Changes in the Rhizospheric Microbiota of Pepitilla Maize in Response to Drought: Functional and Taxonomic Analysis

TL;DR

This study explores how drought affects the rhizospheric bacteria of a native maize variety, Pepitilla, and identifies specific bacterial groups that respond to water stress.

Contribution

The study provides new insights into the taxonomic and functional changes in the rhizospheric microbiota of Pepitilla maize under drought conditions.

Findings

Drought significantly altered the abundance of bacterial taxa like Actinobacteria and Proteobacteria in the rhizosphere.

Certain bacterial families, such as Microbacteriaceae and Sphingomonadaceae, showed increased relative abundance under drought stress.

The study highlights the potential role of specific bacterial groups in plant responses to drought stress.

Abstract

Native maize varieties provide important information for counteracting the effects of climate change, which leads to agricultural drought. The native rhizospheric microbiota is an ecological niche that maintains a close relationship with the plant and helps mitigate the effects of drought on it. The objective of this study was to describe the composition and structure of the rhizospheric bacterial communities of the native Pepitilla maize plants under conditions of water stress. An experiment was conducted under greenhouse conditions with three irrigation regimes and a control with normal irrigation. The responses of the plants to drought and the rhizospheric bacterial microbiota were measured before, during, and after the drought. Bacterial diversity was analyzed from rhizospheric soil using massive sequencing of the 16S rRNA gene. The drought model applied in the experiment had a negative effect on the plants, affecting their physiological, morphological, and biochemical functions. Diversity analyses showed statistical differences between the conditions during and after the drought in most cases. A reduction and modification in bacterial abundance was observed during the drought condition across different taxonomic groups, the most representative being the phyla Actinobacteriota, Pseudomonadota, and Acidobacteriota; the families Acidobacteriaceae, Rhodanobacteraceae, Solirubrobacteraceae, Acidothermaceae, and Microbacteriaceae; and the genera Actinobacteria, Sphingomonas, Geodermatophilus, Conexibacter, and Acidothermus. It is worth noting that the taxa Actinobacteria and Proteobacteria, as well as the families Microbacteriaceae, Sphingomonadaceae, and Unclassified_Actinobacteria, were directly associated with the drought condition, as an increase in their relative abundance was observed. This information is very useful for understanding the relationship between certain taxa enriched during stress conditions and the physiology of maize plants.

Figures

Full text

1. Introduction

Drought is a meteorological phenomenon resulting from changes in rainfall patterns, leading to agricultural drought, which is defined as a shortage of water in the soil that affects the optimal development of crops [1]. It is estimated that drought can cause up to a 40% loss in maize production [2]. Currently, alternatives are being sought to optimize the use of water resources in a more sustainable way, including improved and more efficient irrigation systems and the genetic improvement of crops to make them more drought-tolerant [3]. The importance of better water resource management in maize production is due to its significance in the agricultural sector, as well as its notable cultural, culinary, industrial, and nutritional value [4,5]. Therefore, maize is one of the most widely produced and consumed staple crops globally [6]. Many improved maize varieties have positively impacted production despite water scarcity. However, there has also been a significant increase in the inputs required for their production, and in many cases, this improvement is limited to some varieties [7,8,9]. While the plant genome represents a promising alternative for developing drought-tolerant varieties, especially among native maize, from a holistic perspective, another portion of the plant is also subject to study: the plant’s native microbiota, with particular emphasis on the rhizosphere. This close link between plants and their microbiota can reveal numerous key factors that contribute to stress adaptation. The rhizosphere microbiota depends on the plant’s genotype, and it has also been found to depend on the crop’s phenological stages, which are classified and described as beginning with Germination (0), followed by the development of leaves and stem (Vegetative: V1, V2, V3… up to V(n)), emergence of the inflorescence (VT—Tassel), Flowering (6), Fruit Development (7), Maturation (8), and finally Senescence (9) [10,11]. The rhizosphere is a microenvironment comprising the first three millimeters adjacent to the root epidermis [12], an area where a dynamic, symbiotic environment forms with numerous microorganisms that fulfill various functions [13,14], including the fixation and solubilization of essential nutrients for the plant, pathogen suppression, and the production of metabolites that the plant uses for its development. The microbiota also plays a key role in promoting plant tolerance to abiotic stresses, such as drought and salinity [15,16]. According to several studies, plants encourage the assembly and recruitment of microbes in the rhizosphere through the secretion of different metabolites that attract certain microorganisms, depending on the plant’s genotype [17,18]. Understanding the plant-microbe-soil relationship by analyzing rhizosphere bacterial communities is crucial for improving maize drought tolerance. Focusing on native maize varieties, owing to their high adaptability to diverse environments, represents an opportunity to improve agricultural production systems sustainably, save available water resources, and even reduce their use in agriculture. The objective of this study is to analyze the composition and structure of the bacterial communities in the rhizosphere of native Pepitilla maize under drought conditions, and to relate the bacterial genes found to the plant’s vegetative responses and metabolic activity under water stress.

