Mapping and assessing ecosystem service hotspots and bundles in rapidly urbanizing, data-scarce regions: a case study of Kabul, Afghanistan

Mapping and assessing ecosystem service hotspots and bundles in rapidly urbanizing, data-scarce regions: a case study of Kabul, Afghanistan | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Mapping and assessing ecosystem service hotspots and bundles in rapidly urbanizing, data-scarce regions: a case study of Kabul, Afghanistan Mohammad Reza Ansari, Maisam Rafiee, Blal Adem Esmail, Daniela Kempa, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7130094/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Context Urban expansion in the Global South challenges ecological sustainability. Understanding ecosystem service (ES) distribution and bundling in these rapidly changing landscapes is vital for informed urban planning and resilience. However, there is a research gap in applying integrated, high-resolution ES assessments to data-scarce, informally developing urban contexts. Objectives This study investigates ES patterns and trade-offs in the Kabul metropolitan region (KMR) in Afghanistan, a fast-growing, data-scarce city. Objectives are to: (1) map spatial patterns, including hotspots and coldspots, of five key ESs (water yield, heat mitigation, stormwater runoff retention, nature access, habitat quality), and (2) identify distinct ES bundles and their associated synergies and trade-offs. Methods The study focuses on KMR, as a representative case of a rapidly urbanizing, data-scarce region facing significant environmental pressures. Using multi-source data, a high-resolution land-use/land-cover map is created for ES modeling (InVEST), hotspot analysis (Getis-Ord Gi*), ES bundle identification (k-means clustering), and conservation prioritization. Results The assessment reveals significant spatial heterogeneity across five key ESs. ES hotspots corresponded to green/blue infrastructure, while coldspots are prevalent in dense urban areas. Distinct ES bundles are mapped, revealing context-dependent synergies (e.g., heat mitigation, stormwater runoff retention) and trade-offs (e.g., water yield, stormwater runoff retention). Finally, conservation priority areas are mapped based on multi-service supply. Conclusions This study provides spatially explicit information for KMR's sustainable urban planning, using a procedure transferable to other cities in the Global South. Findings enable targeted interventions to enhance urban resilience by identifying priority zones for ES protection. Future research should prioritize model validation with local data. Data scarcity Global South Spatial analysis Sustainable spatial planning Urban ecosystem services Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Introduction The mapping and assessment of ecosystems and their services (MAES) is increasingly recognized as fundamental to sustainable spatial planning and environmental governance, providing critical tools to harmonize urban development with ecological preservation (e.g., Albert et al. 2014, 2016; Geneletti 2016; Maes et al. 2012). Ecosystem services (ESs)—the manifold benefits humans derive from healthy functioning ecosystems, such as water supply, climate regulation, flood mitigation, and recreational opportunities (Costanza et al. 1997; Daily 2010; Millennium Ecosystem Assessment 2005)—are pivotal for urban resilience (e.g., McPhearson et al. 2015) and for enhancing the quality of life in increasingly urbanized landscapes (e.g., Bolund and Hunhammar 1999; Kowarik et al. 2017). While ecosystem service (ES) research has expanded significantly, robust spatially explicit ES assessments remain crucial for translating conceptual understanding into actionable planning strategies. A key challenge in ES assessment is understanding their spatial heterogeneity—where services are abundant, where they are scarce, and how they co-vary. Methodologies such as hotspot and coldspot analysis (e.g., Getis-Ord Gi*) have become instrumental in identifying statistically significant clusters of high and low ES provision, thereby guiding targeted interventions (Adem Esmail et al. 2025; Bagstad et al. 2017). Furthermore, recognizing that ESs rarely occur in isolation, the concept of ES bundles—recurrent sets of associated ESs—has emerged as a powerful analytical framework for understanding landscape multifunctionality and identifying synergies and trade-offs (Queiroz et al. 2015; Raudsepp-Hearne et al. 2010). Such analyses are vital for developing integrated conservation and restoration plans that optimize multiple benefits and address complex environmental challenges (Spake et al. 2017; Wu et al. 2022). Despite these methodological advancements, ES research has predominantly remained concentrated in the Global North, leaving the Global South underexplored (Gangahagedara et al. 2021; Seppelt et al. 2011; Vihervaara et al. 2010). These areas often face the dual challenge of intense ecological pressure from urban expansion and severe data scarcity, hindering the effective integration of ES information into planning processes (e.g., Mahendra et al. 2021; e.g., Shackleton et al. 2021). The Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) regional assessment report on biodiversity and ESs for Asia and the Pacific (2018) underscores this imbalance, identifying urbanization, climate change, and habitat degradation as primary threats, yet noting a deficit in localized, high-resolution ES assessments that can effectively inform policy in these dynamic contexts. While studies in the Global South have begun to address these issues (Adem Esmail et al. 2023, 2025; Akhtar et al. 2022; Najmuddin et al. 2022; Srivathsa et al. 2023), many rely on coarser-resolution data that may not capture the fine-scale landscape heterogeneity critical for urban planning, especially in areas with significant informal development. The Kabul metropolitan region (KMR), Afghanistan, exemplifies these challenges. Characterized by swift, largely informal urban expansion over the past two decades—its population doubling with approximately 70% of this growth occurring outside formal planning (Chaturvedi et al. 2020; NSIA 2004, 2023; UN Habitat 2015)—KMR faces escalating environmental hazards such as heatwaves, flooding, and resource scarcity. This context of rapid, under-documented landscape transformation, coupled with profound data limitations (NEPA 2014), makes KMR a critical yet understudied area for applied ES research. The dominance of informal settlements presents a distinctive opportunity to examine ES provision in such settings, addressing a significant gap in regional and global research and offering insights applicable to comparable urbanizing areas. The aim of this study is to conduct a high-resolution, spatially explicit assessment of key ESs in KMR to understand their distribution, interactions, and implications for urban and conservation planning. Specifically, this research undertakes a detailed analysis of five key ESs—water yield (WY), heat mitigation (HM), stormwater runoff retention (SWRR), nature access (NA), and habitat quality (HQ)—employing a multi-source data approach and established modeling techniques to overcome common data limitations. The detailed assessment aims to uncover variations often missed by coarser assessments, thereby facilitating the identification of priority areas for conserving important multi-ES supply zones, and informing strategic green infrastructure development to enhance urban resilience and livability. This research addresses three core questions: How are the selected ESs spatially distributed in the KMR, where are their spatial hotspots and coldspots located, and how does ES provision relate to human population density gradients? Where across the study area do distinct ES bundles (i.e., areas with similar characteristics of ES supply) occur, and what patterns of synergies and trade-offs do these spatially defined bundles reveal? Which zones provide multiple ES benefits requiring protection? By tackling these questions, this study delivers actionable insights for KMR’s planners and policymakers to promote sustainable urban development. It further contributes to the broader ES literature by demonstrating the application of integrated assessment methodologies in a data-scarce, rapidly urbanizing, and informally developing context, offering a framework that can inform planning in similar cities across the Global South and supporting the IPBES call for enhanced biodiversity and ES management strategies. Study area The study area is located within the Kabul-Indus River Basin Zone in eastern Afghanistan, focusing on an approximately 2,584 km² region defined by latitudes 34°16′N to 34°40′N and longitudes 68°52′E to 69°31′E (Fig. 1a, b). This strategically demarcated area encompasses Kabul City, the capital and most populous urban center of Afghanistan, alongside its adjacent peri-urban and surrounding Woloswali (provincial districts). These include Bagrami, Musayi, and Chahar Asyab in their entirety, as well as portions of Nirkh, Maidan Shahr, Paghman, Surkhi Parsa, Shakardara, Dih Sabz, Kohi Safi, Surobi, Khaki Jabbar, and Mohammad Agha. The KMR has been purposefully selected as the focal point of this research due to its pivotal role as the primary hub of Afghanistan's accelerating urban transition. Currently, Kabul City accommodates 54% of the nation's urban population (Ellis and Roberts 2016). This urban primacy is further underscored by the city's remarkable demographic growth, with its population more than doubling from 2.8 million in 2004 to 5.8 million in 2023 (National Statistics and Information Authority 2004, 2023). Projections indicate this rapid trend will continue, with Afghanistan's urban population expected to rapidly grow and reach half of the country's total population by 2060 (UN Habitat 2015). Notably, Kabul city was recognized as world's fifth fastest-growing city in 2015 (City Mayors 2015). A critical characteristic of KMR's urban expansion is its largely informal nature, with informal settlements constituting approximately 70% of residential areas (World Bank Group 2006; UN Habitat 2015) and providing shelter for an estimated 80% of the city's residents (Shahraki et al. 2020). This pattern of informal development has been a major driver of substantial land-use/land-cover (LULC) changes across the metropolitan region over the past two decades (DeWitt et al. 2022; Najmuddin et al. 2018, 2022; Zaryab et al. 2022). Consequently, the area is facing a confluence of intensifying environmental challenges, including: A significant increase in land surface temperatures (LST), with studies indicating a rise of +1.45°C to +3.78°C between 2009 and 2019 (Mahmoodi et al. 2019); Rapid depletion of groundwater resources, estimated at an average decline of 0.8 meters per year in the Kabul Plain, with localized areas experiencing drastic drops of up to 60 meters due to over-extraction (Zaryab et al. 2022); A discernible upward trend in aerosol optical depth levels from 2000 to 2022, particularly during the spring and summer seasons, indicating worsening air quality (Torabi et al. 2024); Increasing pressure on water resources due to rising demand from the growing population, coupled with declining rates of groundwater recharge (Gauster 2021; Taher et al. 2014). Given the synergistic effect of rapid and largely informal urbanization and the associated escalating environmental stressors, the KMR presents a highly relevant and crucial case study for a detailed investigation into the current state of ES supply within a data-scarce and rapidly transforming urban environment. Understanding the spatial patterns and interrelationships of these services in KMR can yield critical insights for informing sustainable urban planning strategies and effective ES management interventions tailored to the specific challenges of this region and potentially applicable to other similar contexts in the Global South. Material and methods Research design Over time, a range of conceptual frameworks and tools have emerged, including the Millennium Ecosystem Assessment (MEA; 2005), The Economics of Ecosystems and Biodiversity (TEEB; 2010), the Common International Classification of Ecosystem Services (CICES; 2012), and IPBES (2019). MAES methodological frameworks in particular provide structured approaches to identifying, assessing, quantifying, and valuing ESs. They guide how ESs are linked to human well-being, spatial patterns, trade-offs, and policy decisions (Albert et al. 2014; De Groot et al. 2010; Haines-Young and Potschin 2010). While specific methodological procedures vary depending on the research questions and context, many ES studies follow a general structure that includes three interlinked stages: selection of ESs, assessment and quantification of their supply, and application of the results to inform planning and decision-making. Some studies focus on identifying ES hotspots and coldspots to guide conservation or inform sustainable urban planning (e.g., Adem Esmail et al. 2025; Li et al. 2017), while others delve into understanding the complex interrelationships between services by analyzing ES bundles and their trade-offs and synergies (e.g., Karimi et al. 2021). Furthermore, many studies aim to translate these analytical outputs into practical applications, such as delineating ecological functional zones or identifying priority areas for management interventions (e.g., Deng and Cao 2023; Xu and Duan 2024). In this study, we structured our methodological framework accordingly into three stages: selection of ESs, mapping and assessment, and spatial-statistical analysis for application (Fig.2). Selection of ES: The selection was guided by the IPBES classification of Nature’s Contribution to People (NCP) to ensure representation across material, regulating, and non-material contributions. We selected five services: WY, HM, SWRR, NA, and HQ. This selection ensured the inclusion of diverse functions relevant to urban resilience and planning. ES mapping and assessment: The core of the methodological framework involved the biophysical quantification of ESs using the InVEST ®[1] (Integrated Valuation of Ecosystem Services and Tradeoffs) suite of models (Natural Capital Project 2024). InVEST has become a standard tool in ES research because it is user-friendly, simple, focuses on important ESs, and employs a multi-purpose, peer-reviewed methodology (Nemec and Raudsepp-Hearne 2013). Model outputs were then processed in ArcGIS Pro to generate spatially explicit layers of ES supply. These maps enabled the identification of spatial distribution patterns and are the backbone for the subsequent statistical analysis. Spatial and statistical analysis for application: We conducted a series of spatial-statistical analyses to translate the results into planning-relevant insights. The Getis-Ord Gi* statistic was applied to detect statistically significant clusters of high and low ES supply (Adem Esmail et al. 2025; Li et al. 2017). The k-means clustering was used in R (R Core Team 2024) to identify areas with similar ES profiles, aiding in the recognition of multifunctional zones (Queiroz et al. 2015; Raudsepp-Hearne et al. 2010). Next, to explore interrelationships among services, we used Spearman's rank correlation (Spearman 1904), a robust method for non-parametric association analysis (Li et al. 2024). And finally, to support targeted policy and intervention recommendations, a composite conservation priority map was generated by aggregating areas of high ES supply using mean and standard deviation (SD) thresholds. Overall, our approach follows a structured methodological logic: grounded in conceptual frameworks (IPBES), operationalized through biophysical modeling (InVEST), visualized through spatial analysis (ArcGIS Pro), interpreted statistically (R), and concluded with spatial prioritization (Fig. 2). Overall, this structured methodology leverages advanced geospatial techniques and modeling to generate actionable insights despite data limitations, aiming to support evidence-based decision-making for KMR's policymakers. The following sections explain each of these stages in greater detail. Data sources A comprehensive and multi-faceted dataset was compiled to support this study, encompassing satellite imagery, global repositories, and regional institutions. These data were crucial for the creation of a high-resolution LULC map (Fig. 1c) and the subsequent modeling of key ESs. Data sources for land-use/land-cover creation The foundation for our LULC mapping was the 10m resolution ESA WorldCover 2020 product (Zanaga et al. 2021). This base layer was enhanced by integrating supplementary datasets to produce a detailed 19-class LULC map (Fig. 1c). Key additions included crop-type classifications from Food and Agriculture Organization of the United Nations (FAO; 2020), Sentinel-2-derived canopy height (Lang et al. 2022), OpenStreetMap (OSM) infrastructure data (Geofabrik 2024), and built-up height information from the Global Human Settlement Layer (GHSL; Pesaresi 2023). These datasets were essential for accurately representing the urban and peri-urban landscape of KMR. A complete list of data sources used in the LULC creation process, including their resolution and year, is provided in Table 1. Data sources for ecosystem service assessment The spatial assessment of the five selected ESs relied on a diverse array of biophysical parameters and socio-demographic data. Critical climate inputs included monthly precipitation and temperature data from Fick and Hijmans (2017), daily temperature from Visual Crossing (2020), and extraterrestrial radiation from FAO (1998). Soil characteristics were obtained from International Soil Reference and Information Centre (ISRIC) SoilGrids (Hengl et al. 2017; ISRIC 2017), while topographical data were provided by the United States Geological Survey (USGS; 2018) and watershed boundaries by HydroSHEDS (Lehner and Grill 2013). Additional datasets included land surface albedo from RSLab (2020), rainfall intensity-duration-frequency data from the U.S. Army Corps of Engineers (USACE; 2010), hydrologic soil-cover complexes from U.S. Department of Agriculture-Natural Resources Conservation Service (USDA-NRCS; 2004), and soil hydrologic groups from Simons et al. (2020). For assessing NA, the analysis incorporated LULC naturalness data from Veteikis et al. (2011), GHSL population grids (Schiavina et al. 2023), and socio-demographic statistics from the Afghanistan Central Statistics Organization (CSO). HQ assessment used threat and sensitivity tables from Sallustio et al. (2017). A comprehensive list of all datasets used for the ES assessment, including their specific purposes and sources, is detailed in Table 2. For detailed descriptions of the ES models used, readers are referred to the InVEST user manuals (Natural Capital Project 2024). Ecosystem service selection The selection of ESs for this study is guided by IPBES’s NCP framework, which emphasizes the diverse ways ecosystems support human well-being in specific socio-ecological contexts (Díaz et al. 2018). The NCP classification integrates cultural, regulating, and material contributions, thereby fostering more effective and equitable decision-making for sustainable futures (Pascual et al. 2017). Five ESs were selected based on their relevance to the KMR’s rapid urbanization, climate vulnerabilities, and data availability, informed by expert consultations and a review of regional environmental pressures. These selections align with the IPBES emphasis on context-specific contributions (Díaz et al. 2018) and address KMR’s pressing needs for resource security, climate adaptation, and livability. Five selected ESs are: WY: This service is formally classified under the regulating contribution of “Regulation of freshwater quantity, location and timing.” However, our study focuses specifically on the tangible output of this process. Given that our assessment prioritizes the provision of water as a consumable good, we have classified it as a material contribution, an approach consistent with other recent studies (e.g., Külling et al. 2024). This approach better reflects its critical importance as a finite material resource for human consumption, agriculture, industry, and energy production, particularly within the arid climate and high-demand context of the KMR. HM: Classified as a regulating contribution, this service directly relates to the NCP of “Regulation of climate.” In the KMR, this function is vital for mitigating urban heat island effects at a local scale, a problem exacerbated by high building density and informal urban sprawl. SWRR: This is a regulating contribution corresponding to the NCP “Regulation of hazards and extreme events.” It is crucial for flood risk reduction in a city prone to heavy seasonal rains and characterized by areas with poor drainage infrastructure, thereby protecting property and lives. NA: This represents a non-material contribution, specifically the NCP of “Physical and psychological experiences.” These include the recreational, cultural, and mental health benefits that residents of the densely populated KMR derive from interacting with urban green spaces. HQ: As a measure of the NCP “Habitat creation and maintenance,” this is a core regulating contribution. While officially a regulating service, it is recognized as being foundational to the delivery of almost all other NCPs. By supporting local biodiversity, high HQ is essential for ensuring the long-term resilience and continued functioning of the KMR’s ecosystems. Mapping and assessment of ecosystem service supply potential To map ES supply potential, this study integrates multi-source data to create a high-resolution (10m) LULC map of the KMR (Table 1; Fig. 1b). The LULC map is then imported into InVEST along with other respected biophysical elements (Table 2) to quantify the supply of the five selected ESs. While comprehensive documentation of the ES models is available in the InVEST user manuals (Natural Capital Project 2024), we briefly describe the ES models applied in this study: WY was calculated using the InVEST annual water yield model, which estimates spatial variations in water provision by analyzing precipitation and evapotranspiration based on the Budyko curve. The model output represents the annual WY per pixel. HM was calculated using