Increasing Valley Retention as an Element of Water Management: The Opinion of Residents of Southeastern Poland

Abstract

This study presents the results of an analysis of public perceptions of flood safety and river valley management in southeastern Poland. The aim of the study was to identify sociodemographic and spatial factors influencing preferences for two distinct river valley management models: the traditional, technical model (a strategy to move water away from people, MWAfP), and the ecosystem-based model (leaving space for the river, LSfR). A diagnostic survey was employed using a custom-designed questionnaire completed by 563 respondents residing in southeastern Poland. The research tool enabled the identification of flood risk perceptions and attitudes toward retention and flood control solutions. The collected data were analyzed using descriptive statistics, and exploratory analysis was conducted to identify clusters of respondents and to test for differences between groups. Correlation analysis between items was performed, and a model of determinants of river valley management strategy selection was calculated using logistic regression. The results enabled the identification of three dominant perception clusters, reflecting diverse approaches to hydrological safety and environmental adaptation. The calculated logistic regression model includes a number of factors, among which significant determinants of the LSfR strategy selection include level of education, belief in the need to slow water runoff from the catchment, and support for the cultivation of permanent meadows in floodplains. The applied methodological approach allows for a comprehensive assessment of the social determinants of flood risk perception and supports the development of adaptive water management strategies in flood-prone areas.

1. Introduction

Water is a key resource determining the functioning of societies and economies. Its availability—both quantitatively and qualitatively—is a fundamental condition for civilizational development. At the same time, water can also be a threat factor: excess leads to floods and material losses while deficiency leads to droughts, the consequences of which affect agriculture, the energy sector, and the quality of life of residents [1,2]. In this sense, water is becoming a strategic resource, and its management is a matter of national and regional security [3,4,5,6].

Water-related extreme events are intensifying as a result of climate change. Forecasts indicate that Poland, especially its southeastern part, will experience increasingly frequent floods and prolonged droughts [7,8]. This region is particularly vulnerable due to its geographical conditions: the presence of the Carpathians and the Foothills, with highly variable rainfall, numerous watercourses with dynamic flows, and intensive agricultural use of valley areas.

In this situation, the concept of valley retention, understood as the ability of river valleys to retain water in their channels, coastal zones, and floodplains, becomes particularly important. Valley retention, alongside landscape and reservoir retention, is indicated in the literature as one of the most effective tools for mitigating the effects of both floods and droughts [9,10]. The role of valley retention is not limited to hydrological functions. River valleys also perform ecological, biogeochemical, and social functions [7,11].

The problem of water retention in Poland also has a historical and institutional dimension. Water management in the 20th century was subordinated to the idea of land improvement, which focused on the rapid drainage of water from agricultural fields. As a result, a system was created that accelerated runoff and reduced natural retention. Currently, attempts are being made to reverse this trend such as, for example, through river renaturalization and adaptation projects, such as the “Drought Effects Countermeasures Plan” [12]. Despite these efforts, elements of the old approach still dominate in our society [13,14]. An important aspect of the discussion on water management is the social perception of risk. Residents of regions prone to floods and droughts often base their expectations of the authorities on previous experiences and stereotypes. Studies [8,15,16] indicate that the view that flood safety is primarily provided by embankments and river regulations is still popular in Poland. Meanwhile, contemporary hydrological and ecological approaches emphasize the need to restore space to rivers and use valleys as natural water reservoirs [17,18,19,20].

Table 1. Theories and concepts regarding water retention and river valley development strategies.

Theory/Concept	Brief Description of the Theory/Concept, Along with Literature References
Water retention as an element of security, and adaptation to climate change
Risk theory	Analysis of threats, and ways to minimize socio-economic losses; justifies the need for retention in the context of floods and droughts [21,22].
Resilience theory	The ability of systems to adapt and recover from disturbances; retention increases socio-ecological resilience [23,24].
Environmental economics	Analysis of the economic value of ecosystem services and external costs of disasters; retention as an investment in security [25,26].
Climate adaptation theory	Strategies for adaptation to the effects of climate change; water retention as an adaptation tool [27,28].
Ecosystem services theory	Ecosystems provide benefits (retention, filtration, biomass production); they justify the protection of river valleys [29,30].
Circular economy	Efficient use of natural resources and processes; use of flood meadows for fodder and energy purposes [31,32].
Perception of safety and river valley development strategies
Risk society theory	Modern societies create risks themselves and have to manage them; a change in approach from embankments to polders [33,34].
Adaptive management	Flexible water and landscape management in changing climatic conditions [35,36].
Integrated Water Resources Management (IWRM)	An integrated approach combining environmental, economic, and social aspects in water management [37,38].
Path dependency	Consolidation of old hydrotechnical decisions (e.g., land improvement) in current water management practices [39,40].
Institutional theory	The influence of norms, institutions, and law on the consolidation of traditional solutions in water management [14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42].
River restoration	Restoring natural river processes as an alternative to regulating and straightening rivers [43,44].
Risk perception theory	The public perception of risk differs from the expert perception; it is important for the acceptance of retention measures [17,45].
Framing theory	The way a problem is presented influences its social perception and political decisions [46,47].
Social capital theory	The importance of trust, and cooperation, for the implementation of new solutions in water management [48,49].

presents selected theories and research trends concerning water retention as security and adaptation to climate change, as well as the perception of security and river valley development strategies.

It should be emphasized that resilience theories are central to the analysis of how valley systems can survive disturbances (e.g., floods, droughts) [24,36,50,51]. Risk perception theories, in turn, help to understand how residents perceive hydrological hazards (e.g., floods, droughts) and what factors influence these perceptions [17,52,53,54,55,56]. The concept of social capital, on the other hand, can explain the extent, to which residents will be willing to accept innovative retention solutions or share responsibility for water management [47,49,57,58].

In the context of theories and trends promoting water management, institutions, legal regulations, and history (e.g., past land improvement) influence current approaches to this issue [25,29,59,60]. Other works, in turn, are related to restoring river functions and their retention and analyze how to design interventions that are resistant to climate change [24,51,59]. In this context, the theory of ecosystem services (circular economy) analyzes multiple benefits in this regard (e.g., retention, filtration, crops, energy) related to valley systems [1,7,19,60].

In this study, we focused on highlighting respondents’ expectations regarding river valley development and their perception of flood safety. We attempted to identify characteristics grouping respondents according to their preferences for water retention management in the context of sociodemographic characteristics. Finally, we aimed to identify, using a logistic regression model, the determinants of river valley development choices consistent with the strategy of leaving space for rivers.

2. Literature Review

2.1. Water Retention as an Element of Security and Adaptation to Climate Change

Hydrological security is one of the fundamental dimensions of environmental and economic security. Water shortages lead to losses in the agricultural, energy, and municipal sectors, while excess water leads to catastrophic floods. As elsewhere in the world, in Poland, both phenomena occur simultaneously. Intense floods particularly affect mountainous and foothill regions, while droughts regularly affect agricultural areas [7,61]. The economic and social consequences of both phenomena are enormous and mitigating them is one of the main tasks of contemporary water management [17,29,45].

The concept of risk provides a fundamental framework for the analysis of natural hazards. Beck [33], in his “risk society” theory, argues that modern societies generate risks themselves and then learn to control them. Floods and droughts, exacerbated by climate change, fit into this paradigm as structural threats requiring appropriate management strategies. On the other hand, the resilience theory of socio-ecological systems emphasizes the ability of systems to adapt to disturbances and maintain their basic functions despite shocks [23,62]. In the context of water management, this means building systems resistant to extreme events by increasing retention, diversifying water sources, or developing green infrastructure [63].

Floods also generate direct material losses (e.g., damaged infrastructure, houses, crops) as well as indirect costs, such as decreased production, job losses, and health costs. Environmental economics studies indicate that these costs often exceed expenditures on preventive measures such as the construction or modernization of retention systems [1,64]. Drought-related losses are equally serious. In Poland, according to data [14,65], agricultural droughts occur on average every 2–3 years, leading to losses counted in billions of zlotys. Water shortages during the growing season result in reduced grain, potato, and corn yields, and on a national scale translate into increased food prices and problems in the energy sector, which is dependent on water cooling [66].

Hydrological literature [67,68] currently emphasizes that one of the most effective ways to reduce the risk of floods and droughts is to increase retention at the landscape and river valley scale [9]. Valley retention acts as a buffer because it captures excess water during floods and makes its resources available during periods of shortage. Studies conducted in Carpathian river basins have demonstrated that retention systems can significantly flatten flood waves and stabilize flows [7]. The importance of retention is also evidenced by reports [69] which indicate that Poland retains only about 15% of the average annual runoff, which is one of the lowest rates in the European Union. Insufficient retention weakens the country’s resilience to droughts and floods, which translates into increased economic and social risk [70].

The southeastern region of Poland is particularly vulnerable to hydrological hazards. It experiences heavy rainfall, leading to river floods as well as periodic water shortages during the summer season. The floods of 1997 and 2010 highlighted the scale of the threat, with losses estimated in billions of zlotys and long-lasting social consequences [71,72]. Research [16] indicates that communities in southern Poland demonstrate a relatively high level of mobilization in the face of floods, yet expectations regarding outdated forms of protection (embankments, river regulation) still prevail. Meanwhile, the hydrological and ecological literature increasingly emphasizes the need to implement nature-based solutions (NBS) that increase water retention and the regions’ resilience to climate change [73,74].

