Rhizobacteriome Diversity and Morphophysiological Features of Three Tomato Plant Varieties Under Nitrogen Deficiency

Abstract

The increasing biogeochemical imbalance of nitrogen (N) heightens the importance of studying rhizosphere bacteria, which aid crop nutrient uptake, and their responses to N deficiency. The aim of the study was to assess variety-specific responses of the tomatoes and their associated rhizobacteriome to low N availability. Three determinant varieties of Solanum lycopersicum L. were cultivated in pot-scale experiment during 10 weeks on low-fertility substrate (sod-podzolic soil:peat:clay:sand, 1:1:1:2, v/v), half of which were supplemented with ammonium nitrate (60 mg N kg⁻¹ in total). A comprehensive methodology was employed, including 16S rRNA metagenomic Nanopore sequencing, quantitative assessment of N-cycling bacteria, and analysis of plant growth, photosynthetic pigments, total N in biomass, and fine root architecture. Results demonstrated that N deficiency significantly reduced plant biomass and photosynthetic pigments. The rhizosphere contained a diverse community of N-transforming bacteria (38 identified genera), whose composition and relative abundance were strongly influenced by both tomato variety and N fertilization. Nitrogen application increased the abundance of N-fixers and altered alpha-diversity in a variety-dependent manner. Significant correlations were found between the abundance of key bacterial genera (e.g., Stenotrophomonas, Rhizobium) and N parameters in plants and substrates. The study concludes that the response of the tomato rhizobacteriome to N availability is variety-specific, which is important for the development of microbiome management strategies for enhancing N use efficiency.

1. Introduction

Nitrogen (N) is an essential biophilic element required for plant growth, as it is a fundamental component of vital biomolecules like proteins, nucleic acids, phospholipids, etc. [1,2,3]. As N nutrition is a major factor limiting crop productivity, the use of N fertilizers constitutes a significant portion of costs in the agricultural sector [2,4].

Soil N consists of compounds from three groups, differing in their availability to plants. The first group includes mineral N compounds (ammonium and nitrates), which are directly accessible to plants. The second group consists of easily hydrolyzable organic N compounds (such as amino acids, urea, and uric acid), which are readily mineralized into plant-accessible forms. However, most of the soil N, is a component of non-hydrolyzable organic compounds like humic substances, which are resistant to mineralization [1,2,3]. The balance of N in arable soils is governed by opposing processes. Inputs occur through the mineralization of organic matter (including humus) and the biological fixation of atmospheric N. These are counteracted by outputs, including crop uptake, microbial immobilization, leaching, and gaseous losses via denitrification [5,6]. Although the atmosphere is a vast reservoir of N, plants cannot utilize it directly. Consequently, the microbial processes that convert this inert gas into usable soil compounds are a major focus of agricultural research [6,7].

Rhizosphere bacteria can promote plant growth by increasing the availability of mineral elements, among other mechanisms. Therefore, the rhizosphere microbiome is considered an important tool for enhancing nutrient uptake in agricultural crops [2,8]. Bacteria involved in the N cycle are particularly important for the biogeochemical cycling of N and for providing N nutrition to plants [1,6,7]. These N-transforming microorganisms are typically classified by their primary process (e.g., N-fixers, nitrifiers, denitrifiers, etc.). Some can perform multiple redox reactions, while others specialize in a single process [6].

The composition of bacterial communities is influenced by both plant species, genotype, physiological state, and age, as well as soil physicochemical characteristics [4,9,10]. By providing carbon through root exudates, litter, and plant residues, plants can regulate the abundance, diversity, and functional activity of microorganisms that transform N into plant-accessible forms [1,2]. Previous studies have reported that N fertilizer application alters the composition and abundance of taxa within these communities, reducing bacterial diversity in the rhizosphere of common beans [11], wheat [12], and corn [13]. Under optimal N fertilization in the rice rhizosphere, a decrease in specific denitrifying bacteria was observed, alongside an increase in nitrifying bacteria, a rise in the relative abundance of Proteobacteria (syn. Pseudomonadota) and Actinobacteria, and a decline in Firmicutes and Bacteroidetes (syn. Bacteroidota) [14].

Efficient N uptake and assimilation are critical for plant vitality and productivity [2,15]. However, crop plants typically utilize only about 30% of the N applied to soil, while over 60% is lost through leaching, runoff, and denitrification [16]. One approach to addressing N deficiency is to identify crop species/varieties with root systems capable of efficient N absorption [2,17], particularly under conditions of low N availability. Root system architecture, which determines root placement, is a major factor regulating N uptake. Crops with a greater N uptake capacity are more productive in N-limited agroecosystems [2,8]. Furthermore, root structure influences the supply of carbon compounds to the rhizobiome, which affects the abundance and functional composition of microorganisms. However, plant–microbial interactions in this context remain insufficiently studied [1,16].

Tomato (Solanum lycopersicum L.) is a widely cultivated crop due to its high content of vitamins, minerals, and antioxidants, including phytochelatins [10,18]. Tomatoes can be grown in open fields, greenhouses, and hydroponic systems, enabling resource-efficient production [4,19,20]. Additionally, the tomato serves as a convenient model organism in scientific research [4,18].

The bacterial microbiome of S. lycopersicum has been the focus of numerous international studies in recent years [4,10,20,21,22,23]. It has been shown that tomatoes host a diverse bacterial community inhabiting not only the roots but also other organs, with Proteobacteria being the most dominant phylum [4,20,21]. However, the contribution of N-cycling bacteria to the diversity of the rhizosphere bacteriome is poorly understood. Furthermore, information on how bacterial communities change in response to N fertilizer application is contradictory. Previous studies on the morphophysiological parameters of tomato adventitious roots revealed that roots of different orders vary in their capacity for nutrient absorption and transport [24]. In contrast, the structure of lateral fine roots has been less studied [25], despite their greater importance for plant mineral nutrition and involvement in carbon translocation into the rhizosphere bacteriome.

The specific objectives of this study are: (a) to analyze the rhizobacteriome diversity of three tomato varieties grown in a low-fertility substrate with and without N application; (b) to quantitatively assess N-transforming bacteria and determine their relative abundance among the total bacterial genera in the rhizosphere substrate; and (c) to conduct a comparative assessment of plant growth parameters, photosynthetic pigment content, structural characteristics of fine roots, and N accumulation in aboveground and underground organs.

