Appraisal of Combined Applications of Trichoderma virens and a Biopolymer-Based Biostimulant on Lettuce Agronomical, Physiological, and Qualitative Properties under Variable N Regimes

The current research elucidated the agronomical, physiological, qualitative characteristics and mineral composition of lettuce (Lactuca sativa L. var. longifolia) after treatments with a beneficial fungus Trichoderma virens (TG41) alone or in combination with a vegetal biopolymer-based biostimulant (VBP; ‘Quik-link’). The experiment consisted of lettuce plants grown in three N conditions: sub-optimal (0N kg ha−1), optimal (70N kg ha−1), and supra-optimal (140N kg ha−1) N levels. Lettuce grown under 0N fertilization showed a significant increase in fresh yield when inoculated with TG41 alone (45%) and a greater increase with TG41 + VBP biostimulant (67%). At 48 days after transplanting, both the TG41 alone or TG41+VBP biostimulant induced higher values of CO2 assimilation in comparison to the control. The mineral concentrations in leaf tissues were greater by 10% for K and 12% for Mg with the TG41+VBP treatments compared to the untreated lettuce. The lettuce plants receiving either TG41 alone or TG41+VBP biostimulants had a significantly lower nitrate content than any of the untreated controls. In non-fertilized conditions, plants treated with TG41+VBP biostimulants produced lettuce of higher premium quality as indicated by the higher antioxidant activity, total ascorbic acid (+61%–91%), total phenols (+14%) and lower nitrate content when compared to the untreated lettuce.

of industries, but they have also been used as fertilizers in crops [45]. They have proven auxin and gibberellin-like activities, probably due to the biological action of phenol metabolites able to interact with plant phytohormones and enzymes affecting carbon-nitrogen metabolism [45]. Lucini et al. [46] indicated that when the vegetal-based biopolymer was applied as a drench to melon, it altered the plant hormone profile by inducing an increase in ABA intermediates, brassinosteroids, and cytokinins in a dose-dependent manner. This mechanism stimulated root growth and consequently resulted in a 'nutrient acquisition response' improving resources use efficiency (RUE), thus enhancing plant biomass production and resistance to transplant stress. In addition, the authors reported that brassinosteroids may play a key role both in root system architecture changes as well as in shoot interference with hormone signaling and secondary metabolites, such as phenolic acids and carotenoids, plus the modulation of photosynthesis.
Romaine lettuce requires varying levels of N during the 65-75-day production cycle that depends upon the plant growth and development stages, plus N availability in the rhizosphere. N availability affects the morphological and physiological plant attributes [47] that influence the marketability of the leafy produce (i.e., leaf size) and consumer perception (i.e., visual green color). From this perspective, depending on the farming conditions, growing season, and genotypes, the combined application of Trichoderma and vegetal-biopolymer biostimulants could be particularly useful to enhance lettuce production due to their abilities to increase NUE, favoring nutrient uptake and utilization efficiency. Furthermore, the appropriate incorporation of N in the plant is important since the nitrate content in vegetable products must be within the limits established by the market according to EU regulation no. 1258/2011, whereby the levels should not exceed 3000-5000 mg kg −1 fw.
In a recent opinion, Rouphael and Colla [23], indicated that the scientific community and private companies should focus on exploiting the potential synergistic biostimulatory action of microbes with non-microbial PBs combinations to design and develop second-generation plant products (biostimulant 2.0) with specific targeted biostimulant actions. A few experimental investigations have demonstrated the beneficial effects on crop performance of combining microbial inoculants (i.e., Rhizophagus intraradices or plant growth promoting bacteria or R. irregulare and T. atroviride) with humic acids [48,49] or protein hydrolysates [50]. Previous indications by Fiorentino et al. [25] suggested that the nutrient content of leafy horticulture crops could vary according to cultivation in diverse fertilizer conditions and in the presence/absence of a fungal inoculant. However, to date, nothing is known about the effects of Trichoderma alone or in combination with a vegetal biopolymer based-biostimulant on the agronomical, physiological and qualitative responses of an important leafy vegetables, such as Romaine lettuce (Lactuca sativa L. var. longifolia). This study will investigate the effect of a beneficial microbe (T. virens TG41) when used alone or in combination with a VBP biostimulant ('Quik-link'), under supra-optimal, optimal, and suboptimal N regimes, on Romaine lettuce production and marketability characteristics. This study will increase understanding of the processes involving these two different types of plant biostimulants and the effects on plant N acquisition response, for which the comprehension is pivotal to increasing NUE, as well as attempting to decrease N environmental inputs and reduce risks to consumer health.