In the literature on climate change [61,63,75], two approaches are distinguished: mitigation (limiting the causes of change) and adaptation (adapting to the effects of change). In the case of water management, the emphasis is primarily on adaptation because phenomena such as the increased frequency of droughts and floods are already observed and require practical responses [1]. Water retention, both in the scale of large reservoirs and in the form of small-scale landscape retention, is one of the key adaptation tools [63,76]. It has been known for many years that the operation of large dams is controversial. Dams, one of the purposes of which is flood protection, create a false sense of security in society. As a result, floodplains become built up, which intensifies flood losses during the next, inevitable flood. Consequently, large dams paradoxically contribute to increased flood losses [14,16,52,77,78].

Valley retention, on the other hand, involves utilizing the natural capacity of river valleys to retain water through floodplains, polders, oxbow lakes, and riverside meadows. Small-scale retention, on the other hand, involves dispersed activities such as the construction of small reservoirs, the restoration of wetlands, or the creation of ponds in agricultural landscapes [9].

Polish literature indicates that small retention is often more effective than large reservoirs because it operates locally, reducing the risk of both floods, and droughts [14]. It turns out that the concept of ecosystem services provides a theoretical framework for assessing the value of water retention. River valleys provide a number of services that can be divided into [1] (i) regulatory: slowing down runoff, water filtration, and nutrient retention; (ii) provisioning: providing crops from floodplain meadows and biomass for energy; (iii) cultural: demonstrating landscape and recreational values; and (iv) sustaining: maintaining biodiversity and biogeochemical processes.

Alluvial processes are particularly important in agriculture. The deposition of fertile sediments improves soil quality, and floodplain meadows and can be used to produce feed and energy biomass [79,80]. Such solutions also align with the idea of a circular economy, in which resources are reused and recycled [75]. In Poland, the topic of retention as an adaptation tool has been widely discussed in the public debate in the context of the “STOP DROUGHT” project, implemented by the State Water Management Company “Wody Polskie-Polish Waters.” Although the project aimed to mitigate the effects of drought, it was criticized for its excessive focus on the construction of large reservoirs and the continuation of a technical approach. Non-governmental organizations, including some scientific publications [78], have emphasized the need to promote landscape retention, encompassing small reservoirs, wetlands, and dispersed activities. This discussion demonstrates that adaptation to climate change requires a paradigm shift from centralized, large-scale investments towards more sustainable, local, and ecosystem-based solutions [63].

The southeastern region of Poland is particularly vulnerable to hydrological variability. Both mountain floods and summer water deficits occur here. Valley retention is particularly important in this region, as the Carpathian valleys serve both agricultural and ecological functions. The literature indicates that the use of floodplain meadows for the production of feed and energy biomass can be an element of adaptation, combining economic and environmental aspects [7,76].

2.2. Perception of Safety and River Valley Development Strategies

2.2.1. River Valley Development Models

The literature on the discussed subject [16,52,53,61,70,73] most often identifies two basic models of river valley management. The first can be described as technical, aiming to divert water away from people. This strategy is based on the construction of flood embankments, regulation of river channels, and deepening and straightening rivers. The logic of this strategy assumes subordinating the environment to human needs, and ensuring human safety, through maximum control over the flow of storm water [9,81]. The second model is the ecosystem approach, which involves leaving space for rivers. In this approach, river valleys are perceived as natural floodplains that can act as a safety buffer. Such solutions include the creation of polders, restoration of floodplains, and the restoration of degraded sections of river channels [24,51]. This model is increasingly recommended within the EU water policy [82] and the concept of nature-based solutions [73]. This issue will be discussed in more detail later in the article.

River valley management fits well with Beck’s concept of risk society [33,34]. In his view, modern societies generate risks—among other things, through interference with ecosystems—and must learn to manage them. Although the construction of embankments and river regulation were intended to minimize flood risk, in practice they led to the displacement of hazards and increased vulnerability of systems to water failures. Other contemporary literature [59,63] points out that technical solutions cannot completely eliminate risk, they can only shift and transform it. Therefore, approaches that integrate flood protection with the preservation of ecosystem functions are increasingly promoted [60].

The concept of adaptive management emphasizes the need to flexibly adapt water policies to changing environmental and social conditions [36]. In practice, this means combining technical elements (e.g., infrastructure protection) with ecological solutions (e.g., maintaining floodplains or wetlands). Hydrological studies [10,19] show that polders and restored floodplains can effectively reduce flood waves while simultaneously increasing landscape retention. This approach is more sustainable because it allows the water system to perform multiple functions simultaneously, from flood protection to supporting biodiversity [24,83,84].

River valley development strategies should also be considered within the broader framework of Integrated Water Resources Management (IWRM). This approach, promoted by the UN (United Nations) and the European Union, assumes that water management should balance social, environmental, and economic goals [85]. In the Polish context, this means the need to reconcile flood safety, the protection of valley ecosystems, and economic interests related to agriculture, and energy [59].

Meanwhile, a technical approach, focusing on embankments and river regulation, still dominates in Poland. However, state strategic documents, such as those indicated by the Senate of the Republic of Poland [3], point to the need to increase valley retention and more efficient use of floodplains. At the same time, non-governmental organizations, including some scientific publications [78], emphasize that effective adaptation to climate change requires a shift away from large reservoirs and a focus on dispersed, natural forms of retention. In southeastern Poland, where mountain floods are particularly dangerous, such an approach has significant practical importance. Floodplains, such as those in the San Valley [76], can function as polders, and the renaturalization of river valleys can contribute not only to flood risk reduction but also to the preservation of the region’s natural values [26].

2.2.2. Social Expectations Regarding the Development of River Valleys and the Perception of Flood Safety

The public perception of rivers is complex and depends on many factors. Local communities are sensitive to the hazards associated with periodic floods. However, there is a noticeable lack of knowledge about what constitutes a high-quality river [86]. Public expectations primarily concern increasing river embankments [17], while the level of knowledge regarding non-technical flood protection measures is low [15].

It should be emphasized that risk perception is one of the key factors determining the effectiveness of water management activities [21,22,33,34]. Research in the psychology of risk shows that subjective perception of threats often differs from expert assessments [56]. According to [45], people tend to overestimate spectacular risks (e.g., floods) and underestimate chronic risks (e.g., droughts). In the context of water management, this means that local communities may prefer investments, such as flood embankments, even if ecosystem-based solutions are more effective in the long run [17,54]. In turn, the framing theory [87] emphasizes that the way a problem is presented influences its social perception. The media, and politicians, using a specific language can reinforce technical patterns of flood protection, emphasizing the need to build embankments or regulate rivers. In Poland, as research [15] indicates, society still largely equates hydrological security with technical investments. Cleaning drainage ditches or strengthening embankments are perceived as basic protective measures, even though contemporary hydrological literature indicates the need for wider use of floodplains and renaturalization activities [63,88].

A significant factor determining the effectiveness of social adaptation to water hazards is the concept of social capital. It refers to the level of trust, cooperation networks, and social norms that enable people to act together [48,49]. Lechowska’s [16] research, conducted in southern Poland, showed that communities characterized by high levels of social capital responded more effectively to floods by organizing neighborly assistance, mobilizing local resources, and cooperating better with public institutions. In this sense, social capital not only strengthens resilience to disasters, but also increases acceptance for new forms of valley development, such as polders or floodplain meadows [57,58]. Where social ties are weaker, the expectation of simple, technical solutions implemented by the state prevails [47]. Furthermore, public opinion analyses provide additional evidence of the importance of the social perception of water risk. Research [15] showed that the majority of Poles had flood victims in their surroundings, and fear of further hydrological events remains high in flood-prone areas [14,15]. At the same time, a significant percentage of respondents believe that the responsibility for flood protection rests with the state, and institutions, not with local communities [16,17]. Other studies [59,60] indicate that people with higher education are more likely to believe that floods are the result of both natural processes and climate change [89]. This shows that education, and environmental awareness, can influence the perception of threats and increase openness to nature-based solutions [73].

In this context, it is important to understand public expectations regarding river valley development models. Public perception of this issue can vary greatly. Some people believe in the need for technical development of rivers, others favor maintaining the river in a state close to its natural state, and still others only expect local authorities to ensure safety [15,16,90,91]. At the same time, understanding public expectations regarding river valley development may be crucial in developing intelligent and innovative safety systems in accordance with the NBS principle [20].

Most available studies, both in Europe and elsewhere, focus only on selected aspects of management such as perceptions of flood embankments [92], acceptance of individual flood policy instruments [93], or opinions on natural flood management and floodplain restoration measures [94,95]. These approaches are fragmented and primarily assess a single protection tool without reference to alternative systemic strategies. Furthermore, previous studies have rarely considered complex sociodemographic factors as variables explaining strategy choice, treating them rather as descriptive data.

