Integrating Local Perceptions in Bayesian Belief Network for Watershed Planning: A Case in the Colombian Andes

Integrating Local Perceptions in Bayesian Belief Network for Watershed Planning: A Case in the Colombian Andes | 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 Integrating Local Perceptions in Bayesian Belief Network for Watershed Planning: A Case in the Colombian Andes Cesar Rojas, Riascos-Ochoa Javier, Nicola Clerici, Magnolia Longo This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7382296/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 04 Apr, 2026 Read the published version in Modeling Earth Systems and Environment → Version 1 posted 7 You are reading this latest preprint version Abstract Addressing global water security challenges requires effective watershed management that seamlessly integrates ecosystem service (ES) assessments with governance and socio-economic aspects of water resource utilization. However, most existing modeling approaches focus narrowly on biophysical processes and lack mechanisms for incorporating local knowledge under uncertainty. Unlike previous ES-watershed models that focused on biophysical indicators, our approach integrates ecological modeling with local perceptions and governance probabilistically, representing uncertainty and socio-ecological interactions that are often omitted from ES tools. The rural water supply system (RWSS) supports ES for water access, biodiversity, and social cohesion but is absent from watershed planning. To address this gap, we developed a spatial Bayesian Belief Network (BBN) model using data from RWSS organizations to support participatory zoning. The spatial BBN framework was applied to the Coello River Basin, Colombia, integrating high-resolution land use, ecosystem service multifunctionality indices, and multidimensional poverty indicators with perception data from the community leaders. The model generated spatial probability maps of multifunctional ES zones, incorporating governance capacity and conservation willingness. Key variables affecting ES multifunctionality include participation, restrictions, and poverty. Areas of high multifunctionality are aligned with zones of strong governance and high ecological integrity. This BBN enhances planning by integrating community insights into ecological modeling. By quantifying both the probability of ES supply and the uncertainty of predictions, the framework provides actionable insights for integrated watershed management (IWM) and supports the localization of global Sustainability Development Goals (SDG) (SDG 6-clean water and sanitation and SDG 15-life on land) in diverse socio-ecological contexts. Ecosystem service multifunctionality watershed planning local perception mapping watershed environmental zoning Mesli Index Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 INTRODUCTION Water security remains a pressing global challenge, particularly in mountainous regions, where climate variability, land-use change, and socioeconomic pressures intersect. Achieving sustainable water governance in such contexts requires approaches that integrate ecological processes with social and institutional systems that mediate water access. Rural water supply systems (RWSS) in the Andean Mountains are central to this challenge. They operate in ecologically valuable micro-watersheds and provide multiple ecosystem services (ES), such as water provision, sediment retention, and recreational opportunities. In many regions, RWSS are important socio-ecological systems that connect biodiversity conservation and community-based management. However, complementary methods for integrative spatial analysis of ecological functions and socioeconomic characteristics are needed (Delucchi 2024 ; Omarova et al. 2019 ). The implementation of ES–watershed modeling approaches has different limitations that reduce their use in IWM in rural areas. First, most frameworks, including widely utilized tools such as the Integrated Ecosystem Services and Trade-offs Tool (InVEST) (Posner et al. 2016 ), Artificial Intelligence for Ecosystem Services (ARIES) (Martínez-López et al. 2019 ), and Soil and Water Assessment Tool (SWAT) (Francesconi et al. 2016 ), predominantly emphasize biophysical processes, with minor consideration of socioeconomic conditions such as governance and local stakeholder perspectives. Second, decision-making confidence is limited in data-scarce regions because of the underrepresentation of uncertainty in model outputs, which can be caused by scale mismatches or data quality. Finally, these ES models seldom integrate ecological and social variables within a unified analytical framework, resulting in spatial priorities that may not align with the local community’s interests. It is necessary to have flexible methods for advancing the limits of ES mapping that explicitly incorporate governance, socioeconomic indicators, and ecosystem functions in a spatially explicit manner that includes uncertainty in their outcomes. Bayesian Belief Networks (BBNs) are an integrative modeling framework for studying complex socio-ecological systems such as watersheds. BBNs are probabilistic graphical representations that employ nodes and conditional probability tables (CPTs) to describe the relationships among variables. They are valuable analytical tools for synthesizing diverse quantitative (e.g., ecological indicators) and qualitative (e.g., stakeholder perceptions) data, enabling the representation of uncertainty and nonlinear relationships (Landuyt et al. 2016 ), which characterize socio-ecological systems (Scutari and Denis 2021 ). These BBN models can support scenario development (Phan et al. 2019 ; Rumbold 2023 ), identify ES trade-offs and synergies (Höfer et al. 2020 ), and promote participatory planning (Daniel et al. 2023 ; Salliou et al. 2017 ). Nevertheless, further investigation is needed to effectively apply BBNs for mapping ES multifunctionality, including local conservation preferences and governance forms in rural Andean contexts (Pham et al. 2021 ; Scrieciu et al. 2021 ; Seifert-Dähnn et al. 2015 ). Integrative spatial analysis contributes to identifying strategic zones in watersheds that deliver multiple ES supplies (ES multifunctionality) in RWSS. In this context, BBNs are a relevant framework that supports the integration of ecological models with community knowledge, creating maps that are both scientific and locally relevant to IWM. Studying ES multifunctionality in RWSS shows the ecological, social, and economic values of ecosystems (Hölting et al., 2019 ). In this context, mapping ES multifunctionality supports IWM by identifying how ES synergies and trade-offs are included in water-use decisions (Cole et al. 2023 ; Pilogallo and Scorza 2022 ). However, little research has been conducted on the implementation of ES multifunctionality in RWSS zoning, specifically in mapping high-value zones of supply and demand for ES (Rodríguez-Loinaz et al., 2015 ; Ruhl et al., 2021 ). In this context, according to recent studies, it is necessary for integrative analysis of ES better spatial ES data, community input, and land-use patterns (Bruen et al. 2022 ; Forio et al. 2020 ; Zeng and Li 2019 ). Similarly, global sustainability goals, such as SDG 6 (Clean Water and Sanitation) and SDG 15 (Life on Land), and local governance both benefit from addressing this issue. Watershed-level tools are necessary to achieve these objectives. The ecological and social facets of water management require comparable approaches (Shao et al., 2023 ). This challenge is particularly relevant in Colombia. In rural Colombia, the water supply relies predominantly on local community organizations to ensure an adequate water supply and timely management of water resources (Blanco-Moreno et al. 2024 ; Delacámara et al. 2020 ; Platarrueda 2022 ). Water-use management in the Andean region of Colombia, is characterized by complex relationships between ecosystem benefits, traditional practices, and socioeconomic conditions that underpin the well-being of river basin water users (Cardenas et al. 2011 ; Johnson et al. 2008 ). The objective of this study is to develop and apply a participatory spatial BBN framework that integrates ecological indicators, socio-economic metrics, and governance perceptions to support multifunctional ES zoning in RWSS areas of the Coello River Basin, Colombia We posed the following research question: Do local conservation preferences within multifunctional zones in RWSS improve IWM in watershed planning? To address this, we aimed to (i) develop a spatial BBN for RWSS to assess ES multifunctionality. (ii) integrate ecological and socioeconomic attributes derived from local perceptions, and (iii) represent ES multifunctional zones within land-use planning instruments in watersheds. By bridging ecological modeling with participatory governance, our approach aims to advance inclusive, data-informed watershed planning in rural Andean landscapes and offers a replicable method for other socio-ecological systems. The resulting framework not only quantifies multifunctional ES zones but also identifies areas where conservation priorities align or conflict with governance capacities, offering decision-relevant insights for IWM. MATERIALS AND METHODS Study area The Coello River watershed is located in the Andean region of Colombia, South America, and has an area of 1,817.7 km 2 , and is located within the Department of Tolima (Fig. 1 ). Land use and land cover (LULC) were distributed as follows: 88.7% forest and semi-natural areas, 0.39 water bodies, 9.41% agricultural land, and 1.42% artificial surfaces. The Watershed Management Plan (POMCA) designates up to 25% of the areas of high conservation importance that are protected under schemes such as natural parks, bird sanctuaries (AICAS), and forest reserves. These areas contain tropical dry forest relics, paramo ecosystems, and wetlands are present in these areas (Cortolima, 2019 ). The land-use patterns in the Coello River watershed directly influence water supply and demand, which are crucial for sustainable watershed management. The basin's tropical dry forests, high mountain ecosystems, and wetlands significantly contribute to their ecological value. The main ES in the basin is water yield, which varies between 17.98 l/s/km² during the dry season and 40.63 l/s/km² in the rainy season. The total water demand is 28,233.76 l/s, distributed among 881 concession users. The watershed supports a population of approximately 264,717, mainly concentrated in the municipalities of Ibagué and Cajamarca, which together host 91% of the population (Cortolima, 2019 ). This includes 156 RWSS, which serves 23.97% of the rural population. These RWSS play a critical role in decentralized water governance by supplying both domestic and agricultural water. 1.1 Overview of the Methodological Framework This study followed a four-stage methodology to integrate quantitative ES indicators and qualitative stakeholder perceptions into a spatial BBN to map ES multifunctionality: (1) ES indicator modeling, quantification of ES supply, demand, and socio-ecological context; (2) local perceptions mapping, stakeholder interviews to understand community knowledge, values, and concerns; (3) BBN development and validation: two models were developed by combining quantitative and qualitative data; and (4) zoning integration: BBN results were combined with planning instruments to delineate ES multifunctionality zones (Fig. 2 ). This approach supports sustainability planning by combining geo-data and participatory knowledge for IWM in watershed governance. Methods Indicators of Ecosystem Services and Spatial Modeling To characterize ES multifunctionality, we selected indicators based on the relevance to represent three dimensions: supply, demand, and socio-ecological access (Table 1). Spatial modeling was conducted using the InVEST and ARIES tools to map ES. Indicators included water yield, nutrient retention, carbon sequestration, pollination, habitat quality, and recreational value. The Mesli Index (Berrío and Montenegro, 2022 ) and species distribution models were also included to reflect the ecological conditions (N. N. et al. 2024). The demand-side indicators were derived from the LUCC data (Cortolima, 2019 ), a water-use index (Cortolima, 2020 ), and risk management layers (e.g., volcanic and flood zones). The socioeconomic context was captured using the Multidimensional Poverty Index and spatial distribution of public services (schools and hospitals). All variables were resampled to 30 m resolution, normalized (0–1 scale), and classified into ordinal states (e.g., low, medium, and high) for BBN input. This selection ensures that the BBN represents the coupled human–natural system rather than focusing solely on biophysical processes. The spatial data were managed using QGIS 2.3 (QGIS.org, 2023 ). Code Node Description States Source AWY Annual water yield The InVEST model was used to determine the relative contributions of water production in different watershed areas Very High (> 75th), High (50–75%), Moderate (25–50%), Low (10–25%), None (< 10%) Berrío and Montenegro ( 2022 ) CA Crop areas Agricultural production zones in watershed according to LULC maps (Cortolima, 2019 ) Yes (1), No (0) CS Carbon storage InVEST model for estimating carbon storage by land cover High (> 75 Mg C/ha), Moderate (40–75), Low ( 0.8), High (0.6–0.8), Moderate (0.4–0.6), Low (0.2–0.4), None ( 0.7), Moderate (0.4–0.7), Low (0.2–0.4), None ( 0.66), Moderate (0.33–0.66), Low ( 0.66), Moderate (0.33–0.66), Low ( 0.75), High (0.6–0.75), Moderate (0.4–0.6), Low (0.2–0.4), None ( 1.5), High (1.0–1.5), Moderate (0.5–1.0), Low (< 0.5), None (no data) Cortolima ( 2020 ) LUCCI Index of Land use land cover The zones of LULC in the watershed are forest, agriculture, and urban Agriculture, Forest, Urban LWC Liters of water Quantity of water granted at each water intake High (> 15 L/s), Moderate (5–15 L/s), Low (< 5 L/s) AMO Types of aqueduct management organizations Types of aqueduct management: private and public Community (Yes), Private (No) LPO Local participation scenarios Existence of local participation instruments in the area, in accordance with the management of the aqueduct Yes, No Alcaldía de Ibagué ( 2022 ) EB Equity in the use of benefits Census information with a representation of areas with water service access Yes (access ≥ 80%), No (access < 80%) DANE ( 2018 ) SPG Supply of goods for public use Rural areas with educational or health services Yes (present within 1 km), No (absent) MPI Multidimensional Poverty Index Percentage of people living in multidimensional poverty Yes (≥ 33%), No (< 33%) NTR Natural territorial hazards Land Use Restricted Zones related to hazards, flooding, torrential flooding, erosion, etc. Yes, No Cortolima ( 2019 ) ZPMES Zone of multifunctional Ecosystem Services ES Multifunctionality with improved ES supply and area for species distribution Very High (> 0.8), High (0.6–0.8), Moderate (0.4–0.6), Low (0.2–0.4), None (< 0.2) N.N et al. (2024) Local perception and mapping sampling A two-step strategy was used to map the local perceptions of the RWSS. (a) Demographic context: Population data for RWSS were obtained from an analysis of the catchment points in the Coello River Basin, as reported by Cortolima ( 2019 ). This secondary data source was used to collect RWSS demographic information. After completing the demographic analysis, in-depth interviews were conducted with the RWSS presidents to enrich the demographic data with qualitative insights into local conservation perceptions. (b) Stakeholder Interviews: Fifteen RWSS presidents were selected using stratified snowball sampling, covering altitudinal gradients, management types, and multifunctionality values. The interviews were conducted in a semi-structured format (Adeoye-Olatunde and Olenik, 2021 ), combining a Likert-scale and open-ended questions. The themes included: ecosystem benefits, conservation roles, threats, organizational capacity, and equity in access. The conservation attitudes were quantified using Likert responses (1–5) and then discretized to match the BBN state definitions. These leaders were selected to represent diverse altitudinal gradients (high, middle, and low zones), aqueduct management types (community-managed vs. private), and ecosystem service contexts. The questionnaire included Likert-scale items and open-ended questions designed to elicit the following responses: a) recognition of the role of forests and riparian areas in water regulation; b) level of community participation in conservation or reforestation activities; c) awareness of local environmental risks and land-use pressures; and c) organizational experience with education, governance, or water access equity. Data were collected between March and May 2023. 