Linking topography, soil physicochemical traits and bacterial community dynamics in a managed tea landscape

Abstract

Topography significantly shapes soil properties and microbial community structure, influencing ecosystem function and agricultural productivity. This study examined the upland-lowland gradient in a tea-growing landscape to assess how topographic variation affects soil physicochemical and biological properties, and bacterial community composition. Soils were sampled from upland tea fields, lowland paddy fields, and an intermediate transition zone. Upland tea soils were highly acidic (pH 3.7–4.5) with higher organic carbon content, favoring the presence of Acidobacteria, Patescibacteria, and other oligotrophic taxa known for roles in organic matter turnover. Lowland paddy soils, in contrast, exhibited higher moisture and enzymatic activities corresponds to anaerobic and facultative anaerobic taxa such as Moraxellaceae and Rhodobacteraceae. The transition zone, characterized by moderate pH, organic carbon, and minimal human disturbance, exhibited the highest bacterial diversity, including nitrogen-cycling taxa like Burkholderiaceae. Organic carbon was strongly correlated with community composition, particularly with the abundance of Actinobacteria. These findings highlight the strong interdependence between topographic variation, soil health, and microbial functional processes. Implementing topography-informed soil management practices is critical for preserving microbial diversity, enhancing nutrient cycling efficiency, and sustaining long-term soil fertility within agricultural ecosystems.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10123-026-00777-5.

Introduction

Tea (Camellia sinensis), as a globally significant cash crop, is deeply intertwined with human civilization and plays a critical role in driving regional economies and maintaining ecological balance. Among the diverse environmental factors influencing tea cultivation, land topography stands out as a key determinant of soil properties, microbial diversity, and ultimately, crop productivity. Across major tea-producing regions worldwide, elevation, slope, and drainage patterns have shaped tea quality and yield. Proper topographical conditions optimize sunlight exposure and temperature regulation, which are essential for producing high-quality tea liquor. However, excessively steep slopes can exacerbate challenges such as soil erosion, nutrient leaching, and water runoff, leading to declines in soil fertility and crop resilience. Consequently, tea plantations have historically been established in landscapes where the natural terrain supports sustainable growth and minimizes environmental degradation stability (Wen et al. 2020; Jayasinghe et al. 2020).

The influence of topography on tea cultivation is further exemplified by geographical variations in tea varieties and cultivation strategies. In China and Japan, Camellia sinensis var. sinensis predominates, thriving at higher altitudes characterized by cooler temperatures and well-drained soils. Tea-growing areas such as mountain slopes, coastal regions, and river basins create distinct microenvironments that promote optimal tea production. Conversely, Camellia sinensis var. assamica, better adapted to lower elevations and humid conditions, is extensively cultivated across India, Sri Lanka, Vietnam, and Indonesia (Durighello et al. 2021). These regional differences underscore the intricate relationship between topography, climate, and tea cultivation practices.

Beyond shaping plant growth, topographical features exert a profound impact on soil physicochemical properties, hydrological processes, and microbial community structures. Variations in elevation and slope directly affect water transport, soil aeration, and nutrient distribution, thereby influencing tea plant physiology and productivity (Baltensweiler et al. 2020). On finer spatial scales, microtopographical features generate heterogeneity in soil moisture, organic matter content, and microbial activity even across short distances. Moreover, vegetation, particularly tea roots, modifies the soil environment through the release of root exudates that selectively shape microbial community composition and is associated with nutrient cycling processes (Pawlik et al. 2015; Bacilio-Jimenez et al. 2003).

Microorganisms are indispensable components of tea garden ecosystems, serving as critical drivers of organic matter decomposition, nutrient mineralization, and soil health maintenance. Their diversity and functional activity are closely tied to environmental parameters such as organic matter availability, soil moisture, and plant diversity, with notable shifts occurring even over microtopographical gradients (Xu et al. 2015; Shigyo et al. 2022). Growing evidence from agroecosystem research indicates that fine-scale variation in soil conditions associated with vegetation differences is linked with changes in microbial community composition and functional profiles, emphasizing the importance of vegetation-mediated heterogeneity in shaping soil ecosystem processes (Li et al. 2024a, b, 2025). Understanding the complex interactions among topography, soil properties, and microbial diversity is thus essential for sustaining soil fertility, enhancing nutrient cycling, and improving the resilience and quality of tea crops.

However, despite increasing recognition of these relationships, comprehensive studies systematically examining the synergistic responses of soil chemical properties and microbial communities across slope gradients remain limited, particularly in topographically complex tea-growing regions. This study investigates the soil physicochemical characteristics and microbial community structures across different slope positions within a tea plantation, focusing on their interactions and functional implications. By elucidating slope-induced variations in soil chemistry and microbial dynamics, our findings aim to provide practical insights for slope-specific soil management, targeted fertilization strategies, and the promotion of sustainable tea plantation practices.

