Improving Wastewater Management for Sustainable Environmental Conditions: A Flexible Semantic Network-Based Approach

Highlights

What are the main findings?

• The application of semantic networks based on a holistic approach to model human factors and underlying causes, improving wastewater management.

What are the implications of the main findings?

• It allows addressing the problem at its root, transforming complex data into a visual structure that facilitates decision-making to mitigate negative environmental impact.
• Its flexibility allows for the identification of specific causes and critical variables that affect waste management performance, thereby adapting to the particularities of each context.
• It enables modelling of processes that hinder effective management, facilitating strategic improvements that directly reduce environmental damage in these sensitive ecosystems.
• This approach aims to reduce environmental damage to coastal ecosystems and safeguard biodiversity and water quality from the adverse impacts of inadequate management.

Abstract

Anthropogenic activities interact through multifactorial processes that generate harmful factors, hindering wastewater management (WWM) and causing environmental degradation, particularly in rapidly urbanizing coastal regions. Understanding how these processes operate is essential for identifying effective interventions under data-limited conditions. This study introduces a process-based and context-driven approach whose main contribution lies in the construction of semantic pathways that represent how indirect anthropogenic drivers give rise to direct harmful factors affecting WWM and the environmental state. Semantic networks are used as a formal representation tool to model these pathways, where nodes represent factors and directed arcs represent causal relationships. Harmful semantic pathways are evaluated within a multidisciplinary decision-making framework supported by multi-criteria decision-making (MCDM) methods that account for environmental, social, economic, and sustainability criteria. The approach is applied to a coastal tourist municipality on the Mexican Pacific coast, where rapid urban expansion and insufficient basic services have severely constrained wastewater management and contributed to environmental damage. Results show that constructing and analyzing semantic pathways enables decision-makers to identify critical harmful factors and prioritize viable pro-environmental actions. The resulting priorities (range 0–1) highlight the restoration of wastewater treatment plants and improved urban planning as the most effective interventions, followed by mangrove reforestation and changes in agricultural practices. The proposed approach supports transparent, context-sensitive decision-making and is transferable to coastal tourist municipalities facing similar wastewater management challenges.

1. Introduction

In general, waste management systems aim to reduce the negative impacts of waste on the environment by implementing strategies such as reduction, reuse, prevention, and responsible disposal [1,2,3,4,5,6,7]. These systems play a key role in minimizing risks to public health and environmental quality, while also supporting the transition toward resource-efficient and circular economies [8]. Improperly managed waste can lead to air, water, and soil pollution, as well as the spread of disease and other serious health risks.

Decision-making processes play a critical role in supporting waste management systems through the selection of effective pro-environmental actions (PEAs) that protect both environmental integrity and public health. In particular, wastewater management systems (WWMs) are designed to collect, treat, and safely dispose of wastewater generated by households, businesses, and industries. Effective wastewater management generates social, environmental, and economic benefits and represents a key enabling action toward achieving the 2030 Agenda for Sustainable Development [9].

1.1. Wastewater Management Challenges

Low water quality resulting from the poor performance of wastewater treatment plants (WWTPs), many of which were built decades ago and now require repair or replacement, remains a persistent challenge. Inadequate infrastructure can lead to leaks, overflows, and treatment failures that negatively affect ecosystems and public health [10,11,12,13]. Wastewater discharges containing high nutrient loads contribute to eutrophication, harmful algal blooms, and other water quality problems [14,15,16].

In many developing countries, wastewater infrastructure has not kept pace with rapid population growth and urban expansion, resulting in insufficient treatment capacity and service coverage [17]. Agricultural runoff further exacerbates these challenges by introducing pesticide residues and nutrients into rivers, lakes, lagoons, and groundwater systems [18,19].

It has been widely documented that population growth is closely linked to increased consumption patterns worldwide [20,21,22], leading to higher demand for water, food, energy, and raw materials [23,24,25]. This growing demand intensifies production processes, increases wastewater generation, and amplifies pressures on water resources, often resulting in water scarcity and environmental degradation [26,27,28].

Governance represents another critical challenge for wastewater management under sustainability objectives [29,30,31]. Within this context, water governance encompasses the social, political, economic, and administrative institutions and processes that shape water resource management [32]. Core principles such as legitimacy, transparency, and accountability are essential for effective governance. Consequently, decision-making processes related to wastewater management must address sociopolitical dimensions, including institutional coherence and anti-corruption measures, in order to design and implement legitimate and effective strategies [33,34,35].

Overall, wastewater management effectiveness is strongly influenced by a set of interacting anthropogenic activities (indirect drivers), including population growth, socioeconomic and sociopolitical factors, technological capacity, and environmental culture [36,37,38,39]. These indirect drivers often give rise to direct drivers—such as nutrient loading, untreated discharges, and other harmful factors—that directly damage waterbodies and associated ecosystems [40,41].