1.1.1 Bayesian Belief Network Model 1: Quantitative ES indicators Following Marcot et al. ( 2006 ), the BBN was built in three steps: i) structure and nodes: The model included 12 input nodes grouped into supply, demand, and access dimensions, all of which influenced the final ES multifunctionality output node. These were grouped into three intermediate components: supply, demand, and access. These components influence the final output node representing ES multifunctionality. ii) State Definitions: Based on the data distributions, each variable was classified into three to five ordinal states (Table 1). Thresholds for classification were defined using quantiles or expert-defined cutoffs depending on data availability and distribution. iii) Derivation of Conditional probability tables (CPT) derivation: We used hybrid learning: structure learning (hill-climbing + BIC) and parameter learning (MLE) on raster inputs via the bnlearn R package. Input data were extracted from 30 m layers and converted to categorical values using the thresholds listed in Table 1. iv) Sensitivity and Validation: The influence of variable in the BBN was assessed using GeNIe Modeler (v.2.1) and the logic and outcomes were validated through expert consultations. All spatial layers were rasterized to a 30 m resolution and intersected using QGIS. The thresholds listed in Table 1 were used to extract pixel-level values and categorize them into node states. Bayesian Belief Network Model 2: Local Perceptions Model 2 extended Model 1 by incorporating variables on conservation awareness, institutional trust, and perceived threats that are informed by stakeholders. BBN Model 2 followed the following steps: i) Structure and Nodes: The second model was structured around qualitative data derived from interviews and participatory mapping. Nodes representing perceived threats, values assigned to specific ES (e.g., water quality and forest cover), and trust in institutions were included. It included nodes representing perceived threats, values assigned to specific ES (e.g., water quality and forest cover), and trust in institutions. The output node represented the perceived ES condition. ii) State Definition: Qualitative responses were coded and aggregated into three states per variable (e.g., not valued, moderately valued, and highly valued) using content frequency and coding triangulation. iii) CPT Derivation: The perception data collected from interviews were systematically coded to inform the CPT in our BBN model due to the qualitative nature of the inputs, Due to the qualitative nature of the inputs, the perception data collected from interviews were systematically coded to inform the CPT in our BBN model, thus directly translating local insights into the model’s structure. iv iv) Spatial BBN: BBN nodes were linked to 30 m raster layers using the bnSpatial package (Masante, 2016 ) in R. Software as described by González-Redin et al. ( 2016 ). iv) Uncertainty analysis and validation: The spatial uncertainty was assessed using the Shannon entropy index, which was applied to posterior probabilities at the pixel scale. This metric quantifies the degree of uncertainty on the information content of each prediction across the spatial domain. For each pixel, the entropy H was calculated using Eq. 1: H(X) = - Σ P(x i ) · log₂ P(x i ) (1) Where: P(x i ) is the posterior probability of state x i , Where n is the number of states of the target node (e.g., Very High, High, Moderate, etc.). Incorporating local conservation perceptions into environmental planning areas We overlaid the BBN outputs with existing zoning instruments: the POMCA plan, complementary conservation measures (OECM), and ES multifunctionality zones. Each layer received equal weight using the terra package of R software (Hijmans et al. 2022 ) to compute ZEMSE for strategic conservation planning as indicated in Eq. 2 . By applying uniform weights, we ensured that no single data source or conservation instrument dominated the zoning synthesis, thus preserving the multidimensional integration required for equitable and evidence-based IWM in watershed: $$\:ZEMSE=\sum\:_{i}^{n}{w}_{i}\text{*}{x}_{i}$$ 2 where, ZEMSE: ES multifunctionality zones for strategic conservation planning. w i weight of each variable. x i Environmental zoning (from the POMCA instrument, complementary conservation areas). RESULTS Model 1. Bayesian network for ES multifunctionality zones The output of Model 1 was constructed to model the ES multifunctionality probabilities of in the Coello River Basin resulting from the ES supply and demand interaction (Fig. 3 ). A BBN provides high-probability ES distribution for water, carbon sequestration, nutrient delivery, and recreation. ES supply nodes that contribute more with a high probability to an ES multifunctional zone include water supply, habitat quality, forest carbon storage, and nutrient retention nodes. A more probable ES supply affects the probability that an RWSS will exhibit high multifunctionality. The high probability of the ES supply highlights its ecological significance as an area for species habitat distribution in watersheds, with a higher probability of being a zone of ES multifunctionality. The ES demand node has a high probability distribution for the water and forest land-use indices. The node of liters of water concession has a low probability. A high probability of forest land use and water use index is likely to exhibit an ES multifunctionality zone. Additionally, the BBN considers the socioeconomic characteristics associated with the population’s current living conditions, such as the multidimensional poverty node, which indicates a high probability of unsatisfactory basic needs. The type of organization and local participation scenario revealed a high distribution of occurrences. Similarly, BBN incorporates territorial elements, such as land-use restrictions, which represent natural limitations on watershed use. Finally, a territorial equity node was included, indicating a significant probability that the population in the area would not have access to the natural and public land-use benefits. Sensitivity analysis The sensitivity analysis showed that the variables with the highest influence on the ES multifunctionality output node were local participation in water planning, territorial access restrictions, and multidimensional poverty, each contributing to more than 60% uncertainty reduction. Additional variables with notable sensitivity included land-use type, equity in benefit distribution, supply of public goods, the Water Use Index, and the Mesli Index. The aggregated posterior probabilities for the ES multifunctionality output were 44%, 18%, and 10% for high, low, and for high multifunctionality, 18% for low multifunctionality, and 10% for zones lacking multifunctionality, respectively. In Fig. 4 , the dark red variables significantly affect the variables shown in dark red significantly affect the output node. The variables in the red shade have less influence on the end node. The variables that most influenced the uncertainty of ES Multifunctionality Zones (probability ≥ 60%) were local participation in local planning, territorial access restrictions, and multidimensional poverty. Other variables with significant impacts include land use, equity in the use of benefits, supply of public goods, the water use index, and the Mesli index. The model showed a 44% probability of high ES multifunctionality, 18% probability of low ES multifunctionality, and 10% probability of zones without ES multifunctionality. Model 2 Preferences for conservation in ES multifunctionality zones Model 2 (conservation preferences, Fig. 5 ) was updated and adjusted to Model 1 with new nodes acquired from local stakeholders’ interviews. The RWSS serving 50–100 people (29.34% of the total) had the highest participation in conservation activities, whereas those serving less than 50 people had the highest participation (23.7%). In this area, 75% of the rural aqueduct residents engage in agriculture, 23% engage in urban activities, and 5% participate in tourism. Among the aqueduct administrations, 70% are managed by local communities, and 15% are either shared or owned by private companies. Among the stakeholders, 17.4% participated in local conservation and restoration activities. 23% agreed with conservation activities in the area, and 47% agreed with the quality of the water supplied by the aqueduct, which is related to the variable “efficiency of water service provision”. Aqueduct organizations make collective efforts to manage their activities. Then operation involves self-financing the cost of its activities. In Model 2, when local perceptions were included, the probability of areas with ES multifunctionality increased from 35–44% (with respect to Model 1), the probability of areas without multifunctionality decreased from 10–8%, and the probability of areas with low MF increased from 18–15%. Regarding the node on conservation perceptions, 25% of the respondents agreed with conservation, whereas 35% disagreed with conservation activities in RWSS. Figure 6 shows the mapping of ES multifunctionality zones. The map reveals that the upper section of the watershed, characterized by a greater extent of forest cover and regions situated close to the riverbed, exhibits the highest probability values for ES multifunctionality. The lower watershed includes forest areas that remain after clearing, and the present water catchment contributes to a higher probability of sustaining multiple ES. The rural water supply system, in which local administration leads to organization and land ownership, is important for water supply area conservation activities. These areas have a high likelihood of multifunctionality because of a positive combination of socioeconomic and ecological factors. In contrast, areas with a medium likelihood of multifunctionality were found in the watershed region, whereas remnant forests and areas with a low likelihood of multifunctionality were characterized by soil transformation caused by urban and industrial activities. Model 2, uncertainty. The uncertainty map (Fig. 7 ) generated from Model 2 indicated spatial heterogeneity in the posterior probability distributions for ES multifunctionality across the Coello River watershed. Higher uncertainty values, represented by elevated entropy values, were concentrated in the lower watershed, particularly in urban expansion zones and areas dominated by agricultural activities. These areas exhibit greater divergence in the probability distribution among multifunctionality states, reflecting a lower model confidence in assigning a definitive ES condition. Instead, lower uncertainty values were associated with the upper watershed, where land cover is predominantly forested and includes protected ecosystems, national parks, and areas with territorial use restrictions (e.g., volcanic hazard zones). These zones showed a clear convergence toward the “high multifunctionality” state, with strongly skewed posterior probabilities. This pattern reflects stronger biophysical signals and more consistent stakeholder preferences. Intermediate uncertainty values appear in transition zones—particularly in the mid-basin—where agricultural land-use and forest remnants coexists. Here, the spatial heterogeneity in ES provision, land-use, and local governance practices results in a more even distribution of probabilities across ES multifunctionality states. Overall, the uncertainty map provides a critical complement to the most-likely state map by revealing areas where decision-making should be accompanied by additional stakeholder engagement, field validation, or targeted data collection due to model uncertainty. The most likely state of ES multifunctionality map was confirmed by identifying multifunctional zones and comparing them with current land use in the watershed based on LUCC in the watershed (Cortolima 2019 ). Persistence of land-use areas and likelihood of being highly functional according to key actors in the watershed. The water catchment areas, forested zones, and areas designated for conservation by local communities exhibited high ES multifunctionality. In contrast, areas in the lower watershed with a low probability of multifunctionality corresponded to zones with some fragments but without designated conservation areas without designated conservation areas but with some fragments. Furthermore, a local environmental authority was consulted in the basin regarding the spatial output to confirm that the areas displaying multifunctional ES on the map were indeed within the RWSS with forest vegetation. Mapping local perceptions of the multifunctionality of ecosystem services zones in environmental planning instruments The inclusion of node status zones with high multifunctionality complements the current watershed environmental zoning of the watersheds. This zoning is related to management plans, watershed management, and OECM and is called the Convention of Biodiversity in Decision 18/4 (CBD 2018 ). Figure 8 shows the regions with the highest probability of high ES multifunctionality based on the spatial results of the BBN. Figure 8 shows the regions with the highest probability of high ES multifunctionality. The regions with the highest multifunctionality occurrence of multifunctionality were in the higher watershed segment, which had considerable forest cover. These areas also display different forms of community organization regarding water use because agricultural activities are prevalent. Additionally, a more even-handed distribution of benefits was reported, and the provision of a water supply service was viewed as favorable. Moreover, these areas have topographic limitations that hinder changes in natural land use. In terms of the middle watershed areas, those with high probabilities of multifunctionality comprise zones with agricultural activity while simultaneously preserving forest cover. Similarly, basin riparian zones are noteworthy for their multifunctionality because they cover a substantial amount of land. Conversely, the lower part of the basin has the lowest ES multifunctionality. The proposed zoning method considers the areas influenced by the relevant urban zones. In addition, areas containing forest remnants were identified in the lower basin, which is fundamental for generating restoration strategies. DISCUSSION This study introduced a spatial BBN participatory framework that uses ES supply indicators, socioeconomic metrics, and local conservation perceptions into multifunctional ES zoning for the RWSS in the Coello River Basin, Colombia. The formulated models highlight the relationships between ES supply and demand in environmentally important areas, providing relevant local information of water uses for ES assessment (Carriger et al. 2016 ; Vollmer et al. 2022 ), collaborative spatial planning for water planning (Bruen et al. 2022 ), and community inputs in the implementation IWM (Guo et al. 2024 ). The spatial BBN results indicated that incorporating governance capacity and conservation preferences into spatial modeling changes the location and extent of priority areas compared to biophysical-only approaches (Metzger et al. 2021 ). The output maps provide decision-makers with an integrated view of ecological importance, social readiness, and governance potential. The spatial BBN flexible method used to analyze ES supply and demand relationships in areas with a high ES demand, such as the middle and lower parts of the watershed, demonstrated its applicability to include in informed decision in environmental planning of watersheds (Hölting et al. 2019 ; Landuyt et al. 2016 ). ES modeling frameworks, such as InVEST, have improved the quantification of biophysical processes but often overlook socioeconomic complexity and local stakeholder perceptions (Verburg et al., 2016). The BNN model addresses these gaps by linking ecological and social indicators within a probabilistic framework. The use of indices, such as the ES multifunctionality index, further refines our understanding of the RWSS ecological and social conditions, such as the water-use index and multidimensional poverty index, supporting its application in IWM. Such integrated analyses support the evaluation of use conflicts (Ricart & Rico-Amorós, 2021 ), identification of nature-based solution zones (Maragkaki et al., 2024 ), and determination of biodiversity conservation priorities (Wu et al., 2022 ). Local watershed conditions enable the identification of areas where ecological priorities align or conflict with local governance capacities, a dimension that is not captured by most traditional models. The BBN implemented in the RWSS promoted participatory analysis, which is central to water governance in the Andean region. This participatory dimension of the model improves the output’s relevance and legitimacy, contributing to more equitable and effective watershed planning. The spatialization of participatory outputs in the BBN allows for the integration of data on conservation preferences, organizational presence, and