### BISA92 — Salomon
**CRD page:** [Salomon Islands Coral Survey April 2015](https://coralreefdiagnostics.com/salomon-201504)

**25 Apr 2015**
**GPS:** 5.32886°S, 72.28001°E
**Setting:** exposed (windward) emergent forereef, outside Salomon lagoon, 34.0 m
**Coral cover:** 24.5% (5 depth strata over 19 formal transects, D1–D5; **algal cover 41.3%**) · **published image bleaching (`Mean(%bleaching)` column):** 9.1%
**Live hard coral cover across methods:** formal transects 24.5% · CoralNet points 27.5% · Coralscapes masks 29.5% (21 frames). These products have different sampling footprints and denominators and should not be averaged.
**Temperature:** 29.8 °C · **Salinity:** 34.3
**Photo review:** 20 of 39 frames re-examined for this rebuild — a systematic every-second-frame sample spanning the full sequence (P4250843, 845, 852, 854, 856, 858, 860, 867, 876, 885, 892, 901, 904, 911, 913, 915, 917, 920, 922, 924). An earlier reviewer worked the complete 39-frame set, and every observation from that pass is preserved below.
**Representative photos:** P4250843.jpg, P4250860.jpg, P4250885.jpg, P4250901.jpg, P4250911.jpg, P4250924.jpg

BISA92 is the deepest site in the Chagos 2015 set at **34.0 m**, on a windward (revised: exposed) emergent forereef outside Salomon lagoon. It is also the clearest example in this group of a reef whose visual grandeur and whose benthic-cover number point in opposite directions: **coral cover is only 24.5%, ranking 87th of 109 Chagos sites — bottom quartile — while genus richness is 31 nonzero genera, 13th-highest of all 109 sites.** The reason the frame looks full is not coral. **Non-coral invertebrates (PITS) occupy 26.0% of total benthos here — more than the coral does** — and crustose coralline algae add a further 20.6% (21st of 109 sites). Algal cover overall is 41.3% (47th of 109), of which turf is 10.5%, encrusting algae 4.6%, turf-on-sand 3.5% and macroalgae only 2.1%. Bare substrate is 8.3%. In short, this is a deep, exposed wall whose surfaces are held by sponges, ascidians, gorgonians and coralline pavement, with scleractinian coral present as a large number of small, scattered colonies rather than as continuous cover.

The 31 nonzero genus percentages sum to 24.474% of the benthos, an exact match to the published cover figure. All values below are percent of total benthos, not percent of coral. ***Porites* leads at 7.00% of the benthos — 29% of all coral present** — followed by *Pocillopora* 3.71% (15%), *Montipora* 2.23% (9%), *Acropora* 1.44% (6%) and *Pavona* 1.17% (5%); those five hold 64% of the coral. *Stylophora* (1.17%), *Pachyseris* (1.08%), *Favites* (1.07%) and *Montastrea* (0.99%) follow, and the long tail runs to *Favia*, *Hydnophora*, *Platygyra*, *Astreopora*, *Millepora*, *Galaxea*, *Echinopora*, *Isopora*, *Gardineroseris*, *Leptastrea*, *Leptoseris*, *Physogyra*, *Goniastrea*, *Heliopora*, *Goniopora*, *Cyphastrea*, *Lobophyllia*, *Mycedium*, *Fungia*, *Symphyllia*, *Acanthastrea* and *Oulophyllia*. That tail is the site's real distinction: 22 genera below 1% of the benthos, on a reef with a quarter of the cover of the best Chagos sites.

**The five depth strata do not describe a gradient — cover rises, falls, falls again and then rebounds, and the deepest stratum rests on a single transect that supplies most of three site-level genus figures.** D1 (2 transects) is the poorest stratum at **18.0% coral against 57.2% algae and 24.6% non-coral invertebrates**, with *Porites* 6.00%, *Montipora* 3.10%, *Pocillopora* 2.80% and *Millepora* 2.10% — and *Millepora* occurs essentially nowhere else on the profile, so 99% of the site's 0.43% *Millepora* figure comes from these two shallow transects. Cover peaks at D2 (6 transects) with **35.3% coral against 42.8% algae**, the most coral-abundant stratum, led by *Porites* 15.10%, *Acropora* 4.87% and *Stylophora* 4.57% — and 78% of the site's 1.17% *Stylophora* comes from D2 alone. D3 (7 transects, the best-replicated stratum) falls back to **23.4% coral against 37.9% algae**, and carries **35.7% non-coral invertebrate cover — the highest of any stratum at any of these four sites**; the coral lead passes from *Porites* to *Pocillopora* (4.69%, tied with *Porites* at 4.69%). D4 (3 transects) is the trough at **17.7% coral against 48.6% algae and 26.4% invertebrates**, with *Porites* 6.20%, *Favites* 3.27% and *Pavona* 2.20%. D5, a **single transect**, then rebounds to **28.0% coral with only 20.0% algae but 29.0% bare substrate and 23.0% invertebrates**, and carries just 8 genera in a completely different order — ***Pocillopora* 9.00% (32% of the coral there), *Montastrea* 4.00%, *Pachyseris* 4.00%, *Hydnophora* 3.00%.** That one transect supplies **100% of the site's *Hydnophora* (0.60%), 81% of its *Montastrea* (0.99%) and 74% of its *Pachyseris* (1.08%)**. The overall effect on the headline number is modest — transect-weighting the same data gives 25.92% against the published 24.47%, a 1.5-percentage-point swing, the smallest of the four sites in this group — but the effect on the genus ordering is severe, and any downstream use of BISA92 genus proportions should work from the depth rows.