the InVEST urban cooling model, which assesses the capacity of urban landscapes to mitigate heat by quantifying cooling services from shade, evapotranspiration, and albedo. The output is a HM index for each pixel. SWRR was calculated using the InVEST urban flood risk mitigation model, which quantifies the influence of land use and soil types on runoff via the Curve Number (CN) method. The model estimates the annual volume of runoff avoided (retained) by the current land cover relative to a reference (impervious) scenario, with values provided per pixel. NA was calculated using the InVEST urban nature access model, which evaluates the availability of urban natural spaces for recreation via the Two-Step Floating Catchment Area (2SFCA) method. The model output quantifies the supply of nature space per population pixel. HQ was calculated using the InVEST habitat quality model, which assesses biodiversity by evaluating HQ and rarity based on LULC, and threats. Inputs comprised the LULC map, a threat table, and a sensitivity table. The model calculates a HQ score for each parcel. It is important to note that for the ESs selected in this study, the InVEST model provides assessment results in two distinct types. First, where biophysical parameters allow, it yields results in quantifiable units, facilitating comparisons across different case studies (e.g., WY in mm/m² per year, NA in m²/capita per pixel, and SWRR in m³/pixel). Second, for other services like HM, and HQ, InVEST generates relative indices, typically scaled between 0 and 1. These indices are invaluable for intra-study comparisons, allowing for the identification of areas with relatively higher or lower service provision within the study area. For instance, a HM index of 0.2 in one location and 0.4 in another indicates that the latter has twice the relative capacity for HM compared to the former. While absolute, unit-based values would ideally allow for direct comparisons with other urban areas globally, the primary objective of this study is to delineate the spatial variations and identify critical areas of high and low ES supply within KMR itself. Therefore, these relative indices serve as robust indicators for achieving this aim, highlighting areas requiring targeted interventions. Mapping and analyzing hotspots and coldspots of ecosystem service supply For mapping and analyzing hotspots and coldspots of ES (potential) supply, we resampled the original 10m x 10m square grid to a regular hexagonal grid with 1-hectare cells. Hexagonal grids provide more uniform target-neighbor relationships, as the nearest neighborhood definition is simpler and less ambiguous than in rectangular grids, which enhances the robustness of spatial autocorrelation results, particularly when connectivity or movement paths are relevant (Birch et al. 2007; Esri). Spatial patterns of ES supply were analyzed using hotspot and coldspot detection techniques. The Getis-Ord Gi* Statistics (Ord and Getis 1995) was applied to each ES layer to identify statistically significant clusters of high (hotspots) and low (coldspots) supply. This method calculates z-scores and p-values for each grid cell, highlighting areas where ES provision deviates from random distribution. The Getis-Ord Gi* Statistics (Ord and Getis 1995) was conducted in ArcGIS Pro based on a contiguity edges corner conceptualization of spatial relationships. Ecosystem service supply across population density gradients To extend the analysis beyond landscape composition and explore the direct interface with settlement patterns, we further investigated the relationship between ES provision and human population density. This analysis utilized the GHSL population grid (Schiavina et al. 2023). The continuous population data from this grid was reclassified into five discrete zones (Fig. 3) based on density thresholds informed by the classifications in the 2019 Master Plan for Kabul city (Sasaki 2019). The density zones are defined as follows: high population density (more than 250 people/hectare), medium population density (50–250 people/ha), low population density (5–50 people/ha), very low population density (1–5 people/ha), and unpopulated areas (0 people/ha). Following the reclassification, Zonal Statistics in ArcGIS Pro was employed to calculate the mean and total supply of each of the five selected ESs (WY, HM, SWRR, NA, and HQ) within each of the five population density zones. This approach allows for a direct comparison of ES supply potential across a gradient of urbanization and settlement intensity. Identification of ecosystem service bundles and interrelationships This step examines the interrelationships among the five ESs to identify distinct zones with similar supply profiles, known as ES bundles, and to detect synergies and trade-offs. To identify ES bundles, we applied k-means clustering algorithm (Hartigan and Wong 1979) to the standardized supply maps of the five ESs, grouping areas based on their multi-ES characteristics (Sun et al. 2022; Wu et al. 2022). The dataset includes 258,394 regular 1-hectare hexagonal grids. The optimal number of clusters (k) was determined using both the elbow and average silhouette width methods on a 10% random sample of the dataset. Based on a clear inflection point in the elbow plot and a peak average silhouette score, k=3 was selected as the optimal value. The final three-bundle groups were then generated by applying the algorithm to the full dataset. The entire clustering process was conducted in R (R Core Team 2024) using functions from the cluster (Maechler et al. 2025) and ClusterR (Mouselimis 2024) packages. Subsequently, to quantify the interrelationships, a Spearman’s rank correlation analysis (Spearman 1904) was performed. This analysis identified pairwise positive (synergies) or negative (trade-offs) associations between the five ESs. These calculations were carried out using the rcorr function from the Hmisc package (Harrell Jr 2025). These analyses reveal how ESs co-occur or conflict spatially, informing targeted planning strategies in KMR’s heterogeneous urban landscape. Identifying these ES bundles helps delineate ecologically distinct zones characterized by specific combinations of high and low service provision (e.g., Mouchet et al. 2014). Recognizing these spatial patterns of multifunctionality is crucial for tailoring management interventions (e.g., Queiroz et al. 2015; Spake et al. 2017); for example, understanding the typical ES profile of different zones can inform the strategic placement and type of nature-based solutions or green infrastructure projects. Mapping recommended areas for conservation Priority areas for conservation were identified based on the potential supply levels of the five assessed ESs (WY, HM, SWRR, NA, HQ). A two-step classification approach was used to identify areas providing multiple benefits. First, a primary classification based on the mean supply of each ES across the study area was performed. For each spatial unit (1 hectare hexagon), the number of ESs providing supply levels above their respective area-wide mean was counted. This count determined the primary conservation priority category. Highest priority: areas where all five ESs are above the mean supply; High priority: areas with four ESs above the mean; Medium priority: areas with three ESs above the mean; Low priority: areas with two ESs above the mean; and Lowest priority: areas where only one ES is above the mean supply. Second, these primary priority categories were refined using SD thresholds to assess the magnitude of ES supply relative to the mean. Spatial units were further sub-categorized based on how far the supply of contributing ESs exceeded the mean: High goes to locations where relevant ES supply is substantially above the mean (e.g., >1 SD above the mean); Moderate are locations where relevant ES supply is moderately above the mean (e.g., between 0.5 and 1 SD above the mean); and Low represent the areas where relevant ES supply is slightly above the mean (e.g., <0.5 SD above the mean). [1] https://naturalcapitalproject.stanford.edu/software/invest Results Ecosystem service assessment Our analysis revealed a significant spatial heterogeneity in the supply of ESs across the KMR, with distinct patterns of high and low provision (Fig. 4) and different contributions from areas within and outside the Kabul municipality (Table 3). High supply areas are predominantly located in the more natural and less disturbed landscapes in the northwestern, eastern, and southeastern parts of the study area, while low supply areas are concentrated within the central, densely urbanized core (Fig. 4). Overall, areas outside the Kabul municipality account for approximately 73% of the total aggregated ES potential supply in the study area, while areas within Kabul municipality contribute the remaining 27%. Table 3 Summary statistics of ecosystem service supply within and outside the Kabul municipality Ecosystem Service Statistic Overall Study Area Within Kabul Municipality Outside the Municipality Units/Range Water yield Mean 2.3 1.5 2.7 mm/m² per year SD 1.9 1.3 2.0 Sum (million) 5.8 1.6 4.2 m 3 /year Heat mitigation Mean 0.47 0.40 0.51 Relative Index SD 0.28 0.27 0.29 Stormwater runoff retention Mean 2.10 2.02 2.16 m 3 /pixel SD 0.59 0.56 0.60 Sum (million) 66.3 25.58 40.74 m 3 Nature access Mean (thousand) 243.10 90.84 345.59 m²/capita per pixel SD (thousand) 282.55 181.02 384.08 Range: 0.00 -2.95M Sum (trillion) 7.66 1.15 6.51 Habitat quality Mean 0.19 0.08 0.27 Relative Index SD 0.26 0.15 0.29 The relative indices for heat mitigation and habitat quality were scaled from 0 to 1, allowing for a comparison of relative service provision within the study area. Stormwater runoff retention values were calculated relative to a reference scenario (impervious surface). All analyses and mapping were conducted at a spatial resolution of approximately 10 m x 10 m. The abbreviation SD represents the standard deviation. The material contribution, represented by WY, is dominated by landscapes outside the Kabul municipality. These areas contribute nearly three-quarters of the total water supply and exhibit a significantly higher average yield (1.8 times) compared to areas within the municipality (Table 3; Fig. 4a). The supply of regulating contributions (HM, SWRR, and HQ) varies. For HM and SWRR, the supply is more balanced, with the average supply in areas outside the municipality being approximately 28% higher for HM and the total retention for SWRR being 1.6 times greater (Table 3; Fig 4b, 4c). In contrast, HQ shows a more pronounced spatial disparity, with the vast majority of high-quality habitat areas found outside the municipality (Table 3; Fig. 4e). The non-material contribution, represented by NA, also shows a significant spatial disparity. The vast majority of this service is supplied by areas outside the municipality, which provide around 85% of nature access (Table 3; Fig. 4d). Overall, the spatial patterns of ES supply are strongly correlated with the underlying LULC distribution. Areas with higher values for regulating contributions (HM, SWRR, HQ) predominantly overlap with less disturbed land covers (e.g., tree cover, shrubland, grassland) found largely outside the Kabul municipality. Material (WY) and non-material (NA) contributions also show higher values in these less urbanized zones. Conversely, areas within Kabul municipality consistently exhibited lower values across most assessed ESs, highlighting the trade-offs associated with dense urban development and landscape modification. While localized green spaces within the municipal boundary provide some benefits, their extent appears insufficient to counteract the broader impacts of urbanization on ES provision across this zone. Hotspot and coldspot analysis of ecosystem services Statistically significant spatial clustering of high (hotspots) and low (coldspots) ES provision revealed distinct patterns for each of the five assessed ESs across the KMR (Fig. 5). These patterns, varying notably between areas within Kabul municipality and areas outside the Kabul municipality, reflect the influence of underlying environmental and anthropogenic factors (Fig. 5; Fig. 6). Spatial analysis of the material contribution, WY, revealed significant clustering. Hotspots of high WY (99% confidence) covered 13% of the study area (33,405 ha; Fig. 6) and were predominantly located in areas outside the Kabul municipality (Fig. 5a), reflecting higher elevation characteristics. Conversely, coldspots of low WY (99% confidence) covered a smaller portion of the study area (7%; Fig. 6) but were significantly concentrated within the central, urbanized parts of the municipality (Fig. 5a). Regulating contributions showed a clear pattern of co-location and urban impact. For both HM and SWRR, hotspots of high supply covered a similar extent (Fig. 6)—21% of the study area (54,959 ha and 54,304 ha, respectively)—and were mainly concentrated in the less developed areas outside and on the fringes of the municipality (Fig. 5b, 5c). In stark contrast, coldspots were strongly aligned with the urban core. This was particularly evident for HM, where coldspots covered 24% of the study area (62,600 ha; Fig. 6), highlighting the significant impact of impervious surfaces on creating urban heat islands. For HQ, an interesting pattern emerged: while no statistically significant coldspots were identified at the highest confidence levels (99% or 95%), vast coldspots emerged at a lower confidence level (90%), covering an extensive 42% of the study area (Fig. 6). These were overwhelmingly concentrated within the Kabul municipality (Fig. 5e). For the non-material contribution, NA, a similar pattern of widespread deficit was observed. No statistically significant coldspots were identified at the highest confidence levels (99% or 95%), but at a lower confidence level (90%), vast coldspots emerged, covering 27% of the study area (Fig. 6). These low-confidence coldspots were also overwhelmingly concentrated within the Kabul municipality, covering 54% of its area (Fig. 5d). These spatially distinct clusters underscore the profound influence of LULC patterns on the heterogeneity of ES provision, indicating a widespread and pervasive deficit in accessible nature and quality habitat across the urban landscape, rather than isolated problem areas. Areas within Kabul municipality consistently exhibited lower ES values and higher concentrations of coldspots across most services, reflecting the impacts of urbanization. Conversely, areas outside the Kabul municipality contributed disproportionately to ES hotspots, aligning with the spatial gradients observed in the previous section. Ecosystem service supply across population density gradient The analysis of ES supply across population density gradients reveals a general inverse relationship, where service provision is highest in unpopulated areas and decreases with rising population density (Fig. 7). However, this trend manifests with different magnitudes and patterns across the services. The relationship is strong and direct for some regulating services, while for others it is less pronounced or exhibits more complex patterns. This disparity is starkly evident as unpopulated areas, covering 83% of the landscape, provide a disproportionately high share of the total supply for most services, including 92% of WY, 87% of HM, 97% of HQ, and nearly all NA (99.9%). The only exception is SWRR, where supply is largely proportional to land area across all zones. The trend of decreasing supply with increasing population density is most pronounced for the regulating contributions of HM and SWRR, which both show a clear and steady decline with rising population density (Fig. 7b, 7c). In contrast, the relationship is less direct for the material contribution (WY). While its mean supply is highest in unpopulated areas (approx. 2.5 mm/m² per year), the trend across the populated zones is not as obvious, and the supply drop to its lowest in the very low population density zone (1.0 mm/m² per year). The non-material contribution (NA) presents the most extreme disparity, with nearly all of its supply concentrated in unpopulated areas, making its mean supply drastically higher there than in any populated zone (Fig. 7d). HQ displays the most complex pattern. While it follows the overall trend of declining from unpopulated to populated areas, it shows a surprising inverse pattern within the populated zones, where its supply slightly increases with population density (Fig. 7e). The analysis also shows that the variability of supply (interquartile range) for services like HM and HQ is widest in unpopulated areas and narrows significantly in denser zones, suggesting a more uniformly low supply in settled landscapes. Overall, the inverse relationship between population density and ES supply (Fig. 7) underscores the impact of human settlement intensity on the landscape's capacity to provide essential ESs. Again, the most urbanized zones, characterized by high and medium population densities, correspond to the areas with the most pronounced and consistent deficits in ES provision. Ecosystem service bundles and trade-off/synergy analysis Ecosystem service bundles The k-means clustering analysis of the five standardized ESs (WY, HM, SWRR, NA, and HQ) across the 258,394 hexagonal grids identified three distinct ES bundles. These bundles represent unique combinations of ES provision levels. Based on their characteristic ES supply profiles (Table 4), descriptive names were assigned to each bundle to aid in their interpretation. Table 4 Characteristics of the three ecosystem service bundles identified in Kabul metropolitan region Bundle Water Yield Heat Mitigation Stormwater Runoff Retention Nature Access Habitat Quality Area Coverage (ha) Area Coverage (%) Bundle 1 0.00 -0.83 -0.83 -0.40 -0.25 130,422 51% Bundle 2 -0.63 0.77 0.84 -0.14 -0.51 82,183 32% Bundle 3 1.13 0.98 0.87 1.39 1.64 45,789 18% Values for water yield, heat mitigation, stormwater runoff retention, nature access, and habitat quality represent the standardized (z-score) values of each attribute at the centroid of each k-means cluster (bundle). A value of 0 indicates the mean, positive values indicate above-average ecosystem service provision, and negative values indicate below-average provision. Percentages are based on the total study area (258,394 ha). Bundle 1 (Low-Service Zone) is characterized by approximately average WY, while exhibiting notably low provision of HM (-0.83 SD) and SWRR (-0.83 SD). It also shows moderately below-average levels of NA (-0.40 SD) and HQ (-0.25 SD). This bundle is the most dominant, covering 130,422 ha (51%) of the study area (Table 4). Spatially, it aligns strongly with the central, densely urbanized core of the Kabul municipality (Fig. 8). Bundle 2 (Regulating-Service Zone) presents a distinct profile with significantly above-average provision of HM (+0.77 SD) and SWRR (+0.84 SD). However, it shows low WY (-0.63 SD) and HQ (-0.51 SD), with slightly below-average NA (-0.14 SD). It covers 82,183 ha (32%) of the study area (Table 4). This bundle is primarily located in the peri-urban and agricultural landscapes surrounding the urban core (Fig. 8). Bundle 3 (High-Service Multifunctional Zone) demonstrates the highest provision across multiple services. It features significantly above-average levels of WY (+1.13 SD), HM (+0.98 SD), SWRR (+0.87 SD), NA (+1.39 SD), and exceptionally high HQ (+1.64 SD). It is the least extensive bundle, covering 45,789 ha (18%) of the study area (Table 4). Spatially, it is concentrated in the outer, more natural landscapes in the eastern, southeastern, and northwestern parts of the study area (Fig. 8). Analysis of Spearman's rank correlations within each of the three identified ES bundles revealed context-dependent relationships between services, highlighting how synergies and trade-offs vary across the different landscape types (Fig. 9). A strong and consistent synergy between the two regulating contributions, HM and SWRR, was observed across all bundles (Fig. 9). The intensity of this synergy varied, being strongest in the High-Service Multifunctional Zone (Bundle 3; ρ = 0.64) and the Low-Service Zone (Bundle 1; ρ = 0.51), and slightly weaker, though still moderate, in the Regulating-Service Zone (Bundle 2; ρ = 0.40). This indicates that across the entire study area, interventions to increase vegetation for cooling would also yield significant benefits for stormwater management. A strong and consistent synergy between the two regulating contributions, HM and SWRR, was observed across all bundles (Fig. 9). The intensity of this synergy varied, being strongest in the High-Service Multifunctional Zone (Bundle 3; ρ = 0.64) and the Low-Service Zone (Bundle 1; ρ = 0.51), and slightly weaker, though still moderate, in the Regulating-Service Zone (Bundle 2; ρ = 0.40). This indicates that across the entire study area, interventions to increase vegetation for cooling would also yield significant benefits for stormwater management. The interrelationships involving the non-material contribution (NA) and the regulating contribution of HQ were the most variable. A strong synergy between NA and HQ was found in the Regulating-Service Zone (Bundle 2; ρ = 0.57), suggesting that in these semi-natural landscapes, high-quality habitats are also highly accessible. However, this synergy weakened significantly in the Low-Service Zone (Bundle 1; ρ = 0.28) and flipped to a weak trade-off in the High-Service Multifunctional Zone (Bundle 3; ρ = -0.13), indicating a potential conflict between maximizing biodiversity and recreational use in these prime areas. The relationship between HM and NA also shifted dramatically, from a strong synergy in the High-Service Multifunctional Zone (Bundle 3; ρ = 0.56) to a weak synergy in the Low-Service Zone (Bundle 1; ρ = 0.20) and no significant relationship in the Regulating-Service Zone (Bundle 2; ρ = 0.00). Relationships involving the material contribution, WY, also showed context dependency, primarily exhibiting trade-offs with other regulating services. The trade-off between WY and SWRR, expected from hydrological principles, was most pronounced in the Regulating-Service Zone (Bundle 2; ρ = -0.29) and weaker in the Low-Service Zone (Bundle 1; ρ = -0.21) and High-Service Multifunctional Zone (Bundle 3; ρ = -0.15). Similarly, a weak