### BITB54 — Three Brothers
**CRD page:** [Three Brothers / Central Chagos Coral Surveys (25–28 Mar 2015)](https://coralreefdiagnostics.com/three-brotherscentral-chagos-20150325-to-0328)

**26 Mar 2015**
**GPS:** 6.16601°S, 71.52676°E
**Setting:** intermediate-exposure submerged fringing forereef, 19.2 m
**Coral cover:** 49.5% (5 formal transects, D1–D5; **algal cover 32.1%**) · **published image bleaching:** 4.3%
**Live hard coral cover across methods:** formal transects 49.5% · CoralNet points 21.4% · Coralscapes masks 20.3% (98 frames). These products have different sampling footprints and denominators and should not be averaged.
**Temperature:** 29.8 °C · **Salinity:** 34.3
**Photo review:** all 99 of 99 jpg stills examined (the folder also holds 3 unreviewed video clips — see Flags)
**Representative photos:** P3260489.jpg, P3260518.jpg, P3260559.jpg, P3260562.jpg, P3260515.jpg

BITB54 sits on a fringing forereef off South Brother at 19.2 m under intermediate wave exposure, and it carries above-average coral cover for the archipelago: **49.5% coral against 32.1% algae, ranking 27th of 109 Chagos sites for coral cover and in the lower quartile (83rd of 109) for algal cover.** The algal budget itself is unusual: crustose coralline algae (PCCATS) alone accounts for 19.5% of total benthos — 26th-highest of 109 sites — while fleshy turf (PTTS) sits at just 6.0%, 89th of 109 and among the lowest values recorded in the dataset. Bare substrate (10.0%) and non-coral invertebrates (8.3%) are both close to the site median. This reads as a hard, well-grazed reef where "algae" mostly means encrusting coralline pavement rather than turf or macroalgal growth, a pattern that holds at every depth transect (see below).

Coral cover is spread across an unusually large number of taxa: **34 nonzero genera — joint 2nd of 109 Chagos sites for genus richness**, behind only BIBL87's 36 and tied with BITB39, BISA93 and BIBL82 (the central ranks file records rank 2 for all four; an earlier draft of this shard said "third-highest", which mis-stated the tie), and the 34 genus percentages sum to 49.53% of the benthos — an exact match to the published cover figure. All values below are percent of total benthos, not percent of coral. *Acropora* leads at 15.93% of the benthos (32% of the coral present), followed by *Porites* 7.14% (14%), *Pocillopora* 5.52% (11%), *Favia* 3.41% (7%) and *Pachyseris* 3.38% (7%) — together these five genera account for 71% of all coral recorded. *Montipora* (2.33%) and *Pavona* (2.29%) follow, then *Millepora* (1.43%), *Platygyra* (1.40%), *Stylophora* (0.97%) and *Gardineroseris* (0.97%). A long tail of 23 further genera — *Favites*, *Symphyllia*, *Echinopora*, *Seriatopora*, *Montastrea*, *Leptastrea*, *Podabacia*, *Goniastrea*, *Astreopora*, *Echinophyllia*, *Plesiastrea*, *Psammocora*, *Fungia*, *Heliopora*, *Oxypora*, *Leptoseris*, *Cyphastrea*, *Cantharellus*, *Hydnophora*, *Alveopora*, *Galaxea*, *Acanthastrea* and *Lobophyllia* — each contributes under 1% of the benthos but together adds real diversity.

