Sacred heritage and environmental sustainability: carbon and biodiversity insights from Taiwan’s Fo Guang Shan Monastery and Buddha Museum

Religious landscapes have historically played a vital role in cultural, spiritual, and environmental stewardship, serving as centers for social cohesion, ecological conservation, and long-term sustainability practices^1,2. While their significance has been well-documented in heritage studies, their contributions to carbon sequestration and climate change mitigation remain underexplored in contemporary environmental discourse³. As global frameworks such as the United Nations Sustainable Development Goals (SDGs) and the Paris Agreement emphasize the urgent need for carbon sequestration strategies, sacred landscapes—including monastery forests, temple ecosystems, and church woodlands—present a unique but underutilized opportunity for nature-based climate solutions^3,4.

Studies have increasingly highlighted the biodiversity conservation value of religious landscapes. For example, recent research has emphasized the role of sacred natural sites in maintaining ecological integrity, with studies documenting how these landscapes serve as biodiversity hotspots and contribute to ecosystem resilience^5,6. Furthermore, monastic forest management practices in regions such as Southeast Asia have demonstrated significant contributions to long-term habitat preservation and carbon sequestration⁷. These findings underscore the broader conservation benefits of religious landscapes, reinforcing their potential for integration into global climate adaptation strategies. Ethiopian church forests function as biodiversity refuges, where native species are preserved under religious protection, even amidst widespread deforestation pressures in the region^8,9. Similarly, sacred groves in India have been maintained for centuries through religious taboos and traditional ecological knowledge, demonstrating high levels of plant diversity and passive carbon storage¹⁰. Beyond Africa and South Asia, Tibetan Buddhist sacred landscapes, including Mount Kawa Karpo and Iran’s Zagros forests, illustrate how religious worldviews influence sustainable ecological practices^1,6

Despite these precedents, there remains limited empirical research quantifying the carbon sequestration potential of Buddhist temple ecosystems in East Asia. This study addresses this gap by integrating high-resolution remote sensing, landscape carbon modeling, and comparative conservation frameworks to assess the climate benefits of Buddhist temple landscapes. Unlike previous research that has primarily focused on qualitative assessments or regional biodiversity surveys, this study employs advanced remote sensing techniques and standardized carbon measurement methodologies to provide a rigorous, data-driven evaluation of the sequestration potential of religious landscapes. Particularly in the context of contemporary environmental policy and carbon accounting frameworks⁵.

This study evaluates Fo Guang Shan Monastery (FGSM) and Fo Guang Shan Buddha Museum (FGSBM)—two interconnected religious sites in Kaohsiung, Taiwan—to assess the carbon sequestration potential of Buddhist temple landscapes (Fig. 1).

• FGSM, established in 1967, serves as the spiritual and administrative headquarters of Fo Guang Shan Buddhism, comprising a monastery, pilgrimage zones, and extensive forested land¹¹.

• FGSBM, completed in 2011, is one of the largest Buddhist cultural institutions in East Asia, featuring extensive gardens, tree-lined pathways, and managed green spaces designed to harmonize religious heritage with environmental conservation¹².

a Aerial view of the Fo Guang Shan complex, showing the monastic grounds, Buddha Museum axis, and surrounding green infrastructure. b Front elevation of the Buddha Museum entrance, highlighting the Eight Pagodas and the symmetrical layout of the pilgrimage plaza. Images (a, b) sourced from Wikimedia Commons and used under the Creative Commons Attribution-Share Alike 4.0 license (https://commons.wikimedia.org/wiki/File:Aerial_View_of_Fo_Guang_Shan.png; https://commons.wikimedia.org/wiki/File:Fo_Guang_Shan_Buddha_Museum_%E4%BD%9B%E5%85%89%E5%B1%B1%E4%BD%9B%E9%99%80%E7%B4%80%E5%BF%B5%E9%A4%A8_.jpg).

Full size image

These sites create an extensive religious and ecological landscape, seamlessly integrating afforestation, cultural heritage preservation, and sustainable landscape management. Unlike Ethiopian church forests^2,9, which rely on passive conservation, and Indian sacred groves, which are community-driven^13,14,15, the FGSM and FGSBM represent an actively managed conservation model, combining afforestation, reforestation, and strategic landscape planning to ensure long-term ecological sustainability.

In this study, temple forest refers to vegetated areas actively managed within monastic domains for spiritual, ecological, and cultural purposes^2,6,13. These include afforested lands under religious protection, typically maintained by Buddhist institutions through routine maintenance, selective planting, and symbolic use. Temple landscape, on the other hand, denotes the broader spatial configuration encompassing sacred architecture (e.g., stupas, halls), ritual paths, cultivated greenery, and conservation zones that reflect religious land use and environmental ethics^1,14. This dual framework captures both the ecological structure and spiritual function embedded in Buddhist-managed territories and provides the spatial basis for our assessment of carbon storage and biodiversity.

This study employs a geospatial and remote sensing-based approach combined with ecological modeling techniques to evaluate the carbon sequestration potential of FGSM and FGSBM. We first assessed vegetative health and canopy coverage using the Normalized Difference Vegetation Index (NDVI), which provided insights into vegetation’s spatial distribution and density^16,17. Carbon stock was subsequently estimated through three complementary methods¹: the InVEST Carbon Storage and Sequestration model, which incorporated spatially explicit land-use data and four major carbon pools^2,18; the i-Tree Canopy model, which calculated aboveground biomass and estimated air pollutant removal based on land cover types¹⁹; and³ the Intergovernmental Panel on Climate Change (IPCC) Tier 1 methodology, applying standardized coefficients to estimate total biomass carbon, including root-shoot ratios and carbon fractions^20,21. Carbon sequestration estimates were then translated into economic values using international carbon market prices (e.g., EU ETS, CARB, ACCU), enabling a comprehensive evaluation of climate benefits^19,21,22. Geographic Information System (GIS) tools supported the integration of spatial datasets and analytical procedures, ensuring accurate georeferencing, land cover classification, and zonal carbon distribution analysis²³.

In addition to modeling vegetation and soil carbon, we analyzed landscape fragmentation using FRAGSTATS, a widely used tool for quantifying spatial configuration and ecological connectivity^24,25,26. This model evaluates key fragmentation metrics, such as Patch Density (PD) and Edge Density (ED), to assess the spatial arrangement of afforested areas and green spaces within the study site^27,28,29. FRAGSTATS helps determine how spatial heterogeneity influences carbon retention efficiency by measuring habitat connectivity and fragmentation^30,31. This analysis is essential because landscape fragmentation can disrupt carbon sequestration by increasing edge effects, reducing biomass continuity, and altering microclimatic conditions. The inclusion of FRAGSTATS in this study provides quantitative evidence of how green corridors and afforested zones contribute to optimizing carbon storage efficiency³². Integrating these methodological tools ensures a comprehensive assessment of carbon sequestration across the FGSM and FGSBM landscapes, positioning the study within a broader comparative framework of religious conservation strategies worldwide.

The findings of this study carry significant implications for global climate policy, heritage conservation, and sustainable site management. The research provides empirical evidence of carbon sequestration within Buddhist temple ecosystems, reinforcing the potential for integrating religious landscapes into carbon offset programs. This issue is particularly relevant to Taiwan’s 2024 Carbon Fee Policy, which underscores the need for nature-based solutions in climate mitigation³³. Furthermore, this study aligns with international conservation frameworks, particularly the UNESCO World Heritage Convention, which recognizes cultural landscapes’ ecological significance and heritage value³⁴. UNESCO policies emphasize the protection and sustainable management of sacred natural sites, integrating biodiversity conservation with cultural preservation. These policies support the inclusion of religious landscapes in environmental conservation initiatives, acknowledging their role in carbon sequestration and climate adaptation strategies. Additionally, these policies align with emerging carbon credit markets, advocating for recognizing cultural landscapes as carbon sinks.

