Development of a PGPB-based Biofertilizer to Optimize Strawberry Cultivation in Semiarid regions: Screening, Validation and Scaling up to Commercial Production

Research Paper Development of a PGPB-based biofertilizer to optimize strawberry cultivation in semiarid regions: Screening, validation and scaling up to commercial production Enrique Mateos-Naranjo a,b,* , Jesús V. García-L´ opez b , Noris J. Flores-Duarte c , Elena Romano-Rodríguez a , Ignacio D. Rodríguez-Llorente c , Jesús A. P´ erez-Romero a , Eloísa Pajuelo c , Susana Redondo-G´ omez a a Departamento de Biología Vegetal y Ecología, Facultad de Biología, Universidad de Sevilla, 41012 Seville, Spain b Servicio General de Invernadero, Centro de Investigaci´ on, Tecnología e Innovaci´ on (CITIUS), Universidad de Sevilla, 41012 Seville, Spain c Departamento de Microbiología y Parasitología, Facultad de Farmacia, Universidad de Sevilla, 41012 Seville, Spain ARTICLE INFO Keywords: Beneficial microbes Bioinoculant Chlorophyll fluorescence Intensive farming practices Metabolites Technology readiness level (TRL) Production ABSTRACT The interest in the use of PGPR-based biofertilizers has increased in the last few years, since they may allow crops to increase their productivity through alleviating environmental stress. However, this promising technology is still at an early experimental stage since the majority of evidence has been obtained under controlled conditions. Therefore, the technology readiness levels (TRL) of PGPR-based biofertilizers is in low phases (1–4; laboratory environment), so it is necessary to focus on higher phases to achieve real implementation. In this study, our aim was to reach levels framed between TRLs 5–6, from relevant to the real environment, which addresses the design and development of a definitive prototype of PGPR-based biofertilizer to improve strawberry production under two agronomic managements [FS1 (100 % application of evapotranspired water and conventional fertilizer application) and FS2 (70 % application of irrigation and fertilization reduction to 70 %)] through three experimental phases (1: biofertilizer screening; 2: validation under greenhouse; and 3: trial validation in a commercial strawberry production facility). Phases 1 and 2 allowed us to select biofertilizer 2 (PGP strains SDT3, HPJ40, SMT38, SRT15 and S110) which was able to increase production c. 13 % and 23 % under the FS1 and FS2 treatments, respectively. Furthermore, a significant relationship was also found between biofertilizer supply and the accumulation of primary metabolites. These positive effects were associated with the higher plant carbon assimilation capacity and photosystem energy efficiency. Commercial facility trial validation results showed an increase of 10 % and 8 % in inoculated plots with respect to non-inoculated plots under FS1 and FS2 treatments, respectively. Likewise, this positive effect was related to positive physiological responses. Although the biofertilizer effect was less acute than under laboratory conditions, the magnitude of the percentages obtained was important enough to validate the positive impact of biofertilizer 2 on strawberry yield in the real environment to be able to verify the development of this technology up to level 6. 1. Introduction Several studies suggest that world crop production will need to roughly double to meet future demands from population growth, dietary changes and rising bioenergy use (Foley et al., 2011; Hunter et al., 2017). The challenge becomes all the more daunting if we consider that agriculture itself is the dominant force responsible for such environmental threats as climate change, biodiversity loss and degradation of soils and water resources (Foley et al., 2011; Curtis et al., 2018). Therefore, one of the main challenges facing 21st century society is the need to meet the growing demand for food and resources while greatly reducing agriculture’s environmental footprint in order to guarantee environmental sustainability. To achieve a balance between food security and environmental sustainability of agricultural practices, the search for techniques that facilitate a more efficient management of water resources and optimize * Corresponding author. E-mail address: [email protected] (E. Mateos-Naranjo). Contents lists available at ScienceDirect Scientia Horticulturae journal homepage: www.elsevier.com/locate/scihorti https://doi.org/10.1016/j.scienta.2024.113929 Received 14 May 2024; Received in revised form 18 December 2024; Accepted 19 December 2024 Scientia Horticulturae 340 (2025) 113929 Available online 26 December 2024 0304-4238/© 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC license ( http://creativecommons.org/licenses/bync/4.0/ ). fertilization in agricultural practices is needed. Among these potential technical solutions, the use of beneficial microorganisms naturally present in the soil or supplied in the form of biofertilizers is gaining momentum (Zaidi et al., 2015; Ruzzi and Aroca, 2015; Mahanty et al., 2017). In particular, the use of beneficial bacteria capable of directly or indirectly