### BISA91 — Salomon
**CRD page:** [Salomon Islands Coral Survey April 2015](https://coralreefdiagnostics.com/salomon-201504)

**24 Apr 2015**
**GPS:** 5.33965°S, 72.26332°E
**Setting:** protected (leeward) emergent back reef, inside Salomon lagoon, 18.2 m
**Coral cover:** 52.4% (4 depth strata over 15 formal transects, D1–D4; **algal cover 36.3%**) · **published image bleaching (`Mean(%bleaching)` column):** 12.4%
**Live hard coral cover across methods:** formal transects 52.4% · CoralNet points 27.2% · Coralscapes masks 28.5% (18 frames). These products have different sampling footprints and denominators and should not be averaged.
**Temperature:** 30.8 °C · **Salinity:** 34.2
**Photo review:** all 18 of 18 frames examined (the entire retained set for this site)
**Representative photos:** P4240862.jpg, P4240876.jpg, P4240897.jpg, P4240906.jpg, P4240913.jpg, P4240914.jpg

BISA91 is an emergent back reef inside Salomon lagoon, surveyed at 18.2 m under leeward (revised: protected) exposure, and it is the highest-cover and the warmest of the four Salomon/Speaker's Bank sites in this group. **Coral cover is 52.4% against 36.3% algae, ranking 21st of 109 Chagos sites for coral cover and a mid-table 65th of 109 for algal cover.** The site row records a water temperature of **30.8 °C** — 0.9 to 1.2 °C above the other three sites in this group and high in absolute terms for the archipelago in April 2015. That number matters here, because BISA91 also carries the highest published bleaching figure of the four: **12.4%, ranking 17th of 98 Chagos sites with a value — inside the worst-fifth of the archipelago for bleaching severity.** Within the algal budget, fleshy turf (PTTS) is 14.6% of total benthos (19th of 109), coralline algae 11.7% (69th of 109), turf-on-sand 8.0% and macroalgae only 1.9%. Bare substrate is 8.6% and non-coral invertebrates are very low at 2.7%.

Coral cover is spread across **27 nonzero genera (41st of 109 Chagos sites for genus count)**, and the assemblage is, at site level, remarkably even — five genera between 7% and 10% of the benthos with no clear leader. All values below are percent of total benthos, not percent of coral. ***Lobophyllia* leads at 10.14% (19% of all coral present)**, then *Porites* 8.55% (16%), *Montipora* 7.70% (15%), *Goniopora* 7.38% (14%) and *Acropora* 7.23% (14%) — five genera splitting 78% of the coral almost equally between them. *Galaxea* (1.55%), *Turbinaria* (1.15%), *Favites* (1.07%) and *Astreopora* (0.98%) follow, with a tail of 18 further genera including *Platygyra*, *Fungia*, *Pavona*, *Echinopora*, *Favia*, *Herpolitha*, *Halomitra*, *Ctenactis*, *Pachyseris*, *Goniastrea*, *Cyphastrea*, *Merulina*, *Hydnophora*, *Stylocoeniella*, *Millepora*, *Leptastrea*, *Isopora* and *Physogyra*.

**That site-level evenness is an artefact of averaging, and it is the most important finding at BISA91: each of the four depth strata is dominated by a different single genus, and no stratum looks like the site row.** D1 (shallowest, 2 transects) carries **58.3% coral against 39.3% algae** and is *Acropora* ground — ***Acropora* 22.80% of the benthos, 39% of the coral on that stratum** — with *Lobophyllia* 16.60%, *Turbinaria* 4.50% and *Galaxea* 3.70%. D2 (6 transects) carries **57.3% coral against 31.2% algae** and the lead passes entirely to ***Goniopora* at 20.50% (36% of the coral there)**, with *Porites* 9.30%, *Lobophyllia* 9.03% and *Acropora* falling to 3.90%. D3 (6 transects) is the highest-cover stratum at **60.2% coral against 33.3% algae**, and ***Porites* takes over at 20.30% (34% of the coral)**, with *Lobophyllia* 14.93%, *Goniopora* 6.93% and *Acropora* all but gone at 0.23%. D4 (1 transect) collapses to **33.8% coral against 41.6% algae and 24.2% bare substrate**, with only 8 nonzero genera, and is effectively a single-genus transect: ***Montipora* alone is 26.00% of the benthos there — 77% of all the coral on that transect** — with nothing else above 2.6%. The consequences for the site row are direct and should be carried downstream: **84% of the site's 7.70% *Montipora* figure comes from that one D4 transect; 79% of the site's 7.23% *Acropora* figure comes from the two D1 transects; and 69% of the site's 7.38% *Goniopora* figure comes from D2 alone.** Only *Lobophyllia* is genuinely abundant across the profile (16.60%, 9.03%, 14.93%, 0%), which makes it the one genus the site figure honestly describes. Because site cover is the unweighted mean of per-stratum means, the thin D1 (2 transects) and D4 (1 transect) strata carry half the headline number between them; transect-weighting the same data gives **57.00%** against the published **52.38%**, a 4.6-percentage-point swing.