This study builds on the methodological and policy framework outlined above to address key research questions that highlight the role of Buddhist temple landscapes in carbon sequestration and sustainable heritage management. Doing so aims to provide empirical insights into the intersection of religious heritage and climate mitigation strategies.

The following key research questions guide this study:

1. 1.

   How do Buddhist temple landscapes contribute to carbon sequestration, and how do they compare with other religious conservation models?

2. 2.

   What is the carbon sequestration potential of FGSM and FGSBM, and how does it align with global climate action goals?

3. 3.

   How do institutional conservation strategies at Buddhist sites influence carbon storage efficiency compared to community-driven or naturally preserved religious landscapes?

Study site and cultural-ecological context

The FGSM and FGSBM, located in Dashu District, Kaohsiung, Taiwan, form an extensive religious, cultural, and ecological complex. Established in 1967 by Venerable Master Hsing Yun, FGSM serves as the spiritual and administrative headquarters of the Fo Guang Shan Buddhist Order. The FGSBM, inaugurated in 2011, extends this mission by integrating Buddhist teachings, cultural heritage preservation, and religious tourism. These sites exemplify a holistic approach to monastic practice, heritage conservation, and sustainable landscape management^11,12. FGSM and FGSBM sit at 22.7276° N and 120.4042° E within a tropical monsoon climate zone, where hot, humid summers and mild winters prevail. The area receives an average annual rainfall of 2000 mm between May and September. This climate fosters lush vegetation and diverse ecological landscapes, reinforcing the sacred atmosphere of the complex while supporting its environmental conservation initiatives.

Land cover composition and spatial framework

Remote imagery used in Fig. 1 provides a visual overview of the FGSM and FGSBM landscape. Image (a), an aerial view of the complex, and image (b), a frontal perspective of the Buddha Museum, were retrieved from Wikimedia Commons and are used under the Creative Commons Attribution-Share Alike 4.0 license. These images serve as orientation references and contain no identifiable individuals. All visual materials comply with the journal’s image publication policy.

The FGSBM spans approximately 100 hectares, integrating traditional Buddhist architectural elements with contemporary cultural and educational functions. The Buddha Memorial Center (BMC), positioned at the heart of the museum, enshrines a sacred Buddha tooth relic, serving as a focal point for devotional practice and pilgrimage. Eight pagodas surround the central structure, each symbolizing an aspect of the Noble Eightfold Path, which reinforces the philosophical and didactic dimensions of the museum. The Great Buddha Hall, dominated by a 108-meter bronze Buddha statue, is a defining feature of the complex, reflecting architectural grandeur and religious symbolism (Fig. 2). Figure 2 presents images obtained from two sources. Image (b), showing the Main Shrine of FGSM, comes from Wikimedia Commons under the Creative Commons Attribution-Share Alike 3.0 license. Images (a), (c), and (d) originate from the FGSBM and reflect authorized use under formal written permission. All images underwent ethical review to confirm the absence of recognizable individuals. The visual materials align with the journal’s image publication policy.

a The 108-meter Bronze Buddha statue (佛光大佛), a monumental structure symbolizing wisdom, compassion, and enlightenment. b The Main Shrine of Fo Guang Shan (佛光山大雄寶殿), serving as the spiritual and cultural center of Buddhist practice. c The Sacred Buddha Tooth Relic (佛牙舍利), a revered relic embodying spiritual legacy and devotion. d The Buddhist Botanical Garden (佛教植物園), a serene landscape dedicated to the preservation of sacred plant species and ecological harmony. Image (b) sourced from Wikimedia Commons and used under the Creative Commons Attribution-Share Alike 3.0 license (https://commons.wikimedia.org/wiki/File:Fo_Guang_Shan_Monastery_13.jpg). Images (a, c, d) are provided by the FGSBM and used with written permission.

Full size image

The FGSM spans 50 hectares, incorporating monastic residences, administrative buildings, meditation halls, and sacred shrines. Although the combined area of the temple and its associated museum covers 150 hectares, official tourism maps indicate that only 55 hectares are designated for public access, tourism, and pilgrimage (Fig. 3). This distribution suggests that most of the site is allocated for monastic activities, ecological conservation, and restricted areas, preserving its spiritual and environmental integrity.

This map was used to delineate the study boundaries of the FGSM and the FGSBM.

Full size image

Although precise woodland data is unavailable, qualitative insights from key informant interviews indicate that built-up structures occupy less than 20% of the total area. In contrast, greenspace, including landscaped gardens, forested areas, and vegetative land covers, comprises more than 50%. These green spaces contribute significantly to carbon sequestration, biodiversity conservation, and microclimate regulation, aligning with Buddhist ecological ethics emphasizing harmonious coexistence between human activities and nature^1,2,6,14,35.

Beyond its monastic and architectural significance, FGSM and FGSBM are interdisciplinary research hubs, fostering academic inquiries into heritage conservation, sustainable tourism, and environmental management¹². The museum’s integration of cultural heritage preservation with contemporary sustainability practices positions it as a model for heritage site conservation and religious landscape management. The surrounding forested and landscaped areas, designed by Buddhist ecological principles, play a crucial role in carbon sequestration, habitat preservation, and climate resilience.

To provide spatial clarity, Fig. 4 presents high-resolution satellite imagery of FGSM and FGSBM, illustrating the distribution of built structures, green spaces, and ecological buffer zones. This geospatial analysis enables a quantitative evaluation of the site’s cultural and ecological functions, ensuring analytical coherence with broader conservation and sustainability research^36,37.

The inset maps provide regional context, showing the site’s ___location in Taiwan and its relation to Kaohsiung. This map serves as a reference for defining the study area.

Full size image

In this study, GIS-based classification and remote sensing techniques—including NDVI and supervised land cover classification—were applied to refine estimates of green space coverage and assess its ecological contributions^38,39,40,41. These analyses provide further insights into the role of religious landscapes in sustainable environmental stewardship and heritage conservation.

Remote sensing data acquisition and temporal framework

To assess land cover changes, this study employs Pleiades-1A and Pleiades-1B high-resolution satellite imagery, operated by Airbus Defence and Space, due to their 0.5-meter spatial resolution and stereo imaging capabilities, which facilitate detailed land-use classification and vegetation analysis⁴². Compared to alternative satellite sensors, such as Sentinel-2 (10 m resolution) and Landsat-8 (30 m resolution), Pleiades provides superior spatial precision, enabling the accurate detection of micro-scale land-use changes and vegetation dynamics. This higher resolution is particularly beneficial for assessing subtle variations in vegetation density, anthropogenic land modifications, and conservation effectiveness.

Three datasets—2016, 2019, and 2022—were selected to assess temporal changes in vegetation cover, land-use transitions, and conservation practices within the study site (Fig. 5, Table 1). The selection of 2016 as the baseline year was determined by data availability, as this was the earliest Pleiades satellite imagery accessible from the Center for Space and Remote Sensing Research, National Central University, Taiwan (CSRSR). To ensure a comprehensive longitudinal assessment, subsequent datasets from 2019 and 2022 were incorporated, allowing for the analysis of long-term environmental trends in relation to heritage site management. This temporal framework enables a detailed evaluation of the impact of sustainable conservation strategies, afforestation efforts, and climate variability on landscape stability and ecological resilience⁴³. As the analysis was concluded in 2024, this dataset selection provides a methodologically robust and temporally consistent basis for monitoring heritage landscapes through remote sensing techniques.

Satellite imagery of FGSM and FGSBM captured by Pleiades-1A and Pleiades-1B, illustrating its spatial extent, landscape features, and surrounding environment as a geospatial reference for the study.

Full size image

These datasets provide multi-temporal insights into vegetation health, built environment expansion, and conservation effectiveness, supporting a comprehensive geospatial analysis of the FGSM and FGSBM landscape evolution.