promoting plant development, the so-called plant growth-promoting rhizobacteria (PGPR), has increased substantially in recent years to improve various crops. PGPR allows crops to increase their productivity values (Zaidi et al., 2015; Tsegaye et al., 2022), to alleviate environmental stress (Nadeen et al., 2014; Paul and Lade, 2014; Etesami and Maheshwari, 2018; Khan et al., 2021; Shultana et al., 2022) and to improve the management efficiency of agricultural inputs (Otaiku et al., 2019). In fact, positive studies have been carried out in the production of crops of global importance, such as corn (Zea mays; Riggs et al., 2001; Krey et al., 2013; Di Salvo et al., 2018), rice (Oryza sativa; Islam et al., 2009; Abd El-Magged et al., 2022), wheat (Triticum aestivum; M¨ ader et al., 2011; Roesti et al., 2006; Nawaz et al., 2020) or sugar beet (Beta vulgaris; Çakmakçi et al., 2006; Artyzak and Gozdowski, 2020). Nevertheless, although this is a most promising technology, it is still at an earlier experimental stage and has a low level of commercial implementation on a global scale. Only a small number of bacterial strains with PGP properties have been commercialized for agricultural purposes. Therefore, there is a need for a wider selection of PGP strains with biological activities suitable for specific agronomic contexts (Glick, 2012; Ansari et al., 2017). Another challenge regarding the application of this technology is the lack of knowledge of the adaptability capacity of the microorganisms to the particular conditions in the rhizosphere. In this sense, although inoculations could be performed with exogenous microorganisms, isolated in very different geographic areas, many studies have recommended the use of indigenous microorganisms, which show better competitiveness and adaptability to specific agronomic context (El Fantroussi and Agathos, 2005; Afzal et al., 2019). Another recommendation refers to the combination of strains with beneficial properties, even overlapping beneficial traits, to ensure the maintenance of those positive features in the soil to improve growth of crops (Etesami and Maheshwari, 2018; Morales-García et al., 2019). In addition, it is necessary to use microorganisms capable of growing and maintaining their PGP properties under more extreme climatic conditions (for instance, with more extended and/or severe episodes of water deficits and/or salinity), similar to the conditions likely to be present in a future climate scenario, or those which would arise from a restriction in input supplies (Orhan and Demirci, 2020). Finally, these experiments should be conducted over the whole crop cycle, on a commercial scale, and under conditions with variations in environmental factors, which also include variations in inputs and their application, in order to achieve a reliable assessment of the potential benefits of this technology. Taking into account the premises described previously, a study was designed and developed with the aim of providing a new biotool based on PGP biofertilizer with multiple properties and elevated tolerance to environmental stress to improve the sustainability of agricultural intensity practices. To accomplish this aim, the focus was on one single crop, i.e. strawberry (Fragaria x ananassa). Strawberry was selected as a model crop because it is an intensive crop, which requires a large amount of inputs, such as fertilizers and water in large quantities, since this species has a superficial root system, a large leaf surface area and produces fruits with high water content (Krüger et al., 1999; Grant et al., 2010; Martínez-Ferri et al., 2016). Furthermore, the strawberry industry is highly dependent on the availability of high-quality water. This fact will make the strawberry industry one of the agricultural sectors most seriously affected by future climate conditions, making it a suitable model crop to study the effects of potential management limitations imposed by future climatic conditions, and how they could be modulated by the application of a PGPR based biofertilizer. Regarding the use of microorganisms in strawberry crops, several authors have identified its potential to improve strawberry tolerance to some environmental stresses (water deficit: Erdogan et al., 2016, García-L´ opez et al., 2024; salt stress: Karlidag et al., 2010, 2013 García-L´ opez et al., 2024; nutrient limitation: García-L´ opez et al., 2023; Valle-Romero et al., 2023). Furthermore, studies have been conducted evaluating the effect of different inoculants created by combining AMF and bacterial strains with positive outcomes (Scmitzer, 2023; Tadeschini, 2018), or solely with PGPR strains isolated from the rhizosphere of various strawberry cultivars (Sangiorgio, 2023), or from different industrial processes (De Andrade, 2019). Despite this effort, it should be noted that most of the mentioned studies were conducted under controlled conditions in greenhouses, with a low number of samples and short sampling periods, which limits the ability to extrapolate their conclusions to a real agronomic scale. Furthermore none of these studies