Methodology

Description of the selected study area

The study area is located within the Experimental Garden for Plantation Crops within Assam Agricultural University, Jorhat Campus, Assam. It encompasses 22 hectares of land with latitude of 26.716215˚, longitude of 94.19937˚ and is 91 m above mean sea level. The area is undulating in nature characterised by rolling hills and gentle slopes, and variations in elevation and slope. Surrounding these tea gardens are natural depressions intersected by streamlets or tea garden drains. The lower-lying areas with poor drainage are often used for paddy cultivation, which faces occasional flooding. The fertilizers used in the tea gardens are applied at a ratio of 100:25:75 kg/ha of nitrogen (N), phosphorus (P), and potassium (K), respectively. In contrast, the paddy fields receive fertilizers at a ratio of 60:40:40 kg/ha of N, P, and K, respectively. Besides tea and paddy, the garden also has a rubber plantation, coffee plantation and a pepper vine plantation. The area is located in the Upper Brahmaputra Valley (UBV) agro-climatic zone which experiences both hot summers (22–32 °C) and cold winters (6–12 °C) with an average annual rainfall of 2244 mm. The relative humidity remains above 70% throughout the year.

Topographic characterization of sampling zones

In this present study, topography is defined as a fine-scale upland–lowland gradient characterized by relative height differences, slope position, drainage, and surface hydrological movement, rather than broad altitudinal/mountainous elevation variations. The upland tea, transition, and lowland paddy zones were distributed along a gently undulating hillslope with an approximate vertical relief of ~ 2 m over a horizontal distance of ~ 30–40 m, corresponding to a low-gradient slope. Based on field measurements and visual surveying, the slope gradient ranged from approximately 1–3%, typical of gently undulating tea landscapes in the Upper Brahmaputra Valley of Assam. The tea plantations occupied the upper slope with well-drained soils, the transition zone represented the mid slope, and the paddy fields were situated at the foot slope, receiving lateral runoff and surface water from higher elevations.

Soil sample collection

Soil samples were collected from the 0–10 cm layer within a 7 m × 6 m area across these three land-use zones. For the purpose of topographic characterization, the upland tea plantation, transition zone, and lowland paddy fields were designated as upper slope, mid slope, and foot slope positions, respectively, along a gentle hillslope. From each zone, soil was sampled at three designated collection points. At each point, multiple soil cores were combined to form one composite sample, resulting in a total of three composite soil samples per zone (9 composite samples in total). Subsamples of the composite soils were stored at − 80 °C for amplicon sequencing, while the remaining portions were used for soil physicochemical analyses.

Analysis of soil properties

Soil pH was measured using a 1:2.5 soil-to-water suspension, while electrical conductivity was assessed in a 1:2 soil-to-water suspension. Soil moisture content was determined following the protocol by Benke and Kearfott (1999). Soil organic carbon was quantified using the Walkley-Black method, where 0.5 g of soil was combined with 10 ml of 1 N K₂Cr₂O₇ and 20 ml of concentrated H₂SO₄ in a 250 ml conical flask. The mixture was swirled, incubated for 30 min at room temperature, and then diluted with 200 ml of distilled water. Phosphoric acid (10 ml) and 1 ml of diphenylamine indicator were added, followed by titration with 0.04 N ferrous ammonium sulfate until the solution turned green (Walkley and Black 1934). The procedure was performed in triplicate, with blanks prepared without soil. Available nitrogen in the soil samples was measured using the Subbiah and Asija (1956) method. Five grams of soil were placed in an 800 ml Kjeldahl digestion flask along with 100 ml of 0.32% KMnO₄, 100 ml of 2.5% NaOH, and 20 ml of water. The flask was connected to a Kjeldahl distillation unit, and 75 ml of distillate was collected. To this, 25 ml of boric acid and a mixed indicator (bromocresol green and methyl red) were added. The absorbed ammonia was titrated with 0.02 N H₂SO₄ (Bremner 1960). For available phosphorus, 2 g of soil was extracted using Bray’s solution (a mixture of 0.03 N NH₄F and 0.025 N HCl). The soil suspension was shaken for one minute and centrifuged at 3000 rpm for 5 min. The supernatant was filtered through Whatman No. 42 filter paper, and 25 ml of filtrate was diluted with an equal volume of distilled water. Next, 10 ml of 1.5% ammonium molybdate, 7.5 ml of 0.8 M boric acid, and 1 ml of 40% stannous chloride were added until the solution turned blue. Phosphorus concentration was determined using a standard curve made with potassium dihydrogen phosphate (0–5 ppm), and absorbance was measured at 880 nm using a UV-Vis spectrophotometer, as described by Bray and Kurtz (1945). All analyses were performed in triplicates.

