Willingness-to-Pay for Blue Ecosystem Services of Natural Pools in Sri Lanka: A Discrete Choice Experiment

Abstract

Ecosystem services offered by freshwater ecosystems, in the form of natural pools have not been fully realized by the public, which has led to limited attention on the conservation of these pools. This study therefore was conducted to investigate the user preferences for sustainable enhancement of recreational values of natural pools and their immediate environment. A total of 320 local users of natural pools located in Rangala and Nillambe were surveyed as the sample. A Discrete Choice Experiment (DCE) was employed to elicit the preference and user Willingness to Pay (WTP). The Marginal Willingness to Pay (MWTP) for the improvement of recreational values of natural pools was estimated using a conditional logit (CL) model. Outcomes of the WTP disclosed a clear preference hierarchy for various enhancements and contributions. Users were willing to pay Sri Lankan Rupees (LKR) 297.6 to reduce the environmental damage by 50% and LKR 84.4 to raise the community income by 20%. The option to have a higher number of recreational activities was highly valued. Respondents were willing to pay a value of LKR 554.8 per year for the multiple benefits provided by the pool ecosystems. Conclusively, the study suggested that efforts to upgrade these natural pools should prioritise income generation, broadening of recreational activities and environmental conservation, in line with respondents’ inclinations.

1. Introduction

Blue ecosystem services refer to the benefits derived from marine and freshwater ecosystems. These services are crucial for human well-being and include a variety of functions provided by aquatic environments such as oceans, rivers, lakes, wetlands, and estuaries. Natural pool ecosystems, often referred to as freshwater pools, are small bodies of standing water found in various terrestrial landscapes. It is a type of pool that uses a natural water purification system to maintain the quality of water [1]. These ecosystems are characterised by their relatively small size, shallow depth, and often isolated nature. Natural pools are typically fed by surface water runoff, groundwater seepage, or precipitation. They may also receive inputs from nearby streams or rivers during periods of high flow [2]. Despite their small size, natural pools can support a diverse array of aquatic and semi-aquatic species, including fish, amphibians, aquatic insects, and various types of vegetation. These ecosystems provide important habitat for both resident and migratory species [3]. Further, natural pools play a key role in nutrient cycling and energy flow within ecosystems [4].

The hydrology of natural pools can vary seasonally and in response to changes in precipitation and temperature. Therefore, these may experience fluctuations in water level, temperature, and dissolved oxygen concentrations, which can influence the distribution and abundance of aquatic species [3]. Natural pools provide a range of important ecological functions and ecosystem services; provisioning, regulating, supporting and cultural such as fish, habitat provision, water storage, groundwater recharge, and flood mitigation and recreation [5]. They also contribute to the overall landscape heterogeneity and biodiversity of terrestrial ecosystems. From a human use perspective, these ecosystems can serve as recreational sites for activities such as swimming, fishing, and wildlife observation [6,7].

Recreation, which refers to an activity of leisure and the refreshment of mind and body, is a broad field of study that encompasses various theoretical and practical aspects. It is considered as a key human activity that shapes individual leisure time behaviours [8]. These recreational activities are essential for maintaining health, physical and social development of people. Engagement in leisure activities, together with recreation, has been linked to weight reduction, improved cardiovascular health, positive mental health outcomes and subjective well-being [9]. Recreation activities can fulfill individual and societal needs, channel excess energy into socially acceptable activities, and promote social interactions [10]. Engagement in outdoor recreation contributes to economic growth, job creation, and increased economic activity [11,12]. Valuing recreational services provided by freshwater ecosystems involves assessing the economic benefits users and visitors derive from activities such as fishing, boating, and swimming. These recreational services are a significant component of the overall value of freshwater ecosystems, influencing both individual well-being and local economies. However, if under studied and mismanaged, unregulated recreational activities could lead to negative environmental externalities such as pollution, habitat destruction, and introduction of invasive species, which deletariously influence these fragile ecosystems [13,14,15]. Recreation in natural pools can significantly influence water quality and biodiversity, particularly, when there is an increased number of visitations [16,17]. Thus natural pools, as common-pool resources, are often subject to challenges related to exploitation and management due to their open-access nature [18].

Sri Lanka is blessed with 103 rivers that carry water from the central highlands into the western, southern, and eastern coasts [19] resulting in biodiversity-rich natural pools or nature pools, which can offer multiple benefits. Among these uses, recreational benefits are commonly enjoyed by both villagers and visitors from outside of the village. Natural ecosystems in Sri Lanka provide opportunities for tourists to engage in these activities and experience the diverse ecological landscapes. Currently, there seems to be a potential shift towards natural pools as a preferred recreational resource among tourists visiting tropical destinations [20]. Tourism is a major contributor to Sri Lanka’s economy, representing over 10% of Gross Domestic Production (GDP) and creating employment opportunities for around 500,000 people [21,22]. Therefore, conservation and restoration of natural pools can be economically advantageous for the recreational industry. However, a strategic and sustainable approach is needed to overcome challenges and capitalise on the potential future trends in the recreational services sector [23,24].

