### BITB46 — Three Brothers
**CRD page:** [Three Brothers/Central Chagos Coral Surveys BITB38-BITB49 (Mar 2015)](https://coralreefdiagnostics.com/three-brotherscentral-chagos-20150321-to-0324)

**23 Mar 2015**
**GPS:** 6.09610°S, 71.51430°E
**Setting:** protected lagoonal submerged bank reef, 27 m (inside the lagoon; `exposure-original` "leeward", `exposure-revised` "protected")
**Coral cover:** 25.8% (11 formal transects, D3–D4; **algal cover 55.3%**) · **published image bleaching (`Mean(%bleaching)` column):** 2.02%
**Live hard coral cover across methods:** formal transects 25.8% · CoralNet points 18.8% · Coralscapes masks 16.2% (**scored on 11 frames, not the 29 in the photo folder** — see Flags). A fourth figure, the published image-cover product, gives 27.5%. These products have different sampling footprints and denominators and should not be averaged.
**Temperature:** 30.0 °C · **Salinity:** 34.4
**Photo review:** all 29 of 29 jpg stills examined (camera timestamps 15:21:56–16:00:47, a single continuous dive; no sub-dive folders exist for this site)
**Representative photos:** P3230401.jpg, P3230410.jpg, P3230443.jpg, P3230447.jpg, P3230451.jpg

BITB46 is an algae-dominated, patchy bank reef inside the Three Brothers lagoon at 27 m, and it is the low-cover site of this group: **25.8% coral against 55.3% algae — 84th of 109 Chagos sites for coral cover and 13th of 109 for algal cover**, so it sits in the bottom quartile for coral and the top eighth for algae. **But "algae-dominated" here means something very specific, and the photographs make it unmistakable: crustose coralline algae alone is 29.1% of the total benthos — the 4th-highest CCA value of all 109 Chagos sites.** Turf adds 12.6% (33rd of 109), sediment-laden turf 7.8%, macroalgae 5.7%, erect algae 0.11% and cyanobacteria 0.00%. So more than half the algal budget is hard pink-mauve coralline pavement, not fleshy weed, and frame after frame shows exactly that: a continuous, well-grazed pink-and-lavender crust binding rubble and dead framework. Bare substrate is 6.3% and non-coral invertebrates 12.6%. This is a low-coral reef, but it is not a degraded, weedy one.

All genus values below are **percent of total benthos, not percent of coral**, and the 22 nonzero genera sum to 25.817% — an exact match to the published cover figure. ***Porites* dominates at 11.35% of the benthos, which is 44% of every coral point recorded at the site** — nearly three times the next genus. *Acropora* follows at 3.88% (15%) and *Pocillopora* at 3.83% (15%), then *Montipora* 2.09%, *Favia* 0.93%, *Favites* 0.74% and *Pavona* 0.66%. Everything else is under 0.4%: *Goniastrea*, *Leptastrea*, *Platygyra*, *Stylophora*, *Psammocora*, *Astreopora*, *Echinopora*, *Montastrea*, *Leptoseris*, *Cyphastrea*, *Acanthastrea*, *Ctenella*, *Gardineroseris*, *Symphyllia* and *Cantharellus*. **Genus count is 22, 69th of 109 Chagos sites** — the lowest of the four sites in this rebuild and below the archipelago midpoint, though the assemblage is still moderately even for its size: Shannon H′ = 1.89, evenness J′ = 0.61 and an effective diversity of 6.64 genera. The picture is a *Porites*-carried reef with a modest supporting cast.

**The two depth strata differ more sharply than anywhere else in this group, and the difference is driven almost entirely by *Acropora* falling out of the assemblage.** D3 (n = 9 transects) gives 35.13% coral against 49.64% algae across 19 genera: *Porites* 13.20%, *Acropora* 7.36%, *Pocillopora* 5.36%, *Montipora* 2.78%, *Favia* 1.36% and *Favites* 0.98%. D4 (n = 2) gives **16.50% coral against 60.90% algae across 14 genera — less than half the coral cover of the shallower stratum** — and there *Porites* holds 9.50%, which is 58% of all coral present, ahead of *Pocillopora* 2.30%, *Montipora* 1.40%, *Pavona* 0.60%, *Favia* 0.50% and *Favites* 0.50%. ***Acropora* collapses from 7.36% at D3 to 0.40% at D4, a 95% drop, while *Porites* falls only 28%*** — so the deep stratum is not a scaled-down version of the shallow one, it is a *Porites*-monoculture-leaning assemblage with the acroporids removed. Bare substrate roughly doubles (3.8% → 8.8%) and non-coral invertebrates rise slightly (11.4% → 13.8%). **All of this rests on two transects, and that is the central caveat of the site** (see Flags): the D4 figure is arresting, but it is descriptive, not a demonstrated depth effect. Individual transects across both strata range from 12.8% to 60.0% coral, so within-site patchiness is far larger than the stratum difference.

