### BITB39 — South Brother
**CRD page:** [Three Brothers/Central Chagos Coral Surveys BITB38-BITB49 (Mar 2015)](https://coralreefdiagnostics.com/three-brotherscentral-chagos-20150321-to-0324)

**21 Mar 2015**
**GPS:** 6.17145°S, 71.53494°E
**Setting:** intermediate-exposure emergent fringing forereef outside the lagoon, 28.6 m (published `exposure-original` is windward; `exposure-revised` is intermediate)
**Coral cover:** 50.3% (19 formal transects, D1–D4; **algal cover 36.3%**) · **published image bleaching (`Mean(%bleaching)` column):** 4.2%
**Live hard coral cover across methods:** formal transects 50.3% · CoralNet points 29.1% · Coralscapes masks 26.7% (61 frames). These products have different sampling footprints and denominators and should not be averaged.
**Temperature:** 29.6 °C · **Salinity:** 34.5
**Photo review:** **38 of 61 frames examined** — a systematic every-other-frame sample spanning the full sequence P3210492 → P3210586 (31 frames), plus 7 further frames named in an earlier review of this site (P3210504, 527, 535, 537, 555, 561, 575) so that no previously recorded observation was dropped
**Representative photos:** P3210503.jpg, P3210511.jpg, P3210561.jpg, P3210566.jpg, P3210578.jpg

BITB39 sits on the fringing forereef off South Brother at 28.6 m, the deepest and highest-cover of the three Three Brothers sites in this group, and it is **among the top quartile of the archipelago for coral cover: 50.3% coral against 36.3% algae, ranking 24th of 109 Chagos sites for coral cover and 66th of 109 for algal cover** — a mid-range algal load, not a low one. The composition of that algal budget is what marks the site out. **Crustose coralline algae (PCCATS) is 18.8% of total benthos, 31st-highest of 109 sites**, and fleshy turf (PTTS) is 12.2%, 38th of 109 — so unlike the CCA-dominated hard pavement of some neighbouring forereefs, BITB39 carries a genuinely substantial turf component alongside its coralline crust. Turf-on-sediment adds 2.4%, macroalgae 2.7% and encrusting algae 0.3%; cyanobacterial cover (PCYTS) is recorded as exactly zero. Bare substrate is low at 8.7% and non-coral invertebrates lower still at 4.7% — this is a well-covered reef with very little open ground at site scale.

Coral cover is spread across an exceptional number of taxa: **34 nonzero genera — the equal-2nd-highest genus count of the 109 Chagos sites in the central rank table**, behind only BIBL87 (36) and tied with BITB54, BISA93 and BIBL82, all also at 34. The 34 genus percentages sum to 50.26% 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 19.69% (39% of all coral present), then *Porites* 8.61% (17%), *Pocillopora* 4.30% (9%), *Favia* 3.50% (7%) and *Gardineroseris* 2.87% (6%) — those five hold 78% of the coral. *Montipora* 2.32%, *Pavona* 1.63% and *Favites* 1.36% follow, with *Stylophora* 0.85% and *Platygyra* 0.82% closing the group that clears half a percent. The tail of 24 further genera — *Turbinaria*, *Symphyllia*, *Leptastrea*, *Echinopora*, *Montastrea*, *Goniastrea*, *Cyphastrea*, *Pachyseris*, *Leptoseris*, *Psammocora*, *Millepora*, *Astreopora*, *Hydnophora*, *Ctenella*, *Echinophyllia*, *Acanthastrea*, *Fungia*, *Galaxea*, *Podabacia*, *Cantharellus*, *Oxypora*, *Goniopora*, *Lobophyllia* and *Scolymia* — contributes under 3% of the benthos in total but carries the genus count to the top of the archipelago. The presence of *Ctenella* (0.13%), the Chagos-endemic brain coral, is worth noting even at trace level.

**The depth profile carries the real finding at this site, and it is a genus handover at the top of the reef rather than at the bottom.** Coral cover rises from **42.3% at D1 (the shallowest stratum, 2 transects) to 54.1% at D2 (7 transects) and 54.6% at D3 (6 transects), then eases to 50.1% at D4 (4 transects)** — a modest 12-point spread. The genus order, however, inverts completely at D1. **At every stratum from D2 down, *Acropora* is the clear leader — 30.40% of the benthos at D2 (56% of all coral there), 19.20% at D3, 17.95% at D4. At D1 it is not: *Porites* leads at 17.70%, against *Acropora* at only 11.20%**, with *Pocillopora* 3.00%, *Favites* 1.90%, *Platygyra* 1.90% and *Turbinaria* 1.60% behind. The shallow stratum is a massive-*Porites* zone; everything deeper is an *Acropora* zone. The substrate flips with it: **D1 carries by far the highest crustose coralline cover of the profile (30.0% of benthos, against 14.6–16.0% at D2–D4) and the lowest turf (7.2%, against 18.0% at D2 and 18.4% at D3)** — so the shallow zone is hard, well-grazed coralline pavement with massive corals on it, while the mid-depths are turf-and-*Acropora* ground. The deepest stratum is different again: turf collapses to 5.1%, macroalgae rises to its profile maximum of 6.9%, and bare substrate doubles to 16.2% (against 3.5–8.4% shallower), while *Gardineroseris* (4.05%) and *Favites* (2.45%) climb into the leading ranks. **The site-level figures of 19.69% *Acropora* and 8.61% *Porites* therefore describe no single stratum**: they average a *Porites*-and-coralline shallow zone against three deeper *Acropora* strata. Individual transects span **33% to 70% coral**, so within-stratum patchiness (37 points) is three times the depth signal (12 points) and both the dense table fields and the open-framework patches visible in the photographs are genuine parts of the site.

