### BISP115 — Speaker's Bank
**CRD page:** [Speaker's Bank Coral Survey Data](https://coralreefdiagnostics.com/speakers-bank)

**03 May 2015**
**GPS:** 4.91949°S, 72.43696°E
**Setting:** protected submerged bank reef inside the lagoon, 20.1 m
**Coral cover:** 42.2% (2 depth transects, D3–D4; **algal cover 44.8%**) · **published image bleaching: the cell is BLANK** — one of 11 of 109 Chagos sites with no `Mean(%bleaching)` value. No substitute computed. **The photographs nonetheless show unambiguous, colony-scale bleaching (see below).**
**Live hard coral cover across methods:** formal transects 42.2% · **no CoralNet or Coralscapes product exists for this site** (BISP115 has no row in `chagos_multimethod_cover_by_site.csv`). These products have different sampling footprints and denominators and should not be averaged — including any CV figure generated for this site later.
**Temperature:** 29.795 °C · **Salinity:** 34.11
**Photo review:** all 20 of 20 stills examined (folder `BiSp115-finished`; single dive, 11:00:10–11:39:02 local from `DateTimeOriginal` EXIF, Olympus TG-2)
**Representative photos:** P5030910.jpg, P5030933.jpg, P5030912.jpg, P5030907.jpg, P5030918.jpg

BISP115 is a submerged bank reef inside the Speaker's Bank lagoon at 20.1 m, surveyed on **3 May 2015 — the latest date in this batch by twelve days**. It carries **42.2% coral against 44.8% algae, ranking 44th of 109 Chagos sites for coral cover and 35th of 109 for algal cover** — both mid-pack. The algal budget is dominated by crustose coralline algae: PCCATS alone is **25.3% of total benthos, 10th-highest of 109 sites**, against fleshy turf (PTTS) 10.3% (50th of 109), turf-plus-sediment 6.65%, macroalgae (PMTS) 2.35%, cyanobacteria 0.2% and erect algae (PETS) **0.0%**. Bare substrate is 8.45% and non-coral invertebrates 4.55%; coral, algae and the two residual categories sum to exactly 100.0% of benthos. **The macroalgae and erect-algae figures are hard to reconcile with the imagery and are flagged below** — several frames are dominated by a dense brown strap-bladed algal carpet that the transects appear not to have crossed.

Coral cover is concentrated in **only 14 nonzero genera — 98th of 109 for genus richness, in the bottom decile of the archipelago** — and the 14 genus percentages sum to exactly 42.20% of the benthos, matching the published cover figure. All values below are percent of total benthos, not percent of coral. *Porites* leads at 15.70% (37% of all the coral present), then *Acropora* 11.50% (27%), *Montipora* 4.60% (11%), *Platygyra* 4.10% (10%), *Pocillopora* 2.90% (7%) and *Stylophora* 2.25% (5%); **those six genera hold 98% of the coral at the site.** The remaining eight are trace: *Millepora* 0.30%, *Montastrea* 0.20%, *Psammocora* 0.20%, *Ctenella* 0.15%, *Cyphastrea* 0.15%, and *Favia*, *Goniastrea* and *Leptastrea* at 0.05% each. The Chagos-region endemic *Ctenella* is present but barely.

**Only two depth transects exist here, and they disagree with each other about almost everything — including which genus dominates.** D3 carries 36.4% coral against 49.6% algae and is emphatically ***Acropora*-led: that genus alone reaches 20.00% of the benthos, 55% of all the coral on the transect**, with *Porites* 7.40%, *Pocillopora* 2.80%, *Montipora* 2.20%, *Stylophora* 1.50% and *Millepora* 0.60% behind it. D4, the deeper station, **carries more coral, not less — 48.0% against 40.0% algae — and the assemblage inverts completely: *Porites* takes the lead at 24.00%** (50% of that transect's coral), followed by *Platygyra* 8.00%, *Montipora* 7.00%, and then *Acropora*, *Pocillopora* and *Stylophora* tied at 3.00% each. **From D3 to D4, *Acropora* falls from 20.0% of benthos to 3.0% while *Porites* rises from 7.4% to 24.0% — a near-perfect exchange of dominance across a single depth step**, with *Platygyra* going from 0.2% to 8.0% alongside it. The site-level ordering that puts *Porites* (15.7%) ahead of *Acropora* (11.5%) is therefore **entirely an artefact of the deep transect**: at D3, *Acropora* outweighs *Porites* by nearly three to one. With only two transects there is no third station to break the tie. Coral cover rising with depth is itself unusual in this dataset and, at n = 2, is not established as a gradient.