Several activities can be planned and implemented for the purpose of improving recreational activities around freshwater ecosystems. Community involvement in natural pool management can enhance both natural and social systems, leading to improved health and well-being [25]. Engaging communities in such projects can increase environmental consciousness and promote local environmental action [26]. Community-driven approaches to develop water trails and recreational areas can improve health, quality of life, and community engagement. Developing access points, monitoring water quality, managing woody debris, and implementing safety plans for water trails are few key actions that can be taken by the communities to improve recreation around the freshwater ecosystems, as it plays a crucial role in supporting and maintaining watershed functions and conservation [27]. Integrated watershed development models have shown a 28% increase in family incomes in villages located in watersheds, compared to non-watershed villages, with a significant contribution from crop cultivation [28].

Willingness to Pay (WTP) is a concept used to measure the value individuals place on certain goods or services, specifically in the context of enhancing recreational facilities. It is measured using contingent valuation methods and is influenced by factors such as perceived value, income levels, and alternative recreational options. Considering WTP in decision-making processes for enhancing recreational facilities can facilitate resource allocation, evaluate welfare benefits, and support sustainable development [29,30]. Discrete Choice Experiments (DCEs) and Contingent Valuation Approach (CVA) are two of the commonly used methods to assess WTP and estimate individual choice behaviour. Both of these approaches are based on stated preference [31].

In environmental and natural resource literature, various methodologies and approaches have been employed to assess the economic value of freshwater ecosystems in general and particularly to understand how these ecosystems contribute to human well-being and the economy. The majority of these studies centre around benefits such as clean water, biodiversity, flood regulation, and recreational opportunities offered by freshwater ecosystems. A recent study conducted by Sierra et al. has assessed the economic value of key natural benefits of freshwater ecosystems using a market value method considering transfer of benefits and direct uses. Results of this study have revealed that users consider ecotourism and prevention of flooding to be the most important values provided by freshwater ecosystems [32]. Since there are complex relationships in freshwater ecosystem services, the value of these ecosystem services often depends upon other ecosystem services provided [33,34]. A meta-analysis conducted by Brouwer et al. has found that services provided by freshwater ecosystems such as lakes are greatly valued than the values reported for wetlands by the users [35]. A recent study conducted by Ureta et al. has employed economic valuation techniques such as the payment card approach to explore the WTP towards protection of blue water ecosystem services in South Carolina. According to the findings of Ureta et al. people are willing to pay more for improvements that increase recreational activities, such as better fishing, presence of more types of species, and better access to water areas [36]. Therefore, assessment of WTP has assisted to guide the policy makers on prioritisation of river protection programmes in South Carolina [36].

Another study conducted in five Kenyan counties (Migori, Siaya, Busia, Kisumu, and Homa Bay) has evaluated how much the residents are willing to pay for the conservation of the Lake Victoria Ecosystem. The findings of this study have reported that around 40.9% of respondents are willing to contribute approximately Kenyan Shilling (KES) 500 each, resulting in an estimated total annual WTP of KES 616,279,069 [37]. Using a contingent valuation approach, Bueno et al. has attempted to investigate how local communities value efforts to restore water quality in Sampaloc Lake in Philippines, amid urbanisation pressures. The findings have revealed that households around the lake are willing to pay Philippine Peso (PHP) 177.09 per month, totaling PHP 7,102,017 annually, for these improvements [38].

Many studies have evidenced that people are willing to pay for services and contribute towards conservation of freshwater ecosystems, which directly benefit them or improve their quality of life [36,38]. Further, implementing an environmental users fee system to fund restoration of such freshwater ecosystems has been identified as an effective sustainable solution for management of such sensitive ecosystems [38]. However, such attempts require coordinated efforts on elevating environmental education and comprehensive policies to manage freshwater ecosystems [36,38]. Further, WTP for ecosystem services often varies with numerous factors, including socio-economic status, cultural values, awareness of ecosystem benefits, and local environmental conditions [39]. Previous studies have shown that people are more likely to pay for ecosystem services when they understand their importance and when there are clear mechanisms in place to link their payments to tangible outcomes, such as improved environmental quality or increased access to resources [37].

Utilizing nature-based solutions to ecosystem management have been proven to be sustainable and result in positive impacts on the socio-economic conditions of rural communities, including enhanced income distribution [40,41]. Particularly, participatory management programmes tend to lead to positive changes in the lives of marginal communities, while ensuring the sustainable conservation of freshwater ecosystems in different parts of the world [42]. For this, governments should facilitate policy-making, policy-lobbying and provide resources, and establish partnerships to promote integrated blue water ecosystem management [40,42,43]. However, designing such programmes requires an in-depth understanding on the perceptions of resource users and local communities on different values of freshwater ecosystems, along with their willingness to contribute towards such efforts either financially or from labour [44,45]. To the best knowledge of the authors, no studies have valued the confounded multiple benefits of natural pools, particularly the non-market values, applying a DCE in Sri Lanka, which is a developing country. Therefore, this study provides a significant contribution to this knowledge gap in literature from a developing country perspective.