trade-off between WY and HM was observed in the Low-Service Zone (Bundle 1; ρ = -0.20) but became negligible in the other two bundles. These bundle-specific correlation patterns underscore that ES interrelationships are not uniform across the KMR. Management strategies aimed at enhancing multiple services or mitigating trade-offs must therefore consider the specific landscape context, as represented by the ES bundles, to be most effective. For instance, in the Regulating-Service Zone (Bundle 2), management actions could effectively leverage the strong synergies between WY, NA, and HQ. Conversely, in the High-Service Multifunctional Zone (Bundle 3), while most services are synergistic, careful planning is needed to manage the slight trade-off between providing NA and maintaining the highest levels of HQ. Recommended areas for conservation Building upon the ES assessments, a conservation priority analysis was conducted to identify zones warranting different levels of attention based on the provision of multiple ES benefits (Fig. 10). This classification evaluates the number of ESs performing above their respective area-wide means, refined by SD thresholds to indicate the magnitude of the benefit. The analysis reveals a clear spatial hierarchy of conservation needs. Areas designated as highest conservation priority, where all five ESs are supplied above their mean, cover 301 km² (12% of the study area). These are complemented by high conservation priority areas (four ESs above mean), which account for another 253 km² (10%). Spatially, these top-tier priority zones are concentrated in the eastern, southeastern, and northwestern and northeastern sections outside the municipality, corresponding to more natural land covers. While the majority of these areas (502 km²) provide benefits of low to moderate magnitude, a combined 52 km² within these two top tiers show high-magnitude benefits (greater than 1 SD above the mean). Representing landscapes with moderate multifunctionality, Medium Conservation Priority areas (three ESs above mean) cover 264 km² (10%). These are predominantly located outside the municipality near higher-priority zones, though some patches exist within the municipality away from the urban core (Fig. 10). The most extensive category is Low Conservation Priority (two ESs above mean), covering 654 km² (25%) of the landscape. These areas, which include agricultural lands and less dense vegetated areas, are found in large portions both within and outside the municipality, often forming a transition between more natural and highly developed zones. Finally, a significant portion of the KMR shows a profound deficit in multiple ES benefits. The Lowest Conservation Priority areas (one ES above mean) cover 496 km² (19%), while a further 604 km² (23%) was classified as a No Conservation Priority zone, having no services performing above the area-wide mean. These two lowest categories, together comprising 42% of the study area, strongly align with the central urban core and other degraded landscapes, reflecting the lowest levels of multiple ES provision. Discussion This study provides a critical, spatially explicit assessment of five key ESs (ESs)—WY, HM, SWRR, NA, and HQ—within the rapidly urbanizing, data-scarce context of the KMR. Our integrated approach, combining high-resolution LULC mapping with ES modeling, bundle analysis, and conservation prioritization, reveals profound spatial heterogeneity in service provision, intricate and context-dependent interrelationships between services, and clear spatial patterns of conservation needs. These findings offer crucial insights for sustainable urban planning in a challenging environment, contributing valuable knowledge relevant to KMR and other cities facing similar pressures in the Global South. Spatial heterogeneity and ecosystem service bundles Consistent with findings from diverse urban contexts (Liu et al. 2019; Bai et al. 2019; Cozzi et al. 2022), our results demonstrate a marked spatial gradient in ES provision, with significantly lower supply concentrated in the densely built-up municipal areas compared to the municipal periphery and surrounding non-municipal areas. This urban-rural dichotomy underscores the substantial impact of urbanization, particularly the proliferation of impervious surfaces and loss of vegetation, on local ecological functions. However, moving beyond a simple gradient, our identification of three distinct ES bundles provides a more nuanced understanding of KMR's landscape structure (Raudsepp-Hearne et al. 2010). These bundles represent unique landscape archetypes, ranging from the service-deficient Low-Service Zone (Bundle 1) dominating the more urbanized municipal areas, to the specialized Regulating-Service Zone (Bundle 2) in the municipal periphery, and culminating in the High-Service Multifunctional Zone (Bundle 3) in the most intact outer zones. Recognizing these distinct bundle types, each with a characteristic signature of high and low service provision, is essential for moving beyond generalized planning approaches towards spatially tailored strategies that leverage or mitigate the specific ES profile of different landscape units (Mouchet et al. 2014; Queiroz et al. 2015). Our findings on the relationship between ES supply and population density, which serves as a proxy for land-use intensity, can be contextualized using the stylized trajectories proposed by Locatelli et al. (2017). The strong inverse relationship we identified for HM and SWRR aligns well with their model, which posits that most regulating services decline as land-use intensity increases from natural to urban states. However, our analysis also reveals patterns with more nuance than these general models suggest. For instance, the supply of the material contribution, WY, did not follow the typical trajectory of peaking at an intermediate intensity; instead, it was highest in unpopulated areas with no clear trend across the populated zones. More surprisingly, the regulating service of HQ, while highest in unpopulated areas, exhibited a slight increase with population density within the populated zones. This is a notable deviation from the expected steady decline and may reflect the specific ecological characteristics of the remaining green spaces within Kabul's informally developed landscapes. These unique patterns underscore the assertion by Locatelli et al. (2017) that while stylized models are useful, they must be adapted to account for local context and the specific nature of land-use transitions. Context-dependent ecosystem service interrelationships A key contribution of this study is the demonstration that relationships between ESs are highly context-dependent, varying significantly across the three identified bundles. While the overall analysis revealed expected patterns, the bundle-specific correlations uncovered important nuances. For instance, the synergy between HM and SWRR was present across all bundles but was strongest in the High-Service Multifunctional Zone (Bundle 3). The trade-off between WY and SWRR was most pronounced in the Regulating-Service Zone (Bundle 2). Furthermore, the relationship between NA and HQ shifted from a strong synergy in the Regulating-Service Zone (Bundle 2) to a weak trade-off in the High-Service Multifunctional Zone (Bundle 3), suggesting a potential conflict between maximizing habitat integrity and providing human access in these highest-value areas. This context-dependency highlights the limitations of relying solely on landscape-wide correlations and emphasizes the need for management interventions to be tailored to the specific bundle type to effectively manage trade-offs and leverage potential synergies (Spake et al. 2017). Implications for conservation planning The prioritization analysis translates the complex ES patterns into actionable spatial guidance for conservation planning. The results reveal a stark spatial segregation of needs: the highest conservation priorities, representing areas providing multiple ES benefits simultaneously, are overwhelmingly concentrated in the remaining natural and semi-natural landscapes, mostly corresponding to the High-Service Multifunctional Zone (Bundle 3) and parts of the Regulating-Service Zone (Bundle 2). Conversely, the areas with no conservation priority, characterized by deficits across multiple ESs, are heavily clustered within the central municipal areas, aligning with the Low-Service Zone (Bundle 1). This clear spatial dichotomy underscores the urgent need for a dual planning strategy that moves beyond static conservation. This strategy must involve both protecting the vital, multifunctional green and blue infrastructure and, crucially, implementing restoration interventions in the lower-priority urban matrix. Such interventions, for example establishing green corridors or "stepping-stone" habitats, would not only improve local livability but could also enhance ecological connectivity between the fragmented high-priority zones. The magnitude analysis within each priority level further allows for finer-grained prioritization, directing initial conservation and restoration efforts towards areas with the most substantial multi-service benefits or the greatest potential for improving connectivity. Finally, it is imperative that any such planning explicitly incorporates principles of environmental justice to ensure the equitable distribution of these benefits, a particularly critical consideration in the socio-political context of Kabul, which has a history of intense ethnic competition and civil war. Methodological significance and contribution This study demonstrates the utility of integrating multi-source remote sensing data, process-based models (InVEST), and spatial analysis techniques (clustering, prioritization) to assess ESs in data-scarce environments. By developing a high-resolution (10m) LULC map, our analysis captures finer-scale landscape heterogeneity compared to previous coarser assessments in the region (e.g., Najmuddin et al. 2022), providing more relevant information for urban planning. The application of ES bundle analysis and context-specific correlation assessment offers a more nuanced understanding of ES interrelationships than typically achieved through landscape-wide correlations alone. The findings provide a crucial baseline for KMR, addressing the call by IPBES (2018) for better biodiversity and ES information in the rapidly changing Asia-Pacific region. Furthermore, the methodological framework and insights are potentially transferable to other cities in the Global South facing similar challenges of rapid informal urbanization, data limitations, and escalating environmental pressures. Limitations and future research directions Despite its contributions, this study has limitations. The reliance on globally available datasets (e.g., SoilGrids) introduces uncertainties, as these may not fully capture local variations in climate and soil properties critical for ES modeling. While the LULC map represents a significant improvement, potential classification errors remain, and the InVEST models, like all models, rely on simplifying assumptions and parameters (e.g., threat/sensitivity tables for HQ) that require local calibration and validation for enhanced accuracy. Ground-truthing of both LULC and ES model outputs was beyond the scope of this study but represents a crucial next step. Furthermore, the analysis provides a static snapshot (circa 2020), and its focus on five ESs, while relevant, does not capture the full spectrum of services. The study also does not explicitly model the dynamic processes of informal settlement expansion or climate change impacts on future ES provision. Future research should prioritize acquiring local climate, soil, and hydrological data to refine ES models and validate their outputs. To achieve a more holistic understanding, future assessments should also consider expanding the analysis to include a broader range of ESs, such as food production, air purification, and other services, to better capture the entire spectrum of benefits. Dynamic LULC modeling, incorporating scenarios of future urbanization (both formal and informal) and climate change, is needed to assess potential future trajectories of ES supply and interrelationships. Investigating the socio-economic dimensions, including ES demand, access equity across different demographic groups, and community perceptions, is essential for developing socially just and effective management strategies. Exploring the governance challenges and opportunities for implementing conservation recommendations, particularly within informally settled areas, is also critical for translating research into tangible outcomes. Finally, comparative studies with other cities facing similar contexts would help generalize findings and refine methodologies for ES assessment in data-scarce environments. Conclusion This research provides a comprehensive and spatially explicit baseline understanding of ES provision, heterogeneity, interrelationships, and priorities within the KMR. By identifying distinct ES bundles, revealing context-dependent synergies and trade-offs, and delineating clear priorities for conservation, the study offers valuable, actionable insights for evidence-based urban planning and environmental management. Addressing the significant spatial disparities and leveraging the multi-functional benefits of remaining green and blue infrastructure are critical for enhancing KMR's resilience, sustainability, and livability in the face of rapid urbanization and environmental change. This work underscores the importance of tailored, high-resolution ES assessments, particularly in data-scarce regions of the Global South, to guide development towards more sustainable pathways. Declarations Acknowledgments This paper is part of the first author's PhD project at the Institute of Environmental Planning, Leibniz University Hannover. The author(s) would like to express their sincere gratitude to the German Academic Exchange Service (DAAD) for providing the Hilde Domin student at-risk PhD scholarship, and to the Institute of Environmental Planning, Leibniz University Hannover for their support, which made this research possible. We also acknowledge the valuable contributions of all public data source providers who made their data publicly available, enabling this research. Funding Funding for this research was provided by the German Academic Exchange Service (DAAD) through the Hilde Domin Programme PhD, 2023 (57615866) under the award number 91853250. Author contributions M.R.A. wrote the main manuscript text. The study was conceptualized by M.R.A., M.R., B.A.E., L.B.F., and C.A. The methodology was designed by M.R.A., M.R., B.A.E., and C.A., while M.R.A. also performed the data curation. Formal analysis was conducted by M.R.A. and M.R. Visualizations were prepared by M.R.A., D.K., and C.A. Additionally, C.A. was responsible for funding acquisition, project administration, and supervision. All authors reviewed and edited the manuscript. Declaration of competing interest The authors declare no competing interests. References Adem Esmail B, Cortinovis C, Geneletti D, et al (2025) Mapping and Analyzing Ecosystem Services Hotspots and Coldspots for Sustainable Spatial Planning in the Greater Asmara Area, Eritrea. Environ Manage 75:551–567. https://doi.org/10.1007/s00267-024-02078-x Adem Esmail B, Cortinovis C, Wang J, et al (2023) Mapping and assessing ecosystem services for sustainable policy and decision-making in Eritrea. Ambio 52:1022–1039. https://doi.org/10.1007/s13280-023-01841-4 Akhtar M, Zhao Y, Gao G, et al (2022) Assessment of spatiotemporal variations of ecosystem service values and hotspots in a dryland: A case‐study in Pakistan. Land Degrad Dev 33:1383–1397. https://doi.org/10.1002/ldr.4245 Albert C, Aronson J, Fürst C, Opdam P (2014) Integrating ecosystem services in landscape planning: requirements, approaches, and impacts. Landsc Ecol 29:1277–1285. https://doi.org/10.1007/s10980-014-0085-0 Albert C, Bonn A, Burkhard B, et al (2016) Towards a national set of ecosystem service indicators: Insights from Germany. Ecol Indic 61:38–48. https://doi.org/10.1016/j.ecolind.2015.08.050 Allen RG, Pereira LS, Raes D, Smith M (1998) Crop evapotranspiration: Guidelines for computing crop water requirements (Chapter 8, Table 22). In: FAO Irrigation and Drainage Paper No. 56. FAO Bagstad KJ, Semmens DJ, Ancona ZH, Sherrouse BC (2017) Evaluating alternative methods for biophysical and cultural ecosystem services hotspot mapping in natural resource planning. Landsc Ecol 32:77–97. https://doi.org/10.1007/s10980-016-0430-6 Bai L, Xiu C, Feng X, Liu D (2019) Influence of urbanization on regional habitat quality:a case study of Changchun City. Habitat Int 93:102042. https://doi.org/10.1016/j.habitatint.2019.102042 Birch CPD, Oom SP, Beecham JA (2007) Rectangular and hexagonal grids used for observation, experiment and simulation in ecology. Ecol Model 206:347–359. https://doi.org/10.1016/j.ecolmodel.2007.03.041 Bolund P, Hunhammar S (1999) Ecosystem services in urban areas. Ecol Econ 29:293–301. https://doi.org/10.1016/S0921-8009(99)00013-0 Central Statistics Organization (2015a) Kabul Socio-Demographic and Economic Survey. Central Statistics Organization Central Statistics Organization (2015b) Parwan Socio-Demographic and Economic Survey. Central Statistics Organization Chaturvedi V, Kuffer M, Kohli D (2020) Analysing Urban Development Patterns in a Conflict Zone: A Case Study of Kabul. Remote Sens 12:3662. https://doi.org/10.3390/rs12213662 City Mayors (2015) City Mayors: World’s fastest growing urban areas (1). http://www.citymayors.com/statistics/urban_growth1.html. Accessed 26 Sep 2024 Costanza R, d’Arge R, de Groot R, et al (1997) The value of the world’s ecosystem services and natural capital. Nature 387:253–260. https://doi.org/10.1038/387253a0 Cozzi M, Prete C, Viccaro M, et al (2022) Understanding the Role of Nature in Urban-Rural Linkages: Identifying the Potential Role of Rural Nature-Based Attractive Clusters That Serve Human Well-Being. Sustainability 14:11856. https://doi.org/10.3390/su141911856 Daily GC (ed) (2010) Nature’s services: societal dependence on natural ecosystems. Island Press, Washington, DC De Groot RS, Alkemade R, Braat L, et al (2010) Challenges in integrating the concept of ecosystem services and values in landscape planning, management and decision making. Ecol Complex 7:260–272. https://doi.org/10.1016/j.ecocom.2009.10.006 Deng Z, Cao J (2023) Incorporating ecosystem services into functional zoning and adaptive management of natural protected areas as case study of the Shennongjia region in China. Sci Rep 13:18870. https://doi.org/10.1038/s41598-023-46182-0 DeWitt JD, Boston KM, Alessi MA, Chirico PG (2022) Quantifying and visualizing 32 years of agricultural land use change in Kabul, Afghanistan. J Maps 18:352–361. https://doi.org/10.1080/17445647.2022.2063079 Díaz S, Pascual U, Stenseke M, et al (2018) Assessing nature’s contributions to people. Science 359:270–272. https://doi.org/10.1126/science.aap8826 Ellis P, Roberts M (2016) Leveraging Urbanization in South Asia: Managing Spatial Transformation for Prosperity and Livability. World Bank, Washington, DC Esri Why hexagons?