Taxonomy & naming
Retroculus septentrionalis was described by the Belgian ichthyologist Jean-Pierre Gosse in 1971, in his revision of the genus Retroculus published in the Bulletin de l'Institut Royal des Sciences Naturelles de Belgique (vol. 47, no. 43, pp. 1–13, plates 1–5). The holotype (IRSNB 515) and a substantial paratype series came from the Oyapock River at Saut-Alikoto, French Guiana — the type locality — with paratypes split between the Brussels institute and the Museu de Zoologia da Universidade de São Paulo. The specific epithet septentrionalis, Latin for 'northern', marks it out as the most northerly of the genus: while its two congeners live in the central Amazon's clearwater tributaries, septentrionalis occupies the rivers of the Guiana Shield and the Amapá coast, well to the north of the others.
Gosse's 1971 paper did three things at once: it designated a neotype for the long-confused Retroculus lapidifer (Castelnau, 1855) — Castelnau's original type from the Rio Araguaia having been lost by the Paris museum — and it described two new species, septentrionalis and xinguensis. In the same paper Gosse also settled a piece of nomenclatural bookkeeping for the genus's oldest member: he showed that Chromys lapidifera Castelnau, 1855 is a senior synonym of Retroculus boulengeri Eigenmann & Bray, 1894, so that the valid name for that species is properly Retroculus lapidifer rather than boulengeri, the combination that had been in use up to that point. The genus Retroculus itself had been erected by Eigenmann & Bray in 1894, originally for a fish (Retroculus boulengeri) later shown to be a synonym of lapidifer. For much of the twentieth century Retroculus was treated as an eartheater, allied with Geophagus and Satanoperca on the basis of a lobed first gill arch with well-developed rakers — Steindachner had moved lapidifer into Geophagus on exactly those grounds in 1875. That placement has since been overturned. Kullander's morphological phylogeny (1998) recovered Retroculus not among the geophagines at all but as a primitive, basal lineage near the root of the South American cichlid tree; the resemblance to eartheaters is convergence, not common descent. Modern classifications accordingly place the genus in its own tribe Retroculini (subfamily Cichlinae), and FishBase even lists it in a subfamily Retroculinae of its own. The species remains valid as Retroculus septentrionalis Gosse 1971, with no junior synonyms. A fourth species, Retroculus acherontos, was added to the genus from the Tocantins basin in 2015, underscoring that the group is still being worked out.
Morphology
Retroculus septentrionalis is a robust, low-slung, somewhat goby-like cichlid — stockier and flatter-bellied than a typical open-water cichlid, an obvious adaptation to life pressed against the substrate in current. FishBase gives a maximum size of about 7.5 in standard length (Kullander 2003), so a large adult is a substantial fish of roughly 8–9.5 in total length. The head bears the feature that names the genus: the eye sits far back, giving Retroculus (Latin retro, 'backward', + oculus, 'eye') its name — Gosse and later authors read this posterior eye position as a diagnostic of the genus rather than of any one species. The mouth is subterminal and protrusible, set under a fleshy snout, suited to picking small prey from between stones.
Coloration in life is a warm sandy-brown to greyish base broken by darker blotches and bars along the flank, with rows of small pearly or golden spangles on the body and fins and, in good condition, iridescent glints on the cheek and operculum; juveniles photographed in French Guiana show a mottled, cryptic pattern that blends against gravel. In overall body shape septentrionalis reads as closest to Retroculus lapidifer: the upper head profile rises comparatively gently, and the head outline traces something close to an isosceles triangle, unlike the steeper, more rounded head of xinguensis. Weidner describes the body and fin coloration as, in effect, a link between the genus's other two species, and the pattern changes markedly with age: juveniles carry vertical banding on the tail reminiscent of xinguensis and a body pattern closer to lapidifer, while adults lose the tail's stripe pattern altogether — a wild-caught adult examined by Stawikowski showed only vertical body bars and no tail striping, with the upper edge of the caudal instead taking on a reddish hue, the same feature seen in lapidifer. The single most useful diagnostic Weidner identifies for telling septentrionalis apart from its congeners is the genus-typical dorsal-fin spot: in this species it is only faintly visible and appears broken up into individual short stripes rather than forming one solid blotch, and — as in xinguensis — it sits visibly offset from the ridge of the back rather than hugging it. Beyond this, the species are separated on fine counts — longitudinal scale rows and caudal-peduncle scale numbers — as set out in Gosse's revision. Sexual dimorphism is weak and not well documented for this species specifically: in captive Retroculus the dominant, presumed-male fish tends to grow largest and to colour up most intensely, while ripe females become noticeably rounder in the body, but the sexes are not reliably told apart by colour pattern or fin shape outside breeding condition.