socioeconomic conditions, facilitating the identification of priority areas for environmental management (Laurila-Pant et al., 2019 ; Zhang et al., 2023 ). Similarly, linking multidimensional poverty metrics to ecological attributes, such as forest cover, yields critical insights into the territorial dynamics that influence conservation decision-making in complex socio-ecological systems (Sun et al., 2022 ). Thus, spatial prioritization shifted toward areas that, while ecologically valuable, also had strong community organization, established water governance practices, and a willingness to engage in conservation efforts. In contrast, biophysical models often emphasize remote areas with high ecological potential but low governance viability, thus limiting their implementation impact (Castro et al. 2016 ; Pham et al. 2021 ). Extending the integration of local knowledge, the spatial BBN framework also clarifies conservation preferences within RWSS zones, which are closely linked to livelihoods, institutional presence, and participation in local water planning (Johnson et al., 2008 ). In this sense, water access increases community participation in conservation, suggesting that IWM strategies should consider factors such as land tenure security and co-designed management plans (Colón López and Restrepo, 2019 ). Therefore, stakeholder perceptions of BBNs are key to promoting conservation policies that recognize shared benefits, ensure effective participation, and strengthen public water management strategies (Hallberg-Sramek et al., 2023 ). This reinforces the need for co-created water assessments in watershed management using local knowledge (Crevier and Parrott, 2019 ; Pereira et al., 2025 ). These findings contribute to the broader field of integrated socio-ecological earth system modeling, which emphasizes the coupling of human and natural systems to understand synergies and trade-offs in ES demand (Meraj et al. 2022 ). Similar to integrated assessment models and coupled human–natural systems modeling, our BBN framework explicitly links biophysical indicators (e.g., water yield and habitat quality) with social variables (e.g., governance participation, poverty, and equity). Unlike large-scale Earth system models that often operate at coarse spatial resolutions, this framework applies those principles at a decision-relevant local scale, making it more actionable for place-based policy and planning for effective IWM in the future. From an IWM perspective, spatial BBN output enables stakeholders to jointly evaluate the trade-offs between ecological sustainability, social equity, and governance feasibility. Therefore, recognizing high-probability areas in ES use decisions is essential for fostering a collective interest in ES multifunctionality. However, the establishment of conservation areas of collective interest that promote environmental management remains challenging (Cosens et al. 2021 ; Di Cintio et al. 2024 ). The assessment of ES multifunctionality using integrative methods provides additional knowledge on territorial management, complementing hydrological modeling in water resource management (Barraclough et al. 2022 ). Despite its strengths, this framework relies on the availability and quality of both ecological and perceptual data. Model adaptation is essential in areas with different governance structures or limited perception data. Future research should integrate climate change projections, hydrological extremes, and dynamic land use change scenarios to improve temporal foresight. Comparative studies of watersheds with different governance regimes could further test the generalizability of the method and refine its integration into broader socio-ecological modeling efforts. CONCLUSIONS This study developed a participatory spatial BBN for watershed planning that integrates data on ES supply and demand, socioeconomic indicators, and local perceptions into ES multifunctionality zoning for RWSS. The BBN model identified areas with high ES multifunctionality linked to equitable access, governance, and ecological integrity. Sensitivity analysis indicated that land-use restrictions, poverty, and participation influenced the ES multifunctionality outcomes in the watershed. This method advances IWM modeling by incorporating local forms of water governance and conservation perceptions into spatial maps. These model frameworks highlight the relevance of spatial planning methods in watershed environmental planning, which incorporate not only static assessments but also local community preferences. This approach is useful for river basins with limited data and low institutional presence. In addition, this method could contribute to strategic IWM planning, allowing planners to prioritize areas where conservation actions are likely to receive community support. In contexts such as the Coello River Basin, where local water organizations drive governance, this method can support decision-making, reduce conflicts, and improve local environmental instruments such as the POMCA. BBN’s participatory spatial framework is a suitable approach for other watersheds characterized by complex socio-ecological systems, whose environmental management seeks to balance ecological priorities with socio-economic realities. The probabilistic structure of the BBN model allows its adaptation to diverse sources of information and uncertainty analysis, making it more accepted for regions with low data availability and different water governance conditions than other ES models. Future analysis will focus on integrating broader IWM frameworks and linking them to global sustainability agendas, such as Sustainable Development Goals 6 and 15. Declarations Author Contribution Rojas C. wrote the main manuscript. Riascos- Ochoa, Longo and Clerici supervised the investigation. All authors reviewed the manuscript Acknowledgement We acknowledge Jiménez, C., for his support in collecting data. Rubiano, J. of the Semilla de Agua Foundation, for his interest in this process and support in collecting the data. Funding: This study was funded by the Research Office of Universidad District, Francisco José de Caldas, Bogotá, Colombia. The WWB Foundation for Financial Support in the Collection of Information. Ministry of Science, Technology, and Innovation (Colombia) for a Bicentennial Grant from the 2021 Fund. Data Availability The data are available from the corresponding author upon reasonable request. References Adeoye-Olatunde, O. A., & Olenik, N. L. (2021). Research and scholarly methods: Semi-structured interviews. Journal of the american college of clinical pharmacy, 4(10), 1358-1367. Alcaldía de Ibagué. (2022). Directorio Juntas de acción Comunal[WWW Document]. URL https://ibague.gov.co/portal/seccion/contenido/index.php?type=2&cnt=72#gsc.tab=0 Barraclough, A. D., Cusens, J., & Måren, I. E. (2022). Mapping stakeholder networks for the co-production of multiple ecosystem services: A novel mixed-methods approach. Ecosystem Services, 56, 101461. BayesFusion, L. L. C. (2017). Genie modeler. User Manual. Available online: https://support. bayesfusion. com/docs/(accessed on 21 October 2019), 16, 30-32. Berrío, M. F., & Montenegro, N. (2022). Evaluación de servicios ecosistémicos y su incidencia en el valor del suelo en la cuenca del río coello, Tolima (Bachelor's thesis). Universidad Distrital. Berrouet, L., Villegas-Palacio, C., & Botero, V. (2020). Vulnerability of rural communities to change in an ecosystem service provision: Surface water supply. A case study in the northern Andes, Colombia. Land Use Policy, 97, 104737. Blanco-Moreno, C., Amaya-Domínguez, J., & Bernal-Pedraza, A. Y. (2024). Research Paper Experiences of community-based water supply organizations partnerships in rural areas of Colombia. Journal of Water, Sanitation and Hygiene for Development, 14(4), 332. Bruen, M., Hallouin, T., Christie, M., Matson, R., Siwicka, E., Kelly, F., Bullock, C., Feeley, H. B., Hannigan, E., & Kelly-Quinn, M. (2022). A Bayesian Modelling Framework for Integration of Ecosystem Services into Freshwater Resources Management. Environmental Management, 69(4), 781–800. https://doi.org/10.1007/s00267-022-01595-x Cardenas, J. C., Rodriguez, L. A., & Johnson, N. (2011). Collective action for watershed management: field experiments in Colombia and Kenya. Environment and Development Economics, 16(3), 275-303. Carriger, J. F., Barron, M. G., & Newman, M. C. (2016). Bayesian networks improve causal environmental assessments for evidence-based policy. Environmental science & technology, 50(24), 13195-13205. Castro, A. J., Vaughn, C. C., Julian, J. P., & García-Llorente, M. (2016). Social demand for ecosystem services and implications for watershed management. JAWRA Journal of the American Water Resources Association, 52(1), 209-221. CBD (2018). Expert workshop on marine protected areas and other effective area-based conservation measures for achieving aichi biodiversity target 11 in marine and coastal areas (Montreal, Canada: Convention on Biological Diversity (CBD)). Available at: https://www.cbd.int/doc/c/d1d1/ff04/1adecc670d5f789c47c747b7/mcb-em-2018-01-03-en.docx. 6-9 February 2018CBD/MCB/EM/2018/1/3. Cole, B., Bradley, A. V., Willcock, S., Gardner, E., Allinson, E., Hagen-Zanker, A., & Whelan, M. (2023). Using a multi-lens framework for landscape decisions. People and Nature. Colón López, J., & Restrepo, C. (2019). Water Quality and Socio-Economic Indicators are Linked in a Tropical Watershed: Emerging Implications for the Sustainable Management of Waterscapes. Wetlands, 39(6), 1303–1316. https://doi.org/10.1007/s13157-019-01232-0 Cortolima. (2019). Ajuste parcial al Plan de Ordenación y Manejo de la subzona hidrográfica del río Coello. Corporación Autónoma Regional del Tolima - CORTOLIMA [WWW Document]. URL https://cortolima.gov.co/planes-y-programas/gestion-integral-del-recurso-hidrico/3377-plan-de-ordenamiento-y-manejo-de-cuenca-pomca-rio-coello Cortolima. (2020). Corporación Autónoma Regional del Tolima - CORTOLIMA [WWW Document]. URL https://datosabiertos.cortolima.gov.co/datasets/9659bd2b87e44869b05402f61e914789_0/explore Cosens, B., Ruhl, J. B., Soininen, N., Gunderson, L., Belinskij, A., Blenckner, T., ... & Similä, J. (2021). Governing complexity: Integrating science, governance, and law to manage accelerating change in the globalized commons. Proceedings of the National Academy of Sciences, 118(36), e2102798118. Crevier, L. P., & Parrott, L. (2019). Synergy between adaptive management and participatory modelling: The two processes as interconnected spirals. Ecological Informatics, 53, 100982. DANE. (2018). DANE Geoportal [WWW Document]. URL https://geoportal.dane.gov.co/geovisores/sociedad/cnpv-2018/ DANE. (2022). DANE Geoportal [WWW Document]. URL https://www.dane.gov.co/index.php/estadisticas-por-tema/pobreza-y-condiciones-de-vida/pobreza-multidimensional Daniel, D., Al Djono, T. P., & Iswarani, W. P. (2023). Factors related to the functionality of community-based rural water supply and sanitation program in Indonesia. Geography and Sustainability , 4 (1), 29-38. Delacámara, G., O’Higgins, T. G., Lago, M., & Langhans, S. (2020). Ecosystem-based management: moving from concept to practice. In Ecosystem-based management, ecosystem services and aquatic biodiversity: Theory, tools and applications (pp. 39-60). Cham: Springer International Publishing. Delucchi, A. A. S. (2024). Unpacking ‘community water management’ in rural Chile: An institutional ethnography. Journal of Rural Studies, 110, 103345. Di Cintio, A., Fernandes-Salvador, J. A., Puntila-Dodd, R., Granado, I., Niccolini, F., & Bulleri, F. (2024). Socio-economic factors boosting the effectiveness of marine protected areas: A Bayesian network analysis. Ecological Informatics , 84 , 102879. Forio, M. A. E., Villa-Cox, G., Van Echelpoel, W., Ryckebusch, H., Lock, K., Spanoghe, P., ... & Goethals, P. L. (2020). Bayesian Belief Network models as trade-off tools of ecosystem services in the Guayas River Basin in Ecuador. Ecosystem Services, 44, 101124. Francesconi, W., Srinivasan, R., Pérez-Miñana, E., Willcock, S. P., & Quintero, M. (2016). Using the Soil and Water Assessment Tool (SWAT) to model ecosystem services: A systematic review. Journal of Hydrology, 535, 625-636. https://doi.org/10.1016/j.jhydrol.2016.01.034 González-Redin, J., Luque, S., Poggio, L., Smith, R., & Gimona, A. (2016). Spatial Bayesian belief networks as a planning decision tool for mapping ecosystem services trade-offs on forested landscapes. Environmental Research, 144, 15–26. Guo, X., Wang, L., Fu, Q., & Ma, F. (2024). Ecological Function Zoning Framework for Small Watershed Ecosystem Services Based on Multivariate Analysis from a Scale Perspective. Land, 13(7), 1030. Hallberg-Sramek, I., Nordström, E.-M., Priebe, J., Reimerson, E., Mårald, E., & Nordin, A. (2023). Combining scientific and local knowledge improves evaluating future scenarios of forest ecosystem services. Ecosystem Services, 60, 101512. https://doi.org/10.1016/j.ecoser.2023.101512 Hijmans, R. J., Bivand, R., Forner, K., Ooms, J., Pebesma, E., & Sumner, M. D. (2022). Package ‘terra.’ Maintainer: Vienna, Austria. Höfer, S., Ziemba, A., & El Serafy, G. (2020). A Bayesian approach to ecosystem service trade-off analysis utilizing expert knowledge. Environment Systems and Decisions, 40(1), 67-83. Hölting, L., Jacobs, S., Felipe-Lucia, M. R., Maes, J., Norström, A. V., Plieninger, T., & Cord, A. F. (2019). Measuring ecosystem multifunctionality across scales. Environmental Research Letters, 14(12), 124083. https://doi.org/10.1088/1748-9326/ab5ccb Johnson, N., García, J., Rubiano, J. E., Quintero, M., Estrada, R., Mwangi, E., ... & Granados, S. (2008). Watershed management and poverty alleviation in the Colombian Andes. Spatial & Economic Analysis for Policy and Decision Support in Agriculture and Environment, 67. Landuyt, D., Broekx, S., & Goethals, P. L. (2016). Bayesian belief networks to analyse trade-offs among ecosystem services at the regional scale. Ecological Indicators, 71, 327–335. Laurila-Pant, M., Mäntyniemi, S., Venesjärvi, R., & Lehikoinen, A. (2019). Incorporating stakeholders' values into environmental decision support: A Bayesian Belief Network approach. Science of the Total Environment , 697 , 134026. Maragkaki, A., Koukianaki, E. A., Lilli, M. A., Efstathiou, D., & Nikolaidis, N. P. (2024). Optimizing the water-ecosystem-food nexus using nature-based solutions at the basin scale. Frontiers in Water, 6, 1386925. Marcot, B. G., Steventon, J. D., Sutherland, G. D., & McCann, R. K. (2006). Guidelines for developing and updating Bayesian belief networks applied to ecological modeling and conservation. Canadian Journal of Forest Research. https://doi.org/10.1139/x06-135 Martínez-López, J., Bagstad, K. J., Balbi, S., Magrach, A., Voigt, B., Athanasiadis, I., ... & Villa, F. (2019). Towards globally customizable ecosystem service models. Science of the Total Environment, 650, 2325-2336. Masante, D. (2016). Package ‘bnspatial.’