One Ecosystem : Research Article
PDF
Research Article
National-scale mapping of ecosystems to improve ocean accounting for marine and coastal management in Indonesia
expand article infoTeguh Satria Gunawan‡, Marsha Hamidah§, Agavia Kori Rahayu§,|, Nabila Nur Septiani§, Jessica Pingkan§, Agus Hermansyah, Muhammad Farhan§, Rusdatus Sholihah‡, Angela Belladova Arundina‡, Diah Retno Minarni#, Rahmatia Susanti#, Gin Gin Gustiar#, Desi Nurulita Kusumastuti¤, Gabriella Rosya Maharani¤, Muhammad Dimas Nurhakim¤, Putri Vency Khalishah¤, Irgi Fadilah Rahman¤, Nicky Nugianto¤, Amehr Hakim‡, Firdaus Agung‡, Annisya Rosdiana§, Heidi Retnoningtyas§, Intan Destianis Hartati§,¤, Efin Muttaqin§, Sophia German«, Jordan Gacutan«, Irfan Yulianto§,¤
‡ Directorate of Aquatic Ecosystems and Species Conservation, Ministry of Marine Affairs and Fisheries, Jakarta, Indonesia
§ Fisheries Resource Center of Indonesia of Rekam Nusantara Foundation, Bogor, Indonesia
| Faculty of Law, University of New South Wales, Sydney, Australia
¶ DHI Indonesia, Jakarta, Indonesia
# Geospatial Information Agency, Cibinong, Indonesia
¤ Faculty of Fisheries and Marine Sciences, IPB University, Bogor, Indonesia
« Centre for Sustainable Development Reform, Faculty of Law, University of New South Wales, Sydney, Australia
Corresponding author: Irfan Yulianto (irfan@rekam.org)
Academic editor: Joachim Maes
Received: 08 Apr 2025 | Accepted: 11 May 2025 | Published: 21 May 2025
© 2025 Teguh Satria Gunawan, Marsha Hamidah, Agavia Kori Rahayu, Nabila Nur Septiani, Jessica Pingkan, Agus Hermansyah, Muhammad Farhan, Rusdatus Sholihah, Angela Belladova Arundina, Diah Retno Minarni, Rahmatia Susanti, Gin Gin Gustiar, Desi Nurulita Kusumastuti, Gabriella Rosya Maharani, Muhammad Dimas Nurhakim, Putri Vency Khalishah, Irgi Fadilah Rahman, Nicky Nugianto, Amehr Hakim, Firdaus Agung, Annisya Rosdiana, Heidi Retnoningtyas, Intan Destianis Hartati, Efin Muttaqin, Sophia German, Jordan Gacutan, Irfan Yulianto
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation: Gunawan TS, Hamidah M, Rahayu AK, Septiani NN, Pingkan J, Hermansyah A, Farhan M, Sholihah R, Arundina AB, Minarni DR, Susanti R, Gustiar GG, Kusumastuti DN, Maharani GR, Nurhakim MD, Khalishah PV, Rahman IF, Nugianto N, Hakim A, Agung F, Rosdiana A, Retnoningtyas H, Hartati ID, Muttaqin E, German S, Gacutan J, Yulianto I (2025) National-scale mapping of ecosystems to improve ocean accounting for marine and coastal management in Indonesia. One Ecosystem 10: e155166. https://doi.org/10.3897/oneeco.10.e155166
 Open Access

Abstract

This study presents a comprehensive national-scale mapping of Indonesia's coastal ecosystems — coral reefs, mangroves and seagrass — using Sentinel-2 and SPOT satellite imagery. The mapping covers 2018 and 2021, validated with ground-truthing and secondary data and spatial analyses were conducted using modelling and digitisation. This study revealed changes in ecosystem extent where coral reefs increased from 1,212,207.46 ha in 2018 to 1,216,249.74 ha in 2021, while seagrass expanded from 273,122.60 ha to 273,950.87 ha and mangroves expanded from 3,329,459.72 ha to 3,364,769.05 ha. Despite these overall increases, localised declines were observed due to human-driven degradation, particularly in Fisheries Management Areas 571, 573, 711, 713, 714, 715 and 716. The study highlights the importance of accurate spatial data for ocean accounts, aiding in the calculation of ecosystem services and providing information for marine spatial planning, marine protected areas and fisheries management. We also addressed challenges, including data limitations, technological infrastructure, methodological advancements and time constraints. These findings underscore the need for integrated management and conservation efforts to maintain and enhance the resilience of coastal and marine ecosystems.

Keywords

ecosystem extent, ecosystem services, geographic information system, spatial analysis, geospatial application

Introduction

Global frameworks, including the Sustainable Development Goals and Global Biodiversity Framework, are driving momentum for integrated approaches to coastal and marine management that balance environmental, social and economic priorities (Carroll et al. 2022). These commitments emphasise the need to incorporate biodiversity and ecosystem health into decision-making processes. Countries with extensive maritime areas have seen an increased demand for diverse data streams, to implement frameworks, such as Integrated Coastal Zone Management (ICZM) and Marine Spatial Planning (MSP) (Gacutan et al. 2022). Diverse datasets are needed to align the environmental, economic and social targets within the ocean space and measure and report progress towards targets and commitments.

