Payment for Urban Mangrove Forest Conservation in Vietnam: A Community Case Study of Can Gio Biosphere Reserve, Ho Chi Minh City

Abstract

Mangroves play an important role in providing livelihoods, ecological services, and values to communities in Vietnam. However, mangrove forests are being significantly reduced due to socioeconomic development. This study assessed the awareness of mangrove forests and willingness to pay communities in the buffer zone of Can Gio Biosphere Reserve, Ho Chi Minh City for mangrove conservation. The model used was the contingent valuation method, through surveying 700 households in five communes in the buffer zone of the Biosphere Reserve. The results show that local people have a fairly good awareness of the values of mangroves and are willing to participate in conservation activities for the benefit of their families and future generations. The average WTP of a household is VND 529.6 thousand/year (USD 22) in the parametric model and VND 597.8 thousand/year in the non-parametric model (USD 25). Factors that significantly affect WTP include household income, number of people in the household, education and payment level. The study results provide implications for forest resource management activities in Vietnam, especially the financial mechanism for mangrove conservation with the participation of the community.

1. Introduction

Mangrove forests are one of the most important resources for the existence and development of several communities worldwide [1,2]. Mangroves provide many benefits and values to the local communities and society. These are direct benefits, such as livelihood promotion, providing fuel wood, fisheries, medicinal herbs and tourism, and indirect values such as disaster prevention, CO₂ absorption, filtration and regulation of groundwater, or non-use values such as heritage values, culture, and biodiversity conservation [3,4]. However, mangrove forests have followed a significant trend of decline worldwide in recent years; one of the reasons is that people have not fully appreciated the economic value of mangroves, leading to the use of resources. This resource usage is ineffective and unsustainable for the local communities in the present as well as in the future [5,6,7].

In Vietnam, mangroves occupy a vital position in the national territorial structure, having particular functions in the country’s development, and are a source of livelihood for several local communities [8,9]. Vietnam has a coastline of 3250 km with many estuaries leading to the sea, so there is a relatively large area of mangrove forest. However, mangroves in Vietnam have been significantly degraded over the past 20 years due to economic development and poor management. It is estimated that in the period 2000–2020, about 200,000 hectares of mangroves were destroyed [9]. Although the importance of mangroves is recognized in many conservation policies and regulations, they are still being destroyed at a remarkable rate due to the activities of agriculture, construction, transport, tourism and urbanization. In addition, due to the impact of natural disasters in the context of climate change, mangroves are also being significantly degraded, thereby causing a decrease in the quantity and quality of ecological services and economic values provided to local communities [10,11].

In Ho Chi Minh City, the largest city in Vietnam, the mangrove forest in Can Gio Biosphere Reserve (CGBR) plays particularly important roles in areas such as disaster prevention, air conditioning, groundwater filtration and regulation, and CO₂ absorption. Mangrove forests in Can Gio account for nearly 30% of the city’s natural area and bring substantial economic and ecological benefits to people and communities [12,13]. However, urbanization and industrialization are seriously threatening the mangroves here. The conservation of mangroves in Can Gio plays an important role in the city’s development. The city has recently implemented a number of measures to mobilize people’s participation in mangrove conservation.

Around the world, there are success stories about effective management of mangrove resources, maintaining ecological values and supporting sustainable livelihood development for local communities [9,11]. According to [14], information on the economic value of mangroves is an important input in the design and implementation of solutions to conserve and sustainably develop this resource. To estimate mangrove economic values, many approaches were employed, of which the contingent valuation m (CVM) is one of the most commonly used methods, especially in quantifying the conservation and biodiversity values of mangroves [15,16,17]. In CVM, a survey will be conducted to estimate the willingness to pay (WTP) of individuals for the conservation and improvement of environmental values, with hypothetical scenarios [17,18].