1.2. Wastewater Management as a Multifactorial and Multidisciplinary Problem

The multiple interactions among indirect and direct drivers significantly complicate the understanding and analysis of wastewater management systems. As a result, decision-making processes are frequently based on incomplete or fragmented representations of reality, reducing their reliability in selecting appropriate pro-environmental alternatives. This difficulty is further exacerbated in developing-country contexts, where reliable and consistent data are often scarce.

To effectively address wastewater management challenges and reduce pollution, it is therefore necessary to consider the interactions among environmental, social, economic, technological, and institutional factors. Decision-making processes supporting wastewater management must be framed within a multifactorial and multidisciplinary perspective capable of integrating these diverse dimensions.

Such an approach enables decision-makers to consider multiple criteria simultaneously—environmental, social, and economic—and to evaluate their relative importance within a given context [42]. In multidisciplinary settings, the involvement of experts and stakeholders from diverse fields becomes essential. Their practical experience with real-world problems related to water pollution and environmental degradation provides valuable qualitative knowledge that can compensate for data limitations and improve decision-making outcomes.

Assessing decision-support approaches in developing-country contexts based on assumptions from high-income settings can result in inaccurate conclusions about their novelty or effectiveness. In these environments, methodological value is found not in optimization under ideal conditions but in the ability to structure complex processes, identify underlying drivers, and facilitate prioritization amid uncertainty and institutional constraints.

1.3. Semantic Networks as Knowledge Representation Tools

Semantic networks provide a formal representation of knowledge in which nodes represent concepts and directed arcs represent relationships between them. They constitute a well-established area within artificial intelligence and knowledge representation research. Several models of semantic networks have been developed and applied across different domains. In this work, we draw on the following models:

• Definitional networks, which represent hierarchical relationships from general to specific knowledge. In this study, multiple levels of abstraction are used to represent how indirect and direct drivers affect environmental outcomes through cause–effect relationships. While highly abstract representations can capture general system behavior, complex environmental systems require more specific levels of knowledge to improve understanding and analysis.
• Learning networks, in which network structures evolve through the incorporation of new information and experiences, enabling continuous knowledge refinement.
• Hybrid networks, which combine multiple network models to represent complex and dynamic systems more effectively.

It is important to emphasize that semantic networks themselves are not a novelty in this study; rather, they serve as a formal representation tool supporting the methodological approach introduced here.

1.4. Proposal and Contribution of This Study

The main objective of this work is to improve decision-making processes supporting sustainable wastewater management in coastal tourist municipalities through the selection of appropriate pro-environmental actions or alternatives that mitigate environmental degradation.

We argue that multifactorial interactions between anthropogenic activities (indirect drivers) and direct drivers generate complex processes that are often poorly understood. This limited understanding leads to unreliable decision-making in wastewater management and hampers the effective selection of pro-environmental alternatives.

To address this challenge, we introduce a process-based and context-driven approach that models how indirect drivers give rise to direct harmful factors affecting the environmental state, including waterbodies, ecosystem services, and vulnerable urban areas. The central contribution of this study lies in the construction of semantic pathways—ordered sequences of concepts connected by causal relationships—that explicitly represent these processes.

Qualitative assessments supported by semantic pathways enable the identification of harmful factors and guide the selection of pro-environmental alternatives. While quantitative assessments depend heavily on data availability, semantic pathways provide a structured and interpretable representation of system dynamics, which is particularly valuable in developing-country contexts where data are limited.

The approach is implemented within a multidisciplinary decision-making framework, where experts from diverse fields evaluate semantic pathways using environmental, social, and economic criteria within a sustainability perspective. Multi-criteria decision-making methods support the prioritization of harmful pathways and their associated pro-environmental alternatives.

This study involved ten experts specializing in environmental management systems, socioecological systems, biodiversity and conservation, urban engineering, wastewater management, agriculture, environmental modeling, regional development, and social anthropology. Although the approach is applied to a specific case study, it is transferable to coastal tourist municipalities facing similar challenges related to anthropogenic pressures, wastewater management deficiencies, and environmental degradation.

2. Materials and Methods

Study context. This study focuses on improving WWM in coastal tourist municipalities characterized by sustained population growth, internal migration, and rapid and uncontrolled urban expansion—conditions commonly observed in developing countries. Tourism plays a central role in local economic development but also intensifies pressure on urban infrastructure and basic services, particularly wastewater treatment systems.

The municipality of Acapulco (Mexico) was selected as a representative case study due to (i) its long-standing dependence on tourism, (ii) rapid demographic growth, (iii) uncontrolled urban sprawl, and (iv) recurrent failures in wastewater treatment plants (WWTPs). These characteristics make it a suitable context for testing approaches aimed at understanding and managing complex, multifactorial environmental problems.

Secondary data related to tourism flows, population growth, and urban expansion were obtained from official governmental sources and previous studies. These data were used to contextualize the construction of the semantic network rather than to perform direct statistical modeling.

