ITS2 amplicon profiling reveals site-specific mycobiomes in Avicennia marina rhizospheres along the red sea coast

Abstract

Despite their ecological significance, the rhizosphere mycobiomes of Egyptian Red Sea Avicennia marina remain largely unexplored. This study presents the first ITS2 amplicon-based characterization of these mangrove rhizosphere fungi at two sites: Al-Kilo 17 south of Safaga (K17) and Ras Mohamed National Park (RM). Alpha diversity showed higher richness in K17, while PCoA showed clear beta diversity differences. Taxonomic profiling indicated significant site-specific communities; Linear Discriminant Analysis Effect Size identified significant biomarkers across all hierarchical levels Ascomycota dominated K17 samples (59–79%), whereas Basidiomycota prevailed in RM (59–93%). At class level, Dothideomycetes prevailed in K17 and Agaricomycetes in RM. Aspergillus (up to 27%) characterized K17, while Malassezia unexpectedly dominated RM (48–62%). Other genera (e.g., Hortaea, Penicillium, Hypocrea) were less abundant. Soil analysis highlighted sodium gradients as a potential environmental driver, showing distinct positive correlations (r = 0.73–0.86) with alpha diversity and sequencing depth. Functional annotation using FungalTraits inferred putative ecological roles and predicted stress-associated traits based on taxonomic assignments rather than direct functional measurements. Notably, human skin-associated fungi were abundant at RM (> 40%), a tourism-impacted area. This study suggests unique fungal diversity often overlooked by culture-based approaches, highlighting Red Sea mangroves as promising reservoirs of ecological and biotechnological potential.

Introduction

Mangroves are unique ecosystems composed of trees, shrubs, ferns, and palms that grow in wetlands at the interface between land and sea, primarily in tropical and subtropical regions. These ecosystems originated approximately 114 million years ago and are currently found in 112 countries, covering about 181,000 km² globally, which accounts for nearly 25% of tropical coastlines^1–3. They are well adapted to harsh environmental conditions, such as high salinity, extreme tidal fluctuations, strong winds, elevated temperatures, and anaerobic, muddy soils^3,4. In Egypt, mangrove forests occupy approximately 700 hectares along the Sinai Peninsula and the Red Sea coastlines; these ecosystems are primarily characterized by two species: Rhizophora mucronata and Avicennia marina^4,5. While R. mucronata is largely restricted to the southern reaches of the coast (e.g., the Halaib Triangle), A. marina is the most widely distributed and dominant species, frequently forming dense, mono-specific stands or “mono-patches” in the northern and central Red Sea regions^4,5.

Mangroves are ecologically significant and provide a range of essential ecosystem services, ranking among the most biologically productive and structurally complex ecosystems on the Planet. They play a vital role in coastal protection, biogeochemical processes, and the sequestration of blue carbon. Additionally, they supply valuable goods such as fuel, timber, food, and construction materials to both local and global communities. Their ability to thrive in extreme environmental conditions has led to the discovery of various bioactive compounds with antifungal, antiviral, anticancer, and insecticidal properties^6,7.

These forests are sheltered by fringing coral reefs, creating a stable environment for a diverse array of flora and fauna⁷ These ecosystems serve as vital nursery grounds that bolster coastal productivity by hosting a rich biodiversity of marine life, reptiles, and amphibians^5,7. They provide critical nesting sites and conservation support for numerous migratory bird species^5,7. Within these ecosystems, microbial communities, particularly bacteria and fungi, represent more than 90% of the total microbiome biomass. Ecosystem services provided by mangroves are largely shaped by their associated microbial communities, particularly bacteria and fungi^8–10. While mangrove bacterial communities are well documented and have been shown to play diverse roles in nutrient cycling and stress mitigation^11–13, the fungal communities (mycobiomes) associated with mangroves remain relatively understudied^14,15.

Despite this gap in knowledge, mangrove-associated fungi play essential roles in biogeochemical cycling and plant health. Through symbiotic interactions with mangrove plants, they enhance resistance to pathogens and increase tolerance to environmental stressors^16–18. Acting as major decomposers of organic matter in mangrove sediments, fungi contribute to nutrient recycling, providing a food source for benthic fauna in mangroves and adjacent marine ecosystems. In addition to their ecological importance, mangrove fungi produce a range of bioactive compounds with pharmaceutical potential, including antibiotics, antivirals, and anticancer agents. They also synthesize enzymes used in various biotechnological applications, such as bioremediation, biodiesel production, and biopesticide development¹⁹.

However, these vital fungal communities face multiple threats from human activities. The degradation of mangroves, along with the discharge of pollutants such as industrial wastewater, sewage, heavy metals, and hydrocarbons, introduces excess nutrients that disrupt the ecological balance between mangroves and their microbial communities, including fungi. Tourism-related anthropogenic disturbances, including trampling, soil compaction, and litter deposition, can also alter the composition of soil microbiomes^20–22. Additionally, climate change factors such as rising sea levels, increased salinity, and changed rainfall patterns exert pressures that threaten the mangrove ecosystems and their associated fungi^23,24.

Therefore, assessing the diversity and assembly patterns of mycobiomes is crucial for setting ecological baselines and guiding sustainable management strategies to preserve the mangrove ecosystem. Although increasing global research on the mycobiome in mangroves in recent years^22,25,26, the rhizosphere mycobiomes of these dominant A. marina stands in Egypt remain largely unexplored. Furthermore, current studies have been limited to culture-based methods, which capture only limited diversity. Hence, there is a need to shift to advanced molecular techniques, employing high-throughput metagenomic approaches, to fully unravel these communities^27,28.

The present study was designed to investigate the diversity of fungal communities associated with the rhizosphere of Avicennia marina along the Egyptian Red Sea coast using high-throughput ITS amplicon sequencing (Illumina MiSeq). Given that both environmental factors and human activities could influence fungal diversity and community assembly, the study aimed specifically to: (i) assess overall fungal diversity across two distinct mono-specific stands of A. marina in Egypt, Al-Kilo 17, south of Safaga (a potentially polluted site), and Ras Mohamed National Park in South Sinai (impacted by tourism); (ii) evaluate how physicochemical soil properties shape mycobiome assembly; and (iii) functionally annotate fungal communities based on taxonomic profiles derived from ITS2 sequencing, enabling the linkage of community composition to inferred ecological strategies. To the best of our knowledge, this is the first study to employ ITS amplicon sequencing to characterize mangrove-associated fungal communities in Egypt.