Habitat
This is a fish of running water on the Guiana Shield and the adjacent Amapá coast, not of the wider Amazon lowlands. Its range covers the Oyapock (Oiapoque) River drainage — the river that forms much of the border between French Guiana and the Brazilian state of Amapá, plus its tributary the Camopi — and the Araguari River drainage, including the large Rio Amapari, entirely within Amapá. In the freshwater-ecoregion scheme of Abell et al. (2008), that places it across two units: the Guianas ecoregion (the Oyapock/Camopi) and the Amazonas Estuary & Coastal Drainages ecoregion (the Araguari/Amapari system). It is one of the characteristic cichlids recorded in dry-season inventories of Amapá's rivers.
Within these rivers Retroculus is a rapids specialist: FishBase notes the species is 'captured mainly in the rapids of rivers', and the genus as a whole lives almost exclusively in fast, broken, well-oxygenated stretches over rock, cobble and coarse gravel. The fullest field descriptions of this habitat come from the two river systems where the species has actually been collected. Von Drachenfels and Bitter, who took septentrionalis in the Oyapock at Saut Maripa, describe water so fast it was difficult to keep footing in the current, running over rocks thickly grown with Mourera fluviatilis — the rheophytic, current-adapted river-weed typical of Guiana Shield rapids — though calmer pockets in the lee of large boulders or tangled roots are also used, provided the bottom there is mainly soft material rather than bare rock. Stawikowski's independent collections from the Araguari upstream of Porto Grande describe a channel braided around small islands whose edges are mostly sand or gravel, only rarely muddy, with accumulations forming preferentially on the side of each island that faces into the current. These are clear, comparatively cool, shield-drainage waters, low in dissolved minerals and acidic to near-neutral — field and aquarium reports for the genus cluster around pH 6.0–7.0 and soft water. The fish hold station on the bottom among stones, perching and hopping from rock to rock rather than cruising; the high current and turbulence keep oxygen saturation high, which is the single non-negotiable feature of their environment. Depth in such rapid habitat is shallow, on the order of a metre or two over the rocky riffle zones the fish frequents, rather than the deeper pools. Water temperature in these tropical shield rivers runs broadly in the high 20s Celsius; captive maintenance of the genus has succeeded at roughly 81–84 °F.
Feeding
Retroculus species are benthic invertivores that work the spaces between stones in the current for small live prey, and septentrionalis is expected to feed the same way as its better-studied congeners. The defining dietary study is Moreira & Zuanon's (2002) analysis of Retroculus lapidifer in the Rio Araguaia, which found the diet dominated by the aquatic larvae of rheophilic insects — chironomids (midges), trichopterans (caddisflies) and ephemeropterans (mayflies) — exactly the immature insects that live clinging to rock surfaces in fast water. FishBase places the genus at a trophic level around 3.4, consistent with a specialist insectivore rather than a detritus-grazer or piscivore.
The subterminal, protrusible mouth and the lobed, well-rakered first gill arch suit a fish that picks and sifts small invertebrates from gravel and biofilm. In the wild community of a shield-river rapid, Retroculus functions as one of the bottom-foraging insectivores partitioning the rich invertebrate fauna of the riffle zone alongside loricariid catfishes and rheophilic characins. In captivity the genus readily takes substitute live and frozen foods of the same character — bloodworms (chironomid larvae) are a proven staple, supplemented with brine shrimp and prepared sinking foods — which is how aquarists have maintained xinguensis and lapidifer in condition.
Mating
Outside of breeding Retroculus live as loose aggregations on the riffle bottom, and a dominance hierarchy forms when several are kept together: in the documented captive accounts of the closely related Retroculus xinguensis, the largest and most intensely coloured individual — presumably a male — became dominant, and courtship began with reciprocal lateral displays and fin-flaring between the eventual pair. The species is a biparental substrate spawner with a genuine pair bond formed for the duration of a breeding cycle rather than a harem or mouthbrooding system.
Pair formation is followed by site preparation: the courting pair excavate a shallow pit in the substrate, the male and female taking turns at digging and display. In the xinguensis observations this pit-digging escalated over weeks before the first spawning, with an excavated depression roughly 8 in across scooped into the sand. Because the fish defend a small patch of bottom rather than a cave or burrow, the spawning site is exposed, and conspecifics and tankmates are a serious threat to the eggs — in captivity, rival fish repeatedly ate the eggs until the pair was isolated. Direct field observations of courtship in septentrionalis itself are scarce; the account here is drawn from the genus's well-documented reproductive behaviour, which is consistent across the studied species.