. Spatial Implementation of Bayesian Networks and Mapping. http://github.com/dariomasante/bnspatial Meraj, G., Singh, S. K., Kanga, S., & Islam, M. N. (2022). Modeling on comparison of ecosystem services concepts, tools, methods and their ecological-economic implications: A review. Modeling Earth Systems and Environment, 8(1), 15-34. Metzger, J. P., Villarreal-Rosas, J., Suárez-Castro, A. F., López-Cubillos, S., González-Chaves, A., Runting, R. K., Hohlenwerger, C., & Rhodes, J. R. (2021). Considering landscape-level processes in ecosystem service assessments. Science of The Total Environment, 796, 149028. Omarova, A., Tussupova, K., Hjorth, P., Kalishev, M., & Dosmagambetova, R. (2019). Water Supply Challenges in Rural Areas: A Case Study from Central Kazakhstan. International Journal of Environmental Research and Public Health, 16(5), Article 5. https://doi.org/10.3390/ijerph16050688 Pereira, P., Inacio, M., Barcelo, D., & Zhao, W. (2025). Ecosystem services mapping and modelling. Where is the validation?. Geography and Sustainability, 6(3), 100286. Pham, H. V., Sperotto, A., Furlan, E., Torresan, S., Marcomini, A., & Critto, A. (2021). Integrating Bayesian Networks into ecosystem services assessment to support water management at the river basin scale. Ecosystem Services. https://doi.org/10.1016/j.ecoser.2021.101300 Phan, T. D., Smart, J. C., Stewart-Koster, B., Sahin, O., Hadwen, W. L., Dinh, L. T., ... & Capon, S. J. (2019). Applications of Bayesian networks as decision support tools for water resource management under climate change and socio-economic stressors: a critical appraisal. Water, 11(12), 2642. Pilogallo, A., & Scorza, F. (2022). Ecosystem services multifunctionality: An analytical framework to support sustainable spatial planning in Italy. Sustainability, 14(6), 3346. https://doi.org/10.3390/su14063346 Platarrueda, I. (2022). Community aqueducts in Colombia and their struggle for legal recognition: A political ecology approach. Perfiles Económicos. https://doi.org/10.22370/pe.2022.13.3445 Posner, S., Verutes, G., Koh, I., Denu, D., & Ricketts, T. (2016). Global use of ecosystem service models. Ecosystem Services, 17, 131-141. https://doi.org/10.1016/j.ecoser.2015.12.003 QGIS.org (2023). QGIS Geographic Information System. QGIS Association. http://www.qgis.org Ricart, S., & Rico-Amorós, A. M. (2021). Water for food, water for birds: How to manage conflicting rural-natural interfaces? Deepening on the socio-ecological system of El Hondo Natural Park (Alicante, Spain). Journal of Rural Studies, 86, 24-35. Rodríguez-Loinaz, G., Alday, J. G., & Onaindia, M. (2015). Multiple ecosystem services landscape index: A tool for multifunctional landscapes conservation. Journal of Environmental Management, 147, 152–163. https://doi.org/10.1016/j.jenvman.2014.09.001 N.N. Ruhl, J., Salzman, J., Arnold, C. A., Craig, R., Hirokawa, K., Olander, L., Palmer, M., & Ricketts, T. H. (2021). Connecting ecosystem services science and policy in the field. Frontiers in Ecology and the Environment, 19(9), 519–525. https://doi.org/10.1002/fee.2390 Rumbold, D. G. (2023). Use of a Bayesian network as a decision support tool for watershed management: A case study in a highly managed river-dominated estuary. Environmental Monitoring and Assessment , 195 (6), 741. Salliou, N., Barnaud, C., Vialatte, A., & Monteil, C. (2017). A participatory Bayesian Belief Network approach to explore ambiguity among stakeholders about socioecological systems. Environmental Modelling & Software, 96, 199–209. https://doi.org/10.1016/j.envsoft.2017.06.050 Scrieciu, A., Pagano, A., Coletta, V. R., Fratino, U., & Giordano, R. (2021). Bayesian belief networks for integrating scientific and stakeholders’ knowledge to support nature-based solution implementation. Frontiers in Earth Science, 9, 674618. Scutari, M., & Denis, J. B. (2021). Bayesian networks: with examples in R. Chapman and Hall/CRC. Seifert-Dähnn, I., Barkved, L. J., & Interwies, E. (2015). Implementation of the ecosystem service concept in water management - Challenges and ways forward. Sustainability of Water Quality and Ecology, 5, 3-8. https://doi.org/10.1016/j.swaqe.2015.01.007 Shao, Y., Xiao, Y., Kou, X., & Sang, W. (2023). Sustainable land use scenarios generated by optimizing ecosystem distribution based on temporal and spatial patterns of ecosystem services in the southern China hilly region. Ecological Informatics, 78, 102275. Sun, X., Ye, D., Shan, R., Peng, Q., Zhao, Z., & Sun, J. (2022). Effect of physical geographic and socioeconomic processes on interactions among ecosystem services based on machine learning. Journal of Cleaner Production, 359, 131976. https://doi.org/10.1016/j.jclepro.2022.131976 UNEP-WCMC (2023). Protected Area Profile for Cuenca hidrográfica del río Coello from the World Database on Protected Areas, September 2023. Available at: www.protectedplanet.net Vollmer, D., Burkhard, K., Adem Esmail, B., Guerrero, P., & Nagabhatla, N. (2022). Incorporating Ecosystem Services into Water Resources Management—Tools, Policies, Promising Pathways. Environmental Management, 69(4), 627–635. https://doi.org/10.1007/s00267-022-01640-9 Wu, C., Qiu, D., Gao, P., Mu, X., & Zhao, G. (2022). Application of the InVEST model for assessing water yield and its response to precipitation and land use in the Weihe River Basin, China. Journal of Arid Land, 14(4), 426–440. https://doi.org/10.1007/s40333-022-0013-0 Zeng, L., & Li, J. (2019). A Bayesian belief network approach for mapping water conservation ecosystem service optimization region. Journal of Geographical Sciences, 29, 1021-1038. Zhang, W., ElDidi, H., Masuda, Y. J., Meinzen-Dick, R. S., Swallow, K. A., Ringler, C., ... & Aldous, A. (2023). Community-based conservation of freshwater resources: learning from a critical review of literature and case studies. Society & Natural Resources, 36(6), 733-754. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 04 Apr, 2026 Read the published version in Modeling Earth Systems and Environment → Version 1 posted Editorial decision: Revision requested 30 Sep, 2025 Reviews received at journal 22 Sep, 2025 Reviewers agreed at journal 26 Aug, 2025 Reviewers invited by journal 23 Aug, 2025 Editor assigned by journal 18 Aug, 2025 Submission checks completed at journal 18 Aug, 2025 First submitted to journal 15 Aug, 2025 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-7382296","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":506986765,"identity":"5ed10caf-d95b-49f3-81cc-14358395b1a1","order_by":0,"name":"Cesar Rojas","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABFElEQVRIie2RsWrDMBBAzwjsRUXrGUq/QWAohZbmVxQEzuKhUMhUjKZkKc2a3yj9AQtBpzSzpxJT8JwuxR1aehoDtpuxUL3p0N3T3UkAgcBfBQGYAJAVxZxiAxAfoaTGKwo4h+gIxUP1EkiBXxWxfGl2F3dlktX60X52r6eTxBrYzx2lTP9Mm1km8dmx8zq/cVzdcs6nJlpvHaWq/nkgjxHjipRCOlCKdpkadrJwIFH1K6Il5btk2bqQtvOKaAz7GlOQuqQLxiQWsuJeQeoSjShYt0ymD47hpqVdclLqxtj77YwP7SJWefSGH6UWS/303l2pSbLSdtfNL8+GXszD6Cv1wYm/ng/WE9Ee4HqsIBAIBP45PwuvWNRCObR7AAAAAElFTkSuQmCC","orcid":"","institution":"Universidad Distrital Francisco José de Caldas","correspondingAuthor":true,"prefix":"","firstName":"Cesar","middleName":"","lastName":"Rojas","suffix":""},{"id":506986766,"identity":"0fbc51ba-7c5e-4e57-acc9-f0dee47fe724","order_by":1,"name":"Riascos-Ochoa Javier","email":"","orcid":"","institution":"Universidad de Bogotá Jorge Tadeo Lozano","correspondingAuthor":false,"prefix":"","firstName":"Riascos-Ochoa","middleName":"","lastName":"Javier","suffix":""},{"id":506986767,"identity":"fb00b39e-f2e3-49c9-9ae0-0558b63acef7","order_by":2,"name":"Nicola Clerici","email":"","orcid":"","institution":"Universidad del Rosario","correspondingAuthor":false,"prefix":"","firstName":"Nicola","middleName":"","lastName":"Clerici","suffix":""},{"id":506986768,"identity":"63377e9a-b5fe-42fd-8907-e79b3476a695","order_by":3,"name":"Magnolia Longo","email":"","orcid":"","institution":"Universidad de Bogotá Jorge Tadeo Lozano","correspondingAuthor":false,"prefix":"","firstName":"Magnolia","middleName":"","lastName":"Longo","suffix":""}],"badges":[],"createdAt":"2025-08-15 14:53:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7382296/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7382296/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s40808-026-02735-6","type":"published","date":"2026-04-04T15:58:51+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":90381345,"identity":"90c4c479-623f-480b-b186-c27130a8c671","added_by":"auto","created_at":"2025-09-02 06:47:36","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":554759,"visible":true,"origin":"","legend":"\u003cp\u003eCoello River Basin and Rural Water Supply Systems (blue dots) in Tolima, Colombia.\u003c/p\u003e","description":"","filename":"image1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7382296/v1/cc8636115b18e3c655f9e289.jpeg"},{"id":90381344,"identity":"8a1e93cd-6d37-4b18-bffb-a914ff48cad4","added_by":"auto","created_at":"2025-09-02 06:47:36","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":67632,"visible":true,"origin":"","legend":"\u003cp\u003eMethodological approach for incorporating local perceptions into multifunctional zones. (Bayesian Belief Networks-BBN). Blue boxes = Data input stages; Orange boxes = Analytical/modeling stages; Green boxes = Outputs/products.\u003c/p\u003e","description":"","filename":"image2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7382296/v1/73202003368d3409486cc312.jpg"},{"id":90383763,"identity":"3feead3c-07d3-4794-8559-4f2f1fc5a901","added_by":"auto","created_at":"2025-09-02 07:03:36","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":172768,"visible":true,"origin":"","legend":"\u003cp\u003eModel 1 BBN of multifunctional ecosystem service zones for RWSS in the Coello watershed, Colombia. (AWY) Annual water yield, (CA) crops, (CS) Carbon storage, (ND) Nutrient delivery ratio, (HQ) Habitat quality, (P) Pollination, (R) Recreation, (IM) Mesli Index, (WUI) Water use Index, (LUCCI) Land use Index, (LWC) Liters of Water concessioned, (AMO) Type off aqueduct management, (LPO) Local participation scenarios, (EB) Equity in the use of benefits, (SPG) Scenario of local participation, (SPG) Supply of goods for public use, (MPI) Multidimensional poverty index, (NTR) Natural territorial limitation. Green squares: Supply of ES, Orange squares: Demand of ES; Blue: Socioeconomic variables; Yellow: zones of multifunctional of ecosystem services, (ZPMES).\u003c/p\u003e","description":"","filename":"image3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7382296/v1/9ad8366e5098972cbc088027.jpeg"},{"id":90383303,"identity":"ea13977f-2df5-438f-802a-dcc908cc8cd4","added_by":"auto","created_at":"2025-09-02 06:55:36","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":164978,"visible":true,"origin":"","legend":"\u003cp\u003eBayesian network sensitivity analysis of RWSS in the Coello River watershed. Red: More sensitive nodes.\u003c/p\u003e","description":"","filename":"image4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7382296/v1/efcaf97a71cb808ae97413e0.jpeg"},{"id":90384657,"identity":"52657e21-21d8-4b5b-a9c7-67222e27a01a","added_by":"auto","created_at":"2025-09-02 07:11:36","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":252955,"visible":true,"origin":"","legend":"\u003cp\u003eModel 2 BBN uses perceptions of conservation in the rural aqueduct zones of the Coello river watershed. (AWY) Annual water yield, (CA) crops, (CS) Carbon storage, (ND) Nutrient delivery ratio, (HQ) Habitat quality, (P) Pollination, (R) Recreation, (IM) Mesli Index, (WUI) Water use Index, (LUCCI) Land use Index, (LWC) Liters of Water concessioned, (AMO) Type off aqueduct management, (SLP) Local participation scenarios, (EB) Equity in the use of benefits, (SPG) Supply of goods for public use, (MPI) Multidimensional poverty index, (NTR) Natural territorial limitation. (P1) Range of aqueduct users, (P2) agricultural activity (P3) Perception of frequency of water supply, (P4) Perception of quality of water supply, (P5) conservation of nature in rural aqueduct, (ZPMSE) Zones of Multifunctional Ecosystem Service. Green squares: Supply of ES; Orange squares: demand for ES; Blue: Socioeconomic aspects; Red: local perceptions, and yellow: zones for multifunctional ecosystem services.\u003c/p\u003e","description":"","filename":"image5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7382296/v1/c335b805ce75037dadd15574.jpeg"},{"id":90381347,"identity":"f2797a8a-90f5-4a24-8222-45b531eb77d1","added_by":"auto","created_at":"2025-09-02 06:47:36","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":611328,"visible":true,"origin":"","legend":"\u003cp\u003eRegions within the Coello River watershed with the most likely ES-multifunctionality states.\u003c/p\u003e","description":"","filename":"image6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7382296/v1/81701036960b6c77c09cf0ea.jpeg"},{"id":90383765,"identity":"42cac0c3-fb29-44c3-a745-1f5ce6cb6d4a","added_by":"auto","created_at":"2025-09-02 07:03:37","extension":"jpeg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":716582,"visible":true,"origin":"","legend":"\u003cp\u003eUncertainty map of ES multifunctionality in the Coello River watershed\u003c/p\u003e","description":"","filename":"image7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7382296/v1/3889c129653ad8e1ee2f931b.jpeg"},{"id":90381365,"identity":"2715eee9-ffef-4472-9aa2-26bfeffb6a54","added_by":"auto","created_at":"2025-09-02 06:47:37","extension":"jpeg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":495311,"visible":true,"origin":"","legend":"\u003cp\u003eAreas of higher conservation relevance in ES multifunctionality zones in the Coello River watershed.\u003c/p\u003e","description":"","filename":"image8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7382296/v1/79d8cf94a47051a2624bafc4.jpeg"},{"id":106344233,"identity":"d3bee196-d59d-41e2-80d9-2c70ff7e798c","added_by":"auto","created_at":"2026-04-07 16:12:53","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3593684,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7382296/v1/d867d3f9-abd2-455f-8115-631eeb9707c6.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eIntegrating Local Perceptions in Bayesian Belief Network for Watershed Planning: A Case in the Colombian Andes\u003c/p\u003e","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eWater security remains a pressing global challenge, particularly in mountainous regions, where climate variability, land-use change, and socioeconomic pressures intersect. Achieving sustainable water governance in such contexts requires approaches that integrate ecological processes with social and institutional systems that mediate water access. Rural water supply systems (RWSS) in the Andean Mountains are central to this challenge. They operate in ecologically valuable micro-watersheds and provide multiple ecosystem services (ES), such as water provision, sediment retention, and recreational opportunities. In many regions, RWSS are important socio-ecological systems that connect biodiversity conservation and community-based management. However, complementary methods for integrative spatial analysis of ecological functions and socioeconomic characteristics are needed (Delucchi \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Omarova et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe implementation of ES\u0026ndash;watershed modeling approaches has different limitations that reduce their use in IWM in rural areas. First, most frameworks, including widely utilized tools such as the Integrated Ecosystem Services and Trade-offs Tool (InVEST) (Posner et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), Artificial Intelligence for Ecosystem Services (ARIES) (Mart\u0026iacute;nez-L\u0026oacute;pez et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), and Soil and Water Assessment Tool (SWAT) (Francesconi et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), predominantly emphasize biophysical processes, with minor consideration of socioeconomic conditions such as governance and local stakeholder perspectives. Second, decision-making confidence is limited in data-scarce regions because of the underrepresentation of uncertainty in model outputs, which can be caused by scale mismatches or data quality. Finally, these ES models seldom integrate ecological and social variables within a unified analytical framework, resulting in spatial priorities that may not align with the local community\u0026rsquo;s interests. It is necessary to have flexible methods for advancing the limits of ES mapping that explicitly incorporate governance, socioeconomic indicators, and ecosystem functions in a spatially explicit manner that includes uncertainty in their outcomes.