Coastal areas pose a particular challenge for spatial planning and management, as there are a diverse range of human activities that depend on ecosystems such as coral reefs, mangroves and seagrasses. These ecosystems provide critical services beyond direct resource use. For example, coral reefs support coastal protection against wave erosion (Yuan et al. 2024); seagrass beds store approximately 10 – 18% of the total Corg buried in the global ocean (Fu et al. 2023) and mangroves enhance coastal resilience through flood control and shoreline stabilisation (Koch et al. 2009). However, these non-use benefits are often poorly measured and inadequately considered in decision-making, leading to management approaches that may overlook their full value to society. Additional frameworks are needed to go “Beyond GDP’ in understanding the contribution of ecosystems to society and the economy in an integrated manner (Dasgupta 2021).

Ocean accounting offers a solution by providing a comprehensive framework to integrate environmental, economic and social data related to ocean resources and ecosystems (GOAP 2021) and its use is endorsed by the Convention on Biological Diversity Decision 15/24 (CBD 2022). Implementation has emerged globally, driven by the need to communicate science to policy and management in a manner aligned with traditional economic accounts responsible for GDP and other economic indicators (Gacutan et al. 2022, Trueb et al. 2024). Ocean Accounts (OA) are aligned with the UN System of Environmental Economic Accounting Ecosystem Accounting (SEEA EA), which provides guidance for measuring ecosystems and generating accounts and statistics within a spatial context (UNSD 2021), enabling the assessment of both direct use values and broader ecosystem services.

Geospatial information forms a critical foundation for OA by enabling the spatial integration of data across various domains, i.e. environmental, social and economic data. Unlike broader natural capital approaches, OA explicitly links the distribution of ecosystems with the location of ecosystem service provision and economic activities (GOAP 2021). Ecosystem accounts are built on geospatial data describing ecosystem extent and condition, which, when combined with socioeconomic data, reveal where and how these ecosystems benefit society and specific economic sectors (Guannel et al. 2016). This spatially-explicit framework enables decision-makers to monitor changes in both ecosystem health and service provision over time while understanding the spatial dependencies between ecosystems and their beneficiaries (Marfai 2023). When maintained over time, these spatially-integrated accounts could facilitate the rapid assessment of environmental change and its implications for dependent communities and sectors (Fenichel et al. 2020).

The development of comprehensive OA requires accurate spatial delineation of coastal ecosystems, yet this process presents significant challenges. The calculation of ecosystem assets is complex, requiring precise data collection and subsequent analysis (Meraj et al. 2021), particularly when accounting for dynamic coastal environments. These complexities are especially apparent when linking physical ecosystem data with economic valuations and information about their social-ecological functions (Wahyudin et al. 2023). While accurate economic valuation of ecosystem services is crucial for illustrating their economic significance (Phelan et al. 2020), this valuation depends heavily on the quality and resolution of underlying spatial data.

Technical advances in Geographic Information Systems (GIS) and remote sensing have enabled higher resolution ecosystem mapping (Gutierres et al. 2016), yet significant challenges remain in mapping seagrass, mangroves and coral reefs globally. Large-scale efforts at resolutions adequate for accounting are limited by access to high-resolution imagery (i.e. beyond 30 m Landsat data), inadequate hardware and gaps in advanced classification methods (Purwanto et al. 2022, Karang et al. 2024). These technical limitations are compounded by insufficient funding and expertise (Fortes et al. 2018, Hafizt et al. 2024), spatial data that is fragmented across institutions and gaps in remote sensing knowledge (Triana and Wahyudi 2021).

Several countries have undertaken national-scale coastal ecosystem mapping initiatives with varying success. The United States' National Oceanic and Atmospheric Administration (NOAA) has mapped benthic habitats in shallow-water coral reef ecosystems, though focused on specific remapped areas (Monaco et al. 2012). Australia has developed the Seamap National Benthic Habitat including underwater communities and salt marshes (Lucieer et al. 2023), while China (Hu et al. 2020) and Belize (Cissell et al. 2021) have conducted national mangrove mapping. However, these efforts often represent a specific ecosystem at a single point in time, rather than temporal monitoring of the extent of multiple ecosystems.

In Indonesia, ecosystem mapping has been conducted across local and regional scales, reflecting the nation's commitment to understanding and managing its marine biodiversity (Triana and Wahyudi 2021, Purwanto et al. 2022, Hafizt et al. 2024). Pilots to develop OA have emerged in 10 areas and have advanced these mapping efforts through multi-institutional collaboration between key agencies, helping to overcome traditional barriers of fragmented spatial data and limited technical resources. By establishing frameworks for collaborative mapping and monitoring of coastal ecosystems, the OA approach has enabled more systematic data collection and validation across multiple regions, supporting both national accounting needs and local management objectives.

This study aims to support OA implementation in Indonesia by conducting a comprehensive national-scale mapping of the temporal and spatial distribution of mangroves, coral reefs and seagrasses. Building on previous mapping efforts, we assess changes in ecosystem extent for the years 2018 and 2021, generating baseline data for the three ecosystems, that can provide information for both ecosystem accounting and marine resource management. This study:

  1. presents the inter-institutional approach for nation-wide mapping;
  2. assesses changes in coastal ecosystem extent across Indonesia's marine jurisdiction by Fisheries Management Area (FMA);
  3. identifies links between the national extent maps and the measurement of ecosystem services and supporting evidence-based marine management in Indonesia.