Analysis of soil extracellular enzymes

Soil enzymes, including phosphomonoesterase, glucosidase, urease, and arylsulfatase, were measured and calculated in the collected samples following established protocols. For enzyme assays, 1 g of soil was placed in a 50 ml Erlenmeyer flask, followed by the addition of 0.25 ml toluene and 4 ml of Modified Universal Buffer (MUB), which was prepared by combining Tris (12.1 g/L), maleic acid (11.6 g/L), citric acid (14 g/L), and boric acid (6.3 g/L) in 500 ml NaOH and diluted to 1 L at pH 6.5. For arylsulfatase, acetate buffer (0.5 M, pH 5.8) was used instead. The enzymatic reactions were initiated by adding 1 ml of specific substrates: p-nitrophenyl phosphate (PNP) for phosphomonoesterase, p-nitrophenyl-β-D-glucoside (PNG) for glucosidase, and p-nitrophenol sulfate for arylsulfatase. The reaction mixtures were thoroughly mixed, incubated at 37 °C for 1 h, followed by the addition of 1 ml of 0.5 M CaCl₂ and 4 ml of 0.5 M NaOH. After thorough mixing, the mixtures were filtered through Whatman No. 2 filter paper. Controls were prepared by adding the substrate after CaCl₂ and NaOH. Absorbance readings were taken at 400 nm using a UV-Vis spectrophotometer, and a calibration curve was constructed using p-nitrophenol standards (0–3 ppm). All measurements were performed in triplicates.For urease activity, 5 g of soil was treated with 2.5 ml of a freshly prepared urea solution (2.4 g urea in 500 ml distilled water) in a 100 ml Erlenmeyer flask. The sample was incubated for 2 h at 37 °C, after which 50 ml of KCl solution (prepared by dissolving 74.6 g KCl in distilled water, adding 10 ml of 1 M HCl, and diluting to 1 L) was added and shaken for 30 min. The filtrate was analyzed for ammonium content. Blanks were prepared by adding urea after the incubation. Ammonium was determined by combining 1 ml of the filtrate with 9 ml of distilled water, 5 ml sodium salicylate/NaOH solution (0.3 M), and 2 ml sodium dichloroisocyanide (0.1%), followed by incubation at room temperature for 30 min. Absorbance was measured at 690 nm using a UV-Vis spectrophotometer, and a calibration curve was constructed using ammonium standards (0–2.5 µg NH₄⁺-N/ml).For fluorescein diacetate (FDA) hydrolysis, 1 g of soil was combined with 50 ml of 60 mM sodium phosphate buffer (pH 7.6) and 0.50 ml of FDA substrate solution (6.0 µM final concentration). The mixture was incubated at 24 °C for 3 h under shaking conditions, and the reaction was stopped by adding 2 ml of acetone. The soil suspension was centrifuged at 8000 rpm for 5 min, and the supernatant was filtered through Whatman No. 2 filter paper. The filtrate’s absorbance was measured at 490 nm using a spectrophotometer. All measurements were conducted in triplicate (Tabatabai and Bremner 1969, 1970; Tabatabai 1982; Kandeler and Gerber 1988; Green et al. 2006).

Analysis of soil metals

A DTPA (diethylene triamine penta-acetic acid) extracting solution was prepared by mixing 0.005 M DTPA, 0.01 M CaCl₂, and 0.1 M TEA (triethanolamine) solutions. Twenty ml DTPA extracting solution was added to a flask containing 10 g of soil and kept in a shaker for 2 h followed by filtering the content through Whatman No. 42 filter paper. Ten standards were prepared using DTPA for each element ranging from 0 to 3 ppm for Zn and K, 0 to 20 ppm for Fe and Mn. A blank solution containing all reagents except soil was run to correct for contamination (Lindsay and Norvell 1978). Each experiment was done in triplicates.

Amplicon sequencing

DNA was isolated from the soil samples using the DNeasy PowerSoil Pro Kit and quantified using the Qubit dsDNA HS Assay Kit. The V3–V4 hypervariable region of the 16 S rRNA gene was amplified using Prokaryote V3-Forward (CCTACGGGNBGCASCAG) and V4-Reverse (GACTACNVGGGTATCTAATCC) primers (~ 460 bp amplicon). Amplicon libraries were prepared using the Nextera XT Index Kit (Illumina), purified using 1× AMPure XP beads, and evaluated using the Agilent Bioanalyzer DNA1000 chip, yielding a mean library size of 626–631 bp. Libraries were quantified with the Qubit dsDNA HS Assay and multiplexed equimolarly.