2. Materials and Methods

2.1. Study Area and Native Soil Sampling

The study area was a field located in the Altos Norte region of Jalisco, in the municipality of Yahualica de González Gallo, Jalisco (−102.869444 N, −21.061667 W), where a farmer has continuously cultivated the native Pepitilla maize variety for approximately ten years, with an average annual rainfall of 570 mm [19]. Using a zigzag sampling pattern across a roughly three-hectare field, 45 samples were collected at a depth of 20 cm, mixed, homogenized, and combined into a composite sample. Plastic bags with a 12 kg capacity (suitable for pots) were filled with this composite.

2.2. Experiment Setup and Soil Analysis

An experiment was established using pots filled with soil following the recommendations of previous authors [20,21] in a completely randomized design distributed across four treatments (T1: irrigated control; T2: three days of drought; T3: five days of drought; and T4: eight days of drought). Each treatment included a total of 11 pots, with a planting density of three plants per pot. A physicochemical soil analysis was previously performed by the company Fertilab^®^, Celaya, Guanajuato, Mexico, using their established techniques [22]. Agronomic management included initial fertilization at sowing with diammonium phosphate (DAP) ((NH_4_)2HPO_4_ 18-46-00) and a second fertilization at the V4 stage with Ammonium Sulfate + Ammonium Nitrate 1:1 at a rate of one gram per pot. The plants were kept under normal irrigation until the flowering stage, as indicated by the emergence of the first silks. Subsequently, the soil in the pots was maintained at 40% field capacity, except for the irrigated control, which was watered normally throughout the crop cycle. The remaining plants were subjected to water withholding for the specified duration. After the corresponding drought period for each treatment, a recovery irrigation was applied.

2.3. Morphological and Physiological Responses of Plants to Drought Conditions

Plant responses were measured in all treatments before, during, and after drought conditions were applied: root biomass (RB)—four plants per drought treatment were selected for each of the three conditions (PRE, DUR, and POST drought). The plants were cut at the base of the stem, excess soil was removed, and only soil adhering to the roots was kept. The roots were then washed, and the procedure proposed by Félix et al., 2023 [23] was followed, using both the fresh and dry weight of the roots to calculate biomass; leaf area (LA) was measured from ligulated leaves in the middle stratum of the plant using a CI-203 Portable Leaf Area Meter (CID Bio-Science, Inc., Camas, WA, USA), with three readings taken per leaf to obtain an average value; chlorophyll content (CC) was measured using a SPAD 502 Plus Minolta [SP02900P] (Osaka, Japan) on ligulated leaves from the middle stratum, with three readings taken at different points on the leaf; relative water content (RWC) was measured following the procedure proposed by Villalobos et al., 2016 [24], using leaf tissue discs to record fresh weight, turgid weight after 24 h of imbibition in distilled water, and dry weight after oven drying. The following formula was then used to calculate the final water percentage: [RWC = (FW − DW)/(TW − DW) × 100].

2.4. Measurement of Biochemical Responses in Plants to Drought Conditions

Free proline levels (PL) were measured following the methodology of Bates [25], using ligulated leaves from the middle stratum of the plant. Absorbance readings were taken with a Thermo Fisher Scientific Multiskan™ GO (Thermo Fisher Scientific, Vantaa, Finland) spectrophotometer at a wavelength of 520 nm, using toluene as a blank. The absorbance values were compared with a standard curve of L-proline based on fresh weight (mg/g of fresh biomass). The total soluble sugar concentration (TSSC) was measured from leaf juice of crushed leaves from the plant’s middle stratum, in °Brix, using an Atago™ MASTER-20T (ATAGO CO., LTD., Tokyo, Japan) refractometer. All values for the plant response variables to drought conditions were analyzed using SAS 9.0 statistical software through an analysis of variance with a significance value (p < 0.05); additionally, a post hoc Tukey test was performed to compare values by treatment.