To date, several studies employing CVM with WTP have been conducted to measure the economic benefits of mangrove protection and conservation [18,19,20]. However, there have not been many studies on community payments for mangrove conservation in Vietnam or assessing the factors affecting WTP levels. Current research studies are limited to marine protected areas (MPAs) or mangroves in south-central Vietnam [20,21] and estimated the mean WTP of visitors to conservation programs in Nha Trang MPA. Ref. [22] also estimated the WTP of households for mangrove restoration in an estuary in central Vietnam. They point out that interviewees’ attitudes toward climate scenarios are important factors influencing household WTP.

The main purpose of this study was to estimate economic values of urban mangrove conservation using the CVM model, and examine the factors affecting WTP for mangrove conservation at CGBR in Ho Chi Minh City. This is one of the first studies in Vietnam attempting to estimate interviewees’ WTP for mangrove conservation and examine how socioeconomic conditions affect their WTP levels. This research can advance the implementation of mangrove restoration and conservation strategies in the context of climate change, such as the REDD+ programs, where information on the economic value of mangroves is essential for cost–benefit calculations of climate change response scenarios at the regional and community levels. In addition, it also makes important comments on ecosystem service payments (PES) for the development of carbon trading markets and provides guidance on mangrove management for sustainable financing mechanisms in CGBR and Vietnam.

2. Methodology and Data

2.1. Valuation Model for Mangrove Conservation

Up to now, many variations of CVM have been implemented worldwide, such as continuous CVM, auction game, payment card and binary CVM (dichotomous). However, the binary CVM approach is most commonly used because of its rigor in terms of the theoretical basis and empirical model and its ability to reduce biases when conducting research investigations [14,23,24]. This study, therefore, used the binary CVM model to estimate the conservation value of mangroves in the study area. This method is theoretically based on the random parametric model of Haab and McConnell (2002) [25].

Random Utility Model

Let V_ij be the utility of the jth household for the conservation (improvement) of mangrove forest under option i, where i = 1 is the improved water quality, and i = 0 is the status quo. V_ij is a function of the properties of the benefits from improvement and the socioeconomic characteristics of the household:

V_ij = V_i (M_j, z_j, ε_ij)

(1)

where M_j is income of household j, and zj is a vector of the socioeconomic characteristics of the household and properties of the mangrove forest. ε_ij. is random error. The binary question requires interviewees to select between improving environmental quality or keeping the status quo at the cost of annual payments.

To measure the welfare of market participants, a logarithmic utility model is used. The random utility model with a linear income function assumes that the marginal utility of income is fixed in all of the alternatives, while the logarithmic utility model allows this variable to change as real income changes [24,26].

The probability that the answer is “Yes” for the proposed “improved” scenario is calculated as follows:

$$\mathrm{Pr}(ye{s}_j)=\mathrm{Pr}(\mathsf{\alpha }{z}_j- \mathsf{\beta }{t}_j+{\epsilon }_j> 0)$$

(2)

$$\mathrm{Or} \mathrm{Pr}(\mathsf{\alpha }{z}_{\mathrm{J}}- \mathsf{\beta }{t}_j+{\epsilon }_j> 0)=\mathrm{Pr}(-(\mathsf{\alpha }{z}_{\mathrm{j}}- \mathsf{\beta }{t}_j<{\epsilon }_j)$$

(3)

Assuming that the random variable ε_j is normally distributed with mean 0 and variance σ², then the distribution function of the answer “Yes” is:

$$\mathrm{Pr}\left(ye{s}_j\right)={\left[1+\exp \left(-\left(\frac{\alpha z_j}{{\sigma }_L}-\frac{\beta t_J}{{\sigma }_L}\right)\right)\right]}^{-1}$$

(4)

The portion $\ln [\left(Mj-tj\right)/Mj]$ is called standard income. The parameter vector of the model can be estimated by running the binary model, allowing the average value of WTP to be calculated as:

$$E_{\epsilon }\left[WT{P}_j\right]=M_j\left[1-\exp \left(-\frac{\alpha }{\beta }{z}_j+\frac{1}{2}\frac{{\sigma }^2}{{\beta }^2}\right)\right]$$

(5)

Non-parametric estimation model

The non-parametric model for estimating the mean value of WTP levels for mangrove conservation is also used in the study following the procedure of [25], which gives the Turnbull lower bound estimate of WTP. In the case where the assumption of monotonicity is violated, the pull-compensated estimator (PAVA) is used [26].

