# Kelp Forests of the World — Methodology and Sources

**Publication:** *The Forgotten Forests of the Sea* (Sunstone Institute / Lou Luddington)
**Date:** June 2026

---

## Overview

This work documents the global distribution, health status, and documented loss of kelp forest ecosystems. It is designed to accompany editorial and long-form journalism and is not a primary research tool. All data is sourced from published scientific literature, government monitoring programmes, or established global biodiversity databases.

The material presents six categories of information, each from a distinct data source:

1. Modelled kelp habitat range
2. Regional health snapshots with canopy monitoring data
3. Documented kelp replacement sites
4. Species occurrence records (OBIS and GBIF)
5. Tasmanian canopy time-series (IMAS Landsat survey)
6. Population proximity estimates

---

## 1. Kelp Habitat (Modelled Range)

### Source
**UNEP-WCMC (United Nations Environment Programme World Conservation Monitoring Centre).** Global distribution of kelp forests. Cambridge, UK: UNEP-WCMC. https://data.unep-wcmc.org/datasets/7

### Description
The kelp biome layer shows the modelled *potential* extent of kelp habitat globally — coastal ocean areas with conditions (water temperature, light, substrate type) considered suitable for kelp forest formation. It does not represent observed kelp canopy or current canopy extent. Some areas within the biome boundary currently lack kelp due to local stressors, overgrazing, or historical removal.

### Assumptions and caveats
The boundary represents suitability, not presence. Large portions of the biome have experienced significant kelp loss that is not reflected in the boundary polygon.

---

## 2. Regional Snapshot Markers

The 22 snapshot regions draw from three primary data sources depending on bioregion.

### A. Kelp Watch — NE Pacific, South America, South Africa, Namibia (18 regions)
**Bell, T., et al. / The Nature Conservancy & University of California Santa Barbara.** Kelp Watch: satellite canopy monitoring programme. kelpwatch.org.

Kelp Watch uses Landsat 5, 7, 8, and 9 satellite imagery (30 m resolution) to map floating kelp canopy surface area annually from 1984 to 2025. The programme covers the full NE Pacific coast (Alaska to Baja California), the Peruvian coast, Patagonian Argentina, and the South African and Namibian west coasts.

Per-region values used:
- **Coverage (hectares)** — modelled canopy area in the most recent survey year
- **Change (percent)** — change from the first survey year to the most recent year
- **Survey period** — first and latest survey years (1984–2025 for most regions)
- **Status** — editorial classification (healthy / stressed / critical)

### B. Kvile et al. 2022 — Norway, Iceland, Greenland, Svalbard (4 regions)
**Kvile, K.Ø., Norderhaug, K.M., Andersen, G.S., Bekkby, T., Gundersen, H., Hancke, K., Hjermann, D.Ø., Moy, F.E., Rinde, E. & Christie, H. (2022).** Scaling up: Kelp forest extent in Norway and surrounding waters. *Frontiers in Marine Science*, 9, 850359. https://doi.org/10.3389/fmars.2022.850359

Provides modelled kelp extent estimates for the Nordic/Arctic region based on habitat suitability modelling combined with observational data. Extent values for Norway (~8,100 km²), Iceland (~1,703 km²), Greenland (~876 km²), and Svalbard (~172 km²) are drawn directly from this study. No annual canopy time-series is available for these regions; change is not reported, and coverage figures represent the modelled extent at the time of publication (2022).

### C. IMAS & EPBC Act — Tasmania (1 region)
**IMAS (Institute for Marine and Antarctic Studies), University of Tasmania.** Tasmanian Giant Kelp Landsat Survey. www.imas.utas.edu.au

**Australian Department of Climate Change, Energy, the Environment and Water (DCCEEW).** Giant Kelp Marine Forests of South East Australia — listed as an Endangered Ecological Community under the *Environment Protection and Biodiversity Conservation Act 1999* (EPBC Act). https://www.dcceew.gov.au/environment/biodiversity/threatened/communities/giant-kelp

The Tasmania snapshot reflects documented long-term decline (approximately 95% since the 1950s) driven by the southward extension of the warm East Australian Current. Quantitative canopy values are not populated for this snapshot. Status is classified as critical.

### Status classification
Snapshot status (healthy / stressed / critical) is an editorial judgement based on: canopy trend direction and magnitude, absolute coverage relative to historical extent, the presence and severity of known stressors, and published scientific assessments of each region. It is not a standardised metric and should not be compared directly across bioregions with different monitoring histories.