2. Materials and Methods

2.1. Plant Material and Experimental Design

Three determinate tomato varieties (S. lycopersicum) suitable for pot cultivation were selected for this study: ‘Balkonnoe Chudo’ (SL-v.1), ‘Komnatniy Sibiryak’ (SL-v.2), and ‘Voroniy Glaz’ (SL-v.3). The first two varieties were purchased from the Agrofarm “Semena Altaya” (Barnaul, Russia), and the third from the Agrofarm “Aelita” (Moscow, Russia). All three are early-ripening, medium-leafed varieties that do not require staking or pruning. They form compact, low-growing bushes, 50–60 cm tall for SL-v.1 and SL-v.2, and 30–35 cm for SL-v.3. These varieties are suitable for both open and protected ground, including container gardening, and produce small, round, sweet, and aromatic cherry-type fruits, valued for their taste and ornamental qualities.

Plants were grown for 10 weeks (March–May 2024) under natural light at a room temperature of 23 ± 3 °C. The low-fertility growth substrate (“−N” treatment) was prepared from sod-podzolic soil (from the humus horizon, pH_H2O = 5.7, total N—0.25%), neutralized high-moor peat (LLC “Seliger-Agro”, Tver, Russia), clay, and river sand in a volumetric ratio of 1:1:1:2 (v/v). The resulting substrate was circumneutral (pH_H2O = 6.7) with a total N content of 0.15%. The substrate was placed in 1.2 L plastic pots, and tomato seeds were sown directly into them. An unplanted control, consisting of substrate only (bulk soil), was also prepared. After germination, six healthy seedlings were left in each pot. This high planting density was used to induce severe N deficiency during plant growth. Phosphorus was applied once to all pots 14 days after sowing as Ca(H₂PO₄)₂ solution (20 mg P per kg of soil). Additionally, half of the pots received N fertilizer (“+N” treatment) in the form of NH₄NO₃ solution (30 mg N per kg of soil), applied twice with a two-week interval (Figure 1).

Figure 1. Schematic of the pot-scale experiment. Three tomato varieties were grown for ten weeks in a low-fertility substrate, with (+N) and without (−N) nitrogen application. An unplanted bulk soil control was also included.

Each treatment was replicated in three independent pots. All pots (with and without plants) were watered with distilled water as needed (2–3 times per week) to maintain substrate moisture. Upon conclusion of the experiment, plants were carefully removed from the pots, and the root system with adhering soil was defined as the rhizosphere. Rhizosphere soil and unplanted bulk soil (composited from the three replicate pots per treatment) were collected in sterile 50 mL polypropylene Falcon tubes for subsequent microbiological analysis and metagenomic sequencing. The remaining substrate from each pot was air-dried, crushed, and sieved (0.25 mm mesh) for the analysis of pH and different N forms. The plants were gently washed to remove soil particles and separated into shoots and roots for morphometric and physiological analyses.

2.2. Soil DNA Extraction, Nanopore Sequencing, and Bioinformatic Analysis

Total DNA was extracted from 250 mg of rhizosphere soil using the SKYAMP Soil DNA Kit (EDC336, “SkyGene”, Moscow, Russia) according to the manufacturer’s instructions. DNA quality and quantity were assessed by electrophoresis on a 1.0% agarose gel and spectrophotometrically using a Nano-500 instrument (Allsheng, Hangzhou, China) by measuring absorbance at 260, 280, and 235 nm. Soil DNA samples were diluted 40-fold, and analytical PCR targeting the 16S rRNA gene (with primers 27F 5′-AGAGTTTGATCCTGGCTCAG-3′ and 1492R 5′-ACGGYTACCTTGTTACGACTT-3′) was performed using an HS Taq DNA polymerase kit (PK018, “Evrogen”, Moscow, Russia) to confirm the absence of PCR inhibitors. DNA samples were stored at −80 °C.

A sequencing library was prepared using the 16S Barcoding Kit 24 V14 (SQK-16S114.24, “Oxford Nanopore Technologies” (ONT), Oxford, UK) with 10 ng of total DNA per sample, following the manufacturer’s protocol. The resulting libraries were sequenced on a R10.4.1 flow cell (FLO-MIN114, ONT) on a GridION instrument (ONT, Oxford, UK) until the desired depth was achieved. Basecalling was performed on a MinION Mk1C (version 22.10.7) using Guppy (version 6.3.9) in high-accuracy mode, with barcode trimming and a minimum quality score threshold of 9. Each sample was sequenced independently three times to ensure reproducibility.

Read quality was assessed using FastQC (version 0.12.1) and SeqKit (version 2.4.0). Sequences shorter than 600 nucleotides were removed, and reads with a quality score below 15 were trimmed using Trimmomatic (version 0.39) and fastp (version 0.23.4). Chimeric reads were removed using Vsearch (version 2.29.2). The final read lengths ranged from 1431 to 1466 nucleotides, with 74–78% of nucleotides achieving a Q20 quality score or higher. Taxonomic annotation was performed using Kraken2 (version 2.1.1) with the Silva reference database (version 138.1). Bracken (version 3.0) was used to estimate genus abundance in each sample, applying a relative abundance threshold of 0.05%. Heatmaps were generated in Python 3.10 using the pandas, numpy, seaborn, and matplotlib packages. Alpha diversity indices, including Chao1 [26] and Shannon [27], were calculated in R (version 4.1.2) using the phyloseq package.

2.3. Isolation and Enumeration of Microorganisms, Including N-Cycling Rhizobacteria

The total quantity of mesophilic aerobic and facultative anaerobic microorganisms (QMAFAnM) and N-fixers in the rhizosphere and unplanted bulk soil was determined using the plate count method [28]. A microbial suspension was prepared by mixing 10 g of soil with 90 mL of sterile 0.85% NaCl solution and shaking at 180 rpm for 10 min at 28 ± 2 °C. After serial ten-fold dilutions, 1 mL aliquots were plated onto solid media and incubated for 3–5 days at 28 ± 2 °C. Inoculations were performed using 4–6 dilutions, with two technical replicates per dilution. QMAFAnM was determined on meat-peptone agar (“HiMedia”, Mumbai, Maharashtra, India), and NF were enumerated on N-free Ashby’s medium, containing (g L⁻¹): K₂SO₄—0.1; K₂HPO₄—0.2; MgSO₄ × 7H₂O—0.2; NaCl—0.2, CaCO₃—5.0, agar—15.0, sucrose—20.0. QMAFAnM and N-fixers were expressed as colony-forming units (CFU) per gram of dry soil. To calculate this, separate substrate samples were dried at 105 °C for 5 h to a constant mass to determine the fresh weight to dry weight (FW/DW) ratio.

The most probable number (MPN) of ammonifying, nitrifying, and denitrifying microorganisms was determined using the limiting dilution method in liquid media [5,28]. For each microbial group, 9 mL of the specific medium was dispensed into test tubes, and a series of ten-fold dilutions of the soil suspension were prepared (four independent dilution series per variant). Tubes were incubated for 5 days (denitrifiers, ammonifiers) or 14 days (nitrifiers) at 28 ± 2 °C. Growth presence was recorded, and the MPN per gram of dry soil was calculated using McCready tables [28].

