Dammed River, Rewired Guts: How Diets Shape Fish Microbiomes in a Channelized Yangtze Tributary

Deep inside the intestines of fish swimming in one of China’s most heavily engineered rivers, an invisible ecosystem is telling a surprising story. A new study of the Jialing River—the largest tributary of the Yangtze by drainage area—has revealed that the gut microbes of carnivorous, omnivorous, and herbivorous fish follow the dietary rules scientists expected, but with one striking twist: predatory fish in the channelized river carry unusually rich machinery for digesting carbohydrates, a capacity that conventional wisdom says they should not need. The finding, published in Ecology and Evolution, suggests that thirteen cascade hydropower dams may be reshaping not just the river’s flow and food webs, but the metabolic lives of the microbes living within its fish.

The Jialing River stretches 1,345 kilometers and drains roughly 160,000 square kilometers of southwestern China. Its middle reaches, running 633 kilometers from Zhaohua to Hechuan, have been fully channelized by a staircase of thirteen hydropower dams. This cascade construction has slowed the current, homogenized habitats, and transformed prey communities from species adapted to fast-flowing water to those suited to still, reservoir-like conditions. Such profound disturbance raises an obvious question for ecologists: when a river’s plumbing is rebuilt, what happens to the microbial communities that help fish digest their food and fend off disease?

To find out, researchers sampled eight dominant fish species representing three feeding strategies from the channelized midstream between June and July 2024. The carnivorous lineup included the catfish Hemibagrus macropterus and the mandarin fish Siniperca chuatsi. Omnivores comprised Leiocassis crassilabris, crucian carp (Carassius auratus), the spotted barbel Hemibarbus maculatus, and common carp (Cyprinus carpio). Herbivores were represented by Xenocypris davidi and grass carp (Ctenopharyngodon idellus). In total, 38 specimens were collected, dissected under aseptic conditions, and their intestinal contents rinsed, flash-frozen in liquid nitrogen, and stored at minus 80 degrees Celsius until analysis.

The team then sequenced the V3–V4 hypervariable region of the bacterial 16S rRNA gene on an Illumina platform, yielding more than one million high-quality reads. After quality filtering and clustering into operational taxonomic units at 97 percent similarity, the dataset contained 5,784 OTUs, with every sample rarefied to a common depth of 28,274 sequences. Rarefaction curves plateaued at around 15,000 sequences, confirming that the sequencing captured the bulk of microbial diversity. Strikingly, only 98 OTUs—just 1.69 percent of the total—were shared across all eight fish species, underscoring how species-specific these internal ecosystems are. The herbivorous Xenocypris davidi alone harbored 959 unique OTUs, while the predatory mandarin fish had just 22.

Community composition told a clear dietary story. Firmicutes dominated the carnivore guts, exceeding 93 percent relative abundance in both predator species, with the protein-fermenting genus Clostridium_sensu_stricto_1 reaching nearly 70 percent in Hemibagrus macropterus. Omnivorous and herbivorous fish instead showed higher proportions of Proteobacteria, along with notable levels of Actinobacteria. Herbivorous fish were significantly enriched in Cetobacterium, a genus known to synthesize vitamin B12 in freshwater fish intestines—a potentially crucial service, since plant-based diets are typically deficient in this vitamin. Omnivores displayed the most taxonomically balanced communities, with genera such as Romboutsia and unclassified Peptostreptococcaceae consistently abundant across all samples.

Diversity metrics reinforced the pattern. Herbivorous fish posted the highest Shannon-Wiener index (mean 4.42) and Chao1 richness (mean 1,836.7), omnivores fell in between, and carnivores had the lowest values on both counts (Shannon-Wiener 1.21; Chao1 173.99). The differences were statistically significant, with post hoc tests separating carnivores from both other groups. Permutational multivariate analysis of variance showed that feeding habit explained 18 percent of the variation in community structure, while host species identity explained an even larger share at 42 percent. Principal coordinates analysis confirmed that carnivore samples clustered most tightly, suggesting their gut communities are strongly canalized by diet and less sensitive to environmental fluctuations than those of plant-eaters and generalists.