—ArcGIS Pro | Documentation. https://pro.arcgis.com/en/pro-app/latest/tool-reference/spatial-statistics/h-whyhexagons.htm. Accessed 17 Apr 2025 Fick SE, Hijmans RJ (2017) WorldClim 2: new 1‐km spatial resolution climate surfaces for global land areas. Int J Climatol 37:4302–4315. https://doi.org/10.1002/joc.5086 Food and Agriculture Organization of the United Nations (2021) Crop type (Afghanistan - 10m - 2020 - 6 classes) - EOSTAT Food and Agriculture Organization of the United Nations (1998) Daily extraterrestrial radiation (15th day of the month). In Annex 2. Meteorological tables of Crop evapotranspiration - Guidelines for computing crop water requirements (FAO Irrigation and Drainage Paper 56) Gangahagedara R, Subasinghe S, Lankathilake M, et al (2021) Ecosystem Services Research Trends: A Bibliometric Analysis from 2000–2020. Ecologies 2:366–379. https://doi.org/10.3390/ecologies2040021 Gauster M (2021) Ecological threats to security and state resilience in Afghanistan. Secur Def Q 33:31–40. https://doi.org/10.35467/sdq/132023 Geneletti D (2016) Ecosystem services analysis for Strategic Environmental Assessment: concepts and examples. In: Geneletti D (ed) Handbook on Biodiversity and Ecosystem Services in Impact Assessment. Edward Elgar Publishing Geofabrik (2024) Afghanistan latest shapefiles Haines-Yong R, Potschin M (2012) Common International Classification of Ecosystem Services (CICES) Version 4. European Environment Agency Haines-Young R, Potschin M (2010) The links between biodiversity, ecosystem services and human well-being. In: Raffaelli DG, Frid CLJ (eds) Ecosystem Ecology, 1st edn. Cambridge University Press, pp 110–139 Harrell Jr FE (2025) Hmisc: Harrell Miscellaneous Hartigan JA, Wong MA (1979) Algorithm AS 136: A K-Means Clustering Algorithm. Appl Stat 28:100. https://doi.org/10.2307/2346830 Hengl T, Mendes De Jesus J, Heuvelink GBM, et al (2017) SoilGrids250m: Global gridded soil information based on machine learning. PLOS ONE 12:e0169748. https://doi.org/10.1371/journal.pone.0169748 IPBES (2019) Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. IPBES Secretariat IPBES, Karki M, Sellamuttu SS, et al (2018) The IPBES regional assessment report on biodiversity and ecosystem services for Asia and the Pacific. Zenodo. URL https://zenodo.org/record/3237373 ISRIC - World Soil Information (2017) SoilGrids250m: Derived available soil water capacity (volumetric fraction) until wilting point at seven standard depths (0 cm, 5 cm, 15 cm, 30 cm, 60 cm, 100 cm, 200 cm) Karimi JD, Corstanje R, Harris JA (2021) Bundling ecosystem services at a high resolution in the UK: trade-offs and synergies in urban landscapes. Landsc Ecol 36:1817–1835. https://doi.org/10.1007/s10980-021-01252-4 Kowarik I, Bartz R, Brenck M, Hansjürgens B (2017) Ecosystem services in the city: protecting health and enhancing quality of life: summary for decision-makers. Naturkapital Deutschland – TEEB DE, Berlin Leipzig Külling N, Adde A, Lambiel A, et al (2024) Nature’s contributions to people and biodiversity mapping in Switzerland: spatial patterns and environmental drivers. Ecol Indic 163:112079. https://doi.org/10.1016/j.ecolind.2024.112079 Lang N, Schindler K, Wegner JD (2022) ETH_GlobalCanopyHeight_10m_2020_version1: Global canopy height map for the year 2020 derived from Sentinel-2 and GEDI Lehner B, Grill G (2013) Global river hydrography and network routing: baseline data and new approaches to study the world’s large river systems. Hydrol Process 27:2171–2186 Li T, Ren Y, Ai Z, et al (2024) Revealing the Spatial Interactions and Driving Factors of Ecosystem Services: Enlightenments under Vegetation Restoration. Land 13:511. https://doi.org/10.3390/land13040511 Li Y, Zhang L, Yan J, et al (2017) Mapping the hotspots and coldspots of ecosystem services in conservation priority setting. J Geogr Sci 27:681–696. https://doi.org/10.1007/s11442-017-1400-x Liu W, Zhan J, Zhao F, et al (2019) Impacts of urbanization-induced land-use changes on ecosystem services: A case study of the Pearl River Delta Metropolitan Region, China. Ecol Indic 98:228–238. https://doi.org/10.1016/j.ecolind.2018.10.054 Locatelli B, Lavorel S, Sloan S, et al (2017) Characteristic trajectories of ecosystem services in mountains. Front Ecol Environ 15:150–159. https://doi.org/10.1002/fee.1470 Maechler M, Rousseeuw P, Struyf A, et al (2025) cluster: Cluster Analysis Basics and Extensions Maes J, Egoh B, Willemen L, et al (2012) Mapping ecosystem services for policy support and decision making in the European Union. Ecosyst Serv 1:31–39. https://doi.org/10.1016/j.ecoser.2012.06.004 Mahendra A, King R, Du J, et al (2021) Towards a More Equal City: Seven Transformations for More Equitable and Sustainable Cities. World Resour Inst. https://doi.org/10.46830/wrirpt.19.00124 Mahmoodi S, Dutta K, Basu D, Agrawal S (2019) UNDERSTANDING LINK BETWEEN LAND SURFACE TEMPERATURE AND LANDSCAPE HETEROGENEITY: A SPATIO-TEMPORAL AND INTER-SEASONAL VARIABILITY STUDY ON KABUL CITY, AFGHANISTAN. ISPRS Ann Photogramm Remote Sens Spat Inf Sci IV-5/W2:57–65. https://doi.org/10.5194/isprs-annals-IV-5-W2-57-2019 McPhearson T, Andersson E, Elmqvist T, Frantzeskaki N (2015) Resilience of and through urban ecosystem services. Ecosyst Serv 12:152–156. https://doi.org/10.1016/j.ecoser.2014.07.012 Millennium Ecosystem Assessment (2005) Ecosystems and Human Well-being: Synthesis. Island Press, Washington, DC Mouchet MA, Lamarque P, Martín-López B, et al (2014) An interdisciplinary methodological guide for quantifying associations between ecosystem services. Glob Environ Change 28:298–308. https://doi.org/10.1016/j.gloenvcha.2014.07.012 Mouselimis L (2024) ClusterR: Gaussian Mixture Models, K-Means, Mini-Batch-Kmeans, K-Medoids and Affinity Propagation Clustering Najmuddin O, Deng X, Bhattacharya R (2018) The Dynamics of Land Use/Cover and the Statistical Assessment of Cropland Change Drivers in the Kabul River Basin, Afghanistan. Sustainability 10:. https://doi.org/10.3390/su10020423 Najmuddin O, Li Z, Khan R, Zhuang W (2022) Valuation of Land-Use/Land-Cover-Based Ecosystem Services in Afghanistan—An Assessment of the Past and Future. Land 11:. https://doi.org/10.3390/land11111906 National Environmental Protection Agency (2014) Fifth National Report to the United Nations Convention on Biological Diversity – Afghanistan. United Nations Convention on Biological Diversity (CBD) National Statistics and Information Authority (2004) Afghanistan Statistical Yearbook 2004. National Statistics and Information Authority of Afghanistan, Kabul, Afghanistan National Statistics and Information Authority (2023) Afghanistan Statistical Yearbook 2023-24. National Statistics and Information Authority of Afghanistan, Kabul, Afghanistan Natural Capital Project (2024) Integrated Valuation of Ecocystem Services and Tradeoffs (InVEST) Nemec KT, Raudsepp-Hearne C (2013) The use of geographic information systems to map and assess ecosystem services. Biodivers Conserv 22:1–15. https://doi.org/10.1007/s10531-012-0406-z Ord JK, Getis A (1995) Local Spatial Autocorrelation Statistics: Distributional Issues and an Application. Geogr Anal 27:286–306. https://doi.org/10.1111/j.1538-4632.1995.tb00912.x Pascual U, Balvanera P, Díaz S, et al (2017) Valuing nature’s contributions to people: the IPBES approach. Curr Opin Environ Sustain 26–27:7–16. https://doi.org/10.1016/j.cosust.2016.12.006 Pesaresi M (2023) GHS-BUILT-H R2023A - GHS building height, derived from AW3D30, SRTM30, and Sentinel2 composite (2018) Queiroz C, Meacham M, Richter K, et al (2015) Mapping bundles of ecosystem services reveals distinct types of multifunctionality within a Swedish landscape. AMBIO 44:89–101. https://doi.org/10.1007/s13280-014-0601-0 R Core Team (2024) R: A Language and Environment for Statistical Computing Raudsepp-Hearne C, Peterson GD, Bennett EM (2010) Ecosystem service bundles for analyzing tradeoffs in diverse landscapes. Proc Natl Acad Sci 107:5242–5247. https://doi.org/10.1073/pnas.0907284107 RSLab (2020) True Land Surface Albedo Sallustio L, De Toni A, Strollo A, et al (2017) Assessing habitat quality in relation to the spatial distribution of protected areas in Italy. J Environ Manage 201:129–137. https://doi.org/10.1016/j.jenvman.2017.06.031 Sasaki (2019) Kabul Urban Design Framework (KUDF). Sasaki Company, USA Schiavina M, Freire S, Carioli A, MacManus K (2023) GHS-POP R2023A - GHS population grid multitemporal (1975-2030) Seppelt R, Dormann CF, Eppink FV, et al (2011) A quantitative review of ecosystem service studies: approaches, shortcomings and the road ahead: Priorities for ecosystem service studies. J Appl Ecol 48:630–636. https://doi.org/10.1111/j.1365-2664.2010.01952.x Shackleton CM, Cilliers SS, Du Toit MJ, Davoren E (2021) The Need for an Urban Ecology of the Global South. In: Shackleton CM, Cilliers SS, Davoren E, Du Toit MJ (eds) Urban Ecology in the Global South. Springer International Publishing, Cham, pp 1–26 Shahraki SZ, Hosseini A, Sauri D, Hussaini F (2020) Fringe more than context: Perceived quality of life in informal settlements in a developing country: The case of Kabul, Afghanistan. Sustain Cities Soc 63:102494. https://doi.org/10.1016/j.scs.2020.102494 Simons G, Koster R, Droogers P (2020) HiHydroSoil v2.0 – High Resolution Soil Maps of Global Hydraulic Properties. FutureWater, Wageningen, The Netherlands Spake R, Lasseur R, Crouzat E, et al (2017) Unpacking ecosystem service bundles: Towards predictive mapping of synergies and trade-offs between ecosystem services. Glob Environ Change 47:37–50. https://doi.org/10.1016/j.gloenvcha.2017.08.004 Spearman C (1904) The Proof and Measurement of Association between Two Things. Am J Psychol 15:72. https://doi.org/10.2307/1412159 Srivathsa A, Vasudev D, Nair T, et al (2023) Prioritizing India’s landscapes for biodiversity, ecosystem services and human well-being. Nat Sustain 6:568–577. https://doi.org/10.1038/s41893-023-01063-2 Sun X, Yang P, Tao Y, Bian H (2022) Improving ecosystem services supply provides insights for sustainable landscape planning: A case study in Beijing, China. Sci Total Environ 802:149849. https://doi.org/10.1016/j.scitotenv.2021.149849 Taher MR, Chornack MP, Mack TJ (2014) Groundwater levels in the Kabul Basin, Afghanistan, 2004–2013. United States Geological Survey TEEB (2010) The Economics of Ecosystems and Biodiversity: Mainstreaming the Economics of Nature. Earthscan Torabi SE, Amin M, Phairuang W, et al (2024) High-Resolution Characterization of Aerosol Optical Depth and Its Correlation with Meteorological Factors in Afghanistan. Atmosphere 15:849. https://doi.org/10.3390/atmos15070849 UN Habitat (2015) The State of Afghan Cities. United Nations Human Settlements Programme (UN-Habitat), Kabul U.S. Army Corps of Engineers (2010) AED design requirements: Hydrology studies (Provisional). U.S. Army Corps of Engineers, Afghanistan Engineer District U.S. Department of Agriculture (2004) Chapter 9 - Hydrologic Soil-Cover Complexes. In: National Engineering Handbook. Part 630 – Hydrology U.S. Geological Survey (2018) Shuttle Radar Topography Mission (SRTM) 1 Arc-Second Global Veteikis D, Šabanovas S, Jankauskaitė M (2011) Landscape structure changes on the coastal plain of Lithuania during 1998–2009. Baltica 24:107–116 Vihervaara P, Rönkä M, Walls M (2010) Trends in Ecosystem Service Research: Early Steps and Current Drivers. AMBIO 39:314–324. https://doi.org/10.1007/s13280-010-0048-x Visual Crossing (2020) Weather Data & API Global Forecast & History Data - Historical weather data for kabul and Paghman (2020) World Bank Group (2006) Why and how should Kabul upgrade its informal settlements (English). World Bank Group, Washington, D.C. Wu L, Sun C, Fan F (2022) Multi-criteria framework for identifying the trade-offs and synergies relationship of ecosystem services based on ecosystem services bundles. Ecol Indic 144:109453. https://doi.org/10.1016/j.ecolind.2022.109453 Xu N, Duan H (2024) Ecological Functional Zoning in Urban Fringe Areas Based on the Trade-Offs Between Ecological–Social Values in Ecosystem Services: A Case Study of Jiangning District, Nanjing. Land 13:1957. https://doi.org/10.3390/land13111957 Zanaga D, Van De Kerchove R, De Keersmaecker W, et al (2021) ESA WorldCover 10 m 2020 v100 Zaryab A, Nassery HR, Alijani F (2022) The effects of urbanization on the groundwater system of the Kabul shallow aquifers, Afghanistan. Hydrogeol J 30:429–443. https://doi.org/10.1007/s10040-021-02445-6 Tables Tables 1 and 2 are available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files Table1and2.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7130094","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":499470253,"identity":"9fb25ce6-56ee-452b-95f6-c05e0830f410","order_by":0,"name":"Mohammad Reza Ansari","email":"data:image/png;base64,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","orcid":"","institution":"Leibniz University Hannover","correspondingAuthor":true,"prefix":"","firstName":"Mohammad","middleName":"Reza","lastName":"Ansari","suffix":""},{"id":499470254,"identity":"6ec33d2a-3827-4176-8067-c31100336392","order_by":1,"name":"Maisam Rafiee","email":"","orcid":"","institution":"Leibniz University Hannover","correspondingAuthor":false,"prefix":"","firstName":"Maisam","middleName":"","lastName":"Rafiee","suffix":""},{"id":499470255,"identity":"8e694e75-0233-432d-b92f-990718bcf154","order_by":2,"name":"Blal Adem Esmail","email":"","orcid":"","institution":"Ruhr University Bochum","correspondingAuthor":false,"prefix":"","firstName":"Blal","middleName":"Adem","lastName":"Esmail","suffix":""},{"id":499470256,"identity":"40de9be9-b90c-40c5-89d1-0f326941bd54","order_by":3,"name":"Daniela Kempa","email":"","orcid":"","institution":"Leibniz University Hannover","correspondingAuthor":false,"prefix":"","firstName":"Daniela","middleName":"","lastName":"Kempa","suffix":""},{"id":499470258,"identity":"469ae526-7547-4cb4-8d9d-d702b9e9728a","order_by":4,"name":"Tinka Kuhn","email":"","orcid":"","institution":"Leibniz University Hannover","correspondingAuthor":false,"prefix":"","firstName":"Tinka","middleName":"","lastName":"Kuhn","suffix":""},{"id":499470259,"identity":"688f6a48-964c-486f-92cd-3d557fabc92d","order_by":5,"name":"Lisa Biber-Freudenberger","email":"","orcid":"","institution":"University of Bonn","correspondingAuthor":false,"prefix":"","firstName":"Lisa","middleName":"","lastName":"Biber-Freudenberger","suffix":""},{"id":499470260,"identity":"c7381322-a1dd-42e8-bd20-594492b05434","order_by":6,"name":"Christian Albert","email":"","orcid":"","institution":"Leibniz University Hannover","correspondingAuthor":false,"prefix":"","firstName":"Christian","middleName":"","lastName":"Albert","suffix":""}],"badges":[],"createdAt":"2025-07-15 11:38:09","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7130094/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7130094/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":89559016,"identity":"60080d6d-311c-44ae-979c-d91b9842665d","added_by":"auto","created_at":"2025-08-21 09:59:11","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":3537273,"visible":true,"origin":"","legend":"\u003cp\u003eStudy area and land-use/land-cover (LULC) map.\u003cstrong\u003e(a)\u003c/strong\u003e Location of the study area within Afghanistan, with the Kabul-Indus River Basin Zone highlighted in dark gray. \u003cstrong\u003e(b)\u003c/strong\u003e Study area overlaid on the Kabul-Indus River Basin Zone, showing provincial boundaries. (\u003cstrong\u003ec)\u003c/strong\u003e The detailed LULC map of the study area, developed by integrating the European Space Agency WorldCover 2020 product with Global Human Settlement Layer built-up height, OpenStreetMap data, Global Canopy Height, and Food and Agriculture Organization of United Nation’s crop type data, with legends illustrating various land uses including cultivation frequency (single or double) for specified crop types.\u003c/p\u003e","description":"","filename":"Picture1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7130094/v1/c4a7e74de34f85734fceee5b.jpg"},{"id":89559012,"identity":"d2edc299-275d-463b-851b-66b24beba1d8","added_by":"auto","created_at":"2025-08-21 09:59:11","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1163644,"visible":true,"origin":"","legend":"\u003cp\u003eThe conceptual framework of the research design, outlining the three main stages. Abbreviations are as follows: WY, water yield; HM, heat mitigation; SWRR, stormwater runoff retention; NA, nature access; HQ, habitat quality.\u003c/p\u003e","description":"","filename":"Picture2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7130094/v1/1ad2a25c121b17c63f81bef1.jpg"},{"id":89560019,"identity":"0dc52e4c-6d1a-4bf9-8497-b652ec7b955c","added_by":"auto","created_at":"2025-08-21 10:15:11","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1481561,"visible":true,"origin":"","legend":"\u003cp\u003ePopulation density zones within the Kabul metropolitan region, illustrating the five categories derived from the GHSL population grid and reclassified based on thresholds informed by the classifications in the 2019 Master Plan for Kabul city.\u003c/p\u003e","description":"","filename":"Picture3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7130094/v1/3f451c244bcb1616f1a85fdf.jpg"},{"id":89559021,"identity":"79211692-651a-4722-84a6-33d273e4e8b9","added_by":"auto","created_at":"2025-08-21 09:59:11","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":3384205,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial distribution of the supply potential for the five assessed ecosystem services in Kabul metropolitan region. The heat mitigation and habitat quality maps show a relative index from 0 to 1.\u003c/p\u003e","description":"","filename":"Picture4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7130094/v1/3d028c71a7eba4238952a8ea.jpg"},{"id":89559017,"identity":"66803a81-f5e2-4699-930c-28d5bafe783e","added_by":"auto","created_at":"2025-08-21 09:59:11","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":3476961,"visible":true,"origin":"","legend":"\u003cp\u003eHotspot and coldspot analysis for the five assessed ecosystem services in Kabul metropolitan region, showing statistically significant spatial clusters of high (hotspot) and low (coldspot) service provision based on the Getis-Ord Gi* statistic.\u003c/p\u003e","description":"","filename":"Picture5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7130094/v1/923fb3c8997430f1d47824fa.jpg"},{"id":89559292,"identity":"1b975015-6516-42df-b0b9-45f08d13c052","added_by":"auto","created_at":"2025-08-21 10:07:11","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":752490,"visible":true,"origin":"","legend":"\u003cp\u003ePercentage of the study area covered by statistically significant hotspots and coldspots for each ecosystem service at three confidence levels (99%, 95%, and 90%). For clarity, labels are not shown for areas covering less than 5% of the total study area. One percent of the study area corresponds to approximately 2,584 ha.\u003c/p\u003e","description":"","filename":"Picture6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7130094/v1/72050bf4dc413b2a35f5fb68.jpg"},{"id":89559018,"identity":"e8b252be-8883-420f-91b3-dfeffbc156cb","added_by":"auto","created_at":"2025-08-21 09:59:11","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":794859,"visible":true,"origin":"","legend":"\u003cp\u003eBoxplots showing the distribution of supply for each ecosystem service across the five population density zones, from unpopulated areas to high-density urban zones. The number of 1-hectare hexagonal grid cells (n) is indicated for each category. \u003cstrong\u003e(a)\u003c/strong\u003e water yield, \u003cstrong\u003e(b)\u003c/strong\u003e heat mitigation, \u003cstrong\u003e(c)\u003c/strong\u003e stormwater runoff retention, \u003cstrong\u003e(d)\u003c/strong\u003e nature access, and \u003cstrong\u003e(e)\u003c/strong\u003ehabitat quality\u003c/p\u003e","description":"","filename":"Picture7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7130094/v1/04b8d3de4d804dc79a677627.jpg"},{"id":89560018,"identity":"bd8aa6c0-4385-451b-8cf5-114e7786bd45","added_by":"auto","created_at":"2025-08-21 10:15:11","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":1201905,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial distribution of the three identified ecosystem service bundles across the Kabul metropolitan region using k-mean clustering: Low-Service Zone, Regulating-Service Zone, and High-Service Multifunctional Zone.\u003c/p\u003e","description":"","filename":"Picture8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7130094/v1/fb5e11344c27dd060c383633.jpg"},{"id":89559014,"identity":"68b96416-0388-4f29-9564-68275a4e1bb0","added_by":"auto","created_at":"2025-08-21 09:59:11","extension":"jpg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":574710,"visible":true,"origin":"","legend":"\u003cp\u003eSpearman's rank correlation matrices showing synergies (positive correlations) and trade-offs (negative correlations) between the five ecosystem services within each of the three bundles: \u003cstrong\u003e(a)\u003c/strong\u003e Low-Service Zone, \u003cstrong\u003e(b)\u003c/strong\u003e Regulating-Service Zone, and \u003cstrong\u003e(c)\u003c/strong\u003e High-Service Multifunctional Zone.\u003c/p\u003e","description":"","filename":"Picture9.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7130094/v1/928c860ea250b486c901ada3.jpg"},{"id":89559290,"identity":"c77db4fa-f713-4b33-bc41-ab83bc4be71f","added_by":"auto","created_at":"2025-08-21 10:07:11","extension":"jpg","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":2741341,"visible":true,"origin":"","legend":"\u003cp\u003eMap of recommended conservation priority areas in Kabul metropolitan region based on the provision of multiple ecosystem services. Priority levels are determined by the number of services supplied above the area-wide mean, with magnitude sub-categories indicating the level of supply relative to the standard deviation.\u003c/p\u003e","description":"","filename":"Picture10.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7130094/v1/25b84666f0eabaf97921ed0e.jpg"},{"id":94043287,"identity":"c55cbb79-07ec-4081-a78e-048fcef9523b","added_by":"auto","created_at":"2025-10-21 19:31:26","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":20077624,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7130094/v1/aba6934b-7359-4838-9539-20eaec6926f3.pdf"},{"id":89559013,"identity":"457cee09-87f1-4652-96dd-bc51142b5613","added_by":"auto","created_at":"2025-08-21 09:59:11","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":28414,"visible":true,"origin":"","legend":"","description":"","filename":"Table1and2.docx","url":"https://assets-eu.researchsquare.com/files/rs-7130094/v1/b231a30b4c05036d8202f54b.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Mapping and assessing ecosystem service hotspots and bundles in rapidly urbanizing, data-scarce regions: a case study of Kabul, Afghanistan","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe mapping and assessment of ecosystems and their services (MAES) is increasingly recognized as fundamental to sustainable spatial planning and environmental governance, providing critical tools to harmonize urban development with ecological preservation (e.g., Albert et al. 2014, 2016; Geneletti 2016; Maes et al. 2012). Ecosystem services (ESs)\u0026mdash;the manifold benefits humans derive from healthy functioning ecosystems, such as water supply, climate regulation, flood mitigation, and recreational opportunities (Costanza et al. 1997; Daily 2010; Millennium Ecosystem Assessment 2005)\u0026mdash;are pivotal for urban resilience (e.g., McPhearson et al. 2015) and for enhancing the quality of life in increasingly urbanized landscapes (e.g., Bolund and Hunhammar 1999; Kowarik et al. 2017). While ecosystem service (ES) research has expanded significantly, robust spatially explicit ES assessments remain crucial for translating conceptual understanding into actionable planning strategies.\u003c/p\u003e\n\u003cp\u003eA key challenge in ES assessment is understanding their spatial heterogeneity\u0026mdash;where services are abundant, where they are scarce, and how they co-vary. Methodologies such as hotspot and coldspot analysis (e.g., Getis-Ord Gi*) have become instrumental in identifying statistically significant clusters of high and low ES provision, thereby guiding targeted interventions (Adem Esmail et al. 2025; Bagstad et al. 2017). Furthermore, recognizing that ESs rarely occur in isolation, the concept of ES bundles\u0026mdash;recurrent sets of associated ESs\u0026mdash;has emerged as a powerful analytical framework for understanding landscape multifunctionality and identifying synergies and trade-offs (Queiroz et al. 2015; Raudsepp-Hearne et al. 2010). Such analyses are vital for developing integrated conservation and restoration plans that optimize multiple benefits and address complex environmental challenges (Spake et al. 2017; Wu et al. 2022).