**The five depth transects tell a story of reef structure building to a peak and then collapsing at the deepest station, with the algal budget staying CCA-dominated throughout.** D1 (shallowest) carries the lowest coral cover of the profile at 30.6% against the site's highest algal cover, 41% — but 34.2 of those 41 points are crustose coralline algae and only 3.8 are turf, so even the shallow zone reads as coralline pavement rather than weedy ground. *Acropora* already leads here at 15.20% (half the coral present), with *Pocillopora* 5.40%, *Millepora* 4.40%, *Porites* 2.00%, *Montipora* 1.40% and *Favites* 1.00%. Cover climbs through D2 (53.3% coral, 40% algae; *Acropora* 17.87%, *Porites* 7.33%, *Favia* 6.87%, *Pocillopora* 4.97%, *Stylophora* 2.83%, *Montipora* 2.63%) and D3 (63.0% coral, 29.4% algae; *Acropora* 16.57%, *Porites* 9.75%, *Pocillopora* 6.22%, *Favia* 5.97%, *Platygyra* 3.92%, *Pachyseris* 2.70%) to peak at D4, **63.8% coral cover with *Acropora* alone at 30.0% — 47% of all the coral on that transect** — plus *Porites* 10.60%, *Pocillopora* 8.00%, *Gardineroseris* 3.00%, *Symphyllia* 2.20% and *Montipora* 2.00%. Then D5, the deepest station, **drops sharply to 37.0% coral, and *Acropora* falls out of the leading ranks entirely**: *Pachyseris* takes over at 13.00% (35% of the coral present), with *Pavona* 6.00%, *Porites* 6.00%, *Favia* 3.00%, *Montipora* 3.00% and *Pocillopora* 3.00%. D5 also carries the lowest algal cover of the profile (22%, none of it recorded as turf), yet coral and algae together only account for 59% of the benthos there — roughly 41% is unaccounted for by this depth-level summary (likely bare pavement, sand or non-coral invertebrate cover not itemized by depth), versus just 6–8% unaccounted for at D2–D4. **The site-level figures of 15.93% *Acropora* and 3.38% *Pachyseris* describe no single transect**: they average a shallow-to-mid reef growing increasingly *Acropora*-dominated with depth up to D4, against a deepest station where *Acropora* nearly disappears and *Pachyseris* becomes dominant instead.

The photographs — all 99 stills in the folder — corroborate a structurally complex, high-relief forereef: broken pavement and low ridges alternating with sand/rubble channels, coral cover ranging from near-continuous thickets (e.g., P3260528, P3260538, P3260541–542) to more open, patchy ground (P3260536, P3260539) consistent with the cover swings documented across the five transects. Growth-form diversity in the imagery matches the genus list: branching and corymbose colonies consistent with *Acropora* dominate many frames, tabular forms appear on the upper slope, massive/boulder colonies consistent with *Porites* or faviid genera sit beside brain-pattern colonies consistent with *Platygyra* or *Favia*, and large tan-olive foliose plates consistent with *Pachyseris* or *Pavona* recur through the set — P3260562, a close-up of a large plate colony with fine surface ridging, is a plausible photographic counterpart to the deep-transect *Pachyseris* dominance at D5. Purple-pink crustose coralline algae is visible encrusting rock and dead framework in numerous frames, consistent with the substrate data showing CCA as the dominant algal category at every depth. Water clarity is very good throughout (deep blue backscatter, strong strobe-lit foreground color), and fish life is a conspicuous, near-constant element: dense schooling fusiliers and small silvery baitfish form a visible "curtain" over the reef in roughly a third of the frames (e.g., P3260559–560, P3260564–575), at times all but obscuring the underlying structure.

**Condition reads as good across both the transect and photographic evidence.** The published image-based bleaching figure is **4.3%, ranking 75th of 98 Chagos sites with a value — in the lower quartile of bleaching severity** — and the 99-frame review found no colony with unambiguous, colony-scale tissue pallor. Two frames carry a caveat: P3260515 and P3260512 each show a large, pale, rounded coral/sponge structure notably lighter than surrounding pigmented colonies; this is consistent with either partial bleaching or an exposed dead-skeleton surface and cannot be resolved from a single still, so it is flagged rather than read as active bleaching. No disease lesions, tissue-loss margins, Drupella aggregations, parrotfish bite scarring, or COTS individuals/feeding scars were observed anywhere in the set — a clean bill of health consistent with the low published bleaching figure.

Fauna in the imagery is abundant and varied. Two whitetip reef shark encounters are documented — a single individual photographed resting under a coral ledge across three consecutive frames (P3260488–490) and a shark cruising past a bommie in P3260484 and P3260486. Reef fish include repeated emperor angelfish sightings (P3260477, P3260495), several moorish idols (P3260481, P3260513, P3260517–518), powder-blue surgeonfish (P3260544, P3260547), a probable sea cucumber on open substrate (P3260502, P3260508), and a mottled, camouflaged animal on rubble in P3260543 consistent with a scorpionfish- or stonefish-type ambush predator (hedged, not confirmed). Diver and camera-rig presence is visible in a number of frames (e.g., P3260464, P3260467, P3260498, P3260508–511, P3260514, P3260520, P3260553–554, P3260559–560), and a clear survey quadrat appears explicitly in P3260500 — standard field documentation of the survey rather than a data concern.