Materials and Methods

Study Area

The mapping process was carried out in phases from 2022 to 2024 across all Indonesian regions, based on the eleven FMAs of the Republic of Indonesia’s for coral reef, seagrass and mangrove ecosystems (Fig. 1). Field surveys were conducted to validate the digitised data in three provinces, comprising Central Java (Kebumen, Demak, Cilacap and Jepara), Central Papua (Mimika) and Maluku (Tanimbar Islands and Aru Islands), as well as five national marine protected areas/MPAs (Pieh, Gili Matra, West Waigeo, Raja Ampat, Banda Sea, Padaido Islands and Anambas Islands) (Table 1). Secondary data were used as input for areas not directly surveyed (Table 2).

Table 1.
Download as
CSV
XLSX

Site, ecosystem and number of observation points per field survey. Field surveys were conducted to assess the extent and condition of mangrove, coral reef and seagrass ecosystems. The field survey was collaboratively conducted by the MMAF, BIG and Rekam Nusantara Foundation (RNF) during Indonesia’s OA Pilot in 2021-2023.

No.

Survey sites

Ecosystems

Number of observation points

1

Kebumen, Demak, Cilacap, dan Jepara (Central Java)

Mangrove

15

2

Mimika (Central Papua)

105

3

Tanimbar Islands (Maluku)

Coral reef and seagrass

211

4

Aru Islands (Maluku)

343

5

West Waigeo Islands and Surrounding Waters MPA

433

6

Raja Ampat Islands and Surrounding Waters MPA

337

7

Banda Sea MPA

112

8

Palaido Islands and Surrounding Waters MPA

132

9

Anambas Islands and Surrounding Waters MPA

1,594

10

Pieh MPA

128

11

Gili Matra MPA

Coral reef, seagrass and mangrove

361

Total

3,771
Table 2.
Download as
CSV
XLSX

The location, number of observation points and source of secondary data on Indonesia’s coastal ecosystem conditions from 2018 to 2021. Data were sourced from relevant scientific and technical reports.

No.

Locations

Number of observation points

References

1

Banggai, Banggai Laut, and Banggai Kepulauan District, Central Sulawesi

10

Sallata et al. (2022)

2

Batanglampe Island, Sinjai District, South Sulawesi

8

Permatasari et al. (2023)

3

Kedindingan Island, Bontang City, East Kalimantan

4

Samin et al. (2023)

4

Situbondo, East Java

20

Fuad et al. (2022)

5

Morowali Waters, South Sulawesi

6

Salanggon et al. (2022)

6

Liukang Tuppabiring Region Waters, South Sulawesi

6

Tenri et al. (2020)

7

Menjangan Island, Bali

30

Lingga (2016)

8

Kutai Kartanegara District, East Kalimantan

12

East Kalimantan Marine and Fisheries Office (2022)

9

Ternate and Tidore Islands, North Maluku

11

Patty et al. (2024)

10

Matahora Island, Wakatobi District, Southeast Sulawesi

4

Didi et al. (2018)

11

Panjang Island, Jepara, Central Java

14

Suryono et al. (2022)

12

Kelapan Island, South Bangka District, Bangka Belitung

4

Amrillah et al. (2019)

13

Tanimbar Islands, West Maluku Tenggara District, Maluku

14

Anggraeni et al. (2019)

14

Kapoposang Island, South Sulawesi

4

Rosalina et al. (2022)

15

Beloreng Waters, Tembeling, Bintan District

31

Putri et al. (2018)

16

Panjang Island, Central Bangka, Bangka Belitung

4

Septiani et al. (2022)

Total

182
Figure 1.

Extent mapping area coverage of Indonesia’s coastal ecosystems. Indonesia’s eleven fisheries management areas (FMAs) are distributed across the Indian Ocean (571, 572, 573) and the Pacific Ocean (711, 712, 713, 714, 715, 716, 717, 718). Locations of ground-truth data are denoted by dots.

Mapping Strategy

Data Inventory

Data inventory involved collecting and organising information to ensure availability and provide information for decision-making. This included gathering data from relevant ministries and institutions, specifically for mangrove, seagrass and coral ecosystems, with the goal to identify available data and sharing procedures. Key data types included vector data (ecosystem boundaries and survey points) and raster data (satellite and aerial imagery). In Indonesia, five key ministries/institutions manage relevant data: the Ministry of Environment and Forestry (KLHK) for mangroves, the Centre for Oceanographic Research of the Indonesian Institute of Sciences (LIPI) (now The National Research and Innovation Agency of The Republic of Indonesia (BRIN)) for seagrasses and coral reefs, the Ministry of Marine Affairs and Fisheries (MMAF) for coastal ecosystems, the Indonesian Space Agency of the National Research Agency and Innovation (ORPA-BRIN) for remote sensing data and the Geospatial Information Agency (BIG) for spatial data integration under the OneMap Policy.

Establishing Governance Framework

The coastal ecosystem mapping process involved the formation of a task force made up of several ministries and institutions to coordinate marine data resources. According to the Decree of the Minister of Marine Affairs and Fisheries No. 44/2023, this task force consists of five working groups and a secretariat, two of which were relevant to this study. The Thematic Geospatial Information Group (TGIG) for Marine Resource Accounting and the Data Management Group focused on ecosystem mapping. These groups, which include BIG, MMAF, BRIN and RNF, are responsible for gathering, compiling, analysing and presenting data, as well as developing standards and procedures for marine resources management. The Data Management Group ensured the coordination and consolidation of all data resources.

