Fecal samples were collected from wild oriental storks in Tianjin Qilihai Wetland (39°16′N-39°19’N, 117°27′E-117°38′E), which is located in the migratory flight route of East Asia-Australasia migratory birds and rich in wildlife resources, so it is an important resting and migrating place for oriental storks. Within the Tianjin Qilihai Wetland, the work of habitat micromodification that include dredging of river channels and reinforcing of corresponding embankments across some water areas and were carried out in 2023. This study collected samples across three time periods including Pre-change group (2022, 10), Under-change group (2023, 10), and Post-change group (2024, 8) (Supplementary Table S1).

Fecal sample collection were performed only on oriental storks that were continuously observed in Tianjin Qilihai Wetland for no less than 7 days. A total of 28 fecal samples were collected immediately after confirming the excretion behavior and stored in 5 mL sterile EP tubes. Then, samples were stored at −80 °C until further processing. After sample collection, it was essential to document habitat characteristics of the sampling sites and collect/identify plant species distributed within the immediate vicinity of the samples.

The TIANamp Stool DNA Kit (TIANGEN, China) was used to extract genomic DNA from fecal samples. After extraction, the Nano Drop 2000 spectrophotometer (Thermo Fisher Scientific, United States) was used to determine the concentration and quality of DNA. The V3–V4 hypervariable regions of the 16S rRNA gene were amplified using universal bacterial primers (338F 5’-ACTCCTACGGGA GGCAGCA-3′ and 806R 5’-GGACTACHVGGGTWTCTAAT-3′). The PCR reaction was prepared in a 20 μL mixture containing 4.0 μL of 5 × Fast Pfu Buffer, 2.0 μL of 2.5 mmol/L dNTPs, 0.8 μL each of forward and reverse primers (1 μmol/L), 0.4 μL Fast Pfu DNA Polymerase, 0.2 μL BSA, and 10 ng template DNA, with the final volume adjusted to 20 μL using ddH₂O. PCR amplification was performed under the following conditions: initial denaturation (95 °C, 3 min); 27 cycles of denaturation (95 °C, 30 s), annealing (55 °C, 30 s), and extension (72 °C, 45 s); final extension (72 °C, 10 min). The purified PCR amplicons were sequenced on the Illumina MiSeq platform by Majorbio Bio-pharm Technology Co., Ltd. (Shanghai, China).

Raw paired-end sequencing data underwent quality control using Fastp, followed by sequence assembly with FLASH v1.2.11.¹ The workflow implemented: Sliding-window scanning (10-bp window) starting from the 3′-end of reads, truncating subsequent bases when the average Q-score within the window fell below 20; Removal of reads shorter than 50 bp post-truncation and those containing ambiguous bases (N); Assembly with a maximum overlap mismatch rate threshold of 0.2.

Raw sequencing data were processed using QIIME2 (35). After quality filtering and denoising, sequences were clustered into Operational Taxonomic Units (OTUs) at 97% similarity using the UPARSE algorithm (36). Taxonomy was assigned using the SILVA database.

Diversity and statistical analyses Alpha diversity indices (Ace, Chao, and Shannon) were calculated to assess gut microbiota richness and diversity. Beta diversity was evaluated using unweighted and weighted uniFrac distances, visualized through Principal Coordinate Analysis (PCoA). Kruskal-Wallis rank-sum tests were used for multi-group comparisons, while Wilcoxon rank-sum tests were used for pairwise comparisons. PICRUSt2 was utilized to predict the functional potential of the microbial communities based on 16S rRNA gene sequences (37). The resulting functional profiles were analyzed at the KEGG pathway level (38). BugBase was employed to predict the potential pathogenicity of the gut microbiota and analyze species contributions to pathogenicity (39). All statistical analyses and visualizations were performed using R (version 4.0.0) with relevant packages (e.g., vegan, ggplot2).