Ammonifiers were cultured in meat-peptone broth (“HiMedia”, Mumbai, Maharashtra, India) containing 3% peptone. A strip of filter paper soaked in 5% lead acetate and a strip of litmus paper moistened with distilled water were suspended in the tube neck. The presence of ammonifiers was indicated by the release of ammonia, hydrogen sulfide, and mercaptans, detected by paper color change [28]. Nitrifying bacteria were cultured in Winogradsky’s medium (g L⁻¹): (NH₄)₂SO₄—2.0; K₂HPO₄—1.0; MgSO₄ × 7H₂O—0.5; NaCl—2.0; FeSO₄ × 7H₂O—0.05; CaCO₃—5.0. Denitrifiers were cultured in a modified Luria–Bertani (LB) medium supplemented with 1.0 g L⁻¹ NaNO₃. Nitrate and nitrite ions in the medium were detected using Griess reagent (“LenReaktiv”, Saint Petersburg, Russia) and zinc powder (“Sigma-Aldrich”, Taufkirchen, Germany) [5,28].

2.4. Estimation of Plant Morphometric and Physiological Characteristics

At the end of the experiment, the following growth parameters were measured: shoot height, leaf number, and fresh and dry biomass of shoots and roots. Photosynthetic pigments were extracted from approximately 50 mg of fresh leaf tissue in chilled 80% acetone. The contents of chlorophyll a (Chl a), chlorophyll b (Chl b), and carotenoids (Car) were determined spectrophotometrically (PD303UV, “Apel”, Saitama, Japan) by measuring absorbance at 470, 647, and 663 nm. Pigment concentrations were calculated according to Lichtenthaler [29] and expressed as mg per gram of dry weight (DW).

Lateral roots with fine roots were separated from the main root and fixed in 70% ethanol for structural analysis. A standard set of anatomical traits was recorded [30], including fine root and stele diameter; cross-sectional area of the roots and stele; thickness of the cortical parenchyma and cortex; number of cortical parenchyma cell layers; and the presence of aerenchyma, endodermis, and exodermis.

2.5. Assessment of pH, Different N Forms in Plant and Soil Samples

For total N analysis, plant samples (shoots and roots) were dried at 75 °C for 48 h, homogenized, and three composite samples were formed per pot. Total N content in plants and substrate was determined by the Kjeldahl method using a Heating Digestor DK 20 and Distillation Unit UDK 12 (“Velp”, Usmate, Italy) with titrimetric finish [31]. Nitrogen content in plants was expressed as mg per g DW and recalculated as mg per plant. Soil N was expressed as mg per g of air-dry substrate (ADS).

Alkaline-hydrolyzable nitrogen (N_alk.h.), representing ammonium and potentially mineralizable organic N, was determined according to [32] and expressed as mg per g ADS. This method involves hydrolyzing soil organic compounds with 1 M NaOH. The released ammonia (including exchangeable ammonium) is absorbed in a 5% boric acid solution and titrated with 0.005 M H₂SO₄ standard solution.

The pH and the concentrations of water-soluble nitrate anions (NO₃⁻), ammonium cations (NH₄⁺) were determined in a 1:2.5 (w/v) aqueous substrate extract. Measurements were performed by direct potentiometry using a universal ion meter (I160MI, “Izmeritelnaya Tekhnika”, Moscow, Russia) with an ion-selective electrode (ELIS-121NH4, Moscow, Russia) and a silver chloride reference electrode (ESr-10101/3.5K80.4, Moscow, Russia). Ion concentrations were quantified using calibration curves (range 10⁻⁵ to 10⁻¹ M) and expressed as mg per g ADS [33].

2.6. Statistical Analysis

Data were processed using MS Excel 16.0, Statistica 13.0 (StatSoft, Inc., Tulsa, OK, USA), and R software (version 4.1.2). The mean values (Means) with their standard errors (SE) and the number of biological replicates (n) are presented in the tables and figures. Data normality and homogeneity of variances were verified using the Shapiro–Wilk and Levene’s tests, respectively. Significant differences between treatments were determined by one-way and two-way analysis of variance (ANOVA) followed by Tukey’s HSD post hoc test. Relationships between bacterial abundance and plant/soil parameters were assessed using Spearman’s rank correlation. Significant differences (p < 0.05) between treatments are indicated by different lowercase or uppercase letters in figures and tables. All chemical reagents not otherwise specified were of domestic Russian production.

3. Results

3.1. Rhizobacteriome Diversity and Composition

The taxonomic composition was characterized based on 16S rRNA sequencing data. After quality and length filtering, a total of 111,320 high-quality sequences were obtained, with lengths ranging from 1431 to 1466 base pairs. Between 77.04% and 79.06% of all reads had a quality score of Q20, and 57.04% to 59.83% had a score of Q30. These data were used for taxonomic classification.

A total of 14 bacterial phyla with a relative abundance greater than 1% were identified. Together, Proteobacteria, Acidobacteria, Bacteroidetes, Planctomycetes, Gemmatimonadetes, Cyanobacteria, Verrucomicrobia, Armatimonadetes, Actinobacteria, Chlorobi, Chloroflexi, Firmicutes, Nitrospirae, and Elusimicrobia accounted for over 99% of the total sequences (Figure 2a). The dominant phyla in both the rhizosphere of the three tomato varieties and the unplanted bulk soil were Proteobacteria (49.0–60.0%), Acidobacteria (17.2–24.8%), Bacteroidetes (8.0–17.8%), Gemmatimonadetes (1.3–6.7%), and Planctomycetes (0.0–5.4%). Under N deficiency, the relative abundance of Proteobacteria decreased in the rhizosphere of all tomato varieties and in the bulk soil, while the abundance of Planctomycetes increased. The relative abundance of Acidobacteria in the rhizosphere of all varieties remained largely unchanged. Bacteria from the Chlorobi phylum were detected only in the unplanted bulk soil and the “–N”-rhizosphere of SL-v.1.

Figure 2. The results of metagenomic analysis of rhizosphere soil of three tomato plant varieties grown on a low-fertility substrate with (+N) and without (–N) nitrogen application: (a) The relative abundance at the phylum level (* The ‘Others’ category represents the sum of all remaining bacteria, each with a relative abundance of less than 1%); (b) Box-plots showing the alpha diversity indexes (Chao1 and Shannon) across treatments, indicating variations in microbial diversity at genera level (box—25–75% quartile, line—median, whiskers—range; different alphabetical letters indicate significant differences between the treatments according to Tukey’s test at p < 0.05); (c) The heatmap of relative abundance (RA) of genera associated with N transformation.