2.1. Tourism, Population Growth and Urban Expansion as Indirect Drivers

Tourism: the most important economic activity in Acapulco. A study covering data related to tourist visitors, hotel occupancy, and the economic impact in Acapulco shows significant changes in these variables (see Table 1) during the period 2018–2023, which include the events of COVID-19 and Hurricane OTIS [43].

Table 1. Effects of COVID-19 and Hurricane OTIS on tourism related to the number of tourists, hotel occupancy, and the impact on the economy due to these events.

Year	Number of Tourists	Hotel Occupancy (%)	Economic (Millions of Mexican Pesos [mp]) Current Exchange: 1 usd = 18.23 mp)
2018	9,891,776	49.3	34,275.0
2019	10,228,539	51.7	35,441.9
2020	6,192,520	31.3	21,456.5
2021	6,405,048	34.4	28,822.7
2022	8,172,730	47.6	62,656.2
2023	6,198,391	53.5	47,520.0

A state of emergency was declared due to COVID-19 on 30 January 2020. The COVID-19 pandemic ended on 5 May 2023. Meanwhile, hurricane OTIS struck the area (from 22 to 25 October 2023). The values of the variables shown in Table 1 decreased during the COVID-19 and OTIS events. Subsequently, Hurricane John struck the area in September 2024. Despite these three events occurring sequentially, Acapulco achieved a 65% tourism recovery by the end of 2024 (https://forbes.com.mx/acapulco-cierra-2024-con-recuperacion-turistica-de-65-tras-huracanes/, accessed on 5 December 2025). Finally, hotel occupancy during a recent long weekend (14–17 November 2025) was 78.5%, with 61,112 visitors, based on data provided by the government of the state of Guerrero, where the municipality of Acapulco is located (https://www.guerrero.gob.mx/2025/11/guerrero-y-acapulco-superan-expectativas-con-exitoso-ultimo-puente-turistico-2025/, accessed on 5 December 2025).

Population Growth and Urban Expansion: Population growth has led to the expansion of formal and informal human settlements, often located along riverbeds, coastal lagoons, and environmentally sensitive areas. These settlements frequently lack adequate drainage and sanitation services, resulting in untreated wastewater discharges and solid waste accumulation.

Urban expansion, tourism activities, and human settlements were considered intermediate process components linking indirect drivers to direct environmental pressures. This intermediate representation is essential to avoid oversimplified cause–effect interpretations.

Key data: Years vs. Population Growth & Years vs. Expansion of the Urban Areas. Figure 1 shows a table with three key data points: the period from 1930 to 2020 (in the first column), population growth (in the second column), and the expansion of urban areas (in the third column). The “Years versus Population-Growth” graph shows a clear upward trend in population between 1970 and 2010. However, population decreased by 158,105 inhabitants between 2010 and 2020 (see the graphics Years–Population Growth). Among the possible causes of this decrease are insecurity caused by criminal groups, the search for other destinations, and the return to their places of origin. In contrast, the expansion of urban areas continued to increase significantly (see the graphics Years–Expansion of the Urban Areas).

Figure 1. Years vs. Population Growth and Years vs. Expansion of the Urban Areas, period 1930–2020 [44].

The subsequent population growth during the 1970s and 1980s was due to the national and international reputation of this tourist destination. At that time, Acapulco was considered Mexico’s premier coastal resort, where tourists from Mexico City spent weekends, summers, Easter and New Year’s holidays. As a result, this flourishing tourist destination became an attractive source of employment for a significant number of people seeking jobs unavailable in their hometowns. This led to what is known as internal migration.

It is clear that this constant internal migration has led to a search for housing, whether in formal or informal settlements. As a result, basic needs such as housing, sanitation, drainage, garbage collection, transportation, and clean water have increased in tandem with population growth and the expansion of urban areas. Despite these complex problems, the expected urban planning has been lacking, and this is one of the most important concerns that must be addressed.

Figure 2 shows a map illustrating the expansion of different zones (in various colors) within the municipality of Acapulco during the period from 1930 to 2014. The area with the most recent expansion is shown in gray and comprises the newer low-income neighborhoods (most of them are part of internal migration) and the Zona Diamante (Diamond Zone). The latter is a luxury residential area for residents of Mexico City, located at three-hours away by car. It can be observed that the gray area has extended into an significant mangrove area surrounding the Tres Palos coastal lagoon. In addition to the damage caused to these areas, a significant impact is on the extension of urban areas on Veladero National Park, because these areas are located in the foothills, where the main hazards are landslides and flooding during the rainy season and hurricanes.

Figure 2. Expansion of the urban areas from 1930 to 2014 [45].