Therefore, when analyzing the strategy of “moving water away from people” one can encounter, among others, studies analyzing a fragmented part of this strategy.

The authors [92] conducted a survey among 828 residents and users of coastal and riverside areas in France, presenting them with five scenarios for the development of dikes, from maintaining their current state, through strengthening, through “opening/lowering” the dikes to more “natural” variants (vegetation, greater integration with the surroundings). The results indicate that the respondent’s relationship with nature (understood as anthropized/civilized) significantly influences the respondents’ choices. The authors of this publication emphasize that communication and public participation will be crucial in the evolution of dike strategies in the context of climate change. Another study [93] involved 2650 residents of high-risk flood areas in Canada. The survey covered, among others, the following policy instruments: flood hazard maps, post-event flood assistance, flood insurance, flood risk disclosure and liability, and property buyouts. The authors argue that even with classic structures (embankments, dams), the importance of social/behavioral instruments and social acceptance is growing. In turn, the study [94] showed that higher levels of knowledge and awareness of respondents positively influenced risk perception and willingness to support protective measures. The authors of this publication emphasize that in the context of protecting people and infrastructure, education and communication are crucial, as they increase acceptance of measures that “move water away from people.” Another study [96] was conducted in the Layyah region (Pakistan) after the 2010 flood. Approximately 200 surveys were collected in four local units (union councils). The study explored the flood risk management and resilience-building practices used by local authorities, organizations, and communities. The conclusions indicate the need to better combine technical measures with local practices and knowledge to “move water away from people” in a sustainable manner. Furthermore, the study conducted in Shenzhen, China [97] found that the importance of integrating technical measures with local practices and knowledge is crucial to “move water away from people.” In 2020, 339 respondents were surveyed in five districts prone to urban flooding. The results indicate that emotions and risk perception strongly influence the willingness to adopt measures that “move water away from buildings/people.” The study highlights the need to consider psychological aspects in planning protective infrastructure.

When analyzing the strategy of “free floodwater spreading,” one can, among others, come across some interesting studies that analyze some parameters of this phenomenon without a comprehensive approach.

For example, an experiment was conducted in Denmark that is important for implementing strategies to “accept water” rather than simply withdraw it. The results showed that farmers are willing to participate in such strategies, but the conditions must be clear (inundation periods, compensation amounts, impact on crops) [98]. The article [99] examines the concept of storing floodwater on agricultural land (“flood storage option”) and compares it to experiences in drought management. The author presents a survey and discussion with farmers and land managers in the UK about preparing such land to function as a flood buffer and delay runoff. He also points out that while the “give water space” approach is promising, it requires a clear legal framework, compensation, and communication with farmers. Other studies from the UK [100], based on surveys and interviews with farmers, landowners, and water authorities, show that respondents appreciated the “natural flood management” (NFM) strategy but identified barriers to this model: income uncertainty, lack of clarity in the compensation mechanism, and the risk of reduced agricultural activity. Yet another study from the UK [101] shows that residents more often see environmental and landscape benefits rather than purely technical ones, while practitioners focus on hydraulic functionality. Both perspectives must be taken into account to effectively design and implement “return water to nature” strategies. In an online study also conducted in the UK, the authors [95] suggest that communicating its socio-ecological, not just hydrological, benefits is crucial for the success of a free-flowing strategy.

The southeastern region of Poland, due to its history of major floods in 1997 and 2010, is particularly sensitive to hydrological safety issues. At the same time, as research conducted in this publication [102] indicates, stereotypes persist in flood-affected communities regarding the need to strengthen embankments and regulate rivers. Implementing innovative retention solutions, therefore, requires not only institutional changes but also educational activities and building social trust [61,63].

2.3. Water Management in Poland

Polish water management was largely shaped in the 20th century, when the dominant paradigm was land improvement [16,17]. During the centrally planned economy, land improvement was primarily understood as a system for quickly removing water from arable fields, which was intended to increase agricultural production, and protect against excessive soil moisture [65,103,104,105]. This approach led to straightening riverbeds, deepening drainage ditches, and sealing drainage [70]. This phenomenon can be explained in terms of path dependency, which is dependence on the development path [25,37]. Once adopted, the water management model became entrenched in administrative institutions and practices, which hinders the introduction of alternative solutions [15]. Institutions responsible for water management focused on production goals, omitting the environmental and retention functions of ecosystems. Drainage projects have resulted in a number of negative consequences [15,16]: (i) accelerated water runoff: shortening the retention time in the landscape, which exacerbates droughts; (ii) loss of natural floodplains: reducing the capacity of rivers to intercept flood waves; (iii) ecosystem degradation: loss of wetlands and riverside meadows, decreasing biodiversity; and (iv) deterioration of water quality: rapid runoff of pollutants into major streams. Furthermore, the construction of wetlands and improved connectivity to floodplains also reduce flood peaks. Floodplain restoration efforts are important to create space for floodwater and reduce exposure to them by relocating people from the hazard zone. Floodplain restoration also provides access to the river, which brings many benefits, including recreation, access to water for domestic use, and other cultural ecosystem services. A key adaptation strategy is to reduce riverbank erosion (due to high peak flow) by using riparian vegetation to stabilize riverbanks during floods [28,30,104,105].

Studies show that retention in Poland is exceptionally low; only about 15% of the average annual runoff is retained [13,65]. This low retention level is a direct consequence of the historical water management model, which promoted drainage rather than retention. However, in recent years, the need for a paradigm shift in water management has been increasingly emphasized. Polish literature [9,103] indicates that land improvement should be understood more broadly, not only as a drainage system, but also as a tool for regulating water conditions and enabling water retention in the landscape. Some authors [13] reveal, however, that many measures implemented within water policy still replicate old patterns, focusing on water drainage. The “STOP DROUGHT” project has also been criticized for its continued technical approach and insufficient consideration of landscape retention [14,30]. However, amid criticism of land improvement, the concept of river renaturalization is gaining increasing popularity. In practice, this means restoring natural river processes: restoring meanders, oxbow lakes, floodplains, and wetlands [79,105]. Restoration is consistent with the ecosystem approach and allows for the improvement of both flood safety and water balance at the catchment scale [24,106,107]. European examples, such as actions on the Rhine in Germany or the Ebro in Spain, show that abandoning excessive regulation and restoring river space brings long-term hydrological and ecological benefits [108]. In Poland, this process is only just beginning, but the first restoration initiatives are already visible, for example, in river valley protection programs in the Podkarpacie region [79].

Polish water management faces a significant institutional challenge. The legacy of land improvement and river regulation is consistent with the “path dependency” logic and hinders a change in approach [25,37]. Although new strategies and programs are emerging, such as the “Retention Development Program,” elements of the old technical paradigm still dominate. Attempts to change the direction, for example, through river renaturalization demonstrate that it is possible to implement solutions that are more consistent with the principles of sustainable development [78,109]. The key factor determining the success of new strategies remains the public perception of risk and safety, and acceptance for nature-based solutions is limited [15,16,65,110].

In summary, it should be stated that valley retention is a key element of contemporary water management, combining hydrological, ecological, and social functions. In the context of increasing climate change leading to more frequent floods and droughts, retention is becoming, not only a technical solution, but, above all, a tool for adaptation and increasing the resilience of social and ecological systems [1,23,28,104]. Literature analysis indicates the existence of two competing models of river valley management. The traditional, technical model is based on the construction of embankments and channel regulation, focusing on diverting water away from people. An alternative is the ecosystem approach, based on restoring space to rivers through polders and floodplains [9,24,103,108]. The second model is consistent with the concept of adaptive management and integrated water resources management and considers the river and valley as a dynamic system whose functioning should be supported rather than completely controlled [28,111,112]. The literature also indicates that effective water management in Poland requires:

(a)

Increasing valley and landscape retention as a fundamental adaptation tool [28,29,30];

(b)

Paradigm shift: moving away from land improvement towards solutions that support natural river processes [113,114];

(c)

Integrating approaches: combining technical and ecosystem-based solutions within IWRM (Integrated Water Resources Management) [85]; and

(d)

Building public awareness and social capital, which are essential for implementing new strategies [65].

Valley retention is therefore not only a technical tool but also a bridge between hydrological security, adaptation to climate change, and social acceptance. In the context of southeastern Poland, which is particularly vulnerable to extreme phenomena, its role cannot be overestimated. Therefore, the aim of this study is to understand the expectations of residents of southeastern Poland, regarding river valley management strategies, in the context of ensuring flood safety.

The uniqueness of our study, compared to other survey studies [92,93,94,95,96,97,98,99,100,101], on approaches to flooding and its effects stems from several issues:

• Our study simultaneously compares two water management models (MWAfP and LSfR) and such a systems approach is rare in the literature.
• Our research covers the broad context of river valley development (social and natural functions, retention and flood safety, and spatial planning), whereas many studies focus on a single aspect, either the acceptance of technical solutions or the effects of water spillage into nearby areas.
• In our study, sociodemographic factors are treated as an explanatory variable, not merely a descriptive one, as is the case in other similar studies.
• Our research questions also encompass perceptions of flood effects and attitudes toward retention, making our questionnaire more comprehensive than in other articles.