2.5. Rhizospheric Soil Sampling and DNA Extraction

Soil surrounding the sampled roots for root biomass measurement was collected, mixed, and homogenized by pooling soil from treatments, and separated into periods before, during, and after the drought condition. Soil DNA extraction was performed using the protocol suggested by Wilson (2001) [26] with some modifications. Lysis was performed from 0.5 g of soil sample diluted in 250 µL of TE 50:20 buffer, using 0.5 g of sterile beads, and 250 µL of lysis buffer [1M Tris HCl pH = 8, 0.5 M EDTA pH = 8, 2.5 M NaCl, 10% CTAB, and 20% SDS], vortexed for 30 min, followed by enzymatic lysis with 10 µL of proteinase K and incubation (57 °C for 45 min). DNA purification was continued, adding phenol-chloroform, followed by incubation (65 °C/10 min), a thermal shock step (–80 °C, and centrifugation (10,000 rpm/ 10 min). A final purification step was performed using chloroform-isoamyl alcohol, followed by centrifugation (10,000 rpm/ 10 min) three times. Finally, the aqueous phase was recovered and mixed with 1 mL of cold absolute isopropanol stored overnight (20 °C), centrifuged (10,000 rpm/ 10 min), decanted, and washed with 500 µL of 70% ethanol. The pellet was air-dried at room temperature and hydrated with 50 µL nuclease-free sterile water. Finally, DNA quantification was performed using a Nanodrop, and its integrity was verified by 1% agarose gel electrophoresis.

2.6. 16S rRNA Library Preparation and Sequencing

Libraries were generated by amplifying the hypervariable regions of the 16S rDNA gene (V2, V3, V4, V6, V7, V8, and V9) using the Ion 16S Kit (Thermo Fisher Scientific, Waltham, MA, USA) in two separate reactions using a Verity™ thermocycler (Thermo Fisher Scientific, Waltham, MA, USA). DNA was quantified with a Qubit Fluorometer (Invitrogen™ Q33216) (Thermo Fisher Scientific, Singapore, Singapore), and samples were adjusted to 50 ng for constructing 16S rDNA libraries with the Ion Plus Fragment Library Kit™ (Thermo Fisher Scientific, Waltham, MA, USA) and Ion Xpress barcode adapters (Thermo Fisher Scientific, Waltham, MA, USA). The libraries with integrated barcodes were purified using the Agentcourt AMPure XP system according to the manufacturer’s instructions (Beckman Coulter, Brea, CA, USA). Once purified, the libraries were quantified using a Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA), adjusted to 26 pM, and then pooled. 25 µL of the pooled library was used for emulsion PCR with the OneTouch Enrichment System (Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer’s instructions. Finally, a 316 v2 silicon chip with microwells (Thermo Fisher Scientific, Waltham, MA, USA) was loaded with the final sample, and sequencing was performed on the Ion S5™ system (Thermo Fisher Scientific, Waltham, MA, USA).

2.7. Data Analysis

Root biomass, leaf area, chlorophyll, relative water content, proline, and sugar data were analyzed using PROC ANOVA, and means were compared using Tukey’s test (p < 0.05) in SAS 9.0 statistical software (SAS Institute Inc., Cary, NC, USA).

Bioinformatics analysis was performed using the nf-core Ampliseq pipeline v2.9.0 [27,28], including the removal of chimeric and low-quality sequences using the MultiQC and DADA2 software packages, respectively [29,30], followed by taxonomic assignment in the DADA2 module using the SILVA v-132 database as reference [31]. The alpha diversity profile was assessed using the QIIME2 [32] module; meanwhile, the beta diversity was measured based on ASV abundance using the Bray–Curtis Index and visualized through the principal coordinate analysis (PCoA) plot. Results were visualized in the Marker Data Profiling module of Microbiome Analyst version 2.0 [33]. Functional prediction based on 16S DNA sequences was evaluated using the PICRUSt2 software (v. 2.4.1) to explore possible metabolic mechanisms associated with rhizospheric bacterial communities under three conditions: PRE, DUR, and POST-drought [34]. Predictions were generated by matching marker gene data to reference genomes in databases, including MetaCyc [35].