have evaluated the effect of bacteria highly resistant to environmental stress derived from the rhizosphere of halophytes. Taking these aspects into account, our working hypothesis is that the design and validation of PGPR-based biofertilizer, integrated by selfcompatible strains with high tolerance to environmental stress with multiple PGP properties, would be a useful biotool to increase the production of this intensive crop and will also improve its tolerance to the limited input management scenarios. To test this hypothesis, three different experimental phases were developed to select and validate the potential of a previously designed multifunctional PGPR-based biofertilizer to improve strawberry plant production efficiency under conventional and input restriction agronomic management (Fig. 1). In addition, as this work addresses the development of a new technological tool, we have attempted to follow the Technology Readiness Levels (TRL) nomenclature in each of the experimental phases (EU, 2020), which is a globally accepted nomenclature to frame the progress resulting from experimentation in the appropriate TRL to determine the degree of maturity of this new technology. 2. Material and methods 2.1. PGPR-based biofertilizers The two biofertilizers tested were integrated by five self-compatible bacterial strains isolated from the rhizosphere of halophytes from the Odiel River marshes, (Huelva, Spain: Spartina densiflora, Spartina maritima, Salicornia ramosissima and Halimione portulacoides) and an endophyte strain from the Belgian Coordinated Collection of Microorganisms (BCCM) with high multi-stress resistance and a variety of complementary plant growth-promoting (PGP) (see Fig. 2 for details of the PGPbased biofertilizers tested). These biofertilizers were used on the commercial strawberry cultivar Rociera throughout the study. For technical details on inoculum preparation and strains compatibility evaluation, see Navarro-Torre et al. (2016) and Flores-Duarte et al. (2022a). 2.2. Experimental phases 1 and 2: PGPR-based biofertilizers screening, selection and validation under greenhouse conditions (Technology readiness levels 4–5) Two experimental phases were conducted in the University of Sevilla greenhouse (Research, Technology and Innovation Centre, CITIUS II; 37◦240 N, 6◦00 W; Southwest Spain) to select the best biofertilizer for future agronomic validation (TRL6). The first phase served to select the inoculant that would be developed based on the results of crop production under simulated environmental conditions (TRL4). In the second phase, production monitoring was complemented with measurements of fruit quality and plant physiological performance throughout one crop cycle in a simulated environment for selected biofertilizer validation (TRL5). For phase 1 development, at the beginning of the crop cycle, seedlings of strawberry plants were obtained and transplanted into 3.5 L plastic pots filled with soil from a commercial farm located in Palma del Condado (Huelva, Spain 37◦361 N, 6◦56 W; Southwest Spain). The physic and chemical properties of the soil are given in Table 1. Before E. Mateos-Naranjo et al. Scientia Horticulturae 340 (2025) 113929 2 transplantation, the soils were sterilized by chemical fumigation (H 2 O 2 , 12 % v/v) and solarization. Sowed plants were placed in a greenhouse with 16–24 ◦C, 40–50 % RH, and subjected to a day/night regime of 14 h of light with a maximum photosynthetic photon flux density (PPFD) incident on leaves of 1000 μ mol m -2 s -1 , 10 h of darkness and with adequate fertigation (100 mL N/P/K of 20/20/20 per plant every two days) until complete implantation. Then, they were allocated to three inoculation treatments (non-inoculation, control and monthly inoculation throughout the experiment, with the different previously designed PGPR-based biofertilizers) in combination with two fertigation Fig. 1. Scheme of the experimental framework, relating the Technology Readiness Level (TRL), the different experimental phases (Phases 1–3), experimental tasks, agronomic management scenarios tested and images of the proposal. Fig. 2. Plant growth-promoting rhizobacterial traits (PGPR) for the strains of biofertilizer used in this study. Information adapted from García-L´ opez et al. (2023). E. Mateos-Naranjo et al. Scientia Horticulturae 340 (2025) 113929 3 scenarios: i) 100 % application of evapotranspired water and conventional fertilizer application, FS1; and ii) 70 % application of irrigation and fertilization reduction to 70 %, limited input scenario FS2. This resulted in a factorial experiment set up as a randomized complete block design with 3 biofertilizer treatments (including an untreated control) x 2 fertigation levels with 12 replications (). During the fruiting period (i. e. January-June), ripe fruits from each plant were collected weekly and individually weighed to calculate plant dependent and treatmentdependent yield. After the development of this initial screening phase, the biofertilizer that gave the best results, in terms of fruit production, was selected for the development of