2.2. Data Collection

2.2.1. Focus Group Discussions (FGDs)

In order to develop a questionnaire suited to the research conditions, two focus group discussions were implemented in May 2022 in the study area.

The first FGD was conducted with state managers and local environmental managers (Can Gio district), specifically including: the Department of Agriculture and Rural Development, Department of Natural Resources and Environment, Department of Land, and Management Board of CGBR. The purpose of this discussion was to provide a platform for managers to discuss problems related to local mangrove protection and the pressures and threats to mangroves in Can Gio, as well as difficulties in mangrove management. Managers were also provided with a draft of the questionnaire for comments and corrections.

The next FGD was conducted with 15 households in 5 buffer zone communes. Those households mainly have their livelihoods based on mangroves. In the FGD, the people were asked about their conservation attitudes and perceptions of the ecological value of mangrove forests in Can Gio, identifying threats, providing initial payments (bids) and suggested means of payment, as well as WTP and non-WTP reasons.

2.2.2. Pre-Test

In order to increase the skills of the enumerators and continue to collect more feedback from the people to complete the questionnaire, the study conducted a pilot survey in all 5 communes in the study area. The staff of the Environment Department together with other investigators went to the interviewees’ homes to assess the actual conditions, advantages and problems during the actual investigation. From there, feedback channels were used to adjust the questionnaire in terms of format, questions, wording, and order of questions. There were 30 local people involved in the pilot survey at the site.

In this survey, collected bid levels were used. In addition, in order to allow people to speak for themselves, surveyors also asked more open-ended questions about the willingness to pay. This resulted in 9 bid levels, of VND 100,000, 200,000, 300,000, 400,000, 500,000, 600,000, 700,000, 1,000,000 and 2,000,000, for each household to pay annually for mangrove conservation.

2.2.3. Payout Levels (Bids)

In the 9 Bid levels used in the FGD and pilot phase, 6 were chosen to be used in the official questionnaire. According to [23], with the binary CVM method, approximately 6 to 8 bid levels should be used, of which the highest is the level that only about 10% of people can agree to pay [27]

From the FDGs and the pilot investigation, the cumulative rate of 6 bids, of VND 100,000, 200,000, 300,000, 400,000, 500,000 and 600,000, accounted for 88% of the total choices. Although there were other levels that people were willing to pay, they were very small and thus not used in the official survey.

2.2.4. Sampling Selection

According to Can Gio District data (2021), in 5 communes of the buffer zone, there are about 60,000 people, with an average of about 5 per household. So, there are about 12,000 households in the Can Gio Reserve area. The sample size was calculated with the following formula:

n = N/(1 + N × e²)

where: n is the sample size

N is the total number of households in Can Gio District

e is the acceptable error level.

From the above formula with 5% error, the sample size to ensure reliability was 657 households. The study collected 700 questionnaires with 140 questionnaires each, from 5 buffer zone communes, namely An Thoi Dong, Binh Khanh, Long Hoa, Ly Nhon and Tam Thon Hiep. To select households in each commune, a list of households was provided by the Can Gio District People’s Committee, then households were randomly selected from these resident lists. Interviews were conducted in the evening when the heads of households were usually at home. When one household was absent, an adjacent household was selected for the interview. The formal survey was conducted in July and August 2022.

2.2.5. Questionnaire

The research questionnaire had four main parts. First, the introduction presented by interviewers explained the purpose and confidentiality of the data collected; also, part 1 provided the interviewees with an overview of the CGBR and the value of mangrove forests, and at the same time asked people about their awareness of and attitudes toward conservation of the ecological value of mangroves, as well as their participation in the conservation management process.