**The photographs are a wall set, and they explain the numbers directly.** Twenty frames were re-examined here across the full sequence; the earlier full-set pass is preserved. **P4250843, P4250852 and P4250854** establish the setting: a steep, high-relief face dropping into deep blue water, with multiple large gorgonian sea fans a metre or more across projecting from the wall, pink and gold whip corals, green foliose plate colonies, coralline-encrusted framework and small scattered hard corals. **P4250845** adds black wire/whip corals (antipatharian or gorgonian) in the foreground of a shaded, mauve-lit section — the earlier review's "black branching hydroid- or octocoral-like growth", confirmed here. **P4250860 is the single most diagnostic frame in the set for the benthic numbers**: a shaded wall panel carpeted edge to edge in salmon-pink clustered tubular organisms — sponges or colonial ascidians, not safely separable from a still — interleaved with red encrusting sponge, chartreuse-yellow lobed sponge and purple encrusting growth, with essentially no scleractinian coral visible. That frame is the 26.0% non-coral invertebrate figure made visible, and it is why a spectacular wall carries bottom-quartile coral cover. **P4250858, P4250867 and P4250904** show the same surfaces at wider scale, with conspicuous smooth white lobed masses that read as white sponges rather than bleached coral (no corallite structure resolvable). **P4250892, P4250922 and P4250924** move onto the reef top and shoulder, where the character changes entirely: P4250922 is a flat sand-and-rubble plain with widely spaced small colonies and pale yellow-green ruffled foliose growth, and P4250924 is a broad reef-top view of moderate cover in mixed small colonies — mauve *Pocillopora*-type heads, cream corymbose colonies, orange-brown massive mounds — with a diver silhouetted in the upper right. **P4250911** gives the best genus-level photographic anchor at the site: a large tan laminar plate colony with fine concentric surface ridging and white growing margins, consistent with the *Pachyseris* the transects record at depth.

**Fish and macrofauna are the other reason this folder is labelled "awesome site" — a field note, not a formal condition category.** Anthias swarm the wall in **P4250845, P4250854, P4250876, P4250885 and P4250904**, with P4250885 showing a cloud of several thousand individuals hanging off the drop-off; **P4250858 and P4250913** show dense aggregations under overhangs, P4250913 a shimmering cloud of golden glassy sweepers; **P4250867** carries a school of dark-bodied, pale-tailed planktivores off the wall edge; **P4250901** is a dark overhang packed with silversides. **P4250892** holds two bannerfish and several red soldierfish- or squirrelfish-type individuals. **P4250917 records a moray eel** — a dark-headed *Gymnothorax*-type animal looking out from among foliose colonies and a large pale meandroid mussid — a fauna record not captured in the earlier pass. **P4250915 and P4250920** are macro frames on the same encrusting *Porites*-type pavement, each showing a banded cleaner shrimp with long white antennae, small crevice fishes, a patch of green bubble alga (*Caulerpa*-type) and a fungiid or mussid disc; the earlier review's note that P4250916–P4250921 revisit small crevice animals and coral polyps is confirmed. **P4250876 and P4250861 show divers and the survey line**, tying the scenic imagery to an active survey dive. These images document habitat use, not standardised counts, and neighbouring frames may repeat the same schools.

**Condition: localised bleaching, plus one observation worth separating out.** The published image bleaching value is **9.1% (`Mean(%bleaching)` column; 35th of 98 Chagos sites with a value — upper-middle of the archipelago for bleaching severity)**, and the two CoralNet runs agree to four decimal places, so this is not one of the divergent bleaching-onset rows. The earlier review's two bleaching flags are preserved: **P4250924**, where several compact branching colonies are conspicuously pale to white among more pigmented neighbours (high confidence — this re-review concurs; the pale colonies sit beside normally pigmented mauve and orange-brown colonies in the same light), and **P4250845**, where broad plate-like colonies are pale (medium confidence — kept at medium, because in a shaded, blue-lit wall frame old mortality, coralline overgrowth and illumination are all live alternatives). Small white patches in P4250911 and P4250915 remain too localised and morphologically ambiguous to classify. **Separately, P4250901 shows two large sea anemones with stark white tentacles** beside the sweeper aggregation. Anemone bleaching is a genuine thermal-stress observation and is logged as such, but anemones are not scleractinian and contribute to neither the coral-cover nor the published bleaching figures, so this is recorded as an independent condition signal rather than as support for the 9.1%. **No crown-of-thorns starfish, advancing disease band, tissue-sloughing margin or discrete fresh feeding scar linked to a predator was identified** in the 20 frames re-examined here, and the earlier full-set pass reported the same across all 39.