Experimental Setup, Design, and Crop Management
An experiment was performed on lettuce (Lactuca sativa L. var. longifolia cv. 'Romana Bionda Lentissima a Montare'-Esasem, Casaleone, Verona) from November 4, 2015 to January 19, 2016, in a protected greenhouse structure (unheated) at the Department of Agricultural Sciences, University of Naples Federico II located at Portici, Italy. The soil was classified as a sandy loam texture (73% sand, 19% silt, 8% clay), with a pH of 7.0, electrical conductivity of 0.5 dS m −1 , an organic matter of 1.25% (w/w) and a total N of 1.1 g kg −1 . The NO 3 -N, NH 4 + -N, available P, and exchangeable K were 95, 7, 35, and 950 mg kg −1 , respectively.
A split-plot design with three replicates (randomized blocks) was adopted with fertilization (3 levels) as the main factor and biostimulant applications (3 levels) as the sub-factor. The three N fertilization levels were suboptimal (0 kg ha −1 ; 0N), optimal (70 kg N ha −1 ; 70N) and supra-optimal (140 kg N ha −1 ; 140N), while the three biostimulant applications were non-inoculated control, inoculated Trichoderma virens G41 (TG41), and T. virens + vegetal biopolymer-based biostimulant (TG41 + VBP). The cultivated area of each experimental plot (27 experimental plots in total) was 3.5 m 2 . Lettuce were transplanted on November 4th (at the 3 true-leaf stage) in double rows with a plant density of 14 plants per square meter. A biodegradable black mulch film (15 µm thick MaterBi ® , Novamont, Novara, Italy) was used and maintained throughout the entire greenhouse experiment.
N total amount was applied as ammonium nitrate (NH 4 NO 3 34%) into two identical doses, at 6 and 27 days after transplanting (DAT) by fertigation using a drip irrigation system with in-line emitters (flow rate: 3.3 L h −1 ; distances: 35 cm). Foliar pests, such as cutworms, were controlled with two applications of Decis Evo (active ingredient 25 g L −1 of deltamethrin-Bayer Crop Science, Milano, Italy) at the rate of 0.4 L ha −1 , whereas a copper-based fungicide (Cupravit 35 WG containing 350 g kg −1 of copper as copper oxychloride-Bayer Crop Science, Milano) was sprayed twice at the rate of 2.5 kg ha −1 to control downy mildew caused by Bremia lactucae Regel.

Fungal and Vegetal Biostimulants
A spore suspension of T. virens strain G41 (final concentration 1 × 10 7 spores mL −1 ; TG41) was used to inoculate the lettuce seedlings at time of transplant by using a root dip method (with submergence for 10 min); then a repeated inoculation was conducted at 18 DAT by watering 25 mL of the inoculum plant −1 . The vegetal biopolymer-based (VBP) biostimulant ('Quik-link ® ', Italpollina, Rivoli Veronese, Italy) was used in the current experiment. The product has a density of 1.21 kg L −1 , a pH (1:5) of 4.7, an electrical conductivity; EC (1:5) of 20 mS cm −1 , 25 g kg −1 of organic N as peptides and free amino acids, 160 g kg −1 of organic C, lignosulphonates, and micronutrients, such as iron, manganese, zinc, copper, and molybdenum, in the following concentrations 10.0, 7.0, 3.0, 1.0, and 0.2 g kg −1 , respectively [46]. Peptides and free amino acids were obtained through enzymatic hydrolysis of a vegetal source of proteins, as reported by Carillo et al. [7]. The peptides in the product have a high biological activity being signaling molecules (e.g., lateral root promoting peptides-LRPP). The commercial product was applied at the base of each plant (100 mL, containing 6 L ha −1 of 'Quik-link') at transplant, plus 17 (stage BBCH41-head beginning to form) and 45 DAT (stage BBCH45%-50% of the expected head size).

Fungal Colony Forming Units in Soil Rhizosphere and Trichoderma-VBP Compatibility
Soil samples were collected from the plant rhizosphere at the time of harvest. The number of fungal colonies forming units was determined, as indicated in Fiorentino et al. [25]. Briefly, a 1% (w/v) soil suspension was prepared in water, in serial dilutions, then 100 µL aliquots of each sample were spread on the surface of 90 mm culture plates containing Rose Bengal-Chloramphenicol agar (HiMedia Pvt. Ltd., Mumbai-India) supplemented with 0.1% (v/v) Igepal (Sigma-Aldrich, Milano, Italy), and incubated for 3-7 days at 25 • C. The emerging fungal colonies were counted daily.
In vitro tests were performed with varying doses of the VBP biostimulant and the Trichoderma inoculum, including the doses used for the field treatments to determine if the 'Quik-link' product inhibited the germination and growth of the fungus.

