### BITB57 — Three Brothers
**CRD page:** [Three Brothers/Central Chagos Coral Surveys (Mar 25–28, 2015)](https://coralreefdiagnostics.com/three-brotherscentral-chagos-20150325-to-0328)

**27 Mar 2015**
**GPS:** 6.15979°S, 71.5271°E
**Setting:** protected submerged channel reef (lagoon channel), depth not reported
**Coral cover:** 40.1% (4 formal transects, D1–D4; **algal cover 46.8%**) · **published image bleaching:** not reported
**Live hard coral cover across methods:** formal transects 40.1% · CoralNet points 32.3% · Coralscapes masks 34.1% (110 frames). These products have different sampling footprints and denominators and should not be averaged.
**Temperature:** 29.8 °C · **Salinity:** 34.2
**Photo review:** all 118 frames examined across the three folders — 51 from BiTb57a-great dive-finished, 19 from BiTb57b-finished, 48 from BiTb57c-finished
**Representative photos:** P3280478.jpg, P3280454.jpg, P3270417.jpg, P3280422.jpg, P3270480.jpg

BITB57 is a mid-cover channel reef inside a lagoon passage, carrying **40.1% coral against 46.8% algae** — the two together with 8.3% non-coral invertebrates and 4.7% bare substrate sum to exactly 100% of the recorded benthos. Of the algal fraction, crustose coralline algae is by far the largest single component at 22.65%, well ahead of turf (14.06%), turf+sediment (6.04%), macroalgae (2.09%), erect algae (1.97%) and cyanobacteria (0.04%) — this is a CCA-heavy reef, not a fleshy-macroalgae-heavy one.

Twenty-nine genera are non-zero, but the assemblage is carried by two. ***Acropora* holds 18.10% of the benthos and *Porites* 5.42%** — together well under half the coral, because this site's genus list has an unusually long, even tail: *Turbinaria* 4.44%, *Pocillopora* 2.01%, *Heliopora* 1.98%, *Montipora* 1.71%, *Favia* 0.83%, *Pavona* 0.65%, *Goniastrea* 0.64%, *Leptastrea* 0.53%, *Astreopora* 0.52%, *Gardineroseris* 0.50%, *Favites* 0.46%, *Stylophora* 0.46%, *Pachyseris* 0.37%, and fifteen further genera below 0.3% down to *Acanthastrea*, *Galaxea* and *Leptoseris* at 0.006%. The genus sum (40.1125%) matches the published site-level coral cover figure exactly, so no genus mass is missing from the headline number.

**The depth structure at BITB57 is a clean, monotonic gradient rather than a flip between shallow and deep zones, and it is the most important structural finding at this site.** Coral cover rises steadily with depth — D1 10.6%, D2 31.35%, D3 51.3%, D4 67.2% — while algae falls in near lockstep: 64%, 58.3%, 41.3%, 23.7%. D1, the shallowest transect, is thin and mixed: *Pocillopora* 3.20%, *Porites* 2.20%, *Acropora* 2.00%, *Montipora* 1.00%, *Stylophora* 1.00%, *Leptastrea* 0.40%, on a bed that is nearly two-thirds algae. D2 introduces *Heliopora* as a genuine codominant — 7.90%, ahead of *Acropora* at 5.55% and *Porites* at 4.35%, with *Favia* 2.70% and *Montipora* 2.52%. By D3, *Acropora* pulls decisively ahead at 27.60% against *Porites* 9.73%, *Montipora* 2.33%, *Goniastrea* 2.17%, *Pocillopora* 2.17% and *Astreopora* 1.50%. D4, the deepest and highest-cover transect, is a two-genus reef: **Acropora 37.27% and Turbinaria 17.73% together account for 55.00 of the 67.2% total coral cover — 82% of everything growing there** — with *Porites* a distant third at 5.40%, then *Pachyseris* 1.47%, *Montipora* 1.00% and *Hydnophora* 0.73%. *Turbinaria* is essentially absent from D1–D3 and appears only at depth, which matches the dramatic columnar/whorled formations photographed in the deeper frames of the third folder. The site-level figures of 18.10% *Acropora* and 4.44% *Turbinaria* describe no single transect: they average a thin, algae-dominated shallow flat against a high-cover, two-genus deep bank.

The photographs corroborate this gradient closely. The great majority of frames in the first folder (BiTb57a) are wide-angle vistas of near-continuous tabular and plating *Acropora*-type colonies running to the edge of visibility, tiled together with sea fans, large schools of fusiliers and rabbitfish working over the structure, and occasional massive tan-brown lobate domes consistent with *Porites* — the visual signature of the D3/D4 *Acropora* buildup. The third folder (BiTb57c) documents the deep *Turbinaria* component directly: several frames (including P3280452–455 and P3280484–486) show dramatic dark, columnar, whorled formations with brown-pigmented tops and deep maroon-red bases, standing in dense ranks across the reef — a growth form consistent with *Turbinaria* in its columnar/vase habit, not documented anywhere in the shallower first folder. The same folder also holds extensive continuous foliose banks (P3280402–406, P3280420–421) that read as a distinct plating community, and repeated massive tabular *Acropora* fields (P3280488–492) matching D4's dominance pattern. The second folder (BiTb57b) is almost entirely close-focus macro work rather than wide reef structure — small coral recruits on coralline pavement, individual colonies, and reef fauna — and adds little to the cover/structure picture but a great deal to the fauna record below.

