A stakeholder-driven holistic approach to mitigate drought in Ramsar wetlands: evaluation of nature-based solutions using interval analytic hierarchy process

The term ‘wetlands’ encompasses a wide range of habitats and environments, typically describing both periodically and permanently flooded lands and shallow water areas with abundant plant growth that may periodically dry up. These areas present highly productive ecosystems that, when not under external threats, can offer a great array of ecosystem services compared to terrestrial ecosystems (Convention on Wetlands 2021). Although wetlands cover between 5 and 8% of the Earth's land surface, wetlands supply 50% of the global ecosystem services that support human welfare, climate change mitigation, and biodiversity preservation on local to global scales (Ballut-Dajud et al. 2022). Wetlands are one of the three major ecosystems on Earth (Fang et al. 2023) that are worldwide threatened by both anthropogenic and natural drivers. Authors Fluet-Chouinard et al. (2023) point out that wetland loss has primarily occurred in Europe, the USA, and China since the middle of the twentieth century. Thus, as a result, in the territory of Europe, up to 90% of wetland areas have disappeared (Junk et al. 2013) and it is projected that the intensification of the effects of climate change will increase that percentage. In the review paper, Ballut-Dajud et al. (2022) confirmed that climate changes and human activities, including agriculture, urbanization, and industry exert the most significant impact on natural wetlands. The sustainability of wetland ecosystems, designated as Ramsar wetland sites, is under threat from external stressors such as climate changes, land reclamation and drainage, excessive exploitation of natural resources, and the expansion and intensification of agricultural activities in the bounded areas (Convention on Wetlands 2021). In this paper, the focus will be on Ramsar wetlands due to their significant role in the protection and conservation of endemic and endangered plant and animal species, as well as a sensitive aquatic ecosystem.

An increasing number of challenges due to climate change are exacerbated by people's irresponsible and careless behavior, leading to increasingly extreme weather events like droughts and floods. As it is stated by the Global Commission on Adaptation (2019; p. 61), ‘the effects of climate change will most immediately and acutely be expressed through water’. Thus, wetlands are vulnerable to climate change, especially to rising temperatures, reducing precipitation, and consequently – altering hydrology due to drought events. According to Biswas Roy et al. (2022), recent droughts resulted in the drying of surface water bodies and a reduction of wetlands area. Wetlands present fragile ecosystems and require additional efforts in their preservation. Nowadays, there are global efforts to define suitable approaches and diverse solutions and measures for mitigating drought risk in wetlands and addressing the water challenges (Donatti et al. 2022). Given the importance of the preservation of wetlands and based on the abovementioned interpretation, it can be concluded that it is necessary for the invention of adequate tools and techniques that would reduce the vulnerability of wetlands to external pressures. This could be supported by a shift toward thinking about the multifunctionality of wetlands and the simultaneous provision of multiple ecosystem services by examining the consequent relationships of possible regulation and restoration measures. These measures must consider a range of political, economic, social, technological, biological, and environmental concepts through the holistic approach of integrated assessment.

Nature-based solutions (NbS) have received rapidly growing recognition and enthusiasm for their potential to address climate-related challenges. Decision- and policy-makers are increasingly accepting the contribution of NbS to promote adaptation to climate change and decrease water-related concerns (Pagano et al. 2019) leading to involvement in various global initiatives and European policies (e.g. the Sustainable Development Goals, the Paris Climate Agreement, the Strategic Framework for Climate Change and Development by the World Bank). Specific NbS interventions and infrastructure can deliver simultaneous benefits to people (human well-being) and biodiversity (a healthy environment) and have to fit well with the existing landscape features, land uses, and infrastructure construction to be effective. Furthermore, NbS practices must strive to improve existing environments to build more resilient natural systems (Seddon et al. 2020) and to minimize economic and social difficulties (Ruangpan et al. 2021) due to severe climate changes and climate-related hazards. To promote the acceptance as well as the implementation of NbSs, decision support tools, the perspectives of numerous stakeholders, trade-offs, and practical solutions must be taken into consideration (Giordano et al. 2020). Inclusive processes require balancing of diverse stakeholders’ interests and conflicting demands to contribute to the relevance and acceptance of NbS as well as to ensure equitable citizen engagement in civic decision-making (Naumann et al. 2023). Wetlands represent highly complex and multifaceted systems that require integrative, holistic perspectives that support processing different types of information related to the ecosystem health and functions of wetlands. Despite the simplicity of the NbS concept, there remain gaps because of specific local constraints and socioeconomic circumstances where choosing appropriate NbS is still challenging. There is no particular NbS that can resolve each possible issue and the best measure will depend on the needs and characteristics of Ramsar wetland. Making decisions is challenging in real life and practice due to, e.g. the complex nature of water resources management and the related uncertainty, or any instance whenever both quantitative and qualitative information needs to be processed. Furthermore, the implementation of NbS could be challenging due to a lack of communication and cohesiveness among different cross-sectoral stakeholders Chee et al. (2021). The necessity of inclusion of the opinions of local community groups and other stakeholders into the decision process in environmental sciences is promoted in Huang et al. (2011).

