the biosynthesis of nutraceutical compounds in food plants.

#### **5. Mycorrhizospheric Bacteria and Their Functional Significance**

It has long been known that AMF colonization ability and efficiency may be mediated by a third partner of the symbiosis, the diverse and abundant bacterial communities living in the mycorrhizosphere, i.e., associated with mycorrhizal roots, spores, sporocarps, and extraradical hyphae [74]. Later, by ultrastructural studies, bacteria were detected in spore wall layers, within the peridial hyphae surrounding spores [75,76], and inside the cytoplasm [77–80]. Culture-dependent approaches allowed the isolation of many different bacterial taxa from the mycorrhizosphere of *Glomus versiforme*, *R. clarus*, *G. margarita*, *F. mosseae*, and *R. irregulare* [81–84]. A recent work isolated from *Rhizoglomus irregulare* (formerly *R. intraradices*) spores as many as 374 bacterial strains [85]. Culture-independent methods provided an in-depth description of the different bacterial taxa associated with spores: for example, PCR denaturing gradient gel electrophoresis (PCR-DGGE) identified the bacterial communities associated with *F. geosporus*, *Septoglomus constrictum*, and *G. margarita* spores [23,24], and those strongly associated with the spores of six AMF isolates, three belonging to *F. mosseae*, one to *F. coronatus*, and two to *R. irregulare*—the 48 relevant sequences were affiliated with Actinomycetales, Bacillales, Burkholderiales, Pseudomonadales, Rhizobiales, and Mollicutes-related endobacteria [25].

The mycorrhizospheric microbiota showed different functional activities, ranging from the role of "mycorrhiza helper" (MH) [86] to that of "plant growth promoters" (PGP) (Figure 1). MH bacteria may increase spore germination and mycorrhizal symbiosis establishment: for example, *Streptomyces* spp., *Pseudomonas* sp., and *Corynebacterium* sp. improved the germination of *F. mosseae*, *G. versiforme*, and *G. margarita* spores [81,87–89]. The enhancement of spore germination was ascribed to Actinobacteria, a group of bacteria frequently associated with AMF spores, able to hydrolyze chitin, the main component of spore walls [23,25,76,90]. Other MH bacteria, such as *Klebsiella pneumoniae*, *Trichoderma* sp., and *Paenibacillus validus*, increased germlings hyphal growth [91–93], while one bacterial strain belonging to Oxalobacteriaceae enhanced not only spore germination and germling growth but also root colonization [94]. In addition, the development of AMF extraradical mycelium (ERM) may be promoted by strains of *Paenibacillus rhizosphaerae*, *Azospirillum* sp., *Rhizobium etli*, *Pseudomonas* spp., *Burkolderia cepacia*, and *E. meliloti* [45,95–98] (Figure 1).

PGP bacteria show multifunctional activities, encompassing nitrogen fixation, P solubilization and mineralization, the production of indole acetic acid (IAA), siderophores, and antibiotics while supplying fundamental nutrients and growth factors [8,9]. Such activities represent key characteristics to be taken into account when selecting the best AMF and bacterial combinations for the production of inocula for agricultural use. For example, as P is rapidly immobilized in the soil, forming insoluble compounds with aluminium/iron and with calcium in acid and alkaline soil and thus becoming unavailable to plants, P-solubilizing bacteria may work in synergy with AMF to increase P availability and plant P uptake. Indeed, P-mobilizing bacteria, such as *Streptomyces* spp., *Leifsonia* sp., *Bacillus pumilus*, *Lisinobacillus fusiformis*, and *E. meliloti*, isolated from AMF spores of *R. irregulare*, showed synergistic action with AMF, promoting the mineralization of soil phytate and facilitating P uptake by mycorrhizal plants [45,99]. Similarly, the isolation from the mycorrhizosphere of bacterial strains possessing the *nifH* gene amplicon suggested a possible role in plant acquisition of nitrogen [85]. On the other hand, some PGP bacteria are able to produce IAA, a phytohormone of the auxin class, which plays a key role in the regulation of plant growth, increasing plant cell division and root formation, thus affecting water and nutrient uptake [100–102]. Accordingly, IAA producing bacteria isolated from *R. irregulare* and *F. mosseae*, such as *E. meliloti* and *Paenibacillus favisporus*, enhanced the growth of AMF extraradical hyphae, the fungal structure fundamental for absorbing and translocating P from the soil to plant roots [45,95]. An important role in the promotion of plant growth is played by mycorrhizospheric bacteria able to protect plants against soil-borne pathogens, either by directly producing antibiotics or indirectly producing siderophores, high-affinity iron-chelating compounds which mediate iron acquisition by pathogenic microorganisms [85,103–105]. Moreover, the facilitation of plant iron acquisition by siderophores-producing bacteria represents an additional benefit, as iron is an essential element in key biochemical processes like photosynthesis and respiration [106,107]. Interestingly, many of the bacteria isolated from AMF spores showed multiple PGP activities, i.e., 17 actinobacterial strains were able to produce siderophores and IAA to mineralize phytate and solubilize inorganic phosphate, and ten putative N-fixers to produce siderophore and solubilize P [85]. A recent work confirmed such data, reporting the occurrence of diverse bacterial functional taxa in a commercial AMF inoculum: 14 isolates showed the best combination of PGP traits, such as the production of IAA and siderophores, while 6 of them were also able to solubilize P, i.e., *Bacillus megaterium*, *Streptomyces* sp., and *Enterobacter* spp. [108]. These strains, both as single- and multi-strain inocula, deserve further in-depth studies in order to evaluate their efficiency as biofertilizers and biostimulants, able to boost plant growth, nutrition and health in sustainable food production systems (Figure 1). New remarkable findings showed that several members of the mycorrhizospheric microbiota may establish a more intimate relationship with their host plants as root endophytes [109,110]: considering that they can reach 105–107 CFU per g of root [111,112], their possible beneficial effects should be further investigated in the years to come.

#### **6. Conclusions and Perspectives for Future Studies**

The multiple beneficial activities of AMF and their associated bacteria discussed so far highlight the complex networks of interactions taking place in the mycorrhizosphere, functional to plant growth, nutrition, and health. The tripartite association among host plants, fungal symbionts and their associated bacteria shows beneficial emerging properties that could be efficiently exploited in sustainable food production. Although much is known on a very small number of AMF species, often studied singly in sterile conditions, very little is known about the high physiological and genetic interand intra-specific diversity of AMF and their associated microbiota. Further in-depth studies should be performed on different AMF species and isolates, and on their associated bacteria, both singly and in various combinations, in order to evaluate their colonization ability and efficiency when inoculated with a number of plant hosts. The studies carried out in microcosms should be followed by investigations in the field to assess the ability of the selected AMF and bacteria to compete with native microorganisms and to maintain their beneficial activities. Once detected as the best performing inocula, they could be differentiated by assessing their resilience against diverse environmental conditions, from soil types to drought, salt, biotic stresses, and pesticides. Transcriptomic studies could reveal the expression levels of nutrient transporter genes in fungal absorbing hyphae in the presence of selected efficient bacterial strains, possibly leading to the detection of the best synergistic combinations of AMF and associated bacterial communities, enhancing nutrient availability and plant performance. At the same time, transcriptomics could increase knowledge on the differential expression of genes encoding enzymes relevant to the biosynthesis of nutraceutical compounds in food plants. Eventually, newly designed multifunctional microbial consortia could be commercially reproduced and utilized as biofertilizers and biostimulants in sustainable and innovative production systems.

**Author Contributions:** M.A., A.T. and M.G. conceived the topic of the paper and wrote the original draft. L.G., M.P., L.A. and C.S. participated in the preparation and review of the manuscript. L.G. and M.P. provided editing assistance. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was funded by UNIVERSITY OF PISA, ITALY, grant: Fondi di Ateneo.

**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*

### **The Influence of Bio-Stimulants and Foliar Fertilizers on Yield, Plant Features, and the Level of Soil Biochemical Activity in White Lupine (***Lupinus albus* **L.) Cultivation**

**Alicja Niewiadomska 1, Hanna Sulewska 2, Agnieszka Wolna-Maruwka 1, Karolina Ratajczak 2,\*, Zyta Waraczewska <sup>1</sup> and Anna Budka <sup>3</sup>**


Received: 17 December 2019; Accepted: 17 January 2020; Published: 20 January 2020

**Abstract:** The aim of this study is to assess the effect of two biostimulators (Titanit, Rooter) and six foliar fertilizers (Optysil, Metalosate Potassium, Bolero Bo, ADOB 2.0 Zn IDHA, ADOB B, ADOB 2.0 Mo) on white lupine. In addition, we evaluated the enzymatic activity of dehydrogenase, acid, and alkaline phosphatases, catalase, the level of biological nitrogen fixation, yield, plant biometric, chlorophyll fluorescence and chlorophyll content. A field experiment was conducted between 2016 and 2018 at the Gorzy ´n Experimental and Educational Station, Pozna ´n University of Life Sciences in Poland. The best effects in plant yield were obtained after the application of Optysil or ADOB Zn IDHA. The three years results of dehydrogenase (DHA), alkaline phosphatase (PAL), and the biological index of soil fertility (BIF), show that the bio-stimulants and most of the foliar fertilizers used did not always stimulate the activity of these enzymes and index in the white lupine crops, as compared with the control treatment. Analysis of the results of the acid phosphatase activity (PAC) shows that during the entire white lupine growing season the foliar fertilizers and bio-stimulants decreased the activity of this enzyme. This effect was not observed when the Metalosate potassium foliar fertilizer was applied. The field analyses of biological nitrogen fixation showed that the fertilizers and bio-stimulants significantly stimulated nitrogenase activity under the white lupine plantation. The best effects in plant yield were obtained after application Optysil or ADOB Zn IDHA.

**Keywords:** soil enzymatic activity; biological index fertility; nitrogenase activity; microelements fertilization (Ti, Si, B, Mo, Zn)

#### **1. Introduction**

The degradation of the soil environment, excessive use of chemicals, depletion of natural resources, as well as the decreasing biodiversity instigated the European Union to make a decision about the need for integrated crop cultivation and protection [1]. Since 2014 the recommendations concerning integrated protection and cultivation have been in force in Poland. At present we can see the transitional phase between conventional and sustainable agriculture. In order to meet the assumptions of sustainable agriculture it is necessary to diversify the crop structure and minimize the excessive

share of cereals. It is also necessary to use integrated methods of agricultural production, so it might be particularly important to restore legume plantations [2].

The significance of legumes in sustainable agriculture is increasing because they improve the physicochemical properties of soil, increase the content of organic matter by leaving large quantities of crop residues, and reduce the need to apply nitrogen fertilizers. White lupine (*Lupinus albus* L.) is one of the most important crops in this group of plants in Poland. It has been the longest known crop species of the *Lupinus* genus. Because of its very high content of protein and fat, especially in seeds, it has been used for human nutrition for thousands of years, despite its high content of bitter alkaloids [3]. It was only in 1930 that low-alkaloid forms were obtained. Because of the introduction of new varieties, the cultivation of white lupine with low alkaloid content became popular in Poland. Between 2005 and 2015 the area of cultivation of large-seeded legumes increased almost four times so that in 2015 they covered an area of 407,000 ha [4].

Lupine species have the largest share in this group of crops. On the other hand, the area of plantations with small-seeded legumes, such as clover and alfalfa, did not fluctuate much in that decade and in 2015 they covered an area of 93,000 ha [5].

Legumes are characterized by the ability to coexist with the nitrogen-fixing diazotrophic bacteria (*Rhizobium*). In order to increase the protein content in plants, which depends on the system developed by the plant and rhizobia, it is necessary to find agents improving the efficiency of this symbiosis.

Scientists are more and more interested in bio-stimulants, which are defined as materials containing one or more active substances and/or microorganisms. They improve the uptake of nutrients by plants, their tolerance to abiotic and biotic stress, and the quality of crops [6]. Bio-stimulants also increase the activity of rhizosphere microorganisms and soil enzymes, as well as they stimulate hormone production and photosynthesis [7]. They also promote the overall plant growth, including increased biomass and crop yields [8]. In the group of synthetic bio-stimulants, there are preparations containing growth regulators, phenolic compounds, inorganic salts, and beneficial nutrients [9,10], which naturally occur in plants in trace amounts (e.g., titanium and silicon). They act mainly by the stimulation of numerous physiological processes, which has a positive effect on plant yield and crop quality. Nutrients assimilable by plants, reduces the impact of stress, which affects the growth and development of plants. They regulate the uptake of macro- and microelements, alleviate the negative effects of periodic water shortage, high salinity, as well as activates the natural immune mechanisms of plants. They also strengthen cell walls and reduce the susceptibility of plants to mechanical damage [11]. Microelements regulate biochemical processes occurring in plants, being part of most enzymes or acting as their activators, therefore their deficit may lead to the inhibition of specific enzymatic reactions, which in turn leads to disorders of many biochemical and physiological processes, adversely affecting the growth and plant development [12,13]. There are many fertilizers that are enriched with amino acids, organic compounds, or surfactants. For example, potassium in fertilizer is in the form of very small molecules complexed with a unique set of natural amino acids. In turn, boron in the fertilizer is in the form of sodium pentaborate decahydrate, and the addition of sorbitol ensures rapid uptake of the fertilizer through the leaves of fertilized plants and high efficiency of the fertilizer. Zinc in modern fertilizers is chelated with the biodegradable IDHA chelating agent, because of which it also gains a form that is very well absorbed by plants. This fertilizer increases the plants' resistance to drought and diseases and increases the germination of seeds. It is produced in the form of microgranules, based on modern microgranulation technology. The manufacturer of molybdenum fertilizer has developed a liquid formula of the fertilizer additionally enriched with biodegradable tensides, which decreases the surface tension of the working liquid and increases the efficiency of covering the leaf blade during spraying increases [14].

Essential plant nutrients are mainly applied to soil and plant foliage in order to achieve maximum economic yields. Soil application is more common and most effective for nutrients that are required in high quantities. However, under certain circumstances, foliar fertilization is more economic and effective. Because of the intensified cultivation foliar fertilization has become an indispensable agrotechnical procedure. Plants exhibit the highest demand for potassium and nitrogen (more than 200 kg in terms of the yield per 1 ha), and the lowest demand for zinc, boron, copper, and molybdenum. Plants need only a few grams of molybdenum in terms of the yield per hectare. This means that foliar fertilization is particularly recommended and effective when it is necessary to supply micronutrients to crops [15].

Each agrotechnical treatment, i.e., the use of fertilizers or bio-stimulants, may cause changes in the soil environment. There have been numerous studies showing various effects of these treatments on the count of selected groups of microorganisms and the amount of soil enzymes they secrete [16].

Measurement of the activity of soil enzymes provides information about the quality of soil. This procedure is important as it indicates the biochemical activity of soil. Enzymes are thought to be good and sensitive indicators because they quickly react to changes in soil caused by natural and anthropogenic factors. Apart from that, it is easy to measure their activity, which affects the main microbiological reactions involving the cycles of nutrients in soil. Studies also showed that agrotechnical procedures influence the enzymatic activity more than other biochemical parameters [17].