\u003c/p\u003e\u003cp\u003eBayesian Belief Networks (BBNs) are an integrative modeling framework for studying complex socio-ecological systems such as watersheds. BBNs are probabilistic graphical representations that employ nodes and conditional probability tables (CPTs) to describe the relationships among variables. They are valuable analytical tools for synthesizing diverse quantitative (e.g., ecological indicators) and qualitative (e.g., stakeholder perceptions) data, enabling the representation of uncertainty and nonlinear relationships (Landuyt et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), which characterize socio-ecological systems (Scutari and Denis \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). These BBN models can support scenario development (Phan et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Rumbold \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), identify ES trade-offs and synergies (H\u0026ouml;fer et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), and promote participatory planning (Daniel et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Salliou et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Nevertheless, further investigation is needed to effectively apply BBNs for mapping ES multifunctionality, including local conservation preferences and governance forms in rural Andean contexts (Pham et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Scrieciu et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Seifert-D\u0026auml;hnn et al. \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2015\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIntegrative spatial analysis contributes to identifying strategic zones in watersheds that deliver multiple ES supplies (ES multifunctionality) in RWSS. In this context, BBNs are a relevant framework that supports the integration of ecological models with community knowledge, creating maps that are both scientific and locally relevant to IWM. Studying ES multifunctionality in RWSS shows the ecological, social, and economic values of ecosystems (H\u0026ouml;lting et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). In this context, mapping ES multifunctionality supports IWM by identifying how ES synergies and trade-offs are included in water-use decisions (Cole et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Pilogallo and Scorza \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, little research has been conducted on the implementation of ES multifunctionality in RWSS zoning, specifically in mapping high-value zones of supply and demand for ES (Rodr\u0026iacute;guez-Loinaz et al., \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Ruhl et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). In this context, according to recent studies, it is necessary for integrative analysis of ES better spatial ES data, community input, and land-use patterns (Bruen et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Forio et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Zeng and Li \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Similarly, global sustainability goals, such as SDG 6 (Clean Water and Sanitation) and SDG 15 (Life on Land), and local governance both benefit from addressing this issue. Watershed-level tools are necessary to achieve these objectives. The ecological and social facets of water management require comparable approaches (Shao et al., \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThis challenge is particularly relevant in Colombia. In rural Colombia, the water supply relies predominantly on local community organizations to ensure an adequate water supply and timely management of water resources (Blanco-Moreno et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Delac\u0026aacute;mara et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Platarrueda \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Water-use management in the Andean region of Colombia, is characterized by complex relationships between ecosystem benefits, traditional practices, and socioeconomic conditions that underpin the well-being of river basin water users (Cardenas et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Johnson et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). The objective of this study is to develop and apply a participatory spatial BBN framework that integrates ecological indicators, socio-economic metrics, and governance perceptions to support multifunctional ES zoning in RWSS areas of the Coello River Basin, Colombia\u003c/p\u003e\u003cp\u003eWe posed the following research question: Do local conservation preferences within multifunctional zones in RWSS improve IWM in watershed planning? To address this, we aimed to (i) develop a spatial BBN for RWSS to assess ES multifunctionality. (ii) integrate ecological and socioeconomic attributes derived from local perceptions, and (iii) represent ES multifunctional zones within land-use planning instruments in watersheds. By bridging ecological modeling with participatory governance, our approach aims to advance inclusive, data-informed watershed planning in rural Andean landscapes and offers a replicable method for other socio-ecological systems. The resulting framework not only quantifies multifunctional ES zones but also identifies areas where conservation priorities align or conflict with governance capacities, offering decision-relevant insights for IWM.\u003c/p\u003e"},{"header":"MATERIALS AND METHODS","content":"\u003cp\u003eStudy area\u003c/p\u003e\u003cp\u003eThe Coello River watershed is located in the Andean region of Colombia, South America, and has an area of 1,817.7 km\u003csup\u003e2\u003c/sup\u003e, and is located within the Department of Tolima (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Land use and land cover (LULC) were distributed as follows: 88.7% forest and semi-natural areas, 0.39 water bodies, 9.41% agricultural land, and 1.42% artificial surfaces. The Watershed Management Plan (POMCA) designates up to 25% of the areas of high conservation importance that are protected under schemes such as natural parks, bird sanctuaries (AICAS), and forest reserves. These areas contain tropical dry forest relics, paramo ecosystems, and wetlands are present in these areas (Cortolima, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The land-use patterns in the Coello River watershed directly influence water supply and demand, which are crucial for sustainable watershed management.\u003c/p\u003e\u003cp\u003eThe basin's tropical dry forests, high mountain ecosystems, and wetlands significantly contribute to their ecological value. The main ES in the basin is water yield, which varies between 17.98 l/s/km\u0026sup2; during the dry season and 40.63 l/s/km\u0026sup2; in the rainy season. The total water demand is 28,233.76 l/s, distributed among 881 concession users. The watershed supports a population of approximately 264,717, mainly concentrated in the municipalities of Ibagu\u0026eacute; and Cajamarca, which together host 91% of the population (Cortolima, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). This includes 156 RWSS, which serves 23.97% of the rural population. These RWSS play a critical role in decentralized water governance by supplying both domestic and agricultural water.\u003c/p\u003e\u003cp\u003e1.1 Overview of the Methodological Framework\u003c/p\u003e\u003cp\u003eThis study followed a four-stage methodology to integrate quantitative ES indicators and qualitative stakeholder perceptions into a spatial BBN to map ES multifunctionality: (1) ES indicator modeling, quantification of ES supply, demand, and socio-ecological context; (2) local perceptions mapping, stakeholder interviews to understand community knowledge, values, and concerns; (3) BBN development and validation: two models were developed by combining quantitative and qualitative data; and (4) zoning integration: BBN results were combined with planning instruments to delineate ES multifunctionality zones (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThis approach supports sustainability planning by combining geo-data and participatory knowledge for IWM in watershed governance.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eMethods\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eIndicators of Ecosystem Services and Spatial Modeling\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eTo characterize ES multifunctionality, we selected indicators based on the relevance to represent three dimensions: supply, demand, and socio-ecological access (Table\u0026nbsp;1). Spatial modeling was conducted using the InVEST and ARIES tools to map ES. Indicators included water yield, nutrient retention, carbon sequestration, pollination, habitat quality, and recreational value. The Mesli Index (Berr\u0026iacute;o and Montenegro, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and species distribution models were also included to reflect the ecological conditions (N. N. et al. 2024).\u003c/p\u003e\u003cp\u003eThe demand-side indicators were derived from the LUCC data (Cortolima, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), a water-use index (Cortolima, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), and risk management layers (e.g., volcanic and flood zones). The socioeconomic context was captured using the Multidimensional Poverty Index and spatial distribution of public services (schools and hospitals).\u003c/p\u003e\u003cp\u003eAll variables were resampled to 30 m resolution, normalized (0\u0026ndash;1 scale), and classified into ordinal states (e.g., low, medium, and high) for BBN input. This selection ensures that the BBN represents the coupled human\u0026ndash;natural system rather than focusing solely on biophysical processes. The spatial data were managed using QGIS 2.3 (QGIS.org, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Taba\" border=\"1\"\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCode\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNode\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eDescription\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStates\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eSource\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAWY\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAnnual water yield\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eThe InVEST model was used to determine the relative contributions of water production in different watershed areas\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eVery High (\u0026gt;\u0026thinsp;75th), High (50\u0026ndash;75%), Moderate (25\u0026ndash;50%), Low (10\u0026ndash;25%), None (\u0026lt;\u0026thinsp;10%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\" morerows=\"7\" rowspan=\"8\"\u003e\u003cp\u003eBerr\u0026iacute;o and Montenegro (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCrop areas\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eAgricultural production zones in watershed according to LULC maps (Cortolima, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2019\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes (1), No (0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCS\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCarbon storage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eInVEST model for estimating carbon storage by land cover\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eHigh (\u0026gt;\u0026thinsp;75 Mg C/ha), Moderate (40\u0026ndash;75), Low (\u0026lt;\u0026thinsp;40)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eND\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNutrient delivery ratio\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eInVEST model estimates nutrient sources and transport to the main channel, allowing evaluation of the role of vegetative cover retention in watershed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eVery High (\u0026gt;\u0026thinsp;0.8), High (0.6\u0026ndash;0.8), Moderate (0.4\u0026ndash;0.6), Low (0.2\u0026ndash;0.4), None (\u0026lt;\u0026thinsp;0.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHQ\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHabitat quality\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eInVEST model representing the relationship between LULC maps and threats to biodiversity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eHigh (\u0026gt;\u0026thinsp;0.7), Moderate (0.4\u0026ndash;0.7), Low (0.2\u0026ndash;0.4), None (\u0026lt;\u0026thinsp;0.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eP\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ePollination\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eThe ARIES model evaluates the supply and demand of pollination services by insects based on three components: land cover, crops, and weather conditions\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eHigh (\u0026gt;\u0026thinsp;0.66), Moderate (0.33\u0026ndash;0.66), Low (\u0026lt;\u0026thinsp;0.33), None (no data)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRecreation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eARIES model estimates of ES based on recreational use areas\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eHigh (\u0026gt;\u0026thinsp;0.66), Moderate (0.33\u0026ndash;0.66), Low (\u0026lt;\u0026thinsp;0.33), None (no data)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMesli Index (Multiple ecosystem services)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIndicator of areas with the highest ES\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eVery High (\u0026gt;\u0026thinsp;0.75), High (0.6\u0026ndash;0.75), Moderate (0.4\u0026ndash;0.6), Low (0.2\u0026ndash;0.4), None (\u0026lt;\u0026thinsp;0.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWUI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWater Use Index\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIndicator of areas with differing in water supply and demand.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eVery High (\u0026gt;\u0026thinsp;1.5), High (1.0\u0026ndash;1.5), Moderate (0.5\u0026ndash;1.0), Low (\u0026lt;\u0026thinsp;0.5), None (no data)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\" morerows=\"3\" rowspan=\"4\"\u003e\u003cp\u003eCortolima (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2020\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLUCCI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eIndex of Land use land cover\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eThe zones of LULC in the watershed are forest, agriculture, and urban\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAgriculture, Forest, Urban\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLWC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLiters of water\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eQuantity of water granted at each water intake\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eHigh (\u0026gt;\u0026thinsp;15 L/s), Moderate (5\u0026ndash;15 L/s), Low (\u0026lt;\u0026thinsp;5 L/s)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAMO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTypes of aqueduct management organizations\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eTypes of aqueduct management: private and public\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eCommunity (Yes), Private (No)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLPO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLocal participation scenarios\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eExistence of local participation instruments in the area, in accordance with the management of the aqueduct\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes, No\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eAlcald\u0026iacute;a de Ibagu\u0026eacute; (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEB\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eEquity in the use of benefits\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eCensus information with a representation of areas with water service access\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes (access\u0026thinsp;\u0026ge;\u0026thinsp;80%), No (access\u0026thinsp;\u0026lt;\u0026thinsp;80%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\" morerows=\"2\" rowspan=\"3\"\u003e\u003cp\u003eDANE (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2018\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSPG\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSupply of goods for public use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRural areas with educational or health services\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes (present within 1 km), No (absent)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMPI\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMultidimensional Poverty Index\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003ePercentage of people living in multidimensional poverty\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes (\u0026ge;\u0026thinsp;33%), No (\u0026lt;\u0026thinsp;33%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNTR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNatural territorial hazards\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eLand Use Restricted Zones related to hazards, flooding, torrential flooding, erosion, etc.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes, No\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCortolima (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2019\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eZPMES\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eZone of multifunctional Ecosystem Services\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eES Multifunctionality with improved ES supply and area for species distribution\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eVery High (\u0026gt;\u0026thinsp;0.8), High (0.6\u0026ndash;0.8), Moderate (0.4\u0026ndash;0.6), Low (0.2\u0026ndash;0.4), None (\u0026lt;\u0026thinsp;0.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eN.N et al. (2024)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eLocal perception and mapping sampling\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eA two-step strategy was used to map the local perceptions of the RWSS. (a) Demographic context: Population data for RWSS were obtained from an analysis of the catchment points in the Coello River Basin, as reported by Cortolima (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). This secondary data source was used to collect RWSS demographic information. After completing the demographic analysis, in-depth interviews were conducted with the RWSS presidents to enrich the demographic data with qualitative insights into local conservation perceptions. (b) Stakeholder Interviews: Fifteen RWSS presidents were selected using stratified snowball sampling, covering altitudinal gradients, management types, and multifunctionality values. The interviews were conducted in a semi-structured format (Adeoye-Olatunde and Olenik, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), combining a Likert-scale and open-ended questions. The themes included: ecosystem benefits, conservation roles, threats, organizational capacity, and equity in access. The conservation attitudes were quantified using Likert responses (1\u0026ndash;5) and then discretized to match the BBN state definitions. These leaders were selected to represent diverse altitudinal gradients (high, middle, and low zones), aqueduct management types (community-managed vs. private), and ecosystem service contexts. The questionnaire included Likert-scale items and open-ended questions designed to elicit the following responses: a) recognition of the role of forests and riparian areas in water regulation; b) level of community participation in conservation or reforestation activities; c) awareness of local environmental risks and land-use pressures; and c) organizational experience with education, governance, or water access equity. Data were collected between March and May 2023.