**Condition is not addressed by the published data at this site — the `Mean(%bleaching)` field is blank for BITB57, and no site-level bleaching percentage can be quoted.** The photographs, however, are not silent on the question. Four frames across two different folders show unambiguous colony-scale bleaching: P3270386 (folder b) is a small branching colony that is uniformly stark white with no pigmented tissue; P3270412 (folder b) shows a white branching colony in the foreground beside a cluster of green Christmas-tree worms and otherwise pigmented coral; P3280422 (folder c) shows a bright white branching colony wedged among fully tan-pigmented neighbours; and P3280472 (folder c) shows a smaller white colony in a similarly mixed, mostly-pigmented patch. In every case the bleaching is confined to isolated colonies rather than the reef broadly — the wide-angle vistas throughout all three folders are overwhelmingly tan, brown, olive and purple, with no reef-scale whitening evident. This reads as scattered, low-level colony bleaching rather than a mass event, but since there is no published percentage for this site, the imagery is the only evidence available and should be treated as a qualitative flag, not a quantified rate.

Fauna is unusually well documented for a Chagos site. At least two giant clams (*Tridacna*, mantle visible in P3270410 and again near a columnar coral base in P3280478) appear in the imagery, along with two sea anemones hosting anemonefish (P3270416, P3270417 — consistent with the Indian Ocean's *Amphiprion nigripes*), green Christmas-tree worms (*Spirobranchus*-type, P3270412 and P3270415), an eagle ray in open water (P3280388), and a sea turtle visible in the background of two frames (P3280449, P3280450). Reef fish are abundant and diverse across all three folders: large sheltering schools of fusiliers and small planktivores over gorgonian fans (P3270413, P3270414), and closer to the substrate, butterflyfish, angelfish, pufferfish, unicornfish and surgeonfish threading through the coral heads in the third folder's mid-depth frames. Most consequentially, **P3280478 shows a diver holding a crown-of-thorns starfish (*Acanthaster* sp.)** collected from the reef — direct photographic evidence of COTS presence at this site, worth flagging given the genus's capacity for coral-tissue predation and outbreak-driven mortality.

<b>Assemblage composition.</b> Formal surveys give 40.1% coral against 46.8% algae (CCA-dominated, not macroalgae-dominated) and 8.3% invertebrates, with coral cover rising monotonically from 10.6% at the shallowest transect to 67.2% at the deepest as algae falls from 64% to 24%. *Acropora* (18.10% site-wide) and *Turbinaria* (4.44% site-wide) together supply 82% of all coral on the deepest transect alone, a pattern the photographs confirm directly through tabular *Acropora* fields and columnar *Turbinaria*-type banks concentrated in the deeper imagery. The shallow zone is thinner and more mixed, led by *Pocillopora* and *Heliopora* rather than *Acropora*. Fauna — giant clams, anemonefish, an eagle ray, a turtle, and a crown-of-thorns starfish collected by a diver — is richly documented in the close-focus second folder and scattered through the other two.

**Flags:** (1) **Crown-of-thorns starfish (COTS) observed.** P3280478 shows a diver holding an *Acanthaster*-type starfish collected from the reef — direct evidence of COTS presence at BITB57; no abundance or outbreak-scale information is available beyond this single individual. (2) **Published bleaching percentage is blank for this site** (`Mean(%bleaching)` not reported), yet four photographs across two folders (P3270386, P3270412, P3280422, P3280472) show unambiguous colony-scale white bleaching amid otherwise pigmented coral. Treat this as qualitative photographic evidence only — there is no published rate to compare it against or to supersede. (3) **Filename collision between subfolders a and b.** P3270457.jpg, P3270458.jpg and P3270459.jpg exist under identical names in both `BiTb57a-great dive-finished` and `BiTb57b-finished`, at different content (both are wide reef scenes but from different framing/passes) — confirmed genuinely different images from two cameras, not duplicates. Any tool keying on bare filename alone will conflate these; always carry the full relative path. (4) **Date discrepancy in folder c.** All 48 frames in `BiTb57c-finished` carry filenames dated 28 March (P3280xxx), one day after the site's published transect date of 27 March; the CRD page for this region spans dives from 25–28 March, so this likely reflects a second dive/day at the same site rather than a data error, but this has not been independently confirmed and the transect date is retained as published. (5) **Frame-count mismatch:** `chagos_multimethod_cover_by_site.csv` records `n_photos = 110`, while the three photo folders together hold 118 stills; the CoralNet (32.3%) and Coralscapes (34.1%) figures rest on 110 of the 118 available frames, not the full set. (6) **Reef depth is not reported** in the site-level record (blank `depth` field); only the four per-transect covers are available, not a single site depth. (7) **Point/CoralNet annotations are approximately 76% unconfirmed machine output** (only ~600 of 2,500 Chagos images have been human-confirmed by A. Mayfield); genus identifications quoted here from the transect data are treated as identified survey data, while any genus attribution made from a photograph alone is hedged as "consistent with." (8) **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`). Treat the site bleaching percentage as the weaker of the two published numbers, and never read it as a diver's field estimate.

**Benthic data:** [Download BITB57 benthic survey data (CSV)](../benthic_data/BITB57_benthic.csv) — scleractinian cover by genus with full names, plus substrate and algal categories, from the LOF BIOT-1 survey (Mayfield & Dempsey 2026, *Oceans*, supplemental file S1).