Ruangpan et al. (2021) from the Serbian case study stated that ‘the stakeholders’ weighting of measures is important since it can be used to enhance identification of the suitable measures for the specific case study.’ In general, to successfully adopt NbS, this reference emphasized the value of including stakeholders in projects at an early stage. Although their work focuses on the implementation of flood risk reduction strategies, this viewpoint on stakeholder participation should not be dismissed when addressing drought risk reduction as well. One of the recommendations to engage stakeholders and include local community actors to promote the NbS concept in Serbia is found in the publication by IUCN provided by Popovicki (2022). In Serbia, there is a lack of involvement in this type of climate change strategy in the institutional, policy, and legal framework and it is necessary to engage NbS into national disaster risk policies. The term NbS is not explicitly defined in Serbian legislation; it is in the initial process of recognition where NbS measures or possible solutions that can be partially identified as NbS are proposed as a promising approach to deal with social and climate challenges. Despite the widespread recognition of NbS in global policy and the growing literature on its potential as a mitigation strategy for addressing human-induced climate change, research on wetlands, particularly Ramsar wetlands, remains scarce or nonexistent in Serbia. After reviewing the available scientific literature and publications in the field of NbS application in Serbia it was observed that most of the papers are about urban development and the preservation of urban well-being (e.g. Mitić-Radulović & Lalović 2021; Greksa et al. 2023). Authors have observed that climate change effects are generating immense stress on Ramsar wetlands and it is necessary to react in the field of inclusive participative decision-making of proper management plans. In the group model assessment suggested by Srđević et al. (2021a) concerning perspectives of future planning and management, and prevention of risks, ‘Koviljsko-petrovaradinski rit’ (KPR) was chosen as second regarding the significance of Ramsar sites located in Vojvodina Province, Serbia. Moreover, authors Milentijević et al. (2022) indicated that dry periods can be negatively reflected on KPR based on the statistical analysis of precipitation and air temperature. Drought outcomes were determined in the entire Vojvodina area based on the water deficit in 76% of the summer months, which was more prevalent during the period 1987–2016 compared to the period 1948–1990 (Grujić et al. 2021). Also, it was found that extremely long droughts (up to 75 days) can be expected every 100 years during the April–September period (Srdjevic et al. 2021c). These findings underscore the critical importance of considering KPR and its vulnerability to climate-related factors in the strategic planning and conservation of Ramsar sites in Vojvodina Province.

One of the most favored decision support systems in managing water-related risk management is multi-criteria decision-making (Abdullah et al. 2021). A successful approach to reaching decisions based on multiple considerations in the preservation of wetlands can be proposed through the analytical hierarchy process (AHP) among other methods (Zhang et al. 2013; Talukdar & Pal 2020; Bakirman et al. 2022). The AHP methodology was already in widespread use by the middle of the twentieth century as a structured technique for solving complex decision-making problems. Today, AHP serves as a tool for supporting both individual and group decision-making processes in different real-life circumstances. The decision problem is structured hierarchically at different levels and can contain a finite number of criteria or alternatives; the goal is at the top followed by lower levels with criteria and subcriteria, and the bottom level with alternatives. Decision-makers (DMs), which can be decision- and policy-makers, stakeholders, practitioners, or other experts, provide n(n-1)/2 pairwise comparison matrices demonstrating the relative importance between observed n elements to assign weights. This procedure always concerns the superior elements and higher-ranking aspects in the hierarchy providing crisp values, values expressed in intervals, or even no values for judgments over paired elements. The ratings of the criteria or alternatives are produced as priority weights from a pairwise comparison matrix using either existing prioritization method. Sometimes it is quite difficult to provide a precise numerical value of preference due to the complexity and unpredictability involved in real-world decision problems, as well as possibly missing information or DM's expertise. The success of the method has been the subject of academic and other discussions in a variety of ways, including (1) using scales to compare decision-making components in pairs, (2) fuzzification of scales and the comparison process, and (3) using interval comparison matrices. In this paper, only the last mentioned method is proposed as a possible decision support tool for selecting suitable NbS as a measure in mitigating the effects of drought on the Ramsar site.

