Indonesia - Halmahera: Pretty Shallow
Part of the regional summaries: Indonesia - Halmahera and Indonesia - Maluku
More info
North Maluku, Indonesia (NOAA CRW virtual station)Pretty Shallow is built mainly on massive and foliose coral — large massive coral domes and dense foliose plate clusters alongside staghorn coral stands — and Anderson logged a huge, pregnant turtle on the dive. No bleaching, disease, or crown-of-thorns starfish (COTS) was seen in the photo review, though a computer-vision point count scored 0.2% of coral as bleached. A blind Claude computer-vision point count puts coral cover around 32%.
Pretty Shallow terutama terbangun di atas karang masif dan foliaseus — kubah karang masif besar dan gugusan karang lembaran yang rapat di samping rumpun karang tanduk rusa — dan Anderson mencatat seekor penyu besar yang sedang bunting dalam penyelaman ini. Tidak teramati pemutihan, penyakit, maupun COTS dalam tinjauan foto, meskipun penghitungan titik visi komputer menilai 0.2% karang memutih. Penghitungan titik buta visi komputer Claude memperkirakan tutupan karang sekitar 32%.
Terjemahan ringkasan ini oleh AI; versi bahasa Inggris yang berlaku.
These are unedited originals; edited versions will replace them.
This page shows 50 of the 109 photographs from this site, including every one the text cites.
30 of the 50 photographs on this page are shown beside the text that discusses them under "Site description" below.
Site description
Setting
Several frames in the second half show mild haze/backscatter (e.g., Image 29, Image 30, Image 51).
General benthic features
Pretty Shallow (Halmahera dive 8) lives up to its name in a shallow, well-lit reef built mainly on massive and foliose growth forms, with a wide cover spread (Claude: 32% ± 23, range=0.0–82%) driven by the contrast between large bare-sand-fringed massives and dense coral clusters (e.g., Image 12, Image 20, Image 21, Image 27).
Massive and plate corals dominate the frame count: large smooth brown, olive, and maroon domes and boulders recur throughout, and extensive ruffled brown-tan foliose plates form dense overlapping clusters at several points (e.g., Image 5, Image 13, Image 33, Image 34, Image 37, Image 38, Image 40, Image 43, Image 47, Image 48, Image 53, Image 54).
Dense staghorn/finger coral thickets in mauve, salmon-pink, and cream fill much of the mid-dive stretch, several frames hosting swarms of small red-orange anthias or damsels hovering directly over the branches (e.g., Image 4, Image 9, Image 14, Image 15, Image 25, Image 28, Image 31, Image 42, Image 45, Image 55).
A striking magenta/dark-red carpet or barrel sponge and a large pale anemone with cream-tipped tentacles both appear in close-up (e.g., Image 19).
Coral cover
- Reference: Claude CV point count (this site's photos)
- Coral cover: 32% — a blind Claude computer-vision point count of 105 of Anderson's unedited photographs (20 points per photo; water, fish, and obstructions excluded). Other computer-vision methods are listed in the method table.
Stressors
The photo review of 105 frames saw no bleaching, disease, crown-of-thorns, or predation, and pigmentation (browns, olives, maroons, salmon, mauve) is normal for the growth forms present; however, a computer-vision point count called 0.2% of coral points bleached (1 of 105 photos).
Fish and other fauna
Anderson logged a huge, pregnant turtle on this dive; it does not appear in the published photographs.
Comparisons with other data
We found no other surveys at this site as of 2026. If you would like to share yours, please email us.
Other notes
Divers appear occasionally for scale.
CV model comparison
More info
A few low-cover photos (sand, rubble, or open bottom) pull the mean down; 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 same photos 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 | 32.0 | n = 105 · SE 1.4 |
| CoralSCOP unedited | 30.2 | n = 105 · SE 1.8 |
| Claude | 32.3 | n = 105 · SE 2.3 |
The two computer-vision methods agree within 1.8 pp (30.2–32.0%) and read within 2.1 pp of Claude's blind point count (32.3%).
See a more detailed, stand-alone version of this table.
Data & links
All images from Coral Reef Diagnostics. Data are from Coral Reef Diagnostics.
© Coral Reef Diagnostics.