<b>Assemblage composition.</b> Formal transects give 49.5% coral against 32.1% algae (itself mostly crustose coralline rather than turf) across 34 nonzero genera, led by *Acropora* (15.9%), *Porites* (7.1%), *Pocillopora* (5.5%), *Favia* (3.4%) and *Pachyseris* (3.4%). The five-transect depth profile shows coral cover building from 30.6% at the shallowest station to a 63.8%, *Acropora*-dominated peak at D4, then collapsing to 37.0% at the deepest station where *Pachyseris* replaces *Acropora* as the leading genus. The photographs document this same structural range — branching/tabular *Acropora*-type thickets, massive and brain-pattern colonies, and large foliose plates consistent with *Pachyseris*/*Pavona* — plus abundant reef fish, two whitetip shark sightings, and no clear evidence of bleaching, disease, predation or COTS across all 99 stills reviewed.

**Flags:** (1) **Multi-method deviation is unusually large here.** CoralNet points (21.4%) and Coralscapes masks (20.3%) sit 28–29 percentage points below the formal-transect figure (49.5%, deviation range −29.2 to −28.1 pp) — a bigger CV-vs-transect gap than most Chagos sites reviewed to date. Do not average the three figures; the transect value is treated as ground truth for cover. (2) **Frame-count mismatch:** 99 jpg stills were reviewed in the photo folder, but `chagos_multimethod_cover_by_site.csv` records `n_photos=98` for the CV comparison — a one-frame discrepancy of unclear origin; the CoralNet/Coralscapes figures above rest on that 98-frame subset, not the full 99. (3) **The CoralNet points figure is a CV product, not verified ground truth.** Point-annotation data corpus-wide is roughly 76% unconfirmed machine output; the 21.4% figure should not be read as human-confirmed. (4) **Two frames (P3260515, P3260512) show a pale, whitish coral/sponge structure** that could indicate partial bleaching or simply an exposed dead-skeleton surface; ambiguous from a single still and not resolved here. No other frame showed comparable pallor. (5) **No COTS, disease, or predation scarring was observed** across the full 99-frame review — an explicit absence-of-evidence finding, not merely an omission. (6) **D5, the deepest transect, has roughly 41% of its benthos unaccounted for by coral or algae** in the site-depth summary (vs. 6–8% at D2–D4), likely bare pavement, sand or non-coral invertebrate cover not itemized by depth in this dataset. (7) **Genus richness is exceptionally high: 34 nonzero genera, 3rd of 109 Chagos sites** (behind only BIBL87's 36) — a site strength distinct from cover. (8) **The folder also holds three unreviewed video clips** (P3260551.MOV, P3260558.MOV, P3260572.MOV), outside the scope of this still-photo review. (9) **Diver, camera rig, and a clear survey quadrat are visible in numerous frames** (e.g., P3260464, P3260467, P3260498, P3260500, P3260508–511, P3260514, P3260520, P3260553–554, P3260559–560) — standard field documentation, not a data issue. (10) **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. (11) **NON-CANONICAL BLEACHING RANK — added 2026-08-26.** Any "of 98" bleaching rank above was computed off the site table's `Mean(%bleaching)` column, which equals the **1st** CoralNet training run. Anderson's decision of 2026-08-26 (HOW_TO rule 4e) makes **`% bleaching (2nd training)` canonical**, The file carries TWO pools for it, and they are not interchangeable: a **110-row pool** ranking every dive (revisit dives included, so BIPB70B/72B/69C compete) and a **99-site pool** ranking one value per site. The rank quoted below is on the 110-row dive pool. For this site the canonical value is **4.30%, rank 83 of 110** (1st training: 4.30%, rank 83 of 110). The two runs agree to <0.05 pp on 100 of 110 rows and diverge only on the bleaching-onset dives, so for most sites the distinction changes the rank but not the number.

**Benthic data:** [Download BITB54 benthic survey data (CSV)](../benthic_data/BITB54_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).