Materials and methods

Study sites and sample collection

The present work focused on studying Avicennia marina, the dominant mangrove species along Egypt’s Red Sea coast. It involved two locations: the first is a mangrove stand at Al-Kilo 17 south of Safaga (26°61’63.8"N, 34°01’1389"E), and the second is a protected site within Ras Mohamed National Park in South Sinai (27°72’22.2"N, 34°25’38.9"E) (Fig. 1). Ras Mohamed Park faces threats from tourism-related activities such as trampling and littering, which could adversely affect the mangrove’s fungal community. In contrast, the Safaga mangrove stand is impacted by industrial and domestic pollution from the nearby Safaga port, posing potential risks of heavy metal and sewage contamination^29–31. The selection of these sites enables an investigation into how both environmental and human-induced factors influence mangrove mycobiomes.

Fig. 1.

Map showing the locations of selected study sites along the Red Sea coastline in Egypt, where mangrove rhizosphere samples were collected. (RM) Ras Mohamed Mangrove Reserve in South Sinai (27°72’22.2"N, 34°25’38.9"E) and (K17) Safaga Mangrove stand, located 17 km south of Safaga City (26°61’63.8"N, 34°01’1389"E). The map was constructed by the authors using Google Earth Pro desktop software (Version 7.3.7.1155; Google LLC, 2026, https://www.google.com/earth/about/versions/; accessed April 2026).

The mangrove forests at both sampling locations. were monospecific and consisted exclusively of A. marina^4,5. To ensure rigorous taxonomic traceability, the host vegetation at both sampling sites was identified and verified as A. marina (Forssk.) Vierh. by Prof. Dr. Usama K. Abdel-Hameed (Botany Department, Faculty of Science, Ain Shams University, Cairo, Egypt). To profile the associated fungal communities, sediment samples were selectively collected from the rhizosphere of mature, wild mangrove trees. Because this study focused exclusively on these rhizosphere soils, no plant tissues were harvested or used for molecular analyses. Consequently, whole-plant voucher specimens were not collected or deposited in a public herbarium.

Field sampling and the collection of rhizosphere sediment samples were conducted in May 2022 by authors N.A.I. And S. A. D., with the assistance of Heba Sayed Salem, a teaching assistant at the Department of Microbiology at the same institution. A total of four rhizosphere soil samples were collected from both sites, with two biological replicates at each location: K17.1 and K17.2 from Al-Kilo-17, and RM.1 and RM.2 from Ras Mohamed. Rhizosphere samples were collected aseptically from a depth of 10–30 cm surrounding plant roots. All samples were transported to the laboratory in sterile containers and stored at -20 °C until further processing.

Environmental Compliance and Research Permissions

All fieldwork and sampling procedures were conducted exclusively for academic research purposes in full compliance with relevant institutional, national, and international regulations. Fieldwork activities were authorized and conducted under the official institutional framework of the Faculty of Science, Ain Shams University, which submitted formal authenticated requests to the governing environmental bodies, including the Egyptian Environmental Affairs Agency (EEAA) and the Nature Conservation Sector (NCS). All sampling protocols were non-destructive and performed in accordance with national environmental protection regulations, regional conservation frameworks, the IUCN Policy Statement on Research Involving Species at Risk of Extinction, and the regulatory principles of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES).

DNA extraction, PCR amplification, and Illumina MiSeq sequencing

Metagenomic DNA was extracted from mangrove rhizosphere soil using the QIAGEN DNeasy PowerSoil Pro Kit (Qiagen, Germany). Briefly, 0.25 g of soil was subjected to mechanical lysis via bead-beating in PowerBead tubes to disrupt microbial cells, followed by inhibitor removal using the kit’s proprietary solution to eliminate humic acids and other contaminants. DNA was purified via silica membrane spin columns, washed with ethanol-based buffers, and eluted in nuclease-free water. DNA quality was verified via spectrophotometry (NanoDrop). Thereafter, PCR amplification of the fungal ITS2 region was performed using the universal primer pair ITS3 (5′-GCATCGATGAAGAACGCAGC-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′)^32,33. The thermal cycling conditions included an initial denaturation at 95 °C for 3 min, followed by 25 cycles of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 30 s, with a final extension at 72 °C for 5 min³⁴. Afterwards, sequencing of the ITS2 amplicons was conducted using paired-end sequencing technology (2 × 300 bp) on an Illumina MiSeq platform. Library preparation employed the Herculase II Fusion DNA Polymerase and Nextera XT Index V2 Kit, following the Illumina 16 S Metagenomic Sequencing Library Preparation protocol (Part #15044223 Rev. B). PCR amplification and Illumina sequencing procedures were performed by Macrogen Inc. (Seoul, Korea).

Sequence processing, diversity analysis, and functional annotation

Raw paired-end reads generated from Illumina sequencing were processed and analyzed using the web interface of the SHAMAN server (https://shaman.pasteur.fr/). The SHAMAN server has been introduced as a user-friendly platform for metataxonomic analysis, enabling processing from raw reads to statistical analysis without requiring complex computational resources. Reads normalization was performed using the server’s default settings. The analysis of raw reads was performed according to the workflow described by Volant et al.³⁵. Initial quality control was carried out using AlienTrimmer (v0.4.0) for trimming contaminant sequences and clipping, followed by merging of paired-end reads using PEAR (v0.9.10.1). Subsequently, OTU picking, taxonomic annotation, and quantification were performed using Vsearch (v2.3.4.0). The OTU picking process followed the UPARSE pipeline approach, including five steps: dereplication, singleton removal, chimera detection, clustering, and alignment³⁵.

Taxonomic assignment of OTUs was performed using the Findley database integrated within the SHAMAN pipeline as the reference database for fungal ITS sequences. The Findley database was selected to support a more focused analysis of fungal communities and to minimize the inclusion of non-target eukaryotic sequences. Furthermore, the Findley database, along with other reference databases within the SHAMAN server, is updated bi-monthly via biomaj.pasteur.fr, ensuring that taxonomic assignments are based on up-to-date reference sequences³⁵. Taxonomic annotations were filtered according to sequence identity thresholds as follows: ≥75% for phylum, ≥ 78.5% for class, ≥ 82% for order, ≥ 86.5% for family, ≥ 94.5% for genus, and ≥ 98% for species. Alpha diversity indices, including richness and rarefaction curves, and beta diversity analyses (principal coordinate analysis, PCoA) were conducted within the SHAMAN framework. Statistical analysis of differential abundance was performed using the DESeq2 R package implemented within SHAMAN server, which models OTU count data using a negative binomial distribution and provides robust estimates even for datasets with limited sample size³⁵.