The remainder of the paper is organized as follows. In the following Section 2, the decision-making problem is described in more detail; the AHP hierarchy is presented according to goal and alternatives, the Serbian case study is described, possible nature-based solutions are selected for Ramsar wetland, and experts are defined. The interval AHP methodology is defined in Section 3. Section 4 presents the results of the proposed approach demonstrated through analysis of the ranks and weights provided by I-AHP. Section 5 gathers a summary of key findings and a future research agenda.

Contrary to full AHP implementation, which involves comparing elements at all levels, I-AHP simplifies the process by focusing on the evaluation of alternatives according to the goal. In this paper, an I-AHP is presented as an easy-to-follow procedure of evaluation of six core activities (alternatives) to establish a framework for the effective identification of priority actions and for improving the current and future conditions (the goal). Experts conducted a pairwise comparison of elements at Level 3 with respect to Level 1 of the hierarchy, focusing on the relative importance of alternatives according to the goal.

The population within the protected area's borders is primarily engaged in agriculture, which is an indirect driver of negative changes in water and soil quality along with the effects of drought. Wastewater also has a significant impact on the sustainability of the water resources, since untreated effluent from Novi Sad is released upstream, just a few kilometers from the borders of the protected area. Domestic permeable septic tanks, which discharge into drainage channels, have an impact on the usage and water quality as well (Srdjević et al. 2021b).

Since KPR is located in the lowest geomorphologic complex of the alluvial plain and has a plain character, an increase in the frequency and severity of droughts could be potentially dangerous to aquatic ecosystems. The branches of the Danube River are characterized by slow flow and mild descent which is the reason for the existence of numerous wetland formations that are endangered during dry events. Meteorological droughts gradually emerge providing time for proactive water conservation and the use of sustainable practices (Rezaiy & Shabri 2023). A review paper provided by Moraes et al. (2022) shows that wetlands restoration measures are dominant as solutions for various improvements in the area, while the review paper (Yimer et al. 2024) lists a number of alternatives as an effective NbS measure that can mitigate the negative effects of drought.

In this paper, a set of measures is proposed that fit into the NbS context and may serve as solutions that can be implemented through the engagement of the local community. The measures are discussed in detail and could be considered as solutions applicable to any Ramsar wetland worldwide based on certain characteristics that specific locations have in common. Selected alternatives for the Serbian case study are described later.

A1 – Increasing the water table in waterways: Wetland environments are vital, multifunctional ecosystems that depend on the availability of water supplies. During dry periods when monthly rainfall is reduced, less water is available for natural and biological processes and for providing water-related ecosystem services as well. Increasing the water tables in wetlands and increasing low flows and water levels during droughts (Kadykalo & Findlay 2016) can have an impact on evapotranspiration, infiltration, and groundwater recharge (Ferreira et al. 2023). Given the abundance of different water body types in the KPR, it is necessary to ensure their sustained optimal water supplies to support the ecosystems adapted to local specific characteristics. The following are some practices for increasing the amount of water in wetlands: restoring natural creeks, reconnecting rivers to floodplains, restoration of natural infiltration to groundwater, riverbed material renaturalization, reconnection of oxbow lakes, and restoration and reconnection of seasonal streams.

