Part of the regional summaries: Maldives - Central (Male Atolls / Ari / Vaavu) and Maldives

A dawn dive over table corals left gray and algae-overgrown by the 2016 bleaching and finished off in places by crown-of-thorns starfish, though young table corals rebuilding on the ruins and surviving cauliflower and boulder corals show early recovery. Computer vision cover estimate is 14-16%, averaging 15.2%.
Fifteen independent survey stations shown with the program’s logo — click one to see how far it is from this site.
Three other CRD dive sites within 5 km shown in orange — click a marker to open its page.
17 of these 25 photographs are shown beside the text that discusses them under "Site description" below.
Look closely at the survivors: orange cauliflower corals with a tiny hawkfish tucked in the branches, massive boulder corals still holding ground, and here and there a young table coral rebuilding on the ruins — early signs of a reef trying to restart itself (e.g., Image 3, Image 9, Image 10, Image 20, Image 21).
Coral cover: 15% — photo-derived, and the two methods broadly agree (14% vs 16%). This site's figure is the mean of two independent computer vision passes over Anderson's own frames: a DINOv3 calibrated point classifier reads 14.1% and a CoralSCOP computer vision reads 16.3%, a spread of 2.2 percentage points. The map publishes the mean so sites can be sorted against one another, but the honest statement of what is known here is the 14–16% range: the two products have different denominators and should not be treated as one measurement. He is actively comparing CoralSCOP and DINOv3 against newer CV models and will swap in an updated figure here once a more reliable model is found.
A dawn dive over a reef that tells the story of 2016 in one sweep: whole terraces of table corals stand intact but gray, overgrown with algae a year after the great El Niño bleaching — and the photographer's own caption blames crown-of-thorns seastars for finishing parts of the job (e.g., Image 2, Image 11, Image 12, Image 15, Image 17, Image 25).
Rivers of yellow-striped snapper pour over the old coral tables, a powder-blue surgeonfish patrols the rubble, and a honeycomb moray gapes spectacularly from its crevice (e.g., Image 1, Image 5, Image 6, Image 16, Image 18, Image 19).
No independent survey was made at this exact site; the nearest ones are listed in the table below.
Yet the corner is far from lifeless.
| Method | Hard-coral cover (%) |
|---|---|
| In-house DINOv3 | 14.1 |
| CoralSCOP | 16.3 |
| Mean (± SD) | 15.2 ± 1.6 |
| Range | 14.1–16.3 |
Mean ± SD and range are across the computer-vision methods shown.
| Agency / program | Year(s) | Stations / distance | Link |
|---|---|---|---|
| Montefalcone et al. (2020) | 2015–2019 | 4 stations, 0.60–3.59 km | Records ↓ |
| Re E.* | 2026 | 1 station, 1.72 km | Records ↓ |
| Reef Check | 2012–2025 | 10 stations, 0.73–3.88 km | Records ↓ |
Within 5 km, 15 independent survey stations recorded hard coral cover in 46 records between 2012 and 2026. Of the 9 stations surveyed in more than one year, 2 declined, 3 increased, and 4 changed by less than 5 percentage points between their first and latest surveys. The longest record, Guraidhoo backreef (Reef Check, 1.89 km away), went from 28% in 2019 to 25% in 2025. Surveys since 2021 average 25% (range 16–31%, n=10). This site's own estimate (15%) is indistinguishable from the recent range (16–31%): the 0.4-point gap is inside the ±7 percentage-point noise between methods; 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. *Conducted by citizen scientists), so these are context for the figures above, not a like-for-like comparison. Where a station was surveyed more than once the range is shown. Station names are not dive-site names. Values marked best patch exceed 70%: surveyors usually chose the richest area, so they are not site means.
| Agency | Site / location | Date | Hard coral cover (%) | Distance | Link |
|---|---|---|---|---|---|
| Reef Check | Biyadhoo House Reef | 2012 | 18% (3 surveys) | 3.88 km | — |
| Montefalcone et al. (2020) | Villivaru Kuda Giri | 2015–2019 | 8.0–51% (4 surveys) | 3.59 km | HeatCRD database |
| Montefalcone et al. (2020) | Guraidhoo Kuda Giri | 2015–2016 | 12–83% (2 surveys) best patch | 3.59 km | HeatCRD database |
| Montefalcone et al. (2020) | Maadhu Faru Beyru | 2016–2017 | 18–45% (2 surveys) | 0.60 km | HeatCRD database |
| Montefalcone et al. (2020) | Kandooma Beyru | 2016–2019 | 7.7–34% (3 surveys) | 1.05 km | HeatCRD database |
| Reef Check | Guraidhoo Corner | 2017 | 21% (2 surveys) | 0.78 km | — |
| Reef Check | Kandooma Beyru | 2018–2019 | 18–30% (4 surveys) | 0.73 km | — |
| Reef Check | Villivaru Giri | 2018 | 8.8% (2 surveys) | 3.18 km | — |
| Reef Check | Guridhoo outer | 2019–2022 | 16–20% (3 surveys) | 1.11 km | — |
| Reef Check | Guraidhoo backreef | 2019–2025 | 20–28% (6 surveys) | 1.89 km | — |
| Reef Check | Villivaru Kuda Giri | 2019 | 24% (2 surveys) | 3.45 km | — |
| Reef Check | Guraidhoo Forereef | 2024–2025 | 25–26% (4 surveys) | 1.22 km | — |
| Reef Check | Guraidhoo Banana | 2024–2025 | 25–27% (4 surveys) | 1.39 km | — |
| Reef Check | Guraidhoo Backreef | 2024 | 31% (2 surveys) | 1.97 km | — |
| Re E.* | Banana_Reef | 2026 | 29% (3 images) | 1.72 km | Re E. 2026. AquaReef Project - Maldives. MERMAID |
All images from Coral Reef Diagnostics. Data are from Coral Reef Diagnostics and the organizations named in the data-sources table above.
© Coral Reef Diagnostics.