### BIEA19 — Eagle Island
**Date:** 2015-03-15
**GPS:** 6.202°S, 71.3588°E
**Frames reviewed:** 15 of 15 (100% of the site folder, `BiEa19-finished`)
**Coral cover:** 27.5% (formal transect surveys) · **% bleached:** 2.6% (CoralNet analysis of informal survey images; Mayfield & Dempsey 2026, Oceans)
**Reef type:** bank, protected · **Temperature:** 28.8 °C
**Live hard coral cover across methods** (n=15 photos): transect 27.5% · CoralNet points 18.4% · Coralscapes 21.3%. Deviation vs transect -9.1 to -6.2 pp. **Mean of 3: 22.4% ± 4.6.** **Dropping the four blue-water frames (11 benthic frames remain) narrows this sharply: CoralNet 21.8%, Coralscapes 26.3% — a gap of only −5.7 to −1.2 pp.**
**Coral genera identified:** 9 scleractinian groups + a hydrocoral — *Acropora* (**large tabular plates**, corymbose, digitate and branching), *Porites* (massive/domal, columnar-nodular, and a **lime-green digitate carpet form**), *Pocillopora* (thick-branched), *Montipora* (encrusting and foliose), *Platygyra*/*Favites*-type meandroid, *Turbinaria*/*Pavona*-type foliose plates, plus encrusting and submassive colonies not resolvable to genus; and **upright blade-like pale blue-white plates in P3150371 compatible with the hydrocoral *Millepora* (*M. platyphylla*-type)**. Growth-form calls from photographs, given as "compatible with".
**Representative photos:** P3150385.jpg, P3150384.jpg, P3150371.jpg, P3150368.jpg, P3150369.jpg

BIEA19 is a **protected bank**, and its defining character is **colony density without relief**: P3150394, P3150395 and P3150399 show pavement packed almost edge to edge with massive and submassive *Porites* domes, encrusting sheets, columnar knobs and small plates, none of them rising far off the substrate. Against that low-relief matrix sit two features that carry most of the cover. The first is the **lime-green to yellow-green carpet** of P3150368 and P3150384 — densely packed short, blunt branchlets spreading across metres of bottom. The second is **large tabular *Acropora***: P3150384 and P3150398 show plates up to a metre across, overlapping in tiers, and P3150371 pairs a big brown table with a cream domal *Porites* while a **checkerboard wrasse (*Halichoeres hortulanus*)** works between them.

**On that green carpet, the photographs and the point annotation disagree about genus, and it matters.** The branchlets are short, blunt-tipped and uniformly textured with **no visible axial corallites** — a morphology that reads as *Porites* (*P. rus*/*cylindrica*-type) or *Montipora*, which is what the existing description said. **The 30-point annotation of P3150368 scored eight of its points *Acropora*.** Both cannot be right, and this is a large, repeated, cover-dominant taxon at this site, so the call propagates into any genus-level summary. **Flagged for Anderson's eye.**

**The site's bleaching was missed entirely by the previous description, and it should not have been.** The existing text states "no whitened colonies evident in the reviewed photos" and carries **"Flags: None"**. **P3150385 contradicts both**: a frame-filling macro of a thick-branched *Pocillopora*-type colony that is **strongly mottled — bone-white branch tips and patches interleaved with fully pigmented brown tissue across the same colony**, textbook patchy partial bleaching, and one of the clearest such images in the Chagos set. The annotation agrees: **six bleached points, 20% of that frame, and six of the site's seven bleached points in total.** The reported site figure of 2.6% is not wrong — one strongly affected colony among a great many healthy ones is exactly what 2.6% looks like — but **"no bleaching at BIEA19" is wrong**, and P3150385 belongs with BIDI16.5/P3140329 and BIEG05/P3110481 in the partial-colony-bleaching image set.

**The site's seventh bleached point is more doubtful.** P3150400 shows a large tan digitate colony with several white branch ends — but also **a cleanly snapped white branch stub**, and the white in that frame reads as **fresh physical breakage rather than bleaching**. Score it as damage unless the full-resolution frame says otherwise.