---

## 3. Where Kelp Has Been Lost (Replacement Sites)

### Purpose
This layer documents 25 sites worldwide where kelp forest has been lost and replaced by a persistent alternative ecological state. It is intended to illustrate the mechanisms that drive and maintain kelp loss — and why many affected reefs do not recover passively — rather than to provide a statistically complete inventory of all affected areas.

### Site selection
Sites were selected to represent:

- **Geographic breadth** — coverage across all major kelp bioregions (North-east Pacific, South-east Pacific, North Atlantic, North-east Atlantic/Iberia, Norway/Arctic, Mediterranean, Japan/Korea, Australasia, New Zealand)
- **Mechanistic diversity** — the primary replacement pathways (see classification below)
- **Evidential grounding** — each site is documented in the peer-reviewed literature, grey literature, or official government/agency reporting. Sites with no traceable external documentation were excluded.
- **Scientific significance** — preference given to sites frequently cited in the kelp ecology literature as reference or foundational cases

Sites were not selected based on spatial representativeness; the set should not be interpreted as implying that only 25 such sites exist globally. The IUCN and Kelp Forest Alliance have estimated that kelp decline is measurable across more than a third of all surveyed bioregions.

### Data collection method
Data were compiled through a systematic review of:

1. **Peer-reviewed journal articles** — primary source for mechanism, species identity, and year of documentation where possible
2. **Government and agency reports** — used for the Japanese national *isoyake* programme, Norwegian Environment Agency, New Zealand DOC, and Fisheries Agency of Japan guidelines
3. **Marine research institution syntheses** — IMAS (Tasmania), Bigelow Laboratory for Ocean Sciences (Maine), NIVA/Nofima (Norway), UC Santa Cruz (California)
4. **Invasion databases** — Smithsonian NEMESIS (Non-indigenous Estuarine and Marine Species) database for *Sargassum horneri* and *Undaria pinnatifida* records at Channel Islands and Baja California sites

For each site the following were recorded where available: place name; representative coordinates; country/region; replacement state; the organism or process maintaining the alternative state; year of the key documenting study or first formal observation; an editorial confidence rating; a plain-language description of the documented transition; and bibliographic references.

Barren extent in hectares is *not* provided for any site; consistent hectare estimates across bioregions were not available in the literature at the required geographic precision.

### Replacement state classification
Each site is assigned one of five replacement-state categories.

**Urchin barren.** The most widely documented pathway globally. Dense sea urchin populations maintain a grazing front that prevents kelp recruitment and removes adult canopy. Barrens are self-reinforcing: starved urchins persist in a dormant mode, consuming any settling kelp sporelings, and can maintain the barren state for decades without external intervention. *Key drivers:* removal of predators (sea otters, lobster, large fish); climate stress reducing kelp resilience; trophic cascades.
*Sites:* Fort Bragg–Mendocino (CA), Monterey Peninsula (CA), Adak Island (AK), Haida Gwaii (BC), St. Margaret's Bay (NS), Bicheno (Tasmania), Hammerfest (Norway), Kamoenai (Japan), Shiriya (Japan), Mejillones (Chile), Cape Rodney–Okakari Point (NZ).

**Turf algae.** Low-lying mats of filamentous or articulated coralline and non-coralline algae replace kelp canopy. Turf is maintained by warming, eutrophication, sediment loading, and epiphyte pressure on kelp. Chemical compounds released by certain turf assemblages inhibit kelp spore settlement, creating a biochemical lock-in independent of continued disturbance. *Key drivers:* warming; eutrophication; epiphyte loading; sediment trapping.
*Sites:* Jurien Bay–Kalbarri (WA, Australia), Skagerrak Coast (Norway), Southern Maine (USA), Penobscot Bay (USA), Ría de Vigo (Spain).

**Invasive seaweed.** Non-native seaweeds — primarily *Sargassum horneri* (from Asia) and *Undaria pinnatifida* (wakame, from Japan) — establish during warming events and displace native kelp. Marine heatwaves create recruitment windows that invasives exploit more successfully than native species. *Key drivers:* marine heatwaves; ballast water / hull fouling introductions; warming winter minima.
*Sites:* Todos Santos Islands (Baja California), Anacapa Island (CA), Santa Catalina Island (CA).