<b>Assemblage composition.</b> Formal transects give 24.5% coral — bottom-quartile for Chagos — against 41.3% algae, 26.0% non-coral invertebrates and 8.3% bare substrate, yet across 31 nonzero genera, the 13th-highest genus count of the 109 sites. *Porites* leads at 7.0% of benthos (29% of coral), then *Pocillopora* 3.7%, *Montipora* 2.2%, *Acropora* 1.4% and *Pavona* 1.2%, with 22 genera below 1%. The five-stratum profile is non-monotonic — 18.0%, 35.3%, 23.4%, 17.7%, 28.0% — peaking at D2 and troughing at D4, with invertebrate cover peaking at 35.7% at D3 and the single D5 transect carrying a wholly different, *Pocillopora*/*Montastrea*/*Pachyseris*/*Hydnophora* assemblage. The photographs match this precisely: a gorgonian- and sponge-dominated deep wall with large sea fans, antipatharians and a sponge/ascidian-carpeted panel (P4250860), scattered small hard corals, a flatter low-cover reef top, exceptional planktivore aggregations, a moray, cleaner shrimps, and localised pale-to-white branching colonies at the reef top.

**Flags:** (1) **`n_photos` = 21 in the multi-method table against 39 frames in the folder.** The CoralNet (27.5%) and Coralscapes (29.5%) figures rest on that 21-frame scored subset, not on the full folder, and the two counts must never be printed side by side. Origin of the 18-frame difference is unresolved — plausibly the macro, fish-school and diver frames were excluded from scoring, which would bias the scored subset toward benthic views. (2) **This is the only site of the four where the CV products read HIGHER than the transects** (+3.0 to +5.0 pp: transects 24.5%, CoralNet 27.5%, Coralscapes 29.5%). At the other three sites CV under-reads by 12–25 pp. Still do not average them; the transect value is ground truth for cover. The direction reversal on a low-cover, invertebrate-dominated deep wall is worth a look from whoever is auditing CV-vs-transect bias — it is consistent with the models scoring sponge, ascidian and coralline surfaces as coral. (3) **The single D5 transect supplies most of three site-level genus figures**: 100% of *Hydnophora* (0.60%), 81% of *Montastrea* (0.99%) and 74% of *Pachyseris* (1.08%). *Millepora* (0.43%) is 99% from the two D1 transects, and *Stylophora* (1.17%) is 78% from D2. Use the depth rows, not the site row, for any genus-level statement about BISA92. (4) **Cover swing on depth weighting is small here**: 25.92% transect-weighted against 24.47% published, 1.5 pp — the mildest of the four sites in this batch, despite D5 being a single transect, because the strata are close in value. (5) **BLEACHING — for the running log.** P4250924, several pale-to-white compact branching colonies among pigmented neighbours, high confidence (carried over and confirmed). P4250845, broad pale plate colonies, medium confidence (carried over, kept at medium). **P4250901: two large sea anemones with stark white tentacles — anemone bleaching, NEW in this rebuild.** Anemones are not scleractinian and do not enter the coral-cover or published-bleaching figures; logged as an independent thermal-stress observation. (6) **Non-coral invertebrate cover of 26.0% site-wide, peaking at 35.7% at D3, exceeds the coral cover at every stratum except D2.** This is the defining benthic fact at BISA92 and is directly documented in P4250860. It also means "percent of benthos" and "percent of coral" diverge more sharply here than at any other site in this group. (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 (9.0714 vs 9.0714). 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%). (9) **Photo review was a 20-of-39 systematic sample for this rebuild**, combined with a preserved earlier full-set pass. Frames not re-examined here: P4250844, 851, 853, 855, 857, 859, 861, 868, 878, 891, 893, 903, 905, 912, 914, 916, 918, 921, 923. (10) **Divers and a survey line are visible in P4250861 and P4250876** — standard field documentation, not a data concern. (11) **The folder name "BiSa92-awesome site-finished" is a field note**, used here as an observation about the dive and not as a formal condition class. (12) **No COTS, disease or predation scarring was observed** — an explicit absence-of-evidence finding across the 20 frames re-examined and the earlier 39-frame pass. (13) **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 **9.07%, rank 43 of 110** (1st training: 9.07%, rank 42 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 BISA92 benthic survey data (CSV)](../benthic_data/BISA92_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).