### BITB40 — Middle 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:** 6.14811°S, 71.52852°E
**Setting:** intermediate-exposure emergent fringing reef in a lagoon channel setting, 27.8 m (the published `exposure-original` cell is blank for this site; `exposure-revised` reads intermediate)
**Coral cover:** 44.9% (15 formal transects, D2–D4; **algal cover 32.4%**) · **published image bleaching (`Mean(%bleaching)` column):** 0.0%
**Live hard coral cover across methods:** formal transects 44.9% · CoralNet points 12.4% · Coralscapes masks 30.0% (15 frames). These products have different sampling footprints and denominators and should not be averaged.
**Temperature:** 30.0 °C · **Salinity:** 34.5
**Photo review:** all 15 of 15 frames examined
**Representative photos:** P3210425.jpg, P3210426.jpg, P3210429.jpg, P3210431.jpg, P3210434.jpg

BITB40 sits on the lagoon-channel side of Middle Brother at 27.8 m and carries above-median coral cover for the archipelago: **44.9% coral against 32.4% algae, ranking 37th of 109 Chagos sites for coral cover and 82nd of 109 for algal cover** (i.e. in the lower quartile for algae). The algal budget is unusually evenly split rather than CCA-dominated: crustose coralline algae (PCCATS) 13.3% of total benthos (57th of 109), fleshy turf (PTTS) 8.7% (62nd of 109), turf-on-sediment (PTSTS) 8.6% and macroalgae (PMTS) 1.8%, with encrusting and cyanobacterial categories effectively absent (PETS 0.02%, PCYTS 0.01%). Bare substrate is comparatively high at 13.7% and non-coral invertebrates at 9.1% — the latter is visible directly in the imagery as gorgonian sea fans, wire corals and encrusting sponges, which are a conspicuous element of this site in a way they are not at most Chagos forereefs.

Coral cover is spread across an exceptional number of taxa: **33 nonzero genera, 6th-highest of the 109 Chagos sites**, and the 33 genus percentages sum to 44.88% of the benthos — an exact match to the published cover figure, so nothing is unaccounted for. All values below are percent of total benthos, not percent of coral. *Acropora* leads at 13.68% of the benthos (30% of the coral present), then an unusually flat second tier: *Montipora* 6.87% (15%), *Porites* 6.85% (15%), *Pocillopora* 4.94% (11%) and *Favia* 3.10% (7%) — those five hold 79% of all coral. Below them, *Platygyra* 1.29%, *Leptastrea* 1.24% and *Goniastrea* 1.22% each clear 1%, followed by *Favites* 0.80%, *Echinopora* 0.72%, *Pavona* 0.65%, *Symphyllia* 0.62% and *Cyphastrea* 0.58%. A tail of 20 further genera — *Cantharellus*, *Stylophora*, *Gardineroseris*, *Pachyseris*, *Millepora*, *Montastrea*, *Astreopora*, *Ctenella*, *Herpolitha*, *Hydnophora*, *Galaxea*, *Acanthastrea*, *Fungia*, *Echinophyllia*, *Leptoseris*, *Stylocoeniella*, *Oxypora*, *Isopora*, *Seriatopora* and *Lobophyllia* — each sits under 0.5% but together carries the genus count into the archipelago's top ten. The near-parity of *Montipora* and *Porites* behind *Acropora* is what separates BITB40 from the *Acropora*-then-*Porites* pattern typical of the Three Brothers sites.