A2 – Agroforestry is defined by FAO as ecologically based natural resource management that through the integration of multiple forestry components in the landscape (mostly agricultural) sustains production increasing environmental and social benefits. It involves forestry techniques that allow for the conversion or restoration of wetland landscape components. In the KPR region (where it is permitted, mostly in the II and III levels of protection) and in surrounding areas of the protected area itself, the local community is primarily dedicated to agriculture, forestry, and fishing. As stated in the publication by WWAP (2018), maintaining the integrity of agricultural and forest cover helps to keep the soil's ability to retain water, which enhances water security during droughts. Peak flow control structures in managed forests, removal of dams and other longitudinal barriers, and elimination of riverbank protection are some of the actions that could increase water availability during meteorological or hydrological drought (Martin et al. 2020) while benefiting the local community (Murniati et al. 2022). In wetlands where agricultural activities have a significant negative impact, as is the case in KPR, such solutions could be appropriate and effective in maintaining water resources.

A3 – Climate-smart agriculture presents an integrated landscape management approach that will help to adapt agricultural systems to the effects of climate change. This type of innovation can help not only DMs but also farmers to mitigate the effects of climate change both in the short and long term (Vincent & Balasubramani 2021). Wetlands support agricultural activities as a source of freshwater for both crops and livestock providing fertile land. This group of actions covers all agricultural activities that tend to be environmentally sustainable, such as strip cropping, intercropping, no-till and low-till agriculture, early sowing, and green cover crops. The aforementioned practices improve soil quality (retention of organic matter) and biodiversity (nutrient cycling), minimize soil erosion, help to preserve the natural characteristics of agricultural arable land and increase the overall yield of arable land. For the Ramsar wetlands, the Convention on Wetlands (2021) prefers that such solutions should not adversely affect the ecological character of wetlands ensuring their wise use.

A4 – Natural water reservoirs: Water reservoirs play an important role in the wetland environment by supporting wetland biodiversity. Considering the various performances they have, water reservoirs are mostly significant in drought management and flood protection. During dry periods, water reservoirs could serve a multifaceted role in wetland water resource management in Ramsar. Reservoirs store water to satisfy essential needs, including for irrigation purposes in agriculture and forestry, to satisfy human and agricultural water demands, to support fish farming, and for powering small water plants for economic purposes. Then, certain reservoirs could be designed for recreational and esthetic purposes, enhancing the environment and providing leisure opportunities. Water ecological reservoirs play an important role in the intent of creating enclaves for aquatic flora and wildlife, while also functioning as filtration sites to purify natural water sources (Mioduszewski 2012). Natural water reservoirs can be associated with the idea of building dams on streams, but in the case of KPR, adequate storage volumes could be obtained from the restoration of existing lake or wetland ecosystems. By restoring possible storage water areas in KPR, it is possible to reduce water-related risks such as droughts and floods while improving water quality and quantity simultaneously (Acreman et al. 2021).

A5 – Mulch coverage: Mulch is a type of organic material (wood bark, green waste, crop residues, compost, manure, straw, dry grass, leaves, etc.) that is applied to the top layer of the soil surface (Critchley et al. 2023). Its primary functions include preserving moisture, enhancing soil fertility and health, increasing the capacity of the soil to retain water, reducing weed growth, protecting plant roots from freezing, and enhancing the area's esthetic attractiveness. Mulch can be used to wetlands preserve areas that are most vulnerable to drought, such as smaller water areas that must maintain a certain humidity level to ensure the survival of ecological systems. This measure can be chosen as mulch material has high quality to resist drought, high temperature, and other adverse conditions in the long term (Altieri et al. 2015).

A6 – Wastewater reuse: Wastewater can contain toxic substances that can be harmful to the aquatic ecosystem. The main drivers of wetlands quality decline can be linked to wastewater discharges (de-los-Ríos-Mérida et al. 2021). Agricultural fields, where farming and the use of fertilizers increase the runoff of nutrient-rich wastewater into water bodies, are the main pressures from the boundaries of KPR despite the status of a protected area (Srđjević et al. 2021b). Control of water pollution due to excess nutrients can be achieved through different types of NbS interventions or technologies. When wastewater is properly recycled and reused, it can be used for various purposes (e.g., irrigation, tourism, esthetic value), which is especially important during dry periods.

To collect stakeholders’ perceptions of the best alternative to reduce the drought effect on selected Ramsar sites, three experts were included in the research. Among them were two professors and one Ph.D. candidate, all related to water management with relevant studies from the case study of KPR. Primarily, this research is intended to examine the feasibility of applying I-AHP to a smaller group of experts. After validation of the I-AHP application in selecting NbS, further research will involve a broader expert consensus; the expanded group will encompass local stakeholders, including members of non-governmental organizations, individuals from various associations, local farmers, business owners, etc.