**Kelp loss — fish grazing / CCA barren.** Two related but distinct pathways involving non-urchin grazers or crustose coralline algae (CCA): *fish grazing (tropicalisation)*, where warming allows subtropical herbivorous fish (especially *Siganus* spp., rabbitfish) to extend into formerly temperate kelp zones and graze fronds more intensively than the ecosystem evolved to withstand; and *CCA barren*, where kelp beds are replaced by a smooth carpet of crustose coralline algae following physical disturbance or chronic stress, with urchin grazing not the primary maintaining mechanism.
*Sites:* Geomun-do (South Korea) — fish grazing; Maemul-do (South Korea) — fish grazing; Jeju Island (South Korea) — CCA barren; Cíes Archipelago (Spain) — herbivorous fish; Tosa Bay (Japan) — compound urchin/fish.

**Isoyake barren (Japan-specific).** *Isoyake* (磯焼け) is the Japanese fisheries term for a condition of depleted or absent seaweed beds, often diagnosed as barren or near-barren reef. It encompasses multiple replacement states and is maintained by a combination of urchin grazing, fish grazing, sedimentation, and warming. Reported as a distinct category to preserve the classification used in Japanese government monitoring programmes.
*Sites:* Mio (Wakayama), Tosa Bay (Kochi).

### Confidence ratings
Each site carries an editorial confidence rating reflecting the strength and specificity of the supporting documentation.

- **High** — documented in peer-reviewed literature with a named DOI or widely cited journal article; or the subject of a formal government/institute scientific report that names the site explicitly.
- **Medium** — referenced in synthesis reports, database records, or institutional grey literature; mechanism and organism are consistent with the regional literature, but the primary source for the specific site is not a standalone peer-reviewed paper.
- **Low** — not used; all retained sites have at least medium confidence.

### Limitations
- **Point representation:** Each site is a single coordinate; the actual affected area may span many kilometres of coastline.
- **Temporal snapshot:** The documented year reflects when the transition was first formally documented, not necessarily the onset of decline. Many transitions were gradual over years before formal study.
- **Incomplete coverage:** The set is illustrative, not exhaustive. Large areas of documented kelp decline (e.g., much of southern Australia, Atlantic Morocco, West Africa) lack site-specific published documentation sufficient to include.
- **No area quantification:** Barren extent in hectares is not provided; comparable area estimates across bioregions are not available in the literature at the required precision.
- **Dynamic states:** Some sites documented as barrens have since undergone partial recovery (e.g., Kamoenai, Cape Rodney); the record represents the documented transition event, not necessarily current conditions.

---

## 4. Species Occurrence Records

Two independent global biodiversity databases are presented as opt-in researcher layers, each accompanied by an explicit caveat that record density reflects survey effort, not kelp abundance.

### 4A. OBIS — Ocean Biodiversity Information System
**OBIS.** Laminariales occurrence records. Intergovernmental Oceanographic Commission of UNESCO. www.obis.org

OBIS aggregates verified species occurrence records from research institutions, monitoring programmes, and citizen science globally. Order-level records for **Laminariales** (the kelp order), plus *Durvillaea* (bull kelp, Southern Hemisphere) where appropriate, are aggregated into ~1.40625° hex-bin cells. Bin intensity is scaled to log₁₀(record count) so single-record cells appear faint relative to intensively surveyed areas.

### 4B. GBIF — Global Biodiversity Information Facility
**GBIF.** Laminariales occurrence records. Copenhagen: GBIF Secretariat. www.gbif.org

GBIF aggregates specimen, observation, and literature occurrence records from natural history collections and monitoring programmes. Laminariales records are aggregated to 1° grid cells, with intensity scaled to log₁₀(record count) per cell.

### Caveat
Both layers reflect where surveys have been conducted. High record density in some regions (California, Tasmania, Norway) reflects extensive scientific monitoring, not necessarily higher kelp abundance relative to less-studied coasts.

---

## 5. Tasmanian Kelp Time-Series (IMAS)

### Source
**Johnson, C.R., et al. / IMAS, University of Tasmania.** Tasmanian Giant Kelp Landsat Survey, 1987–2015. www.imas.utas.edu.au

The IMAS Landsat survey mapped giant kelp (*Macrocystis pyrifera*) canopy extent around Tasmania across multiple survey periods from 1987 to 2015 using Landsat satellite imagery. Two spatial products are included: a coarser **statewide** product showing canopy coverage at the coast-wide scale across survey years, and a higher-resolution **fine-scale** product for intensively surveyed locations.