\u003c/p\u003e\n\u003cp\u003eDespite these methodological advancements, ES research has predominantly remained concentrated in the Global North, leaving the Global South underexplored (Gangahagedara et al. 2021; Seppelt et al. 2011; Vihervaara et al. 2010). These areas often face the dual challenge of intense ecological pressure from urban expansion and severe data scarcity, hindering the effective integration of ES information into planning processes (e.g., Mahendra et al. 2021; e.g., Shackleton et al. 2021). The Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) regional assessment report on biodiversity and ESs for Asia and the Pacific (2018) underscores this imbalance, identifying urbanization, climate change, and habitat degradation as primary threats, yet noting a deficit in localized, high-resolution ES assessments that can effectively inform policy in these dynamic contexts. While studies in the Global South have begun to address these issues (Adem Esmail et al. 2023, 2025; Akhtar et al. 2022; Najmuddin et al. 2022; Srivathsa et al. 2023), many rely on coarser-resolution data that may not capture the fine-scale landscape heterogeneity critical for urban planning, especially in areas with significant informal development.\u003c/p\u003e\n\u003cp\u003eThe Kabul metropolitan region (KMR), Afghanistan, exemplifies these challenges. Characterized by swift, largely informal urban expansion over the past two decades\u0026mdash;its population doubling with approximately 70% of this growth occurring outside formal planning (Chaturvedi et al. 2020; NSIA 2004, 2023; UN Habitat 2015)\u0026mdash;KMR faces escalating environmental hazards such as heatwaves, flooding, and resource scarcity. This context of rapid, under-documented landscape transformation, coupled with profound data limitations (NEPA 2014), makes KMR a critical yet understudied area for applied ES research. The dominance of informal settlements presents a distinctive opportunity to examine ES provision in such settings, addressing a significant gap in regional and global research and offering insights applicable to comparable urbanizing areas.\u003c/p\u003e\n\u003cp\u003eThe aim of this study is to conduct a high-resolution, spatially explicit assessment of key ESs in KMR to understand their distribution, interactions, and implications for urban and conservation planning. Specifically, this research undertakes a detailed analysis of five key ESs\u0026mdash;water yield (WY), heat mitigation (HM), stormwater runoff retention (SWRR), nature access (NA), and habitat quality (HQ)\u0026mdash;employing a multi-source data approach and established modeling techniques to overcome common data limitations. The detailed assessment aims to uncover variations often missed by coarser assessments, thereby facilitating the identification of priority areas for conserving important multi-ES supply zones, and informing strategic green infrastructure development to enhance urban resilience and livability. This research addresses three core questions:\u003c/p\u003e\n\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003eHow are the selected ESs spatially distributed in the KMR, where are their spatial hotspots and coldspots located, and how does ES provision relate to human population density gradients?\u003c/li\u003e\n \u003cli\u003eWhere across the study area do distinct ES bundles (i.e., areas with similar characteristics of ES supply) occur, and what patterns of synergies and trade-offs do these spatially defined bundles reveal?\u003c/li\u003e\n \u003cli\u003eWhich zones provide multiple ES benefits requiring protection?\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eBy tackling these questions, this study delivers actionable insights for KMR\u0026rsquo;s planners and policymakers to promote sustainable urban development. It further contributes to the broader ES literature by demonstrating the application of integrated assessment methodologies in a data-scarce, rapidly urbanizing, and informally developing context, offering a framework that can inform planning in similar cities across the Global South and supporting the IPBES call for enhanced biodiversity and ES management strategies.\u003c/p\u003e\n\u003ch2\u003eStudy area\u003c/h2\u003e\n\u003cp\u003eThe study area is located within the Kabul-Indus River Basin Zone in eastern Afghanistan, focusing on an approximately 2,584 km\u0026sup2; region defined by latitudes 34\u0026deg;16\u0026prime;N to 34\u0026deg;40\u0026prime;N and longitudes 68\u0026deg;52\u0026prime;E to 69\u0026deg;31\u0026prime;E (Fig. 1a, b). This strategically demarcated area encompasses Kabul City, the capital and most populous urban center of Afghanistan, alongside its adjacent peri-urban and surrounding \u003cem\u003eWoloswali\u0026nbsp;\u003c/em\u003e(provincial districts). These include Bagrami, Musayi, and Chahar Asyab in their entirety, as well as portions of Nirkh, Maidan Shahr, Paghman, Surkhi Parsa, Shakardara, Dih Sabz, Kohi Safi, Surobi, Khaki Jabbar, and Mohammad Agha.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe KMR has been purposefully selected as the focal point of this research due to its pivotal role as the primary hub of Afghanistan\u0026apos;s accelerating urban transition. Currently, Kabul City accommodates 54% of the nation\u0026apos;s urban population (Ellis and Roberts 2016). This urban primacy is further underscored by the city\u0026apos;s remarkable demographic growth, with its population more than doubling from 2.8 million in 2004 to 5.8 million in 2023 (National Statistics and Information Authority 2004, 2023). Projections indicate this rapid trend will continue, with Afghanistan\u0026apos;s urban population expected to rapidly grow and reach half of the country\u0026apos;s total population by 2060 (UN Habitat 2015). Notably, Kabul city was recognized as world\u0026apos;s fifth fastest-growing city in 2015 (City Mayors 2015).\u003c/p\u003e\n\u003cp\u003eA critical characteristic of KMR\u0026apos;s urban expansion is its largely informal nature, with informal settlements constituting approximately 70% of residential areas (World Bank Group 2006; UN Habitat 2015) and providing shelter for an estimated 80% of the city\u0026apos;s residents (Shahraki et al. 2020). This pattern of informal development has been a major driver of substantial land-use/land-cover (LULC) changes across the metropolitan region over the past two decades (DeWitt et al. 2022; Najmuddin et al. 2018, 2022; Zaryab et al. 2022). Consequently, the area is facing a confluence of intensifying environmental challenges, including:\u003c/p\u003e\n\u003cul class=\"decimal_type\"\u003e\n \u003cli\u003eA significant increase in land surface temperatures (LST), with studies indicating a rise of +1.45\u0026deg;C to +3.78\u0026deg;C between 2009 and 2019 (Mahmoodi et al. 2019);\u003c/li\u003e\n \u003cli\u003eRapid depletion of groundwater resources, estimated at an average decline of 0.8 meters per year in the Kabul Plain, with localized areas experiencing drastic drops of up to 60 meters due to over-extraction (Zaryab et al. 2022);\u003c/li\u003e\n \u003cli\u003eA discernible upward trend in aerosol optical depth levels from 2000 to 2022, particularly during the spring and summer seasons, indicating worsening air quality (Torabi et al. 2024);\u003c/li\u003e\n \u003cli\u003eIncreasing pressure on water resources due to rising demand from the growing population, coupled with declining rates of groundwater recharge (Gauster 2021; Taher et al. 2014).\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eGiven the synergistic effect of rapid and largely informal urbanization and the associated escalating environmental stressors, the KMR presents a highly relevant and crucial case study for a detailed investigation into the current state of ES supply within a data-scarce and rapidly transforming urban environment. Understanding the spatial patterns and interrelationships of these services in KMR can yield critical insights for informing sustainable urban planning strategies and effective ES management interventions tailored to the specific challenges of this region and potentially applicable to other similar contexts in the Global South.\u003c/p\u003e"},{"header":"Material and methods","content":"\u003ch2\u003eResearch design\u003c/h2\u003e\n\u003cp\u003eOver time, a range of conceptual frameworks and tools have emerged, including the Millennium Ecosystem Assessment (MEA; 2005), The Economics of Ecosystems and Biodiversity (TEEB; 2010), the Common International Classification of Ecosystem Services (CICES; 2012), and IPBES (2019). MAES methodological frameworks in particular provide structured approaches to identifying, assessing, quantifying, and valuing ESs. They guide how ESs are linked to human well-being, spatial patterns, trade-offs, and policy decisions (Albert et al. 2014; De Groot et al. 2010; Haines-Young and Potschin 2010). While specific methodological procedures vary depending on the research questions and context, many ES studies follow a general structure that includes three interlinked stages: selection of ESs, assessment and quantification of their supply, and application of the results to inform planning and decision-making. Some studies focus on identifying ES hotspots and coldspots to guide conservation or inform sustainable urban planning (e.g., Adem Esmail et al. 2025; Li et al. 2017), while others delve into understanding the complex interrelationships between services by analyzing ES bundles and their trade-offs and synergies (e.g., Karimi et al. 2021). Furthermore, many studies aim to translate these analytical outputs into practical applications, such as delineating ecological functional zones or identifying priority areas for management interventions (e.g., Deng and Cao 2023; Xu and Duan 2024).\u003c/p\u003e\n\u003cp\u003eIn this study, we structured our methodological framework accordingly into three stages: selection of ESs, mapping and assessment, and spatial-statistical analysis for application (Fig.2).\u003c/p\u003e\n\u003cp\u003eSelection of ES: The selection was guided by the IPBES classification of Nature\u0026rsquo;s Contribution to People (NCP) to ensure representation across material, regulating, and non-material contributions. We selected five services: WY, HM, SWRR, NA, and HQ. This selection ensured the inclusion of diverse functions relevant to urban resilience and planning.\u003c/p\u003e\n\u003cp\u003eES mapping and assessment: The core of the methodological framework involved the biophysical quantification of ESs using the InVEST\u003csup\u003e\u0026reg;[1]\u003c/sup\u003e (Integrated Valuation of Ecosystem Services and Tradeoffs) suite of models (Natural Capital Project 2024). InVEST has become a standard tool in ES research because it is user-friendly, simple, focuses on important ESs, and employs a multi-purpose, peer-reviewed methodology (Nemec and Raudsepp-Hearne 2013). Model outputs were then processed in ArcGIS Pro to generate spatially explicit layers of ES supply. These maps enabled the identification of spatial distribution patterns and are the backbone for the subsequent statistical analysis.\u003c/p\u003e\n\u003cp\u003eSpatial and statistical analysis for application: We conducted a series of spatial-statistical analyses to translate the results into planning-relevant insights. The Getis-Ord Gi* statistic was applied to detect statistically significant clusters of high and low ES supply (Adem Esmail et al. 2025; Li et al. 2017). The k-means clustering was used in R (R Core Team 2024) to identify areas with similar ES profiles, aiding in the recognition of multifunctional zones (Queiroz et al. 2015; Raudsepp-Hearne et al. 2010). Next, to explore interrelationships among services, we used Spearman\u0026apos;s rank correlation (Spearman 1904), a robust method for non-parametric association analysis (Li et al. 2024). And finally, to support targeted policy and intervention recommendations, a composite conservation priority map was generated by aggregating areas of high ES supply using mean and standard deviation (SD) thresholds.\u003c/p\u003e\n\u003cp\u003eOverall, our approach follows a structured methodological logic: grounded in conceptual frameworks (IPBES), operationalized through biophysical modeling (InVEST), visualized through spatial analysis (ArcGIS Pro), interpreted statistically (R), and concluded with spatial prioritization (Fig. 2). Overall, this structured methodology leverages advanced geospatial techniques and modeling to generate actionable insights despite data limitations, aiming to support evidence-based decision-making for KMR\u0026apos;s policymakers. The following sections explain each of these stages in greater detail.\u003c/p\u003e\n\u003cp\u003eData sources\u003c/p\u003e\n\u003cp\u003eA comprehensive and multi-faceted dataset was compiled to support this study, encompassing satellite imagery, global repositories, and regional institutions. These data were crucial for the creation of a high-resolution LULC map (Fig. 1c) and the subsequent modeling of key ESs.\u003c/p\u003e\n\u003cp\u003eData sources for land-use/land-cover creation\u003c/p\u003e\n\u003cp\u003eThe foundation for our LULC mapping was the 10m resolution ESA WorldCover 2020 product (Zanaga et al. 2021). This base layer was enhanced by integrating supplementary datasets to produce a detailed 19-class LULC map (Fig. 1c). Key additions included crop-type classifications from Food and Agriculture Organization of the United Nations (FAO; 2020), Sentinel-2-derived canopy height (Lang et al. 2022), OpenStreetMap (OSM) infrastructure data (Geofabrik 2024), and built-up height information from the Global Human Settlement Layer (GHSL; Pesaresi 2023). These datasets were essential for accurately representing the urban and peri-urban landscape of KMR. A complete list of data sources used in the LULC creation process, including their resolution and year, is provided in Table 1.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eData sources for ecosystem service assessment\u003c/p\u003e\n\u003cp\u003eThe spatial assessment of the five selected ESs relied on a diverse array of biophysical parameters and socio-demographic data. Critical climate inputs included monthly precipitation and temperature data from Fick and Hijmans (2017), daily temperature from Visual Crossing (2020), and extraterrestrial radiation from FAO (1998). Soil characteristics were obtained from International Soil Reference and Information Centre (ISRIC) SoilGrids (Hengl et al. 2017; ISRIC 2017), while topographical data were provided by the United States Geological Survey (USGS; 2018) and watershed boundaries by HydroSHEDS (Lehner and Grill 2013). Additional datasets included land surface albedo from RSLab (2020), rainfall intensity-duration-frequency data from the U.S. Army Corps of Engineers (USACE; 2010), hydrologic soil-cover complexes from U.S. Department of Agriculture-Natural Resources Conservation Service (USDA-NRCS; 2004), and soil hydrologic groups from Simons et al. (2020). For assessing NA, the analysis incorporated LULC naturalness data from Veteikis et al. (2011), GHSL population grids (Schiavina et al. 2023), and socio-demographic statistics from the Afghanistan Central Statistics Organization (CSO). HQ assessment used threat and sensitivity tables from Sallustio et al. (2017). A comprehensive list of all datasets used for the ES assessment, including their specific purposes and sources, is detailed in Table 2. For detailed descriptions of the ES models used, readers are referred to the InVEST user manuals (Natural Capital Project 2024).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eEcosystem service selection\u003c/p\u003e\n\u003cp\u003eThe selection of ESs for this study is guided by IPBES\u0026rsquo;s NCP framework, which emphasizes the diverse ways ecosystems support human well-being in specific socio-ecological contexts (D\u0026iacute;az et al. 2018). The NCP classification integrates cultural, regulating, and material contributions, thereby fostering more effective and equitable decision-making for sustainable futures (Pascual et al. 2017). Five ESs were selected based on their relevance to the KMR\u0026rsquo;s rapid urbanization, climate vulnerabilities, and data availability, informed by expert consultations and a review of regional environmental pressures. These selections align with the IPBES emphasis on context-specific contributions (D\u0026iacute;az et al. 2018) and address KMR\u0026rsquo;s pressing needs for resource security, climate adaptation, and livability. Five selected ESs are:\u003c/p\u003e\n\u003cp\u003eWY: This service is formally classified under the regulating contribution of \u0026ldquo;Regulation of freshwater quantity, location and timing.\u0026rdquo; However, our study focuses specifically on the tangible output of this process. Given that our assessment prioritizes the provision of water as a consumable good, we have classified it as a material contribution, an approach consistent with other recent studies (e.g., K\u0026uuml;lling et al. 2024). This approach better reflects its critical importance as a finite material resource for human consumption, agriculture, industry, and energy production, particularly within the arid climate and high-demand context of the KMR.\u003c/p\u003e\n\u003cp\u003eHM: Classified as a regulating contribution, this service directly relates to the NCP of \u0026ldquo;Regulation of climate.\u0026rdquo; In the KMR, this function is vital for mitigating urban heat island effects at a local scale, a problem exacerbated by high building density and informal urban sprawl.\u003c/p\u003e\n\u003cp\u003eSWRR: This is a regulating contribution corresponding to the NCP \u0026ldquo;Regulation of hazards and extreme events.\u0026rdquo; It is crucial for flood risk reduction in a city prone to heavy seasonal rains and characterized by areas with poor drainage infrastructure, thereby protecting property and lives.\u003c/p\u003e\n\u003cp\u003eNA: This represents a non-material contribution, specifically the NCP of \u0026ldquo;Physical and psychological experiences.\u0026rdquo; These include the recreational, cultural, and mental health benefits that residents of the densely populated KMR derive from interacting with urban green spaces.\u003c/p\u003e\n\u003cp\u003eHQ: As a measure of the NCP \u0026ldquo;Habitat creation and maintenance,\u0026rdquo; this is a core regulating contribution. While officially a regulating service, it is recognized as being foundational to the delivery of almost all other NCPs. By supporting local biodiversity, high HQ is essential for ensuring the long-term resilience and continued functioning of the KMR\u0026rsquo;s ecosystems.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMapping and assessment of ecosystem service supply potential\u003c/p\u003e\n\u003cp\u003eTo map ES supply potential, this study integrates multi-source data to create a high-resolution (10m) LULC map of the KMR (Table 1; Fig. 1b). The LULC map is then imported into InVEST along with other respected biophysical elements (Table 2) to quantify the supply of the five selected ESs. While comprehensive documentation of the ES models is available in the InVEST user manuals (Natural Capital Project 2024), we briefly describe the ES models applied in this study:\u003c/p\u003e\n\u003cp\u003eWY was calculated using the InVEST annual water yield model, which estimates spatial variations in water provision by analyzing precipitation and evapotranspiration based on the Budyko curve. The model output represents the annual WY per pixel.\u003c/p\u003e\n\u003cp\u003eHM was calculated using the InVEST urban cooling model, which assesses the capacity of urban landscapes to mitigate heat by quantifying cooling services from shade, evapotranspiration, and albedo. The output is a HM index for each pixel.\u003c/p\u003e\n\u003cp\u003eSWRR was calculated using the InVEST urban flood risk mitigation model, which quantifies the influence of land use and soil types on runoff via the Curve Number (CN) method. The model estimates the annual volume of runoff avoided (retained) by the current land cover relative to a reference (impervious) scenario, with values provided per pixel.\u003c/p\u003e\n\u003cp\u003eNA was calculated using the InVEST urban nature access model, which evaluates the availability of urban natural spaces for recreation via the Two-Step Floating Catchment Area (2SFCA) method. The model output quantifies the supply of nature space per population pixel.\u003c/p\u003e\n\u003cp\u003eHQ was calculated using the InVEST habitat quality model, which assesses biodiversity by evaluating HQ and rarity based on LULC, and threats. Inputs comprised the LULC map, a threat table, and a sensitivity table. The model calculates a HQ score for each parcel.\u003c/p\u003e\n\u003cp\u003eIt is important to note that for the ESs selected in this study, the InVEST model provides assessment results in two distinct types. First, where biophysical parameters allow, it yields results in quantifiable units, facilitating comparisons across different case studies (e.g., WY in mm/m\u0026sup2; per year, NA in m\u0026sup2;/capita per pixel, and SWRR in m\u0026sup3;/pixel). Second, for other services like HM, and HQ, InVEST generates relative indices, typically scaled between 0 and 1. These indices are invaluable for intra-study comparisons, allowing for the identification of areas with relatively higher or lower service provision within the study area. For instance, a HM index of 0.2 in one location and 0.4 in another indicates that the latter has twice the relative capacity for HM compared to the former. While absolute, unit-based values would ideally allow for direct comparisons with other urban areas globally, the primary objective of this study is to delineate the spatial variations and identify critical areas of high and low ES supply within KMR itself. Therefore, these relative indices serve as robust indicators for achieving this aim, highlighting areas requiring targeted interventions.\u003c/p\u003e\n\u003cp\u003eMapping and analyzing hotspots and coldspots of ecosystem service supply\u003c/p\u003e\n\u003cp\u003eFor mapping and analyzing hotspots and coldspots of ES (potential) supply, we resampled the original 10m x 10m square grid to a regular hexagonal grid with 1-hectare cells. Hexagonal grids provide more uniform target-neighbor relationships, as the nearest neighborhood definition is simpler and less ambiguous than in rectangular grids, which enhances the robustness of spatial autocorrelation results, particularly when connectivity or movement paths are relevant (Birch et al. 2007; Esri).