Southern Gulf of Guinea Drainages

Between the Sanaga River in Cameroon and the Gabon estuary, a band of short, rain-fed rivers cuts through equatorial rainforest on its way to the Gulf of Guinea. FEOW calls this the Southern Gulf of Guinea Drainages (ecoregion 533): the lower Sanaga, Nyong, Ntem and Benito, plus the northern Ogooue tributaries Abanga, Okano and Ivindo, draining parts of southern Cameroon, Equatorial Guinea and northern Gabon. The water is soft, warm and tannin-stained, and the cichlids here are mostly small forest-stream fishes rather than the lake flocks of the Rift. The drama sits just inland, on the Cameroon Volcanic Line, where a string of crater lakes has each grown its own private flock of tilapiine cichlids found nowhere else on Earth.

Geography & hydrology

The ecoregion runs from the lower Sanaga south to just above the Gabon estuary, reaching roughly 249 mi inland. It sits within the West Coast Equatorial (Lower Guinea) bioregion, a low sedimentary coastal plain that climbs to an interior plateau. The named rivers — lower Sanaga, Nyong, Ntem, Benito (Mbini) and the northern Ogooue tributaries — are short by continental standards but carry a lot of water: this is one of the wettest stretches of Africa, and every river fans out into a dense net of forest tributaries (FEOW Ecoregion 533).

FEOW draws the boundaries by fish fauna rather than by watershed. A short run of falls on the Dja River walls off this ecoregion from the Sangha (534) and the Congo basin behind it. The northern Ogooue tributaries are lumped in here, not with the Ogooue mainstem, because a 50-km belt of rapids and falls below the Ivindo — together with past river captures — left the Ivindo fauna looking more like the Ntem in Cameroon than like the river it physically joins (Thys van den Audenaerde 1966; Sullivan et al. 2002). During the Pleistocene the whole region is thought to have acted as a forest refuge, which helps explain its high, patchy endemism.

The water itself does several different things. The Ivindo is largely rocky and deep, broken by rapids as it runs through closed forest. At the head of the Ntem the rivers spill their banks at peak flow and drown the forest into swamp. The Nyong, Doume and Boumba carry a rare blackwater swamp habitat: separate drainages whose headwaters merge into one flooded expanse when the rains come, then break apart again, leaving a wet forest floor stippled with pools through the dry season. Mangroves line the estuaries (FEOW Ecoregion 533).

Inland and uphill, the Cameroon Volcanic Line changes the story. This southwest-northeast ridge has been building calderas for some 25 million years, and roughly 36 crater lakes are now known from Cameroon. Most have no outflow, or an outflow too steep for fish to climb, so each lake is a sealed island. The best known, Lake Barombi Mbo, sits at about 984 ft, covers around 3 mi², and drops to 364 ft deep; its most recent lava flow is dated to about a million years ago, setting a minimum age for the lake (FEOW Ecoregion 519; Cornen et al. 1992). Nearby Barombi Kotto, Dissoni, Bermin, Kotto, Mboandong and the non-volcanic Lake Ejagham complete the cluster.

Water & habitat

The riverine habitats here are classic Lower Guinea forest water: soft, weakly buffered, warm and stained brown by leaf litter and humic acids leaching out of the rainforest floor. Conductivity stays low and pH runs acidic to near-neutral, the chemistry you would expect from rain falling on old, leached soils under closed canopy. The Nyong/Doume/Boumba blackwater swamps are the extreme end of this: tea-coloured, shaded, oxygen-poor backwaters where the forest is flooded for months at a time and small fishes work the submerged leaf litter and stands of Echinochloa grass close to the banks (FEOW Ecoregion 533). Faster, clearer, oxygen-rich water sits at the other end of the spectrum — the rocky, rapid-broken Ivindo and the gravel runs of the Ntem and Monts de Cristal streams, where current and hard substrate take over.

The crater lakes are a different kind of water altogether, and they are the reason ichthyologists keep coming back. Because most have no outflow and a small, steep catchment, they stratify hard. Barombi Mbo is meromictic: its deep water never fully mixes with the surface, so below roughly 131–164 ft the lake is permanently without oxygen and loaded with dissolved gases. The fish are squeezed into a thin oxygenated surface layer over a vast dead deep — 364 ft of water of which only the top third or so is habitable. That, plus a stable tropical temperature and the lake's age, gave evolution a closed arena to work in (FEOW Ecoregion 519; Trewavas et al. 1972).

Lake Ejagham is the odd one out. It is not a volcanic crater at all but a roughly oval solution basin about 1,050 by 2297 ft and only 59 ft deep, probably dissolved out of the bedrock by groundwater and likely formed during the last glaciation. A waterfall cuts its outlet off from the nearby Munaya River, a barrier no cichlid can pass, so despite being young and shallow it has stayed as isolated as the deep craters (Schliewen et al. 2001). The shared theme across all these lakes is the same: a small, stable, sealed body of soft water with a single colonising cichlid lineage and nowhere for it to go.