A Venn diagram was constructed at the genus level to show the unique and shared fungal taxa among the samples using the InteractiVenn online tool (http://www.interactivenn.net/). Functional annotation based on ITS2 taxonomic data at the genus level was conducted using the FungalTraits database³⁶. Functional annotation profiles were visualized as a heatmap using Python 3.11, Pandas v1.5.3, Seaborn v0.12.2, and Matplotlib v3.7.1. To identify taxa differing significantly between sampling stations, Linear Discriminant Analysis Effect Size (LEfSe) was performed on normalized relative abundance data at the phylum, class, and genus levels. Significant differences were first detected using the non-parametric Kruskal–Wallis test (P < 0.05). Subsequently, Linear Discriminant Analysis (LDA) estimated the effect size of these taxa, with an LDA score threshold of > 2.0 used to identify discriminative biomarkers. Statistical analyses and data normalization were conducted using the R software environment (v4.3.1; R Core Team, 2023). This approach integrates statistical significance with biological consistency to highlight features explaining inter-station differences.

Examining soil physicochemical properties and their influence on mycobiome diversity

Soil physicochemical properties of the four collected rhizosphere samples (K17.1, K17.2, RM.1, RM.2) were analyzed. The parameters measured included pH, electrical conductivity (EC), salinity, major cations (Ca²⁺, Mg²⁺, Na⁺, K⁺), anions (HCO₃⁻, Cl⁻, SO₄²⁻), macronutrients (N, P, K), trace metals (Cu, Fe, Mn, Zn), organic matter, and calcium carbonate (CaCO₃). The analyses followed standard procedures based on methodologies outlined by Nelson & Sommers³⁷, Dane and Topp³⁸, and Sparks et al.³⁹. For each sample, three replicates were analyzed per parameter to ensure the reliability of the data. Results were presented as the mean ± standard deviation of the replicates. The significant differences in the measured soil parameters across sampling sites were analyzed using the non-parametric Kruskal-Wallis H test, and Fisher’s least significant difference (LSD) post hoc test, based on one-way ANOVA residual mean square error for calculating the LSD threshold.

Pearson correlation analysis was performed to assess the correlation between soil physicochemical properties and fungal diversity metrics (sequencing reads, alpha diversity, unique OTUs). Due to the limited sample size, emphasis was placed on both effect sizes (r) and P-values, following recommendations that strong correlations can be biologically meaningful even without formal statistical significance⁴⁰. To evaluate the relationship between fungal assemblages and soil properties, we performed pairwise Spearman rank correlations (ρ) between fungal class abundances and physicochemical parameters⁴¹. Exact permutation tests were applied to generate robust non-parametric P-values, to address the small sample size⁴², and the Holm correction was subsequently used to adjust significance levels⁴³. A heatmap was constructed in Python 3.11 to demonstrate correlation patterns, with significance stars annotated on the tiles. Significant correlations were defined as P < 0.05. All statistical analyses were performed using R v3.6.3⁴⁴.

Results

Characterization of rhizosphere-associated fungal communities

In this study, we investigated the fungal community structure of four mangrove rhizosphere soil samples using Illumina MiSeq sequencing of the ITS2 region of rDNA. The sequencing results showed clear variation in read counts, GC content, and total bases among the four samples (Table 1). Samples K17.1 and K17.2 produced the highest numbers of reads (109,771 and 116,285, respectively) and total bases (66.1 M and 70.0 M), while RM.1 and RM.2 yielded fewer reads (80,966 and 62,830) and lower total bases (48.7 M and 37.8 M). The GC content also differed between samples, ranging from 45.9% in RM-2 to 64.8% in K17-2. Overall, the K17 samples showed higher read counts, GC content, and total bases than the RM samples, while these data suggest potential site-specific variation in fungal community composition, these findings remain exploratory due to the small sample size.

Table 1.

Number of reads, GC content, and number of bases identified in the examined mycobiomes.

Sample name	Reads	GC	Bases
K17.1	109,771	63.8%	66.1 M
K17.2	116,285	64.8%	70.0 M
RM.1	80,966	57%	48.7 M
RM.2	62,830	45.9	37.8 M

The rarefaction curves of the four mycobiomes show a clear plateau, indicating that the sequencing depth was generally sufficient to capture most of the fungal diversity present in the samples (Fig. 2). Notably, even samples with comparatively lower read counts exhibit well-defined plateaus samples (Table 1and Fig. 2), suggesting that their lower richness reflects inherent differences in community composition rather than under sampling. Subsequent diversity analyses supported these findings. The K17 samples displayed higher alpha diversity values compared to the RM samples (Fig. 3); however, these differences are descriptive as the limited number of replicates (n = 4) restricts formal inferential generalization. The PCoA plot (Fig. 4) further demonstrated a separation between the K17 and RM sample groups, suggesting distinct microbial community compositions at the two sampling sites within this specific dataset.

Fig. 2.

Rarefaction curves for the mycobiomes identified in this study.

Fig. 3.

Richness (Alpha) diversity analysis of the four mangrove-mycobiomes identified in this study.

Fig. 4.

PCoA analysis of the four mangrove-mycobiomes identified in this study.

Furthermore, the Venn diagram (Fig. 5) showed a single OTU shared among all samples, while 6, 2, 2, and 3 OTUs were unique to K17.1, K17.2, RM.1, and RM.2, respectively, further suggesting the potential site-specific nature of the fungal communities. However, this extremely low overlap should be interpreted with caution, as Venn-based OTU comparisons are highly sensitive to sequencing depth variation, OTU clustering criteria, and filtering thresholds. In addition, the limited number of biological replicates (n = 2 per site) restricts the ecological interpretability of such overlap patterns. Specifically, at the Safaga (K17) station, samples K17.1 and K17.2 hosted 6 and 2 unique OTUs, respectively. These unique taxa were primarily associated with the phylum Ascomycota, including salt-tolerant genera such as Stemphylium and specific unclassified Capnodiales.By contrast, the Ras Mohamed (RM) station exhibited a different set of site-specific taxa, with 2 and 3 unique OTUs identified in RM.1 and RM.2. These were largely representative of the Basidiomycota phylum, specifically the yeast-like genus Malassezia. The minimal overlap between the K17 and RM stations (as shown by the lack of shared OTUs between site, specific groups) suggests the impact of localized environmental filters, particularly the high sodium gradients, in shaping unique fungal assemblages at each location.