2.2. Semantic Network Construction

2.2.1. Methodological Protocol for Semantic Network Construction

Conceptual basis. Semantic networks were used to represent the multifactorial processes through which anthropogenic activities generate harmful environmental factors. In this study, nodes represent concepts (factors), and directed arcs represent causal or influence relationships between them.

The contribution of this work does not lie in the semantic network formalism itself, which is well established but in its process-based and context-driven application to wastewater management under data-scarce conditions typical of developing countries.

The following step-by-step protocol aims to formalize the construction of the semantic network. To ensure transparency and reproducibility, the semantic network was constructed following the protocol. The graphics representing the semantic networks are shown in the appendices.

Step 1. Identification and definition of indirect drivers (Layer 1) and their relationships with the components of the environmental state layer

Corresponding semantic network. Layer-1, indirect drivers, represents anthropogenic activities relevant to wastewater management (population growth, socioeconomic, sociopolitical, technological, and cultural factors). It is linked to the layer called “environmental state”, which includes damage to waterbodies, ecosystem services, and vulnerable urban areas. From a process perspective, the links between these two layers are not convincing enough to demonstrate that the indirect drivers of Layer-1 cause damage to the defined environmental state; therefore, more specific knowledge is required.

These layers were identified and defined through consultation with experts. The network linking both layers is shown in Appendix A (Figure A1).

Step 2. Definition of intermediate process layers

The corresponding semantic network. Intermediate layers representing indirect and direct drivers were added to capture part of the process through which indirect drivers generate environmental pressures. The added layers: Layer-2 of indirect drivers representing human settlements, tourism activities, and urban expansion; Layer-2 and Layer-1 of direct drivers, which represent tangible harmful factors. This more detailed network is now composed of five layers (see Appendix B, Figure A2).

Step 3. Progressive refinement through additional layers

More specific concepts were incorporated to improve the resolution of the process. Layer-3, comprising indirect factors, was added. This Layer-3 includes more specific formal and informal settlements along the river, around the lagoon, and those distributed in urban areas, as well as specific tourism activities, and factors that hinder adequate urban planning. The current network shows that overgeneralization has been overcome (see Appendix C, Figure A3).

Step 4. Identification of intangible (indirect drivers) and tangible (direct drivers) harmful factors

As the construction of the process model advances, tangible harmful factors directly affecting the environmental state were identified (e.g., inoperative WWTPs). These factors belong to the following layers: Layer-2 and Layer-1 of direct drivers (see Appendix B, Figure A2).

Step 5. Definition and construction of harmful semantic pathways

Sequences of indirect drivers linked to direct drivers were extracted as harmful semantic pathways. We have constructed the following harmful pathways: poor agricultural practices; mangrove resource loss; inoperative WWTPs; inefficient urban planning. Figure A4 of Appendix D shows that the number of arcs converging at the input of the node representing the concept of “Inoperative WWTPs” is 6, four of them have a solid arc and two have two dashed arcs. Solid arcs represent a greater impact on this harmful factor than dashed arcs, according to expert consensus. The largest number of arcs converge at the node of “inoperative WWTPs”. A similar situation occurs with the concept of “Inefficient Urban Planning”. These two harmful factors should be addressed before the others to improve WWM. Appendix D (Figure A4) and Appendix E (Figure A5), respectively, show both pathways.

We use a backward tracking to construct semantic pathways. For example, starting at the input of the inoperative WWTPs node, we can identify the upstream nodes that converge to it, and so on until reach the indirect drivers of Layer-1.

Step 6. Expert validation and refinement

Experts reviewed harmful semantic pathways to ensure contextual relevance and semantic coherence.

Step 7. Prioritization using MCDM methods

TOPSIS [46,47] is used to weight pro-environmental actions according to their priority. AHP [48,49] is used to weight the components of the criteria according to their importance. Appendix F (Figure A6) shows the hierarchical network that serves to guide the calculation performed through the TOPSIS and AHP methods. The 10 experts are also involved in assigning weights to pro-environmental actions and to the components of the criteria.

Step 8. Iterative feedback and applicability

The semantic network structure can be updated by adding or removing layers and relationships as new knowledge becomes available.

2.2.2. Methodological Safeguards and Key Points

• Overgeneralization/Over-specification Warning: We must be careful to avoid the following errors: (a) overgeneralization, which could be interpreted as an exaggerated abstract model, such as one composed of only two layers. For example, layer 1, which contains the anthropogenic activities linked to the layer composed of elements of the environmental state, is a highly abstract model; (b) the over specification, which is characterized by a large number of layers, with a specific concepts and their relationships irrelevant to the study.
• Contextual issue: The factors related to indirect and direct drivers, and the layer of environmental state, depends on the context in which this approach is applied. The number of layers also depends on the problem being studied.
• Expert involvement and Validation. Given the lack of reliable quantitative data, a multidisciplinary group of ten experts was involved in the construction, validation, and evaluation of the semantic network. The experts represented environmental management, wastewater engineering, biodiversity conservation, urban planning, agriculture, socioecological systems, regional development, and social anthropology.