The aim of this study is to assess the effect of selected bio-stimulants (Tytanit, Rooter) and foliar fertilizers (Optysil, Metalosate potassium, Bolero Bo, ADOB 2.0 Zn IDHA, ADOB B, ADOB 2.0 Mo) on the yield and plant features, activity of dehydrogenase, acid and alkaline phosphatases, and catalase, as well as the level of biological nitrogen fixation based on the activity of nitrogenase in a white lupine plantation.

#### **2. Material and Methods**

#### *2.1. Experimental Design*

A field experiment was conducted between 2016 and 2018 at the Gorzy ´n Experimental and Educational Station, Pozna ´n University of Life Sciences. The GPS coordinates of the experiment are as follows: N-52.56589, E-015.90556, 65 m AMSL. Each year one-factor experiment was conducted as randomized block design in four replications with the following nine factor levels: 1. control treatment—plants not treated with preparations; 2. Tytanit; 3. Optysil; 4. Metalosate Potassium; 5. Rooter; 6. Bolero Mo; 7. ADOB Zn IDHA; 8. ADOB B; 9. ADOB 2.0 Mo. Each fertilizer was applied in a timely manner, according to the manufacturer's recommendations (Table 1).

An experiment was conducted on white lupine (*Lupinus albus* L.) of the Butan cultivar. The lupine seeds were inoculated with the effective strain of *Bradyrhizobium lupinus* root nodule bacteria directly before sowing by using nitragina. Nitragina is a single-component graft, containing a specific bacterial strain for a specific legume plant, in which peat is a carrier. The Butan cultivar can be grown all over Poland, this variety is insensitive to delayed sowing; its growing period is 2–14 days shorter than that of traditional varieties and it is less susceptible to diseases caused by *Fusarium* fungi. The cultivar is more valuable as a feed and it has high content of protein (32–37%) and fat (10–12%), while the content of alkaloids is about 30–40% lower.

The seeds were sown (4 April 2016, 4 April 2017 and 7 April 2018) on plots with an area of 21 m2, with a distance of rows of 15 cm, and sowing density of 75 seeds per 1 m2.

According to the FAO/WRB classification [18], the soil in the experimental plots is a typical lessive soil formed from light loamy sands, deposited in a shallow layer on light loam (*Haplic Luvisols*) (Table 2). The soil texture was determined by means of a sieve (sand fraction) for the silt and clay fraction [19].

The agrotechnical and cultivation treatments were carried out in accordance with the principles of good agricultural and experimental practice for this species [20]. In the autumn before winter plowing, basic macronutrients were supplied to the soil in the form of multi-component fertilizer Polifoska 4 in the amount of 350 kg ha−<sup>1</sup> (N—4%, P—12%, K—32%). Before sowing, urea in the amount of 30 kg ha−<sup>1</sup> was used.


**Table 1.** The terms and doses of bio-stimulants and fertilizers applied in the experiment.



LS—loamy sand.

The agrotechnical procedures were carried out in accordance with the rules adopted for the species used in the test. White lupine was sown in early April. The following products were used for weed control: Afalon Dispersive 450 EC (1.1 L ha−1) in April, Basagran 480 SL (2.6 L ha−1) and Betanal MaxPro 209 OD (1.25 L ha−1) in May. Fusilade Forte 150 EC (1.0 L ha−1) was additionally applied in June. The following products were sprayed to protect the plants from diseases: Gwarant 500 SC (2.0 L ha<sup>−</sup>1) in May and Korazzo 250 SC (1.0 L ha−1) in mid- and late June.

#### *2.2. Weather Conditions*

During the growing seasons in 2016 and 2017 the weather conditions were similar in terms of temperature and rainfall. During the growing season the highest average air temperature was noted in July both in 2016 (19.5 ◦C) and 2017 (18.9 ◦C), whereas the lowest temperature was noted in April, i.e., 8.7 ◦C in 2016 and 7.5 ◦C in 2017. However, the weather conditions in 2018 were different than in the previous years (Figure 1). The highest average temperature was noted in August (21.2 ◦C), whereas the lowest was noted in May (12.7 ◦C). As far as the average monthly temperature from April to September is concerned, 2018 was the warmest—it was 2.9 ◦C warmer than 2016 and 1.7 ◦C warmer than 2017. In 2016 there was drought only at the end of the growing season. Likewise, in 2017 there was no rainfall deficit. On the contrary, it was a wet year, especially from June to August. On the other hand, in 2018 rainfall was unevenly distributed and there were droughts that were particularly unfavorable for plants in May, June, and August.

**Figure 1.** Climate graphs according to Walter [21] characterizing weather conditions in Gorzy ´n.

#### *2.3. Influence of Fertilizers on Nitrogenase Activity (Diazotrophy)*

Nitrogenase activity was estimated using the acetylene reduction assay (ARA) at the beginning of the plants' flowering [22]. For this purpose, five plants were randomly selected in plots, in a given experimental treatment and directly were placed tightly in sealed test vials (2000 mL) at the field, purified C2H4 was injected to obtain an acetylene concentration of 10% (*v*/*v*) in the gas phase (air). After an hour, 1 mL of the gas phase was taken from inside of the test vessels with a Hamilton gas-tight syringe and stored in small glass vials, which were sealed with rubber septa and aluminum seals. Ethylene concentration was determined using gas chromatograph CHROM 5 (Laboratorni Pˇristroje, Praha, Czech Republic, 1980). Nitrogenase activity was determined based on the quantity of acetylene

reduced to ethylene and expressed in nmolC2H4 produced per plant per hour (nMC2H4 plant−<sup>1</sup> h<sup>−</sup>1). The results are the mean value of five replications from each measurement.

#### *2.4. Plant Biometric Assessment*

The plant height (from soil surface to the highest plant point) and number of pods per plant were measured. Shoot, root, and nodule dry mass were determined after drying for 2 days at 70 ◦C until reaching constant weight. All the biometric traits were measured on 10 randomly selected plants from each object and replication during plant vegetation and before harvest. The total one-sided area of leaf per unit ground surface area expressed by the leaf area index (LAI) and was measured at three randomly selected places of each plot at the BBCH stage 69 using a SunScan Canopy Analysis System type SSI. Lupine was harvested at one stage (BBCH 90–92) with a 1.35-m wide plot combine. The yield of clean seeds was determined in dt·ha<sup>−</sup>1, given at a standardized (15%) water content and thousand seed weight was measured using a seed counter.

#### *2.5. Chlorophyll Fluorescence and Chlorophyll Content Measurements*

A fluorimeter (OS5p; Opti-Sciences, Inc., Hudson, NY, USA) was used to measure the efficiency of the photosynthetic apparatus. Prior to fluorescence measurements, the upper surface of three healthy leaves at the top of one plant from three randomly selected sites for each plot was covered with leaf clips for 30 min. Leaf fluorescence was then measured with a light pulse of 15,000 μmol m−<sup>2</sup> s−<sup>1</sup> at a wavelength of 660 nm The assessed parameter was maximum photosynthetic efficiency of PSII (Fv/Fm), which was calculated using the following formula: Fv/Fm = (Fm − F0)/Fm, on the basis of the measured parameters: minimal fluorescence (F0), maximal fluorescence (Fm), variable fluorescence (Fv) [23].Chlorophyll content meter (CCM-200plus; Opti-Sciences, USA) was used to estimate the chlorophyll content index (CCI) on the same leaves that were used for chlorophyll α fluorescence measurements. CCM-200plus measures the chlorophyll absorbance and calculates the chlorophyll content index, which is proportional to the concentration of chlorophyll in the sample.

#### *2.6. Soil Sampling for Biochemical Analyzes*

Soil samples collected from the arable layer (0–20 cm) were used as the research material for biochemical analyses. Each year they were collected at four terms: First term—at the plants' emergence (BBCH 5–10), Second term—at the plants' full growth (BBCH 35–40), third term—at the at the plants' florescence (BBCH 51–59), fourth term—after harvest.

Soil samples were taken from five places of each experimental plot, in four replications for each of the nine treatments of the experiment. In this way, at each analysis term we received 36 samples of soil, each of 1 kg.

#### *2.7. Soil Enzymatic Activity*

The analyses of soil enzymatic activity in individual treatments were based on the colorimetric method applied to measure the dehydrogenase activity (DHA), where 1% triphenyl tetrazolium chloride (TTC) was used as the substrate. The activity was measured after 24-h incubation at a temperature of 30 ◦C and a wavelength of 485 nm and it was expressed as μmol triphenyl formazane TPF 24 h−<sup>1</sup> g<sup>−</sup>1dm of soil [24].

Apart from that, the biochemical analyses of soil involved the determination of activities of acid (EC 3.1.3.2) phosphomonoesterases (PAC) and alkaline phosphomonoesterases (PAL) with the method developed by Tabatabai and Bremner [25]. The activities were determined with disodium *p*-nitrophenyl phosphate tetrahydrate used as a substrate after 1 h incubation at 37 ◦C and at a wavelength of 400 nm. The results were converted into μmol (p-nitrophenol) PNP h−<sup>1</sup> g<sup>−</sup>1dm of soil.

Catalase activity was measured by means of permanganometry, according to the method developed by Johnsons and Temple [26], where 0.3% H2O2 was the substrate. After 20-min incubation at room temperature (about 20 ◦C) 0.02 M KMnO4 was titrated to a light pink colour and expressed as μmol H2O2 g−<sup>1</sup> dm min<sup>−</sup>1.

#### *2.8. Biological Index of Fertility*

The biological index of fertility (BIF) was calculated using the dehydrogenase activity (DHA) and catalase activity (CAT) according to the Stefanic method [27] using the following formula: (DHA + kCAT)/2, where k was the factor of proportionality which equaled 0.01.

#### *2.9. Statistical Analyses*

The dynamics of changes in the soil enzymatic activity was statistically analyzed. As there were no significant differences between the parameters in the research years, they were treated as replicates and the results were analyzed by two-way ANOVA using Statistica 12.0 software. The fertilization method and the term of analysis were the factors differentiating the traits under study to estimate the soil biochemical activity parameters. Homogeneous subsets of mean were identified via Duncan's test at a significance level of *p* = 0.05. Yield, biometric, physiological traits of plants, and nitrogenase activity were tested once a year for the experiment. Hence, one-way analysis of variance (ANOVA) was used with Duncan's confidence interval, which was applied at a significance level of *p* = 0.05. As there were no significant differences between the parameters in the experimental years, they were treated as replicas.

Principal component analysis (PCA) was used to visualize the multidimensional dependencies between the soil biochemical activity and the types of fertilization [28]. In order to show the existing regularities (correlations) between biometric and physiological parameters of plants in individual years of research, a Pearson correlation matrix was determined, which was illustrated using a heatmap. The colors indicate the correlation coefficient values (from darkest—value −1, to the lightest—value +1). Cluster analysis enables grouping of the studied physiological parameters of the plants in the experiment in such a way that the degree of correlation between parameters within one group was the highest and between groups the smallest [29]. The agglomeration Ward method (Ward Hierarchical Clustering) and the Euclidean distance were used to create a tree diagram.

#### **3. Results**

#### *3.1. Yield, Biometric, and Physiological Traits of White Lupine Plants*

The studied biostimulators/foliar fertilizers modified the yield and yield components of white lupine. The yield of white lupine seeds was low and ranged from 11.67 dt·ha−<sup>1</sup> (ADOB B) to 13.88 dt·ha−<sup>1</sup> (Optysil) and depended significantly on the bio-stimulants or foliar fertilizers that were applied (Table 3). After applying Optysil or ADOB Zn IDHA (13.63 dt·ha−1), the yields were significantly higher when compared to the control plants by 1.82 and 1.57 dt·ha<sup>−</sup>1, respectively.

Thousand seed weight (TSW) was significantly higher than the control plants when ADOB Zn IDHA (322.7 g) was applied. All tested preparations significantly stimulated the height of white lupine. The strongest stimulation was obtained by Metalosate potassium, which increased the height of white lupine (40.5 cm) by 6.2 cm when compared to the control. Apart from these fertilizers, in the group that most strongly stimulated this trait were: Optysil (39.8 cm), ADOB 2.0 Mo (38.9 cm), and Bolero Mo (38.6 cm).

ADOB Zn IDHA (318.4 pc·m<sup>−</sup>2) and Tytanit (300.8 pc·m<sup>−</sup>2) significantly increased the number of pods compared to the control, and the increase was 96.6 and 79 pc·m<sup>−</sup>2, respectively.

Studies have also shown changes in nodulation and physiological parameters of the plant. Dry mass of root nodules was significantly stimulated after application of ADOB Zn IDHA (0.212 g) by 0.067 g when compared to the control treatment.

Chlorophyll fluorescence (Fv/Fm), showing the level of plant stress, was measured in the BBCH 69 (end of flowering) and BBCH 79 phases (75% of the pods reached typical length). At the end of flowering, the best plant condition, expressed by the Fv/Fm parameter, was obtained after the application of ADOB Zn IDHA (0.815) or Metalosate potassium (0.813) and both values significantly exceeded those obtained both with the control and all other treatments. In the assessment made at a later developmental phase, the tested biostimulators/foliar fertilizers did not significantly differentiate this parameter.

ADOB Zn IDHA application significantly stimulated the content of chlorophyll in leaves, expressed in CCI, which was 50.9 and exceeded the control by 17.4, as well as all other objects. In addition, significantly higher CCI values than in the control object were obtained after using Tytanit (46.7), ADOB B (42.9), or ADOB 2.0 Mo (40.7).

In turn, the significantly highest LAI value in the experiments was obtained after application of Rooter. The LAI value was 2.03 and exceeded the control by 0.62, for which the lowest LAI value was determined.


**Table 3.** The influence of the bio-stimulants and fertilizers on yield, biometric, and physiological traits of white lupine.

1. control—no bio-stimulants or foliar fertilizers applied to the plants; 2. plant + Tytanit; 3. plant + Optysil; 4. plant + Metalosate potassium; 5. plant + Rooter; 6. plant + Bolero Mo; 7. plant + ADOB Zn IDHA; 8. plant + ADOB B; 9. plant + ADOB 2.0 Mo; lack of homogeneous groups means no significant differences at the level of *p* < 0.05, a, b, c, d, e, f, g-homogeneous groups (Duncan's test. *p* < 0.05); TSW-thousand seed weight, Fv/Fm—maximum photosynthetic efficiency of PSII, CCI—chlorophyll content index, LAI—leaf area index.

The results of the experiment showed that foliar fertilizers and bio-stimulants affected the enzymatic activity of the soil and the biological index of fertility (BIF), as well as the nitrogenase activity in the white lupine plantation. The two-way analysis of variance showed that the foliar fertilization/bio-stimulants did not have a significant influence on the enzymatic activity and the soil biological index of fertility (BIF). Only the term of the test (development phase, based on BBCH scale) had a highly significant influence on the enzymatic activity and the biological index of fertility (BIF) of the soil (Table 4). One-way analysis of variance showed that foliar fertilization/bio-stimulants had a significant influence on nitrogenase activity.