The collected plant and feces samples of oriental storks were pretreated according to the method from Mo et al. (40). When treating plants, different parts of the plants (such as flowers, seeds, and stems) should be pretreated, respectively, and the plant cells were magnified 100 times to record the morphology by microscope. When examining fecal samples, four slides needed to be prepared. After magnifying 100 times, 10 non-overlapping fields of view were counted. Based on the structure of the plant cells that had been captured, the types of plants in the fecal samples were determined. Statistically identified the species and occurrence frequency (F) of plants, and converted the frequency into average density (D):

Convert the average density (D) into the relative density (RD):

D_i represented the density of recognizable keratin fragments of a certain plant, and ΣD represented the sum of recognizable keratin fragments of all plants. The relative density was used to estimate the actual proportion of plants.

Inter-taxonomic correlations between dominant gut microbiota components (at phylum and genus levels) and dietary plant species of oriental storks were assessed using bivariate Spearman analysis in Origin 2022 (version 9.7).

This study conducted microscopic tissue analysis on the fecal samples of oriental storks. Referring to the microscopic structure of plants in Tianjin Qilihai Wetland, plants in the photographs were identified and classified (Fecal samples from the post-change group were not subjected to microscopic examination due to insufficient sample volume) As shown in Table 1, a total of 8 families, 10 genera, and 10 species of plants were identified. Oriental storks consumed more plant species (9 species) in the under-change group than in the pre-change group (8 species). There were a total of 5 families, 7 genera and 7 species of plants in the two groups. They, respectively, included Phragmites australis, Echinochloa crusgalli, Digitaria sanguinalis, Cynanchum chinense, Myriophyllum spicatum, Bolboschoenus yagara, and Pharbitis nil. In the pre-change group, the proportions of plants except P. australis were all lower than those in the under-change group. In the pre-change group, there was 1 family, 1 genus and 1 species, namely Abutilon theophrasti. In the under-change group, there were 2 families, 2 genera and 2 species, namely Suaeda glauca and Nelumbo nucifera.

This study conducted correlation analysis between the relative density (RD) of consumed plants and the abundances of the top 10 phyla and top 15 genera of gut microbiota in oriental storks during the pre-change and under-change groups. The Spearman correlation results were shown in Figure 6. The results indicated that the abundance of gut microbiota in oriental storks were related to their plant diet. In the pre-change group, Acidobacteriota and norank_f__Vicinamibacteraceae were significantly negatively correlated with M. spicatum (0.01 < p ≤ 0.05); Bacillus and Desulfobacca were significantly negatively correlated with B. yagara (0.01 < p ≤ 0.05); Lysobacter was significantly negatively correlated with C. chinense (0.01 < p ≤ 0.05); Patescibacteria was significantly positively correlated with E. crusgalli (0.01 < p ≤ 0.05); Lysobacter was extremely significantly negatively correlated with A. theophrasti (0.001 < p ≤ 0.01). In the under-change group, Campilobacterota was significantly negatively correlated with D. sanguinalis (0.01 < p ≤ 0.05); unclassified_k__norank_d__Bacteria was significantly negatively correlated with both P. australis and C. chinense (0.01 < p ≤ 0.05); Desulfobacterota was significantly positively correlated with D. sanguinalis (0.01 < p ≤ 0.05); Bacteroidota was significantly positively correlated with S. glauca (0.01 < p ≤ 0.05); Lactobacillus was significantly positively correlated with P. australis and B. yagara (0.01 < p ≤ 0.05); Fusobacterium was significantly positively correlated with P. nil (0.01 < p ≤ 0.05); Eubacterium_nodatum_group was significantly positively correlated with N. nucifera (0.01 < p ≤ 0.05); norank_f__Eggerthellaceae was significantly positively correlated with E. crusgalli (0.01 < p ≤ 0.05); unclassified_k__norank_d__Bacteria was extremely significantly positively correlated with M. spicatum (0.001 < p ≤ 0.01).

By analyzing the fecal microbial composition of oriental storks at different stages across three consecutive years consistently, the results demonstrated that at the phylum level, the top three dominant phyla included Firmicutes and Actinobacteria, consistent with our previous findings (20). Firmicutes is prevalent among numerous migratory bird species including black-necked cranes (Grus nigricollis), mallards (Anas platyrhynchos), and bean geese (Anser fabalis) (41–43). Firmicutes ferments carbohydrates, polysaccharides, and lipids (44, 45), facilitating host nutrient absorption and energy acquisition. Critically, Firmicutes-derived short-chain fatty acids (SCFAs) regulate host metabolism and mitigate intestinal disorders (46).