Data Management and Sharing Plan

Data management and sharing were critical given the involvement of multiple ministries and institutions. Data management involved the collection and analysis of data to ensure accessibility, integrity, quality and secure storage.

In this case, the data collected from various ministries were stored in a cloud-based database. Procedures were established to regulate data-sharing mechanisms between parties, as the data are owned by different custodians. This phase included inventorying available data that can be shared, setting policies for data access and defining the process flow for data distribution.

Monitoring and Evaluation

Monitoring and evaluation were crucial steps to ensure that analysed data met established standards. This process involved identifying indicators to assess the quality of the analysis results. A commonly used tool for this is the quality control form, which helped ensure the completeness of data based on its structure. This was done in reference to the Indonesian National Standard (SNI) 9257:2024, which specifies the metadata standards and the quality of government-acceptable geospatial information — particularly for spatial accounting of coastal and benthic shallow marine habitats.

Data Sources

Geospatial data were compiled from primary and secondary sources. Primary data collection was carried out through field surveys to gather ground-truth data in three ecosystems: coral reef, seagrass and mangrove. Surveyors snorkelled in the observed area, following a survey transect. The transect was determined, based on the indicative variability of habitat classes, as estimated from the preliminary unsupervised classification of satellite imagery. The starting point of the survey transect was adjusted according to the water conditions. The surveyor then swam from the shore parallel to the coastline, marking habitat classes and coverage percentages on a tally sheet, while taking pictures at each stop point. The stop points were determined, based on specific time intervals or distances (ranging from 1 to 3 minutes or 10 m ± 1 m), as well as variations in the coverage conditions and habitat classes encountered, following the methodology outlined in the SNI 9257:2024. Secondary data sources include land data from the Indonesian Topographic Map to determine the land boundaries, imagery from SPOT 6/7 satellites, imagery from Sentinel-2 satellites and ESRI Imagery Basemaps (Table 3).

Table 3.
Download as
CSV
XLSX

Geospatial data and sources used to map the extent of changes in coastal ecosystems in Indonesia between 2018 and 2021. Satellite imagery data were mostly acquired from ORPA-BRIN. See Supplementary Material 1 for details on satellite imagery.

Data

File Extension

Prefix

Data Format

Use

Spatial Resolution

Dataset Creation Year

Source

SPOT 6/7 Satellite Imagery

jp2

IMG_

Raster

Data source for image analysis to determine ecosystem type and extent

5 m

2018, 2021 (Suppl. material 1)

Airbus Defence and Space imagery acquired by Indonesian Space Agency

Sentinel-2 Satellite Imagery

jp2

T43NCC_

Rater

Data source for image analysis to determine ecosystem type and extent

10 m

2018, 2021 (Suppl. material 1)

ESA Copernicus imagery, acquired by the Indonesian Space Agency

Digital Topographic Map

shp

Vector

Boundary for coastlines

2021

Geospatial Information Agency, Indonesia

National Mangrove Map

shp

Vector

Baseline data for mangrove ecosystem extent

2021

Ministry of Environment and Forestry, Indonesia

OneMap Policy

shp

Vector

Boundary for the model builder process

2016

 Geospatial Information Agency, Indonesia

Marine Protected Area Zoning Map

shp

Vector

As input for area of interest for validation

2021

Ministry of Marine Affairs and Fisheries, Indonesia

Spatial Analysis

The coral reef, seagrass and mangrove ecosystems were mapped to account for changes in extent from 2018 (opening year) to 2021 (closing year) using spatial modelling and remote sensing techniques.

Shallow Water Benthic Ecosystems Mapping

Shallow water benthic ecosystems, namely coral reef and seagrass were mapped using SPOT 6 and 7 satellite imagery (Fig. 2). The first step involved the pre-processing of satellite imagery, which includes geometric correction, sun glint correction, as well as water column correction using the Lyzenga algorithm (Lyzenga 1981). These processes improve the image quality by reducing disturbances in the water column so that objects can be more clearly detected, thereby increasing the accuracy of the image classification (Prasetyo and Helmi 2020). In more detail, water column correction using the Lyzenga algorithm utilised the differences in attenuation coefficients of each wavelength to estimate water depth, based on variations in the logarithm of reflectance between spectral bands — typically red, green and blue (Pratomo et al. 2024). In images that had not undergone water column correction, deeper areas appeared darker even when the substrate was the same, making it difficult to accurately distinguish substrate types. After applying water column correction, the influence of water depth was minimised, allowing the reflectance to more accurately represent the characteristics of the substrate. The training dataset, consisting of processed SPOT 6 images and field data from the Numfoor Island, Biak Numfoor Recency, Papua, was then selected and used to train the model. Next, shallow water areas, namely the benthic area from the shoreline up to the reef crest, were extracted from the satellite images by clipping them using the ancillary boundary data. Then, image enhancement was conducted using the density slicing method, resulting in a maximum of 10 distinct segments. Ecosystem maps were then created using the threshold values obtained from the model training to convert probability maps into binary classifications, where 1 represented the intended ecosystem (i.e. coral reef or seagrass) and 0 denoted other areas (i.e. waterbodies, rocks, sand, rubble, macroalgae, mud, mangroves and other substrates). The resulting ecosystem extent map was validated by cross-referencing with ground-truth data, basemaps and pre-existing habitat maps.

