© M. Mamboleo and A. Adem, Published by EDP Sciences 2023

1 Introduction

Humans rely on wetland ecosystems to provide goods and various ecosystem services (Larsen, 2017). Ecosystem services have been defined as the benefits people obtain either directly or indirectly from ecological systems (Costanza et al., 1997; de Groot et al., 2002; Millennium Ecosystem Assessment, 2005). Wetlands provide habitat for both plants and animals in addition to assisting in the support of human lifestyles that depend on their different biological resources. They do this by supplying numerous goods and services and performing various functions (Brouwer et al., 1999) and are regarded as a source of goods and services. Wetland ecosystems provide a variety of commodities and services on which human groups rely heavily, either directly or indirectly, for their survival (Costanza et al., 1997). However, many wetlands have traditionally been deemed wastelands and have been degraded through their conversion for diverse human uses, including agricultural, industrial, and residential ones (Barbier et al., 1997). Numerous scientists have investigated beneficiaries' Willingness to Pay (WTP) for environmental protection in watershed regions using the Contingent Valuation Method (CVM) (Ryan and Watson, 2009). The multiple valuable services produced by these ecosystems are fast degrading owing to a lack of effective conservation models and incentive mechanisms for stakeholders involved in maintaining such ecosystems (Costanza et al., 1997; Karami et al., 2020). A rising scarcity of services, along with increased strain on ecosystems, has resulted in a rise of studies focused on the Willingness to Pay concept (Wunder et al., 2008). The efficient use of natural and environmental resources is dependent on their worth, which may be quantified through economic valuation. The value of water resources may be calculated based on the value of the goods or services that they contribute. Several economic valuation methodologies may now be used to estimate the economic worth of natural ecosystems and inhabitants' WTP for specific ecosystem services.

According to Martin-Lopez et al. (2008), the CVM is the most extensively utilized economic valuation technique to estimate the economic worth of diverse ecosystems. The CVM is built on the ideas of WTP, or readiness to take compensation to improve utility because of damage to a public good. In the absence of a true market, this strategy presents individuals with a hypothetical option to purchase public goods. According to Reynisdottir et al. (2008), both the rational choice and the utility maximization theories underpin the CVM's WTP for non-market goods. CVM is a technique that allows individuals to buy public goods in hypothetical contexts, particularly when there is no genuine market or current knowledge about a real-world issue. In CVM, surveys are used to obtain information on people's maximum WTP for biodiversity protection and conservation (Loomis and White, 1996). The maximum amount that residents or visitors agree to pay is referred to as the WTP (Kyle et al., 2002). WTP is used to assess the value of non-market items, usually linked to socio-economic, institutional, or educational factors (Reynisdottir et al., 2008).

The income is one of the socio-economic factors that has been popular in most CVM studies of residents' WTP. For a long time, the impact of having money on WTP has been hotly debated, but a definitive answer to the question has yet remained elusive. Several studies, however, have revealed that low-income earners are more sensitive to tax increase than high-income ones (More and Stevens, 2000). In other words, regardless of the goal, residents' WTP is mostly determined by their income level (Reynisdottir et al., 2008). As a result, it is plausible to suppose that residents' income has a favorable influence on their commitment to protect their surrounding environment, including biodiversity conservation. Institutions are another important component that influence peoples' WTP. Many studies have shown that environmental views are a major predictor of WTP (Kotchen and Reiling, 2000). On the other side, attitude toward fee policy or perceived fairness has been highlighted as one of the main predictors of WTP (Rosenberger et al., 2012). Before establishing or updating any fee-paying policy, it is critical to evaluate these variables. This will assist with the identification of the key market segments among people who may be able to donate more cash for biodiversity conservation.

Contingent valuation method (CVM) is a non-market valuation method that requires individuals to indicate the maximum amount they are willing to pay for the use or preservation of natural commodities, with the idea that their decisions are contingent on alternative things offered in a hypothetical market (Mamboleo, 2021). This strategy has been proved appropriate for evaluating environmental items that lack market data but have an influence on the respondents' well-being and is commonly used for cost-benefit analysis and for assessing the environmental effect of non-marketable resources (Fogarassy et al., 2016). Several studies have been undertaken to estimate WTP for biodiversity conservation in wetlands and other natural attractions using CVM. WTP seems to vary according to countries and type of wetlands and is determined by the population's socioeconomic condition and ecological knowledge. This technique has been used to determine the worth of wetlands (Siew et al., 2015), renewable energy sources (Botelho et al., 2018), forests (Amirnejad et al., 2006), and other non-market resources (Ramajo-Hernández and Saz-Salazar, 2012). CVM key benefit is that it may be used for a wide range of evaluative circumstances, as being independent of real market or observable behavior (Pearce and Moran, 1994).