A multi-criteria decision-making framework is used for ranking the overall performance of decision alternatives against multiple objectives considering different types of criteria. In the standard approach of the AHP methodology, each rating represents a judgment concerning the relative importance or preference of decision element i over j proposed by a given decision-maker or expert. The scale is an integer provided as a crisp, unique number in the bounded discrete scale (1/9, 1/8, …, 1/2, 1, 2, 3, …, 9) suggested by Saaty (Table 1) in 1980. A rating of 1 from this scale indicates the two elements (subcriteria or alternatives) are equally important, a 9 indicates element i is absolutely more important than element j, and a 1/9 indicates element j is absolutely more important than element i. This scale and reciprocal values are in line with people's psychological habits when making judgments.

The results of each comparison for each matrix are processed further to derive the priority vector where and ⁠. The pairwise comparison allows the DM or experts involved in the evaluation to express their preferences considering two alternatives at a time simultaneously. However, this way restricts the experts’ understanding of the preferences of all the elements in the hierarchy, including alternatives, criteria, and subcriteria (Wu & Xu 2012). AHP is completed after all local priority vectors are synthesized; this approach provides a global vector of priorities of the selected alternatives with respect to the goal, derived with respect to the priorities of the criteria, and, when relevant, subcriteria as well.

In some instances, DMs are faced with situations where they must weigh many different feelings and opinions before making a final decision. Comparing different solutions by breaking down a complex decision into several smaller ones allows easier control of decision problems, which can also be performed using pairwise comparison of elements at one hierarchical level concerning elements at a higher level. Due to the complexity of decision-making issues, it is satisfactory to allow the DM may express his/her preferences through numerical intervals. This way generally defines the AHP approach as well, and Saaty & Vargas (1987) proposed an extension of AHP. Instead of defining preferences through crisp values, the DM can express preferences as a range of scale values. Based on the chosen semantic-numerical scale, DM preference is stated for pairs of decision components and numerical values are placed into the associated comparison matrix. If an interval of values (e.g. from the, 1–9 Saaty's scale) is defined in at least one matrix in place of a single value from the scale, the matrix is defined as an interval, and the standard approach of AHP is changed to an interval. One or more intervals may be present in the interval matrix, and the left and right bounds of each interval denote the minimum and maximum DM preference strengths. This way of expressing DMs’ preferences can reduce the influence of human subjective desire and better reflect the uncertainty of judgment (Gonzalez et al. 2022).

The variable is defined as a virtual variable with a function in constraint relations and this value should be minimized. Model (4) identifies peaks in a two-dimensional space of possible vectors/weights ⁠, and it is possible to obtain more solution vectors if the model is pivoted around that peaks n(n–1) times. Although the interval matrix must be consistent for the model to be computationally effective, the problem can prove unsolved if no solution satisfies all the initial constraints. To calculate the weights of hierarchy elements from inconsistent interval matrices, apart from linear ones, there are also mathematical techniques based on goal programming: min-max, lexicographic, logarithmic, etc. Today, there are a variety of linear programming tools and prioritization methods (e.g. Conde & Pérez 2010; Zadnik & Groselj 2013) that can identify some of these solutions with a certain level of accuracy when they have different minimum values for and different values for the ‘slack’ variables used in the model's constraint relations.

Experts evaluated alternatives and conducted pairwise comparisons of six alternatives by filling out interval-based decision matrices by addressing questions such as ‘How many times is alternative i is dominated by alternative ?’. The optimization model can be illustrated through different methods as demonstrated by previous works (e.g., Sarker et al. 2019; Singhal et al. 2024); in this paper, Arbel-LP model was utilized as given in (5). All interval-based matrices and an example of the Arbel linear model for an interval matrix produced by Expert #2 are given in the Supplementary Material.