Sequencing was performed on the Illumina MiSeq platform using 2 × 250 bp paired-end chemistry, generating approximately ~ 150 Mb of data per sample. Raw FASTQ reads were processed in QIIME2. After primer trimming, joining of paired reads, and Phred quality filtering, reads were denoised using the Deblur algorithm to infer amplicon sequence variants (ASVs). Taxonomic assignment was performed using the SILVA 132 reference database with the classify-sklearn classifier. Mitochondrial and low-abundance ASVs were removed prior to downstream analyses.Across the dataset, a total of 2,983,861 paired-end reads were generated, with 339,538 high-quality ASV reads retained for downstream analyses. The raw FASTQ files have been deposited in the NCBI SRA under accession number PRJNA1169761.

Statistical analyses

All experiments were performed in triplicates. Student’s t-test was performed along with analysis of variance followed by Tukey HSD to determine statistical difference using IBM SPSS Statistics (Version 20). The values in the table are represented as mean ± standard error. Alpha diversity indices were calculated using Past software (Version 4.03). All figures were generated in Rstudio v3.4. Stacked bar plots and boxplots were generated with the ggplot package, while stacked donut charts were created using the plotly and dplyr packages. The RDA analysis was performed with the vegan package. Genus-level OTU tables were imported into MicrobiomeAnalyst for beta-diversity analysis. OTU counts were normalized and Bray–Curtis dissimilarity matrix was computed and was visualized using Principal Coordinates Analysis (PCoA). Statistical significance of group differences was tested using PERMANOVA with 999 permutations on the Bray–Curtis distance matrix. Pairwise PERMANOVA was also applied to assess differences between individual zones. Slope position (upper slope, midslope, and footslope) was included as a categorical proxy for micro-topographic variation in multivariate analyses examining soil properties and bacterial community composition. A phylum-level co-occurrence network was constructed based on Spearman rank correlations among microbial relative abundances across the lowland paddy, intermediate transition zone, and upland tea soils using Cytoscape v.3.10.3.

Results

Soil physicochemical and enzymatic properties

The relative elevation of sampling zones decreased sequentially from upland tea to lowland paddy soils, with an approximate vertical drop of ~ 2 m, sufficient to alter drainage class from well-drained (upland) to periodically waterlogged (lowland). The study site is characterized by a gentle slope, with no abrupt breaks in relief. Although high-resolution Digital Elevation Model (DEM) data were not available, field observations indicate a low-gradient slope (< 3%), typical of managed tea gardens in Assam. The study area did not exhibit a pronounced aspect-related exposure; thus, aspect-driven differences in solar radiation were minimal compared to hydrological effects associated with slope position.

The pH of the entire sampling area ranged from 3.7 to 5.5 (Fig. S1). Moisture was found to gradually increase from upland tea soil to lowland paddy region (3.19 ± 0.16 to 6.82 ± 0.61%). Soil electrical conductivity significantly decreased from higher pH (paddy) to lower pH (tea). Despite the relatively small vertical relief (~ 2 m), clear topography-driven gradients were evident across the upland–lowland system. Soil moisture increased downslope, reflecting enhanced water accumulation at lower slope positions. This micro-topographic control on hydrology strongly influenced soil pH, organic carbon distribution, and enzyme activities, demonstrating that even subtle elevation differences can generate pronounced biogeochemical heterogeneity. Organic carbon was found to be highest in the transition and tea cultivated upland region (5.07 ± 0.11% and 4.48 ± 0.23%, respectively), while available nitrogen, available zinc, and enzymes such as glucosidase, arylsulphatase and urease were found to be highest in the soils from lowland region (729.86 ± 46.85 kg/ha, 4.2 ± 0.67 mg/kg, 36.9 ± 3.42 µg p-nitrophenol g^− 1 h^− 1, 27.5 ± 0.79 µg p-nitrophenol g^− 1 h^− 1, and 834.75 ± 18.09 µg NH₄-N g^− 1 2 h^− 1respectively). The FDA hydrolysis, a measure of soil microbial activity, was found to be higher in the tea zone ((9.35 ± 0.40 µg fluorescein g^− 1h^− 1) (Tables 1 and 2).

Table 1.

The mean values of soil physicochemical properties of tea, paddy and transition zones

Zones	pH	Mo	EC	AP	OC	AN	Fe	Mn	Zn	K
Paddy	5.1 ± 0.02a	6.82 ± 0.61a	0.58 ± 0.05a	23.87 ± 0.28a	3.78 ± 0.13a	729.86 ± 46.85a	7.13 ± 0.68a	23.43 ± 3.67a	4.2 ± 0.67a	177.13 ± 30.75a
Transition	4.62 ± 0.04b	5.51 ± 0.53bc	0.38 ± 0.03ab	24.51 ± 1.11a	5.07 ± 0.11b	419.95 ± 26.88bc	6.56 ± 0.11a	25.89 ± 0.64a	1.61 ± 0.17ab	187.86 ± 18.84a
Tea	4.09 ± 0.11c	3.19 ± 0.16c	0.33 ± 0.07b	26.52 ± 1.1a	4.48 ± 0.23b	358.83 ± 18.52c	6.77 ± 0.08a	31.21 ± 9.03a	1.42 ± 0.14b	164.94 ± 41.69a

Abbreviations: Mo = moisture, EC = Electrical Conductivity, AP = Available Phosphorus, OC = Organic Carbon, AN = Available Nitrogen, Fe = Iron, Mn = Manganese, Zn = Zinc, K = Potassium. Mean separation by tukey’s significant difference Test. In each column, figures followed by similar letters are not significant. P-values significance: ‘**’ p-value < 0.01; ‘*’ p-value < 0.05. n = 3

Table 2.

