Philippines - Anilao: Monte Carlo Hotel

General benthic features

Anderson's own filenames set the depth structure - "Monte Carlo shallows", "view at 30 m", "30-40 m" - and the site is genuinely zoned, with a field of overlapping pale plating and foliose corals (Turbinaria, Echinopora lamellosa, and the Montipora group) forming near-continuous cover at roughly 30-40 m, a mid-slope of sunken objects, and shallows dominated by rubble and turf with only sparse recruits (e.g., Image 1, Image 2, Image 5, Image 6, Image 7, Image 25).

1256725

The artificial reef is the draw: an upright Volkswagen Beetle on sand, a mermaid statue, a large standing cross on a masonry base, and a deliberate tableau of household fixtures - bathtubs and sinks, a strongbox, a desk with a telephone, and toilets, one with a massive coral colony growing in the bowl (e.g., Image 13, Image 20, Image 21, Image 22, Image 23).

1320212223

Deployed coral settlement tiles are also present, one carrying crustose coralline algae and small recruits, with Anderson pointing out a marked plate - a recruitment installation, not incidental debris (e.g., Image 8).

8

Large Euphyllia hammer-coral stands in cream and turquoise are the finest stony corals here (e.g., Image 18).

18

Coral cover

Cover comes from two computer vision methods, neither preferred. DINOv3 point classification gives 22.3% hard coral (95% CI 19.2-25.4) from all 82 frames and 1,640 points. CoralSCOP computer vision, its full run now finished, gives 22.0% across all 82 frames (median=18.3%, bootstrap 95% CI 18.2-26.0), against the 21.8% from the 20-frame subset the map's 22.1% was built on. The published spread is 0.6 points and the completed run closes it to 0.3 - the closest agreement of the seven sites, and the most misleading. The two models correlate at only r = 0.23 across these frames, the weakest of the set, and they reverse on exactly the frames that matter: on the two deep plating-field images DINOv3 reads 44.4% and 43.6% hard coral while CoralSCOP reads 8.2% and 13.9%, because instance segmentation cannot separate overlapping foliose plates and returned only seven masks for the 30-40 m frame. A single site figure here averages two very different habitats and hides that reversal. The two models fail differently. CoralSCOP computer vision reads high on low-cover reefs - a median=1.68x a co-located Reef Check diver survey, 4.6x at worst - yet under-reads dense plating fields. DINOv3 is strong in-domain (6-class macro-F1 0.62, hard coral recall 0.82) but collapses off-domain: cross-domain point accuracy 0.17-0.25, ceiling near 35-50%. Coral cover: 22% — photo-derived, and the two methods broadly agree (22% vs 22%). Two independent computer vision passes over all 82 of Anderson's frames from this dive give a DINOv3 calibrated point classifier reading of 22.3% (95% CI 19–26%, block bootstrap) and a CoralSCOP computer vision of 22.0%, a spread of 0.3 percentage points. The map publishes their mean so sites can be ranked against one another, but the 22–22% range is the honest statement of what is known here: the two products have different denominators and should not be collapsed into one measurement. He continues to test alternative CV models for this task and will rerun the CoralSCOP and DINOv3 number here if a more reliable option turns up.

Stressors

No crown-of-thorns starfish (COTS) or other corallivores evident.

Fish and other fauna

Damselfish and anemonefish have colonized all of it. The macrofauna is the most abundant and the most genus-rich of the Philippine archive: Phyllidia and Phyllidiella nudibranchs, several blue Chromodoris including an apparent mating pair, a tiger cowrie, cleaner shrimp in the crevices, and a granulated sea star (Choriaster) (e.g., Image 9, Image 10, Image 11, Image 12, Image 16, Image 17).

91011121617

A run of five consecutive frames records green turtles foraging over the rubble; the sequence cannot establish how many individuals it shows (e.g., Image 14, Image 15).

1415

Comparisons with other data

No independent survey was made at this exact site; the nearest ones are listed in the table below.

Dive record

The third dive of 30 September 2016, and the most unusual of the trip: a house reef that is substantially man-made. All 82 available frames were reviewed.