Variety-specific changes in the relative abundance of bacterial phyla were observed. In SL-v.1, N deficiency decreased the relative abundance of Bacteroidetes and Firmicutes, did not affect Gemmatimonadetes, and increased the abundance of Cyanobacteria, Verrucomicrobia, Actinobacteria, Armatimonadetes, and Chloroflexi. In SL-v.2, a decrease in the abundance of Gemmatimonadetes, Cyanobacteria, Verrucomicrobia, Actinobacteria, Armatimonadetes, Nitrospirae, and Chloroflexi was observed, along with an increase in Bacteroidetes and Firmicutes. In SL-v.3, the abundance of Bacteroidetes was unchanged, but similar to SL-v.2, there was a decrease in Gemmatimonadetes, Verrucomicrobia, and Nitrospirae, and an increase in Firmicutes. Conversely, and similar to SL-v.1, the abundance of Cyanobacteria, Actinobacteria, and Armatimonadetes increased.

One-way ANOVA revealed that the Chao1 and Shannon indices, calculated at the genus level, differed significantly among SL-v.1, SL-v.2, and SL-v.3 (p < 0.0001) (Figure 2b). In the bulk soil, N application resulted in decreased bacterial diversity, as indicated by lower Chao1 and Shannon index values. In the planted treatments, intervarietal differences were found: N addition did not affect the bacterial diversity in SL-v.1, which was the lowest among the varieties and comparable to the unplanted bulk soil. Nitrogen deficiency caused a significant decrease in bacterial diversity in SL-v.2 (particularly evident from the Shannon index), whereas SL-v.3 maintained a consistently high level of diversity and evenness that was largely unaffected by N application.

Taxonomic classification identified a total of 157 genera across all experimental rhizosphere samples. Of these, 38 genera directly involved in N transformation are presented in Figure 2c. Among these N-transforming bacteria, 95% belonged to the dominant phylum Proteobacteria, while only 2.5% belonged to Bacteroidetes (genus Flavisolibacter) and Firmicutes (genus Bacillus). The dominant genera among N-transforming bacteria in both the rhizosphere and the bulk soil were Stenotrophomonas (2.9–21.5%), Flavisolibacter (4.8–16.4%), and Janthinobacterium (0.0–10.6%). The relative abundance of the remaining 29 N-transforming genera ranged from 0.0 to 7.6%. The dominant genus, Stenotrophomonas, was more abundant in the bulk soil than in the tomato rhizosphere; however, N fertilization increased its abundance more than two-fold in almost all treatments (with the exception of SL-v.3). The unplanted bulk soil was also characterized by a high abundance of the genera Rhodoplanes, Limnohabitans, Sphingobium, Burkholderia, and Pseudomonas, in contrast to the rhizosphere. Furthermore, N application generally had no effect or reduced the abundance of these genera in the rhizosphere.

Variety-specific changes in the relative abundance of N-transforming bacterial genera were also revealed. In SL-v.1, representatives of the genera Rhodoplanes, Limnohabitans, Sphingomonas, Bradyrhizobium, Pelomonas, Rhodoferax, and Mesorhizobium were found only under N deficiency. With N-addition, the abundance of Flavisolibacter, Janthinobacterium, Devosia, and Delftia increased more than two-fold, and representatives of the genera Sphingobium, Phenylobacterium, Novosphingobium, Rhodanobacter, Luteimonas, Bacillus, Agrobacterium, Azohydromonas, and Polaromonas appeared. In SL-v.2, N deficiency conversely caused an increase in the abundance of Flavisolibacter, Devosia, and Sediminibacterium, while in SL-v.3, their abundance was either unchanged or increased. Additionally, in both SL-v.2 and SL-v.3, the genera Limnohabitans, Sphingobium, Burkholderia, Achromobacter, and Mesorhizobium appeared when N was added. Notably, in all studied varieties, N-application induced the appearance of the genus Rhizobium, which was not detected in the unplanted bulk soil.

A comparative analysis of bacterial genera was conducted to assess the effect of N deficiency on the microbial communities in the rhizosphere of the three tomato varieties and in the unplanted bulk soil. Venn diagrams (Figure 3) illustrate the number of common and unique genera. In the “+N” treatment, the highest number of unique bacterial genera (n = 34) was observed in the rhizosphere of SL-v.3. A significant number of unique genera (n = 21) was also found in SL-v.2, while SL-v.1 and the bulk soil had significantly fewer unique genera (4 each). A total of 13 bacterial genera were common to all four communities. Under N deficiency, the highest number of unique genera (n = 37) was again found in SL-v.3. The bulk soil contained 9 unique genera, while the SL-v.1 and SL-v.2 communities had 6 and 9 unique genera, respectively. Furthermore, in the “–N” treatment, only 11 bacterial genera were common to both the rhizosphere and the unplanted soil.

Figure 3. Venn diagram illustrating the distribution of microbial genera in different tomato varieties and unplanted bulk soil, highlighting common and unique communities to emphasize differences in microbial recruitment and establishment under nitrogen application (“+N” treatment) or deficiency (“–N” treatment).

3.2. Total Microbial Counts and N-Cycling Rhizobacteria

The total microbial counts showed that N addition had no effect on the QMAFAnM in the bulk soil, which averaged 2.6 × 10⁷ CFU kg⁻¹ (Table 1). A significant increase in QMAFAnM was observed only in the rhizosphere of SL-v.3 upon the addition of NH₄NO₃, while N deficiency slightly reduced it in all three varieties (on average by 1.5-fold). The analysis of N-cycling bacteria in the rhizosphere of the studied tomato varieties showed a significant predominance of N-fixers and denitrifiers over ammonifiers and nitrifiers (Table 1).

Table 1. The total quantity of mesophilic aerobic and facultative anaerobic microorganisms (QMAFAnM) and total N-cycling bacteria in the rhizosphere of three tomato plant varieties grown on a low-fertility substrate with (+N) and without (–N) nitrogen application.

Tomato Variety	Treatment	1 QMAFAnM, ×107 CFU g−1 of Dry Soil	1 N-Fixers, ×106 CFU g−1 of Dry Soil	2 Ammonifiers, ×106 MPN g−1 of Dry Soil	2 Nitrifiers, ×103 MPN g−1 of Dry Soil	2 Denitrifiers, ×106 MPN g−1 of Dry Soil
SL-v.1	+N	2.409 ± 0.215 bc	16.767 ± 0.731 a	0.115	0.095–20.00	15.0
	–N	1.646 ± 0.058 cd	1.704 ± 0.007 b	1.500	0.15–4.50	11.5
SL-v.2	+N	2.182 ± 0.141 bcd	20.671 ± 1.702 a	2.500	0.01–20.00	25.0
	–N	1.469 ± 0.107 d	3.346 ± 0.047 b	4.500	0.01–1.50	45.0
SL-v.3	+N	4.896 ± 0.110 a	4.766 ± 0.212 b	0.950	0.095–20.00	9.5
	–N	3.323 ± 0.125 b	2.628 ± 0.022 b	2.500	0.095–4.50	25.0
Bulk soil	+N	2.446 ± 0.205 b	0.683 ± 0.058 b	0.75	0.045–0.95	1.5
	–N	2.748 ± 0.199 b	0.612 ± 0.033 b	0.75	0.040–0.95	1.5

¹ Means ± SE (n = 6). Different superscript alphabetical letters indicate significant differences between the treatments (SL-v.1–3 and bulk soil) within each parameter according to Tukey’s test (p < 0.05). ² Most probable number (MPN) of rhizobacteria calculated based on four independent dilutions.