Vegetation health assessment using NDVI

We employed the NDVI to evaluate vegetation health and detect land-cover changes⁴⁴. This widely used spectral vegetation index quantifies plant vigor by measuring the difference in reflectance between near-infrared (NIR) and red light⁴⁵. Healthy vegetation exhibits strong reflectance in the NIR spectrum (700–1100 nm) while absorbing most red light (620–750 nm) for photosynthesis⁴⁶. The NDVI calculation follows the standard formula, as shown in Eq. (1)^47,48:

The index produces values ranging from −1 to 1, where higher values indicate greater vegetation density and vitality. NDVI values above 0.6 typically correspond to dense, healthy vegetation, while values between 0.2 and 0.6 indicate moderate vegetation cover. Values below 0.2 suggest sparse vegetation or non-vegetated surfaces. By offering a quantitative measure of vegetation health, NDVI is instrumental in assessing ecological stability, monitoring land degradation, and evaluating climate-induced shifts in vegetation patterns⁴⁹.

Carbon storage estimation: InVEST, i-Tree canopy, and IPCC approaches

We employed the InVEST Carbon Storage and Sequestration Model, developed by the Natural Capital Project, to assess carbon stock dynamics within the FGSM and FGSBM. The model provides a spatially explicit assessment of carbon storage based on land cover classifications, integrating four major carbon pools: aboveground biomass, belowground biomass, soil organic carbon, and dead organic matter^23,37,50. The model quantifies carbon sequestration potential by analyzing temporal variations in land cover and vegetation, supporting heritage conservation strategies and climate adaptation policies^18,51.

This study utilized high-resolution Pleiades-1A and Pleiades-1B satellite imagery to generate land cover maps for 2016, 2019, and 2022, employing Support Vector Machine (SVM) classification within ArcGIS Pro (3.0) for data processing^38,41. The classification system identified key vegetation types, including evergreen broad-leaved forests and grasslands, both of which function as primary carbon sinks^39,40. To enhance spatial accuracy and boundary delineation, administrative boundary data and conservation zoning layers were integrated into the analysis of the FGSM-FGSBM site.

We assigned carbon stock values to each land cover class based on IPCC (2019) guidelines, supplementing them with regional ecological datasets for improved estimation accuracy. Each land cover type was attributed carbon pool coefficients, representing aboveground biomass, belowground biomass, soil organic carbon, and dead organic matter^{52,53,54,55,56}. These coefficients were applied uniformly across the study area, enabling the calculation of total carbon storage per unit area and facilitating spatial comparisons over time.

The total carbon storage for each land parcel (Ci) was calculated by summing the contributions from four primary carbon pools, as shown in Eq. (2) and detailed in Table 2⁵²:

In this equation, C_i denotes the total carbon stock in land parcel i, measured in tons per hectare (t/ha). The terms C_ia, C_ib, C_is, and C_id represent the carbon stored in aboveground biomass, belowground biomass, soil organic carbon, and dead organic matter, respectively, all expressed in t/ha.

The cumulative carbon storage (C_total) for the entire study area was determined by aggregating the carbon storage values of individual land parcels, as presented in Eq. (3):

In this equation, C_total represents the total carbon stored within the study area, while \(S_i\) denotes the area of each land cover type, expressed in hectares (ha).

The net carbon sequestration (ΔC) between the selected periods was calculated according to Eq. (4):

In this equation, C_t1 denotes the carbon storage in the initial year (2016), C_t2 represents the carbon storage in the final year (2022), and ΔC quantifies the net change in carbon stocks, expressed in tons of CO₂ per year.

This approach detects land-use-induced carbon gains or losses, providing insights into afforestation, deforestation, and conservation-driven carbon sequestration dynamics^54,55. The InVEST carbon model was applied to estimate carbon storage and sequestration using IPCC Tier 1 default values and regional datasets. The resulting carbon stock estimates were analyzed using GIS-based spatial mapping techniques, producing high-resolution carbon distribution maps to visualize sequestration patterns across the FGSM-FGSBM landscape.

The i-Tree Canopy Model, developed by the USDA Forest Service, was used to estimate tree canopy coverage, impervious surface extent, and vegetation composition within FGSM and FGSBM. The model employs randomized point sampling of high-resolution aerial imagery, coupled with manual classification and validation, to ensure high accuracy in land cover estimation⁵⁷.

In addition to estimating biomass, we used the i-Tree Canopy model to calculate annual air pollutant removal by tree canopy. The model assigns pollutant-specific removal rates to each land cover class and quantifies total removal for major urban air pollutants, including ozone (O₃), nitrogen dioxide (NO₂), sulfur dioxide (SO₂), carbon monoxide (CO), and fine particulate matter (PM_2.5). These rates originate from empirical studies conducted by the USDA Forest Service and form part of the model’s default parameters. To assess economic benefits, we converted removal quantities into monetary values based on avoided healthcare costs and environmental damage, following the valuation framework embedded in the model.

GIS software delineated a geo-referenced boundary of the 57.85-hectare study area and imported it into i-Tree Canopy. We systematically generated 1016 randomized sampling points within this boundary and used high-resolution aerial imagery from Google Earth for classification⁵⁸.

We manually classified each sampling point into four primary land cover types: tree canopy, grassland/shrubland, impervious surfaces, and water bodies. Classification accuracy was ensured through systematic verification and cross-referencing with high-resolution satellite imagery and ancillary spatial data^9,59. Any discrepancies were resolved using cross-validation techniques and expert judgment, ensuring consistency and reliability in land cover identification. The model automatically calculated the percent land cover for each category, estimating vegetation distribution and urbanization extent⁵⁷.

We calculated carbon sequestration potential using the i-Tree Canopy tool, incorporating region-specific parameters calibrated to reflect local tree species and environmental conditions in Taiwan^59,60. To ensure consistency in carbon stock assessments, we compared sequestration estimates from i-Tree Canopy with those obtained from the InVEST Carbon Storage and Sequestration Model. Cross-validating these two models enhanced methodological rigor and accuracy in our carbon stock estimations⁶¹.

Carbon storage and sequestration were estimated using the IPCC Tier 1 methodology, a standardized framework for quantifying greenhouse gas (GHG) emissions and removals⁶². This approach ensures comparability with global carbon assessments and aligns with climate policy reporting requirements.

The total biomass carbon stock (C) was estimated using the equation recommended by the IPCC, as shown in Eq. (5)⁶³:

In this equation, C represents the total carbon stored in biomass (tC), while V_t denotes the aboveground biomass volume (m³/ha) derived from regional land cover classifications. BD refers to the biomass bulk density (t/m³), obtained from regional ecological datasets. BEF stands for the Biomass Expansion Factor, accounting for unmeasured components such as branches and leaves. R is the root-to-shoot ratio used to estimate belowground biomass, and CF represents the carbon fraction of biomass, typically valued at 0.47 according to IPCC defaults.

Using regional forest inventory data and land cover classifications, we estimated stand volume (V_t) for different vegetation types, including broadleaf forests and grasslands. Biomass expansion factors (BEF), root-to-shoot ratios (R), and carbon content coefficients (CF) were adopted from Taiwan’s national greenhouse gas inventory, specifically from Table 6.2.3 in Chapter 6 of the Taiwan Greenhouse Gases Inventory: 1990–2022 Summary⁶⁴. This approach ensured that the carbon stock calculations used region-specific, policy-aligned values.

The financial value of carbon sequestration was estimated based on international carbon market prices, specifically incorporating data from three primary carbon trading mechanisms. These included the European Union Emission Trading System (EU ETS) with a rate of €63.50 per ton of CO₂, the California Cap-and-Trade Market (USA) priced at $39.80 per ton of CO₂, and the Australian Carbon Credit Units (ACCU) valued at $39.20 per ton of CO₂.

The total economic value (V) of carbon sequestration was calculated using Eq. (6):

In this equation, V represents the total economic value of carbon sequestration, C_sequestered refers to the total amount of carbon sequestered (tCO₂ per year) as derived from IPCC and InVEST model estimates, and P_carbon denotes the carbon price per metric ton, adjusted for currency conversion.