**Table 4.** The test *F* statistics and the significance levels of the two-way analysis of variance for the soil bioactivity. The traits under analysis were affected by two factors, i.e., foliar fertilization and the term of the test.


*F* test statistics and significance levels of two-way analysis of variance for activity of enzymes associated with herbicides and terms research fixed factors \*\*\* *p* = 0.001, ns—no signification.

#### *3.2. Biological Fixation of Nitrogen under Lupine Plantation*

The field analyses of the biological fixation of nitrogen showed that the fertilizers and bio-stimulants significantly stimulated the nitrogenase activity in the white lupine plantation (Figure 2). During the three years in all the experimental treatments nitrogenase exhibited higher activity than in the control plot and differences were statistically significant. The highest nitrogenase activity was noted after the application of the ADOB B and ADOB Zn IDHA. The activity of the enzyme was respectively six and four times higher than in the control plot. Apart from the control treatment, the lowest biological fixation of nitrogen was noted after the application of Metalosate potassium.

**Figure 2.** The influence of the bio-stimulants and fertilizers on the level of biological fixation of nitrogen. Abbreviation: means values ± standard errors; a, b, c, d, e, f—homogenous groups according to Duncan's test at level *p* = 0.05.

The heat map presents correlations between all biometric and physiological characteristics of white lupine plants studied (Figure 3). Based on this visualization, relatively higher correlations were found between some features, including: PN (number of pods, pc.·m<sup>−</sup>2), TSW (thousand seed weight), H (height plant), PDM (plant dry mass), Y (seed yield), and PDM, Y, LAI (leaf area index) and Fv/Fm<sup>1</sup> (maximum photosynthetic efficiency of PSII BBCH–69). In turn, BNF (biological nitrogen fixation) and RNDM (root nodules dry mass) are negatively correlated with LAI, Y, PDM, H, TSW, PN, and Fv/Fm2 (maximum photosynthetic efficiency of PSII BBCH–78). Additionally, based on cluster analysis, groups of related biometric and physiological traits of plants were determined. Three groups have been designated. The first group that is the most distinct from the others contains: RNDM, BNF, CCI, and Fv/Fm1. The other two groups are: LAI, Y, PDM and H, TSW, PN, Fv/Fm2.

**Figure 3.** Correlations between all biometric and physiological characteristics of white lupine plants. Abbreviation: RNDM—root nodules dry mass, g—BNF—biological nitrogen fixation, CCI—chlorophyll content index, Fv/Fm1—maximum photosynthetic efficiency of PSII BBCH–69, LAI—leaf area index, Y—seed yield, PDM—plant dry mass, g, H—height plant, TSW—thousand seed weight, PN—number of pods, pc.·m<sup>−</sup>2; Fv/Fm2—maximum photosynthetic efficiency of PSII BBCH–78.

#### *3.3. Analysis of Soil Biochemical Activity*

Only the ADOB 2.0 Mo and Metalosate potassium foliar fertilizers stimulated the dehydrogenase activity throughout the growing season, as compared with the control treatment. After the application of the bio-stimulants the level of the enzyme activity was similar to the activity in the control treatment. However, when the Optysil and ADOB B were applied, the activity decreased but not statistically significantly. The experiment also showed that the peak of the dehydrogenase activity significant occurred at the third term of analyses, when the plants began flowering (BBCH 51–59). The results of the analysis of the dehydrogenase activity in the soil under the white lupine plantation are shown in Figure 4.

**Figure 4.** The influence of the bio-stimulants and fertilizers on the dehydrogenase activity. Abbreviation: a, b—homogenous groups according to Duncan's test at level *p* = 0.05; I term—at the plants' emergence (BBCH 5–10), II term—at the plants' full growth (BBCH 35–40), III term—at the at the plants' florescence (BBCH 51–59), IV term—after harvest.

The analysis of the results of the acid phosphatase activity (PAC) shows that during the entire white lupine growing season the foliar fertilizers and bio-stimulants decreased the activity of this enzyme, as compared with the control treatment (Figure 5). This effect was not observed when the Metalosate potassium foliar fertilizer was applied. During the second term of analyses, shortly before flowering, the acid phosphatase activity in all the experimental treatments was higher than in the control treatment. It was very high after the application of the Bolero Mo (0.170 μmol PNP h−<sup>1</sup> kg<sup>−</sup>1dm of soil) and ADOB 2.0 Mo (0.171 μmol PNP h−<sup>1</sup> g<sup>−</sup>1dm of soil).

The bio-stimulants and most of the foliar fertilizers did not increase the alkaline phosphatase (PAL) activity in the white lupine plantation, as compared with the control treatment (Figure 6). The ADOB 2.0 Mo and Bolero Mo stimulated the activity of this enzyme, which respectively increased by

14% and 5%, as compared with the control treatment. The enzyme exhibited statistically significantly increased activity shortly before they began flowering (II term).

**Figure 6.** The influence of the bio-stimulants and fertilizers on the alkaline phosphatase level. Abbreviation: a, b—homogenous groups according to Duncan's test at level *p* = 0.05; I term—at the plants' emergence (BBCH 5–10), II term—at the plants' full growth (BBCH 35–40), III term—at the at the plants' florescence (BBCH 51–59), IV term—after harvest.

All the preparations stimulated the catalase activity, as compared with the control treatment (Figure 7), but not significantly. The enzyme significantly exhibited high activity, i.e., when the plants started flowering (III term) in all the experimental treatments. The catalase activity ranged from 98.510 μmol H2O2g<sup>−</sup><sup>1</sup> dm min−<sup>1</sup> after the application of the Tytanit to 135.819 μmol H2O2g<sup>−</sup><sup>1</sup> dm min−<sup>1</sup> after the application of the Bolero Mo.

**Figure 7.** The influence of the bio-stimulants and fertilizers on the catalase activity. Abbreviation: a, b—homogenous groups according to Duncan's test at level *p* = 0.05; I term—at the plants' emergence (BBCH 5–10), II term—at the plants' full growth (BBCH 35–40), III term—at the at the plants' florescence (BBCH 51–59), IV term—after harvest.

The biological index of fertility (BIF), which was calculated on the basis of the dehydrogenase and catalase activity, was not always higher after the application of the bio-stimulants and foliar fertilizers (Figure 8). The highest value of this indicator was noted after the application of the Optysil and the lowest after ADOB Zn IDHA. The BIF was significantly high at the beginning of flowering, as it ranged from 5.17 after the application of ADOB Zn IDHA to 12.34 after the application of ADOB 2.0 Mo. The indicator was also high after the application of the Bolero Mo and Optysil.

**Figure 8.** The influence of the bio-stimulants and fertilizers on the BIF. Abbreviation: a, b—homogenous groups according to Duncan's test at level *p* = 0.05; I term—at the plants' emergence (BBCH 5–10), II term—at the plants' full growth (BBCH 35–40), III term—at the at the plants' florescence (BBCH 51–59), IV term—after harvest.

Principal component analysis (PCA) was used to show how the foliar fertilizers and bio-stimulants affected the white lupine plantation. The first two principal components accounted for over 89.2% of the total variation (Figure 9). The parameters of the soil biochemical activity in 2018 differed significantly from 2016 to 2017. This effect may have been caused by the weather conditions (Figure 1). In 2018 the season was the warmest of all the research years. The average temperature difference between 2018 and the previous years was 2.9 ◦C in August and 1.7 ◦C in May. As the thermal conditions were very similar in 2016 and 2017, the PCA showed similar dependencies for these two years. In 2016 the fertilizer preparations and bio-stimulants significantly affected the catalytic activity of acid phosphatase (PAC) at all the terms of analyses. This dependency was not observed for the other parameters of soil biochemical activity.

**Figure 9.** The dependence between the soil enzymatic activity and all treatments with fertilizers and bio-stimulants at the terms of analyses. Abbreviation: I term—at the plants' emergence (BBCH 5–10), II term—at the plants' full growth (BBCH 35–40), III term—at the at the plants' florescence (BBCH 51–59), IV term—after harvest. BIF—index of fertility, CAT—catalase activity, PAC—acid phosphomonoesterases, PAL—alkaline phosphomonoesterases, DHA—dehydrogenase activity.

In 2017 the application of the fertilizers did not cause significant differences in the activity of the soil enzymes or the biological index of soil fertility. In dry 2018 the preparations did not significantly affect the catalytic activity of the test parameters only during plants' emergence (I term). However, at the plants' full growth (II term), the foliar fertilizers and bio-stimulants strongly influenced the catalytic activity of catalase (CAT), dehydrogenase (DHA), alkaline phosphatase (PAL), and the biological index of soil fertility (BIF). Apart from that, the principal component analysis showed that in 2018 the indicators of soil biochemical activity were affected most strongly by foliar fertilizers and bio-stimulants the flowering of the plants (III term) and after the harvest (IV term).

#### **4. Discussion**

#### *4.1. Yield, Biometric, and Physiological Traits*

Silicon, iron, manganese, boron, copper, molybdenum, and zinc are the basic micronutrients. The silicon content in most plants is comparable to the content of calcium, magnesium, and phosphorus. Many studies have shown the positive effects of silicon on plants, their development, yield, and sensitivity to biotic and abiotic stress [30]. In many tests, silicon has been shown to significantly influence the regulation of nutrient uptake such as: calcium, magnesium, and phosphorus. In other studies [31], silicon fertilization increased the yield of sugar beet roots by 13.7–15.9%, as well as the yield of many other species [11], especially in the form of spraying plants under stress conditions. According to Fageria and Baligar [32] and Duffy [33] Zn is the microelement most limiting crop yield. Zinc is taken up in small amounts and it participates in all major functions of the plant, increases nitrogen uptake, and activates CO2 binding in later stages. Hence, it is necessary in plant nutrition and its importance in plant production is growing [13]. Similarly, Kaya et al. [34] obtained the highest common bean plants (*Phaseolus vulgaris* L.), with the largest number of pods and seeds per plant after application of a foliar mixture of zinc.

The preparations used in our study also stimulated the tested biometric parameters of the plants. Plant height was stimulated the most after application of Metalosate potassium (by 8.5%) when compared to the control treatments. In turn, the number of nodules was most strongly stimulated by ADOB Zn IDHA (by 68.4%) and LAI by Rooter (by 69.5%). Other fertilizers containing boron, molybdenum, silicon, and titanite also increased the parameters indicated above. These results are consistent with the results of Raj and Raj [12] regarding the beneficial effects of Zn on plant efficiency, physiological parameters, plant height, and nodulation formation. Our results are also consistent with field studies of Omer et al. [35], in which the treatments of molybdenum application did not modify any of the studied lentil characteristics, except for the height of the plant. Also Rahman et al. [36] showed that the use of molybdenum in its deficiency in soil, stimulates the formation of nodules. Of the physiological traits studied, chlorophyll fluorescence (Fv/Fm) was most strongly stimulated by ADOB Zn IDHA (by 3.9%) and Metalosate potassium (by 3.7%). In turn, the CCI index was most strongly stimulated by ADOB Zn IDHA, whose application resulted in an increase of this parameter by 51.9% when compared to the control treatment. The results of research on *Vigna sinensis* [37] and on *Celosia* [38] showed that Zn spraying on plants caused a significant increase in chlorophyll content. In a study conducted by Artyszak et al. [39], foliar fertilization with silicon increased LAI and effective quantum efficiency of PSII—ΦPSII, as well as positively affected the growth and development of many plant species [40,41].

#### *4.2. Biological Fixation of Nitrogen*

The bio-stimulants and foliar fertilizers which improved the biological fixation of nitrogen in the white lupine plantation contained important macro- or microelements. Scientific reports suggest that some elements are particularly significant to the nitrogen fixation process.

Mineral nutrients may influence N2 fixation in legumes at various stages of the symbiotic process: infection and nodule development, nodule function, and host plant growth. For healthy and vigorous growth, intact plants need to take up relatively large amounts of some inorganic elements: ions of nitrogen (N), potassium (K), calcium (Ca), phosphorus (P) and sulphur (S), and small quantities of other elements: iron (Fe), nickel (Ni), chlorine (Cl), manganese (Mn), zinc (Zn), boron (B), copper (Cu), and molybdenum (Mo). Molybdenum and iron are especially important because they are components of the nitrogenase complex in rhizobia which is required for nitrogen fixation. They are components of nitrogenase—the bacterial enzyme that enables the diazotrophy process. The nitrogenase protein consists of two subunits: the larger one containing the FeMo cofactor and the smaller one containing iron alone [42]. Plants growing on acidic, moist, and poorly buffered soils do not have sufficient supply of molybdenum. When molybdenum is applied in a field to the leaves of legumes, the nitrogen fixation of these plants is more efficient, and the mass of their root nodules and the yield of seeds increase [43,44]. The use of ADOB 2.0 Mo with high molybdenum content in our experiment confirmed this fact. There are small amounts of boron in plants, but this micronutrient plays an important role in various physiological processes. It affects the separation of plant tissues and it is necessary for the optimal growth of plants. Boron-deficient plants have less bacteria of the *Rhizobium* genus and fewer infection threads [44]. The significant increase in the level of biological fixation of nitrogen may have been caused by the application of the foliar fertilizer containing boron (ADOB B). Our research also proved that zinc supplied with the ADOB Zn IDHA foliar fertilizer significantly increased the nitrogenase activity. Although plants absorb moderate amounts of zinc, this element has significant influence on bacteria of the *Rhizobium* genus. The research by Wani et al. [45] showed that higher concentrations of this element in soil stimulate bacteria of the *Rhizobium* genus to produce phytohormones (including

indoleacetic acid), which promote the growth of plants by increasing the number of root nodules, their dry mass, and the content of leghemoglobin in the nodules.

Many researchers have studied the role of phosphorus in symbiotic systems. Phosphorus plays a crucial role in the nitrogen fixation process [46,47]. The Rooter bio-stimulant, which contained phosphorus and potassium, stimulated this process considerably. Phosphorus participates in a wide range of molecular and biochemical processes. Apart from that, some phosphate bonds are carriers of the energy used in cells. The presence of phosphorus in soil affects the plant's ability to produce root nodules, especially the weight and the number of nodules [48], which translates into the level of nitrogen fixation.

When the supply of phosphorus is insufficient, plants often suffer from nitrogen deficiency. Sulphur and potassium are less important for symbiotic systems than the aforementioned elements. Nevertheless, potassium ions are very desirable in saline soils because they function as an osmolyte. In view of the fact that nearly half of irrigated soils around the world are considered saline, the addition of potassium helps to maintain the bacteria-plant system [48,49].

#### *4.3. Biochemical Activity*

The activities of soil enzymes are considered sensitive indicators of important microbial reactions involved in nutrient cycles and they respond to changes in the soil caused by natural or anthropogenic factors. In this regard, soil enzyme activities are often used to evaluate the impact of human activity on soil life [50].