3. Materials and Methods

3.1. Purpose and Scope of the Research

The main goal of the study was to identify public perceptions of river valley development strategies, with particular emphasis on retention and flood control solutions. The analysis included the role of sociodemographic factors influencing the perception of two distinct water management models: the technical model (“moving water away from people”—MWAfP) and the ecosystem model (“leaving space for the river”—LSfR). Additionally, the study aimed to identify key predictors of strategy selection and assess the internal consistency of perceptions of flood impacts and attitudes toward water retention [115,116]. Based on this, the following research questions were formulated:

• What are respondents’ expectations, regarding river valley management, in the context of ensuring flood safety?
• Are there sociodemographic patterns associated with the perception of river valley management?
• What factors shape the perception of flood safety strategies among residents of southeastern Poland?

Answers to these questions allowed for the formulation of recommendations regarding actions directing the development of river valleys towards the LSfR strategy.

The study was conducted in southeastern Poland (Figure 1), particularly in the Podkarpackie and Lublin Voivodeships, regions characterized by high flood risk, high hydrological variability, and strong links between local communities and agriculture. This area experienced severe floods in 1997 and 2010, making it suitable for analysis of flood risk perception and hydrological safety [117].

At the conceptualization stage, the research was exploratory in nature, focusing primarily on the perception of river valley management as ensuring flood safety. Our intention was to determine whether respondents associate flood control measures with increased water retention, including valley retention, and what factors determine the choice of LSfR strategies.

The research was conducted in 2024. A diagnostic survey method was used, utilizing the CAWI (Computer Assisted Web Interview) questionnaire. Despite some limitations associated with this method, the online tool was chosen for several practical reasons. First, 95.9% of households in Poland had internet access in 2024 [118]. Second, the chosen method was quick and inexpensive.

The research questionnaire included a set of items, regarding:

(a)

sociodemographic characteristics of respondents,

(b)

assessment of river valley development methods,

(c)

assessment of the subsequent impact of floods on crop yields,

(d)

preferred flood safety strategy.

A link to the profitest.pl portal, where the research tool was hosted, was provided to university students and collaborating businesses, asking them to complete the survey and forward the link to others. Students received guidelines regarding the number of surveys to be completed in separate groups based on gender and age. For those experiencing difficulties with digital technology, the interviewers assisted in completing the form. This procedure ensured that the demographic structure of respondents closely resembled the actual structure in the study area [119]. The survey procedure ensured complete anonymity of respondents. The study was voluntary, and respondents could discontinue completing the survey at any time and delete their responses. The reliability of the obtained results was also enhanced by a mechanism for monitoring the time taken to complete the survey. This allowed for the removal from the database of records completed in a shorter time than previously determined empirically during the trial phase.

The population of the study area numbered approximately 4,200,000 people. The sample size was determined using Cochran’s formula [120], assuming a 95% confidence level (α = 0.05), a maximum error of e = 0.05, and p = 0.5, which resulted in the required sample size of n ≈ 384. However, it should be emphasized that in practice, the probability requirement for inclusion in the sample was not met, and, therefore, the research was not probabilistic; therefore, the obtained results were treated as indicative. A total of 563 correctly completed survey forms were qualified for analysis.

Figure 1 shows the study area. The left part of the figure shows the outline of Poland with major cities marked, while the right part of the figure shows a magnified view of the study area. Major rivers are marked, and the flood risk on the date the map was accessed is indicated in orange.

3.2. Research Tools and Research Procedure

Table 2. Tested items, their coding, and abbreviations.

Abbreviations for Item Names	Item Names	Measurement Scales
Sociodemographic data
Ag	Age	Quantitative variable
Sx	Sex	0 = Female; 1 = Male
Ed	Education	0 = Elementary; 1 = Vocational; 2 = Secondary; 3 = Higher
PoR	Place of residence	0 = Village; 1 = City up to 100,000; 2 = City over 100,000
PS	Professional status	0 = Employed in agriculture; 1 = Employed outside agriculture; 2 = Others (students, retirees, unemployed)
HC	Has children	0 = No; 1 = Yes
DR	Distance to the river	Quantitative variable
OFH	Occurrence of flood hazard	0 = No.; 1 = Yes.
Assessment of river valley development methods
NAFSP	Negative assessment of flood safety in Poland	1 = definitely safe; 2 = rather safe; 3 = neither yes nor no; 4 = rather dangerous; 5 = definitely dangerous
RRR	Rivers require regulation	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
SDFW	Slow down the flow of water	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
EBA	Embankment of built-up areas	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
EAL	Embankment of agricultural land	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
BLD	It is necessary to build large dams on the main rivers	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
ARFU	Allowing the river to flood undeveloped areas	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
LFEAA	Limiting flood embankments in agricultural areas	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
CFP_DRR	Creation of flood polders and construction of dry retention reservoirs	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
FEM	Flooded areas should have meadows that increase water retention and provide fodder or energy biomass.	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
DFEP	Development in flood areas should be prohibited	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
Assessment of the subsequent impact of floods on crop yields
FACIC	Floods on sown area cause increased yields in subsequent years	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
FACRC	Floods in sown areas cause reduced yields in subsequent years	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
FMCIC	Floods in permanent meadows increased yields in subsequent years	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
FMCRC	Floods in permanent meadows reduced crop yields in subsequent years	1 = definitely not; 2 = rather not; 3 = neither yes nor no; 4 = rather yes; 5 = definitely yes
Preferred flood safety strategy
EFPM	Effective flood protection measures	0 = Moving water away from people (MWAP); 1 = Leaving space for rivers (LSfR)

presents the names of individual items and their abbreviations, used throughout the study, and the measurement scales. Some items were quantitative in nature; items related to perceptions of the studied phenomena were measured on a bipolar, five-point scale with a neutral value in the middle, and the item related to preferences for flood safety strategies was a dichotomous variable.

The study utilized a proprietary survey form. A series of items related to the assessment of river valley management methods were contrasted with an EFPM item referring to the preference for one of two strategies for this management. This allowed us to determine the consistency of respondents’ views and identify sociodemographic patterns in approaches to ensuring flood safety. In the context of the need to restore ecosystems and their services related to water retention [111], we also introduced items related to the subsequent impact of floods on crop yields. We assumed that respondents who perceived a positive impact of siltation on the yields of permanent meadows would have a positive attitude toward LSfR. Simultaneously, the dichotomous nature of the EFPM variables enabled us to construct a logistic regression model of the determinants of river valley management model selection.

The study was cross-sectional and considered only selected elements of perception of the economy in river valleys. Due to limitations of the research tool, we did not consider the respondents’ level of hydrotechnical knowledge. The “education” variable was limited to formal education level, which provides general, but somewhat reliable, information about the respondent’s intellectual level.

3.3. Data Analysis Methods

Data were analyzed, using IBM SPSS Statistics (version 29.0.2.0) and the Statistica package (version 13.3), with a significance level of p < 0.05. The following analytical techniques were used:

(a)

Descriptive statistics—determining the sample structure and measures of central tendency [121].

(b)

Mann–Whitney U test—for dichotomous variables (e.g., sex, having children, flood experience) [122].

(c)

Kruskal–Wallis H test—for ordinal variables (e.g., education, place of residence, professional status), with post hoc testing and pairwise comparisons, using the Mann–Whitney U test [123].

(d)

K-means cluster analysis—identifying types of respondents by strategy perception [124].

(e)

Spearman’s correlations—assessing the consistency of flood perceptions and retention strategies [125].

(f)

Logistic regression—identifying factors, determining the choice of the LSfR strategy, to build a model of determinants of flood safety strategy selection [124,126,127].

The results of the analyses were interpreted as indicators of the direction of attitudes and differences between social groups. Mean values for ordinal variables (e.g., education, place of residence) can be presented as indicative trends, but their interpretation should be cautious and considered complementary to the results of significance tests.

The methods employed allowed us to identify three main patterns of perception of flood management strategies related to education level, place of residence, and flood experience. The use of statistical methods such as cluster analysis, correlation, and logistic regression enabled a thorough understanding of the determinants of retention strategies in the context of hydrological security and climate change adaptation.

3.4. Research Limitations

The sampling used was non-probabilistic, which limits the generalizability of the results. Respondents’ responses may have been biased by social desirability bias or limited hydrological knowledge. Furthermore, uneven representation of certain groups (e.g., farmers, floodplain residents), may have influenced the data distribution [126].

4. Results

4.1. Sociodemographic Characteristics of the Study Group

Table 3. Basic statistics regarding quantitative variables: age and distance of residence from the nearest river.

Statistics	Ag 1	DR 1 [km]
Minimum	18	0.001
Maximum	85	92
Range	67	91.999
Mean (M)	44.23	5.75
Median (Me)	47	3
Modal (Mo)	21	1
Modal frequency	48	75
Standard deviation (SD)	18.61	8.549
Quartile 25 (Q1)	23	1
Quartile 75 (Q3)	61	6

¹—The names of the items are in Table 2.

presents basic descriptive statistics for quantitative variables: age and distance from the nearest river.

Table 4. Structure of sociodemographic variables characterizing the respondents.