Carbon sequestration rates were analyzed separately for forests, grasslands, and urban vegetation to ensure an ecosystem-based assessment, reflecting variation in sequestration potential across land cover types^52,56,65. Carbon sequestration values were aggregated within each land cover class to produce a spatially explicit valuation of carbon storage within the heritage landscape.

We obtained all carbon pricing data from verified carbon trading platforms and government agencies to ensure the accuracy and reliability of monetary valuation. To maintain currency precision, we retrieved exchange rates from the Bank of Taiwan (2025).

We conducted a comparative analysis using carbon sequestration estimates derived from the InVEST model to ensure consistency with IPCC-derived values. To validate the final results, we cross-referenced them with Taiwan’s National Carbon Inventory, confirming methodological consistency and reinforcing the robustness of the economic valuation.

Landscape fragmentation and biodiversity metrics

High-resolution Pleiades satellite imagery (2016, 2019, and 2022) was acquired and preprocessed to analyze landscape structural changes and biodiversity implications within the FGSM and FGSBM. We employed supervised classification techniques using an SVM in ArcGIS Pro (3.0) to classify land cover, ensuring consistency in detecting vegetation cover changes, habitat connectivity, and fragmentation patterns²⁵. Classified images were converted into raster format, allowing for comparative analysis of spatial patterns across study years.

The FRAGSTATS model was used to compute key landscape metrics associated with biodiversity conservation, fragmentation, and spatial organization²⁶. The analysis was conducted at three hierarchical levels (class, patch, and landscape) to assess habitat distribution and ecological integrity²⁷:

At the class level, metrics such as Class Area (CA, ha) and Percentage of Landscape (PLAND, %) were used to quantify the total area and proportion of each land cover type, with particular emphasis on tree cover and green spaces vital for biodiversity. The Largest Patch Index (LPI, %) identified the dominance of the largest contiguous natural habitat, which is essential for species movement and overall ecosystem stability. The Aggregation Index (AI, %) measured the degree of clustering within land cover classes, where lower values indicated increased fragmentation and habitat isolation. Additionally, the Normalized Landscape Shape Index (NLSI) assessed habitat shape complexity, recognizing that irregular and fragmented patches could hinder wildlife dispersal and reduce biodiversity richness.

At the patch level, the Patch Density (PD, patches/ha) metric quantified the number of habitat patches, with higher values reflecting greater fragmentation and the potential loss of contiguous habitat necessary for species survival. The Landscape Shape Index (LSI) evaluated the morphological complexity of habitat patches, where more irregular shapes suggested increased disturbance or land-use pressures impacting biodiversity.

Finally, at the landscape level, the Contagion Index (CONTAG) assessed habitat continuity, with lower values signifying higher landscape fragmentation and diminished connectivity of crucial biodiversity corridors. Diversity metrics such as Shannon’s Diversity Index (SHDI) and Simpson’s Diversity Index (SIDI) measured the heterogeneity and ecological diversity of the landscape, where higher values indicated richer and more varied habitats. Additional indices, including Shannon’s Evenness Index (SHEI), Simpson’s Evenness Index (SIEI), and Patch Richness Density (PRD), were also calculated to evaluate the evenness and richness of landscape composition, providing a comprehensive understanding of habitat structure and spatial diversity.

Landscape metrics were computed for each study year (2016, 2019, and 2022) to identify temporal trends in habitat loss, biodiversity-supporting landscape stability, and land-use modifications^66,67,68,69.

We analyzed landscape composition and fragmentation using FRAGSTATS and summarized key metrics in Table 3.

To further assess habitat fragmentation and connectivity, temporal trends in AI, PD, LSI, and CONTAG were analyzed. A decrease in AI and CONTAG, combined with an increase in PD and LSI, would indicate a shift toward higher fragmentation and reduced habitat connectivity, which can negatively impact biodiversity and species movement across the landscape^46,70,71.

The results from FRAGSTATS landscape metrics were integrated into biodiversity conservation planning to support the sustainable management of green spaces within the FGSM-FGSBM site^29,32. By quantifying changes in habitat structure, fragmentation, and spatial diversity, this study provides empirical evidence for land-use policies to preserve ecological integrity^24,28,29. The findings contribute to decision-making in heritage landscape conservation, ensuring that cultural and ecological values remain balanced over time⁶⁹.

Vegetation health dynamics: NDVI trends (2016–2022)

The NDVI is a key metric for assessing vegetation health and ecological stability in Buddhist temple landscapes (Fig. 6). Table 4 presents NDVI variations across 2016, 2019, and 2022, reflecting vegetation coverage and quality fluctuations.

NDVI of the study area, showing vegetation distribution and density changes over different years.

Full size image

Between 2016 and 2019, NDVI values declined by 12.41%, with a decrease in maximum NDVI from 0.80 to 0.71, and an increase in minimum NDVI from −0.24 to −0.17. From 2019 to 2022, the mean NDVI rebounded by 15.78%, with the minimum NDVI improving to −0.06. The maximum NDVI in 2022 (0.74) remained slightly lower than in 2016.

These trends indicate periods of vegetation decline (2016–2019) followed by partial recovery (2019–2022), with an overall NDVI increase of 3.37% from 2016 to 2022.

Carbon sequestration assessment: multi-method estimation

The carbon sequestration capacity of Buddhist-managed forests was quantified using three methodologies: the InVEST Model, which evaluates ecosystem service-based carbon sequestration across diverse land cover types; the IPCC Methodology, providing standardized estimates of biomass accumulation, root-to-shoot ratios, and carbon stock per hectare; and the i-Tree Canopy Analysis, which assesses tree canopy contributions and sequestration efficiency in urbanized sacred landscapes.

Figure 7 shows a net increase in total carbon stock over the six-year period, as estimated by the InVEST model. Total carbon storage increased from 3788.09 tons in 2016 to 4502.39 tons in 2019 before stabilizing at 4,462.48 tons in 2022. CO₂ equivalent sequestration followed a similar trend, reaching 16,508.76 tons in 2019 before stabilizing at 16,362.43 tons in 2022 (Table 5). These findings suggest natural biomass accumulation and potential land-use effects on carbon retention capacity.

Land Use and Land Cover (LULC) classification for the InVEST carbon module, highlighting tree and grassland areas as key carbon storage contributors in the study area.

Full size image

IPCC-based calculations estimate carbon stock at 67.36 tons per hectare, with a CO₂ equivalent of 4218.35 tons (Table 6). These values integrate biomass expansion factors (BEF = 1.40), root-to-shoot ratios (R = 0.24), and carbon fractions (CF = 0.4691), ensuring consistency with international carbon estimation protocols.

The i-Tree Canopy analysis (Table 7) refines these estimates by quantifying tree-specific carbon sequestration. Findings indicate that trees within the 25.25 ha study area store 1,312.55 ± 63.94 tC, translating to 4,812.67 ± 234.44 tons of CO₂ equivalent sequestration. Additionally, the annual sequestration rate is 52.26 ± 2.55 tC (191.63 ± 9.34 tons CO₂ equivalent), reinforcing the role of Buddhist sacred forests as continuous carbon sinks.

Land cover composition and spatial distribution

The i-Tree Canopy land cover classification provides insights into the distribution of vegetation and its role in carbon sequestration. The analysis identifies tree and shrub cover as the dominant carbon sink, occupying 17.08 ha of the total 25.25 ha study area(Table 8, Fig. 8). Grasslands cover 8.17 ha, providing moderate carbon sequestration potential.

Land use classifications derived from the i-Tree Canopy model, illustrating vegetation and land cover types relevant to ecological assessment in the FGSM and FGSBM study area.

Full size image

In contrast, impervious surfaces (10.75 ha) and bare soil (1.67 ha) offer minimal contributions to carbon retention, as these land types lack the biomass necessary for long-term sequestration. The classification also highlights roads (10.81 ha) and built environments (9.14 ha), further restricting sequestration potential due to their non-vegetative nature.

