Berre Lagoon fish kill: what the water reveals

Since Friday 11 September 2026, a substantial number of dead fish have washed ashore in the Saint-Chamas inlet, at the northern end of Berre Lagoon in southern France. According to explanations from the lagoon authority Gipreb, relayed by the municipality and local media on 14 September, fish encountered deep oxygen-free water that had risen towards the surface after strong northerly winds. The incident shows why a conspicuous fish kill should not be labelled as chemical pollution or disease before investigators reconstruct how the water mass moved.

Summer heat had already weakened the lagoon

Berre is a shallow, productive Mediterranean coastal lagoon whose physical structure changes with temperature, salinity and wind. On 13 August, after repeated heatwaves, Gipreb reported water temperatures of 30°C in the centre and 32–33°C at some bathing sites. Data recovered from monitoring probes indicated that most deep layers were already anoxic, while surface water still contained oxygen.

Heat acts through several pathways. Warmer water holds less dissolved oxygen, raises the metabolic demand of many organisms and accelerates microbial processes. Where light and nutrients are available, it may also support high phytoplankton production. Part of that biomass eventually sinks, and bacterial decomposition consumes oxygen close to the sediment.

The water column is not a single, uniformly mixed volume. Differences in temperature and, importantly in Berre, salinity can leave lighter surface water above denser bottom water. This density barrier restricts oxygen transfer from the atmosphere. Surface measurements may therefore appear acceptable while a deeper water mass contains too little oxygen to sustain aerobic animals.

Wind moved the hazard into fish habitat

Strong wind often helps a lagoon by mixing and reoxygenating deeper water. Its immediate effect at Saint-Chamas was more complex. As Gipreb explained in reporting by Maritima, the mistral on 9 and 10 September drove surface water southwards. Deep anoxic and acidified water then rose in the northern lagoon to replace it, a circulation process known as upwelling.

Mobile fish may escape oxygen-poor water when a suitable refuge remains accessible. In this case, a rapid rise brought degraded deep water into near-surface habitat. Not every fish in the inlet could move away in time. The carcasses on shore were therefore the visible endpoint of a process that had developed earlier and below the surface.

Locally, such an episode is described as a malaïgue, an Occitan term meaning “bad water”. It refers to acute deoxygenation in a coastal lagoon. Once oxygen disappears, anaerobic microbial pathways alter water chemistry and may produce sulphides, sometimes giving the water a whitish appearance. The term describes a biogeochemical disturbance. It does not mean an infectious outbreak, nor does it alone demonstrate an external chemical spill.

What the evidence supports — and what remains uncertain

The documented heat, deep-water anoxia measured in August, timing of the wind event and reported circulation change form a coherent mechanism. Public communications attribute the mortality to contact with oxygen-free, chemically altered bottom water. On 14 September, Gipreb reported that the water had reoxygenated and that ecological conditions had improved.

This interpretation should not be applied automatically to every lagoon fish kill. Pollution, a toxic bloom, infection or several concurrent stressors can produce similar observations. Discoloured water, odour and stranded carcasses do not distinguish these causes on their own. Likewise, reported acidification should not be reframed as “acid poisoning” without suitable measurements, depth profiles and analytical evidence.

The available public record has limits. It does not yet provide a detailed species and size-class inventory, a total carcass count, oxygen and pH profiles at the exact time of death, or examinations of affected fish. The sources support the proposed hydrological mechanism, but they do not amount to a complete veterinary or ecotoxicological investigation report.

Practical lessons for lagoon monitoring

For lagoon managers, local authorities, environmental organisations and field teams, the first lesson concerns depth. A surface reading cannot characterise a stratified water column. Temperature, salinity, dissolved oxygen, percent saturation and pH should be measured at several depths, particularly in sheltered inlets and close to the bottom. Time series are more informative than isolated readings because daytime oxygen production, night-time respiration and wind shifts can move risk within hours.

The second lesson is meteorological. Wind should not be treated as uniformly beneficial or harmful. Direction, duration and intensity interact with basin geometry. Wind may reoxygenate deep water, yet it can also bring a degraded water mass into fish habitat during the initial circulation response. Useful alerts therefore combine probe data, wind forecasts, temperature, water-colour observations and fish behaviour.

The third lesson concerns incident investigation. When moribund or dead fish first appear, teams should timestamp and geolocate observations, estimate affected taxa and size classes, photograph the distribution, measure water before conditions change and notify the relevant authorities. Freshly dead or moribund fish may support diagnostic work if a laboratory is contacted before collection and traceability is maintained. Even when an environmental cause looks likely, appropriate samples can help detect a contributing factor or test competing hypotheses.

Keep bathing advice tied to current local decisions

The Saint-Chamas municipality relayed Gipreb’s statement that the bathing-water quality was not affected by this phenomenon and that the water had reoxygenated. It nevertheless advised people to avoid areas with large accumulations of dead fish. That advice was local and time-specific. It does not replace current municipal decisions, official health monitoring or notices displayed at each bathing site.

Professional communication should separate ecological condition, bathing-water safety and the diagnosis of animal mortality. These questions rely on different indicators and responsible authorities. Water that meets bathing criteria may still belong to an ecosystem that has suffered severe damage; conversely, a fish kill does not by itself demonstrate a human-health hazard.

Turn one incident into a preparedness protocol

The Saint-Chamas event illustrates how a visible emergency may originate several metres below the surface and weeks earlier. Heat and organic production prepared an anoxic deep-water mass; wind then moved that water into fish habitat. Reconstructing that sequence prevents premature conclusions and directs measurements to the right locations and times.

Managers can convert these lessons into a practical plan: identify sentinel stations and depths, increase monitoring frequency during hot periods, define site-specific alert criteria, plan the response to wind shifts, and agree water and animal sampling requirements with laboratories in advance. Vetofish can support mortality documentation, veterinary sampling, incident protocols and integrated interpretation of biological and environmental evidence—without assuming a single cause before the investigation is complete.