The mean values of soil biological properties of tea, paddy and transition zones

Zones	Pho	Glu	Ars	Ur	FDA
Paddy	24.82 ± 3.97a	36.9 ± 3.42a	27.5 ± 0.79a	834.75 ± 18.09a	7.82 ± 0.27a
Transition	23.45 ± 5.24a	16.11 ± 2.82bc	23.87 ± 1.33ac	543.96 ± 5.06b	5.58 ± 0.47b
Tea	24.87 ± 1.52a	10.8 ± 1.42c	15.87 ± 2.34b	610.4 ± 36.9b	9.35 ± 0.4ac

Abbreviations: Pho = Phosphomonoesterase, Glu =- β-D-Glucosidase, Ars = Arylsulfatase, Ur = Urease, FDA = Fluorescein Diacetate. Mean separation by tukey’s significant difference Test. In each column, figures followed by similar letters are not significant. P-values significance: ‘**’ p-value < 0.01; ‘*’ p-value < 0.05. n = 3

Composition of soil bacterial communities

The analysis of bacterial communities at the phylum level revealed Patescibacteria (38.58%) to be dominant in the lowland paddy soils, Proteobacteria (48.61%) in the transition zone and Acidobacteria (60.03%) in the upland tea soils. At the class level, Alphaproteobacteria (16.54%) were most prevalent in the lowland region, Gammaproteobacteria (36.89%) dominated the transition zone, and Acidobacteriia (17.54%) were most abundant in the upland region. At the order level, Saccharimonadales (12.96%) were predominant in the lowland soils, Pseudomonadales (36.5%) in the transition zone, while Acidobacteriales (7.35%) was most abundant in the upland soils (Fig. 1). At the family level, the lowland was dominated by Moraxellaceae (24.74%), Rhodobacteraceae (5.14%), and Acetobacteraceae (3.41%); the transition zone by Solibacteraceae Subgroup_3 (7.51%), Burkholderiaceae (4.62%), and Gemmataceae (2.77%); and the upland by Sphingomonadaceae (8.37%), Ktedonobacteraceae (5.96%), and Moraxellaceae (5.39%). At the genus level, Paracoccus (13.23%) was the most abundant followed by Enhydrobacter (9.82%), and Corynebacterium (9.53%) in the lowland soils. The transition zone was dominated by Enhydrobacter (27.36%), Acinetobacter (16.24%), and Pseudomonas (7.81%). In the upland, Sphingomonas (12.16%), Bryobacter (5.43%), and Acinetobacter (3.16%) were the most prevalent genera (Fig. S3).

Fig. 1.

Stacked barplots representing the relative abundance of the bacterial population (top 15) at (a) Phylum level, (b) Class Level and (c) Order level in a tea plantation consisting of lowland paddy, transition, and upland tea soil

Bacterial diversity across the upland-lowland system

Alpha diversity indices, including Shannon, Simpson, Evenness, Margalef, and Chao1, were calculated based on operational taxonomic units (OTU) at the genus level to assess the biodiversity across the lowland paddy, transition, and upland tea zones (Fig. 2). The Shannon index, which measures species richness and evenness within a community, was found to be 3.03 ± 0.8 for the lowland region, 3.89 ± 0.54 for the transition zone, and 2.61 ± 0.22 for the upland region. This index indicates that the transition zone has the highest biodiversity, followed by the upland and lowland regions, with significant differences noted (p = 0.037). The Simpson index, which accounts for the relative abundance of species, showed values of 0.89 ± 0.07 for the lowland, 0.96 ± 0.01 for the transition zone, and 0.88 ± 0.01 for the upland. These results suggest that the transition zone has shown a higher dominance in bacterial population. Chao1, an estimator of total species richness, also revealed the highest richness in the transition zone at 136 ± 76.9, followed by the lowland region at 73.3 ± 45.2, and the upland region at 37 ± 14.4. The equity in species abundance as measured by the Evenness index, was highest in the transition zone (0.42 ± 0.07), followed by the upland region (0.39 ± 0.06), and lowest in the lowland region (0.34 ± 0.05). Lastly, the Margalef index, reflecting species richness, indicated values of 7.55 ± 5.24 for the lowland region, 15.72 ± 9.33 for the transition zone, and 3.64 ± 1.28 for the upland region. PCoA based on Bray–Curtis distances revealed distinct clustering patterns among the three zones. PC1 (48.2%), PC2 (17.1%), and PC3 (12.8%) together explained a substantial portion of the observed variation. Transition soil samples formed a relatively tight cluster, while paddy and tea soils displayed wider dispersion (Fig. 3).