These findings suggest that expanding tree-dominated areas within temple landscapes could enhance overall sequestration efficiency while minimizing impervious land use and mitigating carbon loss.

Economic valuation of carbon sequestration benefits

To evaluate financial outcomes, we assessed the economic impact of carbon sequestration under three global carbon credit trading mechanisms: the European Energy Exchange (EEX), the California Cap-and-Trade Program (CARB), and the Australian Carbon Credit Units (ACCU).

As presented in Table 9, the financial valuation of stored Carbon demonstrates an upward trend over time. In 2016, the total market value ranged from $13.84 million (ACCU) to $16.22 million (EEX). By 2019, this valuation peaked at $16.88 million (ACCU) and $19.44 million (EEX) before stabilizing in 2022, settling between $16.74 million (ACCU) and $19.26 million (EEX).

Additionally, the i-Tree Canopy economic valuation (Table 10) estimates that the tree covers carbon sequestration at $408,853 (EEX), $234,084 (CARB), and $229,672 (ACCU). These figures highlight the monetary significance of faith-managed forests, demonstrating the potential for integration into Carbon offset markets.

To ensure cross-market comparability, we standardized all financial values using 2025 currency exchange rates (1 EUR = 1.046 USD; 1 AUD = 0.6397 USD). This approach enables a consistent assessment of faith-managed sequestration potential within international carbon pricing frameworks.

The multi-method assessment of carbon sequestration confirms that faith-managed forests exhibit a stable and increasing sequestration trajectory, with notable gains from 2016 to 2019, followed by a slight plateau in 2022. The land cover analysis underscores the importance of tree canopy expansion, while the financial valuation highlights the potential integration of Buddhist conservation forests into Carbon offset markets. These results reinforce the significance of sacred landscapes in global climate change mitigation strategies.

Landscape fragmentation, biodiversity, and habitat connectivity

We next analyzed land cover, biodiversity, and habitat connectivity using FRAGSTATS. The model comprehensively assesses land cover changes, biodiversity contributions, and habitat connectivity within the FGSM and FGSBM landscapes from 2016 to 2022. This section presents land cover classifications, evaluates the role of temple forests in supporting biodiversity, and examines landscape fragmentation trends and their impact on ecosystem stability.

The Land Use and Land Cover (LULC) classification (Fig. 9) identifies seven primary land cover types within the study area, each influencing carbon storage and biodiversity conservation, as detailed in Table 11. Temple structures cover 82.83 hectares, accounting for 60.17% of the total landscape, and form the dominant spatial feature. Tree cover occupies 17.54 ± 2.24 hectares (12.75% PLAND), acting as the primary carbon sink and habitat zone. Herbaceous land spans 7.16 hectares (5.20% PLAND), contributing to species diversity and stabilization of ground cover. Road networks cover 19.80 hectares (14.39% PLAND), increasing landscape fragmentation and limiting habitat connectivity. Built environments, excluding temple structures, occupy 4.11 hectares (2.99% PLAND), reducing the availability of natural spaces. Lastly, barren land and paths cover 6.21 hectares (4.51% PLAND), offering minimal ecological contributions.

LULC classification of the study area, serving as the basis for FRAGSTATS analysis to assess landscape patterns and spatial characteristics.

Full size image

Tree cover is characterized by an LPI of 3.18, suggesting that tree patches are spatially dominant but not fully connected. The LSI 24.49 reflects moderate complexity in tree patch configurations, which may facilitate diverse edge habitats. In contrast, expanding impervious surfaces, particularly roads and buildings, contribute to higher habitat fragmentation, potentially reducing ecological connectivity.

Temple forests are critical biodiversity reserves, sustaining native flora and fauna through stable habitat formations. The high AI of 88.73 ± 4.49 for tree cover confirms that tree patches remain clustered, ensuring spatial continuity essential for species survival. However, the gradual decline in AI from 93 (2016) to 87 (2022) indicates a weakening of habitat connectivity, which, if continued, may disrupt species migration and genetic diversity (Fig. 10).

Each sub-panel represents a distinct landscape metric, with different colors indicating various land use types. The Y-axis is scaled independently for each metric to facilitate comparison, and trend lines are included to highlight changes in landscape patterns. The legend at the bottom defines the land use classifications.

Full size image

Tree cover maintains PLAND at 12.75%, reinforcing its ecological significance despite increasing land-use pressures. The LSI for trees has risen from 20 in 2016 to 24.49 in 2022, reflecting growing complexity in patch structure, which may support more excellent microhabitat formation and species adaptability.

Herbaceous land, with an LSI increase from 20 to 40 over the study period, complements tree-dominated areas by providing ground cover, supporting pollinators, and stabilizing the soil structure. Despite occupying a smaller fraction of the landscape (5.20% PLAND), it contributes to enhanced biodiversity richness and functional ecosystem services. Temple structures, while non-vegetative, shape landscape organization and habitat stability. The LPI of 56.60 ± 1.14 confirms their dominance, and their role as stable landscape anchors supports species persistence by reducing landscape disruption from built area expansion.

Landscape fragmentation is a growing concern within the study area, as indicated by key FRAGSTATS metrics (Fig. 11). The Patch Density (PD) increased dramatically from 1,000 in 2019 to 6,000 in 2022, reflecting a substantial rise in fragmentation levels. In contrast, the Contagion Index (CONTAG) decreased from 62 in 2019 to 55 in 2022, suggesting reduced habitat connectivity and greater isolation of ecological patches. The Shannon’s Diversity Index (SHDI) rose from 1.225 in 2019 to 1.325 in 2022, indicating enhanced habitat heterogeneity and a more even distribution of land cover types. Similarly, the Simpson’s Diversity Index (SIDI) increased from 0.585 to 0.610, highlighting growing species diversity but also suggesting potential ecological instability due to shifting habitat distributions. Additional metrics, including the Shannon’s Evenness Index (SHEI) and Simpson’s Evenness Index (SIEI), exhibited slight increases over time, pointing to a more balanced distribution of patch types, while the Patch Richness Density (PRD) remained relatively stable. Notably, the Aggregation Index (AI) for tree cover declined from 93 in 2016 to 87 in 2022, signaling reduced tree patch connectivity and increased spatial isolation.

The top panel presents key metrics (PD, LSI, CONTAG, PRD, SHDI) that highlight significant landscape structure changes, while the bottom panel displays supporting metrics (SIDI, SHEI, SIEI, AI) to provide additional insights into spatial heterogeneity. X-axis labels are included only in the bottom panel to improve clarity and readability. The Y-axis scales vary by metric to optimize trend visualization, reflecting structural landscape shifts influenced by ecological and anthropogenic factors.

Full size image

Air purification and ecosystem service benefits of temple forests

The Buddha Museum’s forested area plays a significant role in urban air quality improvement and climate regulation, with measurable impacts on pollutant removal and temperature moderation. This section presents the quantification of air purification benefits, the climate-buffering function of temple groves, and the long-term resilience offered by faith-driven forest preservation.

The temple forest (17.08 ha) effectively captures and removes key atmospheric pollutants, including carbon monoxide (CO), nitrogen dioxide (NO₂), ozone (O₃), sulfur dioxide (SO₂), fine particulate matter (PM_2.5), and coarse particulate matter (PM₁₀). We estimated pollutant removal using the i-Tree Canopy model, which applies pollutant-specific deposition rates to land cover categories to calculate total annual removal and associated monetary values. Over the study period, the total pollutant removal exceeded 1477 kg annually, generating an estimated economic value of $17,752 per year (Table 12).

Among the removed pollutants, PM_2.5 and PM₁₀ account for the highest economic benefits, emphasizing the role of temple forests in reducing urban air pollution and associated respiratory health risks. Notably, ozone removal exceeded 927 kg per year, reinforcing the importance of temple forests in mitigating smog formation and maintaining atmospheric balance. Removing gaseous pollutants (CO, NO₂, and SO₂) suggests that temple forests also reduce vehicle and industrial emissions.