Fig. 5.

Venn diagram showing shared and unique fungal taxa among the four rhizosphere samples: K17.1, K17.2, RM.1, and RM.2.

Taxonomic profiling of rhizosphere-associated fungal communities

The taxonomic assignment of raw sequences from the four mangrove rhizosphere-associated fungal community samples was conducted using the SHAMAN pipeline. The results a suggested a distinct taxonomic divergence between the two study sites across all taxonomic levels (Fig. 6). At the phylum level, Ascomycota was the dominant group in K17 samples, accounting for 59.4% in K17.1 and 79.9% in K17.2, whereas Basidiomycota prevailed in the RM site, representing 93.1% in RM.1 and 59.3% in RM.2 (Fig. 6A). Sequences that could not be classified were grouped as unassigned, comprising 1–7% of the total community. Analysis of class-level distribution within these phyla showed that Ascomycota in the K17 station was primarily represented by Dothideomycetes (40.7% in K17.1; 42.1% in K17.2) and Eurotiomycetes (16.0% in K17.1; 35.1% in K17.2), while Sordariomycetes was also a major contributor (15.7% in K17.1; 6.9% in K17.2). Notably, Basidiomycota class incertae sedis was a significant component only in the K17.1 sample (32%), further highlighting the lack of taxonomic overlap between the regions. In contrast, the Basidiomycota-dominated RM site was characterized almost by Agaricomycetes, which reached abundances of 89.1% in RM.1 and 49.5% in RM.2 (Fig. 6B).

Fig. 6.

Fungal community composition associated with the four rhizosphere samples at the phylum level (A), class level (B), and genus level (C).

At the genus level, the composition was highly site-specific. In the K17 station, the dominant Ascomycota genera included Aspergillus (14.6% in K17.1; 26.5% in K17.2) and Cladosporium (0.6% in K17.1; 22.7% in K17.2), followed by Stemphylium, Penicillium and Stachybotrys. At the RM site, the mycobiome was dominated by the yeast-like genus Malassezia, which accounted for 48.1% in RM.1 and 61.6% in RM.2, followed by Penicillium and Sporobolomyces (13.1 and 9.3% in RM.1) alongside an unclassified Saccharomycetales genus (11.6% in RM.2) (Fig. 6C). Overall, these results demonstrate a fundamental shift in fungal community composition between the Safaga and Ras Mohamed mangrove ecosystems. The transition from Ascomycota-dominated assemblages in K17 to Basidiomycota -dominated assemblages in RM reflects major site-specific adaptations, likely driven by the distinct environmental filters and ionic gradients present at each location.

LEfSe analysis of fungal community composition

To identify fungal biomarkers associated with each sampling site, Linear discriminant analysis effect size (LEfSe) was performed. The analysis identified distinct fungal biomarkers associated with each sampling station at the phylum, class, and genus levels (Fig. 7). At the phylum level, Ascomycota was more strongly associated with station K17, whereas Basidiomycota showed greater enrichment at station RM (Fig. 7A), indicating that the dominant fungal phyla differed between the two locations. Within the phylum Ascomycota, the classes Dothideomycetes, Eurotiomycetes, and Sordariomycetes were significantly enriched at K17 (Fig. 7B). These classes were represented by several genera, particularly Lewia, Aspergillus, and other unidentified ascomycetous taxa, which exhibited higher LDA scores in K17 samples than in RM samples (Fig. 7C). This suggests that K17 supported a more diverse assemblage of ascomycetous fungi.

Fig. 7.

Differential fungal biomarkers between sampling stations identified by LEfSe analysis.
(A) Phylum-level biomarkers, (B) class-level biomarkers, and (C) genus-level biomarkers distinguishing the fungal communities between stations K17 and RM. Horizontal bars represent logarithmic Linear Discriminant Analysis (LDA) scores, where higher values indicate greater contribution of the corresponding taxon to differences between sampling stations. Taxa with positive LDA scores were considered the most discriminative fungal biomarkers between the two sites.

In contrast, the phylum Basidiomycota was predominantly associated with RM (Fig. 7A). At the genus level, Malassezia was the principal basidiomycetous biomarker detected at this station (Fig. 7C), indicating localized enrichment of basidiomycetous fungi in RM samples. Unlike Ascomycota, fewer discriminative basidiomycetous taxa were observed at the class level (Fig. 7B), suggesting that the main differences within this phylum were better resolved at the genus level.

Soil physicochemistry–mycobiome relationships across mangrove sites

Soil physicochemical analyses (Table 2) revealed significant differences between rhizosphere samples from Al Kilo 17 (K17.1, K17.2) and Ras Mohamed mangrove sites (RM.1, RM.2). Baseline parameters such as pH, electrical conductivity, Mg²⁺, HCO₃⁻, Cl⁻, and CaCO₃ remained stable across all sites, indicating consistent soil buffering capacity. However, salinity, Na⁺, nutrient concentrations, and trace metals showed notable variability. Notably, sodium levels were significantly higher (P ≤ 0.001) at K17.2 (381.1 mg/kg) compared to the RM sites (108–144 mg/kg), indicating substantial ionic stress at the K17 site despite the higher overall salinity recorded at RM (44–46%). These results suggest that ecological differences between sites may be driven by specific sodium gradients rather than total salinity. Despite variations in trace metals such as Cu, Zn, and Fe across sites, they remained low and well below contamination thresholds. Nutrients (nitrogen approximately 54–95 mg/kg; phosphorus 7–10 mg/kg) and organic matter (1.5–3.5%) were within expected ranges for uncontaminated mangrove soils, indicating no evidence of sewage contamination.

Table 2.

Soil physicochemical analyses across sampling sites.