Nitrogen application led to a trend of increasing the number of N-fixers and nitrifiers, while the number of ammonifiers decreased (on average by 6-fold). No consistent, unidirectional change in the number of denitrifiers was observed. The abundance of different groups of N-transforming bacteria in the unplanted bulk soil remained unchanged. The rhizosphere of SL-v.3 seedlings had the smallest number of N-fixing bacteria, while intervarietal differences were less pronounced for the other bacterial groups.

3.3. Plant Morphometric and Physiological Characteristics

In the “–N” treatment, the average shoot height was 7.2 cm and the average number of leaves was 4 (Figure 4a,b). All tomato varieties responded to nitrogen addition: shoot height and leaf number increased by an average of 2.4-fold and 1.8-fold, respectively. Concurrently, the fresh aboveground and root biomass per plant increased by 7.6-fold and 9.6-fold, respectively (Figure 4c). Similar changes were observed in dry biomass: shoot biomass increased by an average of 10-fold, and root biomass by 15-fold (Figure 4d). Under limited N nutrition, no significant differences in biomass were found between varieties. Among the “+N” plants, SL-v.2 seedlings had the greatest shoot and root biomass.

Figure 4. The growth parameters of three tomato varieties grown on a low-fertility substrate with (+N) and without (–N) nitrogen application: (a) Shoot height; (b) Number of leaves; (c) Shoot and root fresh weight; (d) Shoot and root dry weight. Data are presented as Means ± SE (n = 18). Different lowercase and uppercase alphabetical letters indicate significant differences between the treatments according to Tukey’s test (p < 0.05).

By the end of the tenth week, even “+N” plants showed symptoms of N deficiency—leaf chlorosis, particularly in the lower leaves. In the “–N” treatment, leaves were not only chlorotic but also exhibited stunted development (Figure 1). Under N deficiency, SL-v.2 seedlings had the highest Chl content, while SL-v.3 plants had the lowest (Figure 5a). There were no significant differences in Car content between varieties. Nitrogen application resulted in a considerable increase in chlorophyll content (on average 2.9-fold). The greatest increase was observed in SL-v.3 seedlings (5.3-fold for Chl a and 4.2-fold for Chl b). The Car content increased by an average of 1.7-fold (Figure 5a). The Chl a/Chl b ratio remained virtually unchanged with N addition, averaging 1.6 (Figure 5b). However, the ratio of Chl (a + b) to Car increased significantly (on average by 40%), due to the more substantial increase in chlorophyll content (especially Chl a).

Figure 5. The physiological parameters of three tomato plant varieties grown on a low-fertility substrate with (+N) and without (–N) nitrogen application: (a) Leaf photosynthetic pigment content; (b) Photosynthetic pigment ratios. Data are presented as Means ± SE (n = 4). Different alphabetical letters (uppercase—Chl a and Chl a/b; nonitalic lowercase—Chl b; italic lowercase—Car and Chl (a + b)/Car) indicate significant differences between the treatments according to Tukey’s test (p < 0.05).

3.4. Fine Root Structural Characteristics

The fine roots of all studied tomato varieties shared a common structural plan. Aerenchyma was present in the parenchyma of the majority (73%) of roots (Table 2). The thickest absorbing roots were found in SL-v.3; they had significantly larger root and stele diameters (1.44-fold) and greater cortex thickness (1.3-fold) compared to SL-v.1 and SL-v.2. SL-v.3 plants also had the maximum parenchyma thickness (1.5-fold greater than other varieties). Furthermore, aerenchyma was more prevalent in the roots of this variety than in SL-v.1 and SL-v.2 (84% vs. 67%). Under N-deficiency, all varieties showed a tendency toward decreased root and stele diameter, as well as reduced parenchyma and cortex thickness. The proportional area of the stele in the root cross-section did not change. Nitrogen deficiency had little effect on the proportion of aerenchyma in SL-v.3, whereas in SL-v.1 and SL-v.2 it decreased by 40% and 11%, respectively. The vast majority of fine roots (87%) had a thickened exodermis. SL-v.3 was distinguished by the presence of thickenings in the endodermis under both treatments, but N-addition stimulated greater development of the endodermis. The endodermis was not thickened in SL-v.1, while in SL-v.2 this feature was found in 8% of the roots in the “+N” treatment (Table 2).

Table 2. Structural characteristics of fine roots of three tomato plant varieties grown on low-fertility substrate with (+N) and without (–N) nitrogen application.

Parameter		SL-v.1		SL-v.2		SL-v.3
		+N	–N	+N	–N	+N	–N
Diameter, µm	Root	1 290.8 ± 7.7 b	280.6 ± 13.5 b	289.0 ± 11.1 b	276.7 ± 13.9 b	414.4 ± 15.9 a	370.0 ± 18.1 a
	Stele	68.11 ± 2.17 b	65.56 ± 3.48 b	73.29 ± 3.57 b	65.44 ± 5.06 b	105.64 ± 6.22 a	98.72 ± 6.96 a
The stele proportion in the root cross-sectional area, %		5.78 ± 0.33 a	5.75 ± 0.37 a	6.57 ± 0.31 a	5.75 ± 0.54 a	6.59 ± 0.43 a	7.20 ± 0.56 a
Thickness of cortex parenchyma, µm		61.85 ± 2.67 b	57.71 ± 3.88 b	57.55 ± 2.40 b	58.54 ± 2.78 b	94.71 ± 4.29 a	87.86 ± 4.62 a
Number of layers of cortex parenchyma, pcs.		2.51 ± 0.09 a	2.37 ± 0.11 ab	2.03 ± 0.05 b	2.05 ± 0.11 b	2.69 ± 0.09 a	2.30 ± 0.10 ab
Thickness of cortex, µm		111.32 ± 3.41 b	107.52 ± 5.54 b	107.83 ± 4.17 b	105.63 ± 6.19 b	154.39 ± 5.66 a	135.64 ± 6.41 a
Proportion of roots with aerenchyma, %		78.05	55.56	71.05	63.64	85.71	81.48
Proportion of roots with thickened exodermis, %		80.49	77.78	94.74	72.72	100.00	96.30
Proportion of roots with thickened endodermis, %		0.00	0.00	7.89	0.00	14.29	7.41
Two-way ANOVA (Multivariate Tests of Significance)		Factor Varieties [V]		Factor Nitrogen [N]		[V] × [N]
		F = 9.285	p < 0.001	F = 0.968	p = 0.463	F = 1.511	p = 0.093

¹ Means ± SE (n = 30). Different superscript alphabetical letters indicate significant differences between treatments (SL-v.1–3) within each parameter according to Tukey’s test (p < 0.05).