Phase 2. For phase 2 completion, another set of strawberry plants were obtained at the beginning of the crop cycle. The plants were transplanted into 3.5 l plastic pots and allowed to grow in a greenhouse under the same conditions as those described above in the phase 1 section. After that, plants were subjected to two inoculation treatments (non-inoculation, control and monthly inoculation with PGPR-based biofertilizer in combination with the previously described fertigation scenarios (i.e., FS1 and FS2). The Phase 2 was also a factorial experiment established as a randomized complete block design with 2 biofertilizer treatments (including a non-inoculated control) and the 2 fertigation levels with 21 replications. The physic and chemical properties of the soil are given in Table 1. During the fruiting period (i.e. February-June), the photosynthetic variables were monitored bi-weekly using an infrared gas analyser (LI6800–01, LICOR Inc., Lincoln, NE, USA). Quantification net photosynthetic rate (A N ), stomatal conductance (g s ), instantaneous water use efficiency ( i WUE) and intercellular CO 2 concentration (C i ) was performed under the following leaf cuvette conditions: at 1000 µmol photons m −2 s −1 flux light density, 2.0–3.0 kPa vapour pressure, 50±0.5 % relative humidity and leaf temperature of 25 ±0.5 ◦C. In addition, coinciding with the gas exchange measurements, photochemical quantum efficiency of PSII (Φ PSII ) and electron transport rate (ETR) were measured with a portable fluorimeter FMS-2 (Hansatech Instruments Ltd., King’s Lynn, UK) (n =200, 4 treatments x 10 measurements/ treatment x 5 samplings). Furthermore, during the harvest season all the ripe fruits of each plant were collected, twice a week, and individually weighed. These evaluations of fruit production were completed with a fruit quality analysis in fruit samples collected at the end of the fruiting phase (n =20, 4 treatments x 5 measurements/treatment x 1 sampling in mid-May) through the analysis of primary metabolites. For this quantification, samples were processed following the methodology described by Roessner et al. (2001). Metabolite extraction, derivation, quantification and identification (organic acids, amino acids and sugars) was carried out by means of a GC-TOF-MS system (Lisec et al., 2006), consisting of an automated sample dispenser (CTC CombiPAL), a gas chromatograph (Agilent 6890 N) and a mass spectrometer (LECO, Pegasus II), with EI+mode powered on. Metabolites were identified by comparisons with standards (Kopka et al., 2005). During the development phases 1 and 2, irrigation and fertilization were carried out using compensating emitters with a flow rate of 5 L h -1 m -1 . The system was connected to a tank and a control device for adjusting the fertilizer doses and the amount of water applied during all experiments. Fertigation doses varied regardless of the need to apply a greater amount of fertilizer and water due to the greater phenological development of plants and the increase in atmospheric demand throughout the experiment. The water content of each pot was measured regularly using an ML3 ThetaProbe soil moisture sensor (Delta-T Devices Ltd) to check the effectiveness of the two treatments and the water difference between them during the time of the experiment. In addition, fertilizer N/P/K ratio composition was changed throughout the experiment to adjust to the phenological development of the crop and following farming management recommendations as follows: From November to March a fertilizer with N/P/K of 20/20/20 (Químicas Meristem, S.L.; Spain) and from April to June N/P/K proportions were 15/5/35 (Atl´ antica Agrícola.; Spain). Biofertilization was carried out with irrigation at the beginning of the experiment and was repeated once a month with irrigation water with a dose of 5 L ha -1 . After confirmation of the beneficial impact on fruit production of the biofertilizer selected in phase 1, and verification of this positive effect in terms of fruit quality and plant performance, a commercial strawberry production facility validation experiment was carried out to assess the potential of biofertilizer in a real operational agronomic context (TRL6). 2.3. Experimental phase 3: agronomic validation of PGPR-based biofertilizer in a commercial production facility trial (Technology readiness level 6, TRL6) In this experimental phase we evaluated the effect of PGPR-based biofertilizer previously tested on strawberry plant physiological performance and fruit production in a commercial farm located in Palma del Condado (Huelva, Spain 37◦361 N, 6◦56 W; Southwest Spain). This was tested under the following agricultural fertigation management scenarios: i) conventional agronomic fertigation (irrigation and fertilization schedule adjusted to the usual practice developed by the farmer without interference; control) and ii) limited fertigation (30 % reduction of conventional fertigation management carried out by the farmer). Specifically, this experimental monitoring was carried out in metallic macrotunnels covered with polyethylene film in the