3. Results

3.1. Soil Analysis

The results of the soil analysis revealed that the soil has a loam texture, a pH of 5.98, an organic matter content of 2.56%, nitrogen at 28.2 mg/kg, available phosphorus at 61.2 mg/kg, and available potassium at 259 mg/kg.

3.2. Vegetative Response Variables to Drought

Drought conditions significantly affect the physiological, biochemical, and morphological functions of plants. No statistical differences were observed before the onset of the drought. However, vegetative response measurements during drought showed reductions in leaf area, chlorophyll, and relative water content, except for root biomass, which did not differ between treatments under the three conditions. After rehydration for each treatment, measurements of vegetative responses showed differences in leaf area in treatments three and four, which experienced the longest drought periods. Statistically significant differences were also observed in the chlorophyll content in treatments three and four. The relative water content stabilized after rehydration, with no statistical differences observed. Proline levels continued to increase after rehydration in treatments three and four compared to treatments one and two; however, levels remained lower than those under drought conditions. Sugar levels increased in the control treatment after rehydration compared with those in the other treatments. These results were opposite to those observed under drought conditions, where sugar levels increased in the prolonged drought treatment (treatment four) (Table 1).

3.3. Rhizospheric Microbiota

After filtering and removing the chimeric sequences, 1991 amplicon sequencing reads were obtained. No significant differences (p > 0.05) were observed when scores for observed, Chao1, Shannon, and Simpson indexes, corresponding to pre-, during, and post-drought, were analyzed as primary indicators for the alpha diversity profile (Figure 1). However, differences for specific taxa at different taxonomic levels are presented and discussed below.

The sequencing results were taxonomically assigned to 16 bacterial phyla, 29 classes, 59 orders, 85 families, and 150 bacterial genera.

The most abundant phylum in all samples was Actinobacteriota, followed by Pseudomonadota and Acidobacteriota, across all three irrigation conditions (PRE, before drought; DUR, during drought; and POST, after drought) (Figure 2). When comparing the three conditions, no significant differences were observed, nor were there differences between PRE and DUR or PRE and POST, except between DUR and POST, where statistical differences were found for the Gemmatimonadota and Pseudomonadota phyla (p ≤ 0.05) (Table 2). The Actinobacteriota phylum showed a decreasing trend in relative abundance during the DUR and POST conditions compared to PRE (PRE 0.712 vs. DUR 0.670 vs. POST 0.536), whereas the Acidobacteriota phylum showed an increase in relative abundance during DUR and POST compared to PRE (PRE 0.027 vs. DUR 0.040 vs. POST 0.058). The Proteobacteria phylum showed the lowest abundance during the DUR condition but increased in the POST condition (PRE 0.161 vs. DUR 0.141 vs. POST 0.288, respectively) (Table 2) (Table S1).

The most representative bacterial families in all samples and across the three irrigation conditions were Acidobacteriaceae (Subgroup_1), Rhodanobacteraceae, unclassified bacteria, Solirubrobacteraceae, Acidothermaceae, Microbacteriaceae, Acetobacteraceae, Sphingomonadaceae, Geodermatophilaceae, and unclassified Actinobacteria (Figure 2). When comparing the three irrigation conditions, BEF vs. DUR, DUR vs. AFT, and BEF vs. AFT, no significant differences were observed in relative abundance at the family level (p ≤ 0.05). However, in the comparison of BEF vs. DUR, there were differences in the Chitinophagaceae family, and in the comparison of DUR vs. POST, the Gemmatimonadaceae, unclassified Frankiales, Burkholderiaceae, Rhodanobacteraceae, and Acidothermaceae families did show significant differences (p ≤ 0.05) (Table 2). The Rhodanobacteraceae family showed a marked decrease during the DUR condition, whereas in POST, an increasing trend was observed compared to PRE (PRE 0.028 vs. DUR 0.006 vs. POST 0.063). The Acidothermaceae group showed an increase in relative abundance during DUR compared to PRE and POST (PRE 0.031 vs. DUR 0.050 vs. POST 0.017). The Microbacteriaceae and Sphingomonadaceae families had increased abundance in the POST condition compared to PRE and DUR, while the Geodermatophilaceae and unclassified Actinobacteria groups showed higher relative abundance in the PRE condition and a decreasing trend during DUR and POST, respectively (Figure 3) (Table S2).