Fresh and Dry Yield, SPAD index and CIE (lab) Measurements
At harvest (76 DAT), the lettuce fresh yield was assessed in sampling areas of 2 m 2 from the center of the 27 experimental plots. The shoot dry biomass was determined (after oven drying at 80 • C for 72 h). The dried leaf tissues were conserved for mineral analysis. At 45 and 75 DAT, the soil plant analysis development (SPAD) index (i.e., non-destructive measurement of chlorophyll content) was measured on undamaged and expanded lettuce leaves using a portable SPAD-502 chlorophyll meter (Konica-Minolta, Tokyo, Japan). Twelve measurements were conducted on four randomly picked lettuce plants per experimental plot, then averaged to a single SPAD value for each replicate [51]. Subsequently, on the same date, measurements were performed using a Minolta CR-300 Chroma Meter (Minolta Camera Co. Ltd., Osaka, Japan) to evaluate the Commission Internationale de L'Eclairage (CIE) color space parameters for L* (lightness) and chroma coordinates: a* (−a* greenness) and b* (+b* yellowness). In each experimental plot, 10 healthy leaves were measured and averaged to represent a single color value [52].

Mineral Composition Analysis
Plant material was dried and pulverized using a cutting-grinder head (IKA, MF10.1, Staufen, Germany), then the powder was extracted in Milli-Q water (Merck Millipore, Darmstadt, Germany) for 10 min at 80 • C in a thermostatic bath (ShakeTemp SW22, Julabo, Seelbach, Germany) and centrifuged at 6000 rpm for 10 min as indicated in Rouphael et al. [50]. A Dionex ICS-3000 system (Sunnyvale, CA, USA) equipped with suppressed conductivity detection was used to determine the ion content of the samples. The ion separation of the samples was carried out with two different ion-exchange columns: An IonPac CS12A column (250 × 4 mm) was used for the cation separation eluted with 20 mM methanesulfonic acid (flow rate 1 mL min −1 ), and an IonPac AS11-HC column (250 × 4 mm) was used for the anion separation eluted with a potassium hydroxide gradient (flow rate 1.5 mL min −1 ). Nitrogen (total N) concentration in leaf tissue was determined according to the Kjeldahl method [53].

Antioxidant Capacity, Total Phenols, and Total Ascorbic Acid Analysis
Lipophilic and hydrophilic antioxidant capacity and total phenols were determined on freeze-dried tissue samples, whereas the total ascorbic acid was assessed on fresh material and measured using a spectrophotometer (Hach DR 2000, Hach Co., Loveland, CO, USA) according to the protocols of Re et al. [54], Fogliano et al. [55], Singleton et al. [56], and Kampfenkel et al. [57], respectively. Solution absorbances were assessed at 505, 734, 525, and 765 nm for the lipophilic and hydrophilic antioxidant fractions, total polyphenols, and total ascorbic acid, respectively.

Data Elaboration, Statistical Analysis, Principal Component Analysis, and Heat Map
The statistical analyses were all carried out using the software IBM SPSS Statistics 21. All data were subjected to two-way analysis of variance, and mean values were separated according to Duncan test with p < 0.05. Principal component analysis (PCA) was performed on the whole morphological and physiological data set, and the eigen values, total variance of the first three principal components (PCs) as well as the loading scores and plots were determined [58][59][60]. A heat map summarizing the agronomical, physiological, and qualitative responses of lettuce to plant biostimulant applications and N fertilization levels was also generated using the https://biit.cs.ut.ee/clustvis/ online program package with Euclidean distance as the similarity measure and hierarchical clustering with complete linkage [6].

Fungal Concentration in the Soil
The total number of fungal colonies (including Trichoderma) recovered from soil rhizosphere in the nine treatments ranged between 2.0 × 10 5 and 6.5 × 10 5 colony forming units (CFU) g −1 of soil and was significantly (p < 0.05) influenced by the interaction of the two tested factors: N fertilization level (N) and VBP biostimulant application. In particular, results indicated that the highest fungal CFU was observed in soils from lettuce plants inoculated with TG41 under suboptimal 0N conditions (6.5 × 10 5 CFU g −1 of soil), in comparison to any of non-inoculated plants under suboptimal, optimal, or supra-optimal N conditions (average 2.5 × 10 5 CFU g −1 of soil), whereas the treatments with TG41 (at 70N and 140N) or TG41 +VBP biostimulant (at 0N and 70N) exhibited intermediate values (average 3.9 × 10 5 CFU g −1 of soil) (data not shown). Moreover, the in vitro tests performed with the beneficial microbe (TG41) and non-microbial VBP biostimulant at the dose applied in the greenhouse experiment did not demonstrate any inhibition of the germination and growth of the fungi concentration (69.2 CFU in the absence and 68.5 CFU in the presence of the 'Quik-link-product), suggesting compatibility between the two biostimulants.