growing season 2021–2022 (October-June). At the beginning of the seeding season, plants were transplanted into raised beds of two rows, covered with impermeable polyethylene to reduce moisture losses by direct evapotranspiration, with a height of 35 cm and a width of 50 cm. The soil was previously sterilized by chemical fumigation (H 2 O 2 , 12 % v/v) and solarization. The irrigation was applied using T-Tape drip tapes with a flow rate of 5 L h −1 m −1 and connected to tanks for fertilization pulses and water application. The following factorial experiment was established as a randomized complete block design in four independent crop tunnels: 2 biofertilizer treatments (including a non-inoculated control) x 2 fertigation management scenarios. Tunnels were treated as blocks and there were 1800 plants per treatment combination). Tunnel dimensions and photographic material can be seen in the supplementary material (Fig. S1). The physic and chemical properties of the soil are given in Table 1. The levels of fertilization and irrigation of each specific management scenario were controlled and monitored throughout the study. Soil moisture and electrical conductivity were quantified using soil moisture probes (Decagon GS3, Decagon Devices, Inc.) and data were collected with a data logger Em50 (Decagon Devices, Inc.). The biofertilizer was Table 1 Physicochemical properties of soils used in Experimental Phases 1–2 and 3. Physicochemical properties Experimental Phases 1–2 3 pH 7.8 ±0.0 7.5 ±0.0 CE (mS/cm2) 160±4.3 185±1.5 Organic matter oxidable (%) 0.51±0.02 0.51±0.01 C (%) 0.38±0.01 0.48±0.02 N (%) 0.04±0.00 0.05±0.00 S (%) 0.01±0.00 0.01±0.00 NO 3 (mg/Kg) 172±5.7 188±9.7 NH 4 (mg/Kg) 5.9 ±0.6 8.6 ±1.3 P (mg/Kg) 38.2 ±0.3 47.9 ±4.6 CIC (cmol/Kg) 4.6 ±0.1 4.5 ±0.1 Ca (cmol/Kg) 7.5 ±0.1 6.2 ±0.1 Mg (cmol/Kg) 1.1 ±0.0 1.3 ±0.0 K (cmol/Kg) 0.6 ±0.0 0.6 ±0.0 Na (cmol/Kg) 0.1 ±0.0 0.1 ±0.0 Fe (mg/Kg) 16.0 ±0.2 24.0 ±0.2 Mn (mg/Kg) 1.6 ±0.1 ± Zn (mg/Kg) 1.8 ±0.0 1.8 ±0.1 Cu (mg/Kg) 6.2 ±0.1 4.9 ±0.1 Texture (%) 7/8/85 9/9/82 Values represent mean ±SE, n =5. Texture (silt/clay/sand). E. Mateos-Naranjo et al. Scientia Horticulturae 340 (2025) 113929 4 applied with irrigation water monthly throughout the crop cycle and with a dose of 5 L ha -1 . Finally, during the fruiting period (i.e. February-June gas exchange measurements collected during phase 2 were also recorded with 10 measurements per treatment at 5 sampling timepoints per month). Furthermore, simultaneously with these measures, all fruits in the experimental plots were harvested, twice a week, and individually weighed to calculate the plot-dependent and treatment-dependent yield. 2.4. Statistical analysis Software R ver. 4.0.3 (R Core Team, 2020) was used to perform the statistical analyses. Generalized linear models (GLM) were used to analyze the main and/or interactive effects of bacterial inoculation and fertigation management scenarios (as categorical factors) on fruit production and physiological parameters (as dependent variables) of strawberry plants in the different experimental phases. Multiple comparisons were analyzed by an LSD (post hoc) test. Furthermore, a Principal Component Analysis (PCA) was used to identify the main axes of the variation of the metabolites (organic acids, amino acids and sugars) variation within fertigation scenarios with inoculation treatment, using the functionalities of the package R ’funspace’ (Carmona et al., 2023). This analysis was carried out on four subsets corresponding to both fertigation scenarios tested (i.e., plants exposed to FS1 and FS2) in combination with the two inoculation treatments (non-inoculation control and inoculated). This analysis was followed by a downscaling assessment of specific metabolites variations regarding to bacterial inoculation and fertigation management scenarios using GLM analysis. 3. Results 3.1. Results of the experimental phases 1: Biofertilizer screening and selection There were significant effects of the variation in the fertigation level and the bacterial inoculation on the productivity of the strawberry plants. In this sense, focusing on the cumulated fruit production curves along the harvest period of the crop cycle, there were notable differences in the shape of the curves between each treatment combination (Fig. 3). Thus, our results showed a significant differential impact of each designed biofertilizer in the last months of the harvested period (i.e. April to June) under both fertigation management scenarios, while cumulated production did not differ greatly during the period between January and April. In this way, under optimal management conditions (i.e. FS1 treatment), the cumulative production curve of plants treated with biofertilizer 2 displayed significantly higher values than the curve obtained in non-inoculated plants and those inoculated with biofertilizer 1, which was the one that presented the lowest values (GLM: Inoc., p < Fig. 3. Results of the experimental phase 1: Panels (A) and (B) show cumulated production curves for fruit collected from strawberries plants subjected to three inoculation