Soil enzymes are a group of catalysts that significantly affect the ecological properties of the pedosphere. These are both extracellular enzymes and the ones that are present in microorganisms (both in proliferating cells and in endospores). Enzymes control the course of all chemical reactions in microbial cells, e.g., the synthesis of proteins, nucleic acids, and carbohydrates [51]. Soil enzymes are involved in the decomposition of organic substances released into the soil during the plant's growth as well as the formation and decomposition of humus in the soil. They release and transfer minerals to plants. In spite of the dynamic nature of the microbiological and biochemical properties, soil enzymes are accurate and significant determinants of soil fertility, and they are important indicators of changes taking place in the soil [52,53].

Dehydrogenases (DHA) are enzymes belonging to the group of oxidoreductases. They are responsible for catalyzing the oxidation of organic compounds. Active dehydrogenases are present only inside living cells and they indicate the presence of physiologically active microorganisms. Dehydrogenases are commonly found in the pedosphere, where they are involved in the decomposition of organic compounds. Measurement of the dehydrogenase activity in soil shows the intensity of respiratory metabolism of soil microorganisms, mainly actinobacteria and bacteria.

Our research showed that only some foliar fertilizers (ADOB 2.0 Mo and Metalosate Potassium) stimulated the dehydrogenase activity in the white lupine plantation, however, the results were not significant.

Dehydrogenase exhibited high activity at the beginning of the plants' flowering phase (BBCH 51–59). It may have been caused by an increased secretion from the root system during that period [54,55]. In consequence, the count of microorganisms increased [56].

Also macro- and microelements applied in the form of foliar fertilizers and biostimulators could affect dehydrogenase activity. Bieli ´nska et al. [57] observed that fertilizing preparations with nitrogen, phosphorus, and potassium increased the content of these enzymes in the soil. There was a similar effect observed in our study after the application of the Metalosate potassium foliar fertilizer. There were analogous results of experiments on similar bio-conditioners conducted by [58] and [53]. According to Bilen et al. [59], boron improves the dehydrogenase activity. Taran et al. [60] showed that molybdenum stimulated the production of these enzymes by the root nodules of legumes. They also observed that the content of titanium might be positively correlated with the soil biochemistry.

The results of the experiment showed that both the bio-stimulants (Tytanit and Rooter) and foliar fertilizers positively affected the acid phosphatase activity, which was lower than in the control treatment. The Metalosate potassium foliar fertilizer did not cause this effect. This shows that the preparations used in our experiment positively influenced the plants' ability to absorb phosphorus. It is necessary to remember that phosphorus-deficient plants are characterized by increased secretion of acid phosphatase through the root system into the soil. Ciereszko et al. [61] found that the deficit of this macroelement stimulated plants' secretion of acid phosphatases. Lemanowicz et al. [62] and Niewiadomska et al. [56] also suggest these relationships in their studies on the effect of the PRP SOL fertilizer containing phosphorus, potassium, zinc, boron, and molybdenum on the lupine plantation. They observed a decrease in the catalytic activity of this enzyme because of the activation of the compounds that were inaccessible to plants. Bieli ´nska and Mocek-Płóciniak [63] made similar observations. Wang et al. [64] also found that these enzymes exhibited higher activity in the experimental treatment without phosphorus fertilization.

The alkaline phosphatase activity increased significantly only after the application of the ADOB 2.0 Mo and Bolero Mo foliar fertilizers. This effect may have been caused by the increased activity of soil microorganisms, which were stimulated by organic phosphorus compounds secreted into the soil by white lupine plants. Waldrip et al. [65] proved that the content of organic forms of phosphorus was correlated with the activity of alkaline phosphatases in the soil.

All the preparations used in the experiment significantly stimulated catalase activity. As early as 1963, Koter [66] found that the catalase activity increased when plants were fertilized with boron. Hu and Zhu [67] observed that the catalase and dehydrogenase activity increased when plants were fertilized with silicon. Such elements as copper and zinc are essential constituents of physiological processes in all living organisms, including microorganisms. Some soils suffer from zinc deficits, which is why they are enriched with fertilizers containing this element to satisfy the nutritional requirements of crops and improved soil activity [68].

The results of the enzymatic analyses of the dehydrogenase and catalase activities enabled the calculation of the biological index of soil fertility (BIF). The treatments with the Optysil and ADOBE 2.0 Mo preparations had influence on the BIF values, as compared with the control sample. The use of the Optysil preparation resulted in particularly high values in the soil samples collected at the beginning of the flowering phase. The BIF value resulted from the significant influence of these fertilizers on the activity of catalase and dehydrogenase. Siwik-Ziomek and Szczepanek [69] indicated that mineral fertilization, which increases the yield of crops, indirectly affects root secretion, and thus increases the biochemical activity of soil at specific phases of plants' development.

#### **5. Conclusions**

When non-root fertilization is applied to plants, they take up all necessary elements chiefly through their leaves as well as the stalk and the whole aerial system. A strong stimulating effect on the yield of white lupine plants in comparison with the control object was obtained after the application of silicon (Optysil) or chelated zinc (ADOB Zn IDHA). The use of zinc in foliar fertilizers (ADOB Zn IDHA) in comparison with control treatment stimulated most of the tested features/parameters: TSW, number of pods per 1 m2, root nodules dry matter, photochemical efficiency of PSII (Fv/Fm), and chlorophyll content (CCI). However, it is noteworthy that this way of "feeding" cannot substitute soil fertilization. It can only be used to quickly supply necessary nutrients to plants at difficult phases so as not to slow down their growth. The bio-stimulants and foliar fertilizers used in our study improved some of the biochemical parameters of soil activity and the nitrogen fixation process in the white lupine plantation. This effect may have been caused by the higher rate of penetration and better uptake of nutrients applied to the plants' leaves. Although macro- and micronutrients differ in their penetration rates, this process can be accelerated up to about a dozen times by non-root fertilization. The downside of foliar fertilization is the fact that only a limited amount of fertilizer can be supplied to plants in this way. Therefore, this method is particularly effective when plants need to be provided with the elements

they need in smaller amounts, e.g., iron, boron, and molybdenum. Not only is foliar fertilization a more efficient method of supplying micronutrients, but it is also safer for the environment and the plants. The search for methods that improve the yield and biochemical parameters of the soil environment is in agreement with the sustainable agriculture policy.

**Author Contributions:** A.N., H.S., A.W.-M., K.R., and Z.W. conceived and designed the experiments; A.N., H.S., A.W.-M., K.R., and Z.W. performed the field experiments and analyzed the data; A.B. statistical analysis; A.N., Z.W., H.S. wrote the paper; A.N., Z.W. revised the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Ministry of Agriculture and Rural Development of Poland, (it was not a grant. And The APC was funded by framework of Ministry of Science and Higher Education programme Project No. 005/RID/2018/19.

**Acknowledgments:** This research was financed by the Ministry of Agriculture and Rural Development of the Republic of Poland in the framework of the Multi-annual Program "Increasing the use of domestic feed protein for the production of high quality animal products in conditions of balanced development" implemented at the Department of Agronomy of the University of Life Sciences in Poznan. The publication was co-financed within the framework of Ministry of Science and Higher Education programme as, Regional Initiative Excellence" in years 2019-2022, Project No. 005/RID/2018/19.

**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* **Biostimulant Seed Coating Treatments to Improve Cover Crop Germination and Seedling Growth**

**Yi Qiu 1,2,**†**, Masoume Amirkhani 1,\*,**†**, Hilary Mayton 1, Zhi Chen <sup>2</sup> and Alan G. Taylor <sup>1</sup>**


Received: 28 December 2019; Accepted: 19 January 2020; Published: 22 January 2020

**Abstract:** Biostimulant seed coating formulations were investigated in laboratory experiments for their potential to increase maximum germination, germination rate, germination uniformity, and seedling growth of red clover (*Trifolium pratense* L.) and perennial ryegrass (*Lolium perenne* L.) seeds. Red clover and perennial ryegrass seeds were coated with different combinations of soy flour, diatomaceous earth, micronized vermicompost, and concentrated vermicompost extract. Coated and non-coated seeds of red clover and perennial ryegrass were evaluated for germination and growth after 7 and 10 days, respectively. Red clover seed was maintained at a constant 20 ◦C with a 16/8 h photoperiod, whereas for perennial ryegrass seed, the germinator was maintained at 15/25 ◦C, with the same photoperiod as red clover. Coated treatments significantly improved germination rate and uniformity with no reduction in total germination, compared to the non-treated controls in red clover. In contrast, for perennial ryegrass, the total germination percentage of all coated seeds was reduced and displayed a delayed germination rate, compared with the non-treated controls. Shoot length, seedling vigor index, and dry weight of seedlings of coated seed treatments of both crops were significantly higher when compared to controls for both species. In addition to growth metrics, specific surface mechanical properties related to seed coating quality of seeds of both species were evaluated. Increasing the proportion of soy flour as a seed treatment binder in the coating blend increased the integrity and compressive strength of coated seeds, and the time for coatings to disintegrate. These data show that seed coating technologies incorporating nutritional materials and biostimulants can enhance seedling growth and have the potential to facilitate the establishment of cover crops in agriculture and land reclamation.

**Keywords:** seed coating; cover crop; vermicompost; biostimulant; growth enhancement

#### **1. Introduction**

Exponential growth in human global population, from 1.7 billion in 1900 to approximately 7.6 billion in 2019, has led to the over use and degradation of agricultural landscapes, including grasslands used for grazing, forage, and food production [1,2]. The rapid growth of populations in pastoral areas, including Inner Mongolia, China, have caused intensive overutilization of grasslands. Approximately, 40% of land area in China is classified as grassland and accounts for 13% of the world's grassland [2,3]. Overgrazing and conversion of grassland to cropland has led to declines in overall agricultural productivity due to increased soil erosion, degraded soil structure, and reduced soil fertility. Recently, China implemented vegetation restoration programs to improve biodiversity in agriculture environments, soil health, and productivity, and to reduce erosion and desertification [2]. Legumes and ryegrasses are widely used as cover crops to reduce desertification and restore productivity on

degraded grasslands [4] and are commonly used for land reclamation and restoration of abandoned mine land [5,6]. Perennial ryegrass (*Lolium perenne* L.) is a cool season grass native to southern Europe, the Middle East, Central Asia, and North Africa [7]. Ryegrass is often used to stabilize soils for erosion control and is frequently seeded with red (*Trifolium pratense* L.) or white (*Trifolium repens* L.) clovers for increased productivity in grazing and to provide nitrogen and aid in weed suppression [8]. Red clover is particularly tolerant to drought conditions and helps to improve soil structure due to its large, fast-growing (more than 60 cm/year) tap root [8]. The benefits of cover cropping in both organic and conventionally managed systems are well documented [9].

Cover crops increase soil organic matter, soil structure, nutrient retention and cycling, and reduce soil erosion [8]. However, under drought conditions, and in areas with poor soils such as arid degraded grasslands, germination and subsequent growth of cover crops are inadequate, and sowing is often unsuccessful. Seed enhancements, which can include seed priming, coating, and conditioning are frequently used to improve seed delivery during planting, and to increase seed germination, stand uniformity, seedling growth, and suppress disease [10]. Seed priming increased germination rate and overall seedling emergence in a study investigating perennial ryegrass for fall seeding under cool temperatures and improved wheat stand establishment under marginal soil conditions [11,12]. Seed treatments with fungicides and fertilizers enhanced stand establishment of perennial ryegrass in field experiments in New Zealand [13,14]. Seed enhancements via seed coating can also provide micro and macronutrients or biostimulant materials to increase germination, seedling vigor, and stand establishment [15]. Seed coating technology has been used as a promising and effective approach for enhancing establishment and yield of different grass and forage species (*Lolium perenne, Trifolium pratense, Elymus dahuricus*) in semi-arid areas of China such as Inner Mongolia [16–18].

Biostimulants are materials that can augment plant growth when applied to plants and seeds, but are not classified as fertilizers, pesticides, or soil amendments [19]. Commonly applied biostimulants include microbial inoculants, beneficial bacteria and fungi, nitrogen containing compounds, biopolymers, and plant extracts [19]. Research and use of biostimulants in agriculture has increased in recent years in an effort to reduce reliance on less sustainable conventional pesticides and fertilizers, which are often overused, in agricultural cropping systems [15,19–22]. Seed treatments require even smaller amounts of active ingredients per hectare than foliar applied treatments, primarily due to the reduced surface area treated, and increase germination and plant growth when compared to non-treated seed [23].

Modern seed coating technology utilizes different approaches depending on the shape and size of the seed and the type and amount of materials added to seeds [10,24–27]. Currently, seed pelleting, film coating, and seed encrusting are the most common coating/treatment procedures used in the seed industry to enhance plant and seedling performance. While seed pelleting, often employed to develop more uniform seeds for mechanical planting, can increase seed weight from 200 to > 5000%, film coating or encrusting utilizes much smaller quantities of materials resulting in a build-up in seed weight of between 0.5–10% and 20–200%, respectively [24]. The physical properties and thickness of the seed treatment/coating are critical factors that influence seed germination and seedling vigor. A thick hard seed coating can reduce, delay, or cause abnormal germination or may even be toxic, while a minimal, fragile seed coating can break or disintegrate before planting or not have a high enough dosage of an active ingredient to be effective. Therefore, specialized seed coating formulations must be developed and evaluated in order to be utilized effectively for any given plant species and agronomic purpose.

The specific objectives of this research were to explore plant-derived bio-based biostimulant seed coatings to enhance germination and growth of two cover crop species, red clover and perennial ryegrass, as an approach for seeding cover crops for grassland restoration. Previous research on seed coatings of broccoli and tomato with soy flour and compost materials showed promising results related to maximum germination, germination uniformity, and seedling vigor [15,25–27].

#### **2. Materials and Methods**

#### *2.1. Seed and Coating Materials*

Two species of cover crops were selected to evaluate biostimulant seed coatings in this study. Red clover 'VSN-variety not stated' seed was obtained from King's AgriSeeds, Inc., Lancaster, PA, USA, and 'Tetraprime' perennial ryegrass seed was provided by SeedWay, LLC, Penn Yan, NY, USA. The red clover and perennial ryegrass seeds were coated with different combinations of soy flour (SF), diatomaceous earth (DE), micronized vermicompost (MVC-2 and 3), and concentrated vermicompost extract (CVE) to identify the most stable and effective coating formulations (Table 1). Specific treatments and ratios of coating materials evaluated were (SF:DE 30:70, 40:60, 50:50 and 60:40, SF:MVC-2 (30:70), SF:MVC-3 (30:70), SF:DE:CVE (30:70)). A mechanical Ro-Tap shaker (Ro-Tap Testing Sieve Shaker No. 1506; The W.S. Tyler Co., Cleveland, OH) was utilized to sieve the SF to obtain a particle size smaller than (<75 μ), as previous studies have shown that smaller particle size results in more even distribution of the SF coating on seeds [15]. Seed coating biostimulant materials used in this research were previously analyzed by the Cornell Soil Health Nutrient Analysis Laboratories and recently reported in [27].