\u003c/p\u003e\n\u003cp\u003eSpatial patterns of ES supply were analyzed using hotspot and coldspot detection techniques. The Getis-Ord Gi* Statistics \u0026nbsp;(Ord and Getis 1995) was applied to each ES layer to identify statistically significant clusters of high (hotspots) and low (coldspots) supply. This method calculates z-scores and p-values for each grid cell, highlighting areas where ES provision deviates from random distribution. The Getis-Ord Gi* Statistics (Ord and Getis 1995) was conducted in ArcGIS Pro based on a contiguity edges corner conceptualization of spatial relationships.\u003c/p\u003e\n\u003cp\u003eEcosystem service supply across population density gradients\u003c/p\u003e\n\u003cp\u003eTo extend the analysis beyond landscape composition and explore the direct interface with settlement patterns, we further investigated the relationship between ES provision and human population density. This analysis utilized the GHSL population grid (Schiavina et al. 2023). The continuous population data from this grid was reclassified into five discrete zones (Fig. 3) based on density thresholds informed by the classifications in the 2019 Master Plan for Kabul city (Sasaki 2019). The density zones are defined as follows: high population density (more than 250 people/hectare), medium population density (50\u0026ndash;250 people/ha), low population density (5\u0026ndash;50 people/ha), very low population density (1\u0026ndash;5 people/ha), and unpopulated areas (0 people/ha).\u003c/p\u003e\n\u003cp\u003eFollowing the reclassification, Zonal Statistics in ArcGIS Pro was employed to calculate the mean and total supply of each of the five selected ESs (WY, HM, SWRR, NA, and HQ) within each of the five population density zones. This approach allows for a direct comparison of ES supply potential across a gradient of urbanization and settlement intensity.\u003c/p\u003e\n\u003ch2\u003eIdentification of ecosystem service bundles and interrelationships\u003c/h2\u003e\n\u003cp\u003eThis step examines the interrelationships among the five ESs to identify distinct zones with similar supply profiles, known as ES bundles, and to detect synergies and trade-offs.\u003c/p\u003e\n\u003cp\u003eTo identify ES bundles, we applied k-means clustering algorithm (Hartigan and Wong 1979) to the standardized supply maps of the five ESs, grouping areas based on their multi-ES characteristics (Sun et al. 2022; Wu et al. 2022). The dataset includes 258,394 regular 1-hectare hexagonal grids. The optimal number of clusters (k) was determined using both the elbow and average silhouette width methods on a 10% random sample of the dataset. Based on a clear inflection point in the elbow plot and a peak average silhouette score, k=3 was selected as the optimal value. The final three-bundle groups were then generated by applying the algorithm to the full dataset. The entire clustering process was conducted in R (R Core Team 2024) using functions from the cluster (Maechler et al. 2025) and ClusterR (Mouselimis 2024) packages.\u003c/p\u003e\n\u003cp\u003eSubsequently, to quantify the interrelationships, a Spearman\u0026rsquo;s rank correlation analysis (Spearman 1904) was performed. This analysis identified pairwise positive (synergies) or negative (trade-offs) associations between the five ESs. These calculations were carried out using the rcorr function from the Hmisc package (Harrell Jr 2025).\u003c/p\u003e\n\u003cp\u003eThese analyses reveal how ESs co-occur or conflict spatially, informing targeted planning strategies in KMR\u0026rsquo;s heterogeneous urban landscape. Identifying these ES bundles helps delineate ecologically distinct zones characterized by specific combinations of high and low service provision (e.g., Mouchet et al. 2014). Recognizing these spatial patterns of multifunctionality is crucial for tailoring management interventions (e.g., Queiroz et al. 2015; Spake et al. 2017); for example, understanding the typical ES profile of different zones can inform the strategic placement and type of nature-based solutions or green infrastructure projects.\u003c/p\u003e\n\u003ch2\u003eMapping recommended areas for conservation\u003c/h2\u003e\n\u003cp\u003ePriority areas for conservation were identified based on the potential supply levels of the five assessed ESs (WY, HM, SWRR, NA, HQ). A two-step classification approach was used to identify areas providing multiple benefits. First, a primary classification based on the mean supply of each ES across the study area was performed. For each spatial unit (1 hectare hexagon), the number of ESs providing supply levels \u003cem\u003eabove\u003c/em\u003e their respective area-wide mean was counted. This count determined the primary conservation priority category.\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003eHighest priority: areas where all five ESs are above the mean supply;\u003c/li\u003e\n \u003cli\u003eHigh priority: areas with four ESs above the mean;\u003c/li\u003e\n \u003cli\u003eMedium priority: areas with three ESs above the mean;\u003c/li\u003e\n \u003cli\u003eLow priority: areas with two ESs above the mean; and\u003c/li\u003e\n \u003cli\u003eLowest priority: areas where only one ES is above the mean supply.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eSecond, these primary priority categories were refined using SD thresholds to assess the magnitude of ES supply relative to the mean. Spatial units were further sub-categorized based on how far the supply of contributing ESs exceeded the mean: High goes to locations where relevant ES supply is substantially above the mean (e.g., \u0026gt;1 SD above the mean); Moderate are locations where relevant ES supply is moderately above the mean (e.g., between 0.5 and 1 SD above the mean); and Low represent the areas where relevant ES supply is slightly above the mean (e.g., \u0026lt;0.5 SD above the mean).\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003e[1] https://naturalcapitalproject.stanford.edu/software/invest\u003c/p\u003e"},{"header":"Results","content":"\u003ch2\u003eEcosystem service\u0026nbsp;assessment\u003c/h2\u003e\n\u003cp\u003eOur analysis revealed a significant spatial heterogeneity in the supply of ESs across the KMR, with distinct patterns of high and low provision (Fig. 4) and different contributions from areas within and outside the Kabul municipality (Table 3). High supply areas are predominantly located in the more natural and less disturbed landscapes in the northwestern, eastern, and southeastern parts of the study area, while low supply areas are concentrated within the central, densely urbanized core (Fig. 4). Overall, areas outside the Kabul municipality account for approximately 73% of the total aggregated ES potential supply in the study area, while areas within Kabul municipality contribute the remaining 27%.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 3\u0026nbsp;\u003c/strong\u003eSummary statistics of ecosystem service supply within and outside the Kabul municipality\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"624\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003eEcosystem Service\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eStatistic\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003eOverall Study Area\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003eWithin Kabul Municipality\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003eOutside the Municipality\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003eUnits/Range\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003eWater yield\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e2.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e1.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e2.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003emm/m\u0026sup2; per year\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eSD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e1.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e1.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e2.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eSum (million)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e5.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e1.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e4.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003em\u003csup\u003e3\u003c/sup\u003e/year\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003eHeat mitigation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e0.47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e0.40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e0.51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003eRelative Index\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eSD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e0.28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e0.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e0.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003eStormwater runoff retention\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e2.10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e2.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e2.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003em\u003csup\u003e3\u003c/sup\u003e/pixel\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eSD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e0.59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e0.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e0.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eSum (million)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e66.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e25.58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e40.74\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003em\u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003eNature access\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eMean (thousand)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e243.10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e90.84\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e345.59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003em\u0026sup2;/capita per pixel\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eSD (thousand)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e282.55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e181.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e384.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003eRange: 0.00 -2.95M\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eSum (trillion)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e7.66\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e1.15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e6.51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003eHabitat quality\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e0.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e0.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e0.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003eRelative Index\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 24.2376%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.2488%;\"\u003e\n \u003cp\u003eSD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0883%;\"\u003e\n \u003cp\u003e0.26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e0.15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6437%;\"\u003e\n \u003cp\u003e0.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 18.138%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe relative indices for heat mitigation and habitat quality were scaled from 0 to 1, allowing for a comparison of relative service provision within the study area. Stormwater runoff retention values were calculated relative to a reference scenario (impervious surface). All analyses and mapping were conducted at a spatial resolution of approximately 10 m x 10 m. The abbreviation SD represents the standard deviation.\u003c/p\u003e\n\u003cp\u003eThe material contribution, represented by WY, is dominated by landscapes outside the Kabul municipality. These areas contribute nearly three-quarters of the total water supply and exhibit a significantly higher average yield (1.8 times) compared to areas within the municipality (Table 3; Fig. 4a).\u003c/p\u003e\n\u003cp\u003eThe supply of regulating contributions (HM, SWRR, and HQ) varies. For HM and SWRR, the supply is more balanced, with the average supply in areas outside the municipality being approximately 28% higher for HM and the total retention for SWRR being 1.6 times greater (Table 3; Fig 4b, 4c). In contrast, HQ shows a more pronounced spatial disparity, with the vast majority of high-quality habitat areas found outside the municipality (Table 3; Fig. 4e).\u003c/p\u003e\n\u003cp\u003eThe non-material contribution, represented by NA, also shows a significant spatial disparity. The vast majority of this service is supplied by areas outside the municipality, which provide around 85% of nature access (Table 3; Fig. 4d).\u003c/p\u003e\n\u003cp\u003eOverall, the spatial patterns of ES supply are strongly correlated with the underlying LULC distribution. Areas with higher values for regulating contributions (HM, SWRR, HQ) predominantly overlap with less disturbed land covers (e.g., tree cover, shrubland, grassland) found largely outside the Kabul municipality. Material (WY) and non-material (NA) contributions also show higher values in these less urbanized zones. Conversely, areas within Kabul municipality consistently exhibited lower values across most assessed ESs, highlighting the trade-offs associated with dense urban development and landscape modification. While localized green spaces within the municipal boundary provide some benefits, their extent appears insufficient to counteract the broader impacts of urbanization on ES provision across this zone.\u003c/p\u003e\n\u003ch2\u003eHotspot and coldspot analysis of ecosystem services\u003c/h2\u003e\n\u003cp\u003eStatistically significant spatial clustering of high (hotspots) and low (coldspots) ES provision revealed distinct patterns for each of the five assessed ESs across the KMR (Fig. 5). These patterns, varying notably between areas within Kabul municipality and areas outside the Kabul municipality, reflect the influence of underlying environmental and anthropogenic factors (Fig. 5; Fig. 6).\u003c/p\u003e\n\u003cp\u003eSpatial analysis of the material contribution, WY, revealed significant clustering. Hotspots of high WY (99% confidence) covered 13% of the study area (33,405 ha; Fig. 6) and were predominantly located in areas outside the Kabul municipality (Fig. 5a), reflecting higher elevation characteristics. Conversely, coldspots of low WY (99% confidence) covered a smaller portion of the study area (7%; Fig. 6) but were significantly concentrated within the central, urbanized parts of the municipality (Fig. 5a).\u003c/p\u003e\n\u003cp\u003eRegulating contributions showed a clear pattern of co-location and urban impact. For both HM and SWRR, hotspots of high supply covered a similar extent (Fig. 6)\u0026mdash;21% of the study area (54,959 ha and 54,304 ha, respectively)\u0026mdash;and were mainly concentrated in the less developed areas outside and on the fringes of the municipality (Fig. 5b, 5c). In stark contrast, coldspots were strongly aligned with the urban core. This was particularly evident for HM, where coldspots covered 24% of the study area (62,600 ha; Fig. 6), highlighting the significant impact of impervious surfaces on creating urban heat islands. For HQ, an interesting pattern emerged: while no statistically significant coldspots were identified at the highest confidence levels (99% or 95%), vast coldspots emerged at a lower confidence level (90%), covering an extensive 42% of the study area (Fig. 6). These were overwhelmingly concentrated within the Kabul municipality (Fig. 5e).\u003c/p\u003e\n\u003cp\u003eFor the non-material contribution, NA, a similar pattern of widespread deficit was observed. No statistically significant coldspots were identified at the highest confidence levels (99% or 95%), but at a lower confidence level (90%), vast coldspots emerged, covering 27% of the study area (Fig. 6). These low-confidence coldspots were also overwhelmingly concentrated within the Kabul municipality, covering 54% of its area (Fig. 5d).\u003c/p\u003e\n\u003cp\u003eThese spatially distinct clusters underscore the profound influence of LULC patterns on the heterogeneity of ES provision, indicating a widespread and pervasive deficit in accessible nature and quality habitat across the urban landscape, rather than isolated problem areas. Areas within Kabul municipality consistently exhibited lower ES values and higher concentrations of coldspots across most services, reflecting the impacts of urbanization. Conversely, areas outside the Kabul municipality contributed disproportionately to ES hotspots, aligning with the spatial gradients observed in the previous section.\u003c/p\u003e\n\u003ch2\u003eEcosystem service supply across population density gradient\u003c/h2\u003e\n\u003cp\u003eThe analysis of ES supply across population density gradients reveals a general inverse relationship, where service provision is highest in unpopulated areas and decreases with rising population density (Fig. 7). However, this trend manifests with different magnitudes and patterns across the services. The relationship is strong and direct for some regulating services, while for others it is less pronounced or exhibits more complex patterns. This disparity is starkly evident as unpopulated areas, covering 83% of the landscape, provide a disproportionately high share of the total supply for most services, including 92% of WY, 87% of HM, 97% of HQ, and nearly all NA (99.9%). The only exception is SWRR, where supply is largely proportional to land area across all zones.\u003c/p\u003e\n\u003cp\u003eThe trend of decreasing supply with increasing population density is most pronounced for the regulating contributions of HM and SWRR, which both show a clear and steady decline with rising population density (Fig. 7b, 7c). In contrast, the relationship is less direct for the material contribution (WY). While its mean supply is highest in unpopulated areas (approx. 2.5 mm/m\u0026sup2; per year), the trend across the populated zones is not as obvious, and the supply drop to its lowest in the very low population density zone (1.0 mm/m\u0026sup2; per year). The non-material contribution (NA) presents the most extreme disparity, with nearly all of its supply concentrated in unpopulated areas, making its mean supply drastically higher there than in any populated zone (Fig. 7d). HQ displays the most complex pattern. While it follows the overall trend of declining from unpopulated to populated areas, it shows a surprising inverse pattern within the populated zones, where its supply slightly increases with population density (Fig. 7e). The analysis also shows that the variability of supply (interquartile range) for services like HM and HQ is widest in unpopulated areas and narrows significantly in denser zones, suggesting a more uniformly low supply in settled landscapes.\u003c/p\u003e\n\u003cp\u003eOverall, the inverse relationship between population density and ES supply (Fig. 7) underscores the impact of human settlement intensity on the landscape\u0026apos;s capacity to provide essential ESs. Again, the most urbanized zones, characterized by high and medium population densities, correspond to the areas with the most pronounced and consistent deficits in ES provision.\u003c/p\u003e\n\u003ch2\u003eEcosystem service bundles and trade-off/synergy analysis\u003c/h2\u003e\n\u003ch3\u003eEcosystem service bundles\u003c/h3\u003e\n\u003cp\u003eThe k-means clustering analysis of the five standardized ESs (WY, HM, SWRR, NA, and HQ) across the 258,394 hexagonal grids identified three distinct ES bundles. These bundles represent unique combinations of ES provision levels. Based on their characteristic ES supply profiles (Table 4), descriptive names were assigned to each bundle to aid in their interpretation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 4\u003c/strong\u003e Characteristics of the three ecosystem service bundles identified in Kabul metropolitan region\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"623\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8.97436%;\"\u003e\n \u003cp\u003eBundle\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.1795%;\"\u003e\n \u003cp\u003eWater Yield\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6218%;\"\u003e\n \u003cp\u003eHeat Mitigation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0641%;\"\u003e\n \u003cp\u003eStormwater Runoff Retention\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.7821%;\"\u003e\n \u003cp\u003eNature Access\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6218%;\"\u003e\n \u003cp\u003eHabitat Quality\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.1795%;\"\u003e\n \u003cp\u003eArea Coverage (ha)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 10.5769%;\"\u003e\n \u003cp\u003eArea Coverage (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8.97436%;\"\u003e\n \u003cp\u003eBundle 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.1795%;\"\u003e\n \u003cp\u003e0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6218%;\"\u003e\n \u003cp\u003e-0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0641%;\"\u003e\n \u003cp\u003e-0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.7821%;\"\u003e\n \u003cp\u003e-0.40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6218%;\"\u003e\n \u003cp\u003e-0.25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.1795%;\"\u003e\n \u003cp\u003e130,422\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 10.5769%;\"\u003e\n \u003cp\u003e51%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8.97436%;\"\u003e\n \u003cp\u003eBundle 2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.1795%;\"\u003e\n \u003cp\u003e-0.63\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6218%;\"\u003e\n \u003cp\u003e0.77\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0641%;\"\u003e\n \u003cp\u003e0.84\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.7821%;\"\u003e\n \u003cp\u003e-0.14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6218%;\"\u003e\n \u003cp\u003e-0.51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.1795%;\"\u003e\n \u003cp\u003e82,183\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 10.5769%;\"\u003e\n \u003cp\u003e32%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 8.97436%;\"\u003e\n \u003cp\u003eBundle 3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.1795%;\"\u003e\n \u003cp\u003e1.