Figure 2.

Shallow water benthic ecosystems (i.e. coral reef and seagrass) mapping workflow.

Mangrove Ecosystem Mapping

Mangrove mapping was conducted following the standardised methodology outlined in SNI 7717:2020. In this study, we primarily utilised SPOT 6 and 7 satellite imagery. However, if the SPOT imagery were compromised by factors, such as excessive cloud cover and became difficult to map, additional Sentinel-2 imagery was used to help interpret the same location and timeframe. The process began with pre-processing, which included geometric and radiometric corrections. This step was automatically performed because, according to Gascon et al. (2017), Sentinel-2 imagery had undergone a calibration and validation process, indicating that it had been radiometrically and geometrically corrected. Mangrove ecosystems were then delineated through on-screen digitisation using ArcGIS 10.8, guided by the interpretation methods and keys specified in SNI 7717:2020, while also referencing the previously published 2021 National Mangrove Map. False colour composite imagery, which enhanced mangrove ecosystems, was utilised to aid the digitisation process (Fig. 3). Finally, the digitisation results were validated using ground-truth field data to ensure its accuracy.

Figure 3.

False colour composite of Sentinel-2 image highlighting the mangrove distribution in the east coast of Sumbawa Island, Indonesia. The distinct appearance of mangroves, namely in a dark orange/brown colour, served as an identification key for digitising the ecosystem.

Results

The findings highlight changes in extent for three coastal ecosystems between 2018 and 2021 (Figs 4, 5, 6). In 2018, the area of coral reefs was 1,212,207.46 ha. By 2021, this area had increased by 4,042.28 ha, resulting in a total of 1,216,249.74 ha. The area of seagrass in 2018 was 273,122.60 ha, which increased to 273,950.87 ha in 2021, reflecting a rise of 828.27 ha. The area of mangroves increased by 35,309.33 ha, with the 2018 area recorded at 3,329,459.72 ha and the 2021 area at 3,364,769.05 ha (Fig. 7).

Figure 4.

Maps of coral reefs in 2018 and 2021, along with selected sample sites illustrating changes in coral reef extent over those years.

Figure 5.

Maps of seagrass in 2018 and 2021, along with selected sample sites illustrating changes in seagrass extent over those years.

Figure 6.

Maps of mangrove in 2018 and 2021, along with selected sample sites illustrating changes in mangrove extent over those years.

Figure 7.

Comparison of the total area of each ecosystem and a Sankey diagram showing their extent changes between 2018 and 2021.

The expansion of coral reefs, seagrass and mangrove areas did not imply that all regions experienced an overall increase. In some places, these areas decreased. This is reflected in the Sankey diagram (Fig. 7), where some mangrove areas transitioned to non-mangrove between 2018 and 2021 and certain areas of coral reef and seagrass in 2018 became macroalgae and other substrate by 2021. The fisheries management areas experiencing mangrove degradation include FMA 571, FMA 573, FMA 711, FMA 713, FMA 714, FMA 715 and FMA 716. Seagrass degradations were found in FMA 572, FMA 573, FMA 711, FMA 713, FMA 714, FMA 716 and FMA 717, while coral reef degradations were found in FMA 572, FMA 573, FMA 712, FMA 714, FMA 715 and FMA 716. The highest level of mangrove degradation occurred in FMA 716, were seagrass and coral reefs occurred in FMA 711 and 714, respectively (see Table 4).

Table 4.
Download as
CSV
XLSX

Extent of each coastal ecosystem by FMA in hectare (ha): (a) coral reef, (b) seagrass and (c) mangrove.

FMA

Ecosystem Extent in 2018 (ha) Ecosystem Extent in 2021 (ha)