Breeding
Retroculus has one of the more distinctive spawning modes among Neotropical cichlids, and it is the genus, not just one species, that earns the description. The adhesive, ovoid eggs are laid directly into a shallow scooped depression in sand and become coated in grains; rather than guarding them in place, the parents gather up the sand-covered eggs and assemble them into a pile among small pebbles — the behaviour that gave the type species lapidifer ('stone-bearer') its name, after Castelnau watched a wild fish carrying pebbles one by one in its mouth to build a nest. The eggs fall into the interstices of the pebble pile, where the female hovers and fans them vigorously with her pectoral fins while the male patrols the perimeter; the pair may pick up and relocate the whole pile several times during incubation. At around 81–84 °F the eggs hatch roughly three and a half days after spawning. The free-swimming fry are tucked among the pebbles at night and shepherded out to forage by day, and in the documented xinguensis spawning the brood numbered on the order of 200–250 fry; the parents continued to guard them for about two weeks before the next breeding cycle began.
This pit-and-pebble strategy was long thought to make Retroculus effectively unspawnable in captivity — eggs were obtained on both sides of the Atlantic for decades but never hatched, because the missing ingredient was the loose pebble substrate the parents need to assemble and tend the nest. The first fully successful captive breeding of any Retroculus, Retroculus xinguensis by Lee Newman around 2006, turned precisely on adding handfuls of carefully sized small pebbles to the tank, after which the pair completed the cycle to free-swimming fry. Breeding of Retroculus septentrionalis specifically has not been documented in the aquarium literature — wild stock has rarely if ever reached the hobby — but on every line of evidence it should reproduce in the same pit-spawning, pebble-tending manner as its congeners.
In the aquarium
Retroculus septentrionalis is essentially not present in the aquarium hobby: its two central-Amazon congeners (lapidifer and xinguensis) have trickled in sporadically since the early 1990s, but septentrionalis comes from the remote Oyapock and Araguari and has practically never been collected for the trade — reaching it would, as one veteran cichlidophile put it, take a dedicated expedition. What follows is therefore the husbandry profile of the genus, which applies directly to this species; it is a fish for the experienced rheophile keeper, not a beginner.
The single make-or-break requirement is water that is clean, cool-tropical and intensely oxygenated, moving fast. These are rapids fish, and their gills are built for near-saturation oxygen; in still or waste-laden water they fade and die even when superficially 'settled' for months. Successful keepers run strong powerheads or a river-tank flow, heavy aeration and/or trickle/sump filtration, and change large volumes of water often (the documented xinguensis setup used roughly 75% changes every 7–10 days). Aim for around 81–84 °F, soft and acidic-to-neutral water (about pH 6.0–6.8), and obsessively low dissolved waste — high nitrate and organic load are the usual killers. The classic mistake is to treat them as ordinary 'eartheaters' in a low-flow community tank; that has cost a great many fish, including those of well-known aquarists.
Provide a large-footprint tank with open swimming-and-perching floor: fine sand as the base, smooth rounded stones and cobbles to perch and hop on, and — critically, for any breeding attempt — a quantity of small, manageable pebbles the fish can pick up and pile. The documented spawnings happened in tanks of 180–300 US gallons, and the species' size (to ~7.5 in SL) plus its need for current argue for a long tank rather than a tall one. Feed a varied invertebrate diet built around frozen bloodworms (chironomid larvae) with brine shrimp and quality sinking prepared foods. Tankmates should be other clean-water, fast-water species that won't outcompete a bottom-perching insectivore or harass it; other rheophilic Amazonians (rapids-dwelling loricariids, some Teleocichla and the larger eartheaters) have been co-housed with the genus, but note that any tankmate — including conspecifics outside a bonded pair — will eat an exposed spawn, so breeding requires the pair to be isolated.
Conservation
The IUCN Red List assesses Retroculus septentrionalis as Least Concern (assessor Renata Guimarães Frederico; date assessed 21 November 2020), reflecting a reasonably wide distribution across the Oyapock and Araguari drainages and no evidence of a population-level decline. It carries no CITES listing and is of negligible commercial fishing value (FishBase scores its fishing vulnerability as very low). Because it is barely collected for the ornamental trade — unlike its central-Amazon congeners, which were swept up in the pleco rush of the 1990s — aquarium harvest is not a meaningful pressure on this species.
That status should nonetheless be read against the specific vulnerability of a rheophilic, rapids-restricted fish. As a current-and-oxygen specialist, Retroculus is exactly the kind of cichlid most exposed to river impoundment: damming a rapid converts the broken, oxygen-rich riffle the fish depends on into still reservoir, eliminating its habitat outright. The Araguari basin in particular has been heavily dammed for hydropower, and proposed and existing dams on shield rivers across the Guianas and Amapá are the most plausible long-term threat to populations of this and other rapids-endemic fishes. Deforestation, gold-mining and associated sedimentation and mercury contamination are the broader regional pressures on these clearwater drainages. None of these is currently documented as driving a decline in Retroculus septentrionalis, but its narrow habitat preference means it warrants more attention than its present Least Concern listing might suggest.