Growth Responses, SPAD Index and Leaf Colorimetry
A significant (p < 0.01) interaction between N fertilization level and biostimulant application was observed on fresh yield and dry biomass. For instance, the use of the TG41-based biostimulant alone or in combination with the VBP biostimulant positively affected both fresh and dry yield of lettuce plants under both sub-optimal (0 kg ha −1 ) and optimal (70 kg ha −1 ) N conditions, but the beneficial effect was not apparent in the over N fertilization condition (140 kg ha −1 ) ( Figure 1). Lettuce grown in the absence of N fertilization demonstrated a highly significant increase in fresh yield of 67% when inoculated with the combined TG41+VBP biostimulants. Instead, a more moderate increase of 45% was observed over the untreated 0N condition with the inoculation of T. virens G41 alone. Moreover, under optimal N fertilization (70 kg ha −1 ), only lettuce plants inoculated with TG41 alone exhibited significantly higher fresh yields. Treatments with TG41 alone or TG41+VBP increased marketable dry yield by 16% when compared to the untreated control, but no significant differences were noted between the two different biostimulant inoculations ( Figure 1). No effects on lettuce yield were observed with either of the biostimulants at the supra-optimal 140N fertilization.
The SPAD index in Lactuca sativa L. var. longifolia, as an indication of chlorophyll content, was significantly affected by N fertilization levels (at 75 DAT) and by biostimulant applications (at 45 and 75 DAT), with effects in the N × T interaction (Table 1). At 75 DAT, the highest SPAD index values were recorded with TG41 + VBP biostimulant combination (Table 1). The visual appearance, particularly the greenness of leaf color, is a primary parameter used by the consumer in product preference and selection choice [61]. In general, neither the N fertilization level nor biostimulant application had a significant effect on the leaf greenness (−a* values) in lettuce (Table 1). Overall, the N application levels resulted in a greater lightness in the color of the lettuce leaves, with the lowest L* values recorded in the 140 kg N ha −1 treatment, which also corresponded to a decrease in the chroma coordinate (b*; Table 1).

Leaf Gas Exchange: Net CO 2 Assimilation Rate and Stomatal Resistance
The physiological parameters, in particular, the net CO 2 assimilation rate (A CO2 ) and stomatal resistance (r s ) in the Romaine lettuce plants throughout the cultivation cycle in the greenhouse, were evaluated as a function of N fertilization level and biostimulant application as displayed in Table 2. The A CO2 was significantly affected by the biostimulant treatments for all measured data, and to a lesser degree, by the N fertilization level (only at 48 DAT). Irrespective of the N fertilization level (N × B interaction= ns) at 33 and 40 DAT, both the TG41 alone or in combination with the VBP-based biostimulant induced higher values of A CO2 in comparison to the control treatment that was not significantly different between the two biostimulant treatments. At 48 DAT, the A CO2 increased in the following order with the applications: TG41+VBP > TG41 > control ( Table 2). On the other hand, augmenting the N fertilization level resulted in a linear increase in A CO2 from 0 to 140 kg ha −1 but only at 48 DAT (Table 2).
Contrary to A CO2 , the r s was not affected neither by N fertilization level nor by biostimulant application at 33 and 48 DAT, while at 40 DAT, the r s was only influenced by the two biostimulant applications (Table 2). Particularly, on this date, the r s was significantly lower on average by 26% when lettuce plants were inoculated with Trichoderma alone or in combination with the commercial product 'Quik-link' ( Table 2).