3.5. Nitrogen Content in Plant Biomass

With limited N nutrition, the total N content in the aboveground biomass of all the studied tomato varieties was lower than that of the “+N” plants, on average by 1.3 times, and in the underground biomass was lower by 1.9 times (Table 3).

Table 3. Total nitrogen content in the shoot and root of three tomato plant varieties grown on a low-fertility substrate with (+N) and without (–N) nitrogen application.

Tomato Variety	Treatment	Ntot., mg g−1 DW		Ntot., mg Plant−1
		Shoot	Root	Shoot	Root
SL-v.1	+N	1 8.990 ± 0.203 b	18.267 ± 0.159 a	4.746 ± 0.442 bc	4.182 ± 0.276 b
	–N	6.985 ± 0.317 c	8.074 ± 0.078 d	0.257 ± 0.011 d	0.129 ± 0.0373 c
SL-v.2	+N	9.959 ± 0.320 b	16.469 ± 0.296 b	16.551 ± 2.852 a	19.771 ± 2.558 a
	–N	7.725 ± 0.317 bc	8.158 ± 0.101 d	0.918 ± 0.075 cd	0.478 ± 0.019 c
SL-v.3	+N	11.756 ± 0.134 a	13.330 ± 0.035 c	8.139 ± 0.212 b	2.091 ± 0.250 b
	–N	9.255 ± 0.309 b	9.570 ± 0.370 d	0.780 ± 0.033 cd	0.227 ± 0.003 c

¹ Means ± SE (n = 9). Different superscript alphabetical letters indicate significant differences between treatments (SL-v.1–3) within each parameter according to Tukey’s test (p < 0.05).

The highest N content in the shoots was found in “+N” SL-v.3-seedlings, and in the roots—in “+N” SL-v.1-seedlings. At the same time, the amount of N in the aboveground and underground biomass per plant in the “–N” treatment tomato plants was lower than in the “+N” treatment ones, on average by 15 and 31 times, respectively, which is associated with both the reduced N content per unit weight and the smaller biomass. The maximum N content per plant was found in “+N” SL-v.2-seedlings (Table 3).

3.6. Different N Forms and pH Values in Substrate

The amount of N_alk.h. averaged 5.6% of the total N (Table 4). Ammonium nitrogen (N–NH₄⁺) accounted for an average of 0.3% of N_alk.h. in planted soils and 0.4% in the unplanted bulk soil, with the remainder comprising N from easily hydrolyzable organic compounds. The soil nitrate nitrogen (N–NO₃⁻) content dominated over N–NH₄⁺: after growing SL-v.1, it was 15.2 times higher on average, and after growing the other two varieties, it was 34 times higher. Nitrogen application did not significantly affect the amount of N–NO₃⁻, which is consistent with active consumption of this element by plants. In the bulk soil under N deficiency, the nitrate level was similar to the planted treatments. However, with N addition, it was 19 times higher compared to all other treatments (Table 4). Under N-deficient conditions, the substrate pH averaged 6.70 after the experiment. In all treatments where NH₄NO₃ was added, the pH decreased, averaging 6.46 (Table 4).

Table 4. The content of different nitrogen (N) forms and pH values in substrate after the pot-scale experiment.

Tomato Variety	Treatment	Ntot., mg 100 g−1 of ADS	Nalk.h., mg 100 g−1 of ADS	N–NO3−, mg 100 g−1 of ADS	N–NH4+, mg 100 g−1 of ADS	pHH2O
SL-v.1	+N	1 183.82 ± 3.10 a	9.197 ± 0.381 bc	0.608 ± 0.024 b	0.042 ± 0.001 a	6.49 ± 0.04 bc
	–N	163.83 ± 0.65 bc	8.078 ± 0.292 c	0.588 ± 0.005 b	0.037 ± 0.002 b	6.70 ± 0.03 a
SL-v.2	+N	181.14 ± 3.12 a	9.755 ± 0.558 bc	0.689 ± 0.010 b	0.018 ± 0.002 c	6.58 ± 0.03 ab
	–N	184.56 ± 1.03 a	9.540 ± 0.388 bc	0.690 ± 0.019 b	0.023 ± 0.001 c	6.67 ± 0.01 a
SL-v.3	+N	184.91 ± 1.93 a	12.375 ± 0.302 a	0.720 ± 0.022 b	0.020 ± 0.001 c	6.39 ± 0.04 c
	–N	174.63 ± 0.26 ab	11.201 ± 0.381 ab	0.867 ± 0.028 b	0.026 ± 0.001 c	6.73 ± 0.00 a
Bulk soil	+N	181.63 ± 3.54 a	8.900 ± 0.413 c	13.639 ± 0.350 a	0.045 ± 0.002 a	6.40 ± 0.02 c
	–N	158.00 ± 2.48 c	8.725 ± 0.185 c	0.901 ± 0.028 b	0.033 ± 0.001 b	6.73 ± 0.03 a

¹ Means ± SE (n = 6). Different superscript alphabetical letters indicate significant differences between treatments (SL-v.1–3 and bulk soil) within each parameter according to Tukey’s test (p < 0.05). ADS—air-dry substrate.

4. Discussion

The biogeochemical imbalance of N in the biosphere has intensified in recent decades [6,7]. In this context, the rhizosphere microbiome is a key factor in facilitating nutrient uptake by crops under N-deficient conditions. Studying the biodiversity of associative microorganisms is fundamental to understanding the structure of soil microbial communities and their roles in soil formation and nutrient cycling, including the N cycle [34].

In recent years, significant research efforts have been directed toward characterizing the bacterial microbiome of S. lycopersicum. Studies have examined tomatoes grown in fields [20,21,23], greenhouses [22], and pot or hydroponic cultures [4,10]. These investigations have demonstrated that S. lycopersicum hosts a complex and diverse bacterial community, with Proteobacteria typically being the dominant phylum. Our research confirms this, showing that Proteobacteria was the dominant phylum in the rhizosphere of all three tomato varieties tested, as well as in unplanted bulk soil, followed by Acidobacteria and Bacteroidetes. Mejia et al. [4] also found Proteobacteria to be dominant in hydroponic culture, with Bacteroidetes as the second most abundant phylum. Naumova et al. [20] reported that Proteobacteria and Actinobacteria together accounted for 90% of the total sequence reads in roots and 50% in the rhizosphere of S. lycopersicum.

