Hydrogen Sulfide Risk in RAS and Saltwater Systems

Where hydrogen sulfide forms in recirculating systems, why saline water raises the risk, what published studies report on its toxicity to fish, and how solids, biofilters and oxidation limit it.

Why hydrogen sulfide matters in recirculating systems

The FAO guide to recirculation aquaculture warns that hydrogen sulfide (H2S) can form under anaerobic conditions, especially in saltwater systems, that it is extremely toxic to fish even at low concentrations, and that fish will be killed if it is generated in the system. Peer-reviewed work from Norway and Denmark describes H2S as a major concern in marine and brackish land-based RAS and links it to sudden mass mortalities of Atlantic salmon (Salmo salar) in recent years.

How and where it forms

Sulfide is produced by sulfate-reducing bacteria that live in anaerobic zones and use sulfate instead of molecular oxygen for respiration, as C. E. Boyd (Auburn University) explains in the Global Seafood Alliance's Responsible Seafood Advocate. Production therefore needs three things together: sulfate, organic matter as a carbon source, and local anoxic conditions. Studies of RAS identify two main hotspots.

Biofilters: in anaerobic batch-reactor tests (Rojas-Tirado et al., 2021, Aquaculture), biofilm on biofilter media hosted many sulfate-reducing bacteria and produced the most H2S once oxygen was absent and carbon was available. The authors estimated that a moving bed biofilter in brackish water (17 ppt salinity) could produce 139 g of H2S per cubic metre of biofilter if the media is poorly mixed and not oxygenated.
Accumulated solids: settled waste in rearing tanks produced H2S at a lower but steady rate. SRAC Publication 452 also notes that solids left in a system decompose, consume oxygen and release toxic gases including hydrogen sulfide.
Denitrification units: the FAO guide says these reactors should be run to keep about 1 mg/L of oxygen at the outlet, because complete oxygen depletion leads to extensive H2S production.

In fixed bed biofilter reactors, linear H2S production began when bulk-water oxidation-reduction potential (ORP) fell below 0 mV and dissolved oxygen (DO) fell below 1 mg/L (Fernandes et al., 2024, Aquacultural Engineering).

Why saltwater raises the risk

Boyd gives an average sulfate content of seawater of 2,700 mg/L, compared with roughly 5 to 50 mg/L in fresh waters of humid regions. More sulfate means more raw material for sulfate-reducing bacteria.

Sulfide forms, pH and measurement

Dissolved sulfide exists as H2S, HS− and S2− in a pH- and temperature-dependent balance. Boyd states that toxicity comes almost entirely from un-ionized H2S, that H2S and HS− are present in roughly equal proportions near pH 7, that HS− dominates at higher pH, and that S2− does not occur at aquaculture pH values. Because test methods measure total sulfide, the H2S share must be estimated from pH and temperature with published factor tables. In Boyd's worked example, freshwater at pH 7.5 and 26 °C with 0.5 mg/L total sulfide contains about 0.119 mg/L H2S; in seawater the factor is multiplied by 0.9.

Toxicity figures from the literature

General aquaculture (Boyd): 96-hour LC50 values (the concentration killing 50% of test animals in 96 hours) range from 20 to 50 µg/L H2S for freshwater fish and from 50 to 500 µg/L for marine species. Boyd suggests freshwater fish should not be exposed to more than 2 µg/L for long periods and that brackish or seawater ponds probably should not exceed 5 µg/L.
Atlantic salmon post-smolts, acute exposure (Bergstedt and Skov, 2023): the mean critical concentration, at which oxygen uptake fell below standard metabolic rate or fish lost equilibrium, was 1.78 µM H2S, independent of fish size (about 100 to 500 g). Converted with the molar mass of H2S (34 g/mol), this is roughly 61 µg/L (conversion made for this article).
Atlantic salmon smolts, behaviour (Ciani et al., 2024): during daily exposures up to about 68 µg/L (2 µM), fish swam faster and more erratically and lost schooling; responses rose with concentration up to 30 to 40 µg/L and returned to baseline once H2S was gone.
Atlantic salmon post-smolts, 4-week exposure in brackish water at 12 ppt (Lazado et al., 2024): 1 µg/L and 5 µg/L H2S gave cumulative mortality of 4.7% and 16%, with reduced growth. The authors concluded that H2S should be kept below 1 µg/L in land-based salmon systems.

Mass mortality reports

Several papers report that sudden mass mortalities of Atlantic salmon in Norwegian RAS in recent years have been attributed to, or suspected to be caused by, acute H2S exposure. Bergstedt and co-authors describe these events as linked to the sudden release of large amounts of H2S, for example from anoxic parts of a biofilter, rather than slow, steady production. One commercial monitoring study found daily averages of at most 0.6 µg/L across 70 days with no significant mortality, which shows that well-operated systems can keep levels low.

Prevention and response

Remove solids quickly from tanks and clean or backwash filters on schedule, so that sludge does not build up anoxic layers (Rojas-Tirado et al.; SRAC 452).
Keep biofilter media mixed and oxygenated. The FAO troubleshooting table gives the remedy for H2S production, noticed as a rotten-egg smell during cleaning, as more aeration and cleaning of the biofilter.
Maintain efficient degassing and steady flow through filters: in two commercial salmon post-smolt farms, H2S fell after the degassers, rose slightly after fixed bed biofilters were backwashed, and depended more on biofilter design, water exchange and aeration than on feeding or biomass (Fernandes et al., 2024).
Avoid feeding beyond system capacity, as uneaten feed adds organic matter; Boyd lists conservative feeding and strong aeration as the main preventive practices.
Lower sulfate in intake water where saline make-up is used: nanofiltration that cut sulfate about 15 times gave about three times less H2S and delayed its onset by two days in experimental biofilters (Fernandes et al., 2024).
Oxidise H2S when it appears: hydrogen peroxide added at a 1:1 molar ratio to H2S cut the H2S half-life in seawater from about 118 minutes (oxygen alone) to about 30 minutes (Bergstedt et al., 2022, Water Research).
Track ORP and DO in biofilters and sumps: ORP was strongly and inversely correlated with H2S in one exposure trial, and production started below 0 mV and 1 mg/L DO in another. Online H2S sensors are used in research and commercial monitoring.