Physicochemical Parameter	Sampling site				Sig.
	K17.1	K17.2	RM.1	RM.2
pH	7.56 ± 0.95ᵃ	7.66 ± 0.56ᵃ	7.81 ± 0.07ᵃ	7.86 ± 0.17ᵃ	ns
Electrical conductivity (S/m)	3.60 ± 0.16ᵃ	3.39 ± 0.01ᵃ	2.93 ± 0.13ᵃ	2.72 ± 0.17ᵃ	ns
Salinity (%)	37.0 ± 04ᵇ	40.0 ± 01ᵇ	46.0 ± 04ᵃ	44.0 ± 07ᵃ	**
Ca²⁺	120.5 ± 19ᵃ	110.5 ± 13ᵇ	84.5 ± 05ᶜ	90.5 ± 01ᶜ	**
Mg²⁺	80.5 ± 01ᵃ	87.5 ± 03ᵃ	60.5 ± 09ᵃ	71.5 ± 06ᵃ	ns
Na⁺	155.0 ± 16ᵇ	381.1 ± 11ᵃ	144.0 ± 12ᵇ	108.0 ± 09ᶜ	***
K⁺	3.75 ± 0.12ᵃ	2.85 ± 0.66ᵃ	3.95 ± 0.14ᵃ	3.95 ± 0.11ᵃ	ns
HCO₃⁻	8.50 ± 0.03ᵃ	8.50 ± 0.15ᵃ	7.50 ± 0.52ᵃ	8.50 ± 0.09ᵃ	ns
Cl⁻	195.5 ± 07ᵃ	189.5 ± 16ᵃ	159.5 ± 17ᵃ	162.5 ± 19ᵃ	ns
SO₄²⁻	155.9 ± 19ᵃ	140.9 ± 11ᵇ	125.9 ± 11ᶜ	100.9 ± 18ᶜ	**
Nitrogen (mg/kg)	86.0 ± 06ᵇ	90.0 ± 07ᵇ	95.0 ± 17ᵃ	54.0 ± 05ᶜ	***
Potassium (mg/kg)	152.0 ± 11ᵃ	146.0 ± 13ᵃ	126.0 ± 09ᵇ	95.0 ± 07ᶜ	**
Phosphorus (mg/kg)	7.3 ± 0.18ᶜ	7.3 ± 0.27ᶜ	9.7 ± 0.09ᵃ	9.0 ± 0.01ᵇ	**
Copper (mg/kg)	0.67 ± 0.35ᵇ	1.04 ± 0.56ᵃ	0.40 ± 0.01ᶜ	0.17 ± 0.03ᶜ	**
Iron (mg/kg)	38.3 ± 09ᵃ	32.4 ± 07ᵇ	32.1 ± 05ᵇ	10.5 ± 09 ᶜ	***
Manganese (mg/kg)	2.38 ± 0.24ᵃ	0.42 ± 0.15ᶜ	0.52 ± 0.01ᶜ	0.40 ± 0.11ᶜ	**
Zinc (mg/kg)	1.08 ± 0.03ᵇ	0.70 ± 0.37ᶜ	1.19 ± 0.09ᵃ	0.75 ± 0.11ᶜ	*
Organic matter (%)	2.32 ± 0.12ᵇ	3.48 ± 0.15ᵃ	1.53 ± 0.33ᶜ	1.65 ± 0.01ᶜ	***
CaCO₃ (%)	1.26 ± 0.14ᵃ	0.88 ± 0.15ᵃ	0.84 ± 0.38ᵃ	0.84 ± 0.33ᵃ	ns

Values represent mean ± standard deviation. Superscripts (ᵃ, ᵇ, ᶜ) indicate statistically different groups within each row at P ≤ 0.05 according to Fisher’s LSD post-hoc test. (Significance codes: ‘***’ P ≤ 0.001; ‘**’ P ≤ 0.01; ‘*’ P ≤ 0.05; ns = not significant).

An Exploratory Pearson correlation analysis (Table 3) of soil parameters and fungal diversity metrics across the rhizosphere sites revealed complex ecological drivers of mycobiome structure. This exploratory correlation suggested potential positive associations between Na⁺ levels and fungal diversity, with high correlation coefficients (r) ranging from 0.73 to 0.86; yet these associations did not achieve statistical significance (P > 0.05) due to the small sample size, which limits statistical inference.

Table 3.

Pearson correlation coefficients (r) between soil physicochemical parameters and fungal diversity metrics across the four sampling sites (n = 4)⸶.

Soil Parameter	Diversity Metric	Pearson’s r	P -value
Salinity (%)	Reads	0.782	0.218
	Alpha Diversity	0.862	0.138
	Unique OTUs	0.026	0.974
Na⁺ (mg/kg)	Reads	0.727	0.273
	Alpha Diversity	0.745	0.255
	Unique OTUs	-0.382	0.618
N (mg/kg)	Reads	0.964	0.036*
	Alpha Diversity	0.933	0.067
	Unique OTUs	0.479	0.521
P (mg/kg)	Reads	-0.483	0.517
	Alpha Diversity	-0.565	0.435
	Unique OTUs	0.407	0.593
Organic matter (%)	Reads	0.927	0.073
	Alpha Diversity	0.943	0.057
	Unique OTUs	0.608	0.392
Cu (mg/kg)	Reads	0.749	0.251
	Alpha Diversity	0.815	0.185
	Unique OTUs	-0.151	0.849
Zn (mg/kg)	Reads	0.959	0.041*
	Alpha Diversity	0.966	0.034*
	Unique OTUs	0.060	0.940
Mn (mg/kg)	Reads	0.832	0.168
	Alpha Diversity	0.898	0.102
	Unique OTUs	0.033	0.967
Fe (mg/kg)	Reads	0.976	0.024*
	Alpha Diversity	0.960	0.040*
	Unique OTUs	0.068	0.932

⸶: Data points represent site means (3 replicates per site). Significant correlations are marked with stars (Significance codes: ‘***’ P ≤ 0.001; ‘**’ P ≤ 0.01; ‘*’ P ≤ 0.05).

This pattern was notably observed at K17.2, where elevated Na⁺ levels were associated with enhanced alpha diversity. Scatterplots (Fig. 8) suggest Na⁺ as a potential driver of fungal richness, reflecting possible mycobiome complexity under high ionic stress. Furthermore, nitrogen, zinc, and iron showed associations with diversity metrics (P > 0.05); however, given the sample size, these should be interpreted as potential secondary factors that may enhance mycobiome richness rather than definitive ecological rules.

Fig. 8.

Salinity and Sodium as Key Drivers of Mangrove Mycobiomes: Scatterplots showing relationships between soil salinity and sodium with fungal alpha diversity (A–B) and sequencing reads (C–D) across the four rhizosphere samples. Regression lines (red) with 95% confidence intervals (shaded area) illustrate Pearson’s correlations, with coefficients (r) and p-values shown in each panel.

