Aquaculture is one of the fastest growing food sectors on the planet, but raising fish at high density in closed systems creates a stubborn chemical problem: nitrogen. Fish excrete ammonia, beneficial microbes convert it to nitrite and then nitrate, and if nitrate builds up, operators must either dilute the water with large volumes of exchange or invest in bulky denitrification equipment. A research team at Nagaoka University of Technology in Japan has now tested a compact answer to this dilemma, pairing two sponge-based biofilm reactors in sequence to handle both halves of the nitrogen cycle inside a marine recirculating aquaculture system, or RAS, stocked with Japanese flounder.
The technology at the heart of the study is the down-flow hanging sponge reactor, a biofilm system originally developed for municipal wastewater treatment. Wastewater trickles from the top of the reactor over cubes of porous polyurethane sponge, each roughly 33 millimeters across with a porosity exceeding 90 percent. The open, gravity-driven structure draws in oxygen by natural ventilation rather than forced aeration, while the sponge matrix retains an enormous population of slow-growing microbes that would otherwise be washed out. Crucially, the sponges create microscopic oxygen gradients: aerobic zones near the surface where ammonia can be oxidized, and anoxic microzones deep inside where nitrate can be reduced to nitrogen gas. In principle, a single sponge cube can host both nitrification and denitrification at the same time.
In the new experiment, the researchers built a pilot-scale marine RAS with a 500-liter culture tank holding Japanese flounder (Paralichthys olivaceus) at an initial stocking density of about 3.7 kilograms per cubic meter, fed a commercial pellet diet at 2 percent of body weight per day in seawater of roughly 30 parts per thousand salinity. Water was pumped from the tank into an anaerobic down-flow hanging sponge reactor, or AnDHS, a 100-liter vessel packed with 50 liters of sponge carriers and seeded with denitrifying granular sludge. From there, water flowed by gravity into an aerobic DHS reactor installed downstream, which polished the effluent by oxidizing residual ammonia and nitrite and reoxygenating the water before it returned to the fish. A UV unit disinfected the recirculating stream, and make-up water was added only to compensate for evaporation and sampling losses.
The 116-day trial was deliberately staged in phases to probe how two operational levers affected performance. In the first phase, the AnDHS received no external carbon source at all. From day 32 onward, sodium acetate was dosed to maintain a carbon-to-nitrogen ratio of 3, and the aerobic DHS sponge volume was progressively enlarged from 15 liters to 25 liters and finally to 40 liters, arranged in two layers with a hydraulic retention time of 2.0 minutes. In the final phase, the AnDHS carbon-to-nitrogen ratio was raised to 6. Because the modifications were applied sequentially rather than in an independently replicated factorial design, the authors caution that the individual contribution of each change could not be separated; the results describe the integrated response of the whole system.
The early phases revealed exactly why balancing the two reactors matters. When the aerobic DHS held only 15 liters of sponge and the AnDHS ran at a carbon-to-nitrogen ratio of 3, ammonia and nitrite climbed in the culture tank, with nitrite reaching concentrations that would pose a toxicity risk to fish. The team attributed this to partial denitrification in the AnDHS, which converted nitrate only as far as nitrite, combined with insufficient nitrification capacity in the undersized aerobic reactor. Once the aerobic sponge volume was increased to 40 liters, about 4 percent of the tank volume, and the carbon-to-nitrogen ratio was raised to 6, the system stabilized dramatically. Total ammonia nitrogen settled at 0.78 plus or minus 0.98 milligrams of nitrogen per liter, nitrite remained near or below detection, and nitrate accumulated only gradually while staying below the target threshold for marine RAS operation. From that point on, routine water exchange was no longer needed at all.
Two other water-quality benefits emerged that would normally require separate equipment in a conventional RAS. First, the culture tank pH stayed above 7.0 throughout the experiment without any chemical base addition, because denitrification in the AnDHS recovered the alkalinity that nitrification consumed. Second, dissolved oxygen in the tank consistently exceeded 8.0 milligrams per liter even though the tank itself was not directly aerated; the aerobic DHS reoxygenated the water during its gravity-driven passage, while the AnDHS effluent dropped to anoxic conditions appropriate for denitrification. In many existing farms these functions are handled by degassers, aeration tanks and base dosing systems, so folding them into two sponge columns represents a meaningful simplification of the treatment train.
Microbial community analysis based on 16S rRNA gene sequencing added a fascinating layer to the story, though the authors stress that marker-gene data indicate potential functional affiliations rather than proven activity. In the aerobic DHS, ammonia-oxidizing archaea related to Candidatus Nitrosopumilus dominated, particularly in the upper sponge layers where oxygen and ammonia were presumed highest. These archaea possess an exceptionally high affinity for ammonia and tolerate low oxygen and fluctuating pH, traits well suited to marine biofilters. Nitrospira, the nitrite-oxidizing lineage, was not detected until day 49, consistent with its slow growth and sensitivity to salinity and organic loading, and its delayed appearance mirrored the early nitrite accumulation. Denitrifying bacteria were also abundant in the aerobic reactor, often exceeding the nitrifiers in relative abundance, which fits the idea that anoxic microzones inside the sponges support simultaneous nitrification and denitrification within a single biofilm.