During habitat micromodification, Firmicutes abundance exhibited significant dynamic changes. It was substantially higher in the under-change group than in pre-change group and post-change group, while remaining elevated in post-change group compared to pre-change group. This pattern coincided with water replenishment activities at Tianjin Qilihai Wetland, which reduced water levels and suspended fish stocking. As the principal energy source for oriental storks, reduced availability of fish resources may decrease energy intake. PICRUSt2-based functional prediction revealed significantly lower abundance of microbial energy metabolic pathways during the under-change group versus pre-change, with a recovery tendency in post-change group. These functional shifts are temporally aligned with dietary perturbation, suggesting potential host energy adaptation challenges. As facultative piscivores shifting to plant-based foods (e.g., reed rhizomes, aquatic seeds) during fish scarcity (47), storks likely increased consumption of high-fiber vegetation during under-change group, potentially driving Firmicutes proliferation to enhance plant polysaccharide digestion.

Actinobacteria, also a dominant phylum, facilitates the degradation of chitin and cellulose and helps maintain gut-immune homeostasis (48–50). Notably, the dominant phyla in pre-change group were characterized by Firmicutes, Proteobacteria, and Actinobacteria, whereas under-change and post-change group featured Firmicutes, Actinobacteria, and Fusobacteriota. The abundance of Fusobacteriota was significantly higher in under-change group than in pre-change group.

Fusobacteria, which are widely distributed in nature, include species that produce butyrate to aid the host in fat absorption and form biofilms to prevent pathogen colonization and invasion (51, 52). According to previous studies, Fusobacteria is dominant bacteria in Anser anser (53), Calypte anna (54), Gyps himalayensis (55) and other birds. Some studies found that Fusobacteria is harmful to species, such as Fusobacterium nucleatum, which is associated with enteritis (56). Stokowa-Sołtys et al. (57) found that F. nucleatum can produce butyrate and act on the immune system to prevent the invasion of pathogens. This study speculates that the increased abundance of Fusobacteria may reflect the enrichment of potential pathogenic bacteria caused by changes in the wetland environment, or the physiologically adaptive adjustment of the host immune system.

At the genus level, Paeniclostridium and Lactobacillus persisted as dominant taxa across all phases. Paeniclostridium, a known pathobiont linked to immune dysregulation (58), peaked during under-change group. This coincided with increased abundances of other potential pathogens (e.g., Actinomyces) (59). Compensatory increases in beneficial bacteria like Lactobacillus [crucial for gut barrier repair; (60, 61)] was observed during this phase, with its abundance significantly higher in under-change group than post-change group. Furthermore, enteritis-associated genera Peptostreptococcus and Peptococcus (62, 63) peaked in under-change group. Studies confirm that identical pathogenic bacteria exhibit variable pathogenicity and symptom presentation among individual hosts (64). Consequently, establishing clinical relevance requires correlating microbial data with species-specific manifestations. Currently, the pathogenicity of the genera identified in this study remains unclear for oriental storks. While our findings may not directly reflect this species’ clinical health status, they lay essential groundwork for future investigations into stork-specific pathogen dynamics.

Collectively, these microbial shifts may indicate compromised host health during active modification. The post-change group showed reduced pathobiont abundances and increased beneficial taxa like Lactobacillus and Eubacterium_nodatum_group which could enhance SCFA production compared to pre-change group (61, 65). This underscores that habitat micromodification transiently threatens oriental storks, health, necessitating future management protocols that minimize disturbance during critical life-history stages (e.g., migration, breeding) and ensure adequate food provisioning.