Changes

Coral Reef

571

917.34

917.34

0.0%

572

69,187.69

69,184.88

0.0%

573

53,197.89

53,148.25

-0.1%

711

95,224.68

95,611.43

0.4%

712

44,300.23

44,293.51

0.0%

713

189,365.04

192,496.82

1.7%

714

272,790.81

272,150.82

-0.2%

715

180,990.18

180,635.87

-0.2%

716

84,939.80

84,450.94

-0.6%

717

94,525.52

94,595.99

0.1%

718

126,768.29

128,763.87

1.6%

Seagrass

571

131.75

131.75

0.0%

572

6,983.01

6,934.62

-0.7%

573

20,565.00

18,989.13

-7.7%

711

33,534.50

31,063.00

-7.4%

712

8,054.63

8,452.83

4.9%

713

33,552.80

32,242.19

-3.9%

714

97,297.83

96,070.46

-1.3%

715

31,512.84

31,420.16

-0.3%

716

19,002.22

18,509.01

-2.6%

717

13,544.08

13,463.51

-0.6%

718

8,943.94

16,674.20

86.4%

Mangrove

571

157,997.50

157,465.41

-0.3%

572

40,245.62

40,336.21

0.2%

573

38,371.07

37,742.49

-1.6%

711

632,932.24

617,404.21

-2.5%

712

99,245.01

108,811.38

9.6%

713

244,812.43

239,774.90

-2.1%

714

134,751.04

133,632.43

-0.8%

715

555,813.20

542,077.01

-2.5%

716

289,789.07

270,974.97

-6.5%

717

173,518.97

193,489.70

11.5%

718

961,983.57

1,023,060.35

6.3%

Discussion

Marine and Coastal Ecosystem Mapping for Indonesia’s Ocean Accounting

This study resulted in the nationwide extent mapping of key marine and coastal ecosystems in Indonesia in 2018 and 2021 (Table 4). Notable discrepancies in ecosystem extent estimates can be observed between mapping results (Table 5), most likely due to differences in data collection and spatial analysis approaches (e.g. field data curation, satellite image resolution, classification algorithms and GIS processing methods). For mangrove and seagrass ecosystems, maps developed by the Indonesian government produced similar results, yet considerable differences were apparent when compared to independent mapping efforts. In the case of mangrove mapping, the discrepancies might have resulted from the spatial resolution of satellite imageries. The Global Mangrove Watch utilised ALOS/PALSAR, Landsat and Sentinel-2 data (Bunting et al. 2022), with 10 m, 30 m and 10 m spatial resolution, respectively, while this study and KLHK’s national mangrove maps were derived from 6 m resolution SPOT imageries. Meanwhile, this study’s coral reef extents are more aligned to Allen Coral Atlas, as opposed to BIG’s OneMap data and KLHK’s Coral Atlas. This difference of over 1 million ha is possibly the result of classification errors, as OneMap’s spatial analysis may have categorised sand, rubble and seagrass as coral reefs.

Table 5.
Download as
CSV
XLSX

The ecosystem extent (ha) of this study in comparison to previous nation-wide and/or global coastal ecosystem mappings.

Map

Mapped Year

Ecosystem Type

Source

Mangrove (ha)

Seagrass (ha)

Coral Reef (ha)

This Study

2018

3,329,459.72

273,122.60

1,212,207.46

Spatial Analysis

This Study

2021

3,364,769.05

273,950.87

1,216,249.74

Spatial Analysis

Indonesia’s Portion of the Global Mangrove Map

2020

2,901,578.22

-

-

Global Mangrove Watch Data

Indonesia National Mangrove Map

2013-2019

3,311,207.00

-

-

KLHK Data

Indonesia National Mangrove Map

2021

3,364,080.00

-

-

KLHK Data

Allen Coral Atlas

2022

-

564,973.00

1,011,087.00

Allen Coral Atlas (2022)

Indonesia OneMap Policy

2016

-

-

2,517,858.00

BIG Data

Indonesia Coral Monitoring Atlas

2015-2021

-

-

2,530,060.00

KLHK (2021)

Indonesia National Seagrass Map

2018

-

293,464.00

-

LIPI (2018)

Geospatial information is important for providing information for Indonesia's sustainable development. National-scale mapping of coastal ecosystems in Indonesia has been previously attempted by both the government of Indonesia and independent parties with different ecosystem extent results (Table 5). In comparison to other ecosystems, national-scale seagrass mapping in Indonesia remains sparse, with few comprehensive efforts preceding this study. The most notable and only prior Indonesian-led initiative was conducted by the Indonesian government in 2018, employing moderate-resolution imagery from Landsat 8 OLI. This mapping effort was conducted by the LIPI, now restructured under BRIN, in collaboration with the Coral Reef Rehabilitation and Management Program - Coral Triangle Initiative (COREMAP-CTI). In 2023, the governance of thematic geospatial information on seagrass and coral reef underwent a transition, as, according to the BIG Regulation No. 16/2023, the seagrass and coral reef data custodian was officially transferred from the BIG to the MMAF. This study represents a pivotal development as it delivers Indonesia’s first high-resolution national-scale seagrass mapping effort, led by the MMAF. The resulting map, endorsed by the government, is specifically designed to support the nation’s ocean accounting activities.

Implication for Ecosystem Service Estimation

By mapping the extent of these ecosystems across two time periods, we have established a spatial foundation for OA (see La Notte (2024)) that can track changes in ecosystem extent and identify shifts from one ecosystem type to another during an accounting period (UNSD 2021). Then, we can define the contribution factor for each ecosystem change. For example, the degradation of ecosystems in certain regions of Indonesia (as shown in Table 4 and Fig. 7) was primarily driven by human activities. In FMA 711, where most of the mangrove and seagrass degradation occurred, factors such as mangrove conversion contributed to their degradation (Eddy et al. 2021, Yudistira and Agustriani 2023), while pollution, port development and boat transportation contributed to seagrass degradation (Sari et al. 2017, Rosalina et al. 2022). Similarly, in FMA 714, coral reef degradation had been exacerbated by destructive fishing practices and unsustainable tourism (Husseini 2020, Estradivari et al. 2022). Meanwhile, the mangroves in FMA 718 that are the largest in extent and the healthiest in Indonesia, playing a key role as a nursery ground and contributing to the significant fish production in Indonesia (Tirtadanu et al. 2022, Edyvane et al. 2024), received low pressure due to their remote area and less development and, hence, the extent increased from 2018 to 2021.

