### BITB58 — Middle Brother
**CRD page:** [Three Brothers/Central Chagos Coral Surveys BITB50-BITB58 (Mar 2015)](https://coralreefdiagnostics.com/three-brotherscentral-chagos-20150325-to-0328)

**28 Mar 2015**
**GPS:** 6.14790°S, 71.52411°E
**Setting:** leeward channel reef, submerged
**Coral cover:** 67.0% (3 formal transects, D2–D4; **algal cover 29.3%**) · **published image bleaching:** 0.8%
**Photo review:** all 28 frames examined
**Representative photos:** P3280503.jpg, P3280518.jpg, P3280538.jpg, P3280540.jpg, P3280541.jpg, P3280542.jpg, P3280545.jpg, P3280539.jpg

BITB58 is the fourth-highest-cover site in the 109-site Chagos set at 67.0% transect coral cover, and one of the most compositionally lopsided. The formal assemblage is carried almost entirely by two genera: *Heliopora* at 31.33% and *Acropora* at 27.67%, together 88% of all coral recorded. *Goniastrea* (3.00%), *Pocillopora* (2.00%) and *Porites* (2.00%) follow, with *Pavona* and *Favia* below 1%. Only seven coral codes are non-zero — an unusually short list for a site of this cover. Crustose coralline algae hold 21.0% of the benthos, turf 5.33%, bare substrate 4.00%, macroalgae 2.00% and erect algae 1.00%.

**The depth structure is the finding here, and it is extreme.** The three transects are not variations on a theme; they are three different reefs. D2 carries only 28% coral, spread thinly across *Porites* (6%), *Acropora* (3%) and *Heliopora* (1%). D3 jumps to **93% coral — and every coral point on that transect is *Heliopora***. D4 holds **80% coral, every point *Acropora***. Two adjacent depth bands, each effectively a monoculture, of different genera. **This is genuine zonation, not a scoring artefact:** the surveyor recalls the dive as blue coral as far as the eye could see (A. Mayfield, 2026-08-25). The site mean of 31.33% *Heliopora* should still be read as the average of three very different transects rather than a uniform site value.

The reef photographs as a broad, low-relief channel bottom rather than a wall or steep slope, and its dominant architecture is **tabular**. P3280540 is filled by a single enormous table *Acropora*, metres across, its olive surface set off by pale growing margins and a brilliant orange-red encrusting sponge at its rim; P3280542 pulls back to show a continuous field of such tables — olive and yellow-green, tiered and overlapping, running to the limit of visibility with white sand only at the margins. P3280518 and P3280515 show the branching zone: multiple large staghorn thickets standing on clean white sand among massive *Porites*-type domes, reticulate faviids and foliose plates, with clouds of *Chromis* overhead. Together these are the photographic match to the 80%-cover *Acropora* transect.

The site is not uniformly dense. P3280545 looks down on broad white sand flats carrying only isolated coral patches, and P3280517 and P3280519 show pavement where massive *Porites* domes, reticulate faviids and pedestalled plates sit widely spaced on open sand — the ground that produces the 28%-cover D2 transect. Coral cover at BITB58 is therefore a mosaic of near-continuous tabular fields and near-bare sand, not a uniform 67% carpet.

The channel carries substantial fish biomass, and this is the site's most consistent visual feature: eight of the 28 frames are dominated by a very large yellowback fusilier aggregation. P3280502 and P3280503 show it extending to the limit of visibility; P3280499, P3280500, P3280501, P3280504, P3280506 and P3280507 catch it working over the pavement, with survey gear — a transect tape, a slate — visible at the frame edges. Other fauna is sparse but notable: a large golden gorgonian sea fan and a barred grouper in P3280541, a blue *Linckia* sea star on sand in P3280506, black finger sponges, a vivid orange-red encrusting sponge on the rim of the great table in P3280540, and dense *Chromis* clouds over the thickets throughout. P3280522 catches a surveyor working a slate above a large cream-tan foliose colony growing as a ruffled vase.