Mineral Composition in Leaf Tissue
The results regarding the mineral profile in Romaine lettuce leaves are presented in Table 3. For all the macronutrients and sodium analyzed, no significant differences were observed in the N fertilization level and biostimulant application interaction. In particular, neither N fertilization rate nor biostimulant treatment had a significant effect on Ca and Na concentrations in lettuce leaves (average 7.0 and 1.4 g kg −1 dry weight, respectively; Table 3). The concentrations of N and P in leaf tissues were significantly affected by N fertilization rate. Concentrations of N and P increased as the N fertilization level increased, with the highest values recorded at 140 kg ha −1 for N and at 70 and 140 kg ha −1 for P ( Table 3).
The effects of TG41 and TG41+VBP biostimulant, when averaged over all N fertilization rates, affected the K and Mg concentrations in leaf tissues which were higher by 10% and 12%, respectively, than in untreated lettuce plants, but with no significant difference noted between the two biostimulant treatments (Table 3). Table 2. Net CO 2 assimilation rate and stomatal resistance of greenhouse Romaine lettuce plants measured during the production cycle in relation to N fertilization level (0N = 0 kg ha −1 , 70N = 70 kg ha −1 , 140N = 140 kg ha −1 ) and biostimulant application (Untreated=Control, TG41=T. virens G41, and TG41+VBP=vegetal biopolymer-based biostimulant).

Treatments
Net CO

Nitrate, Antioxidant Capacity, and Bioactive Content
The registered nitrate content among the 9 experimental conditions (890-1496 mg kg −1 fresh weight) was within the limits imposed by the European Regulation No. 1258/2011 for the commercialization of fresh lettuce (3000-5000 mg kg −1 on a fresh weight basis). In our study, nitrate content was affected by both N fertilization level and biostimulant application, without significant effects in the N×B interaction ( Table 4). As expected, our results demonstrated that increasing N fertilization from 0 to 140 kg ha −1 elicited a significant increase in nitrate content compared to non-fertilized plants, whereas lettuce plants cultivated under optimal N fertilization (70 kg ha −1 ) exhibited intermediate values (Table 4). Interestingly, the nitrate content was significantly lowered in lettuce plants receiving treatments of either TG41 alone and the combined TG41+VBP biostimulants (not significant between them) compared to the untreated control ( Table 4).
The hydrophilic and lipophilic antioxidant fractions of greenhouse lettuce ranged from 1.44 to 1.61 mmol ascorbic acid eq. 100 g −1 dw and from 2.69 to 4.62 mmol trolox 100 g −1 dw, respectively. Neither N fertilization level nor the biostimulant application had a significant effect on the hydrophilic antioxidant activity. Moreover, significant effects were noted on lipophilic antioxidant activity (LAA) with both N and biostimulant treatments, but not the N×B interaction. Irrespective of N fertilization treatments, the application of TG41+VBP demonstrated a significant increase in LAA (+13%) compared to the treatment of TG41 alone and the non-inoculated control (Table 4). Moreover, antioxidant molecules, in particular, total phenols and total ascorbic acid, were significantly influenced by either tested factors of N fertilization and biostimulant application. When averaged over the nitrogen treatments, the lettuce plants cultivated under supra-optimal conditions (i.e., 140 kg ha −1 ) were characterized by low-quality bioactive compounds in terms of both total phenols and total ascorbic acid (Table 4). Interestingly, the biostimulants-treated plants with TG41 alone and particularly in the combination of TG41+VBP, produced a major amplification of total phenols (+14%) and total ascorbic acid (+61%-91%) in comparison to untreated lettuce plants (Table 4). Table 4. Nitrate content, hydrophilic (HAA), and lipophilic (LAA) antioxidants activities, total phenols and total ascorbic acid (TAA) content of greenhouse Romaine lettuce at time of harvest in relation to N fertilization level (0N = 0 kg ha −1 , 70N = 70 kg ha −1 , 140N = 140 kg ha −1 ) and biostimulant application (Untreated = Control, TG41 = T. virens G41, and TG41+VBP = vegetal biopolymer-based biostimulant).