**Table 1.** Materials used as seed coating biostimulant treatment formulations in this study.


#### *2.2. Seed Coating*

A 15-cm diameter, R-6 (Universal Coating Systems, Independence, OR, USA) laboratory-scale rotary pan coater was used to coat seeds in all experiments (Figure 1). Each seed coating treatment consisted of two components: dry powder and liquid. SF and DE, SF and MVC-2, and SF and MVC-3 were applied as dry powder to the seed surface with distilled water. For the SF, DE, and CVE treatments, water was replaced with liquid compost extract. For each treatment, the powder and liquid were applied to the surface of the seeds in incremental amounts as they rotated in the R6 to achieve uniform results. To clean the residual dust of each coating batch and avoid cross contamination of treatments, the R-6 pan was cleaned with a sponge and hot water-liquid soap solution. This was followed by cleaning with disinfectant wipes (three times), and R-6 with high pressure air flow was applied around the pan and cylinder to ensure completely drying. In this study, coating combinations of SF:DE and SF:MVC for both crops were applied on separate days.

Twenty-five grams of seed were used for red clover (1000 seed weight = 2.5 g) treatments, and 15 g of seed were used for the perennial ryegrass (1000 seed weight = 1.5 g). Therefore, we had an equal amount of treated seeds (~10,000 seeds) for each treatment of each crop. The total build up percentage was approximately 30% for clover and 70% for perennial ryegrass (Figure 2). Variation in the percent build up needed to achieve uniform coverage reflects the need for specifically developed seed coatings for each seed type and species. The size and shape of seeds influences uniformity of treatments on the seeds. After coating, the seeds were dried at room temperature for 24 h (h) until completely dry [15,28]. To improve observation of the seed coating uniformity of the SF:DE and SF:DE:CVE blend, a red dye (Pro-Ized red colorant, Bayer Corp, Research Triangle Park, NC, USA) was added to the binder (1.0 mL dye per 10 mL binder). Due to the dark brown color of the MVC materials, no dye was used for SF and MVC combinations; the natural color showed coating quality and uniformity of application.

**Figure 1.** Figure of seed coating methodology used as an approach for application of biostimulant compounds for sustainable agriculture.

**Figure 2.** Non-coated and coated (SF:DE 30:70) red clover seeds with 30% build up (**a**) and non-coated and coated (SF:DE 30:70) perennial ryegrass with 70% build up (**b**). To improve observation of seed coating uniformity, a red dye was added to the binder.

#### *2.3. Seed Coat Physical Properties*

#### 2.3.1. Seed Coating Integrity Test

The strength of the coating is an important quality as it relates to germination and potential for damage during handling, transportation, and planting. The surface material of coated seeds must have good mechanical properties to ensure that they do not crack or disintegrate before sowing. A Ro-Tap sieve shaker (The W.S. Tyler Co., Cleveland, OH, USA) was used to test the integrity of the coated seeds [15]. Four replications of 1.5 g of coated red clover seeds and four replications of 1.5 g of perennial ryegrass seeds from each coating formulation (treatments listed in Table 2) were tested to assure reliability and reproducibility. Samples were weighed and shaken for 2 min using a standard Ro-tap test shaker with U.S. Standard Testing Sieve No. 25 (0.71 mm opening) and a solid catch pan. Each sample was weighed again, and the percentage of coating loss was calculated according to the weight before and after the Ro-tap procedure. The weight of coating material, which passed through a No. 25 sieve was reported as weight loss (WL %) of material.

**Table 2.** Results of seed coating physical property testing, weight loss (WL, %), disintegration time (DT, min), compressive strength (Force N), time to break (TB) seed coating measured in s, relaxation time (RT) after the seed coating was fractured measured in s for coated seeds of red clover and perennial ryegrass.


\* Different letters within each column for each crop indicate significant differences using a Least Significant Difference (LSD) test at a significance level of *p* < 0.05. Lower case letters represent significant differences in red clover treatments and upper case letters denote perennial ryegrass treatment differences.

#### 2.3.2. Mechanical Property Test

A texture analyzer (TA-XTplusC, Texture Technologies Corp., Hamilton, MA) was used to test the compressive strength of coated seeds. The TA-XTplusC is a precision instrument used to measure the surface mechanical properties of coated seeds and the compressive strength of a single seed. The arm of the texture analyzer containing a weighing sensor moves in a downward motion to compress the coated seed placed on the base of the analyzer and then returns to its original position. Data are assessed as the compressive strength (Force N) and time to breakage (TB, measured in seconds (s)) required to fracture the seed coating. The relaxation time (RT), which is the time required to completely rupture the seed coating was also measured [29,30]. After the seed coating was completely broken, the force (N) increases until the seed embryo was crushed (Figure 3). Texture analyzer software (Exponent Connect, version 7.0.2.0, S. Hamilton, MA, USA, 2018) was used to record force for TB and RT [26]. Ten coated seeds were randomly selected from batches of different formulations (SF:DE = 30:70, 40:60, 50:50 and 60:40, SF:MVC-2 (30:70), SF:MVC-3 (30:70), SF:DE:CVE (30:70)) to test their surface compressive strength for both red clover and perennial ryegrass coated seeds.

**Figure 3.** The values of peak load force required to break the seed coat of a single seed from two different seed coating blends of soy flour (SF) and diatomaceous earth (DE) tested at room temperature (SF:DE 30:70 and 60:40) for red clover are 16.2 and 23.8 N, respectively. The maximum force value (N) is a measure of coating strength and shows the maximum force needed to break the seed coat. Time to break seed coating (TB) and Relaxation Time (RT) after seed coat fracture until the coat is completely broken, both measured in s, are shown for a single seed. Force, TB, and RT data shown in Table 2 are the means of 10 seeds (Equipment: TA-XTplusC, Texture Technologies Corp., Hamilton, MA, USA, Software: Exponent Connect, version 7.0.2.0).

#### 2.3.3. Seed Coating Hydration Test

The wet strength of a seed coating is largely dependent on the adhesion of the components after immersion in water. In theory, the slower the decomposition rate of coated seeds in water is, the more likely it is to delay germination. The hydration test was used to investigate the integrity of the coating materials when immersed in water. Hydration tests were conducted to determine the potential for seed coatings to prevent or delay germination. Four replicates of 100 coated seeds with different proportions of soy flour and diatomaceous earth, SF:DE (30:70, 40:60, 50:50 and 60:40), and the soy flour micronized vermicompost treatments SF: MVC-2, SF: MVC-3 were placed in 5 mL of distilled water to determine disintegration time. Disintegration time was measured in minutes.

#### *2.4. Seed Germination and Seedling Growth Measurements*

Four replicates of 50 treated and non-coated control seeds were placed on two 30 cm × 45 cm moistened germination paper towels (Anchor Paper Company, St. Paul, MN, USA); then, an additional moistened standard germination paper towel was placed on top of the seeds. The towels were rolled and positioned in a germinator (Percival germinator, Model I-36LL, Perry, IA, USA). For perennial ryegrass seed, the germinator was maintained at 15/25 ◦C, with a 16/8 h photoperiod [31]; red clover, was maintained at a constant 20 ◦C with the same photoperiod [31]. Radical emergence (>2 mm) was

used to determine successful seed germination. The number of newly germinated seeds for both red clover and perennial ryegrass was recorded every 24 h. For perennial ryegrass, the total germination percentage (Gmax %) was recorded after 10 days. The Gmax % for red clover was recorded after seven days. Gmax %, the number of germinated seeds and germination uniformity (GU), (GU = time required for 90% germination subtracted by time required for 10% germination) were calculated [32] for each treatment. In addition, germination rate (T50, the time in h to reach 50% total germination) was calculated according to the equation developed by Coolbear et al. [33].

Root and shoot measurements (cm) were conducted in separate roll towel experiments (under the same growing conditions described above for seed germination) for each treatment using four replicates of 50 seeds. The seed vigor index (SVI) was equal to Gmax % multiplied by seedling length (combined root and shoot lengths) divided by 100 [34]. Seedlings were measured a week after full emergence for both crops. After measuring shoot and root lengths, all seedlings from each treatment were dried in an oven at 80 ◦C for 48 h to obtain the dry weight data.

#### *2.5. Statistical Analysis*

In all experiments, normality tests were conducted prior to ANOVA and all data passed the normal distribution test at a significance level of 0.05. Analysis of variance (ANOVA) (α = 0.05) and Fisher's least significant difference test for seed coating physical property and Dunnett test for germination and seedling growth data were performed on each of the significant variables measured by Minitab Express [35]. All Gmax % (Tables 2–4) and WL % (Table 2) data were arcsine transformed for analysis. Data for Gmax % and WL% are presented as non-transformed means (Tables 2–4). Pearson correlation was conducted for coating physical properties data using Minitab Express (Table 5).

**Table 3.** Germination and growth metrics of soy flour formulations as measured by total germination (Gmax %), germination rate (T50) measured in hours (h), germination uniformity (GU) measured in hours (h), shoot and root length (cm), and Seedling Vigor Index (SVI = Gmax % × Seedling length) of different coating formulas of SF:DE for red clover and perennial ryegrass.


\* Different letters within each column for each crop indicate significant differences using a Dunnett test at a significance level of *p* < 0.05. Lower case letters represent significant differences between each red clover seed coating treatment compared with the control and upper case letters denote each of perennial ryegrass treatment differences compared with the control.

**Table 4.** Germination and growth metrics of soy flour/vermicompost formulations as measured by total germination (Gmax %), germination rate (T50) measured in hours (h), germination uniformity (GU) measured in hours (h), shoot and root length (cm), seedling dry weight (DW) recorded in grams (g), and Seedling Vigor Index (SVI = Gmax % × Seedling length) from evaluation of different coating materials for red clover and perennial ryegrass. The proportion of all coating materials is 30:70 (30% SF and 70% of DE or MVC).


\* Different letters within each column for each crop indicate significant differences using a Dunnett test at a significance level of *p* < 0.05. Lower case letters represent significant differences between each red clover seed coating treatment compared with the control and upper case letters denote each of perennial ryegrass treatment differences compared with the control.

**Table 5.** Correlation coefficients between disintegration time (DT, min), weight loss (WL, %), and compressive strength (Force, N) from seed coating applications of soy flour and diatomaceous earth on red clover and perennial ryegrass seeds.


\*\*, \*\*\* Significant at *p* < 0.001, 0.0001, respectively.

#### **3. Results and Discussion**

#### *3.1. Seed Coating Physical Properties*

The integrity and physical properties of coated and pelleted seeds are critical for overall performance. The production of dust can lead to health and environmental risks; therefore, it is important to quantitatively analyze the potential for breakage and weight loss that may occur during transportation and handling. In contrast, a seed coating that is too hard or impermeable to water may hinder germination. In this study, physical properties of the various seed coating formulations were tested by employing three different tests, including mechanical, texture, and hydration analyses.

Experimental results from the seed coating mechanical Ro-tap and Texture analysis (TA-XTplusC) tests are presented in Table 2 and Figure 3. For both crops, increasing the proportion of soy flour in the coating blend increased the compressive strength (Force N) of coated seeds. The time required to break the seed coating (TB), measured in s and relaxation time (RT) after breaking seed coating (Table 2) increased as soy flour proportion increased (Table 2). For example, for red clover, the average force (N) increased from 16.2 to 23.9 N as the soy flour content increased from 30% to 60%, which is an increase of approximately 48%. As soy flour content increased, the TB of coated seeds gradually increased from 4.4 to 5.5 s for red clover and 5.7 to 6.4 s for perennial ryegrass (Table 2). Although the same seed coating blend were used for both crops, the TB ranges were different most likely due to the difference in build-up percentage. There was a 1.1 s delay in breakage time for red clover when the content of soy flour increased from 30% to 60%, and 0.7 s delay for perennial ryegrass when the content

of soy flour increased from 30% to 50%. The force value to break down the coating for perennial ryegrass significantly increased by approximately 37% as soy flour increased from 30% (15.3) to 50% (20.9). Additionally, the weight loss of coated seeds from the Ro-tap test gradually decreased from 1.5% to 0.4% for red clover and 1.4% to 0.5% for perennial ryegrass (Table 2) as the proportion of soy flour in the seed coating increased. Interestingly, the mechanical properties of each seed formulation were not significantly different, even though the shape and surface properties of the seeds of both species differed.

There was no significant difference in both crops in terms of weight loss and disintegration time of coated seeds in water (Table 2) of soy flour and vermicompost. For red clover, when seed coatings with soy flour and vermicompost (30:70) (SF:MVC-2, SF:MVC-3 and SF:DE:CVE) were compared with SF:DE 30:70, the force (N) to break the seed coating increased by 0.6, 0.4, and 1.3 N, respectively and time to break (TB) the seed coating were non-significant, indicating that soy flour could serve as a binder for both types of materials used in this study (DE and vermicompost). There was no significant difference observed on force and TB data among treatments for perennial ryegrass (Table 2). Relaxation time after breaking red clover seed coating (RT) of SF:MVC-2, SF:MVC-3, and SF:DE:CVE increased slightly by 0.08, 0.02, and 0.01s, and the RT of perennial ryegrass increased by 0.04, 0.02, and 0.11s compared with that of SF:DE 30:70. In conclusion, the surface mechanical strength of different seed coating formulations with soy flour and vermicompost blends were non-significant but slightly higher than that of soy flour and diatomaceous earth.

The hydration test measures the time required to dissolve the coating materials. The proportion of soy flour in the seed coating blends had a significant effect on disintegration time (DT) (Table 2). A higher proportion of soy flour in the coating blend for red clover increased the DT from 58 to 103 min. This pattern was also observed for perennial ryegrass, as the proportion of soy flour ratio increased from 30% to 50%, the DT significantly increased from 40 to 90 min (Table 2). There were no significant differences in WL or DT for micronized vermicompost and compost extract (SF:MVC-2, SF:MVC-3 and SF:DE:CVE) seed treatments on either crop species (Table 2).

According to the American Seed Trade Association (ASTA), the key to a successful seed treatment is high physical integrity with low dust production. Determination of mechanical integrity of coated seeds is an important step in order to meet environmental safety standards [36]. Amirkhani et al. [15,27] tested the mechanical properties of several different broccoli coated seeds with Ro-tap and Texture Analyzer methods. The weight loss percentage that Amirkhani et al. [27] reported were slightly higher than the data collected from the red clover and perennial ryegrass seed treatments evaluated in this study for the same seed coating formulations. Total peak load force to break the seed coats for broccoli seeds were slightly lower than the value that was observed in this study for the SF:DE combination. Overall, the mechanical integrity of the coated red clover and perennial ryegrass seeds were more stable for the same seed coating formulation treatments, compared to broccoli. This difference might be because of the size and shape of clover and perennial ryegrass seeds contrasted to the broccoli seeds. Accinelli et al. [37] also attributed differences in seed dust emission for seed coating treatments of maize (*Zea mays* L.) and canola (*Brassica napus* L.) to seed physical characteristics. The mechanical integrity data observed for the different formulations in this study are in accordance with European Standards (Italy and France) and meet the benchmarks for safety of dust production of coated seeds [38].