The functional predictions, generated with PICRUSt2 from the 16S data, were where the study’s real surprise emerged. The three expected diet-function associations held firm: carnivorous fish showed the highest relative abundance of cysteine and methionine metabolism, a pathway central to processing high-protein diets; omnivorous fish were enriched in microbial metabolism in diverse environments, reflecting their metabolic flexibility; and herbivorous fish led in carbon fixation pathways in prokaryotes, an adaptation to fiber-rich plant material. Notably, the functional profiles separated the feeding groups even more cleanly than taxonomic composition did, hinting at functional convergence—different microbial communities performing similar ecological jobs.

But two pathways defied expectations. Carnivorous fish guts showed the highest relative abundances of starch and sucrose metabolism (1.83 percent) and the phosphotransferase system, or PTS, a bacterial transport machinery that efficiently scavenges and phosphorylates sugars. This runs counter to the traditional view that predatory fish have limited carbohydrate-processing capacity. The authors propose a plausible ecological explanation rooted in the dams themselves: cascade channelization has shifted benthic invertebrate communities toward lentic species such as chironomid larvae, which are rich in glycogen and other energy-dense carbohydrates. Predators feeding on these altered prey may thus encounter more fermentable sugars than their ancestors did, and their gut microbes appear to have retained—or expanded—the metabolic toolkit to exploit them. The researchers caution that this link between channelization and carbohydrate metabolism remains a hypothesis requiring verification against non-channelized control systems.

The study’s implications extend beyond fish physiology. Because gut microbial communities track both diet and habitat conditions, the authors argue they could serve as bioindicators of aquatic environmental quality and food web health. Monitoring shifts in the gut microbiota of sentinel fish species might offer early warnings of ecosystem degradation and guide targeted conservation strategies. For reservoir management, the findings point toward a concrete recommendation: maintaining hydrological heterogeneity and diverse prey communities within dammed rivers is essential to preserving the healthy expression of gut microbial functions in fish. Restoring the diversity and stability of riverine food webs, the study suggests, is not just about the fish themselves, but about the trillions of microbes that underpin their nutrition, immunity, and adaptation.

The authors are candid about the study’s limitations. The functional profiles are predictions of metabolic potential, not direct measurements of gene expression, and PICRUSt2’s accuracy depends on reference genomes that remain sparse for taxa like Fusobacteria and Cetobacterium. The design lacks a non-channelized comparison river, sample sizes were uneven across feeding groups—grass carp was represented by only two individuals—and the work captures a single season. Future research integrating metatranscriptomics, metabolomics, stable isotope tracing, and seasonal sampling could confirm whether these predicted pathways are actively expressed and disentangle the effects of diet from host evolutionary history. Even so, the study delivers a compelling baseline: in a river remade by concrete and turbines, the ancient partnership between fish and their gut microbes endures, quietly adapting to a menu that humans have rewritten.

Subject of Research: Gut microbiota functional profiles of fish with different feeding habits in a cascade-channelized river

Article Title: Functional Profiles of Fish Gut Microbiota Associated With Feeding Habits in the Midstream of the Jialing River Under Cascade Channelization

Article References: Hu, M., Chen, C., Feng, Y., Liu, L., Tang, X., Wang, H., & Zhang, F. (2026). Functional Profiles of Fish Gut Microbiota Associated With Feeding Habits in the Midstream of the Jialing River Under Cascade Channelization. Ecology and Evolution, 16(10), Article e74414. https://doi.org/10.1002/ece3.74414

Image Credits: AI Generated

DOI: 10.1002/ece3.74414

Keywords: fish gut microbiota, 16S rRNA sequencing, PICRUSt2, feeding habits, cascade dams, Jialing River, Yangtze tributary, channelization, microbial ecology, KEGG pathways, aquatic conservation, Firmicutes