### BITB42 — Three Brothers
**CRD page:** [Three Brothers/Central Chagos Coral Surveys BITB38-BITB49 (Mar 2015)](https://coralreefdiagnostics.com/three-brotherscentral-chagos-20150321-to-0324)

**22 Mar 2015**
**GPS:** 6.05761°S, 71.52129°E
**Setting:** intermediate-exposure submerged bank forereef, outside the lagoon, 25 m
**Coral cover:** 36.2% (16 formal transects, D2–D4; **algal cover 33.3%**) · **published image bleaching:** 7.7% (`Mean(%bleaching)` column)
**Live hard coral cover across methods:** formal transects 36.2% · CoralNet points 23.3% · Coralscapes masks 22.4% (46 frames). These products have different sampling footprints and denominators and should not be averaged.
**Temperature:** 29.7 °C · **Salinity:** 34.7
**Photo review:** 30 of the 51 frames in `BiTb42-finished`, a systematic sample spanning the whole dive sequence — P3220426, 428, 429, 430, 437, 444, 446, 448, 456, 458, 460, 462, 465, 467, 469, 470, 471, 473, 476, 478, 480, 481, 482, 483, 484, 486, 488, 490, 492, 494. The 21 unexamined frames are interleaved between reviewed frames from the start of the sequence to the end, so no stretch of the dive is unrepresented.
**Representative photos:** P3220481.jpg, P3220456.jpg, P3220471.jpg, P3220465.jpg, P3220448.jpg, P3220473.jpg

BITB42 is a submerged bank forereef at 25 m off Three Brothers, outside the lagoon under intermediate exposure, and it sits close to the archipelago middle on both headline numbers: **36.2% coral against 33.3% algae, ranking 63rd of 109 Chagos sites for coral cover and 78th of 109 for algal cover.** The composition of the algal budget is where it departs from its neighbours. **Turf-with-sediment (PTSTS) alone accounts for 14.8% of total benthos — the single largest algal component here, larger than crustose coralline algae** — while CCA (PCCATS) is only 8.6%, 89th of 109 sites and firmly in the lowest quartile, and clean fleshy turf (PTTS) is 5.4%, 91st of 109. Macroalgae (2.1%), erect algae (2.4%) and cyanobacteria (0.03%) are trivial. Bare substrate takes 13.8% and non-coral invertebrates 16.7%; coral, algae, bare rock and invertebrates sum to exactly 100%. **This is a sediment-loaded reef, not a coralline-pavement one** — a real contrast with high-CCA Chagos sites — and the photographs bear it out, with dead framework carrying a fine sediment-and-turf film in almost every close frame (P3220437, P3220467, P3220483) and open sand ground occupying whole frames elsewhere (P3220429, P3220462, P3220476).

Coral cover is spread across **26 nonzero genera, 45th of 109 Chagos sites**, and the 26 genus percentages sum to 36.21% 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 12.98% of the benthos — 36% of all the coral present** — followed by *Porites* 9.48% (26%) and *Montipora* 4.31% (12%); those three hold 74% of the coral at the site. The next tier is thin and closely packed: *Pocillopora* 1.37%, *Leptastrea* 1.15%, *Favia* 1.14%, *Pavona* 1.08%, *Goniastrea* 0.81%, *Favites* 0.72%, *Pachyseris* 0.53% and *Stylophora* 0.50%. Fifteen further genera — *Astreopora*, *Symphyllia*, *Platygyra*, *Montastrea*, *Lobophyllia*, *Cantharellus*, *Cyphastrea*, *Acanthastrea*, *Galaxea*, *Oxypora*, *Echinopora*, *Millepora*, *Gardineroseris*, *Leptoseris* and *Psammocora* — each contribute under 0.4%.

**The depth profile shows coral cover peaking at mid-depth and *Acropora* dominance collapsing below it, while the substrate turns to bare rock.** The shallowest stratum, D2, carries the **lowest coral cover of the profile at 29.0% against the profile's highest algal cover, 45%** — and 27 of those 45 points are turf-with-sediment, so the shallow zone is the most sediment-laden ground at the site. It is also the poorest in taxa: **just four genera recorded, *Acropora* 13.00%, *Porites* 12.00%, *Montipora* 3.00% and *Stylophora* 1.00%.** Cover then peaks at D3, **42.0% coral with *Acropora* at 19.18% — 46% of all coral on that stratum** — with *Porites* 10.09%, *Montipora* 4.89% and *Pocillopora* 2.27% behind it, across 21 genera and with algal cover down to 30.5%. At D4, the deepest stratum, **cover falls back to 37.6% and *Acropora* collapses to 6.75%, barely ahead of *Porites* at 6.35% and *Montipora* at 5.05%** — 18% of coral rather than 46% — while the mid-order genera all rise (*Leptastrea* 0.76→2.70%, *Favia* 0.93→2.50%, *Pavona* 0.80→2.45%, *Goniastrea* 0.38→2.05%, *Pachyseris* absent→1.60%) and genus count reaches its maximum at 23. **The substrate inverts across the same profile: bare substrate climbs 6.0% → 10.0% → 25.4% with depth while algal cover falls 45% → 30.5% → 24.5% and non-coral invertebrates fall 20% → 17.5% → 12.6%.** So the deep station is the most taxonomically even and the least algal, but a quarter of it is bare rock. The site-level figure of 12.98% *Acropora* describes no single stratum; it averages an *Acropora*-and-*Porites* duopoly shallow, an *Acropora*-dominated mid-depth peak, and an even, mixed deep zone where a quarter of the benthos is bare rock.