Expert participation was used to: validate concepts and relationships; assess the plausibility of semantic pathways; support qualitative judgments grounded in real-world experience; assign weights to harmful semantic pathways to determine an order of priority, and assign weights to criteria components to determine an order of importance, using a MCDM methods.

• Interpretation of relationships. The relationships between the elements belonging to the layer 1 of indirect drivers and those of layer 2 of indirect drivers (see Appendix B, Figure A2) show that, as internal migration increases human settlements, and uncontrolled urban sprawl also increase, contributing significantly to waste generation (solid waste and wastewater) and jeopardizing the effectiveness of waste management. Furthermore, constant tourism also plays a key role in waste generation.
• Urban planning importance. Urban planning is one of the most significant factors hindering the fulfillment of municipal responsibilities, where the wastewater management system plays a crucial role [50,51]. A major concern of urban planning are the irregular or informal settlements due to the lack of access to adequate drainage services, suitable housing, potable water, and functional transportation systems [52,53].

2.3. MCDM Methods

The ten experts are also involved in assigning weights to pro-environmental alternatives in order of priority, using the TOPIS method. In addition, the experts assign weights to the components of the criteria in order of importance using the AHP method. The hierarchical network constructed for this study is shown in Appendix F (Figure A6).

3. Analysis and Discussion of Results

It is important to emphasize that the evaluation of the proposed approach and its results must be interpreted within the context in which it was applied. This study addresses wastewater management challenges typical of developing countries, where rapid population growth, informal urban expansion without control, technological limitations in operation and maintenance, weak governance structures, and limited environmental awareness coexist and interact. In such contexts, pressure factors do not arise from isolated drivers but from multifactorial processes involving population growth, socioeconomic, sociopolitical, technological, and cultural dimensions. Consequently, the prioritization of alternative solutions or pro-environmental actions (PEAs) derived from the semantic network reflects these structural conditions and should not be directly extrapolated to developed-country settings, where institutional capacity, technological robustness, and environmental governance differ substantially.

Conventional decision-support and conceptual frameworks for wastewater management are predominantly developed and validated in data-rich, institutionally stable settings. However, in developing-country contexts, where rapid population growth, informal urbanization, weak governance, technological limitations, and low environmental awareness coexist, such frameworks often fail to capture the processes through which environmental pressures are generated. This study explicitly addresses this gap.

3.1. Key Data on WWTPs, Urban Areas, and the Course of the Sabana River

The development of Acapulco attracted internal migration in search of employment leading to informal settlements [54]. As a result, the growing urban sprawl also generates a greater demand for basic necessities such as housing, drainage, garbage collection, transportation, and availability of potable water, among others.

The municipality is divided into two zones: the beach area, where hotels, resorts and luxury settlements are located, and the city, where most of the low-income residents, including migrants, reside, highlighting inadequate urban planning and service provision [55]. Groundwater is frequently used instead of treated wastewater for various purposes, revealing governance challenges [56,57].

The WWM is contextualized as a complex dynamic system due to its holistic–multifactorial nature, causing environmental degradation.

We have proposed semantic networks involving multifactorial interactions. From this networks, we can identify specific concepts and their relationships that cause damage to the environment. Furthermore, we propose improving the understanding of overall semantic network by constructing the following harmful pathways: pathway of poor agricultural practices; pathway of mangrove resource loss; pathway of inoperative wastewater treatment plants; and pathway of inefficient urban planning. However, we are focused on the pathways of inoperative WWTPs and inefficient urban planning, because they are considered the main harmful pathways in this study.

3.1.1. The Inoperative Wastewater Treatment Plants

Figure 3 displays important numerical and visual data related to wastewater generation. Discharge points are represented by red dots (21 sites) and wastewater treatment plants (WWTPs) by orange dots (18 WWTPs), both distributed throughout the municipality of Acapulco. The problem of waste discharges into the river does not only damage the coastal lagoon but also tourist areas such as beaches, hotels, and restaurants, as well as the urban population. As an example of the prevailing lawlessness in this municipality, we can see that 7 out of 21 discharge points distributed along Acapulco Bay, where most tourist activity is concentrated, are located here.

Figure 3. Wastewater treatment plants (WWTPs, orange dots) and discharge points (red dots) distributed throughout the Acapulco Municipality [58].

There are 18 WWTPs in the Acapulco municipality. However, because the WWTPs are not operating efficiently enough, they discharge untreated wastewater into the Savana River, which flows through urban areas from north to south and feeds the coastal lagoon. The Sabana River suffers from at least four sources of pollution: (1) wastewater discharges; (2) malfunctioning of the WWTPs; (3) lack of basic services such as drainage and a garbage collection; (4) waste dumped by the urban population itself as well as the tourist infrastructure, such as hotels, restaurants, service shops and consumer products.