Lake Victoria is the largest freshwater lake in Africa and the largest tropical lake in the world. The lake is one of the most important freshwater ecosystems in East Africa's biodiversity-rich tropical environment. Apart from providing internal maritime transit and energy, the lake also supports fishing, agriculture, nature, tourism, trade, and wildlife. The Lake Victoria Basin has suffered from ecological deprivation, which has coincided with the rapid growth of settlements surrounding the lake (Fig. 1). As a result, conservation is essential for the long-term survival of biological resources as well as the maintenance of livelihoods and wetland management that is sustainable to limit the rate of biological resources loss and deterioration.

Anthropogenic influences on fisheries, wetland buffer zones, and water quality are causing worry about the health of the Lake Victoria ecosystem. Changes in fish assemblages, deterioration of wetlands, changes in water quality, and the invasion of the water hyacinth Eichhornia crassipes have all placed Lake Victoria's coastline area in jeopardy (Masifwa et al., 2001; Twesigye et al., 2011). Despite its apparent relevance to ecology and biodiversity, substantially less is known about the value of the Lake Victoria ecosystem than just the open lake. Lack of specific understanding adds to the uncertainty generated by fast population expansion, which puts more strain on wetlands and water supplies, contributing to additional changes in Lake Victoria's environment.

Kenya is classified as a water-stressed country by the United Nations. As a result, Kenya regards its wetlands as significant resources for local people in terms of ecosystem services and habitat for coastal and aquatic animals. However, even though wetlands are essential resources for livelihoods and biodiversity conservation, little research on wetland services is being performed in Kenya. Moreover, a lack of appropriate data on publicly owned private firms typically hampers planners and government officials from implementing policies and connecting with environmental stakeholders. As a consequence, Lake Victoria's supplies are depleting. Adopting proper policies to aid in the creation and allocation of greater resources for community-based wetland conservation initiatives can assist wetland managers and policymakers. In Kenya, the lake's resources benefit local communities in Busia, Homa Bay, Kisumu, Siaya, and Migori. According to a recent study, the residents of these local communities are Willing to Pay an average of KES 500, even though many factors influence their WTP (Mamboleo and Adem, 2022). The present paper relies on these previous results to further analyse the various factors influencing fishermen, traders, and homeowners' WTP. According to previous studies, individual's WTP may depend on age, education, gender, and financial situation (Choi and Fielding, 2013; Hultman et al., 2015).

Fig. 1 Lake Victoria Basin, Kenya land use map.

2 Materials and methods

This study was conducted along the whole Kenyan Lake Victoria shoreline in East Africa. Primary data was acquired in the study region from a 2021 residents survey (Mamboleo and Adem, 2022). The questionnaire was developed to collect demographic and social information from the survey inhabitants. This research approach is appropriate since it entails delivering a questionnaire to a sample or the whole population of people in order to characterize the attitudes, views, behaviors, or qualities of the population. The study's data was gathered using a standardized questionnaire (see Mamboleo and Adem, 2022 for further details). The first piece of the questionnaire collected respondents' demographic and socioeconomic data, while the second section solicited information on the likely components that impact the respondent's WTP for a conservation program (Mamboleo and Adem, 2022). The questionnaire was carried out in the counties of Busia, Homabay, Kisumu, Siaya, and Migori.

The Kenya 2019 Census estimated the population of residents within the Migori, Homa Bay, Busia, Kisumu and Siaya counties as 5,290,824 with an average of 4 people per household (KNBS, 2019). Figure 2 shows the population pyramid within the project area.

Equation (1) was used to get the sample size for this research.

$$n=\frac{z^{2}p\left(1-p\right)}{e^2\left(1+\frac{\left(z^{2}p\left(1-p\right)\right)}{N{e}^2}\right)}\text{.}$$(1)

In this equation n represented the sample size, while N was the population size in the study area (5300000). Z was 1.96 for a confidence level (α) of 95%, p represented the proportion (0.6), while e was the margin of error (0.05). Therefore, the sample size was found to be (with finite population correction) 369. The binary logistic regression model, described in equation (2), identifies the parameters influencing households' WTP (Gujarati, 1999):

$$Pi=F\left(\delta +\beta x_i\right)=\frac{1}{1+{\delta }^{-\left(\delta +\beta x_i\right)}}\text{.}$$(2)