Part 2 included questions about people’s WTP to preserve mangroves. First of all, with the help of mangrove experts, one conservation scenario was developed and introduced to the people. The villagers were then introduced to the threats to mangrove conservation. If these risks are not prevented, mangrove forests in Can Gio will continue to decline. Therefore, there is a need for a mangrove conservation program with the participation of the people. With the conservation program, mangroves will be maintained, restored and developed.

After presenting the conservation scenario, people were asked if there was a certain level of WTP for mangrove conservation. WTP levels were randomly selected from the 6 levels determined after the pilot investigation. The vehicle for the contribution was a local environmental fund. Questions to identify reasons for payment and non-payment were presented immediately after the question about WTP. The study also divided the interviewees into two groups. Group 1 was shown pictures of the value of the Can Gio mangrove forest before answering the WTP question (n = 350); the second group was not shown pictures of Can Gio (n = 350). The study expected to determine the impact of providing images on the community’s WTP.

The last part of the questionnaire contained socioeconomic information for the households.

2.2.6. Biases and Treatment

According to [23,27], one of the challenges in implementing the CVM is the presence of biases. To minimize bias, we applied some methods and procedures as follows [27].

To minimize strategic biases, interviewees were given a detailed explanation of the purpose of the study, especially with respect to their participation in conservation activities with the potential for restoration and development of mangroves in Can Gio, where they live. The interviews did not involve measures that may have negatively impacted the interviewees’ family life. Thus, interviewees were more open and honest with their answers. In addition, the interviewers were selected from the local community so that people would feel familiar and able to share [23,27].

To minimize and eliminate the starting point bias, we used the binary CVM technique. This type of CVM both creates simplicity for the answer and helps to eliminate starting point deviations because people are not familiar with how to participate in the hypothetical market and offer unsuitable payment levels. The binary CVM selected the bid levels obtained through the FDGs, as these were expected to better reflect the community’s WTP range.

The binary CVM technique was applied to eliminate the starting point bias. Binary CVM requires a detailed experimental research process, including focus group discussion and field trials to identify and adjust the WTP range, thus integrating this band into the real study. The tested WTP range and the Yes/No binary question helped to reduce starting biases.

Information bias and hypothetical bias were minimized through the design of user-friendly questionnaires, using visual images, and close-by local information on mangrove illustration. The questions were designed in a way that people could easily answer but still ensure the science of the model, checked by experts and tested in informal surveys. In addition, we also used the “cheap talk” (for short) technique, which prompted interviewees to balance their actual socioeconomic factors before answering yes or no to a certain bid level given.

2.3. Empirical Model Estimation and Hypotheses

The empirical model estimates WTP as the probability function of willingness to pay for mangrove conservation (

Table 1. Description of variables in the binary CVM model.

Variable Name	Meaning	Coding
Pr (Yes)	Probability of WTP a certain bid level for mangrove conservation	Yes WTP = 1 No WTP = 0
BID	Bid levels	Get the values 100, 200, 300, 400, 500, 600
EDU	Education (years)
MEMBER	Number of people in the household (person)	Continuous variable
INCOME	Income of household (VND/year)	Continuous variable
AGE	Age of interviewee (age)	Continuous variable
GENDER	Interviewee’s gender	Male = 1 Female = 0

):

Pr (Yes) = a₁ + b₁ BID + b₂ EDU + b₃ MEMBER + b₄ INCOME + b₅ AGE + b₆ GENDER

From the theoretical model and experimental studies, the following hypotheses were proposed to be tested in this study.

First, there is a positive relationship between the WTP level and the following factors:

• Income: households with high incomes will tend to pay higher prices for mangrove conservation
• Education: People with a higher level of education tend to have a better understanding of the environment and mangroves, resulting in higher payments.
• Age: According to some available studies, older people are more financially stable and often WTP to pass on the value of mangroves to future generations.
• Gender: Men are often in charge of the family’s economic affairs (mangrove related) and often have a higher awareness of conservation value and often higher WTP.

Second, there is a negative relationship between willingness to pay and the following factors:

• Bid level: The higher the bid level is set, the lower the probability of agreeing to that payment.
• Household size: When a household has more members, it costs more to spend on other primary activities and reduce spending on additional ancillary activities (consumption of environmental quality).