Heat Map Analysis of all Measured Plant Parameters
An aggregated data heat-map analysis of the measured agronomic and physiological parameters was conducted to produce a visual comparison of the effects determined by the tested treatment factors on the Romaine lettuce plants. In Figure 2, the analysis revealed two dendrograms: on the top (Dendrogram 1), a classification that corresponded principally to the biostimulant applications, and on the left (Dendrogram 2), the parameters that influenced this distribution. Dendrogram 1 revealed two main groups: on the left, the cluster corresponded to controls for each of the three N levels that were all untreated with the biostimulant conditions; then on the right of the heat map, two clusters that contained the other six treatments, consisting of a mix of the N levels receiving the biostimulant applications ( Figure 2).
In particular, in the left cluster of Dendrogram 1, the 140N Control was well separated from the other two controls (0N and 70N) due to the higher r s at 40 and 48 DAT; nitrate, Na, P, and dry biomass (in the first/highest cluster of Dendrogram 2), as well as the lower values for the parameters in the second cluster, mainly for the parameters of L*, total phenols, Mg, and K content. On the right side of Dendrogram 1, two clusters were identified, the first on the left included treatments 70N TG41+VBP biostimulant, separated from the 140N level with the biostimulants TG41 or TG41+VBP, that showed in particular lower Na, r s at 40 and 48 DAT, hydrophilic (HAA), b* and total phenols parameters, but higher LAA, a* value, P and N content, SPAD index and A CO2 at 33 and 48 DAT. The grouping on the right included 70N TG41, 0N TG41, and 0N TG41+VBP treatments. Within this cluster, the 0N treatments with the biostimulants were clearly separated from the 70 N TG41 by higher HAA, leaf number (LN), and lower LAA in this latter treatment. Instead, the two 0N levels receiving the biostimulants were distinguished by the parameter groupings found in Dendrogram 2, whereby 0N TG41 could be attributed to the lower values for the parameters found in the third cluster (mainly due to a*, SPAD Index, N), as well as the lower r s 40 DAT and nitrate, but higher b*; whereas 0N TG41+VBP biostimulant could principally be identified by the all the higher parameters found in the second cluster-specifically total ascorbic acid (TAA). Interestingly, the first cluster in Dendrogram 2 clearly demonstrated the differential effects of the biostimulant treatments (untreated ones had high parameters for all N levels), while the second cluster clearly revealed the consequence of supra-optimal N levels (all parameter values were low), and the outcome of the combined biostimulants in the low N level condition (all parameter values were high), comparatively to the Trichoderma alone (i.e., TG41) at 0N.

Principal Component Analysis of all Measured Plant Parameters
Principal component analysis was carried out on the whole experimental data set, and the loading plot and scores are reported in Figure 3. The analysis indicated that the variables in the first three principal components (PCs) were highly correlated, with eigen values greater than 1, thus explaining for 80.4% of the total variance, with PC1, PC2, and PC3 accounting for 36.1%, 32.0%, and 12.4%, respectively. The variable distribution along PC1 was clearly attributed to the biostimulant treatments, while N fertilization levels contributed to that on PC2 (Figure 3). TG41 and TG41+VBP biostimulant treated plants were distributed in the positive quadrants of PC1 except for 0N TG41, while all control treatments (untreated lettuce plants) were distributed in the negative side of PC1. In particular, 0N TG41+VBP biostimulant and 70N TG41 were in the upper right quadrant, while 70N

Principal Component Analysis of all Measured Plant Parameters
Principal component analysis was carried out on the whole experimental data set, and the loading plot and scores are reported in Figure 3. The analysis indicated that the variables in the first three principal components (PCs) were highly correlated, with eigen values greater than 1, thus explaining for 80.4% of the total variance, with PC1, PC2, and PC3 accounting for 36.1%, 32.0%, and 12.4%, respectively. The variable distribution along PC1 was clearly attributed to the biostimulant treatments, while N fertilization levels contributed to that on PC2 (Figure 3). TG41 and TG41+VBP biostimulant treated plants were distributed in the positive quadrants of PC1 except for 0N TG41, while all control treatments (untreated lettuce plants) were distributed in the negative side of PC1. In particular, 0N TG41+VBP biostimulant and 70N TG41 were in the upper right quadrant, while 70N TG41+VBP biostimulant, 140N TG41+VBP biostimulant, and 140N TG41 were in the lower right quadrant. Moreover, in PC2, 0N TG41 was positioned in the positive side of the upper left quadrant, with 0N and 70N untreated control treatments, while the 140N control was in the lower left negative quadrant (Figure 3). PC1 was positively correlated to A CO2 at 33, 40, and 48 DAT, K, Mg, and Ca content, yield (fresh weight), dry biomass, and SPAD index. PC1 was also negatively correlated with r s at 33, 40, 48 DAT, and also with nitrate content. PC2 was positively correlated with L* and b* colorimetric parameters, total phenols, and TAA, while it was negatively correlated to P content and a* colorimetric parameter. In addition, the treatments with 70N TG41+VBP biostimulant and 140N TG41 produced lettuce with a higher yield, leaf number, SPAD index, and A CO2 at 48 DAT. Interestingly, the non-fertilized 0N lettuce plants treated with TG41+VBP produced lettuce with higher premium quality (higher total phenols and TAA and lower nitrate content) (Figure 3). Finally, the upper and lower left quadrant depicted the three non-treated control treatments with the lowest quality characteristics (high Na and nitrate content; Figure 3).  Figure 3). PC1 was positively correlated to ACO2 at 33, 40, and 48 DAT, K, Mg, and Ca content, yield (fresh weight), dry biomass, and SPAD index. PC1 was also negatively correlated with rs at 33, 40, 48 DAT, and also with nitrate content. PC2 was positively correlated with L* and b* colorimetric parameters, total phenols, and TAA, while it was negatively correlated to P content and a* colorimetric parameter. In addition, the treatments with 70N TG41+VBP biostimulant and 140N TG41 produced lettuce with a higher yield, leaf number, SPAD index, and ACO2 at 48 DAT. Interestingly, the non-fertilized 0N lettuce plants treated with TG41+VBP produced lettuce with higher premium quality (higher total phenols and TAA and lower nitrate content) ( Figure 3). Finally, the upper and lower left quadrant depicted the three non-treated control treatments with the lowest quality characteristics (high Na and nitrate content; Figure 3).

