### BIDI10 — GCB-West (Danger Island)
**Date:** 2015-03-12
**GPS:** 6.39078°S, 71.24618°E
**Frames reviewed:** 31 of 31 (100% of the site folder, `BiDi10-finished`)
**Coral cover:** 23.0% · **% bleached:** 2.3%
**Live hard coral cover across methods** (n=31 photos): transect 23.0% · CoralNet points 21.0% · Coralscapes 22.4%. Deviation vs transect -2.0 to -0.6 pp. **Mean of 3: 22.1% ± 1.0** — **by far the tightest three-method agreement in the archipelago.** (Unlike every other site, this SD really does approximate precision rather than bias; see the assemblage note and CHAGOS_MULTIMETHOD_README.md.)  
**Coral genera identified:** 8 scleractinian + 1 hydrocoral — *Acropora* (digitate, corymbose and tabular), *Porites* (massive domes, broad plates and digitate), *Pocillopora*, *Platygyra*/*Leptoria*, *Favites*/*Goniastrea*, *Echinopora*/*Pavona* (foliose), *Astreopora*, *Montipora*; hydrocoral *Millepora*. Also long-spined urchins compatible with *Diadema*. Growth-form calls from photographs, given as "compatible with".
**Representative photos:** P3120248.jpg, P3120276.jpg, P3120288.jpg, P3120274.jpg, P3120278.jpg

**BIDI10 is the most important site reviewed so far for the multi-method question, and its condition is why.** At 23.0% transect cover it is less than half its neighbour BIDI08 (59.0%), and it is built on **low-relief, largely dead framework** rather than standing tabular plates. **P3120248 is the defining frame: a field of dead branching coral rubble — grey-brown, broken, algal-crusted — running unbroken to the limit of visibility.** P3120247 shows the same substrate in macro, a single small live colony surviving on a pavement of CCA-encrusted dead rubble, and P3120249 shows dead framework being recolonised by soft corals and encrusting growth. This is not 2015 bleaching damage — the rubble is consolidated, coralline-crusted and colonised, so it is **older mortality of unresolved date and cause**, and it should not be read as an impact of the event this expedition was documenting.

Where live coral persists it is diverse and healthy, and it takes forms the tabular sites lack. **P3120276 and P3120270 are the best benthic frames** — dense mosaics of lilac and purple digitate *Acropora*, brown *Pocillopora*, massive and plating *Porites*, dark foliose *Echinopora* or *Pavona*, yellow-green tabular *Acropora* and pink crustose coralline algae packed together at low relief. **P3120256 and P3120269 record a distinctly foliose-dominated zone** — grey-brown ruffled leafy plates covering large areas — and **P3120259 and P3120261 show enormous yellow-gold plating *Porites*** forming tiered ledges and overhangs several metres across. Two macro frames stand out: **P3120280**, an extreme close-up of coral polyps with white-tipped nematocyst batteries clearly resolved, and **P3120278**, a large brown meandroid brain compatible with *Platygyra* or *Leptoria*.

Bleaching is essentially absent and the 2.3% figure holds: no bleached colony appears in any of the 31 frames. Two condition signals are logged instead. **P3120278 shows a clean white lesion running across the surface of the brain coral** — a band of exposed skeleton with live tissue either side, which could be disease, predation scarring or physical damage and cannot be resolved from one frame. And **P3120263 shows a large pale, tiered plating colony** whose surface is grey-tan and dulled, read here as older mortality rather than bleaching. Separately, **long-spined urchins compatible with *Diadema* are conspicuous** — five or more together on open pavement in P3120288, and again around the lesioned brain in P3120278. Given the dead rubble fields, that urchin abundance is worth noting: these are bioeroders and grazers, and their density is a variable in how this reef's framework is being reworked.

Faunally the site is dominated by planktivores and their predators. **P3120274 is the dive's spectacle**: a dense wall of tiny silver baitfish — cardinalfishes or silversides — streaming across the reef with dark snappers and a large jack cutting through them, with a comparable swarm in P3120253 and P3120254. Blue fusiliers compatible with *Caesio* or *Pterocaesio* fill P3120262 through P3120267. Emperors and snappers compatible with *Monotaxis*, *Lutjanus* or *Macolor* school through P3120250, P3120251, P3120272 and P3120273. Reef fish include powder-blue surgeonfish compatible with *Acanthurus leucosternon* (P3120273, P3120276), parrotfishes compatible with *Scarus* (P3120271), butterflyfishes, and red squirrelfishes and bigeyes compatible with *Sargocentron* or *Priacanthus*.

<b>Assemblage composition.</b> The cover line above carries the headline result, and it deserves emphasis: **transect 23.0%, CoralNet 21.0%, Coralscapes 22.4% — a spread of 2.0 pp, against gaps of -14 pp at BIDI09 and -25 pp at BIDI08 on the same bank in the same week.** The frames explain the difference exactly. BIDI08's cover is carried by large horizontal tabular plates that photographs foreshorten; BIDI10's is carried by low-relief encrusting, massive and digitate colonies lying essentially in the plane of the substrate, which a downward photograph and a point-count see the same way a transect tape does. **This is the cleanest natural experiment in the Chagos dataset for the claim that the photo-vs-transect gap is driven by growth form and relief, not by model error** — three sites, one bank, one week, cover from 23% to 59%, and the disagreement scaling with the amount of tabular framework rather than with anything about the methods themselves.

**Flags:** (1) **This site is the low-cover, low-relief end-member of the multi-method comparison** and should be cited alongside BIDI08 (the tabular high-cover end-member) whenever the photo-vs-transect gap is discussed. Its ±1.0 SD is the one place in the archipelago where the SD is plausibly precision rather than bias. (2) **P3120248 documents extensive OLD dead branching rubble** — consolidated and CCA-crusted, so pre-dating 2015; do not attribute it to the event this expedition documented. (3) **P3120278: white lesion on a brain coral** — disease, predation or physical damage, unresolved from a single frame; needs Anderson's eye. (4) ***Diadema*-type urchin aggregation** (P3120288, P3120278) — relevant to bioerosion given the dead framework. (5) P3120263's pale plating colony is logged as older mortality, not bleaching. (6) **Filename collision warning: BIDI08 and BIDI10 share frame numbers** (P3120253, P3120254, P3120277, P3120278 and others exist in both folders). Always qualify Chagos frame IDs with the site code — a bare "P3120278" is ambiguous. (7) The polyp macro P3120280 cannot be assigned to genus from tentacle detail alone. (8) No per-frame depth data.

**Benthic data:** [Download BIDI10 benthic survey data (CSV)](../benthic_data/BIDI10_benthic.csv) — scleractinian cover by genus with full names, plus substrate and algal categories (bare substrate, CCA, turf, macroalgae, invertebrates), from the LOF BIOT-1 survey (Mayfield & Dempsey 2026, *Oceans*, supplemental file S1).