The photographs — all 20 stills in the folder — document a broad, low-relief lagoonal plain of pale encrusted pavement and rubble carrying a discontinuous veneer of acroporid tables and corymbose colonies. **P5030891**, **P5030893** and **P5030908** are the wide establishing views: hazy blue water over an almost flat expanse of rubble and turf-filmed hardground, with tabular colonies standing as isolated features rather than forming a canopy. **P5030901**, **P5030918**, **P5030931** and **P5030932** show the coral-abundant end of the same plain — dense fields of tan, yellow-green and deep-mauve corymbose and tabular *Acropora*, brown foliose discs consistent with *Turbinaria* or *Pavona*, and scattered massive colonies. **P5030918** is the best single structural portrait at the site and also shows the widest colour range, from deep purple-mauve corymbose acroporids to pale cream tables in the same frame. **P5030934** and **P5030927** are vertical frames over sparser ground, showing a brown meandroid brain colony consistent with *Platygyra* or *Ctenella* beside encrusting plates — the *Platygyra* signal that is 8.0% of the benthos at D4. **P5030928** is a single large yellow-green tabular plate on a narrow stalk over otherwise open pavement. Growth-form calls from photographs are given as "consistent with"; the genus percentages above are transect identifications.

**Condition is the finding at BISP115, and the published record cannot carry it because the bleaching cell is blank.** The 20-frame set contains **unambiguous, colony-scale bleaching in at least five frames**, with strong within-frame contrast that exposure or strobe artefact cannot explain:

- **P5030910** — a large tabular *Acropora* that is **uniformly bone-white across its entire surface**, sitting directly on top of a brown, turf-covered dead table of similar size, with pigmented colonies elsewhere in the same frame. The clearest single bleaching image in this batch.
- **P5030933** — a large *Pocillopora*- or *Stylophora*-type colony, **cream-white over every branch to the tips**, immediately beside a fully pigmented dark brown massive colony and a pigmented yellow-green corymbose acroporid. Within-frame controls on both sides.
- **P5030912** — a stark white branching *Acropora* colony sheltering under a fully pigmented tan tabular *Acropora*, with a second small white colony lower right and pigmented lavender *Pocillopora* alongside.
- **P5030926** — a bone-white branching *Acropora* directly abutting a pigmented tan colony of the same growth form, with humbug damselfish sheltering in the white branches.
- **P5030911** — a wide view in which **numerous small white corymbose colonies are scattered across the plain among pigmented ones**, indicating this is not a handful of individuals but a distributed condition across the site.

Additional pale colonies appear in **P5030907**, **P5030930**, **P5030931** and **P5030932**. Two of these carry a caveat: the pale plate in **P5030930** has a dark, algae-filmed centre consistent with a colony that is dead rather than currently bleached, and the large pale-yellow tabular colony in **P5030931** is pale but not white and could be natural pigmentation; both are recorded as ambiguous. Against that, **P5030929** — a close-up of an olive digitate colony with bright white branch tips — is explicitly **not** counted: pale apical tips over pigmented tissue are normal growth and are the single most common false-positive morphology in this corpus. **No crown-of-thorns starfish, no *Drupella* aggregation, no disease lesion or tissue-loss margin, and no predation scarring was seen in any of the 20 frames.** These observations are logged as frame-specific bleaching records; **no site bleaching percentage has been computed from them**, and the blank published cell has not been filled.

Fauna is sparse relative to the other sites in this batch — scattered damselfish, butterflyfish, a moorish idol in **P5030933**, small planktivores over the plain. A diver appears in **P5030900** and bubble trails in **P5030893**; standard field documentation.

<b>Assemblage composition.</b> Formal transects give 42.2% coral against 44.8% algae — the algal side dominated by crustose coralline algae at 25.3% of benthos, 10th-highest of 109 sites — across only 14 nonzero genera, 98th of 109 for richness, with *Porites* (15.7%) and *Acropora* (11.5%) holding two-thirds of the coral and six genera holding 98% of it. The two-transect depth profile inverts between stations: D3 is *Acropora*-dominated at 20.0% of benthos against *Porites* 7.4%, while the deeper and higher-cover D4 reverses that to *Porites* 24.0% against *Acropora* 3.0%, so the site-level genus order describes neither transect. The photographs document a flat lagoonal plain of tables, corymbose acroporids and massive *Porites* over encrusted pavement — and, in at least five frames, **unambiguous colony-scale bleaching on 3 May 2015 at a site whose published bleaching cell is blank.**