Figure 12 illustrates the amount of pollutants removed (in kg) by trees in the study area, ranked from highest to lowest efficiency. Ozone (O₃) exhibits the highest removal rate, followed by particulate matter (PM₁₀) and nitrogen dioxide (NO₂). In comparison, sulfur dioxide (SO₂), fine particulate matter (PM_2.5), and carbon monoxide (CO) show comparatively lower removal rates. These findings highlight the essential role of urban trees in air purification and environmental health.

Air purification benefits of trees analyzed using the i-Tree Canopy model, illustrating their role in removing pollutants and improving air quality within the study area.

Full size image

The analysis of NDVI values from 2016 to 2022 reveals significant fluctuations in vegetation health, reflecting the interplay between environmental stressors and conservation efforts. The decline in mean NDVI by 12.41% between 2016 and 2019, coupled with a decrease in maximum NDVI from 0.80 to 0.71, suggests a reduction in vegetation density and canopy coverage (Table 4). Although minimum NDVI improved slightly from −0.24 to −0.17, the overall trend indicates ecological stress, likely attributable to increased impervious development within the heritage site, land-use modifications, and the natural aging of tree stands.

Conversely, the subsequent period from 2019 to 2022 exhibited a notable NDVI recovery of 15.78%, indicative of successful regrowth efforts. Minimum NDVI improved to −0.06, signifying the revitalization of previously degraded patches. However, the maximum NDVI in 2022 remained slightly below its 2016 level (0.74 vs. 0.80), indicating that while afforestation initiatives have contributed to ecosystem restoration, the expansion of FGSBM and FGSM under new building construction and land use changes has led to lower NDVI values, suggesting that the ecosystem has not yet fully recovered (Fig. 6). These findings parallel broader trends observed in faith-managed conservation areas, where periodic vegetation stress is mitigated through localized stewardship and long-term conservation commitments³⁴.

The observed fluctuations in NDVI correspond closely with the conservation practices undertaken by Buddhist monasteries. These findings underscore the function of religious institutions as ecological sanctuaries, wherein faith-based environmental stewardship actively counteracts urbanization pressures and fosters ecosystem resilience^11,43. The post-2019 NDVI rebound is likely the result of structured afforestation programs, sustainable forestry management, and conservation awareness initiatives spearheaded by monasteries.

Other faith-managed conservation systems have reported similar patterns, such as:

1. (1)

   In China, indigenous ecological beliefs and sacred conservation practices sustain higher NDVI values in Fengshui forests compared to adjacent state-managed areas⁷².

2. (2)

   Due to long-term religious protection, Ethiopian church forests exhibit enhanced vegetation stability and biodiversity compared to surrounding landscapes³⁴.

The significant NDVI recovery (2019–2022) suggests that monastic afforestation initiatives have successfully counterbalanced prior vegetation losses. Reforestation efforts, including selective planting of native species and soil rehabilitation programs, have likely contributed to this recovery. These findings align with studies on temple forests in Thailand, where minimal human intervention and sacred status ensure sustained ecological restoration⁷. In a comparative study of faith-managed reforestation efforts, Ouyang et al. ⁵⁸ found that Japanese temple forests exhibit long-term NDVI stability, a phenomenon attributed to low-impact forest management and religious conservation ethics⁵⁸.

These comparisons highlight the importance of faith-based conservation models in sustaining long-term ecological integrity, particularly in regions experiencing rapid land-use transformations.

The decline in NDVI between 2016 and 2019 aligns with well-documented trends in land-use transformations. The expansion of impervious surfaces, infrastructure development, and rising anthropogenic pressures near Buddhist temple landscapes have directly contributed to vegetation loss. Furthermore, as shown in Fig. 7, the land use and land cover of FGSBM in 2022 reveal a notable increase in barren areas (brown color), indicating continued landscape alterations that further suppress vegetation recovery.

Research indicates that vegetation loss in urbanizing regions often results from progressive land-use changes, a trend corroborated by Nowak et al. ⁶², who reported a 10–15% NDVI reduction within five years in urban forestry zones due to infrastructure expansion and foot traffic.

The observed increase in minimum NDVI values suggests improving the resilience of degraded landscapes. This trend is consistent with findings by Huo and Wang, who demonstrated that faith-managed forests mitigate microclimate fluctuations and maintain long-term ecosystem stability³⁹. Since monastic landscapes are relatively insulated from large-scale disturbances, they serve as strongholds of ecological resilience, buffering against climate variability and external environmental stressors.

A comparative evaluation of Buddhist temple forests and other faith-managed conservation areas further reinforces the efficacy of sacred landscapes in biodiversity preservation and ecosystem stability in Table 13^2,6,13. Studies on Ethiopian church forests and indigenous sacred groves demonstrate that religiously protected landscapes tend to sustain higher NDVI values over time than state-managed reserves⁹.

These comparisons indicate that faith-based conservation models outperform state parks in maintaining vegetation stability, mainly due to religious protection from deforestation, lower infrastructure development, and community-driven stewardship.

Buddhist temple forests function as significant long-term carbon sinks, with sequestration trends from 2016 to 2022 demonstrating sustained carbon accumulation. The InVEST model estimated an increase in total carbon stock from 3788.09 tons in 2016 to 4502.39 tons in 2019, followed by stabilization at 4462.48 tons in 2022. CO₂ sequestration mirrored this pattern, peaking at 16,508.76 tons in 2019 before leveling at 16,362.43 tons in 2022. This trend suggests a shift in biomass accumulation dynamics, similar to those observed in other sacred conservation areas.

A comparative approach using the IPCC methodology estimated 67.36 tC per hectare, corresponding to a CO₂ equivalent of 4218.35 tons. This estimation, derived from standardized biomass expansion factors and root-to-shoot ratios, ensures compatibility with international carbon stock estimation protocols⁷³. Additionally, the i-Tree Canopy assessment, which provides a fine-scale analysis of sequestration potential, estimated that the 25.25 ha study area contained 1312.55 ± 63.94 tC, translating to a CO₂ equivalent of 4812.67 ± 234.44 tons. The annual sequestration rate, recorded at 52.26 ± 2.55 tC, reinforces the role of tree-dominated landscapes in maintaining continuous carbon absorption.

These multi-method assessments underscore Buddhist temple forests’ robust carbon sequestration potential, validating their role in nature-based climate mitigation strategies¹⁹.

The observed stabilization of carbon sequestration post-2019 suggests that the system is approaching equilibrium due to biophysical and climatic factors.

One potential driver is aboveground biomass saturation, as the InVEST model suggests^20,22. Miyawaki forests exhibit similar patterns, where rapid biomass accumulation slows as carbon storage capacity reaches its limit⁷⁴. Without additional afforestation or canopy expansion, sequestration rates naturally plateau.

Additionally, soil carbon fluxes and microbial decomposition influence belowground carbon retention⁷⁵. While the IPCC approach accounts for biomass-derived sequestration, it does not fully capture soil-carbon interactions. Research indicates that microbial decomposition and soil respiration can partially offset carbon storage gains, particularly in humid environments where organic matter decay accelerates⁷⁶.

Climate variability further impacts sequestration potential. Findings from the i-Tree Canopy analysis emphasize that tree canopy density regulates sequestration efficiency¹⁹. Disruptions in monsoon cycles and increased seasonal droughts may have constrained tree growth rates, reducing the system’s ability to sustain additional sequestration gains^22,77. These findings emphasize the necessity of adaptive conservation frameworks considering biomass management and soil carbon stabilization⁷⁶.

Buddhist temple forests are essential biodiversity refuges, sustaining native flora and fauna in peri-urban or semi-rural religious landscapes. However, habitat fragmentation poses increasing challenges to their ecological stability³². The FRAGSTATS model analysis highlights a substantial increase in PD from 1000 (2019) to 6000 (2022), reflecting rising landscape fragmentation⁶⁸. Simultaneously, the AI declined from 93 (2016) to 87 (2022), indicating a reduction in habitat connectivity, which may disrupt species migration, genetic diversity, and ecosystem resilience.