The most representative genera in the samples and irrigation conditions were items from the unclassified Actinobacteria group, followed by Sphingomonas, unclassified bacteria, Geodermatophilus, Conexibacter, Acidothermus, Microbacterium, Jatrophihabitans, unclassified Geodermatophilaceae, and Bacillus (Figure 2). In the comparison of the three irrigation conditions (PRE, DUR, and POST), only the Burkholderia-Caballeronia-Paraburkholderia and Streptacidiphilus groups showed significant differences (p ≤ 0.05). In the comparison of PRE vs. DUR, the unclassified Chitinophagaceae group showed a statistical difference. No statistical differences were observed between the genera in the PRE vs. POST comparison, whereas in the DUR vs. POST comparison, statistical differences were observed in the genera Acidothermus, the Burkholderia-Caballeronia-Paraburkholderia group, Crossiella, Dyella, Granulicella, Terracidiphilus, unclassified Frankiales, and unclassified Gemmatimonadaceae (p ≤ 0.05) (Table 2). The unclassified bacteria group showed a trend toward increased relative abundance in the DUR condition compared to PRE and POST (PRE 0.029 vs. DUR 0.053 vs. POST 0.042), as did Acidothermus (PRE 0.031 vs. DUR 0.050 vs. POST 0.017). Other genera that showed enrichment in relative abundance under the DUR condition compared to PRE and POST were Bacillus (PRE 0.024 vs. DUR 0.032 vs. POST 0.013) and Conexibacter (PRE 0.024 vs. DUR 0.047 vs. POST 0.032). In contrast, the genus Jatrophihabitans showed a decreasing trend in relative abundance during DUR compared to PRE and POST (Table S3). The analyses showed that most statistically significant differences occurred between DUR and POST drought conditions (Table 2).

The principal coordinate analysis (PCoA) corroborated that the samples differed in terms of relative abundance. A clear separation was observed among the three irrigation conditions, with the DUR and POST conditions showing some overlap, suggesting that samples from these conditions were more similar to each other than those from the PRE condition. The separation of points indicates significant differences in the characteristics of the samples and irrigation conditions (Figure 4)

3.4. Functional Prediction of Bacterial Communities

A total of 40 metabolic pathways were predicted as the most relevant to establish the functionality of rhizospheric bacterial communities in native maize plants under the three drought conditions: PRE, DUR, and POST. In the PRE-drought condition, the frequency of metabolic pathways was much higher than that in the DUR condition, where the frequency of most metabolic pathways was drastically reduced, except for the RND superfamily transporter pathway, acyl-CoA dehydrogenases, branched-chain amino acid transport systems, amidase enzymes, major subunits of acetolactate synthase (ALS), and aldehyde dehydrogenase (ALDH) enzymes, which were more highly expressed in the DUR condition (Figure 5).

4. Discussion

4.1. Soil Analysis and Vegetative Responses to Drought

Drought affects the physical, chemical, and microbiological composition of soil. Organic matter makes up approximately 5% of the soil structure and decreases dramatically in soils with low precipitation [36], which is consistent with the results obtained from the soil analysis. The lack of mobility and availability of nutrients and water is also a consequence of drought, due to the accumulation of salts that influence osmotic activity and nutrient translocation [37]. Comparing the results, there was a low availability of most macro- and micronutrients essential for the plants. Stress responses were observed in plants under different irrigation regimes. In the DUR condition, as stress increased, there was a reduction in leaf area and basal stem diameter due to water scarcity, with both characteristics being severely affected. References [38,39] reported an alteration in maize growth conditions, expressed as a reduction in leaf blade and vascular meristematic tissue. The physiological function of the plants was affected by water shortage, as evidenced by reductions in the chlorophyll index and leaf water content. Several authors [40,41] have reported a strong relationship between leaf vigor, plant stability, and early senescence, which directly impacts yield. Water relations showed stability after recovery irrigation, suggesting that the plant has the ability to improve its water balance despite the lack of soil moisture [42,43]. The biochemical responses of plants are also activated under drought conditions. Khan et al., 2025 [44] stated that plant tolerance to water shortage is mediated by osmoprotectants that prevent water loss and maintain cellular turgor. In addition to its antioxidant activity, proline is considered an important osmoprotectant that prevents water loss. The results of this study are consistent with those of Cortés-Patiño et al., 2022 [21,45]. During drought conditions, a reduction in foliar sugar production is observed [46]; however, Pelleschi et al. (1997) [47,48] reported an increase in sugar concentration, as the plant expresses this as a physiological response to water stress. Plant defense mechanisms are related to the increased concentration and translocation of soluble sugars in plants. Increased sugar levels are also reported to have osmotic activity during drought, aiming to regulate water balance and prevent loss of cellular turgor [49,50].