Fig. 2.

Alpha diversity indices of the lowland paddy, transition and upland tea soil- a) Shannon, b) Simpson, c) Chao1, d) Margalef, and e) Evenness

Fig. 3.

Principal Coordinates Analysis (PCoA) of soil bacterial community composition across paddy (orange), transition (green), and tea (blue) zones based on Bray–Curtis dissimilarities. The first three axes explained 78.1% of the total variation (PC1 = 48.2%, PC2 = 17.1%, PC3 = 12.8%)

Co-occurrence network analysis of the bacterial population across the upland-lowland system

A phylum-level co-occurrence network was constructed to examine potential associations among bacterial taxa across the upland-lowland system, where the positive correlations are denoted by blue edges and the negative correlations by orange edges. It was observed from Fig. 4 that Proteobacteria, Acidobacteria, and Actinobacteria, and Firmicutes, emerged as prominent phyla characterized by larger nodes and extensive connections with other taxa. Several other phyla such as Patescibacteria, Gemmatimonadetes, Planctomycetes, and Latescibacteria also occupied central positions, indicating broad ecological connectivity despite moderate abundance levels. On the other hand, several low to moderate abundance phyla such as Rokubacteria, GAL15, Dependentiae, and Dadabacteria showed notable negative correlations.

Fig. 4.

Phylum level co-occurrence network analysis across the lowland paddy, intermediate transition zone, and upland tea soils. Node size and color intensity represent the mean relative abundance of each phylum across three land types; Edges in blue and orange color indicate positive and negative correlation, respectively

Responses of soil microbial community to soil physicochemical properties and enzyme activity

Redundancy analysis (RDA) was performed to examine the relationships between dominant bacterial phyla and soil physicochemical and biological properties across the upland–lowland gradient of the tea plantation (Fig. 5). The first two ordination axes explained a substantial proportion of the total variance, with RDA1 and RDA2 accounting for 43.83% and 27.19% of the variation, respectively. Organic carbon (OC), available nitrogen (AN), and β-D-glucosidase activity exerted the strongest influence on microbial community structure. Actinobacteria and Dependentiae showed a strong positive association with OC, whereas Patescibacteria, Bacteroidetes, and Proteobacteria were closely linked to soil enzyme activities, particularly arylsulfatase and glucosidase. Firmicutes was primarily associated with soil pH and zinc content. In contrast, Planctomycetes and Acidobacteria exhibited negative correlations with pH, OC, and enzymatic activities. Notably, soil moisture was largely regulated by slope position and relative elevation along the micro-topographic gradient and contributed indirectly to microbial differentiation by shaping the distribution of these key soil properties, underscoring the role of fine-scale topography in structuring soil bacterial communities.

Fig. 5.

Redundancy analysis (RDA) illustrating the relationships between dominant bacterial phyla and soil physicochemical and biological properties across the upland–lowland gradient of the tea plantation system. The first two axes (RDA1 and RDA2) explain 43.83% and 27.19% of the total variation, respectively. Soil moisture, which reflects slope position–mediated hydrological variation along the micro-topographic gradient, indirectly captures the influence of relative elevation on microbial community structure

Discussion

Influence of topography on soil physicochemical and biological properties

Although the absolute elevation difference across the study site was limited, previous studies have demonstrated that micro-topographic variation of even 1–2 m can strongly regulate soil moisture dynamics, redox conditions, and microbial community assembly, particularly in humid agroecosystems (Zhang et al. 2024a, b). In the present study, the downslope accumulation of moisture and nutrients amplified the ecological consequences of this small relief, resulting in distinct physicochemical and microbial properties across slope positions. The soils were predominantly acidic, with pH values ranging from 3.7 to 5.5, consistent with earlier reports for Assam soils (Goswami et al. 2017). High annual rainfall (~ 1500 mm) promotes the leaching of base cations, thereby accelerating soil acidification, while in lower-lying positions, groundwater inputs partially buffer soil pH by replenishing basic cations (Blaser et al. 2008). Along the micro-topographic gradient, soil moisture increased from upland tea plantations to lowland paddy fields as a result of downslope water movement and accumulation, creating periodically waterlogged conditions favorable for rice cultivation (Denardin et al. 2020). Soil organic carbon (OC) was highest in the transition and upland tea zones and showed a negative relationship with soil pH (Fig. S2), suggesting that organic matter decomposition and the accumulation of organic acids contribute to localized acidification and influence microbial substrate availability (Baruah et al. 2013; Zhou et al. 2019). In contrast, the comparatively lower OC content in lowland soils may reflect intensive tillage practices in paddy fields, which disrupt soil aggregates and accelerate organic matter mineralization. Enzymatic activities, which serve as sensitive indicators of nutrient cycling and microbial metabolism, were also strongly regulated by topography-mediated soil conditions. Elevated β-D-glucosidase, urease, and arylsulfatase activities in lowland soils indicate intensified organic matter turnover and nutrient release under moist, nutrient-enriched conditions (Karaca et al. 2011). These patterns suggest that microbial-driven decomposition processes are amplified in lower slope positions, reinforcing feedbacks between topography, soil chemistry, and biological function. In contrast, higher FDA hydrolysis rates in tea-growing upland soils likely reflect sustained organic inputs from perennial tea plants, supporting active microbial communities (Tokuda and Hayatsu 2002). Collectively, these findings demonstrate that even modest micro-topographic differences can generate pronounced heterogeneity in soil environments by regulating hydrological pathways, resource availability, and microbial habitat quality, thereby shaping ecosystem functioning.

