South Africa’s sardine saga

Unravelling South Africa’s 2026 sardine mortality event

SOUTH AFRICA: Three and a half months after dead sardines first washed up on the West Coast, the die-off has ended; a virus has emerged as the most likely cause, and attention has now turned to how many fish remain in the water

On 28 July 2026, fishers reported dead sardines off Saldanha Bay. Within weeks, reports were coming in from Hout Bay, Gansbaai, Gqeberha, the Wild Coast and even Namibia, and South Africa found itself dealing with an unusual and highly public marine mortality event.

A virus never before recorded in South African sardines was soon identified causing questions to be raised about the safety of canned fish, and the future of a commercially and ecologically important fishery came under scrutiny.

With no dead sardines of note reported since September and a new set of interim catch limits in place, Maritime Review Africa traces the saga from the first sightings to where matters now stand in mid-November. This draws on the consistent series of updates issued by the Department of Forestry, Fisheries and the Environment (DFFE).

August sees samples collected

The first samples were collected off Saldanha Bay on 1 August with the help of the fishing vessel Shivon, and by 2 August dead and dying sardines had been reported from Saldanha Bay to Gansbaai. Some were drifting at the surface as far as eight nautical miles offshore near Hout Bay. As a response, scientific observers were placed aboard purse-seine vessels to record the distribution and extent of the mortality.

After a brief lull in early August, which initially suggested the event had run its course in the Western Cape, mortalities resurfaced farther east. Isolated sightings at Bakoven and off Tsitsikamma were followed on 27 August by widespread reports of sardines washing ashore around Gqeberha and floating in and around the harbour. Live and dead sardines also turned up in the Koeberg cooling-water intake basin, where seals were feeding on them.

September sardine mortality spreads

By 2 September, reports extended to the East London area and the northern Wild Coast between Port St Johns and Mbotyi, roughly 60 km south of the KwaZulu-Natal border. Strandings were also reported at Langstrand near Walvis Bay, Swakopmund and Henties Bay in Namibia.

The outermost locations spanned about 2,600 km of coastline, although the DFFE stressed that mortality was not continuous over that distance and that the Namibian strandings could not be conclusively linked to the South African event until laboratory comparisons were complete.

The mortality front stopped short of KwaZulu-Natal. One possible explanation is that relatively warm inshore water along that coast limited the northward movement of affected shoals, but the DFFE has described this as a plausible interpretation rather than a confirmed barrier. T

he last notable report was an isolated fresh stranding near Mbotyi on 10 September. Apart from a small, sparse wash-up of juvenile sardines near Mossel Bay on 23 October, no further dead sardines have been reported since.

Identifying the cause

Several possible causes have been investigated. A harmful algal bloom of Alexandrium catenella was present in Saldanha Bay when the mortality began, and this alga has been associated with sardine deaths in South Africa before. Oceanographic data pointed to enhanced upwelling, unusually cold nearshore water and low oxygen in St Helena Bay.

Toxin testing on sardines from Saldanha Bay, the Cape Canyon and Gansbaai found none of the paralytic shellfish toxins linked to the bloom, and two other algal toxins were present only at low levels, highest of all in a comparison sample of apparently healthy fish. No harmful algae were found at significant densities in stomach contents or gill rinses.

Cold, low-oxygen conditions had also been recorded during the July recruitment survey, when no dead sardines were seen. Such conditions occur periodically on the West Coast, but do not usually cause deaths on this scale.

The fact, however, that the mortality was confined almost entirely to sardines also made a general environmental cause less likely.

Virus identified

Instead, attention settled on pilchard herpesvirus (PHV). Mid-August, laboratory sequencing confirmed PHV genetic material in sampled sardines. According to the DFFE, this represents the first confirmed detection in South African sardines and the viral sequence was a 99.7% match to an Australian PHV gene.

In Australia and New Zealand, PHV was associated with two large sardine die-offs in the 1990s, causing gill inflammation and damage that can interfere with oxygen uptake.

Initial testing by the South African Institute for Aquatic Biodiversity (SAIAB) and Rhodes University found consistently high viral levels in fish from the mortality event and lower, more variable levels in fish caught alive and apparently healthy.

PHV was also later confirmed in samples from Gqeberha and Port St Johns, linking the Eastern Cape and Wild Coast mortalities to the earlier Western Cape event. By mid-September, the DFFE described PHV as the most likely primary cause, while noting that environmental conditions could not be ruled out as contributing stressors.

Further testing resumed in October after being temporarily delayed by a shortage of essential reagents. By 19 October, the DFFE laboratory had received around 976 samples and completed analysis of 326. A substantially greater proportion of dead sardines tested positive than apparently healthy fish, and infected fish were, on average, in poorer condition.