Redundancy Analysis (RDA) of the five most abundant bacterial phyla against substrate physicochemical parameters (pH and various N forms) showed that the first two RDA axes together explained 69.0% of the total variation (Figure A1b). An ANOVA-test for the full set of parameters was not significant for these phyla. However, the most influential parameter was the amount of N_alk.h. in the soil, which characterizes ammonium and potentially plant-available organic N compounds. This parameter had a greater impact on Proteobacteria, while Acidobacteria and Gemmatimonadetes were more influenced by soil pH.

A key focus in studying the structural and functional characteristics of tomatoes under N deficiency is the prevalence of N-transforming bacteria. Among the 38 N-cycling genera isolated from the rhizosphere of the three tomato varieties, representatives of different functional groups were identified. These included not only specialists but also bacteria capable of multiple metabolic transformations. RDA of the top ten N-transforming genera against substrate physicochemical parameters showed that the first two axes explained 74.9% of the total variability (Figure A1b). Among these parameters, N forms such as N_tot. and N–NO₃⁻ had the greatest impact on the selected bacterial genera.

The genera Stenotrophomonas, Flavisolibacter, and Janthinobacterium had the highest relative abundance among the identified N-cycling bacteria (Figure 2c). Stenotrophomonas, a genus of Gram-negative bacteria found in soil and plants [35,36], includes species capable of N fixation. For instance, Jia et al. [35] isolated a novel strain, Stenotrophomonas maltophilia DQ01, from landfill leachate, which can simultaneously remove nitrate and ammonium under aerobic conditions. Our study revealed a strong positive correlation between the relative abundance of the Stenotrophomonas genus and the amount of N–NO₃⁻ (0.90, p = 0.002, Figure A2, Table S1). Flavisolibacter, a genus of Gram-negative bacteria, can participate in the N cycle through the dissimilatory nitrate reduction pathway, converting nitrates to molecular N under anaerobic conditions, thus acting as a denitrifier [34]. Janthinobacterium is a genus of soil bacteria whose representatives actively participate in N transformations [37,38]. A strain identified as Janthinobacterium sp. M-11, isolated from the Songhua River, effectively removes ammonia at cold temperatures and possesses the narG gene, which encodes a membrane-bound nitrate reductase [37]. Chen et al. [38] reported that J. svalbardensis F19 can perform simultaneous heterotrophic nitrification and aerobic denitrification under aerobic conditions, with gene-specific PCR confirming the presence of napA, hao, and nirS genes. In our research, a positive correlation was found between the abundance of Janthinobacterium spp. and the number of N-fixers (0.86, p = 0.0065) and denitrifiers (0.73, p = 0.0378).

Overall, N-fixers and denitrifiers were the most common N-cycling bacteria isolated from the tomato rhizosphere. Among the N-fixing bacteria, in addition to Stenotrophomonas, were genera such as Bradyrhizobium, Rhizobium, Polaromonas, Herbaspirillum, Dechloromonas, Novosphingobium, and Azohydromonas [39,40,41]. For example, we found a high correlation between the occurrence of Rhizobium and the number of N-fixers (0.79, p = 0.02) and nitrifiers (0.84, p = 0.08). The abundance of Azohydromonas also correlated with the quantity of N-fixers (0.73, p = 0.039).

In the tomato rhizosphere, denitrifiers such as Devosia, Dokdonella, Thermomonas, Pseudoxanthomonas, Rhodanobacter, Comamonas, Pelomonas, Rhodoferax, and Thauera were found alongside Flavisolibacter [42,43,44]. A strong negative correlation was observed between Devosia and soil N–NO₃⁻ content (–0.85, p = 0.008), and a positive correlation with the number of denitrifiers (0.73, p = 0.041). This aligns with other authors [42] who reported that Devosia sp. carries the nirK gene, key to catalyzing the reduction of nitrite (NO₂⁻) to nitric oxide (NO) during denitrification. A high negative correlation was also found between the abundance of the denitrifier Rhodoferax and total soil N (−0.76, p = 0.0273).

Many microorganisms are metabolically versatile polyfunctional systems. When environmental conditions change, they can adapt, perform different functions, and be reclassified into another physiological group. For example, many aerobic nitrite-oxidizing bacteria can also oxidize ammonia or reduce nitrates [6]. Such polyfunctional bacteria isolated from the tomato rhizosphere include Rhodoplanes, which is associated with N fixation, nitrate reduction, and denitrification [45,46]. We found a high positive correlation between the abundance of these bacteria and the amount of N–NO₃⁻ (0.81, p = 0.015). Nitrospira plays a crucial role in both nitrate reduction and nitrite ammonification [47]. Furthermore, comammox Nitrospira are abundant ammonia oxidizers capable of aerobically oxidizing ammonium to nitrate in a two-stage process [48]. We found a high negative correlation between Nitrospira abundance and the amount of N–NH₄⁺ (−0.86, p = 0.006) and a positive correlation with total soil N (0.88, p = 0.004). Thus, even in a pot-scale experiment with low-fertility soil, we observed a considerable diversity of rhizobacteria involved in various N transformation processes (representing 24% of all identified genera).

Efficient N uptake and assimilation are critical for plant vitality and productivity [2,8]. Nitrogen starvation during early growth stages causes poor development of the assimilatory surface and growth retardation [1,15]. Under insufficient N nutrition, both the aboveground and root biomass of the model tomato plants were significantly lower than when N was applied. Similar results have been reported by other authors studying the production process in these species as a function of N supply [49,50].

Tomato shoot and root biomass correlated with the total quantity of N-fixers (0.94, p = 0.0048 and 0.99, p = 0.0001, respectively). A positive correlation was also found between the number of N-fixers and tomato shoot height (0.837, p = 0.0048).

Plant productivity is determined by photosynthetic intensity, which drives the accumulation of dry biomass. Therefore, assessing the effect of N deficiency on the photosynthetic pigment complex—namely the content of Chl a, Chl b, and Car, and their ratio—was important. Pigment content can also serve as an indicator of stress responses. The Chl and Car content in the leaves of all three tomato varieties, even with N fertilization, was significantly lower (on average 2.5 times) than in the ‘Gruntovy Gribovsky 1180’ variety [51]. This is likely explained by our use of a low-fertility substrate containing sand and clay in addition to peat. Varietal differences may also contribute to the disparity in pigment content. Both factors (N application and tomato variety) significantly influenced morphometric parameters (biomass and shoot height) and the total content of photosynthetic pigments.