Closer frames show more genus-level variety than the seven-code transect list implies. P3280520 holds branching *Acropora*-type thickets alongside large ruffled foliose plates in cream-yellow and one deep maroon, massive meandroid brain-coral heads, smooth *Porites*-type domes and encrusting colonies. P3280524 contains a solitary free-living fungiid mushroom coral, radiating septa clearly visible. P3280516 is a macro frame: a small brown *Acropora* whose branch tips are bright white — these are growing tips, not bleaching — wedged between dense brown-green foliose algae and reticulate faviids. P3280537 is pure substrate: turf, dark red encrusting algae, green *Halimeda*-type segments, small ascidians, pink CCA and a single pale coral recruit — the matrix behind the 21% CCA and 29.3% algal figures. **P3280538 holds the best candidate for the site's dominant taxon**: a robust olive-brown colony with thick, blunt, finger-like branches and a smooth surface, growing beside a delicate fine-branched coral and beneath a large meandroid brain coral. That blunt digitate form is characteristic of *Heliopora*, and it is the only colony in the set whose growth form matches the 93%-cover D3 transect.

Condition is good. The published image-based bleaching figure is **0.8%**, among the lowest in the archipelago (rank 102 of 109). The single clear bleaching subject is P3280539: a corymbose *Acropora* with its outer branches stark white and the colony centre still fully pigmented brown — a partial, colony-scale event on an otherwise healthy reef, and exactly the kind of isolated case a sub-1% site figure describes. Two frames record partial mortality of older colonies rather than bleaching: P3280521 shows patchy dark red algal overgrowth across parts of the branching thickets with pale, algae-bound branches, and P3280523 holds a large standing table whose whole upper surface is dead and algae-filmed, bare skeleton showing through, immediately beside healthy branching thickets. Both read as older loss on large colonies, not an active event. No COTS, no disease fronts and no corallivore scars were seen in the frames reviewed.

<b>Assemblage composition.</b> Formal surveys give 67.0% coral and 29.3% algae, the coral split 88% between *Heliopora* and *Acropora*. Photographs document more genus-level variety than that — foliose plates, meandroid faviids and *Porites* domes are all conspicuous — which is the expected direction of disagreement when three transects sample a heterogeneous channel bottom.

**Flags:** (1) **D3 and D4 are single-genus transects** (93% all *Heliopora*, 80% all *Acropora*). Treat the 31.33% site mean for *Heliopora* with caution; it rests on one transect. (2) **The HELI code means *Heliopora*, not *Heliofungia*.** The LOF codebook maps HELI to *Heliofungia*, but the field data are *Heliopora* (A. Mayfield, 2026-08-25) — a solitary mushroom coral cannot reach 93% cover, and HEFU is used only 3 times across all 3,067 GRE transects. (3) **No large *Heliopora* stand was captured photographically.** Across all 28 frames the only strong candidate is the blunt digitate colony in P3280538; the imagery is dominated instead by tabular and branching *Acropora*. The 93%-cover D3 transect rests on the survey data and the surveyor's recollection of blue coral as far as the eye could see — the photographer evidently did not shoot that zone, which is a real limitation of this photo set rather than a contradiction of the survey. (3b) P3280524 contains a clear free-living fungiid mushroom coral, yet **no `FUNG` code appears anywhere in this site's transect data** — a reminder that seven non-zero codes under-describe the assemblage. (4) **P4280889 is dated 28 April**, one month after this dive, and shows a staghorn thicket with tabular plates behind; it is almost certainly a Peros Banhos frame misfiled into this folder. (5) **P4280889 was removed from this folder** (2026-08-25) — dated 28 April, a month after this dive; moved to `_misfiled_28Apr_needs_site/`. It could not be filed to its true site because BiSa102-finished already holds a *different* image under that filename, and frame numbers 890-895 appear in both BiSa102 and BiSp101 — two cameras with colliding sequences on 28 April. (6) Of the 28 remaining frames, three carry no benthos (P3280546 diver-in-bluewater; P3280544 hazy diver ascent) and eight are fusilier-dominated water-column shots, so the effective benthic sample is roughly 17 frames.

**Benthic data:** [Download BITB58 benthic survey data (CSV)](../benthic_data/BITB58_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).