The 29 photographs corroborate the transect picture closely. **The dominant visual signature of this dive is pink-mauve crustose coralline pavement** — P3230399, P3230437, P3230444, P3230445, P3230446, P3230447 and P3230448 are all substantially CCA-crusted rock and rubble with scattered colonies rising from it, the photographic counterpart to the 4th-highest CCA value in the archipelago. Coral in the imagery is exactly as the transects describe it: **large yellow-gold tabular colonies consistent with *Acropora* recur throughout the set** (P3230410, P3230411, P3230419, P3230420, P3230434, P3230441, P3230446, P3230449) — often the largest structures in frame — alongside massive and encrusting olive-green colonies consistent with *Porites*, brown scroll-shaped *Turbinaria*-type cups, corymbose *Acropora*, small *Pocillopora* clumps, meandroid brain colonies consistent with *Platygyra* (a fine mauve-and-cream one is the subject of P3230418) and foliose *Echinopora*/*Montipora* plates. Orange-red encrusting sponges are conspicuous in P3230400, P3230448 and P3230449. **P3230418 also shows a dark maroon, finely branched, articulated organism sprawling across the CCA pavement — consistent with a crinoid (feather star) or with a fleshy red alga; the two cannot be separated from a single still**, and the distinction matters because one belongs in the 12.6% non-coral invertebrate class and the other in the macroalgal budget.

**Condition is good, and this is the least-bleached of the four sites rebuilt here.** `Mean(%bleaching)` is **2.02%, 89th of the 98 Chagos sites carrying a value — deep in the low-severity tail of the archipelago**, and the two CoralNet training runs agree (2.02 vs 2.0202), so this is not one of the ten sites where the runs diverge. The bleaching estimate is again denominator-dependent: the published product gives 2.02%; the human point audit found **two bleached points, which is 0.61% of all 330 annotated points or 3.13% of the 64 coral points**; the Coralscapes masks give an average 0.001% of image area. **No broad acute bleaching field is visible in any of the 29 frames.** Two features deserve naming. **P3230451 is the site's set-piece: a very large tabular colony with sharply bounded partial mortality across roughly the right-hand 40% of the plate — the dead area is pale mauve-grey, already colonized by crustose coralline algae and filamentous turf, with a clean margin against the intact olive-brown tissue and a small fish sheltering in a hole in the dead zone.** The CCA colonization dates the lesion: it is far more consistent with an older lesion or a breakage scar than with uniform recent bleaching, though its cause and timing remain unresolved. Similar dead-and-colonized surfaces are visible on the large tables in P3230410 and P3230411. Second, **P3230449 contains a strikingly pale, near-white lavender branching colony in the foreground**, notably lighter than every colony around it — unresolvable from a single still between paling, a naturally pale species and an already-dead-and-CCA-crusted skeleton, and flagged rather than counted. **No COTS individuals or feeding scars, no disease front, no *Drupella* aggregation and no clear corallivore scarring were observed anywhere in the 29-frame review.**

Fauna at this site is exceptional and the previous record needs one correction. **Two shark species are present, not one.** P3230400 and P3230401 show a **grey reef shark (*Carcharhinus amblyrhynchos*)** cruising close over the reef — identified on the plain, un-tipped first dorsal, the broad black posterior margin of the caudal fin and the blunt rounded snout. P3230409, P3230410, P3230411, P3230433 and P3230434 show a different animal entirely, a **whitetip reef shark (*Triaenodon obesus*)** — slender, broad-headed, cruising just above the substrate and passing beneath the large tables, with the white-tipped first dorsal and caudal clearly visible. Schooling fish dominate the middle of the dive and sort into three distinct groups: **dense dark surgeonfish (Acanthuridae) in P3230420, P3230433, P3230434, P3230435 and P3230436**, several dozen strong and repeatedly grazing over the large tabular colonies; a **massive silverside/baitfish cloud filling the water column in P3230438**; and **blue-and-silver fusiliers (Caesionidae, *Pterocaesio*-type with yellow-tailed *Caesio* above them) in P3230439, P3230440, P3230441 and P3230442.** Beyond the schools: **a large giant clam (*Tridacna* sp.) with a dark olive, intricately patterned mantle in P3230443**; a spiny lobster sheltering under a ledge in P3230450, identified from its long pale antennae (the body is in shadow, so the painted spiny lobster *Panulirus versicolor* identification previously logged for this frame is plausible but not confirmable); parrotfish in P3230447 and P3230448; and a pale, finely wrinkled animal wedged in a crevice in P3230427 — logged previously as a sea cucumber, though an octopus in its den is an equally defensible read of the same frame and the two cannot be separated here.