4.2. Rhizosphere Microbiota

Drought severely affects the physicochemical conditions of the soil, resulting in changes in bacterial communities [51]. Studies have also confirmed that plant genotype is an important factor influencing the composition and selection of bacterial rhizosphere microbiota because of the type of root exudates secreted by the plant [52]. Various studies on rhizosphere soil have found that the phyla Pseudomonadota, Bacteroidota, and Acidobacteriota are the most abundant [53,54]. These results are consistent with those obtained in this study, except for Bacteroidota, which ranked ninth in relative abundance. Guevara et al. (2024) [55] also reported Pseudomonadota, Actinobacteriota, and Acidobacteriota as the dominant phyla, results that align with our study, where we found an average abundance of 19%, 63%, and 4%, respectively. Several studies have confirmed the enrichment of Actinobacteriota during severe droughts, which decreases once the water supply is restored [56]. This proves that drought induces changes in diversity patterns in certain taxonomic groups, which is consistent with our findings. Other phyla present in the native maize rhizosphere have been reported, including Bacteroidota, Chloroflexota, Verrucomicrobiota, and Gemmatimonadota, with an abundance greater than 2% [57], which is consistent with our results, although the relative abundances of these phyla were below 2%. In this study, Gemmatimonadota and Pseudomonadota showed statistically significant differences in relative abundance, similar to the findings of López et al. (2023) [58]. Differences in the abundance of these phyla were observed between drought and post-drought conditions, revealing that water deficit and subsequent rehydration influenced the populations of Gemmatimonadota and Pseudomonadota [59,60]. Some bacteria benefit plants under stressful conditions. In this study, statistically significant differences were found in the families Chitinophagaceae, Gemmatimonadaceae, unclassified Frankiales, Burkholderiaceae, Rhodanobacteraceae, and Acidothermaceae, mostly during and after drought, except for Chitinophagaceae. All these families have been reported as part of the rhizosphere microbiome of native maize. The Chitinophagaceae family includes bacteria known to promote plant growth through mechanisms such as enhanced nutrient and water absorption. They are also linked to ACC deaminase hydrolase activity, which reduces ethylene levels, thus inducing adaptation and tolerance to water stress [61]. Gemmatimonadaceae is considered part of the rhizosphere bacterial recruitment selected by the plant during water stress, as it is adapted to low-moisture and arid environments and is involved in C fixation and the N cycle [62,63]. There is evidence that Frankiales bacteria establish mutualistic symbiosis with actinorhizal plants and play a role in nitrogen fixation [64]. Some members of Burkholderiaceae and Rhodanobacteraceae are known to promote plant growth under biotic and abiotic stress [65,66]. The Acidothermaceae family is mainly associated with improving soil nutrients through cellulose decomposition and promoting plant resilience to environmental stress [67]. On the other hand, some species of the genus Terracidiphilus, also part of the rhizosphere microbiome, have been attributed with enzymatic functions for degrading organic matter and plant-derived biopolymers [68]. Strains of the genus Dyella promote plant growth by modifying root architecture and improving biomass and shoot weight in Arabidopsis and tomato plants [69]. Several genera mentioned in this study have been reported to promote growth and hormonal modulation, especially abscisic acid (ABA), known as the “plant stress hormone” [70], which is involved in various cell signaling processes under extreme environmental conditions, particularly drought [71].

5. Conclusions

The drought model used revealed stress responses in the plants, affecting their growth capacity and physiological functions. The bacterial diversity analyses demonstrated changes in the composition, structure, and relative abundance of bacterial communities between normal irrigation and water deficit conditions. The phylum-level taxa associated with drought were Actinobacteriota and Pseudomonadota, as well as the families Microbacteriaceae, Sphingomonadaceae, and Unclassified_Actinobacteria, which, although they did not show significant differences, exhibited an increase in their relative abundance under water deficit conditions. This suggests a possible association with the plant’s drought tolerance. The water stress condition revealed diverse metabolic expressions of bacterial genes, which establish connections with the plant mainly through signaling pathways and metabolic precursors. These results provide important information to further the understanding of the relationships between the plant and its bacterial microbiota in situations of varying precipitation.