### BIDI15 — Danger Island
**Date:** 2015-03-14
**GPS:** 6.3531°S, 71.2358°E
**Frames reviewed:** 10 of 10 (100% of the site folder, `BiDi15-finished`) — **one frame (P3140294) is open blue water containing no substrate at all**
**Coral cover:** 20.1% (formal transect surveys) · **% bleached:** 8.8% (CoralNet analysis of informal survey images; Mayfield & Dempsey 2026, Oceans)
**Live hard coral cover across methods** (n=10 photos): transect 20.1% · CoralNet points 17.3% · **Coralscapes 33.9%**. Deviation vs transect -2.8 to **+13.8 pp**. **Mean of 3: 23.8% ± 8.9.** **The two photo methods split by 16.6 pp — the second-largest photo-method disagreement in the archipelago after BIDI12, and the frame set explains it (see the assemblage note). Treat the photo-derived figures for this site with caution.**
**Coral genera identified:** 7 scleractinian + 1 octocoral — *Platygyra*/*Leptoria* (superb development here), *Porites* (massive domes), *Astreopora*, *Pavona*/*Echinopora* (foliose and plating), *Acropora* (small, scattered), *Pocillopora*, *Favites*/*Goniastrea*; plus large gorgonian sea fans. Growth-form calls from photographs, given as "compatible with".
**Representative photos:** P3140268.jpg, P3140247.jpg, P3140270.jpg, P3140261.jpg, P3140291.jpg

BIDI15 is a **protected, low-relief forereef of eroded pavement and rubble**, and the wide frames are consistent and unflattering: P3140246, P3140272, P3140283 and P3140291 all show pale, largely bare carbonate hardground strewn with rubble, across which small colonies, olive-green foliose plates compatible with *Pavona* or *Echinopora*, and domal *Porites* are widely scattered. P3140291 in particular is a broad plain in which live coral is a clear minority of the surface. The site's 20.1% transect cover is a fair description of this.

Against that background the site has two genuinely distinctive features. First, **large gorgonian sea fans** — yellow-brown and finely branched — stand on the pavement in P3140246, P3140247 and P3140261, a growth form scarce elsewhere in the reviewed Chagos frames and typical of current-swept, low-relief hardground. Second, and more striking, **BIDI15 has the finest brain corals in the collection.** P3140268 is a large, near-spherical meandroid colony compatible with *Platygyra* or *Leptoria* filling most of the frame, its valleys tightly and evenly sinuous, flanked by a porous yellow-green dome compatible with *Astreopora*. P3140270 and P3140271 then go closer still — **two extreme macros resolving individual polyps and extended tentacles along the valley walls**, rendered in an orange-to-teal gradient (a strobe and white-balance artefact, not natural pigmentation). These are the highest-magnification coral images in the reviewed set and are the obvious candidates for any illustrative or outreach use.

**Bleaching: I could not corroborate the 8.8% figure from these frames.** No colony in any of the ten is white or obviously paled — the brain corals, *Porites*, foliose plates and gorgonians are all normally pigmented. That is not a contradiction of the CoralNet number, because ten opportunistic frames (one of them empty blue water, three of them macros of a single colony) cannot sample a site's bleached fraction. But it should be stated plainly rather than glossed: **the imagery here neither supports nor refutes 8.8%, and the transect and CoralNet figures remain the only quantitative basis.** The pale tan colony and surrounding whitened substrate in P3140283 are the only ambiguous feature, and read more as old eroded pavement than as recent bleaching.

Faunally the dive is quiet — no schooling aggregations of the kind that dominate BIDI08–BIDI14. **A whitetip reef shark compatible with *Triaenodon obesus* cruises low over the pavement in P3140247** (the third whitetip record of the expedition, after BIDI08 and BIDI13), with scattered damselfishes and small reef fish elsewhere. P3140294 is open blue water with no subject.

<b>Assemblage composition.</b> This site is a **methodological caution, and it is the second of its kind after BIDI12.** Coralscapes reads 33.9% live hard coral against CoralNet's 17.3% and the transect's 20.1% — but the model figure is an artefact of frame composition, not of the reef. **Three of the ten frames (P3140268, P3140270, P3140271) are close macros in which a single brain coral fills most or all of the image**, and an area-based mask will score those at near-100% coral; with only ten frames in the set, three such images can lift the mean by more than 20 points on their own. A fourth frame is empty blue water. In other words, **a ten-frame opportunistic set containing frame-filling macros cannot produce a site-level cover estimate**, whichever method is applied to it. The general rule emerging across BIDI12 and BIDI15 is that **photo-derived cover should be computed only from frames shot at consistent, reef-scale framing** — and that small frame sets should carry a minimum-n and framing check before they enter any archipelago summary.

**Flags:** (1) **DATA-QUALITY — Coralscapes 33.9% is inflated by frame composition.** Three of ten frames are frame-filling coral macros and one is empty blue water. Do not use the photo-derived cover or the mean-of-3 for this site without re-computing on reef-scale frames only. **Second instance of this class after BIDI12** — recommend a framing/minimum-n filter before any archipelago photo-cover summary. (2) **The 8.8% bleached figure is not corroborated by any of the ten frames** — no white or obviously paled colony appears. Stated as a limitation of the frame set, not as a contradiction of the number. (3) **P3140270/P3140271 are the highest-magnification coral macros in the reviewed collection** (polyps and tentacles resolved) — flagged for outreach/illustrative use. Their orange-teal colour is a strobe/white-balance artefact, **not** pigmentation, and must not be read as a health signal. (4) **Whitetip reef shark** in P3140247 — third record of the expedition. (5) Large gorgonian sea fans (P3140246, P3140247, P3140261) — a growth form scarce elsewhere in the Chagos frames. (6) P3140294 contains no substrate and should be excluded from any benthic analysis. (7) No per-frame depth data.

**Benthic data:** [Download BIDI15 benthic survey data (CSV)](../benthic_data/BIDI15_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).