Variables 1	Participation in the Structure	Variables 1	Participation in the Structure
Sx		PS
women	54.88%	0 = working in agriculture	5.33%
men	45.12%	1 = working outside agriculture	48.31%
HC		2 = others (students, retirees, unemployed)	46.36%
0 = No	39.61%	PoR
1 = Yes	60.39%	0 = village	49.73%
Ed		1 = city up to 100 thousand	30.91%
0 = primary	5.33%	2 = city over 100 thousand	19.36%
1 = vocational	23.80%	OFH
2 = secondary	41.74%	0 = No	65.19%
3 = higher	29.13%	1 = Yes	34.81%

¹—The names of the items are in Table 2.

presents the structure of sociodemographic characteristics of the respondents.

The average age of respondents was 44 years, and the median age was 47 years. The most frequent value was Mo = 21, and the mode number was 48. In general, the age structure of respondents was similar to the actual age structure of people living in the study area [119]. The proportion of women, among respondents, was higher than that of men, with 121.7 women per 100 men (Table 4). In the population of the study area, this was approximately 105 women per 100 men [119]. The majority of the study group had children. Among the respondents, the largest group were people with secondary education (41.74%). Regarding professional status, 48.31% of respondents worked outside agriculture, while a small percentage of respondents (5.33%) worked in agriculture. It should be emphasized, however, that in Polish society this percentage was 8.1% in 2023 and has a decreasing tendency [119]. However, the majority of respondents, 49.73%, lived in rural areas. It was assumed that this may be important in the context of the perception of the subsequent impact of flooding on agricultural areas. Rural residents, even if they do not work in agriculture, have closer contact with nature, including riverside spaces, than city dwellers. The median distance from the place of residence to the nearest river was Me = 3 km, the mean was M = 5.749 km, and the standard deviation was SD = 8.549 km. These distances allow for the formation of perceptions of river functioning based on observations of the immediate surroundings.

4.2. Respondents’ Expectations Regarding the Development of River Valleys and Sociodemographic Patterns in the Perception of Ensuring Flood Safety

The core of the study included a series of items related to perceptions of river valley spatial development and the subsequent impact of floods on floodplains. Individual items were assessed using a bipolar, five-point scale with a neutral value. The results of the item rating structure are presented in Figure 2. The overall assessment of flood protection in Poland was negative, with 63.4% of respondents stating that these measures were insufficient.

The statements evaluated by respondents contained theses, representing two different approaches to creating flood protection. The outdated, repeatedly negatively reviewed strategy of moving water away from people (MWAfP) included items RRR, EAL, and BLD (see Table 2). The strategy of leaving space for rivers (LSfR) included items SDFW, ARFU, LFEAA, CFP_DRR, FEM, and DFEP (see Table 2). The item EBA, referring to embanking built-up areas, is an action that finds justification in both alternative strategies and received the highest percentage of positive ratings (84%). The data presented in Figure 2 indicates that respondents positively assessed the actions characterizing both opposing strategies. The negative assessment of the proposal to eliminate embankments on agricultural land (LFEAA) deserves particular attention. In this case, only 21.3% of respondents rated this item positively. And the possibility of floodwaters spilling over agricultural areas may be an element of reducing flood losses.

The four items on the right side of Figure 2 addressed perceptions of the flood’s subsequent impact on sown areas and permanent meadows. In an analysis of all results, 41.7% of respondents denied an increase in sown area yields, while 41.6% confirmed a decrease in these crops. However, different results were obtained for permanent meadows. In this case, 46.2% of respondents confirmed higher green mass yields in the years following the flood, while 42.5% denied a decrease in these yields. The large percentage of neutral responses is noteworthy, which, given the highly emotional nature of flooding, may indicate difficulty for respondents in defining a clear view.

A k-means cluster analysis was conducted on the collected data regarding ratings of individual items related to flood protection and the subsequent impact of floods on floodplains [121]. A preliminary elbow analysis indicated the selection of three clusters. The identified clusters differed in their approach to the studied features. The mean ratings for individual items, broken down by cluster, are presented in Figure 3.

Cluster 0 was characterized by a predominance of women (110 vs. 76 men). Less educated individuals predominated. This cluster reported low ratings for elements of the LSfR strategy, such as allowing floodwaters to flow (ARFU) or limiting the use of embankments on agricultural land (LFEAA). Regarding the subsequent impact of flooding, the prevailing opinions were about the negative impact on crop yields (FACRC, FMCRC) (see Table 2).

Cluster 1 was characterized by a nearly equal sex distribution (84 women vs. 85 men). Respondents mostly had secondary education and were young. This group rated highly the elements of the LSfR strategy. Respondents supported proposals to allow floodwaters to flow into undeveloped riverside areas (ARFU), create polders and dry retention basins (CFP_DRR), and designate riverside areas for the cultivation of permanent floodplain meadows (FEM). In this cluster, respondents perceived the positive effects of flooding on yields of both arable crops (sown area) and permanent grasslands (permanent meadows).

In cluster 2, the sex distribution was slightly favorable to women (115 vs. 93 men). This group had the highest percentage of people with higher education. Respondents supported engineering flood protection measures. The proposal to limit flood embankments (LFEAA) received low ratings, while technical measures, such as the construction of flood embankments (EBA, EAL), the construction of large river dams (BLD), and the perceived need to prohibit development in floodplains (DFEP), received high ratings. Respondents in this cluster pointed to the negative subsequent impact of flooding on crop yields (FACRC).

It can therefore be concluded that in the surveyed group of respondents, three groups of approaches to the issues examined were identified: Cluster 0: low level of confidence in the positive effects of floods, low acceptance of the LSfR strategy; Cluster 1: positive perception of the effects of floods and natural activities that increase water retention; Cluster 2: dominance of the engineering approach, consisting in the construction of dams, river regulation, and strong belief in the harmful subsequent impact of floods on plant yields.

To identify determinants of perceptions of river valley spatial development, a Mann–Whitney U test was used to analyze the ratings for individual items, broken down by sociodemographic characteristics and expressed as binary (Sx, HC, OFH, see Table 2). The comparison consisted of testing the hypothesis that there were no differences between the groups. Statistically significant results of this analysis are presented in

Table 5. Statistically significant results of the Mann–Whitney U test, comparisons of differences in the ratings of the studied items between groups of sociodemographic characteristics.

Variables 1	Total Rank Group 1	Total Rank Group 2	U	Z Corrected	p-Value	n Group 1	n Group 2
Sx	W	M
RRR	91,223.00	67,543.00	35,158.00	2.22479	0.02609 *	309	254
FEM	82,179.50	76,586.50	34,284.50	−2.73666	0.00621 *	309	254
FACIC	83,156.50	75,609.50	35,261.50	−2.12819	0.03332 *	309	254
HC	No	Yes
EBA	100,098.5	58,667.50	33,691.50	2.42656	0.01524 *	340	223
OFH	No	Yes
EAL	98,850.50	59,915.50	31,322.50	−2.64656	0.00813 *	367	196
DFEP	107,523.00	51,243.00	31,937.00	2.29033	0.02200 *	367	196
EFPM	106,734.00	52,032.00	32,726.00	2.64098	0.00827 *	367	196

*—Statistically significant coefficients (p < 0.05). ¹—The names of the items are in Table 2.

. The Z coefficient indicates which ranks are higher in the compared groups. If the Z value is <0, it means that the ratings in Group 1 are lower than in Group 2.

The respondents included 309 women and 254 men. Statistically significant differences between these groups were found for the RRR item, i.e., the demand for river regulation. In this case, men were more likely to advocate for straightening and concreting rivers. Women, on the other hand, gave higher priority to developing river valleys in the form of floodplain meadows, which are used for fodder and energy purposes (FEM). Similarly, women, to a greater extent than men, perceived the positive after-effects of flooding on permanent grasslands (FACIC).

An analysis of respondents’ responses based on whether they had children (HC) revealed statistical differences only in relation to the EBA requirement that built-up riverside areas should be embanked. The group without children comprised 340 people, while the remaining 223 had children. In this case, parents rated the proposal to build embankments around built-up areas more highly. This result was likely related to their responsibility, not only for themselves but also for their families, and in this case it seems that flood embankments are associated with safety.

Item OFH concerned the occurrence of floods in the respondents’ area of residence. Group 1 (367 people) experienced no flooding, while Group 2 (196 people) experienced it. The identified statistical differences concerned item EAL, the requirement to embank agricultural lands, which was rated lower in the group with flood experience. These individuals were less likely to agree with the statement that agricultural riverside areas should be embanked. This indicates a greater acceptance of the natural functions of river valleys as floodplains and not necessarily as areas requiring engineering protection. Item DFEP—development in floodplains, should be prohibited—was rated higher by respondents with flood experience. Awareness of the risk associated with flooding shaped stricter expectations for spatial planning. Individuals from the second group were also more likely to support the LSfR strategy. These are important observations, as respondents who experienced flooding, due to their experience, had a more realistic approach to creating flood protection.