treatments (non-inoculation and monthly inoculation throughout the experiment, with two different PGPR-based biofertilizers, biofertilizers 1 and 2) in combination with two fertigation management scenarios, 100 % application of evapotranspirated water and conventional fertilizer application, FS1 and 70 % application of irrigation and fertilization reduction to 70 %, FS2. The values in each specific time are the sums of fruits collected from the 12 plants exposed to each specific experimental treatment. The curve region bounded between both arrows connected by the dashed line indicates significant differences between inoculation treatments (LSD test, p <0.05). Results of the experimental phase 2: panel (C) shows mean number of fruits per plant and panel (D) displays mean weight for fruits collected during the harvest period of each combination of fertigation and inoculation treatment. Values are mean ±standard error of twenty-one pots per treatment. FS., Inoc., or FS. x Inoc. in the upper part of the panel indicate the existence of principal or synergistic significant effects (GLM test, + p <0.05, + p <0.01). Different letters indicate means that are significantly different from each other (LSD test, p <0.05). E. Mateos-Naranjo et al. Scientia Horticulturae 340 (2025) 113929 5 0.05; Fig. 3A). Thus, total production was significantly higher with biofertilizer 2 compared to control and fertilizer 1 treated plants (GLM: Inoc., p <0.01) in global terms; increasing by 13 % and 23 %, respectively. Furthermore, we found that limited input treatment (i.e. FS2 treatment) led to a clear reduction in plant production; however, this effect was significantly mitigated by the biofertilizer applications, and mainly by biofertilizer 2 which was therefore selected for the second experimental phase development (GLM: Inoc., p <0.05; Fig. 3B). 3.2. Experimental phase 2 results: Biofertilizer validation under greenhouse conditions The development of the experimental phase allowed us to corroborate the positive effect of the application of biofertilizer 2 on the production yield of strawberries. Biofertilizer 2 lead to significant improvements in fruit production of 12 % and 21 % in both grown scenarios, respectively, compared with their non-inoculated counterparts. Furthermore, we were able to corroborate that this positive effect was mediated by the significantly greater number of fruits per plant (GLM: FS. x Inoc., p <0.05; Fig. 3C) and not due to a variation in fruit size characteristics, indicated that the weight of the fruit did not vary between inoculated and non-inoculated plants (GLM: Inoc., p >0.05; Fig. 3D). Furthermore, a positive impact of inoculation could also be identified on the physiological performance of strawberry plants. In this regard, we found that there were major and synergistic significant effects on the net photosynthetic rate (A N ) during the whole sampling period (GLM: FS., p <0.01; Inoc., p <0.01; Fig. 4A). Results revealed that this effect was mitigated to some extent with the application of biofertilizers although overall limited inputs led to reduction in A N in strawberry plants from each specific sampling period (GLM: FS. x Inoc., p <0.05; Fig. 4A). Furthermore, the highest A N values were recorded in plants added with biofertilizer and grown under FS1 treatment (Fig. 4A). A very similar trend was recorded for g s , with higher values with biofertilizer application and FS1 treatment, but differences were only significant compared to plants not inoculated and subjected to FS2 and with the rest of the treatments only in the May sampling (Fig. 4B). Intercellular CO 2 did not differ between FS treatments, but overall plants Fig. 4. Results of the experimental phase 2, net photosynthetic rate, A N (A), stomatal conductance, g s (B), intercellular CO 2 concentration, C i (C), intrinsic water use efficiency, i WUE (D), quantum efficiency of PSII, Ф PSII (E) and electron transport rate, ETR (F) of strawberry leaves throughout the sampling period (January-May) in plants subjected to two inoculation treatments (non-inoculation and inoculated with biofertilizer 2) in combination with two fertigation management scenarios, FS1 and FS2 (for details see caption of Fig. 3). Values are mean ±standard error of ten replicates per treatment and sampling. FS*., Inoc**., or FS. x Inoc***. in the upper part of the panel indicate the existence of principal and/or synergistic significant effects (GLM test, p <0.05). Asterisks above each sampling period indicates significant differences between experimental treatments (LSD test, p <0.05), specifying whether these differences are due to fertigation management (*), to inoculation (**) and/or interaction among both factors (***). E. Mateos-Naranjo et al. Scientia Horticulturae 340 (2025) 113929 6 supplied with bacterial biofertilizer showed the lowest C i (GLM: Inoc., p <0.01; Fig. 4C). Intrinsic water use efficiency ( i WUE) values remained relatively constant throughout the sampling period in inoculated plants, regardless of the type of management, with values ranging between 66 and 55 µmol CO 2 mol