Both show canopy **presence** at each survey period — where kelp was observed, not a calculated change or loss raster. Comparing extents across years reveals the long-term contraction. The data does not encode a binary lost/retained classification.

### Caveat
The survey covers 1987–2015 only. Conditions may have changed since 2015.

---

## 6. Population Proximity Estimates

Reported as "People within 50 km of kelp" for each region.

### Source
**European Commission, Joint Research Centre (JRC).** Global Human Settlement Layer — Population Grid (GHS-POP), 2025 epoch, 1 km resolution, Mollweide projection (ESRI:54009). https://human-settlement.emergency.copernicus.eu/ghs_pop2023.php

### Method
Population within 50 km of kelp habitat was estimated for each of the 22 snapshot regions:

1. **Kelp buffer** — A 50 km buffer was generated around the global kelp distribution and reprojected to the GHS-POP raster coordinate system (Mollweide).
2. **Per-region clipping** — For each snapshot region, a ±5° bounding box around the snapshot coordinate was intersected with the buffer to extract the relevant local buffer parts. This prevents double-counting where buffers from different regions overlap.
3. **Raster extraction** — Population pixel values within each clipped buffer were read from the GHS-POP raster, filtered for the nodata sentinel value, and summed.
4. **Normalisation** — The raw global raster sum (491,637,335,418) exceeds the 2025 world population (~8.2 billion) by approximately 60×, reflecting the raster's arbitrary-unit encoding. Results were corrected by a scale factor of 8,200,000,000 / 491,637,335,418 ≈ 0.0167.

### Interpretation and caveats
Figures are estimates only. They represent people living within 50 km of the *modelled kelp habitat boundary*, not within 50 km of currently observed canopy. They are intended to contextualise the human connection to kelp ecosystems — the coastal population living in proximity to, and potentially dependent on, kelp-associated ecosystem services (fisheries, shoreline protection, cultural value). The ±5° regional clipping may undercount populations in large, continuous kelp zones where habitat extends well beyond the snapshot bounding box (e.g., the full NE Pacific coastline).

---

## General Limitations

- **Kelp habitat vs. canopy:** The UNEP-WCMC biome layer shows modelled potential habitat, not current canopy. Large portions of the boundary have experienced kelp loss not reflected in the polygon.
- **Snapshot status is editorial:** Health classifications (healthy / stressed / critical) are interpretive and not directly comparable across bioregions with different monitoring histories.
- **Survey effort bias:** OBIS and GBIF occurrence records reflect where surveys have been conducted, not necessarily kelp abundance.
- **Population estimates are approximate:** See the regional clipping caveat above.
- **IMAS Tasmania extent:** Covers 1987–2015 only; conditions may have changed since.
- **Kelp Watch licensing:** Kelp Watch data is provided under an as-is research disclaimer. Written permission for editorial publication use should be confirmed with The Nature Conservancy prior to public launch.

---

## Sources Summary

| Topic | Source | Citation |
|---|---|---|
| Kelp habitat (biome) | UNEP-WCMC | Global kelp distribution dataset, data.unep-wcmc.org/datasets/7 |
| NE Pacific, S. America, S. Africa snapshots | Kelp Watch | Bell et al., kelpwatch.org; Landsat 1984–2025 |
| Nordic/Arctic snapshots | Kvile et al. 2022 | *Frontiers in Marine Science*, doi: 10.3389/fmars.2022.850359 |
| Tasmania snapshot | IMAS / EPBC Act | IMAS Landsat Survey; DCCEEW EPBC listing |
| Species occurrences | OBIS | obis.org; Laminariales records |
| Species occurrences | GBIF | gbif.org; Laminariales records, 1° grid |
| Tasmania time-series | IMAS Landsat | Johnson et al., IMAS, 1987–2015 |
| Population proximity | JRC GHS-POP 2025 | European Commission JRC, 1 km, Mollweide |