Tree cover, which accounts for 12.75% of the study area (PLAND), remains a critical ecological component despite built area expansion pressures²⁶. The LSI for trees increased from 20 in 2016 to 24.49 in 2022, signifying a transition toward irregular patch configurations⁶⁸. While this can support edge-adapted species, it reduces interior habitat stability, negatively affecting forest-dependent taxa⁷⁸. The decline in the CONTAG from 62 (2019) to 55 (2022) further reinforces habitat isolation, necessitating targeted conservation interventions^31,32.

Buddhist temple forests retain more biodiversity than state-managed reserves due to religious land protection and long-term ecological stability. However, analyzing other sacred and secular conservation landscapes reveals parallels and key vulnerabilities.

• Ethiopian Church Forests, similarly embedded within human-dominated landscapes, exhibit increasing fragmentation, with patch density rising from 5 patches/km² in 1990 to 12 patches/km² in 2020 and a concurrent decline in the Aggregation Index from 0.85 to 0.60⁹. These patterns closely mirror those in Buddhist temple forests, emphasizing the need for connectivity restoration efforts to prevent further habitat isolation.

• African Sacred Groves face similar fragmentation pressures, with studies reporting a 20% decline in forest cover within a 1 km radius over a decade, leading to biodiversity loss and ecosystem degradation⁷⁹. Fragmentation of temple forests reduces functional diversity and threatens long-term species stability without effective conservation strategies^9,80.

• European Monastic Forests, historically semi-managed conservation reserves, have experienced a tree density reduction from 300 trees/ha to 200 trees/ha, contributing to significant declines in avifaunal populations within fragmented patches. This trend reinforces findings from Buddhist temple forests, where fragmentation has altered habitat availability for specific taxa^80,81,82.

• Indigenous Sacred Groves in India, known for strong belowground carbon sequestration potential, have recorded a 30% decline in pollinator visitation rates in fragmented groves, impacting fruit set and seed production⁸³. A similar pattern may emerge in Buddhist temple forests, where habitat isolation could disrupt pollination networks and associated trophic interactions.

While faith-managed forests demonstrate superior biodiversity conservation compared to state-managed reserves, rising fragmentation trends indicate an urgent need for conservation strategies to ensure their continued ecological functionality.

Buddhist temple forests are pivotal to urban air quality management, functioning as highly efficient natural air filters. In our study, the Buddha Museum’s temple forest—spanning 17.08 ha—removes approximately 1477 kg of airborne pollutants annually, with an estimated economic benefit of $17,752 annually. Notably, ozone (O₃) removal is particularly effective, averaging 927.67 kg per year, substantially contributing to reducing ground-level smog and mitigating associated respiratory health risks. Equally important is the absorption of particulate matter, with PM₁₀ and PM_2.5 being removed at rates of 323 kg and 45 kg per year, respectively—critical processes given the established links between these particulates and cardiovascular and respiratory diseases. Furthermore, removing nitrogen dioxide (NO₂) and sulfur dioxide (SO₂) underscores the capacity of temple forests to lower emissions from traffic and industry, thereby mitigating acid rain and related environmental hazards. When normalized per hectare, the highest removal efficiencies are observed for ozone (54.31 kg/ha) and PM₁₀ (18.92 kg/ha), reinforcing the ecological significance of these sacred green spaces as nature-based solutions for urban air pollution¹⁹.

The air purification function of Buddhist temple forests is not an isolated phenomenon; instead, it is consistent with the performance of other religiously managed landscapes worldwide. For example, Fengshui forests in China have demonstrated comparable pollutant sequestration capabilities, yet they are seldom incorporated into formal air quality policies⁵⁸. Similarly, studies show that urban temple forests in Bangkok significantly reduce particulate concentrations in high-canopy areas⁸⁴. Ethiopian church forests, characterized by dense and relatively undisturbed vegetation, provide robust pollutant absorption and carbon sequestration, further validating the ecological merits of faith-managed green spaces⁹. Although subject to varying degrees of urban encroachment, European monastic forests also retain significant ecosystem service functions; however, their contributions to air quality are often underrepresented in policy frameworks^3,61,63. These comparative insights highlight that faith-managed landscapes, by their long-term protection from deforestation, often exhibit superior air quality benefits compared to conventional municipal parks in urban cores, yet they remain underrecognized in environmental policy⁷².

Built area expansion, particularly the increase in impervious surfaces, has exacerbated habitat isolation⁴⁶. Road networks, which now cover 19.80 ha (14.39%), present significant barriers to species dispersal, increasing the risk of genetic bottlenecks^27,30,85. Despite doubling the herbaceous land LSI from 20 to 40, which supports pollinator populations and soil stabilization, the spatial disconnect between herbaceous zones and core forest patches reduces their effectiveness as functional ecological buffers⁸⁶.

The contradiction between rising biodiversity indices (SHDI, SIDI) and increasing fragmentation (PD increase, AI decline) presents a complex conservation scenario. The increase in evenness indices (SHEI, SIEI) suggests enhanced spatial balance, whereas stable PRD implies compositional consistency across years. While more significant landscape heterogeneity can support generalist species, continued fragmentation may disrupt interior-dependent taxa, leading to long-term biodiversity instability^27,32,87. If habitat connectivity continues to decline, Buddhist temple forests may transition toward fragmented peri-cultural ecological structures, reducing their overall capacity to function as compelling conservation landscapes^24,26,85.

A comparative evaluation reveals that Buddhist temple forests exhibit carbon sequestration characteristics consistent with other faith-managed conservation areas.

Ethiopian church forests, which experience low deforestation rates due to community-led conservation efforts, display sequestration trends similar to IPCC-derived biomass estimates⁹. Indigenous sacred groves in India, known for their deep-rooted vegetation, exhibit higher belowground carbon retention, aligning with i-Tree Canopy findings that emphasize the role of root systems in sequestration dynamics¹³.

A broader comparison underscores that faith-managed landscapes outperform urban parks in sequestration efficiency. While urban parks store an average of 15.3 tC/ha, religious forests surpass 67 tC/ha⁴. This significant disparity highlights the potential of sacred landscapes as high-density carbon sinks and their importance in climate resilience strategies.

In addition to their role in air purification, Buddhist temple forests serve as effective climate buffers in low-density developed cultural landscapes. The removal of PM_2.5 and PM₁₀ not only improves air quality but also reduces atmospheric heat absorption, thereby mitigating radiative forcing and attenuating urban heat island effects⁶⁰. Similarly, ozone and nitrogen dioxide absorption limits smog formation and reduces temperature variability caused by chemical pollutants. Moreover, temple forests’ inherent dense biomass and mature canopy structures enhance thermal regulation, which can decrease the frequency and intensity of extreme heat events—a benefit corroborated by studies on urban forestry in regions such as Canada⁸⁸. Such multifaceted climate mitigation functions underscore the role of temple forests as nature-based solutions that provide sustainable alternatives to energy-intensive cooling systems.

Buddhist temple forests play a vital role in biodiversity conservation and climate mitigation, yet they face vegetation health fluctuations due to development pressures. The NDVI decline observed in FGSM and FGSBM from 2016 to 2019, followed by partial recovery (2019–2022), reflects trends seen in Ethiopian church forests, where afforestation efforts counteract habitat fragmentation. Targeted afforestation programs using native high-carbon sequestration species should be prioritized to enhance long-term resilience, as demonstrated in Chinese Fengshui forests⁸⁹.

Beyond afforestation, carbon sequestration stagnation is a growing concern in old-growth temple forests, where biomass saturation limits further carbon absorption. Similar patterns in Chinese temple forests and Indigenous sacred groves suggest that sequestration equilibrium is inevitable without active management.

Strategies such as:

1. (1)

   Soil carbon stabilization techniques (biochar application, microbial soil enhancement)^3,65.

2. (2)

   Implementing selective thinning and light management enhances biomass renewal^42,56,90.