<b>Assemblage composition.</b> Formal transects give 25.8% coral, 55.3% algae, 6.3% bare substrate and 12.6% non-coral invertebrates across **22 nonzero genera (69th of 109 Chagos sites)**, with *Porites* alone holding 11.35% of the benthos — 44% of all coral present. **The site's real distinction is its algal composition, not its algal quantity: 29.1% crustose coralline algae, 4th-highest of 109 Chagos sites**, and the imagery confirms a hard, well-grazed pink-mauve coralline pavement rather than a weedy reef. Cover more than halves between D3 (35.1%) and D4 (16.5%) as *Acropora* falls 95% and *Porites* comes to hold 58% of the deep stratum's coral — but D4 rests on two transects. The 29 photographs document scattered massive, branching and large tabular colonies rising from that coralline pavement, an unusually abundant fauna (a grey reef shark, a whitetip reef shark, surgeonfish and fusilier schools, a baitfish cloud, a giant clam, a spiny lobster), one large table carrying sharply bounded, algae-colonized partial mortality, and no acute bleaching field, disease or COTS anywhere in the set.

**Flags:** (1) **D4 rests on TWO transects yet carries half the headline cover figure**, because the site value is the unweighted mean of the two stratum means (35.13% and 16.50%). Every D4 statement above — 16.5% cover, 60.9% algae, 14 genera, *Acropora* at 0.40%, *Porites* at 58% of coral — rests on `BITB46_d4_T1` (12.8% coral) and `BITB46_d4_T2` (20.2%). The 18.6 pp shallow-to-deep drop is real in the data but is the thinnest-supported number in this shard, and the site cover of 25.8% would move several points under any other weighting. (2) **Frame-count mismatch: the photo folder holds 29 frames, but `chagos_multimethod_cover_by_site.csv` records `n_photos=11`.** The CoralNet (18.8%), Coralscapes (16.2%) and published-image (27.5%) figures rest on that scored 11-frame subset, and the 330-point human audit is 11 frames × 30 points. **The 29-frame count in the photo-review line above must never be printed beside the CV cover figures.** (3) **Several reviewed frames contain almost no benthos** — P3230435 is surgeonfish silhouettes against the surface and P3230438 is a baitfish cloud in open blue water — so if either fell inside the scored subset it would depress the CV cover figures without saying anything about the reef. Which 11 of the 29 were scored is not recorded here and should be checked before the CV numbers are used quantitatively. (4) **The transects mix 100-point and 500-point totals** and are compared through percentage fields; note that both D4 transects are 500-point and seven of the nine D3 transects are 100-point, so the protocol difference is partly confounded with the depth difference at this site. (5) **Bleaching is quoted four ways with four denominators** — 2.02% published, 0.61% of all 330 audited points, 3.13% of the 64 coral points, and 0.001% of mask area. **Two bleached points underpin the human audit.** (6) **P3230451's partial mortality is an older lesion, not fresh bleaching**: the dead ~40% of the large table is already colonized by crustose coralline algae and turf, with a sharp margin against live tissue. Cause and timing are unresolved. Comparable dead-and-colonized surfaces appear on the tables in P3230410 and P3230411. (7) **P3230449 contains one near-white branching colony** far paler than its neighbours — flagged, not counted; paling, a naturally pale species and a CCA-crusted skeleton cannot be separated from a single still. (8) **Species-identification corrections and open questions from this pass:** the sharks are **two species** — a grey reef shark in P3230400–401 and a whitetip reef shark in P3230409–411/433–434 — not one whitetip throughout; the schooling fish sort into surgeonfish (P3230420, 433–437), a baitfish cloud (P3230438) and fusiliers (P3230439–442), not one "fusilier-like" school; **the animal in P3230427 is ambiguous between a sea cucumber and an octopus in its den**; **the maroon branched organism in P3230418 is ambiguous between a crinoid and a fleshy red alga** (invertebrate vs macroalgal budget); and the lobster in P3230450 is identified only from its antennae. These four are worth a second opinion. (9) **No COTS, disease front or corallivore scarring was observed** across the full 29-frame review — an explicit absence-of-evidence finding. (10) **No D1–D2 transects exist at this site** and the photographs carry no depth assignments, so the shallow reef is unsampled and the imagery cannot be attributed to a stratum. (11) **The published bleaching figure is a COMPUTER-VISION product, not an in-situ observation.** It comes from CoralNet annotation of the informal survey images (Mayfield & Dempsey 2026, *Oceans*). Per-image human-vs-model agreement is substantially weaker for bleaching than for cover: r = 0.40 for bleached against r = 0.72 for live hard coral (`CHAGOS_MULTIMETHOD_README.md`). The site table's `Mean(%bleaching)` column is not a mean — it equals the 1st CoralNet training run — so any rank taken off it is approximate; it also carries at least one known bad value elsewhere in the archipelago (BITB58 listed at 22.22% against a corrected 0.77%). Never read this number as a diver's field estimate.

**Benthic data:** [Download BITB46 benthic survey data (CSV)](../benthic_data/BITB46_benthic.csv) — scleractinian cover by genus with full names, plus substrate and algal categories (bare substrate, CCA, turf, macroalgae, invertebrates), from the LOF BIOT-1 survey (Mayfield & Dempsey 2026, *Oceans*, supplemental file S1).