In this equation the index i means the ith observation, Pi is the possibility of someone making a particular choice, given that X_i (2.718), is the basis of natural logarithms, while δ and β are the model parameters. To assess how each of the explanatory factors link to a dependent variable, a chi-square analysis was employed. The in-person questionnaire was acceptable primarily because the topics at hand are somewhat professional. From a total of 1,232,558 households in the study area (KNBS, 2019), 394 households were interviewed (25 more than the number calculated in Eq. (1)). This was done to ensure that the precision of the estimate corresponded to the width of the confidence interval for the cumulative average. To guarantee that the sample gathered appropriately reflected the lake's economic activity, beaches were sampled using targeted sampling based on the area's economic activity.

Fig. 2 Project area population pyramid. Source: Based on self-calculations adopted from (KNBS, 2019) census report.

3 Results and discussion

3.1 Socio-demographic factors

The socio-demographic characteristics that were captured included county of origin, Gender, Age, Marital status, educational level and household size. The samples picked per county were based on the proportionalities to the shoreline coverage by the county. Results from Table 1 show that Homa Bay County had the highest representation (43.4%) of all the respondents based on the fact that it has the longest shoreline. Busia County had the smallest number of respondents comprising only 8.4%. In terms of gender, majority were male (61.4%). 78% of household heads were over the age of forty. 82.5% of the respondents were married while only 17.5% of the household heads were either single or widowed.

From this study it was clear that the majority of household heads (43.9%) beneficiated from a secondary education level. Other levels of education were Primary school (35.5%), Diploma (13.5%) and others (3%). Less than four-person households (61.42%) outnumbered those with more than four people. This was a clear indication that family planning was making significant progress in the region. These respondents were engaged in the following economic activities: fishing, farming, trade, livestock keeping, sand harvesting and formal employment (Tab. 2).

Results from Table 2 show that 44.9% of the respondents were involved in fishing, 25.6% were doing small scale farming along the lake, 14% were engaged in small scale trade and 9.4% were involved in livestock keeping. The respondents were further asked to give their perceptions on the current problems they face and faced by the lake' ecological system. The greatest concerns of the respondents were the decline in income as reported by 64.2%, the decrease in fish catches (44.9%), water degradation (42.4%), siltation (33.9%) and the decline in biodiversity (25.1%). These results are presented in Figure 3.

Finally, the respondents gave their opinion on the level of importance attached to ecological conservation with 90.1% reporting that it is either very or quite important to conserve the environment, as shown in Table 3.

The socio-demographic aspects influencing WTP included county of origin, gender, age, marital status, level of education and household size. The Chi square association test was performed to test the hypothesis that “there was no significant relationship between WTP and socio-demographic characteristics”. A confidence level of 95% was used in this study. A chi square test p-value of less than 0.05 leads to rejecting the null hypothesis. The results obtained are presented in Table 4.

The socio-demographic parameters impacting significantly WTP were identified using multilevel logistic regression (Tab. 5). According to the model socio-demographic factors accounted for 26.3% of the total variation in the WTP, with respondents' county, age, and educational level significantly influencing WTP (Tab. 5).

Keeping all other factors constant, the proportion of Siaya residents who were willing to pay for a Lake Victoria ecosystem conservation program was higher compared to the other counties. Furthermore, a resident over the age of 40 was 1.829 times more likely to pay than a resident under the age of 40. A previous study conducted by (Acquah and Abunyuwah, 2011) in Ghana, also obtained a similar finding. Finally, when all other characteristics were held constant, the results revealed that a resident with a diploma was the most likely to pay (5.089) compared to others.

The results showed that the participation of older respondents in the water conservation program of Lake Victoria was higher than that of young respondents. A possible reason may be that most elders may be interested in conserving the water resources of the Lake Victoria ecosystem because they wanted this resource to be available for their descendants. The study also noted that older people were probably more closely associated with their villages and traditions, making them willing to pay more than younger ones. Another likely cause may be that fishing activities that bring most of the economic benefits from the waters of Lake Victoria are carried out by older people. The dependent variable and literacy rate were shown to have a positive association. Those residents with a diploma had the highest regression coefficient of 1.627 and a Wald coefficient of 18.331. Many studies have found that education is an important factor of WTP (Baral et al., 2008). WTP has been demonstrated to be positively connected to a greater degree of education. This is due to the fact that educated individuals are more aware of environmental concerns and thus more inclined to participate in conservation actions (Brennan et al., 2007).

Fig. 3 Most serious problems along Lake Victoria.