\u003c/p\u003e\u003cp\u003e1.1.1 Bayesian Belief Network Model 1: Quantitative ES indicators\u003c/p\u003e\u003cp\u003eFollowing Marcot et al. (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), the BBN was built in three steps: i) structure and nodes: The model included 12 input nodes grouped into supply, demand, and access dimensions, all of which influenced the final ES multifunctionality output node. These were grouped into three intermediate components: supply, demand, and access. These components influence the final output node representing ES multifunctionality. ii) State Definitions: Based on the data distributions, each variable was classified into three to five ordinal states (Table\u0026nbsp;1). Thresholds for classification were defined using quantiles or expert-defined cutoffs depending on data availability and distribution. iii) Derivation of Conditional probability tables (CPT) derivation: We used hybrid learning: structure learning (hill-climbing\u0026thinsp;+\u0026thinsp;BIC) and parameter learning (MLE) on raster inputs via the bnlearn R package. Input data were extracted from 30 m layers and converted to categorical values using the thresholds listed in Table\u0026nbsp;1. iv) Sensitivity and Validation: The influence of variable in the BBN was assessed using GeNIe Modeler (v.2.1) and the logic and outcomes were validated through expert consultations.\u003c/p\u003e\u003cp\u003eAll spatial layers were rasterized to a 30 m resolution and intersected using QGIS. The thresholds listed in Table\u0026nbsp;1 were used to extract pixel-level values and categorize them into node states.\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eBayesian Belief Network Model 2: Local Perceptions\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eModel 2 extended Model 1 by incorporating variables on conservation awareness, institutional trust, and perceived threats that are informed by stakeholders. BBN Model 2 followed the following steps: i) Structure and Nodes: The second model was structured around qualitative data derived from interviews and participatory mapping. Nodes representing perceived threats, values assigned to specific ES (e.g., water quality and forest cover), and trust in institutions were included. It included nodes representing perceived threats, values assigned to specific ES (e.g., water quality and forest cover), and trust in institutions. The output node represented the perceived ES condition. ii) State Definition: Qualitative responses were coded and aggregated into three states per variable (e.g., not valued, moderately valued, and highly valued) using content frequency and coding triangulation. iii) CPT Derivation: The perception data collected from interviews were systematically coded to inform the CPT in our BBN model due to the qualitative nature of the inputs, Due to the qualitative nature of the inputs, the perception data collected from interviews were systematically coded to inform the CPT in our BBN model, thus directly translating local insights into the model\u0026rsquo;s structure. iv iv) Spatial BBN: BBN nodes were linked to 30 m raster layers using the bnSpatial package (Masante, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) in R. Software as described by Gonz\u0026aacute;lez-Redin et al. (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eiv) Uncertainty analysis and validation: The spatial uncertainty was assessed using the Shannon entropy index, which was applied to posterior probabilities at the pixel scale.\u003c/p\u003e\u003cp\u003eThis metric quantifies the degree of uncertainty on the information content of each prediction across the spatial domain. For each pixel, the entropy H was calculated using Eq.\u0026nbsp;1:\u003c/p\u003e\u003cp\u003eH(X) = - Σ P(x\u003csub\u003ei\u003c/sub\u003e) \u0026middot; log₂ P(x\u003csub\u003ei\u003c/sub\u003e) (1)\u003c/p\u003e\u003cp\u003eWhere:\u003c/p\u003e\u003cp\u003eP(x\u003csub\u003ei\u003c/sub\u003e) is the posterior probability of state x\u003csub\u003ei\u003c/sub\u003e,\u003c/p\u003e\u003cp\u003eWhere n is the number of states of the target node (e.g., Very High, High, Moderate, etc.).\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eIncorporating local conservation perceptions into environmental planning areas\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eWe overlaid the BBN outputs with existing zoning instruments: the POMCA plan, complementary conservation measures (OECM), and ES multifunctionality zones. Each layer received equal weight using the terra package of R software (Hijmans et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) to compute ZEMSE for strategic conservation planning as indicated in Eq.\u0026nbsp;\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e2\u003c/span\u003e. By applying uniform weights, we ensured that no single data source or conservation instrument dominated the zoning synthesis, thus preserving the multidimensional integration required for equitable and evidence-based IWM in watershed:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:ZEMSE=\\sum\\:_{i}^{n}{w}_{i}\\text{*}{x}_{i}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003ewhere,\u003c/p\u003e\u003cp\u003eZEMSE: ES multifunctionality zones for strategic conservation planning.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003ew\u003csub\u003ei\u003c/sub\u003e\u003c/strong\u003e\u003cp\u003eweight of each variable.\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003ex\u003csub\u003ei\u003c/sub\u003e\u003c/strong\u003e\u003cp\u003eEnvironmental zoning (from the POMCA instrument, complementary conservation areas).\u003c/p\u003e\u003c/p\u003e"},{"header":"RESULTS","content":"\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eModel 1. Bayesian network for ES multifunctionality zones\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe output of Model 1 was constructed to model the ES multifunctionality probabilities of in the Coello River Basin resulting from the ES supply and demand interaction (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). A BBN provides high-probability ES distribution for water, carbon sequestration, nutrient delivery, and recreation. ES supply nodes that contribute more with a high probability to an ES multifunctional zone include water supply, habitat quality, forest carbon storage, and nutrient retention nodes. A more probable ES supply affects the probability that an RWSS will exhibit high multifunctionality. The high probability of the ES supply highlights its ecological significance as an area for species habitat distribution in watersheds, with a higher probability of being a zone of ES multifunctionality.\u003c/p\u003e\u003cp\u003eThe ES demand node has a high probability distribution for the water and forest land-use indices. The node of liters of water concession has a low probability. A high probability of forest land use and water use index is likely to exhibit an ES multifunctionality zone. Additionally, the BBN considers the socioeconomic characteristics associated with the population\u0026rsquo;s current living conditions, such as the multidimensional poverty node, which indicates a high probability of unsatisfactory basic needs. The type of organization and local participation scenario revealed a high distribution of occurrences. Similarly, BBN incorporates territorial elements, such as land-use restrictions, which represent natural limitations on watershed use. Finally, a territorial equity node was included, indicating a significant probability that the population in the area would not have access to the natural and public land-use benefits.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eSensitivity analysis\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe sensitivity analysis showed that the variables with the highest influence on the ES multifunctionality output node were local participation in water planning, territorial access restrictions, and multidimensional poverty, each contributing to more than 60% uncertainty reduction. Additional variables with notable sensitivity included land-use type, equity in benefit distribution, supply of public goods, the Water Use Index, and the Mesli Index. The aggregated posterior probabilities for the ES multifunctionality output were 44%, 18%, and 10% for high, low, and for high multifunctionality, 18% for low multifunctionality, and 10% for zones lacking multifunctionality, respectively. In Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, the dark red variables significantly affect the variables shown in dark red significantly affect the output node. The variables in the red shade have less influence on the end node. The variables that most influenced the uncertainty of ES Multifunctionality Zones (probability\u0026thinsp;\u0026ge;\u0026thinsp;60%) were local participation in local planning, territorial access restrictions, and multidimensional poverty. Other variables with significant impacts include land use, equity in the use of benefits, supply of public goods, the water use index, and the Mesli index. The model showed a 44% probability of high ES multifunctionality, 18% probability of low ES multifunctionality, and 10% probability of zones without ES multifunctionality.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eModel 2 Preferences for conservation in ES multifunctionality zones\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eModel 2 (conservation preferences, Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) was updated and adjusted to Model 1 with new nodes acquired from local stakeholders\u0026rsquo; interviews. The RWSS serving 50\u0026ndash;100 people (29.34% of the total) had the highest participation in conservation activities, whereas those serving less than 50 people had the highest participation (23.7%). In this area, 75% of the rural aqueduct residents engage in agriculture, 23% engage in urban activities, and 5% participate in tourism. Among the aqueduct administrations, 70% are managed by local communities, and 15% are either shared or owned by private companies. Among the stakeholders, 17.4% participated in local conservation and restoration activities. 23% agreed with conservation activities in the area, and 47% agreed with the quality of the water supplied by the aqueduct, which is related to the variable \u0026ldquo;efficiency of water service provision\u0026rdquo;. Aqueduct organizations make collective efforts to manage their activities. Then operation involves self-financing the cost of its activities.\u003c/p\u003e\u003cp\u003eIn Model 2, when local perceptions were included, the probability of areas with ES multifunctionality increased from 35\u0026ndash;44% (with respect to Model 1), the probability of areas without multifunctionality decreased from 10\u0026ndash;8%, and the probability of areas with low MF increased from 18\u0026ndash;15%. Regarding the node on conservation perceptions, 25% of the respondents agreed with conservation, whereas 35% disagreed with conservation activities in RWSS.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e shows the mapping of ES multifunctionality zones. The map reveals that the upper section of the watershed, characterized by a greater extent of forest cover and regions situated close to the riverbed, exhibits the highest probability values for ES multifunctionality. The lower watershed includes forest areas that remain after clearing, and the present water catchment contributes to a higher probability of sustaining multiple ES. The rural water supply system, in which local administration leads to organization and land ownership, is important for water supply area conservation activities. These areas have a high likelihood of multifunctionality because of a positive combination of socioeconomic and ecological factors. In contrast, areas with a medium likelihood of multifunctionality were found in the watershed region, whereas remnant forests and areas with a low likelihood of multifunctionality were characterized by soil transformation caused by urban and industrial activities.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eModel 2, uncertainty.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe uncertainty map (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e) generated from Model 2 indicated spatial heterogeneity in the posterior probability distributions for ES multifunctionality across the Coello River watershed. Higher uncertainty values, represented by elevated entropy values, were concentrated in the lower watershed, particularly in urban expansion zones and areas dominated by agricultural activities. These areas exhibit greater divergence in the probability distribution among multifunctionality states, reflecting a lower model confidence in assigning a definitive ES condition. Instead, lower uncertainty values were associated with the upper watershed, where land cover is predominantly forested and includes protected ecosystems, national parks, and areas with territorial use restrictions (e.g., volcanic hazard zones). These zones showed a clear convergence toward the \u0026ldquo;high multifunctionality\u0026rdquo; state, with strongly skewed posterior probabilities. This pattern reflects stronger biophysical signals and more consistent stakeholder preferences. Intermediate uncertainty values appear in transition zones\u0026mdash;particularly in the mid-basin\u0026mdash;where agricultural land-use and forest remnants coexists. Here, the spatial heterogeneity in ES provision, land-use, and local governance practices results in a more even distribution of probabilities across ES multifunctionality states.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eOverall, the uncertainty map provides a critical complement to the most-likely state map by revealing areas where decision-making should be accompanied by additional stakeholder engagement, field validation, or targeted data collection due to model uncertainty. The most likely state of ES multifunctionality map was confirmed by identifying multifunctional zones and comparing them with current land use in the watershed based on LUCC in the watershed (Cortolima \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Persistence of land-use areas and likelihood of being highly functional according to key actors in the watershed. The water catchment areas, forested zones, and areas designated for conservation by local communities exhibited high ES multifunctionality. In contrast, areas in the lower watershed with a low probability of multifunctionality corresponded to zones with some fragments but without designated conservation areas without designated conservation areas but with some fragments. Furthermore, a local environmental authority was consulted in the basin regarding the spatial output to confirm that the areas displaying multifunctional ES on the map were indeed within the RWSS with forest vegetation.\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eMapping local perceptions of the multifunctionality of ecosystem services zones in environmental planning instruments\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe inclusion of node status zones with high multifunctionality complements the current watershed environmental zoning of the watersheds. This zoning is related to management plans, watershed management, and OECM and is called the Convention of Biodiversity in Decision 18/4 (CBD \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Figure\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e shows the regions with the highest probability of high ES multifunctionality based on the spatial results of the BBN. Figure\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e shows the regions with the highest probability of high ES multifunctionality. The regions with the highest multifunctionality occurrence of multifunctionality were in the higher watershed segment, which had considerable forest cover. These areas also display different forms of community organization regarding water use because agricultural activities are prevalent. Additionally, a more even-handed distribution of benefits was reported, and the provision of a water supply service was viewed as favorable. Moreover, these areas have topographic limitations that hinder changes in natural land use. In terms of the middle watershed areas, those with high probabilities of multifunctionality comprise zones with agricultural activity while simultaneously preserving forest cover. Similarly, basin riparian zones are noteworthy for their multifunctionality because they cover a substantial amount of land. Conversely, the lower part of the basin has the lowest ES multifunctionality. The proposed zoning method considers the areas influenced by the relevant urban zones. In addition, areas containing forest remnants were identified in the lower basin, which is fundamental for generating restoration strategies.