Bacterial community composition across the upland-lowland system

Soil microorganisms are essential to ecosystem functioning, with shifts in microbial community composition often influenced by changes in soil properties along slope gradients. The present study identifies clear variations in soil microbial communities between upland and lowland soils in tea gardens.The variation in bacterial community composition closely mirrors the patterns observed in soil physicochemical and biological properties. In upland soils, the dominance of Patescibacteria and Acidobacteria, including genera like Bryobacter and Ktedonobacteria, reflects their adaptation to acidic, nutrient-poor conditions with low pH and higher organic carbon. The phylum Patescibacteria are believed to generate energy through anaerobic fermentation, utilizing compounds released by the action of the many glycoside hydrolases they encode, which are predicted to break down a wide range of complex carbohydrates (Srinivas et al. 2024). Additionally, they are also reported to be involved in the nitrogen, sulfur, and iron cycles (Herrmann et al. 2019). An increase in the abundance of the phylum Patescibacteria reflects an enhancement of certain microbial groups involved in elemental cycling. This rise in beneficial microorganisms can strongly impact the rhizosphere environment (Li et al. 2024a, b).In lowland paddy soils, the high moisture and reduced oxygen, combined with frequent ploughing, modulated both the microbial and soil structural dynamics. The prevalence of anaerobic and facultative anaerobic taxa such as Moraxellaceae (Paracoccus, Enhydrobacter) and Rhodobacteraceae corresponds with elevated enzymatic activity (e.g., β-D-glucosidase and arylsulfatase), suggesting active carbon and sulfur cycling (Pujalte et al. 2014; Si et al. 2022). The transition zone, with high urease activity, supported Proteobacteria, including nitrogen-cycling families like Burkholderiaceae, and Gemmataceae (Omotayo et al. 2022). Here, the sparse vegetation and low human intervention coupled with moderate nutrient status likely fostered a balanced microbial network capable of supporting both nitrification and organic matter decomposition, enhancing soil functional resilience.

Bacterial diversity and its ecological implications

Previous studies have demonstrated that soil microbial community composition varies across different slope gradients, with microtopographic differences strongly shaping soil physicochemical properties and, in turn, influencing microbial community structure (Li et al. 2024a, b, 2025; Suriyavirun et al. 2019).Bacterial diversity exhibited significant variation across the upland, transition, and lowland zones, with the transition zone displaying the highest alpha diversity indices. Beta diversity analysis revealed the bacterial population in the lowland paddy zone appeared to be spread out indicating greater heterogeneity in microbial community structure. Paddy fields are exposed to episodes of flooding and drying which can alter soil nutrient composition, pH and redox potential. This may influence bacterial communities to shift significantly and stabilise over time via metabolic responses such as increased nitrification (de Freitas et al. 2024). In the upland tea zone, reduced microbial diversity is likely a consequence of continuous monocropping. Such practices limit the diversity of organic inputs and root exudates, thereby suppressing microbial heterogeneity (Gui et al. 2021). Additionally, low pH conditions may select for acidophilic bacteria, narrowing the microbial community dynamics. Conversely, the lowland paddy fields, though rich in moisture, exhibited low bacterial diversity, potentially due to lower levels of soil organic carbon and available potassium. The transition zone, on the other hand, had minimal human disturbance and sparse vegetation. This indicates reduced rhizosphere dominance, which can otherwise favour specific microbial groups and limit community complexity. The more heterogeneous resource environment and moderate disturbance likely created a mosaic of microhabitats, promoting coexistence and ecological resilience.The observed trend aligns with previous studies reporting a positive correlation between bacterial diversity and near-neutral pH, as extreme acidic or alkaline conditions tend to constrain microbial adaptation and enzymatic activity (Zheng et al. 2022; Bartram et al.2014).