As shown in Fig. 9, the heatmap demonstrates Spearman correlation coefficients (ρ) between fungal class abundances and physicochemical parameters. The results suggested that several fungal classes exhibit potential associations with ion concentrations. Eurotiomycetes, for instance, had a positive association with Na⁺; conversely, Sordariomycetes displayed a negative correlation with Na⁺ levels. Trace metals (Cu, Fe, Zn) demonstrated weak or nonsignificant correlations, supporting the inference that heavy metal contamination is absent in these soils. Similarly, the lack of significant associations with nitrogen and phosphorus suggests no evidence of sewage-derived pollution. Overall, these exploratory data suggest that fungal taxonomic assembly patterns in these specific samples are primarily influenced by ionic gradients, particularly sodium.

Fig. 9.

Spearman correlations between fungal class abundances (%) and soil physicochemical parameters in the studied mangrove rhizospheres. In the heatmap, color intensity indicates correlation strength (red = positive, blue = negative). Bright grey stars denote significant associations (*P < 0.05). Saccharomycetes were excluded due to negligible abundance.

Functional annotation based on ITS2 taxonomic data using the FungalTraits database

The heatmap of functional annotations (Fig. 10) revealed distinct variations in the distribution of fungal traits among the analyzed samples. Endophytic fungi were present in K17.1 and K17.2 but absent from RM.1 and RM.2. The extremophilic trait appeared in RM.1, RM.2, and to a lesser extent in K17.1, while it was not detected in K17.2. Plant pathogenic fungi were restricted to RM.1, whereas the plant-associated trait was found exclusively in K17.1. Skin-associated fungi were abundant in RM.1 and RM.2, but less represented in K17.1 and K17.2. Soil fungi dominated RM.1, appeared at lower levels in K17.1 and K17.2, and were absent from RM.2. The ubiquitous trait was strongly expressed in K17.1 and K17.2, moderately present in RM.2, and entirely lacking in RM.1. Wood-decaying fungi were detected at low abundance only in K17.1. Finally, sequences classified as “unknown”, representing those without functional annotation, were more frequent in K17.1 and K17.2 compared with the other samples. Collectively, the observed heatmap patterns suggest potential ecological specialization across the studied sites. However, given the study’s scope, these functional profiles remain preliminary and require validation through broader spatial sampling.

Fig. 10.

Functional annotation of fungal communities based on ITS2 taxonomic data using the FungalTraits database. The heatmap demonstrates the relative abundance (%) of fungal taxa assigned to ecological guilds across the four rhizosphere samples. Color intensity corresponds to the abundance scale (0–100%), with darker shades indicating higher representation within each habitat-associated trait.

Discussion

Mangrove ecosystems support a diverse range of unique fungi that are crucial for biogeochemical processes and the health of surrounding trees. Rhizosphere fungi, in particular, play critical roles in nutrient cycling, plant growth promotion, and stress tolerance, acting as a vital link between mangrove roots and their environment. Understanding the community assembly patterns of these fungi is essential for preserving the ecosystem^19,24,45. Given the scarcity of research on the Egyptian mangrove rhizosphere mycobiomes, exploring their diversity and assembly patterns is crucial. This knowledge will help define the impacts of environmental changes and human activity on mangrove forests in this understudied region, guiding necessary conservation efforts.

Thus, this study employed high-throughput ITS2 amplicon sequencing to characterize the diversity of mycobiomes associated with Avicennia marina rhizospheres along the Egyptian Red Sea coast, focusing on assessing correlations between assembly patterns of fungal community and soil physicochemical parameters. We also conducted functional annotation based on taxonomic profiles from the sequencing data, linking community composition to ecological strategies. Notably, this study provides a preliminary, exploratory insight into the Egyptian mangrove mycobiome, establishing an initial baseline for this understudied region.

This investigation focused on two distinct A. marina mangrove stands along Egypt’s Red Sea coast: Alkilo-17, located south of Safaga (K17), and Ras Mohamed in the Sinai Peninsula (RM). The soil analysis of rhizosphere samples collected from Safaga (K17.1 and K17.2) and Ras Mohamed (RM.1 and RM.2) revealed a consistent geochemical baseline, characterized by stable pH levels, electrical conductivity, carbonate content, and levels of major ions, which are typical for arid coastal mangrove environments of the Red Sea^46,47. Trace metal concentrations (Cu, Zn, Mn, Fe) measured remained well below contamination thresholds^48,49 confirming the absence of heavy metal pollution. All sites exhibited low nutrient levels (N: 54–95 mg/kg; P: 7–10 mg/kg) and organic matter content (1.5–3.5%), consistent with oligotrophic mangrove soils^46,50, ruling out anthropogenic nutrient enrichment. These levels are significantly lower than those typically found in systems affected by sewage^51,52.

Despite a consistent geochemical baseline, distinct physicochemical differences were observed between sites. While overall salinity was higher at the RM sites (44–46%), elevated Na⁺ concentrations were significantly higher (P > 0.001) at the K17 sites, particularly K17.2 (381.1 mg/kg). This indicates that the K17 rhizosphere experiences more intense specific ionic stress, which is a recognized selective pressure in mangrove ecosystems^53–55. In contrast, the RM sites, despite their higher total salinity, showed higher nutrient content, likely due to localized anthropogenic influences such as tourism-related litter deposition⁵⁶. These findings confirm the established pattern that mangrove soils maintain a core geochemistry yet exhibit significant site-specific variation due to factors such as ionic stress, salinity, nutrient dynamics, and human activity^{10,55,57–59}.

In the present study, high-throughput ITS2 amplicon sequencing successfully characterized the mangrove-associated mycobiomes. The analyzed sequence data revealed site-specific fungal community profiles across the studied Avicennia marina rhizosphere samples, indicating that environmental factors may influence community assembly. Specifically, Alpha-diversity (fungal richness) was higher at the Safaga (K17) than at the Ras Mohamed (RM) sites. However, the limited sampling scope makes these findings descriptive rather than inferential, restricting broader generalization across the Red Sea coastline. Beta-diversity analysis (PCoA) revealed clear separation of the sample sites. Although the Venn diagram suggested minimal OTU overlap (with only a single shared OTU), this result is likely influenced by methodological factors such as sequencing depth heterogeneity, OTU definition thresholds, and stringent filtering steps. Given these limitations, and the low number of replicates per site, such overlap metrics have limited ecological resolution and should not be overinterpreted as evidence of absolute community segregation. Accordingly, greater emphasis is placed on beta-diversity patterns and taxonomic composition, which provide more robust indications of site-specific differences within this dataset. These patterns align with global findings on mangrove ecosystems, attributing these compositional differences in fungal community structure to geographic and environmental variation^15,60,61.