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6218%;\"\u003e\n \u003cp\u003e0.98\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15.0641%;\"\u003e\n \u003cp\u003e0.87\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.7821%;\"\u003e\n \u003cp\u003e1.39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 13.6218%;\"\u003e\n \u003cp\u003e1.64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.1795%;\"\u003e\n \u003cp\u003e45,789\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 10.5769%;\"\u003e\n \u003cp\u003e18%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eValues for water yield, heat mitigation, stormwater runoff retention, nature access, and habitat quality represent the standardized (z-score) values of each attribute at the centroid of each k-means cluster (bundle). A value of 0 indicates the mean, positive values indicate above-average ecosystem service provision, and negative values indicate below-average provision. Percentages are based on the total study area (258,394 ha).\u003c/p\u003e\n\u003cp\u003eBundle 1 (Low-Service Zone) is characterized by approximately average WY, while exhibiting notably low provision of HM (-0.83 SD) and SWRR (-0.83 SD). It also shows moderately below-average levels of NA (-0.40 SD) and HQ (-0.25 SD). This bundle is the most dominant, covering 130,422 ha (51%) of the study area (Table 4). Spatially, it aligns strongly with the central, densely urbanized core of the Kabul municipality (Fig. 8).\u003c/p\u003e\n\u003cp\u003eBundle 2 (Regulating-Service Zone) presents a distinct profile with significantly above-average provision of HM (+0.77 SD) and SWRR (+0.84 SD). However, it shows low WY (-0.63 SD) and HQ (-0.51 SD), with slightly below-average NA (-0.14 SD). It covers 82,183 ha (32%) of the study area (Table 4). This bundle is primarily located in the peri-urban and agricultural landscapes surrounding the urban core (Fig. 8).\u003c/p\u003e\n\u003cp\u003eBundle 3 (High-Service Multifunctional Zone) demonstrates the highest provision across multiple services. It features significantly above-average levels of WY (+1.13 SD), HM (+0.98 SD), SWRR (+0.87 SD), NA (+1.39 SD), and exceptionally high HQ (+1.64 SD). It is the least extensive bundle, covering 45,789 ha (18%) of the study area (Table 4). Spatially, it is concentrated in the outer, more natural landscapes in the eastern, southeastern, and northwestern parts of the study area (Fig. 8).\u003c/p\u003e\n\u003cp\u003eAnalysis of Spearman\u0026apos;s rank correlations within each of the three identified ES bundles revealed context-dependent relationships between services, highlighting how synergies and trade-offs vary across the different landscape types (Fig. 9).\u003c/p\u003e\n\u003cp\u003eA strong and consistent synergy between the two regulating contributions, HM and SWRR, was observed across all bundles (Fig. 9). The intensity of this synergy varied, being strongest in the High-Service Multifunctional Zone (Bundle 3; \u0026rho; = 0.64) and the Low-Service Zone (Bundle 1; \u0026rho; = 0.51), and slightly weaker, though still moderate, in the Regulating-Service Zone (Bundle 2; \u0026rho; = 0.40). This indicates that across the entire study area, interventions to increase vegetation for cooling would also yield significant benefits for stormwater management.\u003c/p\u003e\n\u003cp\u003eA strong and consistent synergy between the two regulating contributions, HM and SWRR, was observed across all bundles (Fig. 9). The intensity of this synergy varied, being strongest in the High-Service Multifunctional Zone (Bundle 3; \u0026rho; = 0.64) and the Low-Service Zone (Bundle 1; \u0026rho; = 0.51), and slightly weaker, though still moderate, in the Regulating-Service Zone (Bundle 2; \u0026rho; = 0.40). This indicates that across the entire study area, interventions to increase vegetation for cooling would also yield significant benefits for stormwater management.\u003c/p\u003e\n\u003cp\u003eThe interrelationships involving the non-material contribution (NA) and the regulating contribution of HQ were the most variable. A strong synergy between NA and HQ was found in the Regulating-Service Zone (Bundle 2; \u0026rho; = 0.57), suggesting that in these semi-natural landscapes, high-quality habitats are also highly accessible. However, this synergy weakened significantly in the Low-Service Zone (Bundle 1; \u0026rho; = 0.28) and flipped to a weak trade-off in the High-Service Multifunctional Zone (Bundle 3; \u0026rho; = -0.13), indicating a potential conflict between maximizing biodiversity and recreational use in these prime areas. The relationship between HM and NA also shifted dramatically, from a strong synergy in the High-Service Multifunctional Zone (Bundle 3; \u0026rho; = 0.56) to a weak synergy in the Low-Service Zone (Bundle 1; \u0026rho; = 0.20) and no significant relationship in the Regulating-Service Zone (Bundle 2; \u0026rho; = 0.00).\u003c/p\u003e\n\u003cp\u003eRelationships involving the material contribution, WY, also showed context dependency, primarily exhibiting trade-offs with other regulating services. The trade-off between WY and SWRR, expected from hydrological principles, was most pronounced in the Regulating-Service Zone (Bundle 2; \u0026rho; = -0.29) and weaker in the Low-Service Zone (Bundle 1; \u0026rho; = -0.21) and High-Service Multifunctional Zone (Bundle 3; \u0026rho; = -0.15). Similarly, a weak trade-off between WY and HM was observed in the Low-Service Zone (Bundle 1; \u0026rho; = -0.20) but became negligible in the other two bundles.\u003c/p\u003e\n\u003cp\u003eThese bundle-specific correlation patterns underscore that ES interrelationships are not uniform across the KMR. Management strategies aimed at enhancing multiple services or mitigating trade-offs must therefore consider the specific landscape context, as represented by the ES bundles, to be most effective. For instance, in the Regulating-Service Zone (Bundle 2), management actions could effectively leverage the strong synergies between WY, NA, and HQ. Conversely, in the High-Service Multifunctional Zone (Bundle 3), while most services are synergistic, careful planning is needed to manage the slight trade-off between providing NA and maintaining the highest levels of HQ.\u003c/p\u003e\n\u003ch2\u003eRecommended areas for conservation\u003c/h2\u003e\n\u003cp\u003eBuilding upon the ES assessments, a conservation priority analysis was conducted to identify zones warranting different levels of attention based on the provision of multiple ES benefits (Fig. 10). This classification evaluates the number of ESs performing above their respective area-wide means, refined by SD thresholds to indicate the magnitude of the benefit.\u003c/p\u003e\n\u003cp\u003eThe analysis reveals a clear spatial hierarchy of conservation needs. Areas designated as highest conservation priority, where all five ESs are supplied above their mean, cover 301 km\u0026sup2; (12% of the study area). These are complemented by high conservation priority areas (four ESs above mean), which account for another 253 km\u0026sup2; (10%). Spatially, these top-tier priority zones are concentrated in the eastern, southeastern, and northwestern and northeastern sections outside the municipality, corresponding to more natural land covers. While the majority of these areas (502 km\u0026sup2;) provide benefits of low to moderate magnitude, a combined 52 km\u0026sup2; within these two top tiers show high-magnitude benefits (greater than 1 SD above the mean).\u003c/p\u003e\n\u003cp\u003eRepresenting landscapes with moderate multifunctionality, Medium Conservation Priority areas (three ESs above mean) cover 264 km\u0026sup2; (10%). These are predominantly located outside the municipality near higher-priority zones, though some patches exist within the municipality away from the urban core (Fig. 10). The most extensive category is Low Conservation Priority (two ESs above mean), covering 654 km\u0026sup2; (25%) of the landscape. These areas, which include agricultural lands and less dense vegetated areas, are found in large portions both within and outside the municipality, often forming a transition between more natural and highly developed zones.\u003c/p\u003e\n\u003cp\u003eFinally, a significant portion of the KMR shows a profound deficit in multiple ES benefits. The Lowest Conservation Priority areas (one ES above mean) cover 496 km\u0026sup2; (19%), while a further 604 km\u0026sup2; (23%) was classified as a No Conservation Priority zone, having no services performing above the area-wide mean. These two lowest categories, together comprising 42% of the study area, strongly align with the central urban core and other degraded landscapes, reflecting the lowest levels of multiple ES provision.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study provides a critical, spatially explicit assessment of five key ESs (ESs)\u0026mdash;WY, HM, SWRR, NA, and HQ\u0026mdash;within the rapidly urbanizing, data-scarce context of the KMR. Our integrated approach, combining high-resolution LULC mapping with ES modeling, bundle analysis, and conservation prioritization, reveals profound spatial heterogeneity in service provision, intricate and context-dependent interrelationships between services, and clear spatial patterns of conservation needs. These findings offer crucial insights for sustainable urban planning in a challenging environment, contributing valuable knowledge relevant to KMR and other cities facing similar pressures in the Global South.\u003c/p\u003e\n\u003ch2\u003eSpatial heterogeneity and ecosystem service bundles\u003c/h2\u003e\n\u003cp\u003eConsistent with findings from diverse urban contexts (Liu et al. 2019; Bai et al. 2019; Cozzi et al. 2022), our results demonstrate a marked spatial gradient in ES provision, with significantly lower supply concentrated in the densely built-up municipal areas compared to the municipal periphery and surrounding non-municipal areas. This urban-rural dichotomy underscores the substantial impact of urbanization, particularly the proliferation of impervious surfaces and loss of vegetation, on local ecological functions. However, moving beyond a simple gradient, our identification of three distinct ES bundles provides a more nuanced understanding of KMR\u0026apos;s landscape structure (Raudsepp-Hearne et al. 2010). These bundles represent unique landscape archetypes, ranging from the service-deficient Low-Service Zone (Bundle 1) dominating the more urbanized municipal areas, to the specialized Regulating-Service Zone (Bundle 2) in the municipal periphery, and culminating in the High-Service Multifunctional Zone (Bundle 3) in the most intact outer zones. Recognizing these distinct bundle types, each with a characteristic signature of high and low service provision, is essential for moving beyond generalized planning approaches towards spatially tailored strategies that leverage or mitigate the specific ES profile of different landscape units (Mouchet et al. 2014; Queiroz et al. 2015).\u003c/p\u003e\n\u003cp\u003eOur findings on the relationship between ES supply and population density, which serves as a proxy for land-use intensity, can be contextualized using the stylized trajectories proposed by Locatelli et al. (2017). The strong inverse relationship we identified for HM and SWRR aligns well with their model, which posits that most regulating services decline as land-use intensity increases from natural to urban states. However, our analysis also reveals patterns with more nuance than these general models suggest. For instance, the supply of the material contribution, WY, did not follow the typical trajectory of peaking at an intermediate intensity; instead, it was highest in unpopulated areas with no clear trend across the populated zones. More surprisingly, the regulating service of HQ, while highest in unpopulated areas, exhibited a slight increase with population density within the populated zones. This is a notable deviation from the expected steady decline and may reflect the specific ecological characteristics of the remaining green spaces within Kabul\u0026apos;s informally developed landscapes. These unique patterns underscore the assertion by Locatelli et al. (2017) that while stylized models are useful, they must be adapted to account for local context and the specific nature of land-use transitions.\u003c/p\u003e\n\u003ch2\u003eContext-dependent ecosystem service interrelationships\u003c/h2\u003e\n\u003cp\u003eA key contribution of this study is the demonstration that relationships between ESs are highly context-dependent, varying significantly across the three identified bundles. While the overall analysis revealed expected patterns, the bundle-specific correlations uncovered important nuances. For instance, the synergy between HM and SWRR was present across all bundles but was strongest in the High-Service Multifunctional Zone (Bundle 3). The trade-off between WY and SWRR was most pronounced in the Regulating-Service Zone (Bundle 2). Furthermore, the relationship between NA and HQ shifted from a strong synergy in the Regulating-Service Zone (Bundle 2) to a weak trade-off in the High-Service Multifunctional Zone (Bundle 3), suggesting a potential conflict between maximizing habitat integrity and providing human access in these highest-value areas. This context-dependency highlights the limitations of relying solely on landscape-wide correlations and emphasizes the need for management interventions to be tailored to the specific bundle type to effectively manage trade-offs and leverage potential synergies (Spake et al. 2017).\u003c/p\u003e\n\u003ch2\u003eImplications for conservation planning\u003c/h2\u003e\n\u003cp\u003eThe prioritization analysis translates the complex ES patterns into actionable spatial guidance for conservation planning. The results reveal a stark spatial segregation of needs: the highest conservation priorities, representing areas providing multiple ES benefits simultaneously, are overwhelmingly concentrated in the remaining natural and semi-natural landscapes, mostly corresponding to the High-Service Multifunctional Zone (Bundle 3) and parts of the Regulating-Service Zone (Bundle 2). Conversely, the areas with no conservation priority, characterized by deficits across multiple ESs, are heavily clustered within the central municipal areas, aligning with the Low-Service Zone (Bundle 1). This clear spatial dichotomy underscores the urgent need for a dual planning strategy that moves beyond static conservation. This strategy must involve both protecting the vital, multifunctional green and blue infrastructure and, crucially, implementing restoration interventions in the lower-priority urban matrix. Such interventions, for example establishing green corridors or \u0026quot;stepping-stone\u0026quot; habitats, would not only improve local livability but could also enhance ecological connectivity between the fragmented high-priority zones. The magnitude analysis within each priority level further allows for finer-grained prioritization, directing initial conservation and restoration efforts towards areas with the most substantial multi-service benefits or the greatest potential for improving connectivity. Finally, it is imperative that any such planning explicitly incorporates principles of environmental justice to ensure the equitable distribution of these benefits, a particularly critical consideration in the socio-political context of Kabul, which has a history of intense ethnic competition and civil war.\u003c/p\u003e\n\u003ch2\u003eMethodological significance and contribution\u003c/h2\u003e\n\u003cp\u003eThis study demonstrates the utility of integrating multi-source remote sensing data, process-based models (InVEST), and spatial analysis techniques (clustering, prioritization) to assess ESs in data-scarce environments. By developing a high-resolution (10m) LULC map, our analysis captures finer-scale landscape heterogeneity compared to previous coarser assessments in the region (e.g., Najmuddin et al. 2022), providing more relevant information for urban planning. The application of ES bundle analysis and context-specific correlation assessment offers a more nuanced understanding of ES interrelationships than typically achieved through landscape-wide correlations alone. The findings provide a crucial baseline for KMR, addressing the call by IPBES (2018) for better biodiversity and ES information in the rapidly changing Asia-Pacific region. Furthermore, the methodological framework and insights are potentially transferable to other cities in the Global South facing similar challenges of rapid informal urbanization, data limitations, and escalating environmental pressures.\u003c/p\u003e\n\u003ch2\u003eLimitations and future research directions\u003c/h2\u003e\n\u003cp\u003eDespite its contributions, this study has limitations. The reliance on globally available datasets (e.g., SoilGrids) introduces uncertainties, as these may not fully capture local variations in climate and soil properties critical for ES modeling. While the LULC map represents a significant improvement, potential classification errors remain, and the InVEST models, like all models, rely on simplifying assumptions and parameters (e.g., threat/sensitivity tables for HQ) that require local calibration and validation for enhanced accuracy. Ground-truthing of both LULC and ES model outputs was beyond the scope of this study but represents a crucial next step. Furthermore, the analysis provides a static snapshot (circa 2020), and its focus on five ESs, while relevant, does not capture the full spectrum of services. The study also does not explicitly model the dynamic processes of informal settlement expansion or climate change impacts on future ES provision.\u003c/p\u003e\n\u003cp\u003eFuture research should prioritize acquiring local climate, soil, and hydrological data to refine ES models and validate their outputs. To achieve a more holistic understanding, future assessments should also consider expanding the analysis to include a broader range of ESs, such as food production, air purification, and other services, to better capture the entire spectrum of benefits. Dynamic LULC modeling, incorporating scenarios of future urbanization (both formal and informal) and climate change, is needed to assess potential future trajectories of ES supply and interrelationships. Investigating the socio-economic dimensions, including ES demand, access equity across different demographic groups, and community perceptions, is essential for developing socially just and effective management strategies. Exploring the governance challenges and opportunities for implementing conservation recommendations, particularly within informally settled areas, is also critical for translating research into tangible outcomes. Finally, comparative studies with other cities facing similar contexts would help generalize findings and refine methodologies for ES assessment in data-scarce environments.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis research provides a comprehensive and spatially explicit baseline understanding of ES provision, heterogeneity, interrelationships, and priorities within the KMR. By identifying distinct ES bundles, revealing context-dependent synergies and trade-offs, and delineating clear priorities for conservation, the study offers valuable, actionable insights for evidence-based urban planning and environmental management. Addressing the significant spatial disparities and leveraging the multi-functional benefits of remaining green and blue infrastructure are critical for enhancing KMR\u0026apos;s resilience, sustainability, and livability in the face of rapid urbanization and environmental change. This work underscores the importance of tailored, high-resolution ES assessments, particularly in data-scarce regions of the Global South, to guide development towards more sustainable pathways.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eAcknowledgments\u003c/p\u003e\n\u003cp\u003eThis paper is part of the first author\u0026apos;s PhD project at the Institute of Environmental Planning, Leibniz University Hannover. The author(s) would like to express their sincere gratitude to the German Academic Exchange Service (DAAD) for providing the Hilde Domin student at-risk PhD scholarship, and to the Institute of Environmental Planning, Leibniz University Hannover for their support, which made this research possible. We also acknowledge the valuable contributions of all public data source providers who made their data publicly available, enabling this research.\u003c/p\u003e\n\u003cp\u003eFunding\u003c/p\u003e\n\u003cp\u003eFunding for this research was provided by the German Academic Exchange Service (DAAD) through the Hilde Domin Programme PhD, 2023 (57615866) under the award number 91853250.\u003c/p\u003e\n\u003cp\u003eAuthor contributions\u003c/p\u003e\n\u003cp\u003eM.R.A. wrote the main manuscript text. The study was conceptualized by M.R.A., M.R., B.A.E., L.B.F., and C.A. The methodology was designed by M.R.A., M.R., B.A.E., and C.A., while M.R.A. also performed the data curation. Formal analysis was conducted by M.R.A. and M.R. Visualizations were prepared by M.R.A., D.K., and C.A. Additionally, C.A. was responsible for funding acquisition, project administration, and supervision. All authors reviewed and edited the manuscript.