### Replacement-site bibliography

**Peer-reviewed with DOI**

1. **Rogers-Bennett, L. & Catton, C.A. (2019).** Marine heat wave and multiple stressors tip bull kelp forest to sea urchin barrens. *Communications Biology*, 2, 383. https://doi.org/10.1038/s42003-019-0548-2 — *Fort Bragg–Mendocino; the 2014–2016 California bull kelp collapse.*
2. **Wernberg, T., et al. (2013).** An extreme climatic event alters marine ecosystem structure in a global biodiversity hotspot. *Nature Climate Change*, 3, 78–82. https://doi.org/10.1038/nclimate1723 — *Jurien Bay–Kalbarri; the 2011 Ningaloo heatwave-driven kelp-to-turf transition.*
3. **Félix-Loaiza, A.P., et al. (2022).** Occurrence and dynamics of *Sargassum horneri* in Baja California. *Botanica Marina*, 65(3), 171–185. — *Todos Santos Islands; S. horneri establishment during the 2014–2016 heatwave.*
4. **Choi, C.G., et al. (2024).** Tropicalisation-driven kelp loss in the South Sea of Korea. [Full citation to be confirmed.] — *Geomun-do, Maemul-do.*
5. **Chapman, A.R.O. (1981).** Stability of sea urchin dominated barren grounds following destructive grazing of kelp in St. Margaret's Bay, eastern Canada. *Marine Biology*, 62(4), 307–311. — *St. Margaret's Bay; foundational description of persistent barren state.*
6. **Estes, J.A., et al. (1998).** *Science*, 282, 473–476. https://doi.org/10.1126/science.282.5388.473 — foundational killer whale–otter–urchin trophic cascade work underpinning *Adak Island, Monterey Peninsula, Haida Gwaii.*

**Institutional and grey literature**

- **IMAS eastern Tasmania science synthesis**, University of Tasmania — *Centrostephanus rodgersii* range extension and barren formation. *Bicheno.*
- **Bigelow Laboratory for Ocean Sciences (2025; 2026)**, Maine coastwide kelp monitoring, including turf allelopathy assays. *Southern Maine, Penobscot Bay.*
- **Christie, H., Norderhaug, K.M. & Fredriksen, S. (NIVA)**, with the **Norwegian Environment Agency** Skagerrak coastal assessment. *Skagerrak Coast.*
- **Fisheries Agency of Japan**, guidelines for kelp forest restoration through urchin removal (documents 1–3 ha/year recovery at Tohoku reference sites). *Shiriya.*
- **Watanuki, A., et al.**, urchin removal and kelp restoration at Kamoenai fishing port, Hokkaido. *Kamoenai.*
- **NIVA / Nofima**, northern Norwegian urchin barrens and active restoration via urchin removal. *Hammerfest.*
- **Vásquez, J.A., Piaget, N. & Alonso Vega, J.M.**, long-term kelp and urchin barren time series, northern Chile. *Mejillones Peninsula.*
- **Barrientos, S., et al.**, kelp collapse within the Islas Atlánticas National Park. *Cíes Archipelago.*
- **Smithsonian NEMESIS**, non-indigenous species records for *Sargassum horneri* and *Undaria pinnatifida*, Pacific coast of North America. https://invasions.si.edu/nemesis/ — *Anacapa Island, Santa Catalina Island, Todos Santos Islands.*
- **California Sea Grant**, *Sargassum horneri* invasion biology and management. *Anacapa Island, Santa Catalina Island.*
- **DOC New Zealand**, with **Blue Parks / Marine Conservation Institute** and **NZ Ministry for Primary Industries**, kina barren dynamics and Leigh reserve monitoring. *Cape Rodney–Okakari Point (Leigh).*
- **UC Santa Cruz, Long Marine Laboratory**, Aleutian kelp and trophic cascade research. *Adak Island, Monterey Peninsula, Haida Gwaii (contextual).*
- **Tosa Bay survey reports**, urchin removal and barren-ground monitoring, Kochi coast. *Tosa Bay.*
- **Mio physical-factors study**, Wakayama Prefecture sediment comparison of degraded vs. intact reef. *Mio.*
- **Korean / Frontiers in Marine Science synthesis** of *isoyake* and kelp dynamics around Jeju Island, including CCA barren formation. *Jeju Island.*
- **NW Iberia invasive-turf study**, *Asparagopsis armata* cover on degraded kelp reefs in the Ría de Vigo system. *Ría de Vigo.*

---

*Document prepared by Sunstone Institute research team. For questions about methodology or data sources, contact contact@sunstoneinstitute.ai.*