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eThis study introduced a spatial BBN participatory framework that uses ES supply indicators, socioeconomic metrics, and local conservation perceptions into multifunctional ES zoning for the RWSS in the Coello River Basin, Colombia. The formulated models highlight the relationships between ES supply and demand in environmentally important areas, providing relevant local information of water uses for ES assessment (Carriger et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Vollmer et al. \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), collaborative spatial planning for water planning (Bruen et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), and community inputs in the implementation IWM (Guo et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The spatial BBN results indicated that incorporating governance capacity and conservation preferences into spatial modeling changes the location and extent of priority areas compared to biophysical-only approaches (Metzger et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The output maps provide decision-makers with an integrated view of ecological importance, social readiness, and governance potential.\u003c/p\u003e\u003cp\u003eThe spatial BBN flexible method used to analyze ES supply and demand relationships in areas with a high ES demand, such as the middle and lower parts of the watershed, demonstrated its applicability to include in informed decision in environmental planning of watersheds (H\u0026ouml;lting et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Landuyt et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). ES modeling frameworks, such as InVEST, have improved the quantification of biophysical processes but often overlook socioeconomic complexity and local stakeholder perceptions (Verburg et al., 2016). The BNN model addresses these gaps by linking ecological and social indicators within a probabilistic framework. The use of indices, such as the ES multifunctionality index, further refines our understanding of the RWSS ecological and social conditions, such as the water-use index and multidimensional poverty index, supporting its application in IWM. Such integrated analyses support the evaluation of use conflicts (Ricart \u0026amp; Rico-Amor\u0026oacute;s, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), identification of nature-based solution zones (Maragkaki et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and determination of biodiversity conservation priorities (Wu et al., \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Local watershed conditions enable the identification of areas where ecological priorities align or conflict with local governance capacities, a dimension that is not captured by most traditional models.\u003c/p\u003e\u003cp\u003eThe BBN implemented in the RWSS promoted participatory analysis, which is central to water governance in the Andean region. This participatory dimension of the model improves the output\u0026rsquo;s relevance and legitimacy, contributing to more equitable and effective watershed planning. The spatialization of participatory outputs in the BBN allows for the integration of data on conservation preferences, organizational presence, and socioeconomic conditions, facilitating the identification of priority areas for environmental management (Laurila-Pant et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Similarly, linking multidimensional poverty metrics to ecological attributes, such as forest cover, yields critical insights into the territorial dynamics that influence conservation decision-making in complex socio-ecological systems (Sun et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Thus, spatial prioritization shifted toward areas that, while ecologically valuable, also had strong community organization, established water governance practices, and a willingness to engage in conservation efforts. In contrast, biophysical models often emphasize remote areas with high ecological potential but low governance viability, thus limiting their implementation impact (Castro et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Pham et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eExtending the integration of local knowledge, the spatial BBN framework also clarifies conservation preferences within RWSS zones, which are closely linked to livelihoods, institutional presence, and participation in local water planning (Johnson et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). In this sense, water access increases community participation in conservation, suggesting that IWM strategies should consider factors such as land tenure security and co-designed management plans (Col\u0026oacute;n L\u0026oacute;pez and Restrepo, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Therefore, stakeholder perceptions of BBNs are key to promoting conservation policies that recognize shared benefits, ensure effective participation, and strengthen public water management strategies (Hallberg-Sramek et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This reinforces the need for co-created water assessments in watershed management using local knowledge (Crevier and Parrott, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Pereira et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThese findings contribute to the broader field of integrated socio-ecological earth system modeling, which emphasizes the coupling of human and natural systems to understand synergies and trade-offs in ES demand (Meraj et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Similar to integrated assessment models and coupled human\u0026ndash;natural systems modeling, our BBN framework explicitly links biophysical indicators (e.g., water yield and habitat quality) with social variables (e.g., governance participation, poverty, and equity). Unlike large-scale Earth system models that often operate at coarse spatial resolutions, this framework applies those principles at a decision-relevant local scale, making it more actionable for place-based policy and planning for effective IWM in the future.\u003c/p\u003e\u003cp\u003eFrom an IWM perspective, spatial BBN output enables stakeholders to jointly evaluate the trade-offs between ecological sustainability, social equity, and governance feasibility. Therefore, recognizing high-probability areas in ES use decisions is essential for fostering a collective interest in ES multifunctionality. However, the establishment of conservation areas of collective interest that promote environmental management remains challenging (Cosens et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Di Cintio et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The assessment of ES multifunctionality using integrative methods provides additional knowledge on territorial management, complementing hydrological modeling in water resource management (Barraclough et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eDespite its strengths, this framework relies on the availability and quality of both ecological and perceptual data. Model adaptation is essential in areas with different governance structures or limited perception data. Future research should integrate climate change projections, hydrological extremes, and dynamic land use change scenarios to improve temporal foresight. Comparative studies of watersheds with different governance regimes could further test the generalizability of the method and refine its integration into broader socio-ecological modeling efforts.\u003c/p\u003e"},{"header":"CONCLUSIONS","content":"\u003cp\u003eThis study developed a participatory spatial BBN for watershed planning that integrates data on ES supply and demand, socioeconomic indicators, and local perceptions into ES multifunctionality zoning for RWSS. The BBN model identified areas with high ES multifunctionality linked to equitable access, governance, and ecological integrity. Sensitivity analysis indicated that land-use restrictions, poverty, and participation influenced the ES multifunctionality outcomes in the watershed. This method advances IWM modeling by incorporating local forms of water governance and conservation perceptions into spatial maps.\u003c/p\u003e\u003cp\u003eThese model frameworks highlight the relevance of spatial planning methods in watershed environmental planning, which incorporate not only static assessments but also local community preferences. This approach is useful for river basins with limited data and low institutional presence. In addition, this method could contribute to strategic IWM planning, allowing planners to prioritize areas where conservation actions are likely to receive community support. In contexts such as the Coello River Basin, where local water organizations drive governance, this method can support decision-making, reduce conflicts, and improve local environmental instruments such as the POMCA.\u003c/p\u003e\u003cp\u003eBBN\u0026rsquo;s participatory spatial framework is a suitable approach for other watersheds characterized by complex socio-ecological systems, whose environmental management seeks to balance ecological priorities with socio-economic realities. The probabilistic structure of the BBN model allows its adaptation to diverse sources of information and uncertainty analysis, making it more accepted for regions with low data availability and different water governance conditions than other ES models. Future analysis will focus on integrating broader IWM frameworks and linking them to global sustainability agendas, such as Sustainable Development Goals 6 and 15.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eRojas C. wrote the main manuscript. Riascos- Ochoa, Longo and Clerici supervised the investigation. All authors reviewed the manuscript\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe acknowledge Jim\u0026eacute;nez, C., for his support in collecting data. Rubiano, J. of the Semilla de Agua Foundation, for his interest in this process and support in collecting the data. Funding: This study was funded by the Research Office of Universidad District, Francisco Jos\u0026eacute; de Caldas, Bogot\u0026aacute;, Colombia. The WWB Foundation for Financial Support in the Collection of Information. Ministry of Science, Technology, and Innovation (Colombia) for a Bicentennial Grant from the 2021 Fund.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe data are available from the corresponding author upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAdeoye-Olatunde, O. A., \u0026amp; Olenik, N. L. (2021). Research and scholarly methods: Semi-structured interviews. Journal of the american college of clinical pharmacy, 4(10), 1358-1367.\u003c/li\u003e\n\u003cli\u003eAlcald\u0026iacute;a de Ibagu\u0026eacute;. (2022). Directorio Juntas de acci\u0026oacute;n Comunal[WWW Document]. URL https://ibague.gov.co/portal/seccion/contenido/index.php?type=2\u0026amp;cnt=72#gsc.tab=0\u003c/li\u003e\n\u003cli\u003eBarraclough, A. D., Cusens, J., \u0026amp; M\u0026aring;ren, I. E. (2022). Mapping stakeholder networks for the co-production of multiple ecosystem services: A novel mixed-methods approach. Ecosystem Services, 56, 101461.\u003c/li\u003e\n\u003cli\u003eBayesFusion, L. L. C. (2017). Genie modeler. User Manual. Available online: https://support. bayesfusion. com/docs/(accessed on 21 October 2019), 16, 30-32.\u003c/li\u003e\n\u003cli\u003eBerr\u0026iacute;o, M. F., \u0026amp; Montenegro, N. (2022). Evaluaci\u0026oacute;n de servicios ecosist\u0026eacute;micos y su incidencia en el valor del suelo en la cuenca del r\u0026iacute;o coello, Tolima (Bachelor\u0026apos;s thesis). Universidad Distrital.\u003c/li\u003e\n\u003cli\u003eBerrouet, L., Villegas-Palacio, C., \u0026amp; Botero, V. (2020). Vulnerability of rural communities to change in an ecosystem service provision: Surface water supply. A case study in the northern Andes, Colombia. Land Use Policy, 97, 104737.\u003c/li\u003e\n\u003cli\u003eBlanco-Moreno, C., Amaya-Dom\u0026iacute;nguez, J., \u0026amp; Bernal-Pedraza, A. Y. (2024). Research Paper Experiences of community-based water supply organizations partnerships in rural areas of Colombia. Journal of Water, Sanitation and Hygiene for Development, 14(4), 332.\u003c/li\u003e\n\u003cli\u003eBruen, M., Hallouin, T., Christie, M., Matson, R., Siwicka, E., Kelly, F., Bullock, C., Feeley, H. B., Hannigan, E., \u0026amp; Kelly-Quinn, M. (2022). A Bayesian Modelling Framework for Integration of Ecosystem Services into Freshwater Resources Management. Environmental Management, 69(4), 781\u0026ndash;800. https://doi.org/10.1007/s00267-022-01595-x\u003c/li\u003e\n\u003cli\u003eCardenas, J. C., Rodriguez, L. A., \u0026amp; Johnson, N. (2011). Collective action for watershed management: field experiments in Colombia and Kenya. Environment and Development Economics, 16(3), 275-303.\u003c/li\u003e\n\u003cli\u003eCarriger, J. F., Barron, M. G., \u0026amp; Newman, M. C. (2016). Bayesian networks improve causal environmental assessments for evidence-based policy. Environmental science \u0026amp; technology, 50(24), 13195-13205.\u003c/li\u003e\n\u003cli\u003eCastro, A. J., Vaughn, C. C., Julian, J. P., \u0026amp; Garc\u0026iacute;a-Llorente, M. (2016). Social demand for ecosystem services and implications for watershed management. JAWRA Journal of the American Water Resources Association, 52(1), 209-221.\u003c/li\u003e\n\u003cli\u003eCBD (2018). Expert workshop on marine protected areas and other effective area-based conservation measures for achieving aichi biodiversity target 11 in marine and coastal areas (Montreal, Canada: Convention on Biological Diversity (CBD)). Available at: https://www.cbd.int/doc/c/d1d1/ff04/1adecc670d5f789c47c747b7/mcb-em-2018-01-03-en.docx. 6-9 February 2018CBD/MCB/EM/2018/1/3. \u003c/li\u003e\n\u003cli\u003eCole, B., Bradley, A. V., Willcock, S., Gardner, E., Allinson, E., Hagen-Zanker, A., \u0026amp; Whelan, M. (2023). Using a multi-lens framework for landscape decisions. People and Nature. \u003c/li\u003e\n\u003cli\u003eCol\u0026oacute;n L\u0026oacute;pez, J., \u0026amp; Restrepo, C. (2019). Water Quality and Socio-Economic Indicators are Linked in a Tropical Watershed: Emerging Implications for the Sustainable Management of Waterscapes. Wetlands, 39(6), 1303\u0026ndash;1316. https://doi.org/10.1007/s13157-019-01232-0\u003c/li\u003e\n\u003cli\u003eCortolima. (2019). Ajuste parcial al Plan de Ordenaci\u0026oacute;n y Manejo de la subzona hidrogr\u0026aacute;fica del r\u0026iacute;o Coello. Corporaci\u0026oacute;n Aut\u0026oacute;noma Regional del Tolima - CORTOLIMA [WWW Document]. URL https://cortolima.gov.co/planes-y-programas/gestion-integral-del-recurso-hidrico/3377-plan-de-ordenamiento-y-manejo-de-cuenca-pomca-rio-coello\u003c/li\u003e\n\u003cli\u003eCortolima. (2020). Corporaci\u0026oacute;n Aut\u0026oacute;noma Regional del Tolima - CORTOLIMA [WWW Document]. URL https://datosabiertos.cortolima.gov.co/datasets/9659bd2b87e44869b05402f61e914789_0/explore\u003c/li\u003e\n\u003cli\u003eCosens, B., Ruhl, J. B., Soininen, N., Gunderson, L., Belinskij, A., Blenckner, T., ... \u0026amp; Simil\u0026auml;, J. (2021). Governing complexity: Integrating science, governance, and law to manage accelerating change in the globalized commons. Proceedings of the National Academy of Sciences, 118(36), e2102798118.\u003c/li\u003e\n\u003cli\u003eCrevier, L. P., \u0026amp; Parrott, L. (2019). Synergy between adaptive management and participatory modelling: The two processes as interconnected spirals. Ecological Informatics, 53, 100982.\u003c/li\u003e\n\u003cli\u003eDANE. (2018). DANE Geoportal [WWW Document]. URL https://geoportal.dane.gov.co/geovisores/sociedad/cnpv-2018/\u003c/li\u003e\n\u003cli\u003eDANE. (2022). DANE Geoportal [WWW Document]. URL https://www.dane.gov.co/index.php/estadisticas-por-tema/pobreza-y-condiciones-de-vida/pobreza-multidimensional\u003c/li\u003e\n\u003cli\u003eDaniel, D., Al Djono, T. P., \u0026amp; Iswarani, W. P. (2023). Factors related to the functionality of community-based rural water supply and sanitation program in Indonesia. \u003cem\u003eGeography and Sustainability\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e(1), 29-38.