Additionally, according to [12], one of the technical problems to be dealt with in estimating the CVM is the protest bid, in which interviewees are unwilling to answer the WTP question because of objection to some aspect of the survey or presented scenario. In this study, the debriefing question was followed by the WTP question to identify protest bids, in which those who chose the following options would be considered unwilling: (i) I will not pay because I think the environmental fund will not work; (ii) I will not pay because I am afraid the money will not be used for the right purposes for conservation; and (iii) I will not pay because I see that the State has to pay for conservation, not the people.

3. Results

3.1. Study Site

Located nearly 40 km from the center of Ho Chi Minh City, Can Gio Mangrove Biosphere Reserve has a total area of more than 75,000 ha, of which the core area is 4721 ha, the buffer zone is 41,000 ha, and the transition zone is 29,000 ha. This reserve has populations of terrestrial and aquatic fauna and flora, formed in the large delta of the Dong Nai, Saigon, Vam Co Dong and Vam Co Tay rivers [22] (Figure 1).

Can Gio encompasses diverse habitats including mangroves, wetlands, salt marshes, mud flats and sea grasses. The ecosystem functions as the “green lungs” of Ho Chi Minh City, absorbing carbon dioxide and other polluting agents on a daily basis. The mangroves contain a high diversity of floral and faunal species. These include mangrove species such as Rhizophora apiculate, Thespesia populnea and Acanthus ebracteatus. A high number of faunal species appear in the biosphere reserve, including king cobra (Ophiophagus Hannah), saltwater crocodile (Crocodilus porosus), spot-billed pelican (Pelecanus philippensis) and fishing cat (Felis viverrina) [21,22].

Topographically, the Can Gio mangrove forest forms a basin with a minimum altitude range of 0–1.5 m in the northeastern sector of the forest, with downward inclines from the east, south, and west. The climate has two seasons: the rainy season from May to October and the dry season from November to April [22].

Before the Vietnam war (1955–1975), Can Gio was originally a mangrove forest with a rich flora and fauna. This place was destroyed by bombs and poisons during the national resistance war. Until 1978, when it was merged into Ho Chi Minh City, the People’s Committee of the city took practical actions to restore 31,000 hectares of natural forest.

Can Gio mangrove forest has very special environmental conditions; it is an intermediate ecosystem between aquatic ecosystems with terrestrial ecosystems, freshwater ecosystems and saltwater ecosystems. The mangroves here receive a large amount of alluvium from the Dong Nai River. Along with the influence of the adjacent sea and tidal currents, the ecosystem here is very rich. In terms of plants, the mangrove forests are mainly white cork, white acacia, double mangrove, guava and flora in brackish water such as sour cork, hollyhock, coconut leaf, and gong [8].

In terms of fauna, there are over 700 species of aquatic invertebrates, over 130 species of fish, nine amphibian species, 31 reptile species, and four mammal species. These include 11 species of reptiles named in the Vietnam Red Book, such as geckos, water lizards, ground pythons, brocade pythons, scorpion cobras, king cobras, and lilac crocodiles. The bird fauna comprise 130 species belonging to 47 families and 17 orders, including 51 species of waterfowl and 79 species of non-waterfowl birds living in many different habitats [28].

Can Gio mangrove forest has become a “lung” as well as a “kidney”, with the function of cleaning the air and wastewater from the city, industrial production zones, and at the same time returning a healthy environment, minimizing environmental pollution for Ho Chi Minh City and neighboring provinces. Can Gio mangrove forest also plays an important role in limiting damage caused by storms and floods, reducing up to 50% of the impact energy from waves and preventing sea level rise, as well as contributing to the protection of coastal populations and infrastructure [29].