H 2 O -1 (Fig. 4D). In contrast, this variable decreased in non-inoculated plants during the sampling periods from February to May, with the lowest values recorded in the February, March, and May samplings for plants subjected to FS2 treatment compared to the other treatments (GLM: Inoc., p <0.01; Fig. 4D). In addition, the overall Φ PSII and ETR values were significantly lower in plants subject to a limited nutrient and irrigation input scenario (GLM: FS., p <0.01; Fig. 4E and F). However, the application of biofertilizer contributed to counteract this negative effect and even allowed this group of plants to have significantly values higher than those of the control treatment in all sampling periods (GLM: FS. x Inoc., p <0.01), except in the sampling of May when the highest values were obtained in inoculated plants subjected to FS1 treatment (GLM: Inoc., p <0.05; Fig. 4E and F). Finally, our results revealed significant effects of fertigation management and biofertilizer application on fruit quality. These were seen in terms of metabolite profile pattern variations (i.e. organic acids, amino acids and sugars content) due to experimental treatments combination. Thus, our PCA ordination analysis showed a significant association between metabolites patterns and fertigation level variation, which was modulated by biofertilizer application, explaining 73.5 % of the proportion of variation of the recorded data in each treatment combination (i.e. PC1 and PC2 45.0 % and 28.5 %, respectively; Fig. 5A–D). The bidimensional plots revealed a clear divergence between both fertigation level along PC1 (i.e. FS1 vs FS2). The fruits from plants grown with FS2 treatment were most likely found in metabolite trait space where concentrations of organic acids, amino acids, and sugars were reduced, except for the succinate, which was significantly higher in non-inoculated plants (i.e. FS2 N-In, Fig. 5A–D; GLM: FS., p < 0.01; supplementary Fig. S2 L). However, the addition of biofertilizer 2 modulated this response as shown in the clear divergent position of fruit sampled from inoculated plants (Fig. 5A). This was associated with an overall higher concentration of many of the metabolites evaluated, such as asparagine, GABA, citrate, fructose and glucose, in plants grown under FS1 treatment compared to noninoculated (N-In) plants (GLM: Inoc. P <0.01; supplementary Fig. S2B, C, I, M, N). The response pattern observed was also maintained under nutrient and irrigation limitation scenarios with respect to non-inoculated plants (Fig. 5C) and was even extended for other metabolites such as glutamine, isoleucine, leucine, valine, lactate, malate and sucrose (GLM: Inoc. p <0.01; supplementary Fig. 5. Results of experimental phase 2, representation of the principal component analysis (PCA) biplot obtained for the fruit metabolite content (i.e., organic acids, amino acids and sugars content) collected in strawberries plants in May. Probabilistic distribution of samples within the metabolite trait space is separated for the inoculated (In, A, C) and non-inoculated (N-In, B, D) plants subject to two fertigation management scenarios, FS1 and FS2 (for details, see caption of Fig. 3). The grey labelled lines show the 0.99, 0.50 and 0.25 quantiles of general additive model predictions, corresponding to probability values. E. Mateos-Naranjo et al. Scientia Horticulturae 340 (2025) 113929 7 Fig. S2D, E, F, H, I, J, K, O). 3.3. Results of experimental phase 3: Commercial production facility trial for agronomic validation of biofertilizer The monitoring of fertigation levels of each specific management scenario tested during the commercial production facility trial phase indicated that the levels of soil water content (SWC) and electrical conductivity (CE) did not vary between the two FS treatments in the period between the beginning of the study and the beginning of March. After early March, both variables dropped and increased, respectively, reaching the percentage of variation c. 30 % between both management treatments in the last two months of harvested period (i.e. April and May; supplementary material; Fig. S3A, B). Regarding the yield results, the cumulated fruit production curves of each experimental treatment combination indicated that strawberry production did not significantly vary much during the period between January and early April (Fig. 6A). In this period, the shape of the curves started gradually to change until the end of the collection period, when significant differences were recorded with the highest production in inoculated FS1 plots (GLM: Inoc., p <0.05; Fig. 6A). Thus, our results indicated that biofertilizer applications significantly increased global plot productivity yield during the crop cycle. The percentages of production were 10 % and 8 % higher in plots subjected to FS1 and FS2 treatments, respectively, compared to their non-inoculated plots. However, it should be denoted that these productivity variations were due to the positive biofertilizer effect in the harvested period from April to May, which was indicated by the significantly greater mean