Despite their high sequestration potential, Buddhist temple forests remain excluded from carbon markets, missing a significant financial opportunity. The carbon stock valuation of FGSBM and FGSM increased from $16.74 M (2016) to $19.26 M (2022), yet a lack of certification prevents market participation. Studies on Nepalese and Thai religious forests indicate that including faith-managed forests in voluntary carbon markets can generate conservation funding^7,91. Aligning temple forests with Taiwan’s 2024 carbon pricing policy could enable them to participate in global offset schemes, ensuring financial sustainability⁹².

Ecotourism presents an alternative revenue model for conservation. Chinese sacred groves have successfully integrated faith-based conservation tourism, where visitors support heritage protection and biodiversity conservation^37,38,93. Buddhist temple forests could adopt similar models by integrating guided ecological tours, sustainable visitor facilities, and community engagement⁹⁴. However, careful zoning regulations and impact assessments are needed to prevent over-commercialization, as seen in overdeveloped sacred landscapes in China^38,41,43,95.

A significant challenge in conserving Buddhist temple forests is their lack of formal policy recognition, restricting legal protection and funding access. Faith-managed forests in Ethiopia, China, and India face similar exclusions despite demonstrated ecological benefits. We recommend the following policy measures to address this issue :

1. (1)

   Legal Protection as Conservation Reserves – Temple forests should be designated as protected ecological reserves, ensuring long-term financial and legal safeguards^23,34,96.

2. (2)

   Institutional Collaboration – Partnerships between monasteries, government agencies, and conservation organizations should establish co-managed forest governance, as seen in Chinese Fengshui forests⁸⁹.

3. (3)

   Integration into National Green Infrastructure – Temple forests should be incorporated into urban resilience planning, mirroring China’s recognition of faith-managed landscapes as climate buffers¹⁹.

4. (4)

   Community-Led Conservation – Local communities should actively engage in forest management, following successful Indigenous sacred forest models^34,97.

Further, fostering collaborative research partnerships between monastic institutions and environmental scientists can significantly enhance conservation efforts^94,98. Integrating GIS-based biodiversity monitoring allows for precisely tracking of vegetation changes and habitat health. At the same time, remote sensing analysis provides large-scale insights into forest cover dynamics and environmental stressors. Additionally, data-driven afforestation strategies can optimize reforestation by selecting suitable native species and ensuring long-term ecosystem stability^19,97. By merging scientific methodologies with faith-based conservation practices, Buddhist temple forests can be more effectively protected and sustainably managed as resilient ecological sanctuaries in rapidly urbanizing landscapes^99,100. This study comprehensively assesses the carbon sequestration capacity, biodiversity conservation potential, and environmental benefits of the FGSM and FGSBM temple forest, highlighting the intersection of cultural heritage management and climate resilience. Through quantitative geospatial analysis and ecological modeling, the findings confirm that faith-managed landscapes serve as critical carbon sinks while enhancing biodiversity stability and urban sustainability^38,43,50.

Between 2016 and 2022, carbon storage increased from 3788.09 to 4462.48 tons, with an annual sequestration rate of 52.26 ± 2.55 tC (191.63 ± 9.34 tons CO₂ equivalent). However, a slight decline of 0.89% in sequestration between 2019 and 2022 suggests that biomass saturation, soil carbon flux limitations, and land-use modifications require strategic intervention. The i-Tree Canopy and IPCC models corroborate these estimates, reinforcing the temple forest’s function as a long-term carbon reservoir with an average carbon stock of 67.36 tons per hectare.

Beyond carbon sequestration, the FRAGSTATS landscape analysis revealed that temple forests contribute to habitat connectivity, with an AI of 88.73%, facilitating species migration and ecological continuity. However, the declining CONTAG suggests increasing habitat fragmentation, which may disrupt ecological integrity. In contrast, SHDI and SIDI improvements indicate species diversification and enhanced ecological resilience, further emphasizing the ecological significance of Buddhist temple forests as biodiversity sanctuaries^43,71.

From an economic perspective, the conservation initiatives at FGSBM and FGSM align with Taiwan’s 2024 Carbon Fee Policy, with carbon market valuations increasing from $16.74 million (2016) to $19.26 million (2022). These findings highlight the potential for Buddhist temple forests to be formally integrated into global carbon trading markets, creating sustainable revenue streams for conservation efforts^21,33. Comparative analysis of Ethiopian church forests and Indian sacred groves underscores FGSBM and FGSM’s proactive conservation model, characterized by afforestation, sustainable landscape management, and carbon credit potential⁹².

Despite these ecological and economic benefits, temple forests remain underrepresented in formal environmental and urban planning policies. Their exclusion from urban air quality management, climate mitigation policies, and carbon finance mechanisms limits their full impact. Addressing these gaps requires strategic conservation planning, policy integration, and scientific monitoring. Future efforts should focus on:

1. (1)

   Expanding afforestation and reforestation programs to enhance carbon sequestration and mitigate fragmentation^22,92.

2. (2)

   Implementing soil carbon stabilization mechanisms prevents sequestration stagnation and maximizes long-term carbon storage potential^19,20.

3. (3)

   Strengthening conservation finance models, integrating temple forests into carbon markets and green infrastructure frameworks^21,96.

This study underscores the urgent need for policy frameworks incorporating religious landscapes into climate mitigation strategies. As cultural heritage sites increasingly contribute to biodiversity resilience and climate adaptation, targeted conservation strategies—including afforestation, reforestation, and enhanced ecosystem monitoring—are essential to ensure ecological and economic sustainability. By aligning faith-based conservation with global climate policy, Buddhist temple forests can continue functioning as vital ecological assets, reinforcing their role in urban sustainability, climate change mitigation, and long-term biodiversity conservation.

While this study enhances the understanding of religious landscapes as carbon sinks and biodiversity refuges, several limitations must be acknowledged. These constraints provide a foundation for future research to refine methodologies, expand comparative frameworks, and strengthen policy integration.

One of the primary limitations lies in data resolution and temporal scope. Although high-resolution satellite imagery and geospatial models offer valuable insights, the absence of field-based validation techniques such as tree-core sampling, LiDAR mapping, and soil carbon analysis limits the precision of sequestration estimates. The six-year study period (2016–2022) also provides useful short-term trends. However, longitudinal analyses spanning multiple decades would yield more substantial insights into the long-term effects of climate variability, biomass saturation, and ecosystem adaptation.

A further challenge stems from uncertainties in carbon estimation models. While i-Tree Canopy, InVEST, and IPCC models provide reliable sequestration estimates, regional variability in biomass expansion factors may introduce discrepancies. Developing a calibrated, site-specific biomass model tailored to Buddhist temple forests would improve estimation accuracy, particularly when considering species composition, soil carbon flux, and age-dependent sequestration variations.

Additionally, biodiversity monitoring in this study relies primarily on landscape-level FRAGSTATS metrics, which capture habitat connectivity and fragmentation trends but lack direct species assessments. Future research should integrate faunal surveys, camera trapping, and soil microbial diversity analysis to corroborate species richness patterns and assess ecological resilience in Buddhist temple forests.

From an economic and policy perspective, while this study monetizes carbon sequestration values, the actual financial viability of temple forests in carbon markets depends on policy implementation, legal recognition, and market accessibility. Future research should explore government incentives, regulatory frameworks, and voluntary carbon credit mechanisms to facilitate the inclusion of religious conservation landscapes in global climate finance.

Finally, while this study compares Buddhist temple forests with Ethiopian church forests and Indian sacred groves, a broader comparative framework is needed to assess regional conservation strategies and policy adaptations. Expanding research to Buddhist temple forests in Japan, Thailand, and China would provide a more comprehensive understanding of faith-based conservation models across different cultural and ecological contexts.

Addressing these limitations through expanded temporal analyses, direct biodiversity monitoring, improved carbon modeling, and global comparative studies will further strengthen the role of religious landscapes in climate mitigation and ecological conservation.