### BIBL83 — Blenheim
**CRD page:** [Blenheim, Nelson Island & Victory Bank Coral Surveys (Apr 2015)](https://coralreefdiagnostics.com/new-page-4)

**22 Apr 2015**
**GPS:** 5.21858°S, 72.49256°E
**Setting:** exposed submerged forereef, 33 m
**Coral cover:** 33.1% (16 formal transects, D2–D4; **algal cover 38.5%**) · **published image bleaching (`Mean(%bleaching)` column):** 2.98%
**Live hard coral cover across methods:** formal transects 33.1% · CoralNet points 24.8% · Coralscapes masks 21.6% (16 frames). These products have different sampling footprints and denominators and should not be averaged.
**Temperature:** 29.7 °C · **Salinity:** 34.4
**Photo review:** all 16 of 16 jpg stills examined (camera timestamps 11:04:51–11:50:31, a single continuous dive; no sub-dive folders exist for this site)
**Representative photos:** P4220911.jpg, P4220916.jpg, P4220923.jpg, P4220944.jpg, P4220962.jpg

BIBL83 is the deepest of the Blenheim stations at 33 m, an exposed submerged forereef carrying **33.1% coral against 38.5% algae — 69th of 109 Chagos sites for coral cover and 59th of 109 for algal cover**, so squarely mid-table on both axes. What is not mid-table is the taxonomic breadth: **31 nonzero genera, 12th-highest of the 109 Chagos sites**, and the 31 genus percentages sum to 33.144% of the benthos, an exact match to the published cover figure with no shortfall to reconcile. The substrate budget is dominated by things that are neither coral nor fleshy algae: non-coral invertebrates (PITS) run to 24.2% of the benthos, crustose coralline algae (PCCATS) to 20.3% (23rd of 109), turf (PTTS) to 10.6% (47th of 109), and bare substrate is only 4.1%. Sediment-laden turf adds 5.1%, macroalgae 2.2%, erect algae 0.36% and cyanobacteria 0.08%. This is a hard, fully occupied, well-grazed deep forereef — very little of it is open ground, and the "algal" third of the benthos is more than half coralline pavement rather than weed.

All genus values below are **percent of total benthos, not percent of coral**. *Porites* leads at 8.04% of the benthos (24% of the coral present), then *Pocillopora* 5.23% (16%), *Favia* 2.98%, *Stylophora* 2.66%, *Goniastrea* 2.06%, *Platygyra* 1.88%, *Favites* 1.33%, *Pachyseris* 1.10%, *Pavona* 1.09% and *Montipora* 1.00%. **The striking absence is *Acropora*, at just 0.77% of the benthos — 2.3% of all coral recorded, and ranked below *Gardineroseris* at this site.** On a shallow Chagos forereef *Acropora* is normally the leading genus; at 33 m on this exposed slope it is a bit player, and the assemblage is instead built from massive, submassive and encrusting-plating growth forms. A tail of 19 further genera — *Gardineroseris*, *Echinopora*, *Symphyllia*, *Galaxea*, *Psammocora*, *Astreopora*, *Leptastrea*, *Montastrea*, *Leptoseris*, *Fungia*, *Mycedium*, *Heliopora*, *Isopora*, *Turbinaria*, *Acanthastrea*, *Cyphastrea*, *Lobophyllia*, *Podabacia*, *Scolymia* and *Caulastrea* — each sits under 1% of the benthos but is what carries the site into the top eighth of the archipelago for genus count.