Further, ecosystem services offer opportunities for diversifying livelihoods of local communities, beyond traditional sectors such as agriculture and forestry. Developing natural pool ecosystems as multi-purpose revenue generating entities can lead in to environmental, economic as well as social benefits [44,46]. Existing research on how much local stakeholders are ready to spend for the enhanced enjoyment and services from bettered recreation spots in natural pools of Sri Lanka is sparse. Hence, there is a noticeable demand to create strategies for enduring enhancement of these areas for recreation and their environs. Therefore, this study explores user preferences for improvement of recreational activities of natural pools and elicits user WTP for conservation of natural pools to ensure sustainable development. The insights from this research could support policy-making, enabling the government to refine recreational offerings and sustainably harness the untapped potential of natural pool resources.

2. Materials and Methods

2.1. Study Area

Natural pools located in the Kandy district, specifically in the Nillambe and Rangala areas were considered as the study sites in the current study. The Kandy District (6.93° to 7.50° N and 80.43°) is located in the central highlands of Sri Lanka processing a wide array of natural environmental features. Both Rangala and Nillambe areas are covered with mountain ranges, tea plantations and a diverse array of natural pools (Figure 1), leading these to become major tourist attractions. The local communities in these areas primarily are low-income tea pluckers.

2.2. Data Collection and Survey Design

The proportionate random sampling technique was employed in selecting a total of 320 locals, who were utilising the natural pools in the two study areas for different purposes. A pre-tested structured questionnaire along with a choice experiment was administered to collect the primary data. The survey instrument focused on three major sections, (1) demographic characteristics; (2) knowledge, awareness, and perception of resource users on conservation and amenity improvement of natural pools; (3) Choice experiment and value elicitation. The perception, awareness and knowledge related questions established a baseline for how much the respondents know about the ecological terminology and its association to the economic and well-being improvement. Furthermore, it also elicited their position on some potential actionable issues relevant to implementation of conservation interventions.

2.3. Analytical Framework of the Choice Experiment

This study employed a DCE to elicit the respondent’s preference by presenting sets of choices with varying attributes and the choice set’s corresponding price vehicle. WTP is a concept used to measure the value individuals place on certain goods or services, specifically in the context of enhancing recreational facilities. The DCEs allow for estimation and forecasting of individual choice behaviour and provide a way to measure preferences for nonmarket goods. Further, DCEs can generate more precise preference estimates compared to traditional “pick the best” approaches, since it captures decision-rule heterogeneity, allowing for a more accurate representation of respondent behaviour [47].

Choice experiments are survey-based valuation methods used to estimate the marginal values of individual elements in environmental resources. Choice modeling provides decision-makers with richer information on economic values, enhancing the quality and sustainability of natural resources at local, national, and global levels [48,49]. Often, DCE designs consist of a small number of choice sets with a limited number of alternatives to increase response efficiency. Optimal designs, also known as orthogonal designs, are used when prior information about population preferences is not available. The orthogonalisation procedure works in the context of choice card attributes and levels, by introducing column vectors for alternative-specific attributes and forcing their values to be orthogonal with other generic attributes within the same alternative. This maintains orthogonality within individual alternatives, but not necessarily across alternatives [47,50].

Although several non-market valuation techniques have been used to estimate the monetary value of ecosystem services, stated preference approaches are increasingly used as they rely on preferences or values as stated by individuals [29,49]. The major advantage of the stated preference methods is that they possess flexibility to capture both use and non-use values. Choice experiment models allow for the estimation of the relative importance of multiple environmental attributes and their levels [37]. Conditional logit (CL) model can be effectively used to relate the probability of a choice among the alternatives to the characteristics of the attribute levels defining those alternatives, which is consistent with the random utility theory. Further, unlike other regression models, the Conditional logit model avoids sparse data biases, leading to more accurate interpretations [49].

2.4. Theoretical Background

Theoretically, DCEs are based on the idea that respondents gain satisfaction from the features of things they choose rather than the things themselves. People aim to get the most satisfaction by picking the best option from a range of possible choices, which are described by their features, while each feature may possess different levels. Therefore, respondent’s choices depend on the combination of these features in the options they are presented with [51].

