Mauritius - Reef Garden
Part of the regional summary: Mauritius
More info
Southern Mauritius (NOAA CRW virtual station)A low-relief Mauritius patch reef of coral bommies and rubble mounds rising from white sand. Massive corals, brain corals, and staghorn corals anchor the bommies, joined by garden eels, two lionfish, and dense snapper and anthias schools, with no widespread bleaching; the site's exact position is approximate.
Un récif de patch mauricien à faible relief formé de patates coralliennes et de monticules de débris s'élevant d'un sable blanc. Des coraux massifs, des coraux-cerveaux et des coraux en corne de cerf structurent les patates, accompagnés d'anguilles-jardinières, de deux rascasses volantes et de bancs denses de lutjans et d'anthias, sans blanchissement généralisé ; la position exacte du site est approximative.
Traduction par IA de ce résumé ; la version anglaise fait foi.
Three 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.
22 of the 23 photographs on this page are shown beside the text that discusses them under "Site description" below.
Site description
Setting
Reef Garden is a low-relief patch-reef system: coral bommies and consolidated rubble mounds rising from a white sand flat, with manually scored point-count cover of 26% hard coral (e.g., Image 10, Image 11, Image 13, Image 16).
General benthic features
Massive dome-shaped corals, brain corals, and branching and staghorn corals anchor the bommies, joined by soft-coral clusters, garden eels, two lionfish, and dense snapper- and anthias-type schools; no widespread bleaching was evident (e.g., Image 1, Image 3, Image 4, Image 5, Image 6, Image 7, Image 9, Image 12, Image 14, Image 15, Image 17, Image 18, Image 19, Image 20).
Coral cover
- Reference: Mayfield et al. (2026, Platax), Table 3 (manuscript in preparation)
- CoralNet submission package
- Coral cover: 26% — Anderson Mayfield's manual point count of the photographs (Mayfield et al. 2026 Platax Table 3).
Stressors
A few colonies were bleached or pale: one colony, photographed six times close up, has stark-white upper lobes with polyps visible over a tan lower part (e.g., Image 2), and Anderson's own point annotations label it bleached hard coral.
The annotations also mark paling on one or two large-polyp colonies, labeling their pale tan tissue bleached while the green oral discs stayed healthy. No crown-of-thorns starfish (COTS) or other corallivores were evident.
Fish and other fauna
An Eschmeyer's scorpionfish (Rhinopias eschmeyeri) was photographed on this dive (e.g., Image 21, Image 22, Image 23).
Comparisons with other data
Reef Check surveyed a nearby station (Island Reef), 0.30 km from the pin; we could not confirm that it is the same reef. Within 5 km, three independent stations (Reef Check and WCS via MERMAID) recorded hard coral cover between 1999 and 2016: two by benthic survey, station-year means of 30–61%, and one by MERMAID bleaching-protocol quadrats, 71%; the nearest, Reef Check at Island Reef, 0.30 km away, read 30–61% between 1999 and 2000.
Other notes
Location approximate: the dive center's site board (Anderson's photograph) names Reef Garden but gives no coordinates, so this site is placed near Ti Korail, the same morning's second dive — treat the position as indicative only.
CV model comparison
More info
That convention would raise every Mauritius site by 0–3.7 pp. The Claude row is the manuscript's Claude (with image context) value, scored on the same frames as the manual count; one of its five sets of context images had been seen before, so it is not fully blind. Anemone Garden is anemone habitat rather than coral reef; its manual value stands but carries no error term. CoralSCOP (SAM-based mask-sum), edited photos: n/a — no edited copies exist; all 347 Mauritius photographs are unedited.| Method | Hard coral cover (%) | Photos (n) ± SE |
|---|---|---|
| Manual (reference) | 26.0 | |
| CoralNet | 27.8 | n = 105 |
| Coralscapes v1 | 38.4 | |
| CoralSCOP | 22.2 | n = 105 |
| DINOv3 | 44.6 | n = 105 |
| Claude (context) | 26.7 | n = 105 |
The four computer-vision methods disagree by 22.4 pp (22.2–44.6%) and read on both sides of the manual point count (26.0%); the headline is the manual point count because a person's point count of the site's photographs is the reference the models are measured against.
The Coralscapes v1 row is shown for reference only, outside this comparison.
See a more detailed, stand-alone version of this table.
Independent surveys nearby
| Agency / program | Year(s) | Stations / distance | Link |
|---|---|---|---|
| Reef Check | 1999–2000 | 2 stations, 0.30–3.88 km | Records ↓ |
| WCS via MERMAID (MERMAID bleaching-protocol quadrats) | 2016 | 1 station, 1.59 km | Records ↓ |
Coral Reef DiagnosticsYear-average (illustrative only, programs differ; quadrats left out)Within 5 km, 3 independent survey stations recorded hard coral cover in 4 surveys between 1999 and 2016. The one station surveyed in more than one year declined by 31 points between its first and latest surveys. The longest record, Island Reef (Reef Check, 0.30 km away), went from 61% in 1999 to 30% in 2000. The most recent nearby survey is from 2016, so these values predate recent heat stress. This site's own estimate (26%) sits 4 points below the range of these station-year means up to 2 km away (30–61%, excluding 1 best-patch value above 70% that is not a site mean); 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; MERMAID is a data platform: each MERMAID row names the organization that did the survey, and its method is in the row's hover text), 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. Values above 70% are marked “best patch”: surveys often target the densest coral on a reef, so these may overstate the site as a whole.
| Agency | Site / location | Date | Hard coral cover (%) | Distance | Link |
|---|---|---|---|---|---|
| Reef Check | Island Reef | 1999–2000 | 30–61% (2 surveys) | 0.30 km | — |
| Reef Check | Island Reef, Anse La Raie | 1999 | 46% | 3.88 km | — |
| WCS via MERMAID (MERMAID bleaching-protocol quadrats) | Bain Boeuf | 2016 | 71% best patch | 1.59 km | Briand J, Darling E. 2016. 2016 bleaching surveys. MERMAID |
Data & links
- CoralNet image repository for this region ↗
- Photo portfolio ↗
- Coral Reef Diagnostics page for this area ↗
All images from Coral Reef Diagnostics. Data are from Coral Reef Diagnostics and the organizations named in the data sources above.
© Coral Reef Diagnostics.