**Mortality is present and structural.** P3150384 is the key frame: **a pale mauve-white table that has lost its polyp texture and a uniformly grey, algal-filmed table beside it, both still in growth position, alongside a living green table** — Coralscapes reads 14.9% dead coral here, its highest value at the site, with P3150400 at 8.2% and P3150394 at 5.8%. Living and dead tabular *Acropora* in one frame is a useful image for the mortality-and-persistence question. **P3150398 shows a whole field of overlapping tables in the middle distance, but they are too blue-attenuated to score** — do not read their grey as death.

**Two frames are methodology imagery.** **P3150369 shows a diver at work**, head-down at the substrate in black wetsuit and white gloves, bubbles streaming, with **a second silver cylinder lying on the reef beside them** — a spare tank or a sample container. And **P3150394 has a thin, straight, pale line running horizontally across the upper background that is compatible with a transect tape**, with a blue elongate object at the right that could be the reel. **Verify at full resolution**; if it is a tape, this site joins BIDI08, BIEG04 and BIEA18 in the photographic record of how the mission's numbers were made.

Faunally the site is quiet but distinctive: **blue-and-yellow fusiliers compatible with *Caesio*** stream through P3150368, P3150396, P3150397 and P3150398 in loose schools, a **powder-blue surgeonfish (*Acanthurus leucosternon*)** — a characteristically Indian Ocean species — sits at the centre of P3150384, and there are Moorish idol, damselfishes, dark surgeonfish and a cloud of small silvery juveniles in P3150397. **A diver's fin is visible in the top corner of P3150399.**

<b>Assemblage composition.</b> Across 15 analysed frames the Coralscapes masks put live hard coral at 21.3% of image area, of which massive and meandroid is the largest component at 32.6%; the rest breaks down as other hard coral 30.6%, tabular <i>Acropora</i> 18.5%, <i>Pocillopora</i> 6.8%, <i>Acropora</i> (non-tabular) 5% of live cover. The independent 30-point human annotation of the same frames (450 points) gives live hard coral as 18.4% of all points and apportions it <i>Acropora</i> 39.8%, other hard coral 31.3%, <i>Porites</i> 24.1%, <i>Pocillopora</i> 4.8% of live coral. Non-coral cover is dominated by crustose coralline algae 6.4%, turf 14% of all points. Structural loss registers as dead coral 2.7% of image area and rubble 5.8%. Bleaching stands at 1.6% of annotated points, concentrated in other hard coral 1.3%, <i>Pocillopora</i> 0.2% of points. Conspicuous fauna and sessile invertebrates appear as sponges in 11 frames, sea cucumbers in 3 frames, urchins in 1 frame.

**Flags:** (1) **THE PREVIOUS DESCRIPTION'S "no whitened colonies evident" AND "Flags: None" ARE BOTH WRONG.** P3150385 is an unambiguous partial-bleaching macro carrying 6 of the site's 7 bleached points. Corrected here. (2) **GENUS DISAGREEMENT ON THE COVER-DOMINANT TAXON** — the lime-green carpet in P3150368/P3150384 has blunt tips and no visible axial corallites (reads *Porites*/*Montipora*), but the annotator scored it *Acropora*. Resolve before any genus-level summary. (3) **P3150400's white is most likely BREAKAGE, not bleaching** — there is a cleanly snapped branch stub in frame. Its single bleached point is doubtful. (4) **P3150384: living and standing-dead tabular *Acropora* in the same frame**, Coralscapes 14.9% dead — pull for the mortality/persistence analysis. (5) **P3150398's distant tables are too blue-attenuated to score** — do not read grey as dead. (6) **METHODOLOGY IMAGERY: P3150369** (diver at work, second cylinder on the reef) and **P3150394** (a straight pale line compatible with a transect tape — **verify at full resolution**). (7) **NEGATIVE CONTROL FOR THE BLUE-WATER ANNOTATION PROBLEM:** four frames here are 43–63% water and **all seven bleached points fall on 0%-water frames**. Unlike BIEA18 and BIDI16.5, this site's bleaching signal is clean — which strengthens the case that those were annotator slips rather than a systematic method fault. (8) ***Millepora* record in P3150371** — first hydrocoral of the Eagle Island block (hedged). (9) **Powder-blue surgeonfish (*A. leucosternon*) in P3150384.** (10) Diver's fin in P3150399. (11) No per-frame depth data.

**Benthic data:** [Download BIEA19 benthic survey data (CSV)](../benthic_data/BIEA19_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).