**The depth profile is the real finding here, and it is a clean monotonic decline with a genus handover at the bottom.** Coral cover falls from **54.3% at D2 (5 transects) to 44.2% at D3 (6 transects) to 36.2% at D4 (4 transects)** — an 18-point drop across the profile — and *Acropora* collapses with it, from **30.04% of the benthos at D2 (55% of all coral on that stratum) to 7.50% at D3 and just 3.50% at D4**, a nine-fold reduction. *Montipora* declines much more gently (8.00% → 6.97% → 5.65%), so it overtakes *Acropora* by D3. **At D4 the leading genus is *Porites* at 11.50%, its site maximum and roughly double its site-level figure**, with *Pocillopora* 5.95% and *Montipora* 5.65% behind it and *Acropora* relegated to fourth; *Goniastrea* (1.75%) and *Favia* (1.80%) also reach their profile maxima there. The algal budget does not simply mirror the coral: algae peaks mid-profile at D3 (36.7%) rather than deep, and D4 carries the lowest algal cover of the three strata (28.0%) — so the deep stratum's coral shortfall is made up by bare substrate and invertebrates rather than by algae. **Bare substrate nearly triples down the profile, 8.4% at D2 to 9.1% at D3 to 23.7% at D4**, and non-coral invertebrates follow the same track (4.9% → 10.1% → 12.2%), so the deep stratum is best read as an open, sponge- and gorgonian-bearing pavement with residual massive *Porites* rather than as an algal-overgrown one. **The site-level number of 13.68% *Acropora* therefore describes no single stratum**: it averages an *Acropora*-dominated shallow zone against a deep zone where *Acropora* is nearly gone and massive *Porites* takes over. Note also that there is **no D1 stratum at all** — the shallowest surveyed depth here is D2 — so the site figure is not comparable with four-stratum sites like neighbouring BITB39 without that caveat. Individual transects span **21.2% to 66% coral**, a 45-point range that exceeds the 18-point depth signal, so within-stratum patchiness is real and the depth trend is a trend in means, not a description of any one patch.

The photographs — all 15 frames in `BiTb40-finished`, reviewed in full — document exactly that patchiness. The high-cover end of the set is a near-continuous canopy of corymbose and digitate branching colonies consistent with *Acropora* (P3210429–433, P3210435), with tabular plates set among them (P3210431, P3210434), massive dome colonies consistent with *Porites* and brain-patterned faviid heads scattered through the thickets, and pale sand-and-rubble lanes threading between. **P3210425 is the essential counter-frame**: an open, sediment-dusted pavement-and-rubble slope falling into blue water, with only scattered small colonies, dark encrusting growth and thin whip/wire corals — a defensible photographic analogue of the 21.2% transect. P3210424 is intermediate, showing broken framework and rubble heavily encrusted in pink crustose coralline algae, with a large fine-textured massive head and small branching colonies growing out of it. **P3210426 is the most taxonomically varied frame in the set**, carrying convoluted foliose/encrusting plates consistent with *Montipora*, massive *Porites*, a large gorgonian sea fan, orange and red encrusting sponges and extensive pink-to-magenta coralline crust in a single view — a good match to both the high genus count and the 9.1% non-coral invertebrate figure. Gorgonian sea fans recur prominently in P3210427–428. Water clarity is good throughout, with strong strobe-lit foreground colour and deep blue backscatter.

**Condition reads as good, with no bleaching signal that the imagery corroborates at site scale.** The published image-based bleaching figure is **0.0% from the `Mean(%bleaching)` column, ranking 96th of 98 Chagos sites with a value — effectively the floor of the archipelago-wide range** (and 107th of 110 on both the 1st- and 2nd-training CoralNet runs, which agree exactly here). The 15-frame review is consistent with that: colonies retain normal tan, olive, purple and brown pigmentation across the set, and no frame shows a coherent field of white or fluorescing-pale colonies. The only pallor worth recording is a cluster of pale, knobby, cream-coloured colonies in the lower right of P3210432 and scattered white *Acropora* branch tips in P3210431 and P3210435; white apical tips on branching *Acropora* are normal active growth, not bleaching, and the P3210432 cluster cannot be separated from a naturally pale *Porites* or *Millepora* on a single still, so both are flagged rather than scored. **No COTS individuals, no COTS feeding scars, no disease lesions or tissue-loss margins, and no *Drupella* aggregations or corallivore bite scarring were seen in any of the 15 frames** — an explicit absence-of-evidence finding across the complete set, not an omission.