The role of environmental drivers in community assembly was further examined by assessing the relationship between soil physicochemical parameters and fungal diversity. Exploratory correlations suggested potential positive associations between Na⁺ levels and alpha diversity (r = 0.73–0.86). though these did not reach statistical significance (P > 0.05), due to the restricted number of replicates. Nonetheless, the large effect sizes provide a preliminary indication that sodium-related ionic stress may enhance mycobiome complexity, a pattern consistent with the selection of halotolerant fungal taxa by specific ion concentrations^15,62,63. Additionally, nutrient enrichment correlated with diversity matrices, suggesting a potential dual interaction: while ionic stress appears to be the major environmental determinant, while nutrients and micronutrients function as secondary factors that enhances richness. In agreement with several previous reports^10,59,64, this combined influence helps explain the observed site-specific differences between study sites, despite a consistent geochemical baseline.

In the present study, distinct fungal assembly patterns observed across sites at multiple taxonomic levels indicated ecological divergence and potential functional specialization. At the phylum level, while Ascomycota dominated in K17.1 (59%) and K17.2 (79%), consistent with common mangrove trends^14–16. Basidiomycetes were dominant in RM.1 (93%) and RM.2 (59%), which is in line with a high-throughput sequencing by Haldar and Nazareth¹⁴ on mangrove sediments from Goa, India, which revealed a dominance of basidiomycetes (~ 81%). Notably, Basidiomycota was sporadically reported from marine habitats, including mangroves. Thus, the relatively high incidences observed here (11–39% for K17, and 59–93% for RM, respectively) contrast with previous reports^65–68, likely due to limitations in culture-based studies that underestimated it because of slower growth and fewer diagnostic traits, leading to many taxa being recorded as unidentified or mycelia sterilia. The findings underscore that molecular approaches are crucial for accurately characterizing the true diversity of mangrove mycobiomes.

At the class level, Dothidiomycetes, Eurotiomycetes, and Sordariomycetes were the most prevalent among the Ascomycota, while Agaricomycetes dominated the Basidiomycota. Similar trends were observed in mangrove rhizosphere samples from Saudi Arabia¹⁶ and Kenya¹⁵. Dothidiomycetes and Eurotiomycetes were predominant at the K17 sites. These two classes are common in marine environments, often encompassing stress-tolerant, halophilic, and decomposing fungi capable of metabolizing complex organic matter, supporting marine food webs^60,61,69. In contrast, RM sites were dominated by Agaricomycetes and Sordariomycetes, classes known to include wood-decaying fungi, soil saprotrophs, plant pathogens, salt-tolerant species, and endophytes, which have been previously documented in mangroves and other marine habitats^70–72.

Fungal diversity drivers were evaluated by correlating class-level abundance with soil parameters. Despite a limited sample size, the significance of the correlations was assessed using exact permutation tests and revealed in the heatmap⁴², supporting the reliability of the results.This exploratory analysis suggested that Na⁺ gradients, rather than bulk salinity, may be a more precise determinant of community composition in these samples. These findings were reflected in the assembly patterns of mycobiomes, where halotolerant taxa, particularly members of the Eurotiomycetes, widely recognized for their osmophilic adaptations⁷³, appeared enriched under an elevated Na+ gradient. This finding is consistent with established pattern in coastal ecosystems^15,62,63. It is noteworthy that no significant correlations with trace metals (Cu, Fe, Zn) were detected, underscoring the absence of substantial chemical pollution due to heavy metals.

At the genus level, the contrasting fungal community structures further underscore an ecological divergence across the study sites. The dominance of stress-adapted taxa at K17, particularly K17.2 (showing elevated sodium ions levels), was evident through the high prevalence of Aspergillus (Eurotiomycetes) and unclassified Capnodiales. (the order Capnodiales belonging to Dothideomycetes, includes halotolerant taxa, reflecting an enrichment of organisms specialized in mitigating ionic stress^74–77. This observation aligns with established ecological models linking ionic stress to the selection of specialized microbial taxa^78,79.

In contrast, the mycobiome from Ras Mohamed (RM) sites was notably characterized by a high abundance of Malassezia (49–62%). While often recognized as a human skin commensal, this high abundance may reflect marine-adapted lineages previously reported from ocean environments, coral reefs, and deep-sea sediments^80–82. Although anthropogenic input from tourism cannot be excluded given the high foot traffic at RM^22,56, the presence of Malassezia likely reflects a combination of human-associated microbes and indigenous marine-adapted populations. To our knowledge, this is the first report of Malassezia prevalence in the Egyptian mangrove rhizosphere. Its previous scarcity in culture-based marine surveys is likely attributable to its fastidious nutritional requirements for long-chain fatty acids and slow growth, presenting significant cultivation challenges^28,83. Other frequent taxa included Stemphylium, Aspergillus, Hortaea, and Penicillium, while Cladosporium, Stachybotrys, Hypocrea (a teleomorph of Trichoderma), and Lewia (a teleomorph of Alternaria) occurred at lower abundances, in agreement with earlier Egyptian studies using culture-based methods^84–87. Many of these genera function as root endophytes, enhancing mangrove stress tolerance and pathogen resistance^88–92. Notably, Aspergillus and Penicillium are established phosphate solubilizers and potent producers of various bioactive metabolites^93–95, reflecting considerable bioremediation and biotechnological potential.

Overall, the hierarchical taxonomic divergence between the two sites indicates that environmental filters along the Red Sea coast distinctively shape fungal lineages. At K17, the dominance of phylum Ascomycota (specifically classes Dothideomycetes and Eurotiomycetes) correlates with high sodium levels. Genera such as Aspergillus and Stemphylium are known halotolerants that produce compatible solutes to tolerate ionic stress^74–77. Conversely, the shift to Basidiomycota (class Agaricomycetes) at RM marks a clear ecological transition. The prevalence of Malassezia (up to 62%) is unusual for mangrove mycobiomes; however, its dominance, alongside the halotolerant black yeast Hortaea, suggests a specialized community adapted to the unique sediment chemistry of the Ras Mohamed reserve. This divergence, from phylum to genus levels, confirms that mangrove-associated fungi form distinct taxonomic signatures driven by site-specific environmental factors.

The LEfSe analysis further supported the existence of distinct fungal communities, with significant differences occurring at multiple taxonomic levels. The enrichment of Ascomycota at station K17. suggests that environmental conditions at this site favored Ascomycotous fungi commonly dominant in coastal sediments and involved in organic matter decomposition and nutrient cycling^16,60,61. Within Ascomycota, the predominance of Dothideomycetes, Eurotiomycetes, and Sordariomycetes at K17 further indicates that this station may provide environmental conditions favorable for stress-tolerant saprotrophic fungal groups. The associated genera Lewia and Aspergillus are known for their ecological versatility and ability to colonize nutrient-rich sedimentary habitats. Their higher abundance at K17 may therefore reflect elevated organic matter availability or greater terrestrial input at this station.