production per pot recorded in biofertilizer plots at both FS management scenarios (GLM: Inoc., p <0.05; Fig. 6B). On the other hand, our gas exchange results revealed that overall, A N values tracked in each sampling periods remained relatively constant in inoculated plants grown in plots with conventional fertigation management (i.e. FS1 treatment). Meanwhile, A N values significantly Fig. 6. Results of the experimental phase 3 (agronomic commercial production facility trial): Panel (A) cumulated production curves for fruit collected from strawberries plants subjected to two inoculation treatments (non-inoculation and inoculated with biofertilizer 2) and grown in plots with two different fertigation management scenarios, FS1 and FS2 (for details, see caption of Fig. 3). Values in each specific time are accumulated sums of fruit collected from the 1800 plants of each specific experimental treatment. Panel (B) mean production in strawberries plants per plot of each combination of fertigation and inoculation treatment for fruit collected in the harvest period included between the red arrow onwards (i.e., early April-May). Values are mean ±standard error of four plots per treatment. FS., Inoc., or FS. x Inoc. in the upper part of the panel indicate the principal or synergistic significant effects (GLM test, + p <0.05, + p <0.01). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) E. Mateos-Naranjo et al. Scientia Horticulturae 340 (2025) 113929 8 decreased to a greater extent in non-inoculated plants grown in plots subjected to limited input management throughout the various sampling periods, compared to the other treatments (GLM: FS. x Inoc., p <0.01; Fig. 7A). Likewise, significant differences were recorded in April and May samplings between plants grown in inoculated plots with FS1 management with respect to non-inoculated plants subjected to the FS1 scenario and those inoculated and grown with water and fertilization limitation, which showed similar assimilation values (GLM: FS. x Inoc., p <0.01; Fig. 7A). A very similar pattern was observed for g s . However, the g s values only varied in non-inoculated plants grown in FS2 plots and the rest of the treatments in the April and May samplings (GLM: FS. x Inoc., p <0.01; Fig. 7B). Meanwhile, the values of intercellular CO 2 concentration (C i ) and intrinsic water use efficiency (iWUE) were relatively constant in all sampling periods regardless of the management of the fertigation and the application of biofertilizer treatment, showing mean values c. 225 and c. 105 µmol CO 2 mol H 2 O -1 , respectively (Fig. 7C and D). Finally, Φ PSII and ETR values were overall lower in non-inoculated plants compared to their inoculated counterparts. Nevertheless, there were only significant differences in plots with limited input management in all sampling periods, except in January (GLM: FS., p <0.01; Inoc., p <0.01; Fig. 7E and F). 4. Discussion Systematic addressing of TRLs phases is required to develop a technology from initial conception through the different stages of research and development to achieve a true commercial scale implementation (EU 2020). The interest in the use of PGP bacteria as biofertilizers to improve the efficiency of intensive agricultural practices has increased systematically in recent years. However, most available studies that focus on the interactions of PGPB and plants were carried out in vitro, within growth chambers or within greenhouses. Only a few publications contain field trial data (Jarak et al., 2012; Kumar et al., 2016; Bacilio et al., 2017). This fact has made that technology readiness levels achieved is in medium-low phases (i.e. TRLs 1–3), so it is necessary to focus on higher TRLs phases to achieve true implementation. In this study, we aimed to reach levels framed between TRLs 4–6, which addresses the design and development of a definitive prototype of PGPR-based biofertilizers. These biofertilizers were derived from halophytes and their highly stress-resistant microbiomes, intended to improve the efficiency of agronomic practices in strawberry cultivation. Fig. 7. Results of the experimental phase 3 (agronomic field trial), net photosynthetic rate, A N (A), stomatal conductance, g s (B), intercellular CO 2 concentration, C i (C), intrinsic water use efficiency, i WUE (D), quantum efficiency of PSII, Ф PSII (E) and electron transport rate, ETR (F) of strawberry leaves throughout the sampling period (January-May) from plants subjected to two inoculation treatments (non-inoculation and inoculated with biofertilizer 2) and grown in plots with two different fertigation management scenarios, FS1 and FS2 (for details, see caption of Fig. 3). Values are mean ±standard error of ten replicates per treatment and sampling. FS., Inoc., or FS. x Inoc. in the upper part of the panel indicate the principal or synergistic significant effects (GLM test, + p <0.05, + p <0.01). Different letters indicate means that are significantly different from each other within each specific sampling period (LSD test). E. Mateos-Naranjo et al. Scientia Horticulturae 340 (2025) 113929 9