#### *3.2. Germination and Seedling Growth of Soy Flour and Diatomaceous Earth Seed Coating*

In seed coatings with diatomaceous earth, soy flour served as the biostimulant component of the formulations. The results presented in Table 3 show that all coated treatments of red clover seeds significantly improved T50 and GU with no reduction in Gmax % compared to the non-treated control (Table 3), except for the SF:DE 60:40 treatment. Although soy flour proportions higher than 40% resulted in stronger and more durable seed coating mechanical properties, it had a negative effect on germination parameters (Gmax % and T50). For example, seeds treated with 30% and 40% soy flour had 98 and 99% Gmax % and T50 of 27 and 29 h, respectively, However, increasing the soy flour to

60% resulted in maximum germination of 96% and delayed the T50 to 34 h (Table 3). The negative effect in germination was attributed to the hard mechanical barrier of the seed coating with high soy flour content.

In contrast, for perennial ryegrass seeds, the Gmax % of all coated seeds was slightly reduced and showed delayed germination rates compared with the non-treated control. Control seeds of perennial ryegrass had the greatest Gmax % (85%) and significantly faster T50 (75 h) compared to all coating formulations. Application of 50% soy flour to the seed coating (SF:DE 50:50) reduced the Gmax % to 80% and T50 by 8 h and GU, compared with the non-treated control seeds. Due to the high percentage of coating build up (70%), the delay and a slight reduction in Gmax % was not unexpected. Several studies have indicated that the seed coating can act as a mechanical barrier for water absorbance and radical emergence [10,15].

Shoot and root length and seedling vigor index (SVI) are important indicators that determine whether the treated seeds promote seedling growth. Shoot and root length and seedling vigor index of treated seeds were significantly higher than those of non-treated control seeds for both crops (Table 3). The lowest application of soy flour (SF:DE 30:70) to the red clover seeds resulted in 4.1 cm shoot length, 3.0 cm root length, and 7.0 SVI, respectively, which were 14, 25, and 23 % higher than those of non-treated seeds. The same application rate of soy flour (SF:DE 30:70) to perennial ryegrass seeds improved the shoot growth by 17% and increased both root length and SVI values by approximately 13% compared to the control seeds. Amirkhani et al. [15,25–27] reported similar results for seed coatings that combined soy flour with diatomaceous earth. In their research, the seed coating blends had significant and positive effects on the above and below ground growth parameters of broccoli, tomato, radish, and hemp. They hypothesized that since soy flour is a plant-based protein and a rich source of several amino acids, it may have led to the increase in plant shoot and root growth and dry matter content and influenced uptake of nitrogen.

#### *3.3. Germination and Seedling Growth of Soy Flour and Vermicompost Seed Coating*

In addition to soy flour and diatomaceous earth, co-application of soy flour and vermicompost as rich sources of nutritional materials were tested as seed coatings and their effect on germination and seedling growth were recorded for red clover and perennial ryegrass. Shoot and root length, dry weight, and seedling vigor index of seedlings of all coated seed treatments were significantly higher compared to the non-treated controls (Table 4).

All treated red clover seeds germinated significantly faster (approximately 10 h) and had higher Gmax % (Table 4 and Figure 4) than non-coated seeds. They also germinated more uniformly than the non-treated control seeds. Gmax % was ≥ 98% for all treated seeds, which was significantly higher than control with 94% Gmax %. Red clover data showed that the shoot and root length and seedling vigor index of treated seeds were significantly higher than the non-treated control seeds. For example, compared with non-treated seeds control, the shoot length of SF:DE, SF:MVC-2, SF:MVC-3, and SF:DE:CVE increased by 14%, 19%, 27%, and 22%, respectively. Moreover, the root length of treated seeds increased by 12%, 16%, 20%, and 28%, respectively. All treatments showed a 40 to 60% increase in seedling dry weight (DW) compared with the control. The seedling vigor indexes (SVI) were 15%, 22%, 27%, and 25% higher than control, respectively (Table 4).

**Figure 4.** Cumulative germination percentage of red clover non-treated control seeds versus biostimulant coated seeds. \* Significant at *p* ≤ 0.05.

For perennial ryegrass, application of soy flour (SF:DE) and co-application with vermicomposts (SF:MVC-2 and 3, and SF:DE:CVE) increased shoot length by 22%, 25%, 28%, and 29% and root length by 10%, 15%, 12%, and 12%, respectively, compared to the non-treated control. The highest root length was observed in the SF:MVC-2 treatment. All treatments had higher DW than the control. Additionally, the highest SVI was observed in the SF:DE:CVE treatment, which was approximately 40% higher than the SVI of non-treated control (Table 4).

Statistical analysis (Pearson's correlation) of seed coating formulations and germination showed significant negative correlations between seed coating WL (%) and DT (min) (r = −0.99 \*\*\*). There was also a significant positive correlation between DT (min) and force (N) (r = +0.92 \*\*). A significant negative correlation between WL (%) and force (N) (r = −0.94 \*\*) from SF:DE coating formulations of red clover evaluated was observed (Table 5). For perennial ryegrass, the significant correlation coefficient between WL (%) and DT (min) was r = −0.99 \*\*\* and the correlation between WL (%) and force (N) was r = −0.96 \*\*\*. Lastly, for perennial ryegrass, the correlation between DT (min) and force (N) was r = +0.99 \*\*\* (Table 5). These data indicate that a higher proportion of SF in the seed coating formulation resulted in harder coatings but had only a slight impact on the Gmax %. For red clover, increasing the soy flour from 30% to 60% in seed coating formula reduced the Gmax % by 2%; however, T50 was significantly delayed by 7 h (Table 3). Similarly, for perennial ryegrass, increasing soy flour from 30% to 50% in the seed coating resulted in a 3% reduction in Gmax % and a minor delay on T50 (4 h).

In the present study, the seedling growth data for both cover crops evaluated indicate that seed coating can be an efficient and effective delivery method for application of nutritional biostimulant materials at the time of sowing for rangeland and grassland restoration. Several previous studies showed that applications of plant-based proteins and vermicompost can improve biometric growth parameters, related to production and yield of horticultural, field, and cover crops. Karlsons et al. [39] showed that a 10% addition of vermicompost in pure sand significantly increased fresh and dry weight of winter rye shoots by 578% and 265%, respectively. Tognetti et al. [40] found that application of vermicompost to degraded volcanic soil (to the extent of 20 and 40 g/kg soil) sown with ryegrass (*L. perene*) significantly increased ryegrass yields compared to the control due to the large nutrient concentrations and high microbial populations, when mixed with the soil. The positive effect of vermicompost on plant growth in this study agrees with the results of Alwaneen [41] on alfalfa and Amirkhani et al. [27] on broccoli. Amirkhani et al. [27] found that dairy manure-based vermicompost can supply essential nutrients to plants to enhance growth as well as increase the organic matter contents of soil for higher crop production. Moreover, in the recent decade, several researchers have been working on treating plants with biostimulants to stimulate crop productivity and increase stress tolerance under dynamic abiotic stresses [42–44]. The cover crop seeds treated with biostimulants in combination with other bio-effectors, such as superabsorbent polymers to investigate the response of these plants to drought, can be an area of future studies.

#### **4. Conclusions**

Seed coating technology can be an effective strategy to maximize early stand establishment of cover crops. Biostimulants applied as seed coatings have the potential to effectively promote seedling growth, and early stand establishment of red clover and perennial ryegrass. In this study, biostimulant seed coatings promoted the seedling growth of red clover and perennial ryegrass seeds and accelerated the germination of red clover compared to the non-treated control seeds. More rapid germination could aid in establishment under arid conditions and in areas with poor soils. However, further studies are needed to determine if vermicompost and plant-based proteins can be developed for economical commercial applications as seed treatments. The seed industry commonly includes fertilizers and *Rhizobium*, nitrogen fixing bacterial inoculants for red clover seeding. Additional research will be needed to determine if the biostimulant materials used in these experiments are compatible with seed inoculants. The use of biostimulants in combination with vermicompost and other biofertilizers as seed coatings may offer a great opportunity to increase plant production and long-term sustainability in agricultural landscapes.

**Author Contributions:** Y.Q., M.A., and H.M. contributed equally in writing the manuscript. Z.C. was involved in reviewing the manuscript. A.G.T. assisted with conceptualization of the overall experiments, funding, writing, and reviewing the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This material is based upon work that is supported by the National Institute of Food & Agriculture, US Department of Agriculture, Multi-state Project W-3168, under accession #1007938. Y.Q. was sponsored by China Scholarship Council.

**Acknowledgments:** The authors appreciate the technical assistance of Michael Loos in laboratory experiments and Zhen Wang for assistance with this project.

**Conflicts of Interest:** The authors declare that the research was conducted without any financial relationships that could be construed as a potential 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* **E**ff**ects of Di**ff**erent Microbial Inocula on Tomato Tolerance to Water Deficit**

**Giuseppe Mannino 1,**†**, Luca Nerva 2,3,**†**, Takoua Gritli 4, Mara Novero 1, Valentina Fiorilli 1, Mnasri Bacem 4, Cinzia Margherita Bertea 1, Erica Lumini 3, Walter Chitarra 2,3 and Ra**ff**aella Balestrini 3,\***


Received: 28 December 2019; Accepted: 22 January 2020; Published: 24 January 2020

**Abstract:** Several recent reports have highlighted some of the mechanisms involved in the enhanced tolerance to abiotic stresses induced by root-associated microorganisms, although additional efforts are still required to exploit and optimize these strategies. Particularly, arbuscular mycorrhizal fungi (AMF) play an important role as "bio-fertilizing microorganisms", establishing mutualistic symbioses with the roots of most crops. In this work, different microbial inocula (a single AMF species, a combination of three different AMF species, a combination of two plant growth-promoting bacteria (PGPB) strains and a more complex commercial inoculum) have been used to inoculate tomato plants (cv San Marzano nano), in order to verify their effects on the tolerance to a water deficit condition in pots, through the evaluation of biochemical stress markers and hormonal profiles (ABA and IAA). Results showed differences among tomato responses to water limitation depending on microbial inocula, confirming the importance to characterize the optimal plant/microorganism genotype combination(s) to maximize plant performance and tolerance. These findings open new perspectives for a better exploitation of these microorganisms.

**Keywords:** tomato; AM fungi; PGPB; water deficit; biostimulant

#### **1. Introduction**

Environmental stresses are becoming a serious threat and productivity is declining at an unprecedented level [1]. Water deficit, salinity, extreme temperatures, flooding, nutritional limitations, pest and pathogen attacks are key threats to plant growth and crop productivity and constitute major constraints to actual agriculture [2]. The extent of agricultural soil affected by water stress and exposed to a loss of fertility is predicted to progressively increase due to climate change [3]. Conventional agriculture's dependence on chemical fertilizers and pesticides has encouraged the thriving of industries producing these products that are not only hazardous for human consumption but also exert negative effects on the environment [1]. Biofertilizers, and biostimulants, could help to solve the problem of feeding an increasing global population at a time where agriculture faces several environmental stresses [1]. A number of recent reports have highlighted some of

the mechanisms involved in the enhanced tolerance to abiotic stresses induced by root-associated microorganisms [4–6]; however, additional efforts are still required to exploit the useful aspects of the different root-associated microorganisms and to optimize these strategies, supporting their application to current agricultural practices [1]. Arbuscular mycorrhizal (AM) fungi play an important role as "bio-fertilizing microorganisms" as they establish mutualistic symbioses with the roots of most crops [7,8]. These symbiotic fungi colonize plant roots and enhance the uptake of water and nutrients by the host plants, while, they receive the carbon compounds. These fungi are considered essential elements for plant nutrition, mainly in low-nutrient conditions, as their hyphae can extend for many meters in the ground, helping the plants to acquire mineral soil nutrients. Since AM fungi play an instrumental role in protecting the plants against abiotic stresses such as nutrient deficiency, extreme temperatures [9] and drought [10–14], they can benefit their hosts in both wild and agricultural environments [15]. Consequently, AM fungi are thought to have a great impact in natural environments [16–18], as in managed conditions in agriculture, horticulture and forestry [8]. Although there is no hard symbiont specificity in AM interactions, the efficiency of these associations depends on the interacting partner genotypes and the environmental conditions [19,20]. Recent findings suggest a certain degree of functional specialization in AM interactions [8]. Some plant/fungus genotype combinations are more efficient than others in terms of nutrition or stress tolerance/resistance [13,14,20–25]. Despite the low host specificity of AM under controlled conditions, the presence of several symbionts might result in the most effective mutualistic combination [8]. Berruti and colleagues [26] demonstrated that the AM fungal communities originating from cells containing the arbuscules, which represent the functional structures in AM symbiosis, and the whole root samples of *Camellia* plants (grown in natural soil) differed remarkably. These results suggested that not all the AM fungal isolates present in soil could form a functional symbiosis. Symbiotic fungi, however, are only part of the soil and root-associated microbiota. Plant growth-promoting bacteria (PGPB), genera like *Bacillus*, *Azospirillum* or *Pseudomonas*, also exert beneficial effects on plant metabolism and primes tolerance mechanisms against biotic and abiotic stresses [6]. Interestingly, it was recently demonstrated that grapevine roots differently respond to a pure AM inoculum with respect to a mixed inoculum containing different microbial isolates/strains [27]. Here, we have used different microbial inocula on the commercial tomato cv San Marzano nano to verify the impact on the tolerance to a water deficit condition. One of the most important challenges in this research area is to dissect the actual mechanism of mode of action for different strains/isolates to evaluate their efficacy, alone or in combination, towards its use in sustainable agriculture.

#### **2. Materials and Methods**

#### *2.1. Inoculation of Tomato Plants and Growth Conditions*

Tomato (*Solanum lycopersicum* 'San Marzano nano') seeds were surface sterilized in sodium hypochlorite for 20 min, washed five times in sterile water, and germinated on wet filter paper. For this pot experiment, pots (10 cm × 10 cm × 12 cm) with a volume of 1 L containing substrate (sterilized quartz sand) were arranged on a growth chamber in a randomized block design including five treatments: (i) non-inoculated control (CTRL); (ii) AM fungi mono fungal inoculum (Myc\_Rhizo); (iii) AM fungi multi fungal inoculum (MULTISTRAIN); (iv) PGPB (LC3.5 + 5.2) and (v) a mixed commercial inoculum containing both bacteria and fungi (Commercial MIX). Each treatment was replicated ten times (10 pots), each pot contained one seedling.