Discussion
Our findings indicated that the suboptimal fertilizer condition (0 kg N ha −1 ) sharply reduced yield, dry biomass, and ACO2, particularly at 48 DAT, whereas rs and sodium content increased. In fact, at 0 and 70 kg ha −1 , the lower leaf N availability may affect photosynthetic performance and rate due to N remobilization from photosynthetic enzymes and pigments [62]. The decreased SPAD index, which is significantly correlated to chlorophyll concentration as indicated by absorbance measurements [63], corresponded to the decrease in photosynthetic capacity, and an increase in the sensitivity to photo-inhibition [64]. However, the application of TG41 alone, but especially in combination with the VBP, to lettuce grown in sub-optimal N induced significant changes in morphology and physiology, as noted with increased yield and dry biomass. Therefore, under low-input conditions (0 kg N ha −1 ), the combination of the microbial inoculant with the biopolymer-based biostimulant exhibited an important synergistic effect, thus confirming the

Discussion
Our findings indicated that the suboptimal fertilizer condition (0 kg N ha −1 ) sharply reduced yield, dry biomass, and A CO2 , particularly at 48 DAT, whereas r s and sodium content increased. In fact, at 0 and 70 kg ha −1 , the lower leaf N availability may affect photosynthetic performance and rate due to N remobilization from photosynthetic enzymes and pigments [62]. The decreased SPAD index, which is significantly correlated to chlorophyll concentration as indicated by absorbance measurements [63], corresponded to the decrease in photosynthetic capacity, and an increase in the sensitivity to photo-inhibition [64]. However, the application of TG41 alone, but especially in combination with the VBP, to lettuce grown in sub-optimal N induced significant changes in morphology and physiology, as noted with increased yield and dry biomass. Therefore, under low-input conditions (0 kg N ha −1 ), the combination of the microbial inoculant with the biopolymer-based biostimulant exhibited an important synergistic effect, thus confirming the beneficial effects on crop productivity as previously reported by several authors [48][49][50]. Both PB treatments enhanced photosynthetic activity, SPAD index, and leaf nutritional status, as reflected by higher K and Mg and lower Na concentrations, that indicate a more efficient accumulation and translocation of assimilates to photosynthetic sinks that improve crop performance, but are not associated to the external N fertilization level applied [50]. Under optimal N conditions (i.e., recommended rate of 70 kg ha −1 ), the treatment with TG41 alone had the best effect on fresh yield, combined with high Mg and antioxidant contents, as well as low nitrate and Na. In this N regime, the addition of the VBP-based biostimulant to the fungal inoculant did not improve the morpho-physiological parameters, nor the mineral profile in the leaves. As mentioned above, growth under suboptimal N conditions increased leaf cell susceptibility to light-induced oxidative damages, a condition that plants are not capable of overcoming. However, the application of the combined microbial and VBP PBs induced a strong production of TAA, phenols, and probably glutathione, a metabolite that works cooperatively with ascorbic acid to generate antioxidant effects that safely detoxify accumulated reactive oxygen species (ROS), thus protecting the plant and increasing the photosynthetic rate [6,7].
The application of 140 kg N ha −1 to lettuce was an excess condition that determined a plateau in yield and dry biomass but not in the N content, although there was an increase in nitrate and Na, as well as r s at 40 DAT. This demonstrated that plants supplied with high levels of N were not able to assimilate and reduce all the nitrate supplied, risking negative consequences by the accumulation of these compounds in the vacuoles. This was also reported by Di Mola et al. [65] in rocket plants and by Wang et al. [66] in leafy vegetables, whereby optimal and particularly supra-optimal N treatments were not always characterized by the best quality traits in the produce, but on the contrary, resulted in damage to the commercial, nutritional, and functional quality traits. These effects were similar to those noted in our lettuce plants under supra-optimal N conditions, i.e., low macronutrients and total ascorbic acid, high nitrate and sodium content. The application of both PBs under supra-optimal N level (e.g., 140 kg ha −1 ) significantly enhanced the N content and SPAD index while reducing nitrate content without affecting the CO 2 assimilation rate and the accumulation of beneficial nutrients. The strong increase in the SPAD index at 140 kg ha −1 in plants inoculated with TG41 or TG41+VBP biostimulants was also observed at 70 kg ha −1 , suggesting that the biostimulants were able to increase the number and efficiency of photosynthesis systems and light-harvesting complexes (LHC), that allowed plants to "fine-tune" photosynthesis in the fluctuating spectral quality and light intensity conditions, thus avoiding ROS formation and photo-oxidation. This also allowed a higher use efficiency of nitrate, as confirmed by the lower concentration of this ion in leaf tissues when compared to the untreated control because of a more efficient reduction and assimilation processes [6,7].