\u003c/p\u003e\n\u003cp\u003eDeclaration of competing interest\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAdem Esmail B, Cortinovis C, Geneletti D, et al (2025) Mapping and Analyzing Ecosystem Services Hotspots and Coldspots for Sustainable Spatial Planning in the Greater Asmara Area, Eritrea. Environ Manage 75:551\u0026ndash;567. https://doi.org/10.1007/s00267-024-02078-x\u003c/li\u003e\n\u003cli\u003eAdem Esmail B, Cortinovis C, Wang J, et al (2023) Mapping and assessing ecosystem services for sustainable policy and decision-making in Eritrea. Ambio 52:1022\u0026ndash;1039. https://doi.org/10.1007/s13280-023-01841-4\u003c/li\u003e\n\u003cli\u003eAkhtar M, Zhao Y, Gao G, et al (2022) Assessment of spatiotemporal variations of ecosystem service values and hotspots in a dryland: A case‐study in Pakistan. Land Degrad Dev 33:1383\u0026ndash;1397. https://doi.org/10.1002/ldr.4245\u003c/li\u003e\n\u003cli\u003eAlbert C, Aronson J, F\u0026uuml;rst C, Opdam P (2014) Integrating ecosystem services in landscape planning: requirements, approaches, and impacts. Landsc Ecol 29:1277\u0026ndash;1285. https://doi.org/10.1007/s10980-014-0085-0\u003c/li\u003e\n\u003cli\u003eAlbert C, Bonn A, Burkhard B, et al (2016) Towards a national set of ecosystem service indicators: Insights from Germany. Ecol Indic 61:38\u0026ndash;48. https://doi.org/10.1016/j.ecolind.2015.08.050\u003c/li\u003e\n\u003cli\u003eAllen RG, Pereira LS, Raes D, Smith M (1998) Crop evapotranspiration: Guidelines for computing crop water requirements (Chapter 8, Table 22). In: FAO Irrigation and Drainage Paper No. 56. FAO\u003c/li\u003e\n\u003cli\u003eBagstad KJ, Semmens DJ, Ancona ZH, Sherrouse BC (2017) Evaluating alternative methods for biophysical and cultural ecosystem services hotspot mapping in natural resource planning. Landsc Ecol 32:77\u0026ndash;97. https://doi.org/10.1007/s10980-016-0430-6\u003c/li\u003e\n\u003cli\u003eBai L, Xiu C, Feng X, Liu D (2019) Influence of urbanization on regional habitat quality:a case study of Changchun City. Habitat Int 93:102042. https://doi.org/10.1016/j.habitatint.2019.102042\u003c/li\u003e\n\u003cli\u003eBirch CPD, Oom SP, Beecham JA (2007) Rectangular and hexagonal grids used for observation, experiment and simulation in ecology. Ecol Model 206:347\u0026ndash;359. https://doi.org/10.1016/j.ecolmodel.2007.03.041\u003c/li\u003e\n\u003cli\u003eBolund P, Hunhammar S (1999) Ecosystem services in urban areas. Ecol Econ 29:293\u0026ndash;301. https://doi.org/10.1016/S0921-8009(99)00013-0\u003c/li\u003e\n\u003cli\u003eCentral Statistics Organization (2015a) Kabul Socio-Demographic and Economic Survey. Central Statistics Organization\u003c/li\u003e\n\u003cli\u003eCentral Statistics Organization (2015b) Parwan Socio-Demographic and Economic Survey. Central Statistics Organization\u003c/li\u003e\n\u003cli\u003eChaturvedi V, Kuffer M, Kohli D (2020) Analysing Urban Development Patterns in a Conflict Zone: A Case Study of Kabul. Remote Sens 12:3662. https://doi.org/10.3390/rs12213662\u003c/li\u003e\n\u003cli\u003eCity Mayors (2015) City Mayors: World\u0026rsquo;s fastest growing urban areas (1). http://www.citymayors.com/statistics/urban_growth1.html. Accessed 26 Sep 2024\u003c/li\u003e\n\u003cli\u003eCostanza R, d\u0026rsquo;Arge R, de Groot R, et al (1997) The value of the world\u0026rsquo;s ecosystem services and natural capital. Nature 387:253\u0026ndash;260. https://doi.org/10.1038/387253a0\u003c/li\u003e\n\u003cli\u003eCozzi M, Prete C, Viccaro M, et al (2022) Understanding the Role of Nature in Urban-Rural Linkages: Identifying the Potential Role of Rural Nature-Based Attractive Clusters That Serve Human Well-Being. Sustainability 14:11856. https://doi.org/10.3390/su141911856\u003c/li\u003e\n\u003cli\u003eDaily GC (ed) (2010) Nature\u0026rsquo;s services: societal dependence on natural ecosystems. Island Press, Washington, DC\u003c/li\u003e\n\u003cli\u003eDe Groot RS, Alkemade R, Braat L, et al (2010) Challenges in integrating the concept of ecosystem services and values in landscape planning, management and decision making. Ecol Complex 7:260\u0026ndash;272. https://doi.org/10.1016/j.ecocom.2009.10.006\u003c/li\u003e\n\u003cli\u003eDeng Z, Cao J (2023) Incorporating ecosystem services into functional zoning and adaptive management of natural protected areas as case study of the Shennongjia region in China. Sci Rep 13:18870. https://doi.org/10.1038/s41598-023-46182-0\u003c/li\u003e\n\u003cli\u003eDeWitt JD, Boston KM, Alessi MA, Chirico PG (2022) Quantifying and visualizing 32 years of agricultural land use change in Kabul, Afghanistan. J Maps 18:352\u0026ndash;361. https://doi.org/10.1080/17445647.2022.2063079\u003c/li\u003e\n\u003cli\u003eD\u0026iacute;az S, Pascual U, Stenseke M, et al (2018) Assessing nature\u0026rsquo;s contributions to people. Science 359:270\u0026ndash;272. https://doi.org/10.1126/science.aap8826\u003c/li\u003e\n\u003cli\u003eEllis P, Roberts M (2016) Leveraging Urbanization in South Asia: Managing Spatial Transformation for Prosperity and Livability. World Bank, Washington, DC\u003c/li\u003e\n\u003cli\u003eEsri Why hexagons?\u0026mdash;ArcGIS Pro | Documentation. https://pro.arcgis.com/en/pro-app/latest/tool-reference/spatial-statistics/h-whyhexagons.htm. Accessed 17 Apr 2025\u003c/li\u003e\n\u003cli\u003eFick SE, Hijmans RJ (2017) WorldClim 2: new 1‐km spatial resolution climate surfaces for global land areas. Int J Climatol 37:4302\u0026ndash;4315. https://doi.org/10.1002/joc.5086\u003c/li\u003e\n\u003cli\u003eFood and Agriculture Organization of the United Nations (2021) Crop type (Afghanistan - 10m - 2020 - 6 classes) - EOSTAT\u003c/li\u003e\n\u003cli\u003eFood and Agriculture Organization of the United Nations (1998) Daily extraterrestrial radiation (15th day of the month). In Annex 2. Meteorological tables of Crop evapotranspiration - Guidelines for computing crop water requirements (FAO Irrigation and Drainage Paper 56)\u003c/li\u003e\n\u003cli\u003eGangahagedara R, Subasinghe S, Lankathilake M, et al (2021) Ecosystem Services Research Trends: A Bibliometric Analysis from 2000\u0026ndash;2020. Ecologies 2:366\u0026ndash;379. https://doi.org/10.3390/ecologies2040021\u003c/li\u003e\n\u003cli\u003eGauster M (2021) Ecological threats to security and state resilience in Afghanistan. Secur Def Q 33:31\u0026ndash;40. https://doi.org/10.35467/sdq/132023\u003c/li\u003e\n\u003cli\u003eGeneletti D (2016) Ecosystem services analysis for Strategic Environmental Assessment: concepts and examples. In: Geneletti D (ed) Handbook on Biodiversity and Ecosystem Services in Impact Assessment. Edward Elgar Publishing\u003c/li\u003e\n\u003cli\u003eGeofabrik (2024) Afghanistan latest shapefiles\u003c/li\u003e\n\u003cli\u003eHaines-Yong R, Potschin M (2012) Common International Classification of Ecosystem Services (CICES) Version 4. European Environment Agency\u003c/li\u003e\n\u003cli\u003eHaines-Young R, Potschin M (2010) The links between biodiversity, ecosystem services and human well-being. In: Raffaelli DG, Frid CLJ (eds) Ecosystem Ecology, 1st edn. Cambridge University Press, pp 110\u0026ndash;139\u003c/li\u003e\n\u003cli\u003eHarrell Jr FE (2025) Hmisc: Harrell Miscellaneous\u003c/li\u003e\n\u003cli\u003eHartigan JA, Wong MA (1979) Algorithm AS 136: A K-Means Clustering Algorithm. Appl Stat 28:100. https://doi.org/10.2307/2346830\u003c/li\u003e\n\u003cli\u003eHengl T, Mendes De Jesus J, Heuvelink GBM, et al (2017) SoilGrids250m: Global gridded soil information based on machine learning. PLOS ONE 12:e0169748. https://doi.org/10.1371/journal.pone.0169748\u003c/li\u003e\n\u003cli\u003eIPBES (2019) Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. IPBES Secretariat\u003c/li\u003e\n\u003cli\u003eIPBES, Karki M, Sellamuttu SS, et al (2018) The IPBES regional assessment report on biodiversity and ecosystem services for Asia and the Pacific. Zenodo. URL https://zenodo.org/record/3237373\u003c/li\u003e\n\u003cli\u003eISRIC - World Soil Information (2017) SoilGrids250m: Derived available soil water capacity (volumetric fraction) until wilting point at seven standard depths (0 cm, 5 cm, 15 cm, 30 cm, 60 cm, 100 cm, 200 cm)\u003c/li\u003e\n\u003cli\u003eKarimi JD, Corstanje R, Harris JA (2021) Bundling ecosystem services at a high resolution in the UK: trade-offs and synergies in urban landscapes. Landsc Ecol 36:1817\u0026ndash;1835. https://doi.org/10.1007/s10980-021-01252-4\u003c/li\u003e\n\u003cli\u003eKowarik I, Bartz R, Brenck M, Hansj\u0026uuml;rgens B (2017) Ecosystem services in the city: protecting health and enhancing quality of life: summary for decision-makers. Naturkapital Deutschland \u0026ndash; TEEB DE, Berlin Leipzig\u003c/li\u003e\n\u003cli\u003eK\u0026uuml;lling N, Adde A, Lambiel A, et al (2024) Nature\u0026rsquo;s contributions to people and biodiversity mapping in Switzerland: spatial patterns and environmental drivers. Ecol Indic 163:112079. https://doi.org/10.1016/j.ecolind.2024.112079\u003c/li\u003e\n\u003cli\u003eLang N, Schindler K, Wegner JD (2022) ETH_GlobalCanopyHeight_10m_2020_version1: Global canopy height map for the year 2020 derived from Sentinel-2 and GEDI\u003c/li\u003e\n\u003cli\u003eLehner B, Grill G (2013) Global river hydrography and network routing: baseline data and new approaches to study the world\u0026rsquo;s large river systems. Hydrol Process 27:2171\u0026ndash;2186\u003c/li\u003e\n\u003cli\u003eLi T, Ren Y, Ai Z, et al (2024) Revealing the Spatial Interactions and Driving Factors of Ecosystem Services: Enlightenments under Vegetation Restoration. Land 13:511. https://doi.org/10.3390/land13040511\u003c/li\u003e\n\u003cli\u003eLi Y, Zhang L, Yan J, et al (2017) Mapping the hotspots and coldspots of ecosystem services in conservation priority setting. J Geogr Sci 27:681\u0026ndash;696. https://doi.org/10.1007/s11442-017-1400-x\u003c/li\u003e\n\u003cli\u003eLiu W, Zhan J, Zhao F, et al (2019) Impacts of urbanization-induced land-use changes on ecosystem services: A case study of the Pearl River Delta Metropolitan Region, China. Ecol Indic 98:228\u0026ndash;238. https://doi.org/10.1016/j.ecolind.2018.10.054\u003c/li\u003e\n\u003cli\u003eLocatelli B, Lavorel S, Sloan S, et al (2017) Characteristic trajectories of ecosystem services in mountains. Front Ecol Environ 15:150\u0026ndash;159. https://doi.org/10.1002/fee.1470\u003c/li\u003e\n\u003cli\u003eMaechler M, Rousseeuw P, Struyf A, et al (2025) cluster: Cluster Analysis Basics and Extensions\u003c/li\u003e\n\u003cli\u003eMaes J, Egoh B, Willemen L, et al (2012) Mapping ecosystem services for policy support and decision making in the European Union. Ecosyst Serv 1:31\u0026ndash;39. https://doi.org/10.1016/j.ecoser.2012.06.004\u003c/li\u003e\n\u003cli\u003eMahendra A, King R, Du J, et al (2021) Towards a More Equal City: Seven Transformations for More Equitable and Sustainable Cities. World Resour Inst. https://doi.org/10.46830/wrirpt.19.00124\u003c/li\u003e\n\u003cli\u003eMahmoodi S, Dutta K, Basu D, Agrawal S (2019) UNDERSTANDING LINK BETWEEN LAND SURFACE TEMPERATURE AND LANDSCAPE HETEROGENEITY: A SPATIO-TEMPORAL AND INTER-SEASONAL VARIABILITY STUDY ON KABUL CITY, AFGHANISTAN. ISPRS Ann Photogramm Remote Sens Spat Inf Sci IV-5/W2:57\u0026ndash;65. https://doi.org/10.5194/isprs-annals-IV-5-W2-57-2019\u003c/li\u003e\n\u003cli\u003eMcPhearson T, Andersson E, Elmqvist T, Frantzeskaki N (2015) Resilience of and through urban ecosystem services. Ecosyst Serv 12:152\u0026ndash;156. https://doi.org/10.1016/j.ecoser.2014.07.012\u003c/li\u003e\n\u003cli\u003eMillennium Ecosystem Assessment (2005) Ecosystems and Human Well-being: Synthesis. Island Press, Washington, DC\u003c/li\u003e\n\u003cli\u003eMouchet MA, Lamarque P, Mart\u0026iacute;n-L\u0026oacute;pez B, et al (2014) An interdisciplinary methodological guide for quantifying associations between ecosystem services. Glob Environ Change 28:298\u0026ndash;308. https://doi.org/10.1016/j.gloenvcha.2014.07.012\u003c/li\u003e\n\u003cli\u003eMouselimis L (2024) ClusterR: Gaussian Mixture Models, K-Means, Mini-Batch-Kmeans, K-Medoids and Affinity Propagation Clustering\u003c/li\u003e\n\u003cli\u003eNajmuddin O, Deng X, Bhattacharya R (2018) The Dynamics of Land Use/Cover and the Statistical Assessment of Cropland Change Drivers in the Kabul River Basin, Afghanistan. Sustainability 10:. https://doi.org/10.3390/su10020423\u003c/li\u003e\n\u003cli\u003eNajmuddin O, Li Z, Khan R, Zhuang W (2022) Valuation of Land-Use/Land-Cover-Based Ecosystem Services in Afghanistan\u0026amp;mdash;An Assessment of the Past and Future. Land 11:. https://doi.org/10.3390/land11111906\u003c/li\u003e\n\u003cli\u003eNational Environmental Protection Agency (2014) Fifth National Report to the United Nations Convention on Biological Diversity \u0026ndash; Afghanistan. United Nations Convention on Biological Diversity (CBD)\u003c/li\u003e\n\u003cli\u003eNational Statistics and Information Authority (2004) Afghanistan Statistical Yearbook 2004. National Statistics and Information Authority of Afghanistan, Kabul, Afghanistan\u003c/li\u003e\n\u003cli\u003eNational Statistics and Information Authority (2023) Afghanistan Statistical Yearbook 2023-24. National Statistics and Information Authority of Afghanistan, Kabul, Afghanistan\u003c/li\u003e\n\u003cli\u003eNatural Capital Project (2024) Integrated Valuation of Ecocystem Services and Tradeoffs (InVEST)\u003c/li\u003e\n\u003cli\u003eNemec KT, Raudsepp-Hearne C (2013) The use of geographic information systems to map and assess ecosystem services. Biodivers Conserv 22:1\u0026ndash;15. https://doi.org/10.1007/s10531-012-0406-z\u003c/li\u003e\n\u003cli\u003eOrd JK, Getis A (1995) Local Spatial Autocorrelation Statistics: Distributional Issues and an Application. Geogr Anal 27:286\u0026ndash;306. https://doi.org/10.1111/j.1538-4632.1995.tb00912.x\u003c/li\u003e\n\u003cli\u003ePascual U, Balvanera P, D\u0026iacute;az S, et al (2017) Valuing nature\u0026rsquo;s contributions to people: the IPBES approach. Curr Opin Environ Sustain 26\u0026ndash;27:7\u0026ndash;16. https://doi.org/10.1016/j.cosust.2016.12.006\u003c/li\u003e\n\u003cli\u003ePesaresi M (2023) GHS-BUILT-H R2023A - GHS building height, derived from AW3D30, SRTM30, and Sentinel2 composite (2018)\u003c/li\u003e\n\u003cli\u003eQueiroz C, Meacham M, Richter K, et al (2015) Mapping bundles of ecosystem services reveals distinct types of multifunctionality within a Swedish landscape. AMBIO 44:89\u0026ndash;101. https://doi.org/10.1007/s13280-014-0601-0\u003c/li\u003e\n\u003cli\u003eR Core Team (2024) R: A Language and Environment for Statistical Computing\u003c/li\u003e\n\u003cli\u003eRaudsepp-Hearne C, Peterson GD, Bennett EM (2010) Ecosystem service bundles for analyzing tradeoffs in diverse landscapes. Proc Natl Acad Sci 107:5242\u0026ndash;5247. https://doi.org/10.1073/pnas.0907284107\u003c/li\u003e\n\u003cli\u003eRSLab (2020) True Land Surface Albedo\u003c/li\u003e\n\u003cli\u003eSallustio L, De Toni A, Strollo A, et al (2017) Assessing habitat quality in relation to the spatial distribution of protected areas in Italy. J Environ Manage 201:129\u0026ndash;137. https://doi.org/10.1016/j.jenvman.2017.06.031\u003c/li\u003e\n\u003cli\u003eSasaki (2019) Kabul Urban Design Framework (KUDF). Sasaki Company, USA\u003c/li\u003e\n\u003cli\u003eSchiavina M, Freire S, Carioli A, MacManus K (2023) GHS-POP R2023A - GHS population grid multitemporal (1975-2030)\u003c/li\u003e\n\u003cli\u003eSeppelt R, Dormann CF, Eppink FV, et al (2011) A quantitative review of ecosystem service studies: approaches, shortcomings and the road ahead: Priorities for ecosystem service studies. J Appl Ecol 48:630\u0026ndash;636. https://doi.org/10.1111/j.1365-2664.2010.01952.x\u003c/li\u003e\n\u003cli\u003eShackleton CM, Cilliers SS, Du Toit MJ, Davoren E (2021) The Need for an Urban Ecology of the Global South. In: Shackleton CM, Cilliers SS, Davoren E, Du Toit MJ (eds) Urban Ecology in the Global South. Springer International Publishing, Cham, pp 1\u0026ndash;26\u003c/li\u003e\n\u003cli\u003eShahraki SZ, Hosseini A, Sauri D, Hussaini F (2020) Fringe more than context: Perceived quality of life in informal settlements in a developing country: The case of Kabul, Afghanistan. Sustain Cities Soc 63:102494. https://doi.org/10.1016/j.scs.2020.102494\u003c/li\u003e\n\u003cli\u003eSimons G, Koster R, Droogers P (2020) HiHydroSoil v2.0 \u0026ndash; High Resolution Soil Maps of Global Hydraulic Properties. FutureWater, Wageningen, The Netherlands\u003c/li\u003e\n\u003cli\u003eSpake R, Lasseur R, Crouzat E, et al (2017) Unpacking ecosystem service bundles: Towards predictive mapping of synergies and trade-offs between ecosystem services. Glob Environ Change 47:37\u0026ndash;50. https://doi.org/10.1016/j.gloenvcha.2017.08.004\u003c/li\u003e\n\u003cli\u003eSpearman C (1904) The Proof and Measurement of Association between Two Things. Am J Psychol 15:72. https://doi.org/10.2307/1412159\u003c/li\u003e\n\u003cli\u003eSrivathsa A, Vasudev D, Nair T, et al (2023) Prioritizing India\u0026rsquo;s landscapes for biodiversity, ecosystem services and human well-being. Nat Sustain 6:568\u0026ndash;577. https://doi.org/10.1038/s41893-023-01063-2\u003c/li\u003e\n\u003cli\u003eSun X, Yang P, Tao Y, Bian H (2022) Improving ecosystem services supply provides insights for sustainable landscape planning: A case study in Beijing, China. Sci Total Environ 802:149849. https://doi.org/10.1016/j.scitotenv.2021.149849\u003c/li\u003e\n\u003cli\u003eTaher MR, Chornack MP, Mack TJ (2014) Groundwater levels in the Kabul Basin, Afghanistan, 2004\u0026ndash;2013. United States Geological Survey\u003c/li\u003e\n\u003cli\u003eTEEB (2010) The Economics of Ecosystems and Biodiversity: Mainstreaming the Economics of Nature. Earthscan\u003c/li\u003e\n\u003cli\u003eTorabi SE, Amin M, Phairuang W, et al (2024) High-Resolution Characterization of Aerosol Optical Depth and Its Correlation with Meteorological Factors in Afghanistan. Atmosphere 15:849. https://doi.org/10.3390/atmos15070849\u003c/li\u003e\n\u003cli\u003eUN Habitat (2015) The State of Afghan Cities. United Nations Human Settlements Programme (UN-Habitat), Kabul\u003c/li\u003e\n\u003cli\u003eU.S. Army Corps of Engineers (2010) AED design requirements: Hydrology studies (Provisional). U.S. Army Corps of Engineers, Afghanistan Engineer District\u003c/li\u003e\n\u003cli\u003eU.S. Department of Agriculture (2004) Chapter 9 - Hydrologic Soil-Cover Complexes. In: National Engineering Handbook. Part 630 \u0026ndash; Hydrology\u003c/li\u003e\n\u003cli\u003eU.S. Geological Survey (2018) Shuttle Radar Topography Mission (SRTM) 1 Arc-Second Global\u003c/li\u003e\n\u003cli\u003eVeteikis D, \u0026Scaron;abanovas S, Jankauskaitė M (2011) Landscape structure changes on the coastal plain of Lithuania during 1998\u0026ndash;2009. Baltica 24:107\u0026ndash;116\u003c/li\u003e\n\u003cli\u003eVihervaara P, R\u0026ouml;nk\u0026auml; M, Walls M (2010) Trends in Ecosystem Service Research: Early Steps and Current Drivers. AMBIO 39:314\u0026ndash;324. https://doi.org/10.1007/s13280-010-0048-x\u003c/li\u003e\n\u003cli\u003eVisual Crossing (2020) Weather Data \u0026amp; API Global Forecast \u0026amp; History Data - Historical weather data for kabul and Paghman (2020)\u003c/li\u003e\n\u003cli\u003eWorld Bank Group (2006) Why and how should Kabul upgrade its informal settlements (English). World Bank Group, Washington, D.C.\u003c/li\u003e\n\u003cli\u003eWu L, Sun C, Fan F (2022) Multi-criteria framework for identifying the trade-offs and synergies relationship of ecosystem services based on ecosystem services bundles. Ecol Indic 144:109453. https://doi.org/10.1016/j.ecolind.2022.109453\u003c/li\u003e\n\u003cli\u003eXu N, Duan H (2024) Ecological Functional Zoning in Urban Fringe Areas Based on the Trade-Offs Between Ecological\u0026ndash;Social Values in Ecosystem Services: A Case Study of Jiangning District, Nanjing. Land 13:1957. https://doi.org/10.3390/land13111957\u003c/li\u003e\n\u003cli\u003eZanaga D, Van De Kerchove R, De Keersmaecker W, et al (2021) ESA WorldCover 10 m 2020 v100\u003c/li\u003e\n\u003cli\u003eZaryab A, Nassery HR, Alijani F (2022) The effects of urbanization on the groundwater system of the Kabul shallow aquifers, Afghanistan. Hydrogeol J 30:429\u0026ndash;443. https://doi.org/10.1007/s10040-021-02445-6\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 and 2 are available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Data scarcity, Global South, Spatial analysis, Sustainable spatial planning, Urban ecosystem services","lastPublishedDoi":"10.21203/rs.3.rs-7130094/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7130094/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Context Urban expansion in the Global South challenges ecological sustainability. Understanding ecosystem service (ES) distribution and bundling in these rapidly changing landscapes is vital for informed urban planning and resilience. However, there is a research gap in applying integrated, high-resolution ES assessments to data-scarce, informally developing urban contexts.\n\nObjectives This study investigates ES patterns and trade-offs in the Kabul metropolitan region (KMR) in Afghanistan, a fast-growing, data-scarce city. Objectives are to: (1) map spatial patterns, including hotspots and coldspots, of five key ESs (water yield, heat mitigation, stormwater runoff retention, nature access, habitat quality), and (2) identify distinct ES bundles and their associated synergies and trade-offs.\n\nMethods The study focuses on KMR, as a representative case of a rapidly urbanizing, data-scarce region facing significant environmental pressures. Using multi-source data, a high-resolution land-use/land-cover map is created for ES modeling (InVEST), hotspot analysis (Getis-Ord Gi*), ES bundle identification (k-means clustering), and conservation prioritization.\n\nResults The assessment reveals significant spatial heterogeneity across five key ESs. ES hotspots corresponded to green/blue infrastructure, while coldspots are prevalent in dense urban areas. Distinct ES bundles are mapped, revealing context-dependent synergies (e.g., heat mitigation, stormwater runoff retention) and trade-offs (e.g., water yield, stormwater runoff retention). Finally, conservation priority areas are mapped based on multi-service supply.\n\nConclusions This study provides spatially explicit information for KMR's sustainable urban planning, using a procedure transferable to other cities in the Global South. Findings enable targeted interventions to enhance urban resilience by identifying priority zones for ES protection. Future research should prioritize model validation with local data.","manuscriptTitle":"Mapping and assessing ecosystem service hotspots and bundles in rapidly urbanizing, data-scarce regions: a case study of Kabul, Afghanistan","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-21 09:59:06","doi":"10.21203/rs.3.rs-7130094/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"9e0befcb-7f05-4ade-b834-8f42c896e975","owner":[],"postedDate":"August 21st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-10-21T19:23:11+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-21 09:59:06","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7130094","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7130094","identity":"rs-7130094","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}