In a DCE, the satisfaction (or utility) a person (i) gets from a choice (j) is not just about the obvious features (x) of the choice. There are also hidden factors that can affect their satisfaction. These are represented by a random element (ε) in the calculation. This approach follows the random utility theory suggested by Hensher et al. [52]. The theory suggests that this random element is simply added to the utility calculation. Therefore, the utility that person (i) gets from choice (j) is the sum of the utility from the obvious features and this random element [52]. Therefore, individual (i)’s utility (U) from alternative j can be expressed as follows (Equation (1)).

U_ij = X_ij + ε_ij

(1)

A DCE presents several choice sets (R), each including a status quo (x) and various alternatives (xj). A participant (i) chooses an alternative (j), if its utility (Uij) surpasses that of others (Uij’). The chance that (i) chooses (j) over other alternatives (j’) within a set (R) is denoted as probability Pr(j|R), calculated where Uij > Uij’ for all j’ in R, and j ≠ j’. Responses from a survey can be effectively used to find out which option is most preferred by the people. When it is assumed that the random element in the satisfaction calculation is the same across all choices and that the utility people get from the features is linear, a CL model can be used for estimation [53]. Assuming that the error term is independently and identically distributed, and an indirect utility (V) is linearly related with attributes (x), the CL model estimates can be expressed as below (Equation (2)).

V_ij = ASC + β_ij

(2)

In the equation, V is the indirect utility obtained by the _ith individual for the _jth alternative. β is the marginal utility of the attributes (x). The Alternative Specific Constant (ASC) measures the effect of unobserved factors on the choice of alternatives relative to the status quo.

2.5. Designing of the Choice Experiment

Choice experiments often use cards that list different options with various features for participants to choose from. In the current study, the features to include on these cards were chosen after group discussions and by reviewing literature. Attributes that were used for the DCE were ‘Reduction of environmental damage’, ‘Number of recreational activities’, ‘Community (who are interested to gain income) income increase from recreational values’, ‘User labour contribution (hours per visit) for a conservation programme’, and ‘Willingness to Pay per visit’. Using a design approach that considers all possible combinations, a total of 243 different scenarios were initially generated in SPSS (version 23). Subsequently, a narrow down method was applied to restrict the possible scenarios up to few key combinations, which were used for making choice cards. The final set of options, showcasing five main features and their different levels, are displayed in

Table 1. Attributes in the choice card.

Attributes	Levels
Environmental damage Reduction	50%
	20%
	No reduction *
Number of recreational activities	14
	11 *
	Status quo
Community income increase from recreational values	20%
	10%
	No income *
User labour contribution (hours per visit)	1
	0.5
	No contribution *
Fund per adult per visit (LKR)	100
	50
	No fund *

Note: * Status quo option, Number of recreational activities -7 (Walking, Swimming, Playing games, Sightseeing, Singing, Dancing, and Photography), 11 (Walking, Swimming, playing games, Sightseeing, singing, Dancing, Photography, Arts and crafts, Cycling, Meditation, and Yoga), 14 (Walking, Swimming, playing games, Sightseeing, singing, Dancing, Photography, Arts and crafts, Cycling, Meditation, Yoga, Camping, Star-grazing, and Hiking).

.

The choices included different scenarios that participants could pick from, regarding how to improve recreation in natural pools. Each person could choose from Options A, B, or C, where Option C was to keep things as they are at the pools currently. Options A and B offered different changes or improvements expected in future.

2.6. Data Analysis and Interpretation

All the collected data were double-checked and verified on the same day for completeness and consistency before entering it into Microsoft Excel data sheets (version, 2016). Cross-tabulations and logical checks were done to ensure the accuracy of the data. Any null data were treated as missing data and appropriately removed during the analysis.

2.7. Relative Importance Index

Relative index analysis was selected in this study to rank the criteria according to their relative importance based on participants’ replies and prioritise indicators rated on Likert-type scales, as shown in Equation (3).

$$RII={\sum }_{i=1}^5{\displaystyle \frac{W_i{X}_i}{X_i}}$$

(3)

where, $W_i$ = the weight given to each factor by respondents, ranging from 1 to 5.

2.8. Estimating the Residents’ Willingness-to-Pay

A conditional logit model was fitted to analyse the data using STATA (version 2016) software. The model includes attributes in general linear form as shown in Equation (4).

$$U_i=ASC+{\beta }_{1}In\left(X_1\right)+{\beta }_{2}In\left(X_2\right)+{\beta }_{3}In\left(X_3\right)+{ \beta }_{4}In\left(X_4\right)+{\beta }_{5}In\left(X_5\right)+{ϵ}_i$$

(4)

$U_i$ = Utility of the _ith alternative

ASC = Alternative Specific Constant

$x_1,\dots ,x_5$ = Component attributes of _ith alternative

${ϵ}_i$ = Error component

The coefficients of the CL model cannot be interpreted directly. Therefore, the following formula (Equation (5)) can be used to calculate the marginal welfare measure for a change in attribute (i), by dividing the coefficient estimate for that attribute by the coefficient estimate for the monetary attribute [42].

$$MWTP=-\left({\displaystyle \frac{{\beta }_{attribute}}{{\beta }_{monetary attribute}}}\right)$$

(5)

To estimate the mean MWTP, including the upper bound and lower bound within 95% confidence interval, the Krinsky-Robb (KR) parametric bootstrapping technique was used. Furthermore, the Log-likelihood and McFadden’s pseudo-R-squared were used to assess the goodness of model fit [53].