Our results correspond to previous findings on the plant growth-promoting effect of fungi inoculants containing Trichoderma [25,26,29,30,33,38,67]. The presumed mechanisms behind the beneficial morpho-physiological effects on lettuce plants by TG41 could be due to the release of signaling molecules with auxin and ethylene-like activity [28], in particular, bioactive volatile compounds [43], which increased nutrient bioavailability to the plant, that improved their uptake, translocation, and accumulation within the plant [35]. In addition, it has also been demonstrated that Trichoderma in the rhizosphere stimulates root growth and reshapes its architecture, morphological changes which are pivotal for improving nutrient uptake, in particular, nitrate, Ca, Mg, and K [29,30,35,41]. The synergistic action of TG41 with the VBP biostimulant is of particular interest because it resulted in the production of premium quality lettuce traits, as is clearly exhibited by the PCA. The vegetal-biopolymer biostimulant action was probably due to the presence of phenol metabolites with auxin and gibberellin-like activities, that interacted with phytohormones and enzymes stimulating the activity of carbon-nitrogen metabolism and plant development [45,46]. Another putative mechanism behind the stimulation of plant growth and yield in response to VBP drench application could involve the increased presence of bioactive molecules, such as signaling peptides (LRPP) and lignosulphonates, which are typical compounds present in VBP [46]. A previous study reported that lignosulphonate treatments can improve N uptake and assimilation in plants through the stimulation of glutamate synthase and glutamine synthetase, as well as by triggering photosynthetic activity through the stimulation of both rubisco enzyme activity, thus improving plant performance [45]. The improved NUE in lettuce treated with PBs enhanced not only the chlorophyll content (as represented by the increased SPAD index) but also the synthesis of antioxidant metabolites that were capable of re-activating photosynthetic activity that under sub-optimal N conditions without PBs, would be severely compromised. Finally, the synergistic beneficial effect on root system architecture, as previously shown by Colla et al. [38,68], determined a 'nutrient acquisition response' improving resource use efficiency (RUE) that enhanced plant biomass production and the quality of the produce.

Conclusions
Our study on the leafy vegetable crop Romaine lettuce confirmed that inoculations with Trichoderma TG41 under optimal N conditions (70 kg ha −1 ) were able to improve the leaf nutritional status as indicated with the higher potassium and magnesium content and lower sodium content, plus providing the best yield performance of all tested conditions in terms of plant fresh and dry weight. Interestingly, the combined biostimulant applications of Trichoderma with the vegetal biopolymer-based product, in suboptimal fertilizer conditions of low N availability (0N kg ha −1 ), was more effective than the treatment of the microbial inoculant alone not only in improving yield but also in producing a premium quality marketable lettuce with higher lipophilic antioxidant activity and total ascorbic acid content. Together these biostimulants positively influenced plant morpho-physiological processes that improved the assimilation of nitrate and macronutrients and stimulated root system architecture reshaping, thus permitting increased bioabsorption or 'nutrient acquisition response'. Moreover, the assimilatory pathways were stimulated, for which nitrate was used to synthesize chlorophyll (increased SPAD index) and the antioxidant metabolites, which, in turn, re-activated the CO 2 assimilation activity normally decreased under sub-optimal N conditions. Therefore, the combination of microbial and non-microbial plant biostimulants represents a promising, efficient, and sustainable strategy for improving yield and quality of horticultural crops, such as lettuce, as well as improving cultivation in N compromised fields or low fertilizer input scenarios.