Among the sociodemographic characteristics with more than two levels treated as determinants of the perception of water retention, respondents’ education (Ed), place of residence (PoR), and professional status (PS) were considered. The results of this analysis are presented in

Table 6. Results of the Kruskal–Wallis H test of the influence of sociodemographic characteristics on the evaluation of the studied items.

Variables 1	Ed	Variables 1	PS
CFP_DRR	H (3, n = 563) =9.998676 p = 0.0186 *	RRR	H (2, n = 563) =9.158335 p = 0.0103 *
DFEP	H (3, n = 563) =9.357683 p = 0.0249 *	SDFW	H (2, n = 563) =9.410665 p = 0.0090 *
	PoR	EBA	H (2, n = 563) =14.23217 p = 0.0008 *
CFP_DRR	H (2, n = 563) =7.722789 p = 0.0210 *	ARFU	H (2, n = 563) =6.873955 p = 0.0322 *
FMCRC	H (2, n = 563) =6.717507 p = 0.0348 *	FEM	H (2, n = 563) =11.81338 p = 0.0027 *
		DFEP	H (2, n = 563) =6.194389 p = 0.0452 *
		FACIC	H (2, n = 563) =14.13463 p = 0.0009 *
		FACRC	H (2, n = 563) =7.091708 p = 0.0288 *

*—Statistically significant coefficients (p < 0.05). ¹—The names of the items are in Table 2.

, while the results of the post hoc analysis are presented in

Table 7. Statistically significant results of post hoc analysis, comparisons between groups using the Mann–Whitney U test for statistically significant differences presented in Table 6.

Variables 1	Group 1 vs. Group 2	Total Rank Group 1	Total Rank Group 2	U	Z Corrected	p-Value	n Group 1	n Group 2
CFP_DRR	Ed: 0 vs. 2	465	27,730	2764	−1.925	0.0405 *	30	235
CFP_DRR	Ed: 1 vs. 2	9045	27,730	13,218.5	−2.564	0.0062 *	134	235
DFEP	Ed: 2 vs. 3	27,730	13,530	16,128.5	−2.772	0.0037 *	235	164
CFP_DRR	PoR: 0 vs. 1	39,340	15,225	20,823.5	−2.602	0.0056 *	280	174
FMCRC	PoR: 0 vs. 1	39,340	15,225	21,284.5	−2.263	0.0190 *	280	174
RRR	PS: 0 vs. 2	465	34,191	4908.5	2.276	0.0174 *	30	261
RRR	PS: 1 vs. 2	37,128	34,191	39,482	2.243	0.0190 *	272	261
SDFW	PS: 0 vs. 2	465	34,191	4853.5	2.15	0.0218 *	30	261
SDFW	PS: 1 vs. 2	37,128	34,191	39,800	2.422	0.0105 *	272	261
EBA	PS: 0 vs. 2	465	34,191	4972	2.422	0.0091 *	30	261
EBA	PS: 1 vs. 2	37,128	34,191	40,785	2.976	0.0013 *	272	261
ARFU	PS: 0 vs. 2	465	34,191	5012.5	2.514	0.0086 *	30	261
FEM	PS: 0 vs. 1	465	37,128	5129.5	2.312	0.0143 *	30	272
FEM	PS: 0 vs. 2	465	34,191	5311	3.198	0.0007 *	30	261
DFEP	PS: 0 vs. 2	465	34,191	4763	1.943	0.0436 *	30	261
FACIC	PS: 0 vs. 1	465	37,128	2538.5	−3.396	0.0005 *	30	272
FACIC	PS: 0 vs. 2	465	34,191	2650.5	−2.897	0.0029 *	30	261
FACRC	PS: 0 vs. 1	465	37,128	5223.5	2.519	0.0098 *	30	272
FACRC	PS: 0 vs. 2	465	34,191	4935	2.337	0.0161 *	30	261

*—Statistically significant coefficients (p < 0.05). ¹—The names of the items are in Table 2.

.

Education level influenced only two items. Regarding the CFP_DRR item, higher education led to greater acceptance of leaving space for rivers in the form of floodplain polders and dry retention reservoirs. Similarly, respondents with higher education were more accepting of the ban on development in floodplains (DFEP) than those with secondary education.

Place of residence (item PoR) influenced only the acceptance of polder creation (CFP_DRR) and the perception of the negative impact of flooding on the yields of permanent meadows (FMCRC). People living in rural areas showed a lower acceptance of polder creation compared to those living in small towns. However, residents of small towns emphasized the negative impact of flooding on the yields of permanent meadows to a greater extent.

Professional status (PS) influenced the perception of many items. It is worth mentioning that in Group 2, students constituted 49%, the unemployed 8%, and retirees 43%. Group 2 rated the proposal for river regulation higher than Group 0 and Group 1. At the same time, this group rated the need to slow the rate of water outflow from the catchment area higher than the other groups.

These two types of actions are contradictory, indicating a lack of adequate knowledge regarding water retention and management. Group 2 also advocated for embanking built-up areas to a greater extent than the other groups. However, the group of people working in agriculture (marked Group 0), demonstrated a lower level of acceptance compared to Group 2 regarding leaving agricultural land unembanked. People working in agriculture rated the proposal to use floodplains as permanent meadows higher than both other groups. Group 0, compared to Group 2, rated the proposal to ban development in floodplains higher. An interesting result was that Group 0 rated the positive consequential impact of flooding on permanent meadows lower than both other groups. Similarly, people working in agriculture emphasized the negative consequential impact of flooding on crop yields (sown area) more strongly than the other groups.

4.3. Analysis of the Relationships Between the Studied Elements of Perception of the Economy in River Valleys

The collected research material was analyzed to identify relationships between perceptions of river valley management elements, perceptions of retention, and selected respondent characteristics. Spearman’s ρ (rho) correlation analysis was performed. The results are presented in

Table 8. Spearman’s correlation coefficients ρ between the assessments of flood safety and water retention methods.

Variables 1	DR	NAFSP	RRR	SDFW	EBA	EAL	BLD	ARFU	LFEAA	CFP_DRR	FEM	DFEP
Ag	0.0348	0.0707	−0.0075	0.0020	0.1017	0.0252	−0.0465	0.0877	0.0630	−0.0108	0.0155	0.0385
	p = 0.410	p = 0.094	p = 0.859	p = 0.962	p = 0.016	p = 0.551	p = 0.271	p = 0.037	p = 0.135	p = 0.798	p = 0.713	p = 0.362
DR		−0.0533	−0.0415	−0.0585	−0.0430	−0.0500	0.1276	−0.0133	0.0437	−0.0305	−0.0364	−0.0320
		p = 0.206	p = 0.325	p = 0.166	p = 0.309	p = 0.236	p = 0.002	p = 0.752	p = 0.300	p = 0.470	p = 0.389	p = 0.449
NAFSP			0.2751	0.1608	0.2046	0.1556	0.1859	0.0479	−0.0626	0.1704	0.0548	0.0693
			p = 0.000	p = 0.000	p = 0.000	p = 0.000	p = 0.000	p = 0.256	p = 0.138	p = 0.000	p = 0.195	p = 0.101
RRR				0.1720	0.2245	0.2640	0.3017	0.0116	−0.1056	0.2317	0.0802	0.1111
				p = 0.000	p = 0.000	p = 0.000	p = 0.000	p = 0.784	p = 0.012	p = 0.000	p = 0.057	p = 0.008
SDFW					0.1100	0.0300	0.2109	0.1784	0.1576	0.1966	0.0958	0.1220
					p = 0.009	p = 0.478	p = 0.000	p = 0.000	p = 0.000	p = 0.000	p = 0.023	p = 0.004
EBA						0.3003	0.1878	0.1253	−0.1372	0.2741	0.2050	0.1787
						p = 0.000	p = 0.000	p = 0.003	p = 0.001	p = 0.000	p = 0.000	p = 0.000
EAL							0.2389	−0.1191	−0.3308	0.0998	0.0096	0.0144
							p = 0.000	p = 0.005	p = 0.000	p = 0.018	p = 0.820	p = 0.733
BLD								0.0951	−0.0307	0.1967	0.1820	0.0813
								p = 0.024	p = 0.468	p = 0.000	p = 0.000	p = 0.054
ARFU									0.1812	0.2803	0.2937	0.2679
									p = 0.000	p = 0.000	p = 0.000	p = 0.000
LFEAA										0.0281	0.0453	0.0578
										p = 0.506	p = 0.283	p = 0.171
CFP_DRR											0.4106	0.2516
											p = 0.00	p = 0.000
FEM												0.2800
												p = 0.000
Very weak correlation		Weak dependence		Moderate dependence		Strong dependence		Very strong dependence		Statistically significant coefficients
0.0–0.2		0.2–0.4		0.4–0.6		0.6–0.8		0.8–1.0		p ≤ 0.05

¹—The names of the items are in Table 2.

,

Table 9. Spearman’s correlation coefficients ρ between the assessments of flood safety and water retention methods and the perception of the subsequent impact of floods on crop yields.