3. Results

3.1. Socio-Demographic Information of the Respondents

A total of 320 respondents participated in the survey, providing insights into their engagement with and attitudes towards natural pools. The socio-demographic profile of the respondents and their recreational behaviour concerning natural pools is shown in

Table 2. Socio-demographic information of the respondents.

Category	Sub Category	Percentage (%)
Gender	Female	62.7
	Male	37.3
Age (Years)	20–30	55.3
	31–45	25.0
	46–60	18.0
	>60	1.7
District	Kandy	46.0
	Nuwara Eliya	13.7
	Matale	5.7
	Colombo	10.3
	Gampaha	8.7
	Anuradhapura	1.0
	Badulla	4.7
	Galle	2.7
	Matara	0.3
	Kegalle	2.0
	Kurunegala	3.3
	Puttalam	1.7
	Other districts	0.0
Education level	Primary education	3.7
	Secondary education	16.7
	Diploma	67.0
	Tertiary education (degree and above)	12.7
Employment status	Farming	32.0
	Private sector job	36.0
	Government sector job	4.67
	Self-employed	11.7
	Unemployed	8.7
	Other	7.0
Monthly income category (LKR)	<15,000	2.3
	15,001 to 30,000	21.0
	30,001 to 45,000	31.3
	45,001 to 60,000	32.00
	≥60,001	14.0
Frequency of visiting natural pool and its surrounding in the past 12 months	Yes. But not now	4.0
	Always	39.7
	Sometimes	29.7
	Rarely	13.7
	No	13.0
Standard of recreational facilities	Very good	7.7
	Good	36.7
	Neither good nor poor	40.3
	Poor	2.3
	Very poor	0
	Do not know	13.0
Willingness to improve sustainable recreational services	Yes	89.0
	No	11.0
Preference on institutional support for improving sustainable recreational services	Community based organisation support	12.0
	Government Support	79.0
	Both	9.0

. The majority of respondents were within the 20–30 year age group (55.3%), indicating a young adult demographic predominance. Those above 60 years were minimally represented (1.67%), suggesting potential barriers or disinterest among the elderly population in engaging in recreational pool activities. Respondents were distributed across various districts, with Kandy (46%), Nuwara Eliya (13.7%), and Colombo (10.3%) being the most represented. This wide geographic coverage ensures diverse insights into the recreational use of natural pools across different regional contexts. A significant majority of respondents (67%) held diplomas as their highest educational qualification, followed by those with secondary education (16.7%), as shown in Table 2. This suggests that the sample had a relatively higher educational attainment, which could correlate with the higher awareness of sustainable practices in recreation. The majority (36%) of respondents were employed in the private sector, followed by farming, reflecting a blend of agricultural and private sector occupational engagement. This could also be due to the fact that farming is the key income source of the locals residing in the study area. The monthly income categories of respondents revealed that the largest segment of respondents earned between Sri Lankan Rupees (LKR) 45,001 to LKR 60,000 (32%), which could influence their discretionary spending on recreational activities.

The natural pool and its surroundings have been consistently frequented by a considerable portion of respondents, with 39.7% visiting ‘always’ and 29.7% indicating ‘sometimes’. This consistent engagement underscores the importance of these natural pools for recreational activities of the community (Table 2). The perception of the standard of recreational services was predominantly deemed ‘Neither good nor poor’ (40.3%) or ‘Good’ (36.7%). This indicates the need for improvement in recreational service standards, which aligns with the perception of the substantial majority (89%), who expressed a desire to enhance sustainable recreational services (Table 2). The community’s inclination towards external support for improving sustainable recreational services in natural pools was very much in favour of government support (79%), followed by community-based organisation support (12%). This finding suggests a trust in governmental initiatives and a potential underutilisation of community-based resources in recreational development.

The percentage preference of the respondents over different ecosystem services provided by natural pools are shown in Figure 2. Results indicated that the respondents in general highly prioritise ecosystem services provided by natural pools, as over 50% rated ‘very important’ to all the services. This demonstrates a strong public inclination towards environmental awareness and the necessity for sustainable practices. They tend to believe that maintenance of water quality, availability of recreational activities as well as community benefits to be important factors among the potential ecosystem services.