The model that is created for each interval matrix in non-standard form contains 6 real (⁠⁠) and one auxiliary variable (⁠⁠); 30 constraint relations (two for each interval in the upper triangle of the decision matrix), one constraint relation that expresses the condition that the sum of the weights of the alternatives must be equal to 1, and one constraint relation that indicates that all variables are greater than or equal to 0. The transformation of the initial model into the standard form required the addition of 30 equalizing variables and one artificial variable, which means the expansion of linear space to 6 + 1 + 30 + 1 = 38 dimensions. In the first quadrant of the 38-dimensional space, there is a set of possible solutions and the value of an auxiliary variable is zero or very close to zero. As mentioned, the problem with interval matrices is their inconsistency. Additionally, Kress (1991) noted that linear programming can be useless for calculating weights for inconsistent matrices since there is sometimes no feasible region of solutions – the feasible region becomes empty. An illustration of a linear model that makes it possible to discover a solution for an interval matrix with a six-element dimension is provided in this paper. Because the discrepancy typically develops with the size of the matrix, the author's analyses with interval matrices of a higher order demonstrated that there are challenges where it is difficult to calculate the weight vectors.

Computing priorities of NbS alternatives A1–A6 (weights ⁠) is done by using a prioritization model based on linear programming. The resulting weights are presented in Table 2. The highest priority is assigned to alternatives A1 (increasing the water table in the main waterways; ⁠) by Expert #2, A2 (Agroforestry; ⁠) by Expert #1, and A4 (Natural water reservoirs; ⁠) by Expert #3, which generally represents restoration measures that require careful distribution of available resources to maximize their effectiveness in mitigating the effects of drought.

A closer look at the values of weights reveals that the weights of the top-ranked alternatives (A1, A2, and A4) hold a share of more than one-third of the total (bolded values). Cardinal information obtained by I-AHP holds significant relevance, particularly in applications related to the distribution of resources. Such results can be used, for example, if three solutions are to be implemented, one-third of financial resources should be allocated to each of the alternatives. This allocation can include financial funding for wetland conservation and restoration projects, personnel for monitoring and maintenance, as well as sustainable use of natural resources. The obtained results can assist in resource allocation by considering the perspectives of all stakeholders involved in the decision-making process and the creation of an allocation strategy.

When it comes to ordinal information obtained by I-AHP methodology, the results presented in Table 3 show that two experts (#1 and #3) identified alternative A1 (Increasing the water table in the main waterways) as the top-ranked, followed by A2 (Agroforestry) and A4 (Natural water reservoirs). Expert #1 showed preferences for A2 and A4 to be the top-ranked alternatives. Furthermore, all experts selected alternatives A1, A2, and A4 as the three top-ranked. Ordinal information related to the ranking of decision elements can ensure that limited and often scarce technical, technological, human, or other types of support are channeled toward the most critical needs or to identify the most vulnerable locations. Given that all the chosen alternatives are preferable in reducing the negative effects of drought, the ranking of alternatives provides enough information for this specific case.

The three top-ranked alternatives belong to the group of wetland-related measures that can contribute to the providing and balancing of water resources during dry periods. Li et al. (2023) concluded that increasing water flow in waterways can shorten the duration of low waters, which is a crucial factor to avoid disrupting the natural balance of biodiversity at the Ramsar wetland. These findings are supported by the fact that the same alternatives received the highest rankings in the paper Srđević et al. (2022b) after being evaluated by different stakeholder groups for the selected Ramsar wetland.

Tschora & Cherubini (2020) noted that measures related to agroforestry can reduce pressures in trade-offs for plot selection between forestry and agriculture sectors. The same happened in our research. After speaking with local managers in the KPR area, it appeared that conflicts between these two sectors also exist. Based on numerous consulted references, measures consisting of agroforestry practices can effectively resolve these conflicts and should be considered as the strategic cornerstone of the management plans (e.g. Cammerino et al. 2024).

Natural water reservoirs with a significant potential for redistributing surface water emerge as promising options. This feature proves advantageous in addressing shifting drought patterns and maintaining a stable water balance during dry periods (Wu et al. 2023). The fact that experts in this study agree that this measure is appropriate for the particular Ramsar wetland further supports the solution's viability.

As shown in Tables 2 and 3, the last ranked alternatives are mulch coverage (A5) and wastewater reuse (A6), the same as in the paper by Srđević et al. (2022b). Interesting to note is that the experts who took part in this research did not recognize the potential to reduce drought effects on Ramsar wetland, similar to other stakeholder groups in the aforementioned reference.