Socially, Can Gio has five wards and towns with a population of about 60,000 people (2021). The main livelihoods of local people are rice farming, fishing and aquaculture, salt making, community tourism and small trade. Here, people’s living standards are below the average for Ho Chi Minh City. Over time, Can Gio has implemented many solutions to promote socioeconomic development in the direction of urban forestry and sustainable poverty eradication. Taking advantage of its natural resources, the District has supported farmers in transitioning all inefficient rice lands to aquaculture, production farms, general business operations, and bird nest farming. Currently, Can Gio is focusing on developing aquaculture in the form of extensive farming under the forest canopy to exploit the value from natural resources but without affecting the mangrove ecosystem [29].

Restoration and conservation of the Can Gio mangrove forest plays an important role in environmental protection, livelihood improvement, life stabilization and the attachment of Can Gio’s people and Ho Chi Minh City to mangroves.

3.2. Characteristics of the Study Sample

Table 2. Socioeconomic characteristics of the survey sample.

Commune	Binh Khanh	An Thoi Dong	Long Hoa	Tam Thon Hiep	Ly Nhon	Total	Percentage (%)
Gender
Male	87	92	85	84	102	450	64.3
Female	53	48	55	56	38	250	35.7
Total	140	140	140	140	140	700	700
Education
Elementary school	32	37	50	41	34	194	27.71
Secondary school	71	87	69	60	72	359	51.29
High school	30	14	16	30	26	116	23.2
University/College	7	2	5	9	8	31	4.42
Graduate	0	0	0	0	0	0	0
Total	140	140	140	140	140	700	700
Number of people in family
Mean	4.3	3.5	4.8	4.2	5.1	4.2
Max	7	6	8	7	7	6
Min	2	2	1	2	2	1
Household income (million VND/year)
Mean	198.2	270.1	220.8	244.3	181.2	222.8
Max	400	450	500	450	600	500
Min	75	90	85	120	90	120

Source: Compiled from this study analysis (2022).

summarizes the socioeconomic characteristics of the interviewees by commune and by total sample (five communes). The proportion of men was higher than that of women (males accounted for 64.3%, females accounted for 35.7%). In terms of education level, the numbers of people with secondary school qualifications accounted for a fairly large proportion of the total sample (51.29%). The proportion of people attending primary and secondary schools was quite similar (27% and 23%, respectively). The percentage of people with a university/college degree was quite low (only 4.4%). Thus, it can be seen that the education level of the interviewees in the five buffer zone communes was relatively low.

According to the survey results, there were an average of 4.2 people per household in the buffer zone communes (this variable was quite uniform in all five communes). The largest family had seven members, and the smallest had only one.

The average household income was VND 222.8 million/year. The average household income was highest in An Thoi Dong commune and lowest in Ly Nhon commune. An Thoi Dong commune is the administrative center of Can Gio district, so people have a higher average income due to participation in trade, tourism, service and logistics activities.

3.3. People’s Awareness of and Attitudes about the Ecological Value and Conservation of Mangroves

From the results of FGDs and he survey, we selected five points of view that were ranked in order in the questionnaire to reflect people’s awareness of and attitudes about the ecological value and conservation of mangroves:

• Strongly agree that mangrove forest must be preserved (5)
• Agree with conserving mangroves (4)
• Neither agree with nor oppose mangrove conservation (3)
• Disagree with the conservation of mangroves (2)
• Strongly disagree with mangrove conservation (1)

According to the survey results (

Table 3. Perspectives on mangrove conservation in Can Gio. Unit: person.

Commune	Binh Khanh	An Thoi Dong	Long Hoa	Tam Thon Hiep	Ly Nhon	Total
(5)	116	124	115	128	124	607
(4)	15	10	25	12	16	78
(3)	9	6	0	0	0	15
(2)	0	0	0	0	0	0
(1)	0	0	0	0	0	0
Total	140	140	140	140	140	700

), up to 79.5% of the interviewees in the sample chose the opinion “Strongly agree that mangrove forest must be preserved”. Even in An Thoi Dong and Tam Thon Hiep communes, almost all of the interviewees completely agreed that mangroves should be conserved. Only 15 people had an unclear attitude, and no one opposed the protection of the mangrove forest (Figure 2).