Ecosystem extent affects other accounts, such as ecosystem services accounts, where extent values are often included in calculating ecosystem services (Fauzi et al. 2023, Rahayu et al. 2024). Coral reefs, seagrasses and mangroves provide a diverse array of ecosystem services that are critical for environmental health and human well-being. Each type of service — provisioning, regulating, supporting and cultural — plays a unique role in sustaining ecosystems and supporting livelihoods. The degradation of coral reef, seagrass and mangroves undermines their ability to deliver such vital services. Coral reef loss affects fish populations and coastal protection and can lead to seagrass decline, increased sedimentation, reduced water quality and mangrove deforestation, diminishing flood protection and carbon sequestration (Marcos et al. 2021). According to Fauzi et al. (2023), through a nationwide literature review in Indonesia, coral reefs, mangroves and seagrass provide a range of ecosystem services. These include food provision, recreation (excluding seagrass), education and research, wave attenuation (excluding seagrass), biodiversity support, marine habitat creation and food supply for marine organisms. Additionally, seagrass and mangroves provide unique benefits, such as raw materials and carbon sequestration. The value of national wide ecosystem services for coral reef, seagrass and mangrove are 12,145, 14,953 and 18,544 USD.ha-1.year-1, respectively (1 USD = 15,000 IDR (Fauzi et al. 2023)).

These issues highlight the urgent need for integrated management and conservation efforts. Protecting and restoring coastal and marine environments is vital for maintaining their functions and benefits. These efforts are essential for ensuring that ecosystems can provide invaluable services and for enhancing the resilience of both natural and human systems.

Management Implications

The OA is crucial for assessing the state and value of marine and fisheries resources within MPAs. Effective management of MPAs requires comprehensive data on marine biodiversity, fish stocks and ecosystem health (Deutz et al. 2020). By incorporating detailed OA, policy-makers can better understand the ecological and economic value of these areas (McCrea-Strub et al. 2011). In this case, OA provide a systematic framework for tracking changes in marine and coastal ecosystems' extent and conditions, such as coral reefs, seagrass and mangroves. These data are essential for evaluating the effectiveness of MPAs in achieving conservation goals and ensuring sustainable fisheries management.

The inclusion of economic valuations in OA can underscore the benefits of MPAs beyond immediate conservation gains. The economic valuation of a particular ecosystem can be measured, based on ecosystem services from the specific area of that ecosystem, meaning they need information on the ecosystem extent. By quantifying the economic value of ecosystem services, such as tourism, fisheries and carbon sequestration, OA provide a compelling case for the continued investment in and expansion of MPAs (Ban et al. 2011). This economic perspective helps bridge the gap between environmental and economic goals, highlighting the long-term benefits of preserving marine ecosystems. Well-structured OA support informed decision-making, facilitate stakeholder engagement and promote the sustainable management of marine resources within protected areas.

As OA track progress towards conservation and economic goals, they also support marine spatial planning (MSP) by providing information for decisions on the allocation of space and resources for human activities, as well as guiding the planning, implementation and evaluation of MSP initiatives (Gacutan et al. 2022). In Indonesia’s national context, OA can provide data that would become the basis for marine use permit issuance, assist in the formulation of MSP documents and aid in determining incentives and disincentives for marine resource users and managers (MMAF 2022).

Furthermore, the accounting for fisheries management involves systematically collecting, analysing and application of data on marine ecosystems to support sustainable fisheries. By providing detailed data on ecosystems extent, ecosystem condition and fish stocks, ocean accounting enables more informed and precise management decisions, helping to set appropriate catch limits and protect marine and coastal ecosystems. Additionally, it helps ensure that fishing practices are sustainable by monitoring the health of marine ecosystems and assessing the impacts of fishing activities. This leads to policies that balance fishery exploitation with ecosystem health. The OA can identify trends and changes in marine environments, providing early warnings of potential issues, such as overfishing or habitat degradation, which allows for timely management interventions.

Multi-institutional Collaboration

This study revealed a lesson learned on multi-institution collaboration. The data collection process was carried out by BIG and the MMAF supported by RNF, while BRIN provided the SPOT 6 and 7 images throughout Indonesia with two datasets in 2018 and 2021. The data were shared amongst institutions for further processing to obtain the extent of changes in each ecosystem. Meanwhile, the mapping process involved cross-ministerial or institutional teams including MMAF, BIG, BRIN and RNF. For achieving a standardised mapping result and establishing a skill and knowledge baseline across institutions, the technicians were re-educated in the technical (e.g. the methodology — how to digitise, how to run the programmes or software) and normative (e.g. the identification keys for benthic and coastal habitats) aspects. Furthermore, the team established a mutually agreed spatial scale for visual digitisation. The team also held periodic meetings to monitor progress and evaluate mapping results.

Limitations

Several technical challenges were encountered during the mapping process. Training datasets often differed from satellite imagery in terms of size, resolution and spectral and radiometric characteristics. These discrepancies could lead to classification errors, emphasising the need for temporal, spectral and geometric compatibility between training datasets and satellite images (Malczewska et al. 2023).

Regarding size, spatial resolution differences referred to the smallest object that could be detected in an image, determined by pixel size. The smaller the pixel, the higher the level of detail that could be captured. In this study, the training data were derived from Landsat imagery with a spatial resolution of 30 x 30 m, whereas the imagery used for processing was obtained from SPOT 6/7 satellites, which had a finer resolution of 5 x 5 m. This difference could result in inconsistencies in how objects were represented across datasets.