By contrast, the enrichment of Basidiomycota at RM and the occurrence of the genus Malassezia suggest a distinct environment at this station. While often recognized as skin commensals, species of Malassezia have increasingly been reported from marine sediments and may be associated with lipid-rich organic matter or other specialized substrates^80–83. Their dominance at RM may therefore reflect a specialized community driven by site-specific sediment chemistry or organic substrate composition relative to K17. This high abundance potentially represents a combination of indigenous, marine-adapted lineages and anthropogenic input resulting from high tourism foot traffic at the RM site^22,56. The stronger discrimination observed at the genus level confirms that local environmental variability influences fungal communities more clearly at lower taxonomic resolution. While phylum-level differences describe broad community shifts, genus-level biomarkers provide more sensitive indicators of ecological variation between sampling stations.

The observed taxonomic patterns were emphasized by FungalTraits functional annotation, which further revealed site-specific enrichment of traits, reflecting inferred ecological strategies. It should be noted that these functional traits are inferred from taxonomic assignments and therefore represent predicted ecological roles rather than direct measurements of functional activity. The detected abundance patterns demonstrate broad consistencies with previously reported distributions but should not be interpreted as direct functional confirmation. For instance, the abundance level of extremophiles (3–16%) is consistent with reports from similar environments⁹⁶. The relatively low abundance of endophytes (4–7%) in K17 samples aligns with baseline levels reported by Põlme et al.³⁶ using the FungalTraits database. Similarly, the plant pathogen trait, detected exclusively in RM.1 sample at 5%, closely matches the 4% prevalence noted by Tedersoo et al.⁹⁷, though their specific database methodology remains unspecified.

Overall, the observed functional profile suggests a putative, stress-adapted mycobiome. Specifically, the moderate abundance of extremophiles may indicate a potential association with saline or ionic stress-prone conditions; however, this remains an inferred relationship rather than direct evidence of physiological adaptation. Concurrently, the low representation of plant pathogens and wood-decayers could reflect a predicted community shift toward general stress tolerance, although this interpretation is based on taxonomic inference. Notably, the presence of skin-associated fungi, peaking at over 40% in RM (consistent with the high abundance of Malassezia), might point toward an adaptive shift; however, these functional interpretations remain exploratory and require validation through larger spatial surveys. Finally, the substantial proportion of unknown habitat assignments (7–52%) emphasizes persistent annotation limitations acknowledged by both Põlme et al.³⁶ for FungalTraits and Nguyen et al.⁹⁸ for FUNGuild, highlighting the continued necessity for expanded functional annotation initiatives to enhance the ecological interpretation of fungal community datasets.

The limited number of sampling sites and replicates per site may restrict the generalization of these findings across all Egyptian mangrove ecosystems; in addition, the functional predictions presented here are inference-based rather than directly measured, which may limit the accuracy of functional interpretations. Nevertheless, the present study demonstrates the effectiveness of ITS2 amplicon sequencing in uncovering fungal diversity that is often overlooked by culture-dependent methods, revealing a diverse and stress-adapted mycobiome with a considerable proportion of unclassified taxa, and highlighting the Egyptian Red Sea mangroves as important ecological reservoirs of fungal resources. Future research should therefore expand sampling efforts across a broader spatial scale and include seasonal and temporal replication to better capture community dynamics. Moreover, integrating advanced multi-omics approaches, such as metagenomics, metatranscriptomics, and metabolomics, alongside quantitative PCR (qPCR) and targeted isolation and cultivation of key fungal taxa, will be essential to validate predicted functions, characterize metabolic pathways, and confirm ecological roles. Such approaches will also facilitate the discovery of novel fungal species and bioactive compounds, ultimately strengthening and extending the findings of the present study.

Conclusion

This study provides a first insight into the fungal communities associated with the rhizosphere of Avicennia marina along Egypt’s Red Sea coast, based on ITS2 amplicon sequencing. Fungal diversity and community composition appeared to vary among sites, with sodium gradients likely influencing these patterns. Linear Discriminant Analysis Effect Size (LEfSe) suggest that these communities are defined by distinct taxonomic biomarkers across all hierarchical levels. While Safaga station (K17) was characterized by stress-tolerant Ascomycota (notably Aspergillus), Ras Mohamed (RM) exhibited a unique shift toward Basidiomycota, dominated by the genus Malassezia, potentially reflecting the presence of marine-adapted lineages or a localized human influence. This strong discrimination at the genus level indicates that lower taxonomic ranks serve as more sensitive indicators of site-specific ecological variation. These patterns, further supported by limited OTU overlap in Venn analysis, suggest specialized adaptations to localized sediment chemistry and potential anthropogenic influences, though such results are methodologically sensitive and should be interpreted cautiously. Functional predictions indicated traits related to stress adaptation and anthropogenic impact, but these should be considered putative until validated by other techniques. Overall, the findings highlight Egyptian Red Sea mangroves as promising reservoirs of diverse fungi and underscore the usefulness of ITS-amplicon sequencing for exploring these communities. Future studies integrating multi-omics approaches, culture-based methods, and qPCR targeting specific functional genes will be important to better understand their ecological roles and assess their biotechnological potential.

Abbreviations

ITS

Internal Transcribed Spacer

PCoA

Principal Coordinate Analysis

OTUs

Operational Taxonomic Units

RM

Ras Mohamed

K17

Alkilo-17 south of Safaga

EC

Electrical Conductivity

OM

Organic Matter

Author contributions

N.A.I. and M.I. designed the study. N.A.I. and S.A.D. collected the samples. S.A.D. managed soil physicochemical analysis. A.E.E. performed DNA extraction and managed the sequencing process. M.I., H.M.A., and S.A.D. conducted bioinformatics analyses of alpha and beta diversity, taxonomic assignments, and functional annotation. N.A.I. Interpreted and discussed diversity-related data and analyzed correlations. All authors drafted the manuscript. N.A.I. and M.I. revised the manuscript. All authors read and approved the final manuscript.

Data availability

The raw sequencing data generated in this study have been deposited in the NCBI Sequence Read Archive (SRA) under accession number PRJNA1130041.

Declarations

Competing interests

The authors declare no competing interests.