PGPR strains and culture media

The PGPR strains were isolated from rhizosphere of the grass halophyte Distichlis spicata L. The bacterial strains used were: LBEndo1, NFbEndo2M2, KBEndo3, and KBEcto4. The LBEndo1, KBEndo3, and KBEcto4 strains were characterized previously as PGPR, these strains improved growth under standard and saline conditions, which correlated with IAA and siderophore production, as well as phosphate solubilization (^{Palacio-Rodríguez, 2015}; ^{Espinosa-Palomeque et al., 2017}; Palacio-Rodríguez et al., 2017; ^{González-Rodríguez et al., 2018}; Espinosa-Palomeque et al., 2019). Stock of PGPR strains were stored -70 ºC in 30% glycerol and before being used were grown at 30 ºC and 180 rpm overnight in Luria Broth medium. Bacterial suspensions were adjusted to 1x10⁸ CFU ml^-1 before making the inoculations to tomato plants.

Shade house experiments

The experiments were conducted in the commercial shade houses of the company ‘Agrícola Vigo’, Ejido El Pilar, Matamoros, Coahuila. The site is located at 25º 72’ 38.33” north latitude, 103º 32’ 73.08” west longitude. Tomato seeds (Solanum lycopersicum L. cv ‘Top1182’) were germinated in transplant tray (200 inverted pyramid cells of 2.5 X 6.5 cm, side length X depth) filled with peat moss (Lambert peat moss, Inc, Quebec, Canada). 43 days after germination the plantlets were inoculated by immersion in a bacterial suspension of 1x10⁸ CFU ml^-1 one by one. After three days of inoculation the tomato plants were transplanted to shade house conditions.

The experimental design in shade house was divided into two conditions according to soil preparation (characterized by being a clayey soil): the first one, soil in flat was enriched with coconut fiber (15 kg m^-2) and composted cattle manure (12.5 kg m^-2), there were four repetitions of 200 tomato plants by each PGPR inoculated, plus a control row with 200 tomato plants uninoculated. Beds soils ((30 cm high) were used for the second condition, four repetitions of 200 tomato plants by each PGPR inoculated, plus a control row with 200 tomato plants uninoculated (Table 1).

All the tomato plants were planted in rows with 18 cm between tomato plants and 1.8 m between rows. The tomato plants were drip-irrigated with a fertilizer solution containing (mg L^-1): N 2 311, Ca 626, Mg 1 158, K 723, and micronutrients 15. The plants were re-inoculated two months after the first inoculation at beginning of bloom stage (63 days after transplantation on shade-house conditions), the second inoculation was performed using a manual 20 L knapsack sprayer with 40 ml of 1x10⁸ CFU ml^-1 of each bacterial suspension by tomato plants.

Biometric parameters and yields tomato

Throughout the experiment, two samples were taken, the effect of PGPR in the plant growth was determined in the first sampling; three plants by treatment were harvest and shoot length, root fresh weight, stem diameter, fruit weight, and number of fruits per plants were determined after 59 days of transplanted to shade house (62 days after first inoculation).

Yield and quality evaluation was determined in the second sampling, 87 days after transplanted to shade house (90 days after first inoculation). Twenty red tomato fruits were randomly sampled of harvest total by treatment. Fruit size was determined by measuring fresh weight and equatorial diameter of 20 red fruits harvested. Tomato fruits equatorial diameter was measured with a caliper (Scala, Inox 222B, México) and divided into four categories of marketable classification: extra-large (>67 mm), large (54-67 mm), medium (47-54 mm) and small (<47 mm) according to the ^{Jones (1999}) scale. Two of twenty tomato fruits were cut in half by equatorial diameter to visually examine the size and internal appearance (Table 2).

Statistical analysis

Data were obtained under completely randomized design with four replications number. The data of measured parameters were analyzed using analysis of variance Anova with the help of software Microsoft Excel complement XLSTAT. Means were tested by minimum significant difference (MSD) at p≤ 0.05.