The association of ecosystem services is illustrated by the Principal Component Analysis (PCA) plot, as shown in Figure 3. This ordination biplot is based on the relative importance of services provided by natural pools as perceived by the respondents. The two-dimensional plane of the biplot captures the variance and trade-offs of ecosystem services. PCA1 and PCA2 explained a combined variance of 93.5% of the total variance. The variables, labeled A to L, correspond to different recreational perception-related indicators. The spread of the variables along PCA1 (88.9%), suggested that most variables are strongly associated with this component. Variables closer to the circumference of the circle are well represented in the PCA space. For instance, variable K which is community income generation closely aligns with PCA1, indicating it is a dominant factor in the dataset. Variable A (waste disposal and recycling methods) also extends farther in PCA1, indicating a significant influence. Variables J (security) and L (advertising) extend towards PCA2, which captures a relatively smaller portion of the variance (4.6%), suggesting these factors contribute uniquely to the dataset’s variability.

Factors D, C and E lie closer to each other, suggesting a possible correlation among those (Figure 3). Hence providing knowledge and awareness on the importance of natural pool conservation in the forms of education programmes can have significant implications in inculcating positive attitudes and values among users. This shows a clear spatial pattern of tradeoffs between different ecosystem services. Further, all variables were clustered in the positive quadrant of PCA1 indicating a strong association. Therefore, factors such as cultural benefits, spiritual benefits and adverstising seemed to be correlated indicating the potential for developing the natural pools in such a way to attract the visitors seeking such ecosystem services.

3.2. Marginal Willingness-to-Pay (MWTP) for Ecosystem Services

Table 3. Results of the Conditional Logit Regression.

Attribute	Levels	SE	Co-Efficient	Odds Ratio	MWTP
ASC		0.014	0.018	0.712
Environment damage reduction	20% reduction	0.456	0.568 *	1.080	113.6
	50% reduction	0.788	1.488 *	1.071	297.6
No of recreational activities	14 activities	0.762	0.748 *	1.312	149.6
	11 activities	0.352	−0.536	0.961	−107.2
Community income increase	10% increase	0.244	−0.826 *	0.813	−165.2
	20% increase	0.312	0.422 *	1.490	84.40
Visitors labor contribution per visit	1 h	0.058	0.028	0.988	5.66
	0.5 h	0.324	0.938	0.992	187.60
Willingness to pay per visit	-	0.058	0.005 *	0.961

Note: *—Coefficients that are significant at a 95% level of confidence; SE—Standard Error; MWTP—Marginal Willingness to Pay in LKR; log pseudolikelihood—(−556.46803); sample N = 320.

shows the main effect specifications identified using the CL regression along with standard errors, coefficients, p-values, and the MWTP values for each attribute level. Results show that in terms of environmental damage reduction, a 20% reduction level has a positive coefficient of 0.568, indicating a significant (p = 0.03) preference for this level of environmental improvement. Hence, the MWTP of 113.6 suggests that visitors are willing to pay a considerable premium for a 20% reduction in environmental damage. Meanwhile, the 50% reduction level has a larger coefficient (1.488), which reflects a statistically significant stronger preference. The MWTP for 50% reduction is higher at 279.6, which implies that visitors value a 50% reduction of environmental damage to have more environmental benefits. Where the number of recreational activities is of concern, the level of 14 activities shows a positive and significant co-efficient (0.748), which indicates a strong preference and statistical significance for a higher number of activities. The MWTP of 149.62 suggests that visitors might pay substantially more to have more recreational activities in the studied natural pool sites. In terms of number, the level of 11 activities shows a negative non-significant coefficient (−0.536), suggesting a lower number of activities, which might not significantly and necessarily detract from visitors’ utility (Table 3).

The CL model (fixed effects) assumes homogeneous preferences across respondents as difference in either the strength of the preference or as the direction (i.e., magnitude or sign of the coefficient). ASC in the model represents a composite of respondents’ underlying preference for choice alternatives. Respondents consider income improvement of the community to be significantly important. This fact is also confirmed from the results in Figure 3. Both income levels are significant compared to business-as-usual scenario. However, a 10% income increase has a negative significant coefficient (−0.826) suggesting a disutility associated with this level of income increase, with visitors potentially unwilling to pay more (MWTP of −165.2) or possibility willing to accept. Meanwhile, the 20% income increase shows a significant positive coefficient (0.422) denoting a strong positive utility for this level of income increase, with a high MWTP (84.40). This suggests that visitors might pay significantly more for these community benefits for their economic well-being (Table 3). Labour contribution of 1 h had a positive coefficient (0.028) suggesting a higher affinity for this labour contribution, which was non-significant. However, a significant 0.5 h contribution suggested a higher positive impact on the utility for 0.5 h of labour contribution, with a positive MWTP of 187.6. The significant positive coefficient (0.005) indicates that higher payment requirements per adult visit increase the utility for visitors, suggesting a willingness to pay more. An odds ratio of 0.98 for funds per adult per visit suggested that the odds of preferring or choosing a one unit increase in funds contributed per adult per visit are 0.96 times those of the no funds, indicating a lower preference for much higher contributions.