Fauna is well represented for a 15-frame set. **P3210434 shows two reef sharks cruising in open water over the reef** — robust-bodied, blunt-snouted animals consistent with grey reef shark (*Carcharhinus amblyrhynchos*); an earlier review of this site recorded the same frame and explicitly corrected an initial reading of "three sharks or rays" to two sharks, and that correction is upheld here. A **giant clam (*Tridacna* sp.) is embedded in a massive coral head in P3210429**, its zigzag mantle clearly visible. Reef fish include dense humbug and three-spot damselfish over the *Acropora* thickets (P3210429–430), butterflyfish (P3210424, P3210435), a blue surgeonfish and small wrasse in P3210432, and elongate midwater fish consistent with cornetfish over the canopy in P3210435. Divers and the yellow transect line appear in P3210436–438 — standard field documentation tying the imagery to the formal survey, not a data concern.

<b>Assemblage composition.</b> Formal transects give 44.9% coral against 32.4% algae, 13.7% bare substrate and 9.1% non-coral invertebrates, across 33 nonzero genera (6th of 109 sites) led by *Acropora* 13.7%, *Montipora* 6.9%, *Porites* 6.8%, *Pocillopora* 4.9% and *Favia* 3.1% — all as percent of benthos. The three-stratum depth profile runs 54.3% coral at D2 to 36.2% at D4, with *Acropora* falling from 30.0% to 3.5% of the benthos and *Porites* rising to 11.5% to take over as the deep-stratum leader. The photographs match that range end to end, from continuous *Acropora* canopy (P3210429–433) to an open rubble-and-pavement lane with scattered colonies (P3210425), and add the site's conspicuous gorgonian and sponge component (P3210426–428), two reef sharks (P3210434) and an embedded giant clam (P3210429). No bleaching field, disease or COTS was seen in any of the 15 frames.

**Flags:** (1) **The site has no D1 stratum** — the formal data cover D2–D4 only, so the headline 44.9% is the unweighted mean of three strata starting one depth band deeper than at four-stratum neighbours such as BITB39. Do not compare the raw figures without that caveat. (2) **The deepest stratum rests on 4 transects and D2 on 5**, so each of those transects carries roughly 7–8% of the headline number; with individual transects ranging 21.2–66%, a single deep transect (D4_T4 at 21.2%) moves the site figure materially. (3) **The two CV cover products disagree sharply with the transects and with each other.** CoralNet points read 12.4% and Coralscapes masks 30.0% against the transect 44.9% (deviation range −32.5 to −14.9 pp, sd 16.3 across three methods) — an 18-point spread between the two CV products alone. Do not average them; the transect value is ground truth for cover. (4) **Denominator conflict between the two published CV runs.** `chagos_multimethod_cover_by_site.csv` records `n_photos=15` for the CoralNet cover figure, but `chagos_pct_bleached_recomputed.csv` lists only **5 annotated images / 150 points** for BITB40, of which 29 are live coral (19.3%) — a figure that matches neither the 12.4% CoralNet cover nor the 30.0% Coralscapes value. The 0.0% bleaching therefore rests on a 5-image, 150-point subset, one third of the folder. **Someone should reconcile which annotation set backs which published column.** (5) **The 0.0% bleaching figure is a zero from a small denominator, not a demonstrated absence of bleaching** — with 150 points and 29 coral points, the detectable floor is coarse. The 15-frame visual review independently supports "no site-scale bleaching field", which is the stronger statement of the two. (6) **Two ambiguous pallor observations, not scored as bleaching:** pale cream knobby colonies at the lower right of P3210432, and white *Acropora* branch tips in P3210431 and P3210435 (normal apical growth). Neither is resolvable from a single still. (7) **No COTS, disease or corallivore predation was observed across the full 15-frame review** — recorded as an explicit negative. (8) **`exposure-original` is blank in the published table** for this site; only `exposure-revised` (intermediate) carries a value. The blank is reported, not filled. (9) **Genus richness is a site strength distinct from cover: 33 nonzero genera, 6th of 109 Chagos sites.** (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, never as a diver's field estimate. Note also that `Mean(%bleaching)` is **not a mean** — it equals the 1st CoralNet training run, while the dive-history map publishes the 2nd; here the two runs agree exactly at 0.0, so the distinction does not change this site's value, and any rank off that column is approximate because it carries at least one known bad entry (BITB58 listed at 22.22% against a corrected 0.77%). (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 **0.00%, rank 107 of 110** (1st training: 0.00%, rank 107 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 BITB40 benthic survey data (CSV)](../benthic_data/BITB40_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).