Individual transects span **23% to 65% coral**, so within-depth patchiness comfortably exceeds the separation between depth means, and the 30 photographs document precisely that range. At the open end, P3220429 is a **near-featureless white sand plain** with only scattered rubble clumps, P3220462 and P3220476 are sand-dominated with isolated colonies (P3220476 a single healthy olive staghorn *Acropora* standing alone on sand), and P3220426, P3220437 and P3220444 are low-relief pavement with wide bare and turf-covered lanes. At the closed end, P3220471, P3220456, P3220488 and P3220490 are near-continuous coral: overlapping tabular and corymbose *Acropora*, brown foliose and vase-form plates, knobby digitate *Porites*, encrusting yellow-green sheets and pink and purple corymbose colonies packed edge to edge. P3220473 is the single best structural frame — **one enormous mustard-green tabular colony spanning the whole frame, undercut by white sand** — and P3220480 and P3220486 show large flat *Acropora* tables with shoaling fish sheltering beneath. Growth-form composition across the set matches the transect ranking well: tables and corymbose *Acropora* are the visual signature, massive and encrusting *Porites* the second element, and foliose grey-purple ridged plates consistent with *Pachyseris* or *Merulina* recur as a third. Diver, camera rig and a descending survey line are visible in several frames (P3220429, P3220430, P3220446, P3220460) — standard field documentation, not a data concern.

**Condition is localized rather than uniform, and the imagery shows both very recent and much older mortality.** The `Mean(%bleaching)` column records **7.7% for BITB42, 44th of 98 Chagos sites with a value** — mid-range severity — and both CoralNet training runs return 7.73/7.73, so no run-selection divergence applies here. **P3220481 is the definitive frame: a corymbose *Acropora* colony whose left two-thirds is normally tan-pigmented with pale growing tips, while a sharply demarcated right-hand third is entirely bone-white to the branch base, the skeleton still clean and unfouled.** That sharpness and cleanliness is consistent with either severe localized bleaching or very recent tissue loss, and a single still cannot separate the two. **P3220456 carries an independent second instance** not previously recorded — a conspicuous bone-white tabular/corymbose colony in an otherwise fully pigmented field, again with intact, unfouled structure. Against these, **P3220483 shows the opposite end of the mortality timeline**: a macro of dead framework thickly encrusted with pink coralline algae, turf and sediment, with one small patch of paler, more recently exposed skeleton beside living fine-polyped *Porites* — old loss that has long since been overgrown. A partly dead-topped hemispherical colony in P3220467 belongs to the same older category. Several pale-cream massive colonies (P3220478, P3220484, P3220430) are ambiguous between natural pale pigmentation and paling and are not counted. **Most colonies in most frames are normally pigmented, and the 30-frame review does not show a coherent reef-wide bleaching field.** **No COTS, disease lesion, tissue-loss margin, *Drupella* aggregation or corallivore bite scarring was identified in any reviewed frame** — but note that **P3220492 contains a black long-spined sea urchin (*Diadema*-type) on open pavement**, an animal an automated classifier could plausibly confuse with *Acanthaster*; it is an echinoid, not a crown-of-thorns.

Fauna is abundant and includes several large-bodied animals. **Grey reef sharks (*Carcharhinus amblyrhynchos*-type, identification kept conservative) appear in four reviewed frames — P3220426, P3220460, P3220470 and P3220494** — always in the blue over deeper, hazier ground, and P3220470 pairs a shark with a loose school of about a dozen unicornfish (*Naso*-type). **Large dark, white-spotted groupers (camouflage- or marbled-grouper type, *Epinephelus* sp.) are photographed twice: one cruising mid-water in P3220446 with a diver silhouetted behind, and two individuals resting side by side in close contact on the bottom in P3220448** — a pairing worth noting given the March timing, though no spawning behaviour can be inferred from a still. A large dark barred fish, snapper- or surgeon-type, crosses P3220458. **P3220465 is a close frame of a large, fully intact gorgonian sea fan** with a second fan behind it, healthy and unfouled; gorgonians also appear in the background of P3220448, P3220480 and P3220469. An orange-and-white anemonefish sits at centre-left of P3220480, implying a host anemone in frame. P3220486 documents a mixed reef-fish assemblage — fusiliers, chromis, a yellow longnose butterflyfish (*Forcipiger*-type) and several butterflyfish — sheltering under and around a large *Acropora* table beside a soft coral consistent with *Sarcophyton*. Moorish idols, butterflyfish, parrotfish and damselfish recur throughout.