This paper addresses gaps in the existing literature by actively engaging various cross-sectoral stakeholders with diverse knowledge and experience related to Ramsar wetland management. The proposed methodology ensures comprehensive research of potential mitigation strategies, supporting the collective expertise of stakeholders from different sectors to contribute to a well-informed and robust decision-making process. An extended interval version of the AHP methodology is proposed for choosing the best NbS in reducing the impact of drought in the Ramsar area. During this research, the necessity of proper selection of the group of evaluators is emphasized aimed at ensuring that the evaluation can be considered valid if consistency within the permissible limits is evidenced and approved.

The use of interval AHP is proposed in this research due to the following reasons. When obtaining comparisons, inaccurate comparisons can occur due to incorrect interpretations, a poor understanding of the criteria, the distinctive motives and interests among different groups, or even the knowledge and personal experience of participating individuals in the decision-making process. As noted by Dede et al. (2022) in the process of providing relative relevance of the th element concerning the th one, it is ‘quite natural’ to have uncertainties. Furthermore, using integer numbers could oversimplify reality when the 9-point Saaty scale is used. The interval or fuzzy versions of the AHP methodology are proposed to solve decision-making problems under uncertainty in several references (e.g. Tavana et al. 2023). The advantage of utilizing the interval version of AHP over the fuzzy version of AHP lies in its simplicity of comprehension and practicality when engaging stakeholders with variations in preferences, interests, and knowledge in the decision-making process. On the other hand, fuzzy AHP requires an advanced defining value set characterized by trapezoidal and triangular fuzzy numbers and membership functions, which can be somehow complicate in application. Our experience shows that stakeholders could be more interested to get involved when input and output data are presented in a simple form.

Based on previous experience working with different stakeholders, the authors enhanced research by conducting additional assessments when the standard AHP (S-AHP) was employed as a control mechanism of the proposed I-AHP. The decision matrix presented in Table 4 was provided by an external evaluator and for comparison, an adjustment of I-AHP and expansion of flexibility by applying a ± 1 pivot is presented (Table 5). The selection of the decision matrix was determined by the consistency level of the evaluator, which should be ≤0.1 for S-AHP. In this particular case study, the consistency level was 0.04.

Table 4

The decision matrix provided by using S-AHP

The light marked cell (values of 1) represent the best, first-ranked alternatives from the comparison process. Cells with dark shading (values of 6) represent the last ranked alternatives from the comparison process.

View Large

Table 5

I-AHP decision matrix pivoted ±1 from pairwise comparisons in Table 4 

The light marked cell (values of 1) represent the best, first-ranked alternatives from the comparison process. Cells with dark shading (values of 6) represent the last ranked alternatives from the comparison process.

View Large

After analyzing the outcomes from the S-AHP and I-AHP matrices, it becomes evident that regardless of the number of alternatives being compared, alternative A1 (Increasing the water table in waterways) consistently stands out as the top-ranked alternative, while A6 (wastewater reuse) ranks as the least favorable option. In the context of I-AHP, the distinctions among the top three alternatives are smaller than in the case of S-AHP. On the other hand, the discrepancies among the last two ranked alternatives are more pronounced. For instance, in S-AHP, the difference between A5 and A6 is merely 0.033, whereas in I-AHP, this difference increases to 0.073. If the proposed I-AHP methodology is used for resource allocation, these variations can have a significant impact on the final decision. Therefore, it is necessary, depending on the experience and knowledge of those who are evaluating, to choose a methodology (either standard or interval version of AHP). Note that S-AHP employed as a control mechanism embodied the ‘spirit of AHP’ because there was no typical synthesis just local prioritization within AHP methodology. The results presented in this paper indicate that a smaller group of experts with greater knowledge can work with the standard version as well, which requires more time than the interval version, while the interval version is recommended in the case of working with a larger number of stakeholders with different backgrounds. An optional possibility to avoid the subjectivity of DMs is to add a step in the process itself when filling in the matrices at different time intervals, that is, to avoid favoring certain alternatives for various reasons, for example, economic, political, or personal interests. It should be kept in mind that, even in the case of standard decision-making with ratings without interval freedom, the creation of a method for monitoring the consistency of ratings presents an important obstacle.