Variable 1	Ag	DR	NAFSP	RRR	SDFW	EBA	EAL	BLD	ARFU	LFEAA	CFP_DRR	FEM	DFEP
FACIC	−0.0480	0.0498	−0.0828	−0.1581	0.0664	−0.0484	−0.1680	−0.0303	0.1200	0.2075	−0.0097	0.0713	−0.0171
	p = 0.255	p = 0.238	p = 0.049	p = 0.000	p = 0.116	p = 0.252	p = 0.000	p = 0.473	p = 0.004	p = 0.000	p = 0.818	p = 0.091	p = 0.686
FACRC	0.0456	−0.0366	0.1157	0.1832	0.0218	0.0675	0.1985	0.0953	−0.0343	−0.1614	0.0803	−0.0237	0.0896
	p = 0.280	p = 0.386	p = 0.006	p = 0.000	p = 0.605	p = 0.110	p = 0.000	p = 0.024	p = 0.417	p = 0.000	p = 0.057	p = 0.575	p = 0.033
FMCIC	0.0131	−0.0041	0.0131	−0.1083	0.0708	0.0616	−0.1108	0.0223	0.0918	0.0950	0.0651	0.1135	0.0067
	p = 0.757	p = 0.922	p = 0.757	p = 0.010	p = 0.093	p = 0.144	p = 0.008	p = 0.598	p = 0.029	p = 0.024	p = 0.123	p = 0.007	p = 0.875
FMCRC	−0.0257	0.0180	0.0079	0.1273	−0.0219	−0.0543	0.1381	0.0373	−0.0765	−0.0357	−0.0502	−0.0776	−0.0188
	p = 0.542	p = 0.671	p = 0.851	p = 0.002	p = 0.604	p = 0.198	p = 0.001	p = 0.377	p = 0.070	p = 0.398	p = 0.234	p = 0.066	p = 0.657
Very weak correlation		Weak dependence		Moderate dependence		Strong dependence		Very strong dependence			Statistically significant coefficients
0.0–0.2		0.2–0.4		0.4–0.6		0.6–0.8		0.8–1.0			p ≤ 0.05

¹—The names of the items are in Table 2.

and

Table 10. Spearman’s correlation coefficients ρ between the assessments of the subsequent impact of floods on crop yields.

Variables 1	FACRC	FMCIC		FMCRC
FACIC	−0.8302	0.5212		−0.4316
	p = 0.00	p = 0.00		p = 0.00
FACRC		−0.4633		0.4859
		p = 0.00		p = 0.00
FMCIC				−0.8396
				p = 0.00
Very weak correlation	weak dependence	moderate dependence	strong dependence	very strong dependence	statistically significant coefficients
0.0–0.2	0.2–0.4	0.4–0.6	0.6–0.8	0.8–1.0	p ≤ 0.05

¹—The names of the items are in Table 2.

.

Based on the correlation analysis (Table 8), it can be concluded that respondents’ perceptions of the direction of river valley development were coherent. Those who positively assessed the proposal to allocate floodplains for the cultivation of floodplain meadows (FEM) to increase water retention and the acquisition of fodder and energy biomass simultaneously supported the proposals to allow floodwaters to spread into undeveloped areas (ARFU) and to prohibit development of riverside areas (EBA). Other examples of such coherence were the correlations between the CFP_DRR items and the ARFU and EBA. A positive correlation was also identified between the assessment of the creation of floodplain polders (CFP_DRR) and the proposal to regulate rivers (RRR), which in this case indicates a lack of coherence between these views. However, it can be assumed that respondents misunderstand the concept of river regulation, equating these activities with the scope of work included in the renaturalization process.

The coexistence of high scores for the postulate of allowing floodwaters to spread in undeveloped areas (ARFU) with the postulate of creating polders and dry reservoirs (CFP_DRR), cultivation of floodplain meadows in floodplains (FEM), and prohibition of development in floodplains (DFEP) was also identified.

However, respondents who rated the item “need for river regulation” (RRR) highly also supported embankment construction, both in developed areas (EBA) and in undeveloped areas (EAL), as well as the construction of large dams on major rivers. This indicates that these respondents favor the MWAfP strategy, which prioritizes large-scale engineering.

The conducted research revealed no strong correlations between the perception of the forms and methods of shaping retention and river valley development and the distance from the river to the respondents’ place of residence. Similarly, no strong correlations were identified between the studied characteristics and the respondents’ age.

Table 9 presents the results of Spearman’s ρ correlations between the ratings of items related to the aftereffects of flooding on sown areas and permanent meadows and the remaining items. The calculated coefficients indicated weak correlations between the studied features, but many coefficients were statistically significant. For example, the item “EAL—embankment of agricultural lands” was significantly correlated with all items related to the aftereffects of flooding on crop yields. The direction of the correlations indicated that a higher belief in the need to build flood embankments was accompanied by a belief in the negative aftereffects of flooding on crop yields, both in sown areas and permanent meadows. Similar correlations were also found for the river regulation proposal (RRR). Respondents, who believed in the need to regulate rivers, also believed in the negative aftereffects of flooding on crop yields. However, caution should be exercised when drawing conclusions, as the strength of the calculated correlations was weak.

Table 10 presents Spearman’s correlation coefficients (ρ) between assessments of the subsequent impact of flooding on crop yields. The calculated coefficients indicated strong correlations between the assessments of the FACIC and FACRC items, and between the FMCIC and FMCRC items. These strong correlations resulted from the opposing formulations of the theses regarding the subsequent impact of flooding. However, it is worth emphasizing the convergence of perceptions of the effects of flooding on sown areas and permanent meadows. Respondents, who perceived a negative impact of flooding on sown areas, also had this belief regarding crop yields on permanent meadows. A similar perception was observed for all other item assessments regarding the impact of flooding on agriculture.

4.4. Identification of Determinants of the Choice of River Valley Development Strategies

The item EFPM (effective flood protection measures) was a dichotomous variable. Respondents were presented with two sets of measures ensuring hydrological safety. Set 0 included measures aligned with the MWAfP strategy, while Set 1 included measures aligned with the LSfR strategy. Logistic regression analysis was conducted to identify factors determining the choice of retention strategy and river valley management. Statistically significant variables were included in the final model, and the results are presented in

Table 11. Logistic regression coefficients for statistically significant variables.

Variables 1	B (β) 2	SE 3	z 4	p-Value 5	95% CI (Lower) 6	95% CI (Upper) 6	OR 7	95% CI (Lower) 8	95% CI (Upper) 8
Ed	0.853	0.217	3.925	0.0001 *	0.427	1.278	2.346	1.533	3.591
PoR	−0.460	0.232	−1.983	0.0474 *	−0.915	−0.005	0.631	0.401	0.995
RRR	−1.943	0.232	−8.364	0.0000 *	−2.398	−1.487	0.143	0.091	0.226
SDFW	0.796	0.200	3.983	0.0001 *	0.404	1.188	2.217	1.498	3.281
EBA	−0.706	0.210	−3.362	0.0008 *	−1.118	−0.295	0.494	0.327	0.745
EAL	−0.612	0.172	−3.564	0.0004 *	−0.948	−0.275	0.542	0.387	0.759
BLD	−0.617	0.181	−3.404	0.0007 *	−0.972	−0.262	0.540	0.379	0.770
ARFU	0.720	0.190	3.782	0.0002 *	0.347	1.094	2.055	1.415	2.985
FEM	0.933	0.237	3.937	0.0001 *	0.469	1.398	2.542	1.598	4.045
DFEP	0.342	0.160	2.141	0.0323 *	0.029	0.655	1.408	1.029	1.925
FACIC	0.613	0.244	2.517	0.0118 *	0.136	1.091	1.846	1.146	2.976
FACRC	0.585	0.256	2.281	0.0226 *	1.087	1.087	1.795	1.086	2.967

¹—The names of the items are in Table 2; ²—Regression coefficient; ³—Standard error of the regression coefficient estimate; ⁴—Test statistic (used to calculate the p-value); ⁵—p-value—if p < 0.05, the effect is considered statistically significant (*); ⁶—Confidence interval for the B-ratio (specifies the range within which the true value of the parameter lies with 95% probability); ⁷—Odds ratio (indicates how much a 1-unit change in the predictor variable changes the odds of a given outcome); ⁸—Confidence interval for the OR (if it contains the value 1, the effect may not be significant).

. To facilitate the interpretation of the effect of predictors on the dependent variable, regression coefficients were converted to odds ratios (ORs) [123].

The results of the calculations indicate that the level of education (Ed) was a significant predictor of choosing the reference strategy. Increasing the level of education more than doubled the odds of choosing this strategy (OR = 2.346). The item place of residence (PoR) proved to be a significant predictor of choosing the riverside space management strategy. Increasing the degree of urbanization (from villages to small towns to large cities) was associated with a decrease in the probability of choosing the LSfR strategy. The odds ratio of OR = 0.631 means that each additional category of urbanization decreases the odds of choosing this strategy by approximately 37%.

In the case of the RRR item, B = −1.943, OR = 0.143, this means that the odds of choosing the LSfR strategy decrease by 85.7%, compared to the reference group. Respondents who supported river regulation were over 6 times less likely (1/0.143 ≈ 7) to choose the strategy of leaving space for rivers than those who did not declare such support.