The cichlid fauna

The river fauna is dominated numerically by killifishes — about a third of the ecoregion's endemics are Aphyosemion — with endemic mormyrid 'elephantfishes' and the endemic riverine cichlid genus Parananochromis as the headline cichlids (FEOW Ecoregion 533). Three of the atlas roster's Parananochromis sit here: P. axelrodi and P. caudifasciatus from the Cameroon–Gabon forest streams, and P. gabonicus. P. axelrodi was one of three species Lamboj and Stiassny described from Gabon and southern Cameroon in 2003, slender, secretive cichlids of shaded forest creeks (Lamboj & Stiassny 2003). Sharing those creeks are the small West African mouthbrooders Benitochromis batesii and B. riomuniensis, the chromidotilapiine genus revised by Lamboj, and the jewel cichlids Hemichromis — here the regional endemics H. camerounensis and H. stellifer rather than the widespread aquarium 'jewel.'

The most familiar fishes to aquarists are the krib-group Pelvicachromis. P. kribensis, long confused with P. taeniatus, was revalidated as a Cameroonian species, and P. drachenfelsi was described in 2014 from small coastal rivers of the Wouri system; P. subocellatus rounds out the trio. The larger tilapias of moving and standing water are well represented too: Coptodon guineensis (a wide-ranging West African tilapia), the more local C. cameronensis and C. nyongana of the Nyong, the redbelly relative Pelmatolapia cabrae, and the substrate-spawning Sarotherodon lineages that also seeded the craters.

Those crater-lake flocks are what put this corner of Africa into evolutionary-biology textbooks. Lake Barombi Mbo holds 15 fish species, 12 of them endemic; every endemic except one clariid catfish is a tilapiine cichlid, and four of its five tilapiine genera — Konia, Stomatepia, Pungu and Myaka — are found in this one lake and nowhere else. The roster's Stomatepia mariae is one of three Stomatepia in that predatory lineage; Sarotherodon lohbergeri and S. mvogoi are Cameroonian Sarotherodon tied to the same volcanic-region waters (FEOW Ecoregion 519; Trewavas et al. 1972). Schliewen, Tautz and Pääbo argued in 1994 that each lake's flock is monophyletic — descended from a single colonising species that split into many without any geographic barrier — making Barombi Mbo, Bermin and Ejagham among the cleanest candidate cases for sympatric speciation in vertebrates (Schliewen et al. 1994). Lake Bermin grew nine Coptodon, including a sponge-eater and a phytoplankton specialist; Ejagham produced a Coptodon and a Sarotherodon species pair in a basin you could swim across (Schliewen et al. 2001). Later genome-scale work complicated the tidy picture, finding pulses of gene flow between lake and river cichlids, but the radiations remain a central example in speciation research (Martin et al. 2015).

Conservation

The river fishes face the slow, basin-wide pressure common to Lower Guinea: deforestation for timber, cocoa and oil palm, the silt and agrochemical runoff that follows, and expanding settlement along the Sanaga, Nyong and Ntem. Kamdem Toham's work on the Ntem flagged the potential effects of clearing forest on these fish assemblages decades ago, and the same drivers have only intensified (Kamdem Toham & Teugels 1998; FEOW Ecoregion 533). Because so much of the endemism is narrow — species known from a single river stretch, a single set of rapids, or one set of hills — habitat change does not have to be dramatic to do lasting damage.

The crater-lake endemics are the acute case. Each flock lives in one small lake with no refuge: there is no second population to recolonise from if the lake is degraded. The IUCN now lists the Barombi Mbo cichlids as Critically Endangered, and the threats are concrete — deforestation and oil-palm planting on the steep crater rim, sedimentation and pollution washing into a closed basin, pesticide spray drifting off watershed cocoa, overfishing of a tiny endemic stock, and introductions of non-native fish into a system whose species evolved with no such competitors (Bristol Zoo Project 2026; FEOW Ecoregion 519). Barombi Mbo was listed as Ramsar site 1643 in 2006 and flagged as a freshwater Key Biodiversity Area, but on-the-ground protection remains thin.

Some of these fish have already been written off and then rediscovered. Stomatepia mongo, a Barombi Mbo endemic, was repeatedly assumed extinct before surveys confirmed it was hanging on, which says as much about how little monitoring these lakes get as it does about the fish's resilience (Musilova et al. 2015). And the Cameroon Volcanic Line carries a rarer, sharper hazard. Some of its crater lakes are charged with magmatic carbon dioxide in their stagnant deep water; at Lakes Monoun (1984) and Nyos (1986) that gas erupted out of solution and asphyxiated close to 1,800 people in the surrounding valleys (USGS 1987). A meromictic, gas-accumulating lake like Barombi Mbo is on the same volcanic line, and a sudden overturn would be lethal to its trapped endemic flock as well as to people nearby. Captive-breeding programmes — such as the EAZA effort that Bristol Zoo Project joined — are now treated as an insurance policy for a fauna that could be lost from a single lake in a single event (Bristol Zoo Project 2026).

Where these cichlids live

Geography & hydrology

Water & habitat

The cichlid fauna

Conservation

Sources

Cichlids recorded here

Benitochromis · 2

Coptodon · 3

Hemichromis · 1

Parananochromis · 3

Pelmatolapia · 1

Pelvicachromis · 3

Rubricatochromis · 1

Sarotherodon · 2

Stomatepia · 1