By 2030, climate change is likely to occur, leading to wetter winters and hotter and drier summers. Understanding the various variables that affect the time scales, amplitudes, and frequency of drought becomes crucial when it comes to altered physical processes caused by drought's negative effects (Sarker 2022). Additionally, developing methodological approaches to mitigate or eliminate these effects is crucial. The significance of the AHP in the realm of environmental protection and climate change mitigation is presented in this paper as substantial. AHP can improve the efficacy and efficiency of environmental protection methods by empowering DMs/experts/stakeholders to analyze and prioritize different alternatives (and criteria) systematically, as well as to break down complex problems into more manageable components. Solutions aimed at improving environmental quality and mitigating the effects of drought can be identified with the support of experts and individuals who have extensive expertise and experience in the area under study. Broadly speaking, the application of interval AHP in environmental protection and climate change mitigation can strengthen the decision-making processes, encourage the adoption of viable and sustainable solutions and increase the ability of ecosystems’ and communities’ to cope with the climate change.

To address wetland vulnerability to drought, it is crucial to establish an integrative framework that simultaneously preserves biodiversity and supports human well-being. Wetlands are vulnerable ecosystems that are exposed to a variety of external disturbances and it is desirable to consider the adoption of NbS if there are opportunities that correspond with the Ramsar's environment and location. The establishment of new or strengthened local and national policies is necessary for conservation management. This can be done by developing and building strategic partnerships with multiple groups of interest, including local communities, businesses, government departments and agencies, academic institutions, and non-governmental organizations that can improve the current state of Ramsar wetland.

The widespread of various versions of multi-criteria decision-making methods in many applications can be attributed to their high accuracy and reliability in addressing different problems. A methodological decision-making approach, AHP, is scientifically validated and includes all vital components that enable the synthesis of the combined knowledge, experience, and expertise of DMs in both an individual or group context. In decision-making, people are usually influenced by their personality, knowledge, and social and political background when deciding between options. Due to the complexity of decision-making issues, it is satisfactory to allow the DM may express preferences through a range of numerical values. In this paper, an interval version of AHP, named as I-AHP, is presented as an easy-to-follow procedure of group-based ranking of six core activities (alternatives) to establish a framework for the effective identification of priority actions and for improving the current and future conditions (the goal). Ranking of alternative nature-based measures (solutions) is performed by three experts. The final decision was to determine the priorities of measures regarding the potential of mitigation of the effects of drought in the Ramsar site KPR in Serbia. The I-AHP results can be considered ordinal and cardinal information at the same time, and can be used by decision- and policy-makers. In instances where the allocation of financial investments and human resources is required, cardinal information is the preferred choice; when it is necessary to determine the order of execution of measures ordinal information is recommended.

Experts in the field of water management who participated in this research identified only alternatives that belong to the group of wetland-related measures. Proposed NbSs are deemed most effective in balancing water resources during dry periods, encompassing actions such as increasing water availability, supporting agroforestry practices, and utilizing natural reservoirs. Moreover, measures that took the last places include measures concerning the use of mulch and the reuse of wastewater. The reason why these two alternatives are ranked last is that, in the case of mulch application, it is necessary to determine the specific hotspots where the mulch should be applied; moreover, for the reuse of wastewater, it is necessary to allocate additional investments in the construction of treatment facilities. Given the shared characteristics of Ramsar wetlands, it is important to note that the findings presented in this paper do not automatically imply that the alternatives ranked as the last in the list would not always be effective solutions for Ramsar wetlands situated in diverse climatic and geographical conditions.

Considering the importance of Ramsar site's conservation in mitigating drought impacts, future research will focus on identifying effective strategies to increase water availability during dry periods, and on involving more stakeholders in the evaluation process. The proposed potential avenue direction for future research would challenge an issue of incorporating input from a larger number of participants within moderately sensitive interval matrices. The research agenda would include examining decisions made at the group level and assigning percentage shares as cardinal values to the options that are determined to be the highest ranked. This way, a new evaluation procedure could be created to involve a new iteration of weight calculations carried out by researchers rather than necessarily requiring repeated engagement of stakeholders in the evaluation.

This research was supported by the Ministry of Education, Science and Technological Development of Serbia (Grant No 451-03-65/2024-03/200117 and Grant No 451-03-66/2024-03/200117).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.