Declining water levels in the species’ habitat will impact the diversity and availability of its food resources (66). The results of this study indicated that the diversity of the species’ gut microbiota was related to environmental modifications. In 2023, the water level in Tianjin Qilihai Wetland was the lowest, leading to a decrease in the number of aquatic plants and a scarcity of food resources. The number of fish and benthic organisms available for the oriental stork to feed on also decreased (67). The reduction in food resources increases competition among species with similar dietary preferences, makes it more difficult to hunt fish, and increases energy consumption. Consequently, the oriental stork may shift to consuming more plants, which in turn alters the diversity and functionality of their gut microbiota. For example, compared to the pre-change group, the number of microbial communities such as Cetobacterium, which are obtained from fish (68), decreased in under-change group. Additionally, the abundance of gut microbiota related to energy metabolism was super significantly lower in under-change group compared to the pre-change group and was notably lower than post-change group. This further suggests that fewer fish were available for the oriental stork in the under-change group. In the post-change group, the water level in the wetland recovered, leading to an increase in the number of aquatic plants. During the post-change group, the abundance of Candidatus_Bacilloplasma was significantly higher than in the pre-change group and under-change group. Xiao et al. (68) demonstrated that food-derived microbiota shape host gut communities. Crustaceans, including crabs and shrimps in the oriental stork’s diet, harbor high abundances of Candidatus_Bacilloplasma in their enteric microbiota (69, 70). We therefore propose that increased Candidatus_Bacilloplasma abundance during the post-change group may reflect heightened crustacean availability in the environment.

Ureaplasma can hydrolyze urea and is commonly present in avian intestines [e.g., barn swallow Hirundo rustica, (71); great tit P. major, (72)]. Bodawatta et al. (27) found that under high-protein diets, the abundance of Ureaplasma increased in the intestines of great tits (P. major). Zeber-Lubecka et al. (73) indicated that under high-fat conditions, the abundance of Ureaplasma increased in mouse intestines. The increased abundance of Ureaplasma during the post-change group may result from heightened protein and fat intake due to enriched wetland food resources (e.g., increased consumption of fish and crustaceans), but its association with specific diets requires further study.

Prior studies indicate that groups within a shared environment develop convergent gut microbiota due to similar living conditions, comparable diets, and social behaviors such as mutual preening (74, 75). Our results in pre-change and under-change groups align with this pattern. However, microbiota composition diverged in post-change group, evidenced by greater dispersion among samples and increased inter-individual variability. Although factors like sex, age, and dietary preferences are known to drive microbial differences (76, 77), logistical constraints prevented verification of individual traits (e.g., foraging choices, demographics) during this phase. We propose that physiological status and dietary selection jointly contributed to these outcomes, though mechanistic details require further investigation.

Microscopic analysis of indigestible fragments (microhistology) has long been used to monitor diets of wildlife, and a total of 10 plant species (8 families, 10 genera) were identified from the fecal samples of oriental storks. A. theophrasti was exclusive to pre-change group, while N. nucifera and S. glauca appeared only during under-change group. Seven species persisted across phases, including terrestrial plants (E. crusgalli, D. sanguinalis, C. chinense, P. nil) and aquatic plants (P. australis, M. spicatum, B. yagara). The analysis of the dietary habits of oriental storks feces revealed the presence of common plant species in Tianjin coastal wetlands, such as Phragmites australis, Echinochloa crusgalli, Digitaria sanguinalis.

Water-level declines increased light penetration, boosting M. spicatum biomass and protein content (78), explaining its higher dietary proportion in under-change group (11.9% vs. 1.02% pre-change group). Concurrently, Acidobacteria abundance plummeted. As M. spicatum contains abundant tannins (79), known to suppress Bacteroides and elevate Clostridiales in poultry (80) and reduce Acidobacteria in natural systems (81), we propose tannin-mediated inhibition as a plausible mechanism for this inverse correlation.

Water-level reductions likely exposed lotus (N. nucifera) rhizomes, facilitating storks access. Its high carbohydrate content (82) coincided with increased Eubacterium_nodatum_group, a known carbohydrate-fermenting SCFA producer (83, 84), suggesting microbial adjustments to dietary shifts. The 2023 environmental micromodification at Tianjin Qilihai Wetland involved both water level reduction and levee reinforcement. These interventions directly altered the composition of subaquatic soils and the distribution pattern of aquatic vegetation. Consequently, the observed changes in avian gut microbiota should be interpreted as a collective outcome of the integrated habitat micromodification efforts. While water level decline served as the most conspicuous driver which directly diminishing primary food sources such as fish, ancillary modifications, particularly levee construction, imposed additional selective pressures on microbial communities. Synthetically, the habitat micromodification itself likely acted as the principal catalyst for microbial community restructuring, whereas hydrological alterations primarily influenced microbiota indirectly through trophic resource mediation.