However, a higher assessment of the need to slow down water outflow from the catchment (SDFW) significantly increased the odds of preferring the LSfR strategy. The odds ratio was OR = 2.217, which means that it more than doubled the odds of choosing the natural strategy.

Respondents who advocated for the need to embank built-up areas (EBA) were less likely to choose the LSfR strategy. The odds of choosing this strategy were approximately 50% lower (OR = 0.494). Similarly, the assessment of the need to embank agricultural land (EAL) had a significantly negative impact on the choice of the LSfR strategy (B = −0.612, OR = 0.542).

Respondents who supported the construction of large reservoirs were significantly less likely to choose the LSfR strategy compared to those who do not support this postulate. (B = −0.617, OR = 0.540) means that the odds of choosing the LSfR strategy (i.e., EFPM = 1) are 46% lower in this group. Therefore, people who support the construction of dams are almost half as likely to give space to rivers.

High ratings for the potential for floodwaters to spread onto undeveloped areas (ARFU) significantly increased the likelihood of supporting the LSfR strategy (OR = 2.055). Similarly, a positive attitude towards allocating floodplains for the cultivation of floodplain meadows (FEM) was associated with a 2.5-fold increase in the likelihood of choosing the LSfR strategy (OR = 2.542). High ratings for prohibiting development in floodplains (DFEP) also positively influenced the choice of the LSfR strategy (OR = 1.408).

In the calculated model, two variables related to the perception of flood impact on sown area crop yields were statistically significant: FACIC (the belief that flooding can increase crop yields) and FACRC (the belief that flooding can reduce crop yields). These variables had opposing content yet both showed a positive effect on the dependent variable EFPM. However, it should be noted, that these variables are strongly negatively correlated (ρ = –0.83; see Table 10), indicating their partial overlap and potential multicollinearity in the model. In such cases, a so-called suppression effect can occur, in which one variable strengthens (or suppresses) the influence of the other variable in a multivariate model. Therefore, this result may suggest that regardless of the direction of flood impact perception, assigning a high importance to it (whether positive or negative) is associated with a preference for a spatial strategy, possibly as an expression of general environmental sensitivity or knowledge of the complex effects of flooding.

5. Discussion

Based on the conducted research, it can be concluded that the study groups’ perception of methods for ensuring flood safety, and therefore river valley management, varied. A large group of respondents expected the construction of flood embankments in both developed and agricultural areas. This group believed in the need to regulate rivers and accelerate the rate of stormwater runoff. Our study also identified a group of respondents who understood the need to increase water retention and slow the rate of runoff. These respondents advocated the construction of polders and the development of riverside areas as floodplain meadows. Diversified perceptions of flood safety and the need to consider the socio-cultural context are currently part of the flood risk research paradigm [8,68,69].

Despite the extensive literature indicating the higher effectiveness of flood safety through the use of LSfR strategies [11,62,64,128,129], our respondents preferred the MWAfP strategy. This phenomenon is confirmed by the concept of “risk perception theory” [17,45]. It turns out that long-term implementation of a specific security strategy, and institutional support for such actions, strengthens the public’s belief in the effectiveness of specific technical solutions.

Our research indicates that less educated individuals were proponents of technical strategies, based on the construction of flood embankments, large reservoirs, or river regulation. Similar results, emphasizing the quality of human capital in shaping space, are presented by other authors [130,131]. Therefore, it can be concluded that the level of education in society is a predictor of perceptions of river valley development. Similarly, women, and residents of large cities, demonstrated greater confidence in MWAfP strategies. This phenomenon may stem from the greater sensitivity to crisis situations among women and people with stronger ties to urban areas [132,133].

Flood control strategies based on natural resources often constitute a more effective and cheaper alternative to technical solutions [134]. In our study, respondents supporting such solutions were generally characterized by higher education and lived in rural areas.

The literature also describes the benefits of cultivating permanent grasslands, often in the form of semi-natural floodplain meadows. They provide numerous ecosystem services, including retention, buffering, ecological filtering, ecological corridors, ecotones, and biodiversity concentrations [5,6,135]. Ecosystem services associated with floodplains are gaining importance, especially in the context of climate change, and the need to maintain, or even increase, flood retention areas [88]. The process of siltation plays a key role in the formation of alluvial soils, in which the deposit of silt left after a flood rejuvenates soil profiles, often providing large amounts of biogenic elements. This process can be considered an alternative to mineral fertilization [83,136,137,138]. In our study, respondents did not perceive a positive impact of floods on floodplains, resulting from the abundance of nutrients in the sediments left behind. The overwhelming majority of our respondents stated that the aftereffects of floods are negative. At the same time, these individuals advocated for the construction of embankments on agricultural land, which consequently cuts off the supply of free nutrients and leads to the degradation of alluvial soils [6]. However, the literature indicates that there are threats associated with the negative impact of floods on alluvial soils [84,139,140]. Therefore, it is impossible to clearly assess the respondents’ negative attitudes towards the effects of floods on agricultural lands. Depending on the biogeochemical characteristics of the catchment, its development method, and the characteristics of the sediment deposits left by the flood, the aftereffects of floods may have varying impacts on crop yields.

However, using floodplains as natural flood meadows is a way to utilize the positive potential of alluvial deposits [80]. In light of our research, public awareness of the positive effects of floods is negligible.

The group of respondents who perceived the positive impact of flooding on crop yields also highly valued the proposal to increase water retention, create polders, and use floodplains as riparian meadows. The literature indicates that such solutions are being used with increasing frequency [26].

Based on the results of logistic regression, we can identify factors that positively influence respondents’ preference for the LSfR strategy. These include education level, general understanding of the functioning of natural rivers, and the retention role of floodplains used for agricultural purposes as flood meadows [26,141].

The younger respondents we surveyed were also typically better educated and more likely to support the LSfR strategy. On the other hand, older individuals, with greater life experience, did not demonstrate greater acceptance of this type of strategy. This may be a consequence of the long-standing promotion and implementation of technical flood protection measures [16,17]. However, this is surprising, as it indicates that these individuals fail to recognize the vicious cycle of flood protection [142,143]. In this context, it is worth noting that the belief in the need to regulate rivers, and build large reservoirs, were negative predictors of support for the LSfR strategy. It is worth emphasizing that living in large cities was a negative predictor of choosing the LSfR strategy. Rural residents, one might say, are closer to nature and demonstrate a greater understanding of the role of natural forms of retention. Therefore, awareness-raising education regarding the creation of flood protection based on natural resources should be conducted primarily in large cities. It is necessary to make farmers aware that maintaining security systems, based on taming the river in technical development, is becoming more and more expensive and impossible [74].

A key condition for the development of the LSfR strategy also appears to be convincing farmers of the positive effects of siltation on semi-natural floodplain meadows. Many authors in the literature emphasize the importance of ecosystem services provided by floodplain meadows. These studies emphasize that these services can be treated as public goods [144,145,146,147]. In line with the concept of “public money for public goods” [148], the introduction of systemic incentive mechanisms for farmers cultivating semi-natural floodplain meadows in floodplains should be considered.

6. Summary, Conclusions, and Recommendations

Based on the research, it can be concluded that most respondents perceived deficiencies in ensuring flood safety. Perceptions of this safety varied. Respondents expressed expectations for river regulation, the construction of large dams, and the construction of flood embankments, in both developed and agricultural areas. The prevailing sentiment among respondents regarding the MWAfP strategy is likely a consequence of the long-standing, administratively supported concept of land improvement. Awareness of the benefits of the LSfR strategy existed, though not universally, within the surveyed community.

This study identified certain sociodemographic patterns in the perception of river valley management. More educated individuals recognized the protective potential of the LSfR strategy, recognized the need to slow the rate of water outflow from the catchment area, and demonstrated a greater understanding of nature-based retention methods. Women, on the other hand, demonstrated greater confidence in technical flood protection measures, as did those with children. Residents of large cities also supported the MWAfP strategy through technical flood protection measures, which may be associated with a weakened daily connection with nature and a greater reliance on technical solutions.

Among the determinants of the LSfR strategy, the strongest positive influences were education level, belief in the need to slow water runoff from the catchment area, the need to limit embankments in agricultural areas, and the cultivation of meadows in floodplains. Living in large cities, belief in the need to embank agricultural areas, and support for the construction of large river dams had a strong negative influence.

In view of the above, and based on the calculated model of determinants of the LSfR strategy selection, we propose the following actions:

(a)

In the study area, there is a need to improve social capital through awareness-raising and educational activities, primarily aimed at rural residents. Education should be focused on highlighting the benefits of the alluvial process in riparian areas being used as semi-natural flood meadows. These programs should emphasize benefits such as free fertilization through the alluvial process, the potential for using the resulting green mass for feed and energy, and, above all, improved safety resulting from increased water retention.

(b)

Since most of the study area is used for agriculture, convincing farmers to change their floodplain use is crucial to increasing valley retention. We propose introducing incentives for farmers to engage in agricultural activities that slow down water runoff, primarily by cultivating semi-natural flood meadows. This will require systemic solutions that recognize this form of economic use of floodplains as social goods, co-financed from state budget funds.