\u003c/li\u003e\n\u003cli\u003eDelac\u0026aacute;mara, G., O\u0026rsquo;Higgins, T. G., Lago, M., \u0026amp; Langhans, S. (2020). Ecosystem-based management: moving from concept to practice. In Ecosystem-based management, ecosystem services and aquatic biodiversity: Theory, tools and applications (pp. 39-60). Cham: Springer International Publishing. \u003c/li\u003e\n\u003cli\u003eDelucchi, A. A. S. (2024). Unpacking \u0026lsquo;community water management\u0026rsquo; in rural Chile: An institutional ethnography. Journal of Rural Studies, 110, 103345.\u003c/li\u003e\n\u003cli\u003eDi Cintio, A., Fernandes-Salvador, J. A., Puntila-Dodd, R., Granado, I., Niccolini, F., \u0026amp; Bulleri, F. (2024). Socio-economic factors boosting the effectiveness of marine protected areas: A Bayesian network analysis. \u003cem\u003eEcological Informatics\u003c/em\u003e, \u003cem\u003e84\u003c/em\u003e, 102879.\u003c/li\u003e\n\u003cli\u003eForio, M. A. E., Villa-Cox, G., Van Echelpoel, W., Ryckebusch, H., Lock, K., Spanoghe, P., ... \u0026amp; Goethals, P. L. (2020). Bayesian Belief Network models as trade-off tools of ecosystem services in the Guayas River Basin in Ecuador. Ecosystem Services, 44, 101124.\u003c/li\u003e\n\u003cli\u003eFrancesconi, W., Srinivasan, R., P\u0026eacute;rez-Mi\u0026ntilde;ana, E., Willcock, S. P., \u0026amp; Quintero, M. (2016). Using the Soil and Water Assessment Tool (SWAT) to model ecosystem services: A systematic review. Journal of Hydrology, 535, 625-636. https://doi.org/10.1016/j.jhydrol.2016.01.034\u003c/li\u003e\n\u003cli\u003eGonz\u0026aacute;lez-Redin, J., Luque, S., Poggio, L., Smith, R., \u0026amp; Gimona, A. (2016). Spatial Bayesian belief networks as a planning decision tool for mapping ecosystem services trade-offs on forested landscapes. Environmental Research, 144, 15\u0026ndash;26.\u003c/li\u003e\n\u003cli\u003eGuo, X., Wang, L., Fu, Q., \u0026amp; Ma, F. (2024). Ecological Function Zoning Framework for Small Watershed Ecosystem Services Based on Multivariate Analysis from a Scale Perspective. Land, 13(7), 1030.\u003c/li\u003e\n\u003cli\u003eHallberg-Sramek, I., Nordstr\u0026ouml;m, E.-M., Priebe, J., Reimerson, E., M\u0026aring;rald, E., \u0026amp; Nordin, A. (2023). Combining scientific and local knowledge improves evaluating future scenarios of forest ecosystem services. Ecosystem Services, 60, 101512. https://doi.org/10.1016/j.ecoser.2023.101512\u003c/li\u003e\n\u003cli\u003eHijmans, R. J., Bivand, R., Forner, K., Ooms, J., Pebesma, E., \u0026amp; Sumner, M. D. (2022). Package \u0026lsquo;terra.\u0026rsquo; Maintainer: Vienna, Austria.\u003c/li\u003e\n\u003cli\u003eH\u0026ouml;fer, S., Ziemba, A., \u0026amp; El Serafy, G. (2020). A Bayesian approach to ecosystem service trade-off analysis utilizing expert knowledge. Environment Systems and Decisions, 40(1), 67-83.\u003c/li\u003e\n\u003cli\u003eH\u0026ouml;lting, L., Jacobs, S., Felipe-Lucia, M. R., Maes, J., Norstr\u0026ouml;m, A. V., Plieninger, T., \u0026amp; Cord, A. F. (2019). Measuring ecosystem multifunctionality across scales. Environmental Research Letters, 14(12), 124083. https://doi.org/10.1088/1748-9326/ab5ccb\u003c/li\u003e\n\u003cli\u003eJohnson, N., Garc\u0026iacute;a, J., Rubiano, J. E., Quintero, M., Estrada, R., Mwangi, E., ... \u0026amp; Granados, S. (2008). Watershed management and poverty alleviation in the Colombian Andes. Spatial \u0026amp; Economic Analysis for Policy and Decision Support in Agriculture and Environment, 67.\u003c/li\u003e\n\u003cli\u003eLanduyt, D., Broekx, S., \u0026amp; Goethals, P. L. (2016). Bayesian belief networks to analyse trade-offs among ecosystem services at the regional scale. Ecological Indicators, 71, 327\u0026ndash;335.\u003c/li\u003e\n\u003cli\u003eLaurila-Pant, M., M\u0026auml;ntyniemi, S., Venesj\u0026auml;rvi, R., \u0026amp; Lehikoinen, A. (2019). Incorporating stakeholders\u0026apos; values into environmental decision support: A Bayesian Belief Network approach. \u003cem\u003eScience of the Total Environment\u003c/em\u003e, \u003cem\u003e697\u003c/em\u003e, 134026.\u003c/li\u003e\n\u003cli\u003eMaragkaki, A., Koukianaki, E. A., Lilli, M. A., Efstathiou, D., \u0026amp; Nikolaidis, N. P. (2024). Optimizing the water-ecosystem-food nexus using nature-based solutions at the basin scale. Frontiers in Water, 6, 1386925.\u003c/li\u003e\n\u003cli\u003eMarcot, B. G., Steventon, J. D., Sutherland, G. D., \u0026amp; McCann, R. K. (2006). Guidelines for developing and updating Bayesian belief networks applied to ecological modeling and conservation. Canadian Journal of Forest Research. https://doi.org/10.1139/x06-135\u003c/li\u003e\n\u003cli\u003eMart\u0026iacute;nez-L\u0026oacute;pez, J., Bagstad, K. J., Balbi, S., Magrach, A., Voigt, B., Athanasiadis, I., ... \u0026amp; Villa, F. (2019). Towards globally customizable ecosystem service models. Science of the Total Environment, 650, 2325-2336. \u003c/li\u003e\n\u003cli\u003eMasante, D. (2016). Package \u0026lsquo;bnspatial.\u0026rsquo;. Spatial Implementation of Bayesian Networks and Mapping. http://github.com/dariomasante/bnspatial\u003c/li\u003e\n\u003cli\u003eMeraj, G., Singh, S. K., Kanga, S., \u0026amp; Islam, M. N. (2022). Modeling on comparison of ecosystem services concepts, tools, methods and their ecological-economic implications: A review. Modeling Earth Systems and Environment, 8(1), 15-34.\u003c/li\u003e\n\u003cli\u003eMetzger, J. P., Villarreal-Rosas, J., Su\u0026aacute;rez-Castro, A. F., L\u0026oacute;pez-Cubillos, S., Gonz\u0026aacute;lez-Chaves, A., Runting, R. K., Hohlenwerger, C., \u0026amp; Rhodes, J. R. (2021). Considering landscape-level processes in ecosystem service assessments. Science of The Total Environment, 796, 149028.\u003c/li\u003e\n\u003cli\u003eOmarova, A., Tussupova, K., Hjorth, P., Kalishev, M., \u0026amp; Dosmagambetova, R. (2019). Water Supply Challenges in Rural Areas: A Case Study from Central Kazakhstan. International Journal of Environmental Research and Public Health, 16(5), Article 5. https://doi.org/10.3390/ijerph16050688\u003c/li\u003e\n\u003cli\u003ePereira, P., Inacio, M., Barcelo, D., \u0026amp; Zhao, W. (2025). Ecosystem services mapping and modelling. Where is the validation?. Geography and Sustainability, 6(3), 100286.\u003c/li\u003e\n\u003cli\u003ePham, H. V., Sperotto, A., Furlan, E., Torresan, S., Marcomini, A., \u0026amp; Critto, A. (2021). Integrating Bayesian Networks into ecosystem services assessment to support water management at the river basin scale. Ecosystem Services. https://doi.org/10.1016/j.ecoser.2021.101300\u003c/li\u003e\n\u003cli\u003ePhan, T. D., Smart, J. C., Stewart-Koster, B., Sahin, O., Hadwen, W. L., Dinh, L. T., ... \u0026amp; Capon, S. J. (2019). Applications of Bayesian networks as decision support tools for water resource management under climate change and socio-economic stressors: a critical appraisal. Water, 11(12), 2642.\u003c/li\u003e\n\u003cli\u003ePilogallo, A., \u0026amp; Scorza, F. (2022). Ecosystem services multifunctionality: An analytical framework to support sustainable spatial planning in Italy. Sustainability, 14(6), 3346. https://doi.org/10.3390/su14063346\u003c/li\u003e\n\u003cli\u003ePlatarrueda, I. (2022). Community aqueducts in Colombia and their struggle for legal recognition: A political ecology approach. Perfiles Econ\u0026oacute;micos. https://doi.org/10.22370/pe.2022.13.3445\u003c/li\u003e\n\u003cli\u003ePosner, S., Verutes, G., Koh, I., Denu, D., \u0026amp; Ricketts, T. (2016). Global use of ecosystem service models. Ecosystem Services, 17, 131-141. https://doi.org/10.1016/j.ecoser.2015.12.003\u003c/li\u003e\n\u003cli\u003eQGIS.org (2023). QGIS Geographic Information System. QGIS Association. http://www.qgis.org\u003c/li\u003e\n\u003cli\u003eRicart, S., \u0026amp; Rico-Amor\u0026oacute;s, A. M. (2021). Water for food, water for birds: How to manage conflicting rural-natural interfaces? Deepening on the socio-ecological system of El Hondo Natural Park (Alicante, Spain). Journal of Rural Studies, 86, 24-35.\u003c/li\u003e\n\u003cli\u003eRodr\u0026iacute;guez-Loinaz, G., Alday, J. G., \u0026amp; Onaindia, M. (2015). Multiple ecosystem services landscape index: A tool for multifunctional landscapes conservation. Journal of Environmental Management, 147, 152\u0026ndash;163. https://doi.org/10.1016/j.jenvman.2014.09.001 N.N.\u003c/li\u003e\n\u003cli\u003eRuhl, J., Salzman, J., Arnold, C. A., Craig, R., Hirokawa, K., Olander, L., Palmer, M., \u0026amp; Ricketts, T. H. (2021). Connecting ecosystem services science and policy in the field. Frontiers in Ecology and the Environment, 19(9), 519\u0026ndash;525. https://doi.org/10.1002/fee.2390\u003c/li\u003e\n\u003cli\u003eRumbold, D. G. (2023). Use of a Bayesian network as a decision support tool for watershed management: A case study in a highly managed river-dominated estuary. \u003cem\u003eEnvironmental Monitoring and Assessment\u003c/em\u003e, \u003cem\u003e195\u003c/em\u003e(6), 741.\u003c/li\u003e\n\u003cli\u003eSalliou, N., Barnaud, C., Vialatte, A., \u0026amp; Monteil, C. (2017). A participatory Bayesian Belief Network approach to explore ambiguity among stakeholders about socioecological systems. Environmental Modelling \u0026amp; Software, 96, 199\u0026ndash;209. https://doi.org/10.1016/j.envsoft.2017.06.050\u003c/li\u003e\n\u003cli\u003eScrieciu, A., Pagano, A., Coletta, V. R., Fratino, U., \u0026amp; Giordano, R. (2021). Bayesian belief networks for integrating scientific and stakeholders\u0026rsquo; knowledge to support nature-based solution implementation. Frontiers in Earth Science, 9, 674618.\u003c/li\u003e\n\u003cli\u003eScutari, M., \u0026amp; Denis, J. B. (2021). Bayesian networks: with examples in R. Chapman and Hall/CRC.\u003c/li\u003e\n\u003cli\u003eSeifert-D\u0026auml;hnn, I., Barkved, L. J., \u0026amp; Interwies, E. (2015). Implementation of the ecosystem service concept in water management - Challenges and ways forward. Sustainability of Water Quality and Ecology, 5, 3-8. https://doi.org/10.1016/j.swaqe.2015.01.007\u003c/li\u003e\n\u003cli\u003eShao, Y., Xiao, Y., Kou, X., \u0026amp; Sang, W. (2023). Sustainable land use scenarios generated by optimizing ecosystem distribution based on temporal and spatial patterns of ecosystem services in the southern China hilly region. Ecological Informatics, 78, 102275.\u003c/li\u003e\n\u003cli\u003eSun, X., Ye, D., Shan, R., Peng, Q., Zhao, Z., \u0026amp; Sun, J. (2022). Effect of physical geographic and socioeconomic processes on interactions among ecosystem services based on machine learning. Journal of Cleaner Production, 359, 131976. https://doi.org/10.1016/j.jclepro.2022.131976\u003c/li\u003e\n\u003cli\u003eUNEP-WCMC (2023). Protected Area Profile for Cuenca hidrogr\u0026aacute;fica del r\u0026iacute;o Coello from the World Database on Protected Areas, September 2023. Available at: www.protectedplanet.net\u003c/li\u003e\n\u003cli\u003eVollmer, D., Burkhard, K., Adem Esmail, B., Guerrero, P., \u0026amp; Nagabhatla, N. (2022). Incorporating Ecosystem Services into Water Resources Management\u0026mdash;Tools, Policies, Promising Pathways. Environmental Management, 69(4), 627\u0026ndash;635. https://doi.org/10.1007/s00267-022-01640-9\u003c/li\u003e\n\u003cli\u003eWu, C., Qiu, D., Gao, P., Mu, X., \u0026amp; Zhao, G. (2022). Application of the InVEST model for assessing water yield and its response to precipitation and land use in the Weihe River Basin, China. Journal of Arid Land, 14(4), 426\u0026ndash;440. https://doi.org/10.1007/s40333-022-0013-0\u003c/li\u003e\n\u003cli\u003eZeng, L., \u0026amp; Li, J. (2019). A Bayesian belief network approach for mapping water conservation ecosystem service optimization region. Journal of Geographical Sciences, 29, 1021-1038.\u003c/li\u003e\n\u003cli\u003eZhang, W., ElDidi, H., Masuda, Y. J., Meinzen-Dick, R. S., Swallow, K. A., Ringler, C., ... \u0026amp; Aldous, A. (2023). Community-based conservation of freshwater resources: learning from a critical review of literature and case studies. Society \u0026amp; Natural Resources, 36(6), 733-754.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"modeling-earth-systems-and-environment","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mese","sideBox":"Learn more about [Modeling Earth Systems and Environment](http://link.springer.com/journal/40808)","snPcode":"40808","submissionUrl":"https://submission.springernature.com/new-submission/40808/3","title":"Modeling Earth Systems and Environment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Ecosystem service multifunctionality, watershed planning, local perception mapping, watershed environmental zoning, Mesli Index","lastPublishedDoi":"10.21203/rs.3.rs-7382296/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7382296/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAddressing global water security challenges requires effective watershed management that seamlessly integrates ecosystem service (ES) assessments with governance and socio-economic aspects of water resource utilization. However, most existing modeling approaches focus narrowly on biophysical processes and lack mechanisms for incorporating local knowledge under uncertainty. Unlike previous ES-watershed models that focused on biophysical indicators, our approach integrates ecological modeling with local perceptions and governance probabilistically, representing uncertainty and socio-ecological interactions that are often omitted from ES tools. The rural water supply system (RWSS) supports ES for water access, biodiversity, and social cohesion but is absent from watershed planning. To address this gap, we developed a spatial Bayesian Belief Network (BBN) model using data from RWSS organizations to support participatory zoning. The spatial BBN framework was applied to the Coello River Basin, Colombia, integrating high-resolution land use, ecosystem service multifunctionality indices, and multidimensional poverty indicators with perception data from the community leaders. The model generated spatial probability maps of multifunctional ES zones, incorporating governance capacity and conservation willingness. Key variables affecting ES multifunctionality include participation, restrictions, and poverty. Areas of high multifunctionality are aligned with zones of strong governance and high ecological integrity. This BBN enhances planning by integrating community insights into ecological modeling. By quantifying both the probability of ES supply and the uncertainty of predictions, the framework provides actionable insights for integrated watershed management (IWM) and supports the localization of global Sustainability Development Goals (SDG) (SDG 6-clean water and sanitation and SDG 15-life on land) in diverse socio-ecological contexts.\u003c/p\u003e","manuscriptTitle":"Integrating Local Perceptions in Bayesian Belief Network for Watershed Planning: A Case in the Colombian Andes","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-02 06:47:32","doi":"10.21203/rs.3.rs-7382296/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-09-30T18:05:02+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-23T00:51:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"150027543334590541622720752078593964303","date":"2025-08-26T23:26:30+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-08-23T14:19:02+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-18T11:30:15+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-08-18T11:29:08+00:00","index":"","fulltext":""},{"type":"submitted","content":"Modeling Earth Systems and Environment","date":"2025-08-15T14:40:28+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"modeling-earth-systems-and-environment","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mese","sideBox":"Learn more about [Modeling Earth Systems and Environment](http://link.springer.com/journal/40808)","snPcode":"40808","submissionUrl":"https://submission.springernature.com/new-submission/40808/3","title":"Modeling Earth Systems and Environment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"f148f02d-1c02-4638-be95-754ada4d4452","owner":[],"postedDate":"September 2nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-04-07T16:09:22+00:00","versionOfRecord":{"articleIdentity":"rs-7382296","link":"https://doi.org/10.1007/s40808-026-02735-6","journal":{"identity":"modeling-earth-systems-and-environment","isVorOnly":false,"title":"Modeling Earth Systems and Environment"},"publishedOn":"2026-04-04 15:58:51","publishedOnDateReadable":"April 4th, 2026"},"versionCreatedAt":"2025-09-02 06:47:32","video":"","vorDoi":"10.1007/s40808-026-02735-6","vorDoiUrl":"https://doi.org/10.1007/s40808-026-02735-6","workflowStages":[]},"version":"v1","identity":"rs-7382296","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7382296","identity":"rs-7382296","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}