**The three depth strata are not a gradient but a dip: cover falls at the middle stratum and recovers at the deepest, and the leading genus changes completely across the profile.** D2 (n = 6 transects) carries 37.03% coral against 38.87% algae and is decisively *Porites*–*Pocillopora* ground: *Porites* 13.73% and *Pocillopora* 10.47% together account for 65% of all coral on that stratum, with *Favia* 3.10%, *Stylophora* 2.53%, *Acropora* 1.13% and *Platygyra* 1.03% behind them, across 20 genera. D3 (n = 6) drops to **25.40% coral — the lowest of the profile — against the highest algal cover, 44.5%**, and simultaneously carries the *most* genera of any stratum (25). **The dominance structure inverts here: *Porites* collapses from 13.73% to 1.33% and falls to 7th place, while *Stylophora* takes the lead at 4.70%**, followed by *Favia* 3.73%, *Pocillopora* 3.17%, *Goniastrea* 1.90% and *Pavona* 1.77%. The stratum is flatter and more even — no genus holds even a fifth of the coral — which is exactly the signature of a low-cover, high-diversity, no-single-winner assemblage. D4 (n = 4) recovers to 37.00% coral against the lowest algal cover of the profile (32.25%), and *Porites* returns to the front at 9.05%, but the supporting cast is new: *Goniastrea* 4.25%, *Pachyseris* 3.30%, *Platygyra* 3.25% and *Favites* 2.75% — **all four of them near-absent or absent at D2** (*Pachyseris* records zero at both D2 and D3 and contributes its entire site-level 1.10% from this one stratum). Non-coral invertebrate cover climbs steadily with depth (21.8% → 25.6% → 25.25%) while bare substrate stays trivial (2.3% → 4.5% → 5.5%). The site-level figures of 8.04% *Porites* and 2.66% *Stylophora* therefore describe no single transect: they average a *Porites*-dominated upper stratum, a *Stylophora*-led low-cover middle, and a *Porites*-plus-faviid deep stratum.

The photographs corroborate exactly this: a continuously occupied, high-relief deep forereef with almost no open ground. P4220911–914 and P4220944 show the coral-abundant end of the site — near-continuous cover of massive and submassive heads, foliose and encrusting plates, small corymbose and digitate colonies, and broad sheets of pink-purple crustose coralline algae binding the dead framework between them, matching the 20.3% CCA figure. P4220916 shows the slope profile with several large tabular colonies and a stacked, fluted foliose colony consistent with *Pachyseris* or *Echinopora*. P4220922–923 are close-range frames of a large tan lobate-foliose colony and a field of cream knobby colonies. At the other end, **P4220961 and P4220962 show the low-cover face of the site — small scattered colonies separated by pale, sediment-veneered pavement and rubble, with visibly wider gaps than anywhere else in the set** — a plausible photographic counterpart to the 25.4% D3 stratum. Growth-form composition in the imagery tracks the transect data closely: massive, submassive, encrusting and foliose forms dominate, branching *Pocillopora*- and *Stylophora*-type colonies are frequent, and unambiguous large tabular *Acropora* appears in only a handful of frames — consistent with *Acropora* holding just 0.77% of the benthos here.

**Condition reads as good, and the site's published bleaching figure survives frame-level scrutiny only because it is very low to begin with.** `Mean(%bleaching)` is **2.98%, 81st of the 98 Chagos sites carrying a value — well into the low-severity end of the distribution** (the two CoralNet training runs agree here, 2.98 vs 2.976, so this is not one of the ten sites where the runs diverge). The 16-frame review found no colony-scale, unambiguous bleaching field anywhere on the site. Two frames warrant naming. **P4220923's two annotated bleached points sit on a genuinely pale cream colony of *Porites*-compatible form — this is the one defensible paling candidate in the entire set**, and even it cannot be separated from normal pale *Porites* pigmentation in a single strobe-lit still. **P4220915, by contrast, carries roughly half the site's entire bleaching signal and shows no bleached coral at all**: the frame is white carbonate sand in the foreground, a dense fusilier school, and blue-hazed backscatter — the bright sand and haze are the most plausible thing the annotator's bleached-coral class fired on. Two isolated small white branching colony tips are visible in P4220954, unresolvable as recent bleaching versus recent breakage. No disease lesions, tissue-loss margins, *Drupella* aggregations, parrotfish scarring or crown-of-thorns starfish (COTS) individuals or feeding scars were observed in any of the 16 frames — an explicit absence-of-evidence finding across a full-frame review.

Fauna is the standout of this dive. **A hawksbill turtle (*Eretmochelys imbricata*) is documented at close range across five frames** — P4220911–914 show it resting and then moving across the pavement, and P4220935 shows a turtle again mid-water beside a diver; three further frames (P4220951, P4220953, P4220954) capture a turtle silhouette in open blue water above the reef. The identification is hawksbill, not green, on the tapered beak-like rostrum and the strongly imbricate, serrated posterior marginal scutes visible in P4220911–913; whether one animal or two were present cannot be resolved. **P4220922 and P4220923 show a large pufferfish previously logged here as a moray — the correction stands: it is an *Arothron* pufferfish, the dark-phase white-spotted form consistent with guineafowl puffer *Arothron meleagris***, photographed resting directly on live coral. P4220944 includes a small angelfish with a yellow forebody and dark blue rear consistent with *Centropyge bicolor*. Dense fusilier schools cross P4220915, P4220916 and P4220922; blue chromis and small planktivores are scattered through P4220961–962. A green whip coral (*Cirrhipathes*/*Junceella*-type) appears in P4220922–923. Divers are visible in P4220935 and P4220952 — routine field documentation.