**Flags:** (1) **THE PUBLISHED BLEACHING CELL IS BLANK, AND THE PHOTOGRAPHS SHOW UNAMBIGUOUS BLEACHING.** `Mean(%bleaching)` is empty for BISP115 — it is one of 11 of 109 Chagos sites with no published value — so this site contributes **nothing** to any archipelago bleaching statistic, any date-binned bleaching series, or any bleaching map layer. Yet P5030910, P5030933, P5030912, P5030926 and P5030911 record bone-white colonies with pigmented within-frame controls, and the survey date is **3 May 2015**, the latest in this batch. **A site that is invisible to the published bleaching record while carrying the clearest photographic bleaching evidence in the batch is a real gap, not a rounding error.** No substitute figure has been computed here and none should be back-filled without a deliberate, documented pass. **This is the flag to act on.** (2) **NO MULTI-METHOD ROW EXISTS FOR THIS SITE.** BISP115 is absent from `chagos_multimethod_cover_by_site.csv` — no CoralNet points, no Coralscapes mask, no `n_photos` subset. The 42.2% transect figure stands alone; do not pair it with the folder frame-count (20) as a CV sample size. The missing CV product is also the likely reason flag 1 exists, the published bleaching column being itself a CoralNet product. (3) **ONLY TWO DEPTH TRANSECTS EXIST (`N Rows` = 2) — the thinnest replication possible**, and the two disagree about which genus dominates (*Acropora* 20.0% vs 3.0%; *Porites* 7.4% vs 24.0% of benthos, D3 to D4). Every site-level genus percentage here is a two-point average across a dominance reversal and should not be quoted as a description of the assemblage without the depth split. (4) **No D1 or D2 transect exists** — the profile starts at D3, so the entire shallow half of the reef at this site is unsampled. (5) **THE MACROALGAE FIGURES CONFLICT WITH THE IMAGERY.** The transects record macroalgae (PMTS) at **2.35%** and erect algae (PETS) at **0.0%** of benthos, but **P5030907**, **P5030930** and **P5030900** are dominated by a dense, continuous carpet of brown strap-bladed macroalgae covering most of the frame. Either the transect lines did not cross those beds, or the blades were scored into a turf category. Worth resolving before any BISP115 macroalgal figure is used; the discrepancy is large enough to matter and is not reconciled here. (6) **Genus richness is low and may be effort-limited.** 14 genera on 2 transects, against sites in this corpus carrying 30+ genera on 5 or more. This is plausibly a richness floor set by sampling effort rather than a depauperate assemblage; the 98th-of-109 rank is real but should be read with the transect count attached. (7) **The depth codes D3–D4 carry no metre boundaries** in `Chagos benthic data by site-depth (n=336 rows).csv`; they are ordered shallow-to-deep stations, and the 20.1 m in the header is the site-level survey depth, not the depth of any one transect. (8) **Exposure was revised across two classes.** `exposure-original` records this site as **windward**; `exposure-revised` gives **protected**, and the header above uses the revised value. That is a two-step revision on a site inside a lagoon, and anyone pulling the original column will get the opposite exposure class. (9) **Two pale colonies are recorded as ambiguous, not as bleaching**: P5030930 (pale plate with a dark algae-filmed centre — possibly dead rather than bleached) and P5030931 (large pale-yellow table — possibly natural pigmentation). **P5030929 is explicitly excluded**: white apical tips over pigmented tissue are normal *Acropora*/pocilloporid growth, and this morphology is the commonest false-positive in the Chagos point-annotation set. (10) **No COTS, no disease, no predation scarring was observed across the full 20-frame review** — an explicit absence-of-evidence finding, not an omission. (11) **Where the published bleaching figure does exist elsewhere in Chagos it is a COMPUTER-VISION product, not an in-situ observation** — 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`). That column also carries at least one known bad value (BITB58 listed at 22.22% against a corrected 0.77%), so any rank taken off it is approximate. Never read those percentages as a diver's field estimate. (12) **A diver and bubble trails appear in P5030900, P5030893** — standard field documentation, not a data issue.

**Benthic data:** [Download BISP115 benthic survey data (CSV)](../benthic_data/BISP115_benthic.csv) — scleractinian cover by genus with full names, plus substrate and algal categories (bare substrate, CCA, turf, turf+sediment, macroalgae, erect algae, invertebrates, cyanobacteria), from the LOF BIOT-1 survey (Mayfield & Dempsey 2026, *Oceans*, supplemental file S1).