Tomatoes were inoculated with AM fungi at planting time by placing the inoculum in the planting hole and in contact with the roots, as follows: 10 plants were inoculated with 15 g/pot of a mono fungal inoculum (Myc\_Rhizo) based on *Rhizoglomus irregulare* BEG140 and 10 plants with 15g/pot of multi fungal inoculum (MULTISTRAIN) constituted by *Claroideoglomus claroideum* BEG96, *Funneliformis caledonium* BEG97 and *F. geosporum* BEG199; both the pure AM fungi inocula were provided by Symbiom Ltd., (Lanškroun, Czech Republic). The commercial MIX Opera-Rizon (MsBiotech, Larino, Italy) was used to inoculate other 10 tomato plants. This formulate, as reported in the product label, consists of AM fungal species (*Glomus* spp. 0.001%) and rhizospheric bacteria (1 <sup>×</sup> 10<sup>2</sup> Colony Forming Unit CFU). In addition, two PGPB (LC3.5 and LC5.2) were selected based on their high levels production of auxin: 46 μg/mL and 24 μg/mL for LC3.5 and LC5.2 respectively (Gritli and Bacem unpublished results). Both bacterial strains were isolated from roots of *Lathyrus cicera* in the northern of Tunisia. They were prepared as following: pure colonies of PGPB strains were multiplied separately in Luria Bertani (LB) broth by incubating them in a shaker for 72 and 24 h respectively at 27 ◦C. The optical density was adjusted to 1 (at 660 nm for PGPR. One mL of the bacterial suspension (10<sup>9</sup> CFU/mL) of the two PGPRs (LC3.5 + LC5.2) was inoculated to 10 pots, where 15 g/pot of the carrier material (without AM fungi) was applied.

Ten plants were left as non-inoculated control plants: in these pots 15 g of carrier material (without AM fungi) was applied. Plants were grown in controlled conditions, with a temperature of 23 ◦C/21 ◦C day/night, 16/8-h light/dark photoperiod, and relative humidity 65%. From transplanting to the beginning of the water deficit experiment (after about 6–7 weeks), all the plants were watered twice per week with tap water and, once per week, with a modified Long-Ashton nutrient solution [28] containing 3.2 μM inorganic phosphate.

Out of 50 plants, 25 (five plants for each treatment) were used as controls (irrigated or non-stressed, NS) and maintained in a well-watered state (at container capacity). The remaining 25 plants (five plants for each treatment) were subjected to a water stress (WS) treatment. To this aim, about 6 weeks after fungal inoculation irrigation was withheld and the experiment was stopped when the first plants reached a stress level (measured by infrared gas analyzer, IRGA, ADC-LCPro+ system; Analytical Development Company Ltd, Hoddesdon, UK).

#### *2.2. Miniprep Bacterial Strains DNA Isolation and 16S SSU rRNA Amplification*

The two PGPB strains were subjected to molecular characterization by means of amplification with conventional PCR using DNAs isolated from bacterial strains LC3.5 and LC5.2 as a template. An almost complete small subunit (SSU) bacterial ribosomal RNA gene (16S) was amplified with bacterial universal primers 27F-1492R that amplified a fragment of about 1465 bp [29]. The PCR were carried out in a final volume of 25 μL containing 10 μL of Platinum Hot Start PCR Master Mix (2X), 0.5 μL of each primer (10 μM), template DNA (1 μL) and 13 μL of PCR-grade water. Bacterial PCR amplification was performed using a T3000 thermal cycler (Biometra, Göttingen, Germany) with the following profile: initial denaturation for 5 min at 95 ◦C; 35 cycles of denaturation (60 s at 94 ◦C), annealing (60 s at 58 ◦C) and extension (60 s at 72 ◦C) and a further 7 min at 72 ◦C. All the PCR products were checked using 1.5% (*w*/*v*) agarose gel stained with ethidium bromide (Merck KGaA, Darmstadt, Germany). The two PCR products replicates for each strain were pooled and purified using Wizard SV Gel and a PCR Clean-Up System kit (Promega, Madison, WI, USA). Purified PCR products were sequenced, using either the universal primer 27F or 1492R, by LMU sequencing services (Munich, Germany). The two sequences were deposited at NCBI (accession # MN879506 and MN879507 for LC3.5 and LC5.2, respectively).

#### *2.3. Eco-Physiological Parameters*

Measurements of transpiration rate (E), stomatal conductance (gs) and net photosynthetic rate (AN) were performed on adult, non-senescing leaves at the same physiological age (in the middle part of the plant, considering the third to fourth leaf from the shoot apex). Intrinsic water use efficiency (iWUE) was calculated as the ratio between AN and gs. Measurements were taken with an IRGA instrument. During measurements, light intensity in the leaf chamber was set at 1200 μ mol photons m−<sup>2</sup> s−1, temperature was 25 ◦C, and the concentration of CO2 was maintained between 450 and 470 ppm. Measurements were taken between 10:00. and 13:00. The chlorophyll content index (CCI) was determined at the end of the experiment (about 9 DAT) using the portable chlorophyll meter SPAD 502 (CCM-200; Opti-Sciences, Hudson, NH, USA).

#### *2.4. Assessment of Symbiosis Development*

At the end of the experiments, plants were harvested, and plant height and fresh weight (not shown) were recorded. A part of the root apparatus of at least three plants (up to five depending on the treatment) was stained with 0.1% Cotton Blue in lactic acid. For each plant, sixty randomly chosen 1-cm-long root segments were mounted on slides and fungal colonization was quantified according with the Trouvelot system [30] using MYCOCALC software, while the remaining root systems were stored at −80 ◦C until further analyses.

#### *2.5. Preparation of Extracts and Biochemical Parameter Evaluation*

After the measurement of plant morphological parameters, leaf and root samples were dried by lyophilization, then, grounded and homogenized using a mortar and pestle. About 30 mg of the obtained dried powders were extracted with 90% (*v*/*v*) methanol using a 1:50 (*w*/*v*) ratio. Samples were mixed by vortexing for 5 min, and sonicated for 15 min at 8 ◦C. Following a centrifugation step (10 min at 8000*g*, 4 ◦C), the supernatants were filtered using a filter tips, and directly used for chemical determinations.

#### 2.5.1. Determination of Total Chlorophyll Content (TCC)

The leaf extracts were employed for the determination of the total chlorophyll content (TCC), according to Lichtenthaler and Buschmann [31]. Briefly, 1 mL of appropriate diluted sample was subjected to spectrophotometric measurements at 665.2 nm and 652.4 nm. TCC, expressed as μg per g of dry weight (d.wt), was calculated for each sample using the following equation:

$$TCC = \frac{\left[\left(16.82 \ge Abs \text{\ $s\$ } 2 - 9.28 \ge Abs \text{\ $s\$ }\_{652.4}\right) + \left(36.92 \ge Abs \text{\ $s\$ }\_{652.4} - 16.54 \ge Abs \text{\ $s\$ }\_{665.2}\right)\right] \ge V\_{\text{extr}} \ge DF}{WH} . \tag{1}$$

*Vextr* = volume, expressed as mL, used for the extraction process; *DF* = dilution factor and *WH* = weight of each sample expressed in grams.

#### 2.5.2. Determination of Proline Concentration (TpC)

The proline concentration (TpC) was determined according to Carillo and Gibon [32]. Briefly, 500 μL of undiluted samples were incubated with 1 mL of the reaction mix containing 1% (*w*/*v*) ninhydrin solubilized in 60% (*v*/*v*) acetic acid and 20% (*v*/*v*) ethanol. The mixture was incubated at 95 ◦C for 20 min in the dark, and then centrifuged at 10.000 rpm for 1 min at room temperature in a table microfuge. The absorbance was then measured at 520 nm. Quantification was performed using an external calibration curve prepared using a pure standard of proline, whose concentration ranged from 0.01 to 1.00 mmol.

#### 2.5.3. Determination of Total Polyphenol Content (TPC)

The total polyphenol content (TPC) was evaluated both in leaf and root extracts following the method of Ainsworth and Gillespie [33]. Briefly, each sample was appropriately diluted in 90% (*v*/*v*) methanol and then 930 μL were incubated with 30 μL of Folin–Ciocalteu reagent and 40 μL of 20% (*w*/*v*) sodium carbonate (Na2CO3). The samples were then incubated for 1 min at 80 ◦C and for 20 min at room temperature in the dark. Then the absorbance was monitored at 725 nm. An external calibration curve using gallic acid (GA), ranged between 50 and 400 mmol, was employed to quantify TPC in the samples. The results were expressed as μmol of gallic acid equivalent (GAE) per g of dry weight (d.wt).

#### *2.6. Determination of Abscisic Acid (ABA) and Indole-3-Acetic Acid (IAA) Content*

About 500 mg of homogenized leaf and root samples were freeze-dried and transferred with 1 mL of methanol:water (8:2 *v*/*v*) acidified with 0.1% (*v*/*v*) of acetic acid in an ultrasonic bath for 1 h. Samples were centrifuged for 10 min at 4 ◦C and 15,000 rpm, and the supernatant was analyzed by

high-performance liquid chromatography (HPLC, Agilent, Waldbronn, Germany). Original standards of abscisic acid (ABA, purity ≥ 98.5%, Merck KGaA, Darmstadt, Germany) and indole acetic acid (IAA, purity ≥ 99%, Merck KGaA, Darmstadt, Germany) were used for the identification by comparing retention time and UV spectra. The quantification was made by external calibration method. The HPLC apparatus was an Agilent 1220 Infinity LC system (Agilent R, Waldbronn, Germany) model G4290B equipped with gradient pump, auto-sampler and column oven set at 30 ◦C. A 170 Diode Array Detector (Gilson, Middleton, WI, USA) set at 265 nm was used as detector. A Nucleodur C18 analytical column (250 mm × 4.6 mm i.d., 5 μm, Macherey Nagel) was used. The mobile phases consisted in water acidified with formic acid 0.1% (A) and acetonitrile (B), at a flow rate of 0.500 mL min−<sup>1</sup> in gradient mode, 0–6 min: 30% of B, 6–16 min: from 30% to 100% B and 16–21 min: 100% B; 20 μL was injected for each sample.

#### *2.7. Statistical Analysis*

All measurements are the average of five different biological replicates. Each biological replicate was analyzed three times in each experiment. The content of chlorophylls (TCC), proline (TpC), polyphenols (TPC), ABA and IAA were reported both as relative content (Figures 1–6) and as absolute content (Supplementary Tables S1–S4). The relative content was calculated comparing the content observed in inoculated plants (treated with MULTISTRAIN, Myc\_Rhizo, LC3.5 + 5.2 or the Commercial MIX) with the content of unstressed and/or untreated plants (control plants). In both cases, data are expressed as mean values ± standard deviation (SD). Significant differences were evaluated by performing one-way ANOVA followed by Tukey's HSD test (*p* ≤ 0.05) or *t*-test (*p* ≤ 0.05) using SPSS ver. 24 software.

#### **3. Results and Discussion**

The effect of several microbial inocula on tomato tolerance to a water deficit condition was verified. The beneficial effects of root-associated microbes (i.e., AM fungi and PGPB) on plant growth and performance under water limitation have already been reported for several plant species [5,34], including the tomato genotype considered in the present research [13,14,35]. These previous works already showed a different plant response to a water deficit condition depending on the AM fungal species associated to plant roots. An untargeted metabolomic analysis in tomato roots colonized by three AM fungi of different genera showed that some responses to drought and salt stress were common to all AM fungi tested, while others were specifically related to single isolates [25]. Here, several microbial inocula (a single AM fungal species, a combination of three different AM fungal species, a combination of two PGPB strains and a more complex commercial inoculum) were tested for the effect on tomato tolerance to water limitation. Both the bacterial strains used in this work (LC3.5 and LC5.2) showed sequence identity with *Bacillus* spp. In detail, a sequence identity with *Bacillus subtilis* (first hit: MN704441.1, query cover 100%, *e*-value 0.0, identity 99.78) and *B. megaterium* (first hit: MK791705.1, query cover 100%, *e*-value 0.0, identity 98.64%) was found for LC3.5 and LC5.2, respectively. The AM fungal colonization using several formulates was also evaluated, showing some relevant differences among the two AM fungal inocula (Myc\_Rhizo and MULTISTRAIN), while the presence of AM fungal structures were not observed in the roots inoculated with the mixed inoculum (Commercial MIX; Table 1). A very low colonization by AM fungi was already observed using a commercial mixed inoculum on grapevine rootstocks [27]. In detail, a different grapevine root transcriptome profile was observed after inoculation with a pure AM inoculum (*Funneliformis mosseae*) and the mixed one, although this last elicited an important transcriptional regulation probably due to the predominantly presence of PGPB.

**Table 1.** AM fungal colonization using three different inocula. F%, Frequency of mycorrhization in root system; M%, Intensity of mycorrhizal colonization in the root system; a%, arbuscule abundance in mycorrhizal parts of root fragments; A%, arbuscule abundance in the whole root system. Values are expressed as a mean ± SD (*n* = 3). Data were subjected to statistical analysis using SYSTAT 10 software, applying the nonparametric Kruskal-Wallis test adopting a probability level of *p* < 0.05. Data followed by different superscript letters indicate significant statistical differences among samples.


#### *3.1. Impact of Treatments and Water Stress on Eco-Physiological Parameters*

Eco-physiological parameters were recorded at the end of the experiment, considering both gas exchanges and CCI (Table 2 and Table S1). In general, assimilation rates (A) decreased under water deficit condition with the lowest values in the treatment with bacteria (LC3.5 + 5.2 and commercial mix), while no effect was observed on *e*-values. Brilli et al. [34] already found that the tomato inoculation with a PGPR (*Pseudomonas chlororaphis*) did not affect the physiological parameters. In addition, the highest "A" values under water stress for the AM-inoculated plants are in agreement with the data from Chitarra et al. [13] and Volpe et al. [14] on the same tomato genotype, although a difference between two AM fungal species was observed. As expected, stomatal conductance (gs) decreased under water limitation mainly in inoculated plants (Figure 2). This is not surprising since a different timing in reaching a stress level has been already reported from AM-colonized and non-colonized tomato plants [13]. Similarly, although not statistically significantly different, a decrease in gs was observed in the presence of bacteria inoculation [34]. On the contrary, an increasing trend in iWUE was observed in WS inoculated plants, in agreement with previous works [13,14,34]. Regarding CCI, a general decrease was observed in WS plants. Interestingly, a different impact of the several inocula was observed (e.g., MULTISTRAIN vs. Myc\_Rhizo) both in NS and WS plants, confirming species-specificity in affecting physiological traits. Concerning plant height, no significant results have been obtained among treatments and stress conditions. Taken together our results highlighted that symbiotic fungi (i.e., AM fungi) differently affect plant traits important for the tolerance to stressful conditions with respect to root-associated bacteria.