In terms of spectral resolution, discrepancies concerned the sensors' ability to capture information across various wavelengths of the electromagnetic spectrum. Landsat imagery consisted of more spectral bands (11 bands) compared to SPOT imagery (5 bands), enabling more detailed object detection, based on spectral characteristics. However, these variations in the number and range of bands could affect feature extraction processes and classification performance.

Regarding radiometric resolution, which indicated a sensor's sensitivity to detect subtle differences in brightness levels, it also played a critical role. Both Landsat 8 and SPOT 6/7 possessed 12-bit radiometric resolution, allowing them to distinguish up to 4,096 variations. Despite having equivalent bit depth, differences in sensor type and image acquisition conditions could influence classification accuracy — especially in benthic habitat mapping, where reflectance differences amongst underwater substrates were often minimal due to light attenuation in water.

These technical differences could diminish the effectiveness of classification algorithms, as models trained on one dataset might not generalise well to imagery with differing characteristics. Consequently, classification accuracy could vary across geographic areas and time periods, leading to inconsistencies in the delineation of ecosystem extents.

Other challenges arose due to constraints in field data collection, including budgetary and environmental conditions. Considering budgetary constraints, ground-truthing in this study was only performed in priority sites (e.g. MPAs, important fishery areas) with a limited number of sample points. Furthermore, ground-truth points used for validation were often unevenly distributed, limiting their ability to represent all environmental conditions within certain areas. This implicated both the model builder and visual digitisation processes. The limited data caused model training to be less than optimal, while the limited amount of data and their uneven distribution did not provide sufficient references for GIS technicians in making decisions to classify coastal ecosystems, particularly benthic habitats. Efforts to save data collection expenses produced trade-offs in terms of reduced map accuracy (Tulloch et al. 2017). Additionally, the lack of effective inter-institutional coordination resulted in data being scattered across different institutions and not being properly managed.

Way Forward

In summary, this study provides the first high-resolution national-scale maps of Indonesia’s mangrove, seagrass and coral reef ecosystems, establishing a spatial foundation for integrating ocean accounting into marine conservation planning. By identifying changes in ecosystem extent and linking them to socio-economic conditions, the study supports data-driven decision-making for marine protected areas, fisheries management and spatial planning. These findings represent a crucial step towards aligning Indonesia’s ocean management with sustainable development and climate resilience goals and underscore the importance of ongoing multi-agency collaboration and investment in ocean monitoring systems.

In the future, the focus will be on improving field data collection to ensure the accuracy and relevance of the information gathered. Additionally, data processing methods will be enhanced to optimise the analysis of the information obtained. To maintain the relevance of environmental monitoring, extensive data will be updated regularly, at least once every three years. These efforts will be supported by collaborative partnerships involving both government and private sector entities, ensuring that problem-solving is conducted effectively and comprehensively.

Furthermore, there is a need for regular acquisition of satellite images and increased capacity for image analysis. The government of Indonesia needs to purchase satellite imagery on a regular basis or cooperate with international organisations to ensure the availability of the necessary satellite imagery, given the vastness of Indonesia's seas, which requires collaborative action. Additionally, the capacity to quickly analyse satellite imagery is needed, considering the vastness of Indonesia's seas, for example, by developing site-specific algorithms for coral reefs and seagrasses.

Funding

This work was funded in part by the UK International Development funds as part of the Blue Planet Fund. The funding is provided by the UK Department for Environment, Food and Rural Affairs through the Global Ocean Accounts Partnership.

Data Availability Statement

The data presented in this study are available on request from the corresponding author due to legal conditions. Restrictions apply to the certain data that were obtained from Indonesian Research and Innovation Agency (BRIN) and are available with their permission.

Acknowledgements

The authors would like to express their appreciation for the Global Accounts Partnership (GOAP), the Indonesian National Research and Innovation Agency (BRIN), the Geospatial Information Agency (BIG) and the Indonesian Ministry of Marine Affairs and Fisheries (MMAF) for their continued long-term collaboration and cooperation. Finally, the authors extend their gratitude to everyone at the Rekam Nusantara Foundation, specifically the Fisheries Resource Center of Indonesia unit, who were directly or indirectly involved in this research.

Author contributions

Conceptualisation, I.Y., A.Ha. and F.A.; methodology, T.S.G., M.H., J.P., A.H. and D.R.M.; formal analysis, T.S.G., M.H., J.P., A.H., M.F., R.S., A.B.A, R.S., G.G.G, D.N.K., G.R.M., M.D.N., P.V.K., I.F.R, N.G. and I.Y.; investigation, T.S.G., M.H., A.K.R., N.N.S. J.P. and A.H.; data curation, T.S.G., M.H., J.P., A.H., D.R.M., R.S., G.G,G., I.D.H. and E.M.; writing—original draft preparation, T.S.G., M.H., A.K.R., N.N.S. J.P. and I.Y.; writing—review and editing, T.S.G., M.H., A.K.R., N.N.S. J.P., D.R.M., A.Ha., F.A., A.R., H.R., S.G., J.G. and I.Y.; visualisation, M.H., J.P. and IY; supervision, F.A. and I.Y.; project administration, A.R., H.R. and I.Y.. All authors have read and agreed to the published version of the manuscript.

Conflicts of interest

The authors have declared that no competing interests exist.

References

Supplementary material

Suppl. material 1: List of SPOT 6/7 and Sentinel-2 images used in this study 
Authors:  Marsha Hamidah
Data type:  Table
Download file (40.00 kb)
login to comment