The most recent findings shed light on timing. PHV was detected in apparently healthy sardines collected around Cape Agulhas during the recruit survey on 22 and 29 July, before or at the same time as the first dead fish were reported off Saldanha Bay on 28 July. This suggests the virus was circulating earlier, and over a wider area, than the mortality reports alone indicated. PHV was also found in the small Mossel Bay wash-up in October.

How and when the virus reached South African waters remains unknown. Imported sardines, bait and other biosecurity pathways have been considered. Imported samples supplied by industry have been tested, but no country, company, product or shipping route has been identified as the source. The 99.7% sequence match confirms the virus’s identity but says nothing about its origin.

Mortality rate remains unanswered

The full extent of sardine mortality remains unanswered. Most reports rely on opportunistic sightings, photographs and videos rather than systematic counts, and they tend to capture the densest concentrations. An unknown number of fish may have sunk or remained offshore undetected.

The DFFE has therefore described its mortality estimates as scenario-based and highly uncertain, and has cautioned against adding them together into a single total.

The department also corrected a figure circulating publicly, stating that it never produced or approved an estimate that 1% of the sardine population had died. That number came from a rough illustrative calculation during early deliberations.

Similarly, the 60% to 70% mortality figures from the Australian events cannot be applied directly to South Africa, given differences in survey methods, coastline, oceanography and stock context. The DFFE also noted that footage of unrelated fish kills elsewhere in the world had been shared in connection with the local event, creating a misleading impression of its scale.

Managing the fishery

The mortality event came shortly after the July 2026 decision to increase the sardine total allowable catch (TAC), which drew questions from conservation organisations. The DFFE explained that the July decision was based on the scientific information available at the time and that PHV was new information to be incorporated into ongoing management.

Detection of PHV did not automatically trigger a closure or a TAC cut. Instead, the Small Pelagic Scientific Working Group (SWG-PEL) reviewed the evidence at regular intervals and recommended a series of precautionary, time-limited measures, which the DFFE implemented:

◼︎ From 11 September to 2 October, each right holder could land no more than 25% of its outstanding 2026 directed sardine TAC, west and east of Cape Agulhas.

◼︎ From 3 to 30 October, the limit was raised to 33% of each right holder’s remaining uncaught directed sardine TAC.

◼︎ From 5 November, the 33% interim limit was extended to 30 November, amounting to roughly 2,800 tonnes west and 2,700 tonnes east of Cape Agulhas. If catches in an area exceed half its limit by 15 November, that area’s limit increases by 80%.

In practice, little fishing took place. Strong winds and poor weather limited activity through September and October, with only one directed catch of 26 tonnes recorded in the first weeks of September. This was followed by limited catches at Gansbaai, in Algoa Bay and off Cape Point, and some sardine bycatch with anchovy and redeye. Those catches have nonetheless been valuable as a source of samples for testing.

Model projections considered by the SWG-PEL showed only a modest difference in biomass between a full moratorium and catching the remainder of the 2026 TAC. The stepped limit reflects both the uncertainty about the scale of the mortality and the fact that poor fishing has provided too little information to judge sardine availability. The SWG-PEL will review the limits before they expire.

Food safety and public advice

As public concern grew, the DFFE and the National Regulator for Compulsory Specifications (NRCS) issued a joint statement on 27 August noting that the PHV is a fish virus and is not known to infect people or other animals.

According to the statement, canned sardines represented no threat as they are produced in NRCS-approved facilities, where the validated heat-treatment process used in canning destroys viruses, including PHV.

The public advice for beach-cast fish has remained unchanged throughout: do not collect, eat, feed to animals or use as bait any dead or dying sardines found on the shore or floating nearby. The concern is not the virus itself but the fact that there is no way of knowing how long a stranded fish has been dead or how far it has decomposed.

Current situation

The latest statements from the DFFE indicate that the acute phase of the event is over. PHV is regarded as the most likely cause, with environmental conditions possibly playing a supporting role.  The fishery is operating under cautious, reviewable limits rather than a closure.

The response has also drawn together government, SAIAB, Rhodes University, the fishing industry, oceanographic institutions, conservation bodies and Namibian authorities under a coordinated plan overseen by a high-level PHV task team.

Despite this, several questions remain unanswered. The true scale of the mortality is still unknown, as is the origin of the virus, and testing continues on samples collected before, during and after the event.

The key outstanding piece of evidence is the fisheries-independent hydroacoustic survey, planned to cover the coast from Hondeklip Bay to Port St Johns.

Preparations are under way, and the survey will provide the first estimate of sardine biomass, distribution and size structure since the die-off. Its results will shape the next stage of stock assessment and the management of the fishery in 2027 and beyond.

Beyond the Cape: What shipping route diversification means for South Africa