Molecular identification of PGPR by 16S rRNA

The four rhizobacteria were identified by molecular analysis, the LBEndo1, and KBEcto4 rhizobacteria were reported in ^{Palacio-Rodríguez et al. (2017}). NFbEndo2M2, and KBEndo3 rhizobacteria were subjected to 16S rRNA gene sequence analysis. The DNA was extracted by CTAB technique according to the method of ^{Doyle and Doyle (1990}). Partial amplification of 16S rRNA gene was made by PCR technique using the 27F oligonucleotides 5’AGAGTTTGATCMT GGCTCAG 3’ and 1492R 5’GGTTACCTTGTTACGACTT 3’, the PCR product was purified using the kit AxyPrep DNA gel Extraction kit (Axygen) and then sent to be sequenced to McLAB in San Francisco, CA, USA. The sequences obtained from sequencing underwent compared using Blast (NCBI) (http://blast.ncbi.nlm.nih.gov/Blast.cgi) to determine the taxonomy of the bacterial strains (^{Weisburg et al., 1991}; ^{Altschul et al., 1997}).

Effects on growth promotion of tomato plants

In this study, the effect of four PGPR strains on the growth of tomato plants was investigated under shade house conditions using flatted soil added of coconut fiber and compost and raised beds soil without compost. Significant increases in root fresh weight were observed with the inoculations of PGPR in special with LBEndo1 and KBEcto4 strains. The inoculation of tomato plants with PGPR promoted the vegetative growth, increase of growth was observed on shoot length, root weight, and stem diameter.

Even though no marked differences were observed in the increase in shoot length and root weight, increase in root weight by the strains LBEndo1 and KBEcto4 on composted flatted soil conditions, by 119.3 and 81.9%, respectively, compared to tomato plants control uninoculated (Table 3 and Figure 1). The root of tomato plants inoculated with PGPR showed that they modified the architecture and increased growth of plants roots (Figure 1). Root fresh weight was increased significantly in tomato plants inoculated with LBEndo1, NFbEndo2M2, and KBEcto4 compared to controls uninoculated (Figure 1).

Figure. 1 Root systems of tomato plants: flatted soil preparation (A-E) and raised beds soil preparation (F-J); uninoculated control (A and F); LBEndo1 inoculated (B and G); NFbEndo2M2 (C and H); KBEndo3 (D and I); and KBEcto4 (E and J). 

Our results agree with those reported by ^{Gamalero et al. (2002}, 2004), who worked with Pseudomonas fluorescens 92rk and P. fluorescens A6RI increased the root fresh weight and affected root architecture when they are inoculated in tomato plants. It Also agrees with what was reported with ^{Sharma et al. (2015}) reported a significant increase in plant growth parameters (length and weight of shoot and root) in tomato seedlings treated with Bacillus subtillis strain S25, which is a strain with antagonistic activity to Phytophthora capsici.

Stem diameter of tomato plants was increased on composted flatted soil conditions by PGPR strains LBEndo1 and KBEcto4 inoculation by 22.17 and 22.42%, respectively, for raised beds soil conditions the PGPR strains NFbEndo2M2 and KBEcto4 inoculation increased significantly by 24.71 and 11.11%, respectively, in relation to control plants uninoculated (Table 3). The increased of stem diameter in tomato plants due to inoculation of PGPR agrees with reported by ^{Zulueta-Rodríguez et al. (2020}), who reported an evident increase of up to 15% in stem diameter in tomato seedlings inoculated with Bacillus subtilis in a chapel-like greenhouse under semi-hydroponic conditions.

The PGPR treatments also increased fruit number per plant on both soil condition preparations. KBEcto4 was the treatment with the highest number of fruits with 23 tomatoes per plant, compared with 18.6 fruits per plant control uninoculated on composted flatted soil conditions. The increase in the growth of tomato plants by PGPR inoculations was already reported (^{Gamalero et al., 2002}, ²⁰⁰⁴; ^{Kokalis-Burelle et al., 2002}; ^{Gravel et al., 2007}; ^{Mena-Violante and Olalde-Portugal, 2007}; ^{Felici et al., 2008}; ^{Pastor et al., 2014}; ^{Sharma et al., 2015}). There are few reports about evaluation of crops in shade house conditions, ^{Yu et al. (2011}) reported that Pseudomonas chlororaphis and Pseudomonas fluorescens increased the phosphorous and nitrogen uptake, and height, shoot and root dry weight of walnut seedlings under shade house conditions. However, the present study reports for the first time the effect of rhizobacteria LBEndo1, NFbEndo2M2, KBEndo3, and KBEcto4 on the growth of tomato plants under commercial shade house conditions. Moreover, the soil preparation and organic fertilizer added were also shown to have a stimulating effect on the growth in tomato plants.