One clear conclusion is that a significant portion of the problem of the wastewater generation stems from the fact that a large number of WWTPS are no longer operational due to inadequate operation and maintenance of the WWTPs. In addition, they discharge untreated wastewater into the Savana River, which runs through the municipality from north to south, where important, densely populated urban areas are located, feeding the Tres Palos lagoon with wastewater.

3.1.2. The Inefficient Urban Planning

Acapulco experiences significant internal migration due to shortage of jobs in neighboring towns, leading to irregular settlements that put pressure on urban needs [59,60,61,62]. In this municipality, groundwater is often used instead of using treated wastewater for various purposes.

Based on the above arguments, we can affirm that urban planning together with wastewater treatment technology plays a relevant role, to a large extent, in supporting wastewater management to meet the water demand in the municipality.

The Urban Development Plan (UDP) for the city of Acapulco was approved in 2001, but it was not fully completed. Between 2013 and 2021, the corresponding administrations updated it, but without following-up. Additionally, updating the UDPs requires citizen participation, including from vulnerable groups and government entities, to support the concept of democratic governance in the public policy formulation processes. However, the participation of vulnerable groups has been virtually nonexistent.

The development and updating of UDPs require public consultation, mandated by law, at both the state and municipal levels. However, vulnerable groups do not attend the working meetings that address the planning processes. Instead, specific groups, along with government groups, are overrepresented, acting as representatives of social groups [62].

As we can see the urgency of developing an urban planning is a clear challenge that must be addressed combining population growth, socioeconomic, sociopolitical, technological, and culture factors.

Figure 4 shows the course of the Sabana River through important neighboring urban areas, such as Llano Largo and Ciudad Renacimiento, and other significant urban areas with formal and informal human settlements derived from a lack of planning, leading to a lack of services, which in turn causes significant human discharges of liquid and solid waste directly into the river or through its tributaries. Moreover, along the river, from its source to the lagoon, industries discharge wastewater into the Sabana River.

Figure 4. The course of the Sabana River through important neighboring urban areas, such as Llano Largo and Ciudad Renacimiento [63].

The thick blue lines represent the main course of the Sabana River. The thin blue lines represent its tributaries. It can be easily inferred that the combination of the Sabana River’s and its tributaries, the transport of debris during its flow, the distribution of inoperative WWTPs and discharge points, and the expansion of urban areas due to inefficient urban planning interact to generate waste, hindering proper wastewater management, and consequently damages the environment.

3.2. Results Derived from the Semantic Network and MCDM Analysis

The application of the proposed semantic network-based approach resulted in the explicit identification of harmful semantic pathways linking indirect drivers (anthropogenic activities) to direct drivers (pressure factors) that damage the environmental state, represented by waterbodies, ecosystem services, and vulnerable urban areas.

Through the stepwise construction of the semantic network, multiple abstraction layers were introduced between indirect and direct drivers, allowing for the identification of specific tangible harmful factors, including inoperative wastewater treatment plants (WWTPs), inefficient urban planning, loss of mangrove resources, and bad agricultural practices. These factors constitute the final layer of direct drivers that exert measurable impacts on the environmental state.

Based on these harmful semantic pathways, four pro-environmental actions (PEAs) were derived as alternative solutions:

(i)

restoration of WWTPs;

(ii)

improvement of urban planning;

(iii)

reforestation of mangroves;

(iv)

changes in agricultural practices.

The prioritization of these alternatives was performed using the TOPSIS method. Each PEA was assigned a performance score within a normalized range from 0 to 1, reflecting its relative effectiveness under environmental, social, and economic criteria. The results show a clear ranking, with restoration of WWTPs obtaining the highest priority score, followed by urban planning improvements, mangrove reforestation, and changes in agricultural practices. Table 2 shows the ranking of alternative solutions using the TOPSIS method.

Table 2. Ranking of the alternative solutions through the TOPIS method.

Alternative Solutions	Ranking	Normalized Values
WWTPS restoration	1	1
Urban planning	1	1
Mangrove reforestation	3	0.666
Change in agricultural practices	4	0.586

In parallel, the Analytic Hierarchy Process (AHP) was used to evaluate the relative importance of criteria associated with the identified harmful pathways. The resulting weights indicate that criteria directly linked to water quality degradation and wastewater discharge exert the greatest influence on the prioritization outcomes. Table 3 shows the order of importance of criteria.

Table 3. Criteria ranking through the AHP method.

Criteria	Ranking
Environmental benefits	52.7%
Social benefits	28.4%
Economic benefits	11.7%
Sustainability	5.9%

The consistency ratio related to the AHP method was: CR = (0.021/0.90) = 0.023 < 0.10. Therefore, the consistency between pairwise comparisons test has been overcome.

These results constitute the direct outputs of the proposed method, derived from the semantic network structure and the expert-supported multi-criteria decision-making process.