Although P. australis consumption decreased in under-change group (9.52% vs. 14.25% pre-change group), Lactobacillus abundance increased paradoxically. As cellulose intake elevates Lactobacillus in mammals (85, 86), we attribute this to higher overall plant consumption (increasing total cellulose), and kaempferol (a flavonoid in post-change group) [S. glauca; (87)] which consistently enriches Lactobacillus across vertebrates (88, 89). Increased under-change group consumption of B. yagara (8.33% vs. 1.27% pre-change group), containing immune-modulating resveratrol and betulin (90), further supported Lactobacillus enrichment, as resveratrol ameliorates gut dysbiosis in birds (91), rodents (92), and swine (93). Comparative analyses revealed that declining water levels triggered increased consumption of fibrous plant matter in Grus grus, accompanied by rising Lactobacillus abundance in their gut microbiota (94). This microbial adaptation pattern mirrors our findings in oriental storks during under-change group, demonstrating how wading birds employ functional microbiome restructuring to cope with nutritional changes during habitat fluctuations. When fish meal was partially replaced by Tenebrio molitor meal in the diet of defatted yellow mealworm (Larimichthys crocea), a significant increase in intestinal Lactobacillus abundance was observed (95). The reduction in fish-derived protein during the under-change group may partially explain the elevated Lactobacillus levels. Therefore, we propose that the increased Lactobacillus abundance in our study could be collectively driven by both plant-based and animal-based dietary components. Kaempferol also drove Bacteroidota enrichment during under-change group (96, 97), absent in pre-change group when S. glauca was unconsumed. These results indicate that environmental micromodification affected the type and quantity of available food resources, which in turn influenced the foraging behavior and gut microbiota structure of oriental storks. This demonstrated the oriental stork’s dual adaptive strategies: physiological adjustments and behavioral plasticity in response to environmental changes. From a management perspective, it is recommended that engineering activities avoid migration and breeding seasons, that water levels be maintained to support aquatic prey populations, and that food supplementation be provided during disturbance periods to reduce foraging costs. This study provides a scientific basis for the restoration of degraded habitats of endangered species by elucidating the interactions among diet, habitat, and gut microbiota. Prior to modification, thorough investigation into species-specific ecological niche requirements should be conducted to minimize ecological disruption. During modification, key prey resources should be enhanced to counteract microbiota dysbiosis caused by food web fragmentation. Post-modification, continuous monitoring of microbial diversity and habitat recovery indicators is essential to dynamically adjust management strategies, thereby promoting functional habitat recovery and ensuring suitable conditions for reproduction, roosting, and survival.

While oriental storks primarily consume animal-based foods, the link between plant intake and gut microbiota should not be overlooked. As documented in Marabou storks, a dietary shift toward omnivory due to habitat alteration induced significant microbiota restructuring (98). This demonstrates that food consumption influences microbial communities even in predominantly carnivorous species. Consequently, gut microbiota composition likely represents a composite outcome of both animal and plant dietary inputs. However, the inherent limitations of microscopic analysis, such as its inability to identify specific animal species and inherent quantification biases, prevent comprehensive animal dietary data collection in this study. This gap may confound correlations between plant intake and microbiota composition.

During the post-change group, microscopic analysis could not be performed on some fecal samples due to limited sample volume. Consequently, dietary data only cover 2022–2023, while microbial data span 2022–2024. This discrepancy may affect direct support for dietary recovery conclusions. However, it’s noteworthy that core phyla like Firmicutes and Proteobacteria showed abundance shifts during the under-change group, which then approached pre-change group levels in post-change group. Similarly, key genera such as Paeniclostridium and Lactobacillus exhibited parallel trends. These microbial patterns may indirectly reflect a restoration of dietary preferences toward pre-change group conditions.