<b>Assemblage composition.</b> Formal transects give 36.2% coral, 33.3% algae, 13.8% bare substrate and 16.7% non-coral invertebrates across 26 nonzero genera, led by *Acropora* (13.0%), *Porites* (9.5%) and *Montipora* (4.3%), which together hold three-quarters of the coral. The algal budget is turf-with-sediment dominated (14.8%) rather than CCA dominated (8.6%, lowest quartile). The three-stratum depth profile runs from a four-genus, 45%-algae shallow station at 29.0% coral, through a 42.0%, *Acropora*-dominated mid-depth peak, to a 23-genus deep station at 37.6% where *Acropora* has fallen to a third of its mid-depth share and a quarter of the benthos is bare rock. The photographs corroborate the full 23–65% transect range — bare sand plains at one extreme, near-continuous tabular *Acropora* and plate gardens at the other — plus four grey reef shark encounters, a grouper pair, an intact gorgonian sea fan, and two frames of sharply demarcated bone-white coral against much older, fully overgrown dead framework elsewhere.

**Flags:** (1) **The D2 stratum is a SINGLE transect, and it carries a third of the headline cover figure.** Site cover is the unweighted mean of the three stratum means (29.0, 42.0, 37.6 → 36.2%), so one 100-point transect determines 33% of the published 36.2%. That transect is also the source of the striking "only four genera" and "45% algae" shallow findings — descriptive only, and not to be generalised. (2) **The multi-method row in the current source disagrees with the figure carried in the previous version of this shard.** `chagos_multimethod_cover_by_site.csv` gives `coralnet_points_pct` = **23.3%** over `n_photos` = **46**; the earlier shard reported 22.0% over 51 frames plus a fourth "published image product" figure of 31.1% over 51 rows. The 46-frame CSV row is treated as authoritative here. Someone should reconcile the 46-frame and 51-frame products before either number is republished. (3) **Do not print the 51-frame folder count beside a CoralNet or Coralscapes figure** — `n_photos` = 46 is the scored subset; 30 of the 51 folder frames were reviewed for this description. (4) **Bleaching percentages here use three different denominators and do not reconcile.** Published `Mean(%bleaching)` is 7.7%; `chagos_pct_bleached_recomputed.csv` records 37 bleached points, which is 2.4% of all 1,530 points but **9.89% of the 374 coral points including bleached**, and marks the site `MISMATCH` at +2.19 pp against the published value. None is wrong; they measure different things, and must never be quoted interchangeably. (5) **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 far weaker for bleaching than for cover: r = 0.40 for bleached against r = 0.72 for live hard coral (`CHAGOS_MULTIMETHOD_README.md`). Treat it as the weaker of the two published numbers and never as a diver's field estimate. (6) **`coralnet_points_pct` is a human 30-point read, not a model output** — do not describe it as a CV product. (7) **The mechanism behind the white coral in P3220481 and P3220456 is unresolved** — severe localized bleaching and very recent tissue loss look identical in a single still. Both colonies show clean, unfouled white skeleton, which does place the event close to the survey date; the encrusted framework in P3220483 and the dead-topped colony in P3220467 are by contrast much older. (8) **Filename collision with BITB43.** Frames P3220426, P3220427, P3220428, P3220429 and P3220430 exist in *both* `BiTb42-finished` and `BiTb43-finished` at different byte sizes (two cameras on 22 Mar). Never key on the bare stem — always carry the folder path. (9) **P3220492 contains a *Diadema*-type long-spined urchin that could be misclassified as COTS by an automated pass.** No crown-of-thorns was seen in any reviewed frame. (10) **Several frames are blue-water or open-sand shots with almost no benthos** (P3220429, P3220494, and P3220426 in part). These are the documented confound behind the photo-vs-transect cover gap; `CHAGOS_MULTIMETHOD_README.md` shows that excluding them narrows but does not close that gap, and actually *worsens* method correlation, so they should not be culled to "fix" the CV numbers. (11) **Multi-method deviation is −13.8 to −12.9 pp** (transects 36.2%, CoralNet 23.3%, Coralscapes 22.4%, `sd_all` 7.7), in line with the corpus expectation at this cover level. Do not average the three; the transect value is ground truth for cover. (12) **Formal data mix 100- and 500-point transects** (11 at 100 points, 5 at 500) and are represented here by percentage fields only. (13) **No D1 stratum exists at this site** and no per-frame depths are recorded, so no photograph can be tied to a depth transect. (14) **Fish identifications are kept conservative:** the sharks in P3220426, 460, 470 and 494 are grey-reef-type, the groupers in P3220446 and P3220448 are identified to type only, and the barred fish in P3220458 is not resolved. (15) **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 **7.73%, rank 52 of 110** (1st training: 7.73%, rank 51 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 BITB42 benthic survey data (CSV)](../benthic_data/BITB42_benthic.csv) — scleractinian cover by genus with full names, plus substrate and algal categories (bare substrate, CCA, turf, turf+sediment, macroalgae, erect algae, cyanobacteria, invertebrates), from the LOF BIOT-1 survey (Mayfield & Dempsey 2026, *Oceans*, supplemental file S1).