According to the results, community income increment by 20% and number of activities denoted the highest odds ratios. This shows that for example, when income increases, the number of activities improves by one unit compared to the status-quo option, the visitor utility increases and are willing to pay more for the benefit of the local community. For a 20% reduction in environmental damage, the odds ratio was 1.08. This indicates that the odds of this level being preferred are 1.08 times greater than the no reduction level. The 50% reduction reported an odds ratio of 1.07, which was 1.07 times greater than the no reduction level for 50% level of reduction. Preferences slightly favoured environmental damage reduction initiatives, with a marginal increase in the likelihood of choosing a 20% or 50% reduction, over the no reduction scenario (Table 3). Results revealed a total user MWTP of LKR 554.8 per year per respondent for the multiple benefits provided by the pool ecosystem to conserve the ecosystem for its multiple benefits provided.

4. Discussion

This study applied a Discrete Choice Experiment approach to evaluate the recreational value of visitors of natural pools in Sri Lanka. Recreational values are often classified under cultural values [54,55,56]. Among the multiple ecosystem values provided by the pool ecosystems, cultural ecosystem services are considered as passive use values as these are intangible in nature and can be improved to provide numerous social and economic benefits [54,57]. Results revealed that certain demographic characteristics of users are important factors to assess and understand the recreational value of amenities. The findings of the study are in line with the findings of previous similar studies, which employed popular stated preference techniques such as the Contingent Valuation method [58] and Choice Experiments [55] to explore freshwater ecosystems and their cultural values that provide economic and social benefits to the community.

Horcea-Milcu et al. has shown how cultural uses are strongly dependent on socio-economic characteristics of the users such as education and income, which are comparable to our findings [59]. A previous study conducted by Liu et al. further shows how user attitudes such as bequest motives and environmental consciousness are related to the proximity [60]. The majority of the respondents of this study were from close proximity, which may influence the strong positive coefficients of the choice experiment attributes [55]. According to the findings, respondents preferred a 50% reduction of environmental damage. Therefore, initiatives should target a moderate 50%reduction in environmental damage, as this level significantly enhances the MWTP and underscores the value of respondent’s perceptions on environmental conservation. This result complies with the findings of Wallmo and Lew [61], which explains how diminishing marginal utility is linked to water and environmental quality reduction of streams. Hence, a strategic approach focusing on key areas aligning with user preference is essential for natural pool conservation and development initiatives.

Enhancing the visitor experience by maintaining or improving the number of recreational activities is imperative, as this factor notably increases MWTP. In terms of community income derived from recreational activities, a nuanced approach is justified. A 20% increase in community income was highest positively received over a 10% increase, warranting thorough assessments to align community benefits with respondents’ expectations. Furthermore, reevaluating the necessity and extent of labour contributions is imperative. Tait et al. has also shown how WTP for improvement in swimming water quality can have a positive relationship with the employment and income generation [62]. Aforementioned studies have further emphasised that users are willing to pay more for alternatives with more job opportunities and improved water quality. This study also explored the funding structure, which could be a combination of direct payment as well as labour. Other forms of contribution for a sustainable management and provision of facilities may yield a more favourable reception [63,64].

The findings of PCA suggest that waste disposal, water quality, conservation education, recreational activities, facilities, and community income generation tend to have a strong influence on user preference for a destination, in this case natural pools. Individual and societal benefits such as physical, mental, social, spiritual, and cultural gains are interrelated and contributed more to the second principal component. Security and advertising denoted a moderate combined influence on both components. It suggests that efforts to improve or invest in these areas may positively affect the users of ecosystem services. Demographics show that the majority of users are from surrounding areas, therefore this may have a strong influence on the preference for community improvement.

The relative importance of recreation related factors suggests that factors related to various aspects of recreational perception are most influential among the younger groups (15–30 years old) of visitors and tend to diminish in importance with age. Community income generation factor, however, maintains relevance across a wider age range, suggesting that it is considered relatively more important. This heterogeneity suggests that planning for sustainable ecosystem management initiatives must cater to various age groups effectively. The number of responses for each importance level suggests that the efforts should focus on bolstering community income generation and environmental conservation initiatives, as these are clearly valued by the community. Enhancing recreational activities and their associated benefits should be undertaken with the aim of elevating their importance, perhaps by demonstrating their economic and social value. Several previous studies have emphasised that such improvements can improve public goods to provide more economic and social benefits to the community [60,61].