<b>Assemblage composition.</b> Formal transects give 33.1% coral against 38.5% algae across **31 nonzero genera (12th of 109 Chagos sites)**, with the algal third of the benthos more than half crustose coralline (20.3%) and non-coral invertebrates taking a further 24.2%. *Porites* (8.04%) and *Pocillopora* (5.23%) lead, but **the site's defining feature is how little *Acropora* it holds — 0.77% of benthos, 2.3% of coral** — and how completely the dominance structure reshuffles with depth: *Porites*-led at D2, *Stylophora*-led at a low-cover, maximally diverse D3, then *Porites* with *Goniastrea*, *Pachyseris* and *Platygyra* at D4. The photographs document the full range from near-continuous massive/foliose cover to sparse colonies on sediment-veneered pavement, a hawksbill turtle across five frames, a guineafowl-type *Arothron* puffer, and no evidence of bleaching, disease, predation or COTS across all 16 stills.

**Flags:** (1) **A formal transect at this site recorded ZERO coral.** `BIBL83_d3_T5`, a 500-point transect, returns `sum(coral)=0` against 47.8% coral+algae — the only null-coral transect among the 16 here, and it pulls the D3 stratum mean down by roughly 4 pp on its own. Worth confirming against the raw datasheet that this is a real observation and not a transcription or import failure. (2) **The 100-point and 500-point transects disagree systematically and by a lot.** The ten 100-point transects average 39.9% coral; the six 500-point transects average 20.6% — a 19.3 pp gap, and the pattern holds inside every stratum (D2: 39.8 vs 31.6; D3: 31.8 vs 12.7; D4: 56.5 vs 17.5). The site figure of 33.1% averages the two protocols together via the depth-stratum means. This is a methodological confound, not a depth effect, and it deserves a corpus-wide check. (3) **The `exposure-original` cell is blank** for this site; `exposure-revised` reads "exposed" and that is what is quoted above. The original class is not inferred. (4) **The multi-method figures rest on the same 16 frames reviewed here** (`n_photos=16` matches the folder exactly — no scored-subset mismatch at this site), but CoralNet points (24.8%) and Coralscapes masks (21.6%) sit 8.3–11.5 pp below the transect figure. Do not average them; the transect value is ground truth for cover. (5) **A single frame carries about half the site's bleaching signal and contains no bleached coral.** P4220915 is white sand plus blue haze; the bleached-class points there are best read as a substrate/backscatter false positive. Removing that frame would materially change the site bleaching percentage. (6) **P4220935 is a documented annotation-provenance artefact**: the frame is carpeted in live colonies yet scores 6.7% live coral and 60% "Unknown". Recomputing site cover with Unknown excluded lands at 32.5–33.1%, essentially on top of the transect value — evidence that the CV shortfall at this site is largely an Unknown-class problem, not a real cover disagreement. (7) **P4220923's two bleached points are the only defensible paling candidate**, on a pale cream *Porites*-compatible colony; it cannot be separated from normal pale pigmentation in a single still. Two small white branching tips in P4220954 are likewise unresolved (recent bleaching vs recent breakage). (8) **No COTS, disease or predation scarring was observed** across all 16 frames. (9) ***Pachyseris* is a single-stratum artefact of the site average**: it records zero at D2 and D3 and contributes its entire site-level 1.10% from D4, which rests on four transects. The same caution applies to *Goniastrea* (0.03% at D2, 4.25% at D4). (10) **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`). The site table's `Mean(%bleaching)` column is not a mean — it equals the 1st CoralNet training run — and any rank taken off it is approximate; it also carries at least one known bad value elsewhere in the archipelago (BITB58 listed at 22.22% against a corrected 0.77%). Never read this number as a diver's field estimate.

**Benthic data:** [Download BIBL83 benthic survey data (CSV)](../benthic_data/BIBL83_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).