3.3. Interpretation of Results in the Context of Wastewater Management

The prioritization results reflect the structural logic embedded in the semantic network and the contextual characteristics of the study area. The dominance of WWTP restoration as the highest-ranked PEA indicates that technological failure in wastewater treatment represents the most critical leverage point for improving environmental conditions.

This result emerges naturally from the semantic pathways, where multiple indirect drivers converge on WWTP malfunction as a central pressure factor affecting waterbodies, ecosystem services, and public health simultaneously. From a systems perspective, this confirms that interventions targeting WWTP functionality produce multidimensional benefits, reinforcing both environmental and social outcomes.

Urban planning also appears as another one first influential action, with minimal difference with WWTPS, highlighting the role of uncontrolled urban expansion in amplifying wastewater generation and overloading existing infrastructure. The semantic network explicitly captures how population growth, tourism, and informal settlements interact to produce pressure factors that cannot be effectively mitigated through isolated technological measures alone.

Mangrove reforestation, while ranking lower, plays a complementary role by contributing to ecosystem services that provide resources and as buffer environmental impacts, particularly in coastal and lagoon systems. Changes in agricultural practices, although important, rank lower due to their more indirect influence on wastewater-related pressures within the studied context.

Overall, the interpretation of these results confirms that priority actions emerge from the structure of multifactorial interactions, rather than from predefined assumptions, and that the ranking is strongly context-dependent.

3.4. Methodological Positioning: Comparison with DPSIR-Based Approaches

While the DPSIR framework has been widely used to analyze environmental problems [64,65,66,67]; its application often remains limited to a small number of abstraction levels between driving forces and pressures. In contrast, the proposed semantic network-based approach explicitly models the process through which indirect drivers generate direct drivers by introducing multiple intermediate layers.

This distinction is critical for wastewater management systems, where pressures do not arise directly from broad drivers such as population growth or socioeconomic conditions but from specific processes such as settlement patterns, infrastructure degradation, and institutional weaknesses.

Furthermore, unlike traditional DPSIR applications, the semantic network allows for structural flexibility, enabling the addition or removal of concepts and relationships as new knowledge becomes available. This supports a dynamic feedback mechanism that improves understanding and decision-making over time.

The methodological novelty of this study therefore lies not in replacing DPSIR but in operationalizing its weakest components—namely the driving-force and pressure layers—through a process-based, multi-level abstraction model that enhances causal transparency and practical relevance.

Conceptual frameworks such as DPSIR have been widely used to structure analyses of environmental problems driven by anthropogenic activities. However, several studies have reported limitations in the representation of driving–pressure relationships, particularly regarding process representation, feedback dynamics, and multi-level abstraction (e.g., [68,69,70]). These limitations motivate the exploration of complementary approaches that explicitly model the processes through which indirect drivers generate direct pressures. Table 4 shows a comparative study between the DPSIR and the semantic network approach of this study, highlighting the points analyzed and discussed previously.

Table 4. Comparative study between DPSIR and the semantic network approach.

Key Points	DPSIR Framework	Semantic Network-Based Approach
1. Holistic approach: multifactorial interactions and multidisciplinary context	DPSIR conceptually integrates multiple components but is often applied in a fragmented or sector-specific manner, limiting holistic and multidisciplinary interpretation [68,71].	Explicitly holistic, integrating multifactorial interactions and multidisciplinary expert knowledge.
2. Several levels of abstraction between driving-force and pressure factors	Driving forces and pressures are commonly represented at a single or limited level of abstraction, masking intermediate mechanisms [68,69].	Multiple intermediate layers explicitly represent progressive causal mechanisms.
3. Process-based approach	Frequently used as a descriptive or analytical framework rather than as an explicit process model [70,71]	Explicitly process-based, modeling how indirect drivers generate direct drivers through semantic pathways.
4. Confusion or ambiguity between driving-force and pressure factors	Empirical applications often show overlap or ambiguity between driving forces and pressures, especially in complex socio-environmental systems [69,71].	Clear semantic distinction between indirect (intangible) and direct (tangible) drivers.
5. Flexibility to update structure	Framework structure is generally fixed once defined for a case study [70,71]	Highly flexible; layers and relationships can be added or removed as knowledge evolves.
6. Feedback dynamics	Feedback is conceptually acknowledged but often weakly implemented and typically delayed until the response stage [68,70].	Feedback is intrinsic and can be introduced at multiple layers to refine causal understanding
7. Identification of specific concepts and relationships causing impacts	Often identifies broad categories of drivers and pressures, with limited resolution of specific causal mechanisms [69,72].	Enables identification of specific concepts and causal relationships responsible for impacts.
8. Data scarcity and expert knowledge integration	Largely data-driven, offering limited guidance in data-scarce contexts [68,72].	Designed to operate under data scarcity using expert-driven qualitative assessments.

