St. Maarten - Mike's Maze
More about this place: Sint Maarten on Wikipedia
More info
Leeward Caribbean Islands (NOAA CRW virtual station)Mike's Maze is a playful network of low coral ridges and sandy swim-through corridors, notable for a genuinely thriving stand of elkhorn coral — now scarce across much of the Caribbean — alongside brain and boulder star corals. A blind Claude computer-vision point count puts coral cover around 14%. Reef Check's Mike's Maze 1 station, where the map pin sits, recorded 10% hard coral in 2006.
Mike's Maze is een speels netwerk van lage koraalruggen en zandige doorzwemgangen, opvallend door een werkelijk bloeiende groep elandsgeweikoraal, dat nu in een groot deel van het Caribisch gebied schaars is, naast hersenkoraal en sterkoraal. Een blinde puntentelling met Claude-computervisie schat de koraalbedekking op ongeveer 14%. Het Reef Check-station Mike's Maze 1, waar de kaartpin staat, registreerde in 2006 10% hard koraal.
AI translation of this summary; the English is authoritative.
Five independent survey stations shown with the program’s logo — click one to see how far it is from this site.
One other CRD dive site with its own page within 5 km shown in orange — click a marker to open its page.
These are unedited originals; edited versions will replace them.
20 of the 49 photographs on this page are shown beside the text that discusses them under "Site description" below.
Site description
Setting
Mike's Maze earns its name from a network of low coral ridges laced with sandy swim-through corridors, making for a playful, twisting dive rather than a single open reef face (e.g., Image 16, Image 54).
The maze-like topography and a healthy elkhorn population make this dive equal parts terrain and coral to explore.
General benthic features
The highlight here is a thriving stand of elkhorn coral, its broad antler-shaped branches spreading across several ridge tops — a heartening sight, since elkhorn has grown scarce across much of the Caribbean in recent decades (e.g., Image 7, Image 15, Image 19, Image 21, Image 30, Image 31, Image 36, Image 47).
Around it, brain corals (some strikingly yellow-green) and sturdy boulder star corals anchor the reef, while a soft-coral understory of sea plumes and sea rods fills the spaces between (e.g., Image 3, Image 6, Image 17, Image 26, Image 43, Image 49).
Encrusting sponges in orange, red, and the occasional vivid yellow add color to the rock (e.g., Image 8), and an octopus was photographed at rest beside an orange sponge (e.g., Image 4).
Coral cover
- Reference: Claude CV point count (this site's photos)
- Coral cover: 14% — a blind Claude computer-vision point count of 78 of Anderson's unedited photographs (20 points per photo; water, fish, and obstructions excluded). Other computer-vision methods are listed in the method table.
- Anderson's own dive log for this dive records 4-5 sharks, mostly soft corals across the site, and stony coral cover of 3.0-5.0%; the photo count above reads higher than the in-water estimate.
Stressors
One small knobby massive colony carries about 15 sharply bordered white patches on and between its knobs, with corallite texture still visible on the white areas (e.g., Image 25, Image 53); from photographs, paling cannot be separated from very recent tissue loss.
Comparisons with other data
The pin sits on Reef Check's "Mike's Maze 1" station, whose only survey, on 9 April 2006, recorded 10% hard coral; an XL Catlin Seaview Survey transect 0.06 km away recorded 1.9% in 2013. Both are other organizations' surveys, not measurements of this dive.
CV model comparison
More info
A few colony-focused photos raise the mean; the median is the typical photo. No computer-vision method yet estimates coral cover accurately on new images, so treat this as indicative rather than a measurement; it is revised as the methods improve.More info
DINOv3 and CoralSCOP were run on the site's photographs (see each row's n) and are listed for comparison. No diver transect exists for this trip, so agreement between the rows is not accuracy, and no computer-vision method yet estimates coral cover accurately on new images; treat every row as indicative.| Method | Hard coral cover (%) | Photos (n) ± SE |
|---|---|---|
| In-house DINOv3 | 19.4 | n = 78 · SE 0.9 |
| CoralSCOP unedited | 25.2 | n = 75 · SE 1.6 |
| Claude | 13.5 | n = 78 · SE 1.8 |
The two computer-vision methods agree within 5.8 pp (19.4–25.2%) and read 5.9–11.7 pp above Claude's blind point count (13.5%); the headline is Claude's blind point count because no transect or manual count exists here.
See a more detailed, stand-alone version of this table.
Independent surveys nearby
| Agency / program | Year(s) | Stations / distance | Link |
|---|---|---|---|
| Reef Check | 2005–2007 | 4 stations, 0.00–4.94 km | Records ↓ |
| XL Catlin Seaview Survey | 2013 | 1 station, 0.06 km | Records ↓ |
Coral Reef DiagnosticsYear-average (illustrative only, programs differ)Within 5 km, 5 independent survey stations recorded hard coral cover in 5 surveys between 2005 and 2013. The most recent nearby survey is from 2013, so these values predate recent heat stress. This site's own estimate (14%) sits 1 point above the range of these station-year means up to 2 km away (1.9–13%), inside the 7-point method-noise floor (inconclusive); the methods differ, so the comparison is context, not a like-for-like test.
Collected by other programs using different methods (Reef Check uses point-intercept transects; the XL Catlin Seaview Survey uses photo-transects with automated image annotation), so these are context for the figures above, not a like-for-like comparison. Where a station was surveyed more than once, the cell shows either the range of its values or, as a single value, the mean of surveys made in one year. Station names are not dive-site names.
| Agency | Site / location | Date | Hard coral cover (%) | Distance | Link |
|---|---|---|---|---|---|
| Reef Check | Proselyte Reef | 2005 | 6.3% | 1.78 km | — |
| Reef Check | Mike's Maze 2 | 2005 | 13% | 1.83 km | — |
| Reef Check | Mike's Maze 1 | 2006 | 10% | 0.00 km | — |
| Reef Check | Frenchman Reef | 2007 | 3.8% | 4.94 km | — |
| XL Catlin Seaview Survey | transect 25035 | 2013 | 1.9% | 0.06 km | doi:10.14264/uql.2019.930 |
Data & links
All images from Coral Reef Diagnostics. Data are from Coral Reef Diagnostics and the organizations named in the data sources above.
© Coral Reef Diagnostics.