Results of the relative importance ranking show that it is imperative to consider conservation of natural pool ecosystem services in a way that improves recreational activities, which can enhance economic and social benefits to the community. Natural pools are ecosystems that provide multiple benefits. This study focuses more on the use values of the system. It is evident that through improvement of facilities and effective conservation methods, pool systems can be improved as sustainable revenue-generating entities creating direct and indirect economic benefits. A comprehensive strategy focusing on waste management, ecological conservation, education, and eco-friendly recreation is essential in this regard. Initiating robust waste disposal and recycling methods is paramount, involving the provision of well-marked waste segregation facilities and the collaboration with local waste management services to ensure responsible waste handling and innovative recycling initiatives. Ensuring water quality and soil health involves regular monitoring of water parameters, and the implementation of soil conservation strategies. Such planning must be done with the involvement of the local community for more sustainable freshwater ecosystem service conservation [64].

It is important to sufficiently enforce government regulations for effective environmental conservation as regulation acts as a strong economic incentive in this regard [60]. Willingness to pay of users for improving recreation and cultural values of freshwater ecosystems such as natural pools are comparatively similar to those by Wallmo and Lew [61]. These studies show how cultural values are usually rated relatively low, compared to other ecosystem services but in a medium range, which is comparable with our findings. These findings emphasise on instituting educational programmes aimed at raising environmental awareness, which encompass interactive learning experiences, hands-on experience in conservation activities, and collaboration with environmental experts to instill values of sustainability among visitors. Integrating recreational activities that promote conservation, such as eco-friendly nature trails and incentivised participation in environmental stewardship initiatives, will further embed an ethos of respect and responsibility towards the natural environment.

The findings of the current study revealed a higher willingness to improve sustainable recreational services of the studied natural pools among the local community. This could be caused by a variety of factors such as the higher education level and monthly income, followed by higher environmental awareness [36,38]. Further, positive attitudes of local communities on the importance of these natural pools for their income, health and wellbeing could also contribute to this tendency [41,42]. However, previous studies have suggested that this WTP for sustainable recreational services can vary significantly among different regions of the country due to several factors such as proximity to the freshwater ecosystem of interest, cultural differences across regions, local environmental conditions and regional income differences [38,39,65,66,67]. Further, studies are recommended to evaluate the regional differences among the WTP to improve sustainable recreational services and to estimate the maximum limits. This will be of immense importance in making policy decisions regarding the community based conservation of these ecosystems.

Natural pools also hold a significant importance for these types of non-use values such as existence and bequest values [68]. Enhancing water-based activities is beneficial, entailing safe swimming zones, lifeguard services, and the provision of eco-friendly rental equipment for activities. Integrating educational elements through informational signage and organised workshops or guided tours will enrich the visitor experience [69]. Engaging the community in the planning process and actively seeking visitor feedback will not only ensure that facilities are met but shall exceed user expectations, making the natural pools a sustainably managed revenue generating entity [26].

5. Conclusions

The study findings indicate that visitors of the natural pools place a significant value on higher economic improvements, high availability of recreational activities and environmental damage reduction. There is a clear WTP for these enhancements, as evidenced by positive coefficients and MWTP values. For example, the value attached to further improvements tends to be influenced by diminishing returns, as the MWTP for a 50% damage reduction is more than double that of a 20% reduction. When it comes to paying, visitors believe that the community should be made better-off through higher income generation. One of the reasons for this perception could be that they believe improved income generation could boost greater benefits, facilities and amenities for visitors as well.

The findings of this study revealed that users were also willing to pay through labour contribution. Hence, these findings can be effectively used to guide co-management based initiatives for freshwater ecosystem management, which are found to be sustainable in the long run. Users were also willing to accept a payment for certain income generation activities, which could be through direct cash or in-kind measures. Moreover, the significantly higher preference for environmental damage reduction exhibits the user affinity and inclinations towards conservation. This is an indication of the potential for implementing nature-based solutions for ecosystem conservation. If implemented, such programmes should be voluntary and perhaps incentivised with discounts or other benefits. It is essential to consider the negative utility of increased funds per adult per visit. The willingness to pay indicates the maximum price consumers are willing to pay for the ecosystem service, which helps to derive the demand for an environmental good or service. Therefore, such pricing strategies should be carefully evaluated to ensure that they do not become a barrier to entry or visitation. Considering tiered pricing, discounts for advance booking, or membership programmes that offer added value without significantly increasing the single-visit cost are few strategies can be recommended. Consumer surplus can be estimated as a measure to understand the gains and losses of price setting.

Further studies should be conducted to understand quantified trade-offs between conservation, water quality improvement and economic development of natural freshwater ecosystems such as natural pools. Multidisciplinary studies are warranted to understand how monetary valuation of these ecosystems can be integrated into national policy making. Understanding these aspects better could help in drafting policies that align visitor experiences with conservation and community goals. Strategies that can enhance community income are more likely to gain strong public support. Additionally, initiatives should focus on improving environmental conditions and providing education on conservation, as these are also highly valued. In contrast, while still important, spiritual and cultural benefits are seen as less immediate priorities and could be integrated into broader programmes that primarily address the more pressing economic and environmental concerns.