3.2 Socio-economic factors

Several elements were evaluated to identify the socio-economic aspects impacting WTP. These included fishing, lakefront property ownership, lakefront farming, access to grazing area, sand harvesting, small trading around the lake, business ownership, and average monthly income. The Chi square test of independence was employed to see if there was a link between socio-economic characteristics and WTP. The null hypothesis was that socio-economic conditions and desire to pay are independent. Table 6 shows that fishing status, land ownership, access to grazing land, business ownership, and household monthly income have a statistically significant connection with WTP.

Table 6 reveals that people who fish along the lake, own land next to the lake, have access to grazing area next to the lake, own enterprises close to the lake, and have an average income of more than KES10,000 are more likely to pay for the program than their counterparts. This means that people who profit directly or indirectly from the lake are prepared to pay for the program, as opposed to those who do not have access to any form of service or commodity from the lake. To determine the socio-economic factors significantly influencing residents' WTP, a logistic regression model was fitted to the data. Table 7 shows that socio-economic factors accounted for 28.2% of all the variations in the WTP. The results show that involvement in fishing, involvement in farming next to the lake, access to grazing land, monthly income and business ownership have statistically significant influence on the WTP.

Those who are not involved in fishing, not involved in farming, don't have access to grazing land and those who don't own businesses are respectively 0.308, 0.388, 0.335 and 0.318 times less likely to pay as compared to their counterparts. On the other hand, those who earn more than 10,000 are 3.165 times likely to pay compared to those earning less than 10,000.

The findings show that households that fished, owned land on Lake Victoria's periphery, owned grazing land, had a business near the lake's periphery, and had a high average income were more willing to pay than their counterparts because they were expected to be more involved in activities that directly used Lake Victoria water than their counterparts. A similar tendency was noticed by (Johnson et al., 2008), who stated that water resource management decisions might be dependent on the location of agricultural land or settlements. Households that rely on Lake Victoria's water as their primary supply of water for fishing and other economic activities (in our case grazing) were more ready to pay than those who were not active in agriculture. In a study evaluating Norfolk Broads (Bateman et al., 2000) observed that respondents' socio-economic conditions and distance to the site provide consistent response in terms of respondents to the survey questionnaire, response to the willing to pay principal question, and in determining the amount of WTP. Theoretically, it is expected that families involved in agricultural operations near the Lake Victoria shoreline derive the greatest economic advantage from the utilization of water resources, and hence their chance of payment will be greater than that of their counterparts. The findings are congruent with a conceptual model developed by Leeworthy and Bowker (1997), that describes the relationships between economy and the environment. The model argues that real circumstances on wetlands in terms of the quality and quantity of water resources are crucial elements in affecting human perception of conservation. This explains why the amount of demand for the economic worth of water resources from the standpoint of agriculture has a considerable impact on a person's impression of the lake's value.

4 Conclusion

This research sheds light on the impact of socio-demographic and socio-economic factors on WTP and the use of WTP in sustainable funding to create market-based conservation techniques in under-developed nations to reduce ongoing biodiversity losses. The study used a Logit regression model to examine the impact of socio-economic characteristics on citizens' WTP for a Lake Victoria conservation initiative. The study found that socio-demographic factors accounted for 26.3% of all the variations in the WTP while socio-economic factors accounted for 28.2% of all the variations in the WTP. The study shows that residents' WTP was influenced by their ability to fish along the lake, own land close to the lake, access grazing area adjacent to the lake, own enterprises near to the lake, and their income. County of Origin, gender, age, marital status, level of education, and household size also greatly influenced a resident's WTP for ecosystem conservation program. This paper adds to the current non-market valuation literature by assessing factors influencing the residents' WTP to promote the development of conservation programs, especially where money for conservation is scarce. This study work might serve as a useful reference for future valuation studies in wetland ecosystems in Kenya and other developing countries. Furthermore, education has been demonstrated to be positively related to WTP for conservation efforts. As a result, media-based public education, and the inclusion of environmental management courses in school curriculum can be adopted. This will aid in adequately informing residents about the importance of caring for their environment and using the benefits of water ecosystems. The study further recommends that before giving any policy recommendation aimed at making water resources sustainable for the lake or any water bodies in Kenya, the effects of income distribution among the households on WTP need to be critically studied.

References

All Tables

All Figures

	Fig. 1 Lake Victoria Basin, Kenya land use map.
In the text

	Fig. 2 Project area population pyramid. Source: Based on self-calculations adopted from (KNBS, 2019) census report.
In the text

	Fig. 3 Most serious problems along Lake Victoria.
In the text