The active release of various organic compounds (root exudates) into the soil provides nutrients for rhizosphere microorganisms, creating favorable conditions for their growth [17,52]. The anatomical structure of the absorbing roots, particularly features of the primary cortex such as the degree of endodermal suberization and the presence of an exodermis, is crucial for the effective release of these compounds [53]. Severe suberization of these barrier tissues can significantly reduce root exudation [52]. In two of the studied tomato varieties (SL-v.2 and SL-v.3), endodermal thickening was observed, though the proportion of such roots was low (not exceeding 14%), while the proportion of roots with a thickened exodermis was significantly higher. The functional structure of the microbial community is also affected by soil oxygen content, which can be reduced by root respiration. The development of root cortex aerenchyma helps oxygenate the rhizosphere, which stimulates nitrification [2] and negatively affects denitrification, an anaerobic process. Aerenchyma was present in the roots of all varieties but was most common in SL-v.3. This may explain why the total N content in the shoots of this variety, as well as the nitrate content in the soil after its cultivation, were higher compared to the other two varieties. Under N deficiency, all varieties showed a tendency toward decreased fine root and stele diameter (Table 2). Since smaller root diameter promotes more active radial transport of substances into the stele [2,8], this decrease may be an adaptive response to soil N limitation. A high positive correlation was noted between stele diameter and the amount of N_alk.h. in the soil (0.83, p < 0.0361, Figure A3, Table S2). Two-way ANOVA showed that only the tomato variety had a significant effect on root structure, while N application and the interaction between variety and N were not significant (Table S3).

Different crop varieties differ in the N content of their aboveground and root biomass and, consequently, in their structural and functional mechanisms for adapting to N deficiency. A high positive correlation was found not only between N accumulation in tomato roots and the content of N_alk.h. in the substrate (0.89, p = 0.019, Figure A3, Table S2) but also with the abundance of certain N-fixing rhizobacteria genera (Novosphingobium, Azohydromonas, and Rhizobium), confirming their crucial role in maintaining the N status of the host plant.

The total N content in the shoots of “–N” tomatoes averaged 8 mg g⁻¹ DW, compared to 10.2 mg g⁻¹ DW in “+N” tomatoes. This is slightly lower than values reported elsewhere. For example, Gatsios et al. [19] found the total content in leaves of S. lycopersicum (var. ‘Ekstasis F1’) grown in a protected culture ranged from 13.1 to 19.3 mg g⁻¹ DW. Mawalagedera et al. [49] reported that N content in the leaves of 8-week-old tomato seedlings varied from 1.0 to 3.0% DW (10–30 mg g⁻¹ DW), and in 12-week-old plants from 2.5 to 3.5% DW. The average tissue N concentration in a study of S. lycopersicum by Erabadupitiya et al. [50] was 5.0% DW (50 mg g⁻¹ DW), and increasing the N rate did not increase tissue N concentration at the vegetative stage. The lower N content in our experiment is likely due to the use of a low-fertility substrate especially in the non-fertilized treatments where visual symptoms of N deficiency (light green leaves and lower leaf chlorosis) were evident. It should also be noted that the data from other authors [49,50] refer specifically to leaves, whereas we determined the N content in total aboveground biomass. Nitrogen content in shoots is typically lower than in leaves due to the inclusion of N-poor structural tissues.

Under N deficiency, the production process is severely limited, consequently reducing the flux of assimilates to the roots to support the microflora. Nitrogen fixation is an energy-intensive process and is therefore limited by insufficient carbohydrates [31]. The addition of even a low dose of N fertilizer (a total of 60 mg N per kg of soil) positively affected plant biomass, leaf photosynthetic pigment content, the biodiversity of N-transforming rhizobacteria, and especially the number of N-fixing bacteria. Our previous research [54] also showed an enhanced beneficial effect of Bacillus sp. TO15c on rapeseed plants in the presence of N fertilizers, despite the N-fixing ability of these bacteria.

5. Conclusions

This study investigated the impact of N deficiency on the rhizosphere bacteriome and plant physiology of three tomato varieties grown in a low-fertility substrate. Our findings confirm that even a minimal application of N fertilizer to a low-fertility substrate had a positive effect on the structure and functional potential of the rhizobacterial community. Under N deficiency, all tomato varieties exhibited reductions in growth parameters, biomass, photosynthetic pigment content, and fine root and stele diameter. Varietal-specific responses were confirmed, observed both in the host plant—manifested through changes in biomass, photosynthetic pigments, and root structure—and the associated rhizobacteria. This is evident from the significant intervarietal differences in alpha-diversity, the relative abundance of key bacterial phyla, and, most importantly, the composition of functional groups of N-transforming bacteria.

The dominant phyla in the rhizosphere were Proteobacteria, Acidobacteria, and Bacteroidetes, with N deficiency reducing the relative abundance of Proteobacteria. It was noted that the tomato rhizosphere harbors a considerable diversity of bacteria involved in the N cycle (38 genera), accounting for 24% of all identified genera. Nitrogen application had a selective effect on different functional groups of bacteria, increasing the number of N-fixers and nitrifiers and suppressing ammonifiers. The strong correlations discovered between the abundance of specific genera (such as Stenotrophomonas, Janthinobacterium, and Rhizobium) and the content of various N forms in the soil and plants underscore their crucial role in maintaining the N status of the host plant. This confirms the important role of the rhizobacteriom as a key tool in tomato adaptation to N deficiency. However, its functional capabilities may be limited by a lack of readily available carbohydrates due to disruption of the production process in plants.

Further research involves isolating key N-transforming bacteria to test their functions in controlled plant experiments and using metatranscriptomics to assess the actual activity of key N-cycling genes in situ. Ultimately, field trials are necessary to validate the performance of promising plant varieties and bacterial strains under real-world conditions, with the goal of developing effective microbial consortia that improve N use efficiency and reduce fertilizer dependence in sustainable agriculture.

Appendix A.1

Figure A1. Relationships between substrate physicochemical parameters (pH, nitrogen forms) and the relative abundance of bacteria in the rhizosphere of three tomato varieties, revealed by Redundancy Analysis (RDA): (a) At the level of the five most abundant phyla; (b) At the level of the top-10 genera involved in N-transformation. * Significant parameter at p < 0.05.

Appendix A.2

Figure A2. Heatmap of Spearman’s rank correlations (r_s) between the relative abundance (RA) of N-transforming rhizobacterial genera and key plant and substrate properties. Plant parameters include N accumulation in shoots and roots. Substrate properties include pH, the content of various N forms, and the total number of N-cycling bacteria in the tomato rhizosphere.

Appendix A.3

Figure A3. Heatmap of Spearman’s rank correlations (r_s) between the morphophysiological parameters of three tomato varieties and key substrate properties. Plant parameters include nitrogen accumulation in shoots and roots. Substrate properties include pH, the content of various N forms, and the total number of N-cycling bacteria in the tomato rhizosphere.