#### *3.2. E*ff*ect of Water Deficit on Total Chlorophyll Content (TCC), Total Proline Content (TpC) and Total Polyphenol Content (TPC)*

Water stress implicates morphological, biochemical and molecular changes [36], and may affect plant growth during different developmental stages [37]. As a first point, we evaluated how a water stress (WS) condition could change some biochemical parameters of tomato plants grown in the absence of specific treatments. The biochemical profile of stressed plants was compared to that of unstressed ones grown in well-watered (WW) conditions (Figure 1; Tables S2 and S3). In our experimental conditions, the exposure of tomato plants to WS negatively influenced the TCC and TPC, while the TpC was positively affected. The major effect of drought in plants is correlated with the decrease of photosynthetic processes, leading not only to the reduction of leaf expansion, but also to fruit production [38]. The main reason for a decrease in photosynthesis is due to changes in photosynthetic pigment levels that are part of the photosynthetic apparatus [39]. The decrease in TCC (0.67 ± 0.08), observed here, was already reported in tomato plants subjected to WS [40,41]. Other abiotic stresses such as heat [42] and salt stress [43,44] also affected the amount of these molecules.

**Figure 1.** Effects of water-stress (WS) on the tomato total content of chlorophylls (TCC), proline (TpC) and polyphenols (TPC). Data for each quantification are expressed as relative content, comparing the measurements obtained for water-stressed plants with those of unstressed plants. The dotted line indicates the basal expression of NS. Absolute quantification of each parameter is also reported in Supplementary Table S2. Bars with different lowercase letters indicate significantly different values at *p* ≤ 0.05 as measured by a one way-ANOVA followed by a Tukey's HSD post hoc test (see Supplementary Table S4 for additional information). The symbol "\*" indicates significant differences (*p* ≤ 0.05) between untreated-water stressed and untreated-non stressed plants, as measured by a *t*-test.

On the other hand, the role of polyphenols, which represent the soluble antioxidant defenses of the plant, in stressed samples has been widely discussed and contrasting. Although several studies already reported the increase of polyphenols to contrast the oxidative damage generated after the exposure to different abiotic stresses [45], in other cases a substantial decrease of these molecules was observed [46,47]. This response is probably due to the loss of the plant capability to synthesize *ex novo* the soluble antioxidant defenses. In our experiment, the TPC, evaluated on both leaves and roots, decreased after the exposure to WS. Moreover, a stronger effect was observed in leaves compared to roots (0.72 ± 0.06 and 0.49 ± 0.04, respectively).

Finally, in order to respond to unbalanced water repartition, plants generally accumulate compatible solutes with the aim to raise osmotic pressure and thereby to maintain both turgor and driving gradient for water uptake [32,48]. Among these solutes, proline plays a key role in these processes. The accumulation of proline in leaves can be considered as a strong indicator of abiotic stresses such as drought, salt and heath stresses. In accordance with our results (1.45 ± 0.09), an increase of proline in leaves of stressed plants was previously reported, not only in tomato but also in other plants [32,48–51].

#### *3.3. MULTISTRAIN, Myc\_Rhizo, LC3.5* + *5.2 and the Commercial MIX are Able to Recover the Biochemical Parameters in Water Stressed Plants*

In order to check if the treatments with the different microbial inocula were able to restore the correct plant homeostasis, tomato plants were inoculated with four different inocula. To allow the successfully establishing of the relationship between roots and employed microorganisms, plants were grown in well-watered conditions for a period of about six weeks before to start with water limitation. Figure 2 shows the change in TCC, TpC and TPC values of treated-WS plants compared to untreated-WS plants (dotted-line). All the treatments promoted a recovery of the TCC and TPC amount in WS plants, suggesting beneficial properties of the formulations, and a decrease of water stress in treated plants. However, significant differences (*p* ≤ 0.05) among the four treatments were found. In particular, Myc\_Rhizo (Figure 2B) was the most effective in increasing TCC in leaves (1.86 ± 0.06), followed by the commercial mix (1.37 ± 0.05; Figure 2D). The highest recovery in term of TPC in the leaves was recorded in WS plants treated with Myc\_Rhizo (3.04 ± 0.05; Figure 2B), while the best recovery of TPC in roots was observed with MULTISTRAIN (1.62 ± 0.08; Figure 2A) followed by Myc\_Rhizo (1.44 ± 0.04; Figure 2B).

Finally, TpC was also affected by the different treatments, with Myc\_Rhizo and MULTISTRAIN (Figure 2A,B) that showed again the highest decrease (0.42 ± 0.05 and 0.65 ± 0.03, respectively). Moreover, a very strong and negative correlation was found between TCC/TpC (ρ = −0.89) and leaf-TPC/TpC (ρ = −0.96), as revealed by Pearson analysis (Table S4). On the other hand, no correlation was found between root-TPC and TpC (ρ = −0.15).

**Figure 2.** Effect of the treatment with MULTISTRAIN (**A**), Myc\_Rhizo (**B**), LC3.5 + 5.2 (**C**) and the commercial mix (**D**) on the total content of chlorophylls (TCC), proline (TpC) and polyphenols (TPC) evaluated on water-stressed plants. Data for each quantification are expressed as relative content, comparing the measures obtained by treated-and untreated-water stress plants. The dotted line indicates the basal level of untreated water stressed plants. Absolute quantification of each parameter is also reported in Supplementary Table S2. Bars with different lowercase letters indicate significantly different values at *p* ≤ 0.05 as measured by a one way-ANOVA followed by a Tukey's HSD post hoc test (see Supplementary Table S4 for additional information). The symbol "\*" indicates significant differences (*p* ≤ 0.05) between treated-WS and untreated-WS plants as measured by a *t*-test.

#### *3.4. MULTISTRAIN, Myc\_Rhizo, LC3.5* + *5.2 and the Commercial MIX A*ff*ect Total Chlorophyll, Polyphenol and Proline Content in Absence of Stress*

In order to evaluate the performance of different formulations without a water stress condition, NS plants treated with MULTISTRAIN, Myc\_Rhizo, LC3.5 + 5.2 or of the commercial mix were analyzed. Figure 3 shows the relative content of treated-non stressed plants in comparison to untreated-non stressed plants (dotted-line). As a general trend, the treatments with MULTISTRAIN (Figure 3A) and LC3.5 + 5.2 (Figure 3C) did not change the content of the analyzed biochemical parameters, with the exception of TpC in non-stressed plants treated with LC3.5 + 5.2. A more evident effect was instead observed in non-stressed plants inoculated with Myc\_Rhizo or with the commercial mix (Figure 3B,D). In these cases, TpC and TPC in the leaves statistically (*p* < 0.05) increased with respect to untreated non-stressed plants (1.20 ± 0.18 and 1.20 ± 0.09 for Myc\_Rhizo and commercial mix, respectively). On the other hand, TPC decreased in roots (Figure 3B,D). The slightly significant changes in the biochemical parameters could be associated to the functional traits of the considered

microorganisms that led to a priming status, also in the absence of stress, as previously reported ([6] and reference therein). Physiological, transcriptional and metabolic changes stimulated by the colonization of soil root-associated microorganisms can prime plants for enhanced defense ahead of abiotic and biotic stress occurrence [52]. Evidence of a possible priming of the plant defensive system induced by AM-inoculation was recently suggested in *Arundo donax* [53], where a significant increase in proline accumulation in AM-colonized roots was reported. Brilli et al. [34] suggested that *Pseudomonas chlororaphis* acted as a 'priming stimulus' triggering in inoculated tomatoes enhanced tolerance to water stress. Interestingly, a simultaneous increase in the activity of superoxide dismutase (SOD) and catalase (CAT), and in proline accumulation was observed in tomato leaves from inoculated plants, independently by the stress level (well-watered or water stressed plants). However, the contribution of the root-associated microorganisms in plant adaptation to environmental stress factors needs to be still extensively evaluated, particularly in natural conditions, where a complex soil microbiota is present, and upon multiple stresses.

**Figure 3.** Effect of the treatment with MULTISTRAIN (**A**), Myc\_Rhizo (**B**), LC3.5 + 5.2 (**C**) and the commercial mix (**D**) on the total content of chlorophylls (TCC), proline (TpC) and polyphenols (TPC) evaluated on unstressed plants. Data for each quantification are expressed as relative content, comparing the measurements obtained by treated- and un-treated-non stressed plants. The dotted line indicates the basal level of untreated-non stressed plants. Absolute quantification of each parameter is also reported in Supplementary Table S2. Bars with different lowercase letters indicate significantly different values at *p* ≤ 0.05 as measured by a one way-ANOVA followed by a Tukey's HSD post hoc test (see Supplementary Table S4 for additional information). The symbol "\*" indicates significant differences (*p* ≤ 0.05) between treated-NS and untreated-NS plants, as measured by a *t*-test.

#### *3.5. E*ff*ects of the Di*ff*erent Formulations Applied on ABA and IAA Content*

Regardless of the water regime conditions the pattern of ABA and IAA content were strongly affected by the applied consortia. The plant hormone ABA is a chemical signal produced in leaves and roots, largely studied because of its pivotal roles in stomata movement and molecular-mediated responses under water stress [54]. In general, ABA content was less affected in roots of treated plants with respect to the controls in both WS and NS conditions (Figure 4).

**Figure 4.** Effects of water-stress (WS) on the tomato content of indole acetic acid (IAA) and abscisic acid (ABA) evaluated both on leaves and roots. Data for each quantification are expressed as relative content, comparing the measurements obtained for water-stressed plants with those of unstressed plants. The dotted line indicates the basal expression of non-stressed plants. Absolute quantification of each parameter is also reported in Supplementary Table S2. Bars with different lowercase letters indicate significantly different values at *p* ≤ 0.05 as measured by a one way-ANOVA followed by a Tukey's HSD post hoc test (see Supplementary Table S4 for additional information). The symbol "\*" indicates significant differences (*p* ≤ 0.05) between untreated-water stress and untreated-non stressed plants, as measured by a *t*-test.

Under the NS condition, the ABA content in roots was significantly higher (*p* ≤ 0.05) in the treated plants when compared to their controls, suggesting an ABA-primed status induced by the microorganisms added in the substrates (Figure 5). In NS leaves, MULTISTRAIN, Myc\_Rhizo and LC3.5 + 5.2 showed significantly higher levels of ABA with respect to the controls. As expected, under WS conditions, ABA content was generally higher with respect to NS and only in roots of Myc\_Rhizo and leaves of the commercial mix was significantly higher with respect to their controls (*p* ≤ 0.05), pointing out a microbial-mediated role in WS sensing and in turn ABA synthesis on inoculated plants (Figure 6) [4,34].

In almost all conditions tested, under NS conditions, IAA content showed an opposite trend for ABA) confirming their negative correlation as previously reported by Saeedipour and Moradi [55], with the exception of LC3.5 + 5.2. Interestingly, under WS conditions, all the treatments showed significantly higher IAA levels in leaves whilst lower levels were observed in roots with respect to their controls (*p* ≤ 0.05; Figure 6).

**Figure 5.** Effect of the treatment with MULTISTRAIN (**A**), Myc\_Rhizo (**B**), LC3.5 + 5.2 (**C**) and the commercial mix (**D**) on the content of indole acetic acid (IAA) and abscisic acid (ABA) evaluated both on leaves and roots of unstressed plants. Data for each quantification are expressed as relative content, comparing the measurements obtained by treated- and untreated-non stressed plants. The dotted line indicates the basal expression of untreated-non stressed plants. Absolute quantification of each parameter is also reported in Supplementary Table S2. Bars with different lowercase letters indicate significantly different values at *p* ≤ 0.05 as measured by a one way-ANOVA followed by a Tukey's HSD post hoc test (see Supplementary Table S4 for additional information). The symbol "\*" indicates significant differences (*p* ≤ 0.05) between treated- and untreated-non stressed plants, as measured by a *t*-test.

**Figure 6.** Effect of the treatment with MULTISTRAIN (**A**), Myc\_Rhizo (**B**), LC3.5 + 5.2 (**C**) and the commercial mix (**D**) on the content of indole acetic acid (IAA) and abscisic acid (ABA) evaluated both on leaves and roots of water-stressed plants. Data for each quantification are expressed as relative content, comparing the measurements obtained by treated- and untreated-water stressed plants. The dotted line indicates the basal expression of untreated water stressed plants. Absolute quantification of each parameter is also reported in Supplementary Table S2. Bars with different lowercase letters indicate significantly different values at *p* ≤ 0.05 as measured by a one way-ANOVA followed by a Tukey's HSD post hoc test (see Supplementary Table S4 for additional information). The symbol "\*" indicates significant differences (*p* ≤ 0.05) between treated- and untreated-non stressed plants, as measured by a *t*-test.

#### **4. Conclusions**

In conclusion, our results confirmed the fact that several microbial inocula have different impacts on the tomato's response under a water stress condition. Although aspects related to the persistence of the inocula at the end of the experiment were not considered, our results showed that the biochemical response of tomato to a stressful factor changed depending on the applied consortia of root-associated microorganisms. The latter were also able to induce a different effect on physiological traits. Moreover, the importance of symbiotic fungi, i.e., the AM fungi, in inducing a primed status and, in turn, a tolerance to water deficit was highlighted, reinforcing the consolidated evidence of the positive role played by these "biostimulants". However, many factors can affect the success of inoculation and persistence of inoculated microorganisms in soil, including compatibility with the target environment, the degree

of spatial competition with other soil organisms in the target niche and the timing of inoculation. For this reason, further efforts should be done, mainly for bacteria species, to implement methods for monitoring and characterizing the degree of root/rhizosphere colonization of the microbial inoculants.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4395/10/2/170/s1. Supplementary Table S1. Statistical analysis of absolute values of photosynthetic rate (A), transpiration rate (E), stomatal conductance (gs), intrinsic water use efficiency (iWUE), stem eight and CCI. Supplementary Table S2. Absolute determination of Total Content of Chlorophylls (TCC), Proline (TpC), Polyphenols (TPC), Indole Acetic Acid (IAA) and Abscisic Acid (ABA). Supplementary Table S3. Statistical analysis of absolute determination of Total Content of Chlorophylls (TCC), Proline (TpC), Polyphenols (TPC), Indole Acetic Acid (IAA) and Abscisic Acid (ABA). Table S4. Statistical analysis of the relative content of Total Content of Chlorophylls (TCC), Proline (TpC), Polyphenols (TPC), Indole Acetic Acid (IAA) and Abscisic Acid (ABA).

**Author Contributions:** All authors have read and agree to the published version of the manuscript. Conceptualization, R.B., E.L. and W.C.; investigation, G.M., L.N., T.G., C.M.B., E.L., M.N., W.C. and R.B.; formal analysis, G.M., L.N., W.C.; resources, T.G., M.B.; writing—original draft preparation, G.M., R.B., E.L., W.C.; writing—review and editing, G.M., L.N., C.M.B., M.N., V.F., W.C. and R.B.; supervision, R.B. All authors have read and agreed to the published version of the manuscript.

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

**Acknowledgments:** The authors thank Maria Teresa Della Beffa for the help in plant preparation and maintenance, and Miroslav Vosatka and Aleš Látr for providing the two AM fungal inocula produced by Symbiom.

**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/).