**All 18 retained frames were examined, and they show a densely occupied, structurally varied back-reef slope of large massive, lobate and foliose colonies interspersed with branching coral, vividly pigmented invertebrates — and unambiguous bleaching.** Open sand is scarce in most frames, though crevices, older skeletal surfaces and dead branching rubble remain between colonies. **P4240849** is a shallow frame (surface haze visible at the top) of a corymbose *Acropora*-type thicket, the photographic counterpart to the *Acropora*-dominated D1 stratum. **P4240914** is the set's signature image: a single dome of orange-tan hummocky massive coral, roughly two metres across and entirely consistent with a large *Porites* bommie, with a cloud of small blue-green damselfish over it and tabular colonies at its base. **P4240903, P4240905 and P4240906** show the multi-form slope proper — tan and olive foliose plates, a large funnel-shaped plate colony, tabular *Acropora*, lavender-to-violet corymbose colonies whose blue-purple colour is normal pigmentation rather than a stress signal, and cream-and-white meandroid brain colonies with fine tightly-packed meanders. **P4240906 also carries the best fauna record in the set**: a barred orange-and-white angelfish sheltering in a crevice, most consistent with a regal angelfish, alongside small dark damselfishes; **P4240862** includes two red-and-black banded sea urchins on the framework; **P4240893 and P4240894** show a loose school of dark-bodied surgeonfish or drummer over the slope.

**The vivid red-to-pink masses that recur through this set are, on re-examination, two different things, and the distinction matters.** The earlier photo review recorded them all as "encrusting sponges or other non-coral colonial organisms" and flagged the identity as unresolved; that observation is preserved, and it is correct for some of them — the smooth, amorphous, blaze-orange masses overgrowing framework in **P4240884, P4240895 and P4240904** have no visible corallite structure and read as encrusting sponge. But the large rose-pink colonies in **P4240863, P4240896, P4240897 and P4240906** show fleshy, meandering-to-phaceloid polyps with distinct rounded mouths, a structure diagnostic of *Lobophyllia* or *Symphyllia* — and ***Lobophyllia* is the leading genus on the transects at 10.14% of the benthos**, which the sponge reading leaves unaccounted for in the imagery. **P4240897 is the decisive frame**: a single large colony in the foreground is tan-brown across one half and vivid pink across the other, with continuous polyp structure across the boundary. That is a within-colony colour transition, not two organisms, and vivid pink or purple pigmentation of that kind in mussid corals is a well-known chromoprotein paling response rather than normal colouration or a sponge. Non-coral invertebrates are only 2.7% of the benthos on the transects, which is consistent with sponges being a minor, if visually prominent, component here.

**Bleaching at BISA91 is real, visible at colony scale, and consistent with the high published figure.** The published value is **12.4% (`Mean(%bleaching)` column; 17th of 98 Chagos sites with a value)**, and the two CoralNet runs agree to four decimal places, so this is not one of the ten divergent bleaching-onset rows. Four independent lines of photographic evidence support it. **P4240913** shows a compact branching colony that is uniformly bone-white against dark pigmented framework and a normally-pigmented tan corymbose colony immediately beside it — the clearest single bleaching record in the set, and preserved from the earlier review at high confidence. **P4240862** shows a large, brilliantly white smooth encrusting/massive colony beside fully pigmented brown massive colonies and a pigmented brain coral — a second unambiguous record, and one the earlier review did not capture. **P4240876** shows a large colony with fully expanded polyps, extremely pale across most of its surface while retaining warmer beige tissue centrally; the expanded polyps establish that the white area is living tissue rather than exposed skeleton, and the growth form is consistent with the *Goniopora* recorded at 7.38% of benthos. This was flagged at high confidence by the earlier reviewer and is confirmed here. Beyond those, **pale-to-whitened colonies are scattered widely through the wide-angle frames** — P4240884, P4240895, P4240903 and P4240904 each carry several mint-green, cream or stark-white colonies among pigmented neighbours, and P4240895 shows a broad field of pale grey-white plate and foliose colonies filling the upper slope. Colour correction and strobe lighting make individual wide-frame calls less secure, so those are not counted as confirmed colonies, but the pattern across the set is consistent. Taken with the 30.8 °C water temperature and the pink chromoprotein response described above, the imagery reads as a reef in active thermal stress at the time of the dive. **No crown-of-thorns starfish, diagnostic disease band, active tissue-sloughing front or discrete feeding scar with a visible predator was identified in any of the 18 frames.** Dead branching rubble and pale exposed patches on some massive colonies (P4240863 in particular) may represent older partial mortality, but their age and cause cannot be determined from a single visit.