The photographs — a 38-frame systematic sample of the 61 in `BiTb39-finished` — document a high-relief forereef whose defining architecture is **broad tabular *Acropora*, present in numbers few Chagos sites match.** P3210566, P3210584 and P3210586 look across fields of dozens of overlapping golden-green tables set on pale pavement; P3210503 and P3210504 show the same tables stacked down a slope falling into blue water, interleaved with large brown foliose plate colonies and two prominent yellow gorgonian sea fans; P3210515, P3210524 and P3210555 are single large tables with bright white active growing margins. Between the tables the ground is a mix of corymbose and digitate *Acropora* clumps, knobby brown *Porites* lobes, meandroid brain colonies (P3210575 upper left, P3210572) and extensive pink-to-magenta coralline crust on dead framework (P3210492, P3210501, P3210526, P3210570). **P3210509 is the most structurally complex frame in the sample** — stacked foliose plates, cream ridged plate colonies, columnar and encrusting forms and a golden lobed colony in one view — and is a plausible counterpart to the 34-genus count. Open, lower-cover ground with pale sediment-dusted pavement between colonies is equally well represented (P3210574, P3210575, P3210582, P3210584), matching the 33% end of the transect range. Water clarity is excellent throughout, with deep blue backscatter and strong strobe-lit foreground colour, and the reef is under an almost constant curtain of fish (below).

**Condition reads as good, and the published 4.2% bleaching figure is not corroborated as a reef-wide event by the imagery — a judgement first made in an earlier review of this site and upheld here on a wider frame sample.** The published figure ranks **76th of 98 Chagos sites with a value — in the lower quartile of bleaching severity** (84th of 110 on both the 1st- and 2nd-training CoralNet runs, which agree exactly at 4.19 here). Most colonies across the 38 frames retain normal tan, olive, gold, brown and purple pigmentation. What the frames do show is **pale and white *substrate*, not a field of pale live colonies**: the conspicuous white areas in **P3210575 and P3210576** are principally exposed and partly encrusted flat framework, eroded plate skeletons and rubble, exactly as the earlier review recorded, and the same reading applies to the white pavement in P3210574, P3210582 and P3210584. Three observations are recorded as genuinely unresolved and are flagged rather than scored: **P3210578**, where a large flat sheet mid-frame is bright white with a faintly textured surface — a recently dead or bleached tabular plate, not separable from a single still; **P3210540 and P3210568**, where several large cream-white plate and foliose colonies sit conspicuously paler than their neighbours; and **P3210513**, a macro frame in which discrete clusters of entirely white branches are scattered across an otherwise brown branching colony — a pattern consistent with partial bleaching, with recent corallivore tissue loss, or with normal apical growth, and not resolvable without a closer view. Ordinary white apical tips on branching *Acropora* recur through P3210492, P3210501, P3210526, P3210561 and P3210562 and are read as active growth. **No COTS individuals, no COTS feeding scars, no clear disease front with a live/dead margin, and no *Drupella* aggregations or corallivore bite scarring were seen in any of the 38 frames reviewed** — an explicit absence-of-evidence finding, consistent with the earlier review of this site.

Fauna is abundant and is one of the site's strongest features. **Schooling fusiliers dominate the water column in roughly a quarter of the sampled frames** — blue-and-yellow *Pterocaesio*-type shoals in P3210528, P3210535, P3210536, P3210555 and P3210556, and dense silvery baitfish clouds in P3210527, P3210537 and P3210538, at times all but obscuring the reef behind them. **P3210505 shows two bluefin trevally (*Caranx melampygus*)** — blue-spotted flanks and purple-blue fins — cruising in open water, refining an earlier record of "jacks or trevallies" at this frame. **P3210511 shows a large giant clam (*Tridacna* sp.) embedded in the reef framework**, its dark mantle boldly patterned in cream, surrounded by coralline crust and colonial tunicates. **P3210561 and P3210562 show a moray eel in a crevice**, dark with dense yellow-gold speckling and a yellow eye — consistent with yellowmargin moray (*Gymnothorax flavimarginatus*), recorded as a hedged identification. Reef fish through the set include butterflyfish (P3210574, P3210582), powder-blue surgeonfish (P3210564, P3210568, P3210582), a bannerfish or moorish idol in P3210560, black surgeonfish over the tables in P3210537 and P3210586, and blue-green chromis and damselfish in almost every wide frame. Green colonial tunicates carpet framework in P3210513. Divers, a camera rig and a white survey quadrat frame appear in P3210509, P3210528, P3210537, P3210574 and P3210580 — standard field documentation of the transect work, not a data concern.