The AnDHS community told a complementary story. Sequences affiliated with Sedimenticola, an acetate-utilizing denitrifier, were dominant, matching the sodium acetate feed and the observed decline in nitrate, and pointing to acetate-dependent heterotrophic denitrification as the main nitrate-removal pathway. Ammonia-oxidizing bacteria related to Nitrosomonas and members of the OM190 lineage were also detected, hinting at locally oxygenated niches within the nominally anoxic sponge matrix, although their activity was not directly demonstrated. The authors also flag an intriguing greenhouse-gas angle: ammonia-oxidizing archaea generally emit less nitrous oxide than ammonia-oxidizing bacteria, and comammox Nitrospira, which oxidize ammonia all the way to nitrate without releasing nitrite, could further reduce both nitrite toxicity and nitrous oxide formation. However, standard 16S sequencing cannot distinguish comammox from canonical Nitrospira, so that possibility remains speculative pending functional-gene or metagenomic confirmation.
From an engineering standpoint, the sponge reactors showed impressive operational resilience. Over the full 116 days there was no sponge clogging, no excessive biomass accumulation, no significant pressure loss, and no need for backwashing, sponge cleaning or media replacement; the only maintenance was routine inspection of the recirculation pump. Because oxygen enters by natural ventilation, the system avoids the energy costs of forced aeration and carrier mixing, and the porous sponge structure reduces clogging risk compared with conventional packed-bed biofilters. Previous work in aquaponic systems even showed a DHS reactor maintaining lower ammonia than a moving-bed biofilm reactor with only one-fifth of the effective volume. The authors are careful to note, however, that energy consumption, capital costs and operational expenses were not quantified in this study, so definitive economic claims must await a proper techno-economic assessment.
The study also carries honest caveats. It was a single pilot-scale system evaluated sequentially, not a replicated experiment, and comparisons with other biofilters should be made cautiously because loading rates, species and management practices differ across studies. Still, the headline result stands on its own: by sizing the aerobic sponge volume to roughly 4 percent of the tank and feeding the anaerobic stage at a carbon-to-nitrogen ratio of 6, the team kept ammonia and nitrite low, recovered alkalinity, reoxygenated the water and eliminated routine water exchange in a marine system culturing a commercially valuable flatfish. Future work will need independently replicated designs, quantification of nitrous oxide emissions, strategies to enrich comammox organisms, and validation at commercial scale with different species. If those steps succeed, the humble sponge cube, trickling quietly in a gravity-fed column, could become a cornerstone of water-stingy, low-impact fish farming for a world that needs more protein without more freshwater withdrawals.
Subject of Research: An integrated anaerobic-aerobic down-flow hanging sponge reactor system for nitrogen removal in marine recirculating aquaculture systems
Article Title: Aerobic–anaerobic down-flow hanging reactor for recirculating aquaculture systems: Performance evaluation
Article References: Nguyen, T. Y. P., Watari, T., Akamine, T., Kato, Y., Adlin, N., Hatamoto, M., Shimabukuro, A., Satanwat, P., & Yamaguchi, T. (2026). Aerobic–anaerobic down-flow hanging reactor for recirculating aquaculture systems: Performance evaluation. Case Studies in Chemical and Environmental Engineering, 14, Article 101490. https://doi.org/10.1016/j.cscee.2026.101490
Image Credits: AI Generated
DOI: 10.1016/j.cscee.2026.101490
Keywords: recirculating aquaculture, down-flow hanging sponge reactor, denitrification, nitrification, Japanese flounder, marine RAS, biofilm, nitrogen removal, ammonia-oxidizing archaea, Nitrospira, water quality, sustainable aquaculture
Cite Scienmag News
APA
MLA
Chicago
Alan Morgan. (September 27, 2026). Sponge-Powered Reactor Keeps Fish Farm Water Clean Without Water Exchanges. Scienmag. https://scienmag.com/sponge-powered-reactor-keeps-fish-farm-water-clean-without-water-exchanges/
Alan Morgan. “Sponge-Powered Reactor Keeps Fish Farm Water Clean Without Water Exchanges.” Scienmag, 27 September 2026, https://scienmag.com/sponge-powered-reactor-keeps-fish-farm-water-clean-without-water-exchanges/. Accessed 27 September 2026.
Alan Morgan. “Sponge-Powered Reactor Keeps Fish Farm Water Clean Without Water Exchanges.” Scienmag. September 27, 2026. https://scienmag.com/sponge-powered-reactor-keeps-fish-farm-water-clean-without-water-exchanges/
Copy citation
Download RIS
Tags: ammonia removal in aquacultureammonia-oxidizing archaeaAquaculture nitrogen managementbiofilmcompact fish farm water purificationdenitrificationdenitrification technology for fish farmsdown-flow hanging sponge reactorin-situ nitrogen removal methodsJapanese flounderJapanese flounder aquaculturemarine RASmarine recirculating aquaculture systemsmicrobe biofilm technologynitrificationnitrogen cycle in fish farmingnitrogen removalNitrospiraoxygen gradients in biofilm reactorsrecirculating aquaculturesponge-based biofilm reactorssustainable aquaculturesustainable aquaculture water treatmentwastewater treatment sponge reactorswater quality