Co-occurrence network analysis of the bacterial population across the upland-lowland system

The phylum level co-occurrence network across the lowland paddy, intermediate transition zone, and upland tea soils (Fig. 4) reflects a complex and tightly coupled microbial community, wherein dominant phyla play a central role in shaping community structure. The presence of both cooperative and antagonistic interactions reflects the ecological interdependencies and competitive dynamics shaping soil microbiomes along the land-use gradient. Proteobacteria, Actinobacteria, Firmicutes, and Chloroflexi emerged as prominent phyla, showing numerous positive associations with other taxa. This suggests coordinated responses to soil physicochemical conditions and shared ecological niches across the three zones. These phyla appear to serve as core structural components influencing microbial community stability and functional potential. Several other phyla such as Patescibacteria, Gemmatimonadetes, Planctomycetes, and Latescibacteria also occupied central positions, indicating broad ecological connectivity despite moderate abundance levels. On the other hand, several low to moderate abundance phyla such as Rokubacteria, GAL15, Dependentiae, and Dadabacteria showed notable negative correlations, indicating contrasting ecological strategies or sensitivity to habitat specific constraints. The presence of negative interactions highlights competitive exclusion or divergent nutrient utilization patterns among these phyla, particularly as soil conditions shift from paddy to transition and tea soils.

Influence of soil properties on bacterial community structure

Redundancy analysis (RDA) highlighted the interconnectedness of soil physicochemical parameters and microbial functional dynamics. A strong influence of organic carbon on bacterial community composition suggests that bacterial communities are not only indicators of soil conditions but also integral components in maintaining essential soil processes.Actinobacteria stood out in areas with high organic carbon, reaffirming their well-known role in breaking down organic matter and cycling carbon (Mhete et al. 2020). This indicates that shifts in organic substrate availability directly select for copiotrophic decomposer communities, enhancing soil respiration and aggregation processes. In contrast, Acidobacteria showed up more often in soils with higher electrical conductivity, hinting at their adaptability to more ion-rich or slightly saline conditions (Kim et al. 2016). Patescibacteria showed a strong association with arylsulfatase activity, consistent with their reported role in sulfur and nitrogen cycling (Hermann et al. 2019). Interestingly, their negative correlation with OC suggests a distinct ecological strategy, possibly relying on syntrophic relationships with other microbial taxa rather than direct utilization of carbon substrates (Shen et al. 2021).Meanwhile, Firmicutes and Proteobacteria were frequently found in soils with higher β-D-glucosidase activity, reinforcing their contribution to carbon mineralization and organic matter turnover (Su et al.2021).Thus, microbial community shifts are not passive reflections of soil chemistry but active regulators of ecosystem processes, decomposition rates influencing nutrient dynamics and plant productivity which ultimately impact plant health and productivity.

Conclusion

This study demonstrates that topographic variation in tea-growing landscapes exerts a strong influence on soil physicochemical and biological properties, which in turn shape bacterial community composition, diversity, and function. Upland soils benefit from enhanced stability and microbial turnover but may require external inputs for optimal plant growth. Lowland soils, while microbially active, face structural degradation and nutrient imbalances over time. Distinct microbial communities were associated with the upland, transition, and lowland zones, each reflecting specific soil conditions such as pH, organic carbon, moisture, and enzymatic activity. These microbial communities are not only indicators of edaphic states but also active agents in nutrient cycling, soil structure formation, and plant–soil interactions. Higher microbial diversity and functional potential in the transition zone suggest that moderate and balanced soil conditions, coupled with minimal disturbance, may foster a resilient and ecologically versatile soil microbiome. In contrast, the upland and lowland zones, constrained by acidification and waterlogging respectively, harbour more specialized microbial taxa with targeted functional roles. Overall, these findings highlight the need for topography-informed soil management practices that enhance microbial diversity and function to maintain soil health and support sustainable crop productivity in heterogeneous agroecosystems. Future studies should focus on understanding the specific functional contributions of poorly characterized taxa and their interactions with plants and soil structure under varying land-use intensities.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

M.Barooah: Conceptualization, Supervision, Writing – review and editing. M. Baruah: Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft. U. Sarmah: Data curation, Methodology, Investigation. S.S.Bora: Writing – review and editing, Formal analysis, Methodology, Software. KNS: Formal analysis, Visualization, Methodology. M. Gogoi: Resources, Data curation.

Funding

Not applicable.

Data availability

The metagenomes are publicly available in the GenBank database (PRJNA1169761).

Declarations

Ethics approval

This article does not contain any studies with human or animal subjects.

Consent for publication

All the authors have contributed in this work; read and approved the manuscript and that all co-authors are aware of its submission to including the concerned authorities.

Competing interests

The authors declare no competing interests.