<b>Assemblage composition.</b> Formal transects give 50.3% coral against 36.3% algae, 8.7% bare substrate and 4.7% non-coral invertebrates, across 34 nonzero genera (equal-2nd of 109) led by *Acropora* 19.7%, *Porites* 8.6%, *Pocillopora* 4.3%, *Favia* 3.5% and *Gardineroseris* 2.9% — all as percent of benthos. The four-stratum depth profile runs 42.3% coral at D1 up to 54.6% at D3 and back to 50.1% at D4, and its real signal is compositional: **D1 is a *Porites*-led (17.7%), coralline-pavement zone with only 11.2% *Acropora*, while every deeper stratum is *Acropora*-led, peaking at 30.4% of the benthos at D2.** The photographs corroborate that architecture end to end — extensive tabular *Acropora* fields (P3210566, P3210586), stacked plates and foliose colonies with gorgonian sea fans (P3210503, P3210509), massive *Porites* and brain colonies on coralline pavement (P3210564, P3210575), and open pale pavement matching the 33% transect end (P3210574, P3210582) — and add fusilier and baitfish schools, two bluefin trevally, a giant clam and a moray. No COTS, disease front or predation scarring was seen in any of the 38 frames sampled, and the published bleaching figure is not visually corroborated at site scale.

**Flags:** (1) **D1 rests on only 2 transects (44.4% and 40.2%) yet supplies 25% of the headline cover figure**, because the site value is the unweighted mean of per-stratum means. It is also the stratum with the most distinctive composition at the site — the only one where *Porites* outranks *Acropora* — so the *Porites*-dominance finding is real but rests on two transects and should be treated as provisional. (2) **Within-stratum patchiness far exceeds the depth signal:** individual transects span 33–70% coral (37 points) against a 12-point range between stratum means. The depth trend is a trend in means, not a description of any patch. (3) **The two CV cover products both sit far below the transects:** CoralNet points 29.1% and Coralscapes masks 26.7% against the transect 50.3% (deviation range −23.6 to −21.2 pp, sd 13.0 across three methods). Do not average them; the transect value is ground truth for cover. (4) **The published bleaching figure is not visually corroborated as a site-scale event.** Across 38 frames most colonies retain normal pigmentation, and the conspicuous pale areas in P3210575–576 are principally exposed and partly encrusted framework, rubble and localized colony damage rather than a defensible bleaching field. This was the earlier reviewer's conclusion and it is upheld on the wider sample. (5) **Three unresolved pallor observations, not scored as bleaching:** a bright white flat sheet mid-frame in **P3210578** (recently dead or bleached tabular plate — not separable from one still); conspicuously pale cream-white plate/foliose colonies in **P3210540** and **P3210568**; and discrete clusters of entirely white branches on an otherwise brown branching colony in the macro frame **P3210513** (partial bleaching, recent tissue loss, or normal growth — unresolvable). Someone with the full-resolution originals should adjudicate P3210513 and P3210578 in particular. (6) **No COTS, disease or corallivore predation was observed across the 38-frame sample** — recorded as an explicit negative. (7) **23 of the 61 frames were not reviewed** (the odd-index frames not named above). The sample is systematic and spans the full sequence, but the negative findings in (6) are stated over 38 frames, not 61. (8) **`Mean(%bleaching)` reproduces cleanly here.** `chagos_pct_bleached_recomputed.csv` recomputes 3.96% from 22 bleached points of 1,830 (61 annotated images, 555 coral points) against 4.2% published — a −0.24 pp delta, flagged MATCH. The 533 live-coral points of 1,830 also reproduce the 29.1% CoralNet cover figure exactly, so this site's two published CV products are internally consistent, which is not true everywhere in this group (cf. BITB40). (9) **Genus richness is a site strength distinct from cover: 34 nonzero genera, equal-2nd of 109 Chagos sites** (behind BIBL87's 36; tied with BITB54, BISA93 and BIBL82). Note that an older shard for BITB54 quotes 34 genera as "3rd of 109" — the central rank table used here places all four 34-genus sites at rank 2, and the central table governs. (10) **Cyanobacterial cover (PCYTS) is recorded as exactly 0** in the published substrate row — reported as published, not treated as missing. (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`). Treat the site bleaching percentage as the weaker of the two published numbers, never as a diver's field estimate. `Mean(%bleaching)` is also **not a mean** — it equals the 1st CoralNet training run, while the dive-history map publishes the 2nd; the two agree here (4.19 both ways, both ranked 84th of 110), but the column carries at least one known bad entry (BITB58 listed at 22.22% against a corrected 0.77%), so the 76th-of-98 rank quoted above is approximate. (12) **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 **4.19%, rank 84 of 110** (1st training: 4.19%, rank 84 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 BITB39 benthic survey data (CSV)](../benthic_data/BITB39_benthic.csv) — scleractinian cover by genus with full names, plus substrate and algal categories (bare substrate, CCA, turf, turf-on-sediment, macroalgae, non-coral invertebrates), from the LOF BIOT-1 survey (Mayfield & Dempsey 2026, *Oceans*, supplemental file S1).