Yield and marketable tomato fruit

After 12 weeks of shade house experiments, the size and the weight of the tomato fruit was obtained of 20 randomly red fruits of each treatment and the respective control uninoculated of both soil preparations variants (flatted and raised beds soil). Yield of the tomato plants was increased significantly in two of four PGPR strains used. As shown in Table 4, strains LBEndo1 and KBEcto4 had the ability to increase the fruit weight and equatorial diameter on both soil preparations; raised beds and flatted soil supplemented with organic fertilizer where the increase in size is more evident in tomato fruits.

The average weight of tomato fruits per plant treated with LBEndo1 (236.27 and 225.8 g corresponding to 40.8% and 48.4% respectively more than control) and KBEcto4 (240.22 and 211.41 g corresponding to 43.1% and 38.9%, respectively, more than control) strains in both soil preparations were higher than those of others including the uninoculated control (Table 4). The results in the present study exceed those reported by ^{Katsenios et al. (2021}), where the increased of mean tomato fruit weight per plant by 30.7, 28.81, 27.52, and 26.78% by the inoculation with Bacillus subtilis, Bacillus amyloliquefaciens, Priestia megaterium, and Bacillus licheniformis, respectively compared to control on the cultivation of industrial tomato.

The equatorial diameter of tomato fruit in LBEndo1 inoculated treatment was increased 16.4% (7.623 cm) and 15.2% (7.423 cm) compared to control uninoculated (6.545 and 6.445 cm) for flatted and raised beds soil, respectively (Table 4; Figure 2). The KBEcto4 treatments increased the equatorial diameter by 16.5% (7.625 cm) and 11.1% (7.166 cm) compared to controls of composted flatted soil conditions (6.545 cm) and raised beds soil (6.445 cm), respectively (Table 4 and Figure 2).

Figure. 2 Tomato fruits cut in half: flatted soil preparation (A-E) and raised beds soil preparation (F-J); uninoculated control (A and F); LBEndo1 inoculated (B and G); NFbEndo2M2 (C and H); KBEndo3 (D and I); and KBEcto4 (E and J). 

This effect is also observed in the study carried out by ^{Espinosa-Palomeque et al. (2017}) with the same strains that we report here under greenhouse conditions, where increases in the equatorial diameter of tomato fruits are reported in the treatment with the LBEndo1 strain corresponding to Bacillus paranichelliformis. ^{Gül et al. (2008}) reported the effect of two commercial strains of Bacillus amyloliquefaciens (FZB24 and FZB42) on tomato yield, the application of either strain increased the yield by 8-9%. In a similar way, enhance of yield has been reported in tomato plants inoculated with PGPR Bacillus subtilis BEB-13bs, increasing yield and marketable grade yield (21%) compared to control (^{Mena-Violante and Olalde-Portugal, 2007}).

Marketable tomato fruits yields were increased significantly in all the treatments of composted flatted soil conditions compared to control tomato fruits and for raised beds soil the LBEndo1 and KBEcto4 were higher than control uninoculated. In the same way as in the previous parameters the LBEndo1 was higher with 95% (composted flatted soil conditions) and 85% (raised beds soil conditions) of extra-large marketable grade yield (Table 2 and Figure 2). The Figure 2 showed tomato fruit cut in half by the equatorial diameter, in which the size of the fruit is observed between treatments, the largest fruits are observed for the LBEndo1 and KBEcto4. Although, not enough observations of tomatoes fruits cut in a half, the locules number is higher in PGPR treatments in comparison with tomatoes fruits uninoculated; this is important in the texture and firmness of fruits. ^{Espinosa-Palomeque et al. (2017}) determined the firmness of tomato fruits by penetrometer device, finding that tomatoes from inoculated plants have greater firmness compared to controls without inoculation.

Identification of selected strain by 16S rRNA sequencing

The selected rhizobacteria were identified by PCR amplification and sequencing of 16S rRNA gene. The LBEndo1, and KBEcto4 rhizobacteria were reported in ^{Palacio-Rodríguez et al. (2017}). NFbEndo2M2, and KBEndo3 rhizobacteria were subjected to 16S rRNA gene sequence analysis. The 16S rRNA gene partial sequence obtained of the sequencing were subject to homology analysis in BLAST (NCBI Database), which retrieve taxonomic positions of rhizobacteria strains, LBEndo1 is like Bacillus paralicheniformis with a homology 96%, NfbEndo2M2 was 98% like Acinetobacter guillouiae, KBEndo3 had a similarity of 99% to Aeromonas caviae and KBEcto4 exhibited 99% homology with Pseudomonas lini (Table 5).