<b>Assemblage composition.</b> Formal transects give 52.4% coral against 36.3% algae — turf-led, with very low non-coral invertebrate cover — across 27 nonzero genera whose site-level ordering is unusually even: *Lobophyllia* 10.1%, *Porites* 8.6%, *Montipora* 7.7%, *Goniopora* 7.4% and *Acropora* 7.2% of the benthos. That evenness dissolves at depth: D1 is *Acropora*-dominated (22.8%), D2 *Goniopora*-dominated (20.5%), D3 *Porites*-dominated (20.3%) and the single D4 transect is 77% *Montipora*, with only *Lobophyllia* abundant across the profile. The 18 photographs corroborate the mixed massive/lobate/foliose/branching structure, add a large *Porites*-type bommie, rose-pink mussid colonies consistent with the leading genus, encrusting sponges, blue-green damselfish and a regal angelfish — and document bleaching at colony scale in at least three frames, alongside a within-colony pink chromoprotein transition, on a reef recorded at 30.8 °C.

**Flags:** (1) **BLEACHING — this site should be entered in the running COTS/bleaching/disease log.** Three colony-scale records: **P4240913** (bone-white compact branching colony, high confidence, carried over from the earlier review), **P4240862** (large brilliant-white encrusting/massive colony beside pigmented neighbours, high confidence, NEW in this rebuild), and **P4240876** (large polyp-expanded colony, consistent with *Goniopora*, pale over most of its surface with living polyps visible, high confidence, carried over). Widespread pale colonies in P4240884, P4240895, P4240903 and P4240904 at medium confidence. Site temperature 30.8 °C; published bleaching 12.4%, 17th of 98. (2) **Identification correction, offered for adjudication.** The earlier review recorded all vivid red-pink masses at this site as sponges "or another non-coral colonial organism". This re-review separates them: the smooth amorphous blaze-orange masses (P4240884, P4240895, P4240904) do read as encrusting sponge, but the large rose-pink colonies in P4240863, P4240896, P4240897 and P4240906 show fleshy meandroid polyps with distinct calices and are consistent with *Lobophyllia*/*Symphyllia* — the site's leading genus at 10.14% of benthos. **P4240897 shows one colony that is tan over half its surface and vivid pink over the other half with continuous polyp structure across the boundary**, which is a chromoprotein paling response, not two organisms. Non-coral invertebrate cover on the transects is only 2.7%, consistent with sponges being visually prominent but quantitatively minor. (3) **Two strata rest on 1 and 2 transects.** D4 has a single transect (33.8% cover, 24.2% bare substrate) and D1 has two, yet the site figure weights all four strata equally. Transect-weighting gives **57.00%** against the published **52.38%**. (4) **Three of the four site-level genus figures are dominated by one thin stratum**: 84% of site *Montipora* comes from the single D4 transect, 79% of site *Acropora* from the two D1 transects, 69% of site *Goniopora* from D2. The site row's apparent five-genus evenness is an averaging artefact and should not be quoted as an assemblage description. Use the depth rows. (5) **Genus sum falls short of published cover.** The 27 genus columns sum to **52.258%** against a published cover of **52.383%** — a 0.125 pp shortfall (0.24% of the coral). The same shortfall appears within strata: D2 57.13 vs 57.27, D3 60.00 vs 60.17, D4 33.60 vs 33.80. No unmapped genus column carries the difference. Small, but unreconciled — reported rather than silently absorbed. (6) **Multi-method deviation is large**: CoralNet (27.2%) and Coralscapes (28.5%) sit 23.9–25.2 pp below the transect figure of 52.4%. Do not average them; the transect value is ground truth for cover. The 18-frame scored subset happens to equal the full folder at this site. (7) **The CoralNet points figure is a CV product, not verified ground truth** — point-annotation output corpus-wide is roughly 76% unconfirmed machine output. (8) **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`). Named column: `Mean(%bleaching)`, which equals the 1st CoralNet training run; the two runs agree here (12.3745 vs 12.3745), so BISA91 is not one of the ten divergent rows despite being a bleaching-onset dive. Any rank off that column is approximate — it carries at least one known bad value (BITB58 listed at 22.22% against a corrected 0.77%). At this site, unusually, the independent photographic evidence supports the model's high reading. (9) **No COTS, disease or predation scarring was observed** across the full 18-frame review — an explicit absence-of-evidence finding. Dead branching rubble in P4240863 indicates prior mortality of unknown age and cause. (10) **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 **12.37%, rank 24 of 110** (1st training: 12.37%, rank 21 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 BISA91 benthic survey data (CSV)](../benthic_data/BISA91_benthic.csv) — scleractinian cover by genus with full names, plus substrate and algal categories, from the LOF BIOT-1 survey (Mayfield & Dempsey 2026, *Oceans*, supplemental file S1).