3.5. Implications for Planning and Decision-Making

The results demonstrate that the proposed approach aligns naturally with planning processes by supporting prioritization, sequencing, and feasibility assessment of interventions. Rather than aiming to address all environmental objectives simultaneously, the method highlights the necessity of staged actions, beginning with those that exert the greatest systemic influence.

By revealing which pressure factors dominate environmental degradation, the semantic network provides decision-makers with a transparent justification for prioritization, particularly in data-scarce contexts typical of developing countries.

Importantly, the approach is transferable to other coastal tourist municipalities experiencing similar dynamics of population growth, tourism pressure, and inadequate wastewater infrastructure, reinforcing its broader relevance for wastewater management research and practice.

4. Conclusions

This study addressed wastewater management as a multifactorial environmental problem characterized by multiple interactions of anthropogenic activities that generate harmful pressures on waterbodies, ecosystem services, and vulnerable urban areas. The analysis was grounded in a real-world context representative of coastal tourist municipalities in developing countries, where rapid population growth, weak planning capacity, and limited institutional resources exacerbate environmental degradation. Framing wastewater management within this context is essential to avoid generic solutions that fail to reflect on-the-ground constraints.

The first contribution of this work lies in introducing a process-based and context-driven approach to improve wastewater management by explicitly modeling how indirect anthropogenic drivers generate direct harmful factors. Unlike conventional conceptual frameworks that rely on single-layer or static representations, the proposed approach incorporates multiple levels of abstraction, enabling a clearer understanding of how complex interactions evolve and where interventions can be most effective. This process perspective is critical for addressing environmental damage that cannot be explained through linear cause–effect relationships.

Another contribution is methodological and concerns the use of semantic networks to support decision-making under conditions of complexity and limited data availability. Through the construction of harmful semantic pathways, the approach made it possible to identify specific combinations of drivers and pressures responsible for environmental degradation. These pathways served as an analytical bridge between qualitative expert knowledge and structured evaluation, allowing for the prioritization of pro-environmental actions using multi-criteria decision-making methods. The resulting rankings and weights were therefore outcomes derived from the semantic analysis, rather than assumptions defined a priori.

By positioning semantic networks as a decision-support tool, this study demonstrates how they can enhance wastewater management strategies by improving the transparency, interpretability, and adaptability of environmental knowledge. The flexibility to update network structures, add intermediate layers, and refine relationships provides a practical advantage over traditional frameworks such as DPSIR, particularly in contexts where feedback dynamics and multi-level interactions are critical but often overlooked.

From a practical perspective, the findings highlight that effective wastewater management interventions depend not only on technological solutions but also on governance, planning capacity, and environmental awareness. The results emphasize the importance of prioritizing actions that are both environmentally effective and contextually feasible, especially in developing-country settings where institutional and economic constraints strongly influence outcomes.

Finally, it is important to acknowledge the methodological boundaries of the proposed approach. Excessive abstraction risks overgeneralization, while excessive detail may lead to over-specification without analytical benefit. The number of layers, the selection of drivers, and the structure of semantic pathways must therefore be defined according to the specific problem and context under study. Future research should focus on the integration with quantitative monitoring data, where available, and application to other environmental management challenges characterized by complex anthropogenic–environment interactions.

A central strength of the proposed approach is the explicit and structured involvement of multidisciplinary experts in both knowledge construction and decision prioritization. The selection of anthropogenic drivers (population growth, socioeconomic, sociopolitical, technological, and cultural factors) and the components of the environmental state (waterbodies, ecosystem services, and vulnerable urban areas) was not arbitrary but the result of expert consensus grounded in the realities of developing-country contexts, where data scarcity, institutional limitations, and rapid socio-environmental change are common. Ten experts from diverse disciplines—including environmental management systems, socioecological systems, biodiversity and conservation, urban engineering, wastewater management, agriculture, environmental modeling, regional development, and social anthropology—participated in defining network layers, validating harmful semantic pathways, and evaluating their relative importance.

These experts also applied multi-criteria decision-making (MCDM) methods, using TOPSIS to prioritize pro-environmental actions (PEAs) and AHP to weight the components of the criteria, thereby translating qualitative domain knowledge into structured, transparent, and reproducible results. The combined use of expert judgment, semantic networks, and MCDM techniques ensured that the prioritization of harmful pathways and associated PEAs emerged from the proposed methodology itself, rather than from predefined assumptions. This expert-driven framework is particularly relevant for wastewater management in developing regions, where reliance on purely data-driven models is often impractical, and informed judgment becomes an essential component of robust environmental decision-making.

Evaluating decision-support approaches developed for developing-country contexts using assumptions derived from high-income settings may lead to misleading conclusions regarding novelty or effectiveness. In such contexts, methodological value lies not in optimization under ideal conditions but in the capacity to structure complex processes, reveal hidden drivers, and support prioritization under uncertainty and institutional constraints.