Similar results were obtained when people were asked to rate the importance of ecological functions and services provided by Can Gio mangrove forest (

Table 4. Perceptions of mangrove forest benefits in Can Gio. Unit: person.

Commune	Binh Khanh	An Thoi Dong	Long Hoa	Tam Thon Hiep	Ly Nhon	Total	Percentage (%)
Mangrove forest sustains people’s livelihoods
Very unimportant	0	0	0	0	0	0	0
Not so important	0	0	0	0	0	0	0
Normal	18	12	8	10	19	67	9.57
Fairly important	96	112	98	108	85	499	71.3
Especially important	26	16	30	22	36	130	18.6
Mangrove forest provides recreational and landscape values
Very unimportant	0	0	0	0	0	0	0
Not so important	56	20	22	25	27	192	0
Normal	48	22	26	36	70	202	28.9
Fairly important	32	86	88	51	25	282	40.3
Especially important	4	12	4	28	18	66	9.43
Mangrove forest protects against disasters
Very unimportant	0	0	0	0	0	0	0
Not so important	0	0	0	0	0	0	0
Normal	0	0	0	0	0	0	0
Fairly important	27	78	72	52	28	257	32.7
Especially important	113	62	68	88	112	443	67.3
Mangrove forest preserves biodiversity
Very unimportant	0	0	0	0	0	0	0
Not so important	24	10	20	12	18	84	12
Normal	60	14	26	32	34	166	23.71
Fairly important	40	84	70	66	28	288	41.14
Especially important	16	32	24	30	68	162	23.14
Mangrove forest conservation will benefit future generations
Very unimportant	0	0	0	0	0	0	0
Not so important	4	2	7	25	5	43	6.14
Normal	40	20	17	0	15	92	13.1
Fairly important	72	86	72	46	36	312	44.6
Especially important	24	32	44	54	84	238	34

Source: Compiled from this study analysis (2022).

). The majority of people believed that the value of the mangrove forest in the living area was very important. With 700 people interviewed, 499 thought the mangrove forest in the area played an important role in maintaining their livelihoods, accounting for 71.3%. Additionally, 130 people (18.6%) rated this role as particularly important and 67 people (9.57%) found this role normal.

The recreational and landscape value of the mangrove forest were highly appreciated by the interviewees. Specifically, 282 people rated it as fairly important, accounting for 40.3% of the total sample, while 66 people rated it especially important (9.43%), and 202 thought it was normal (28.94%).

The survey results showed that people in the buffer zone communes especially appreciated the natural disaster prevention value of the mangrove forest. A total of 443 people (67.3%) rated this at an especially important level, and 257 people (32.7%) rated it at a fairly important level. No one considered mangroves to be unimportant in disaster prevention. It is possible that a proportion of households had lived in the Can Gio area for a long time, had experienced many natural disasters and recognized this important ecological value. Therefore, they rated this value at a high level and as essential to the community.

Regarding the function of biodiversity conservation, people had a lower perception of its importance, with 84 considering it not so important (12%). Even so, 41% and 23% of interviewees rated this as fairly and especially important, respectively.

People in Can Gio had a very clear view of the future importance of the mangroves, or they wanted to preserve the forest to benefit future generations. A total of 312 people (44.6%) considered forest conservation as especially important in terms of the opportunities and benefits for future generations. A 34% share (n = 238) considered the forest’s future value to be especially important. In general, people were aware of the roles of mangroves in their lives and livelihoods, and expected the forest to be preserved, not only for the benefit of the present generation but also for future generations.

3.4. WTP Estimation Results

3.4.1. WTP Estimation with Parameter Model

The study used a random utility model to estimate the WTP of households for mangrove conservation. Experimentally, a binary logistic regression model and maximum likelihood estimation were used to estimate the expected value of households’ WTP for forest conservation. In this study, we used three estimation models corresponding to three different data sets to find the range of expected WTP in specific conditions (

Table 5. WTP estimation models.

No.	Model	Name of Model	Number of Observations
1	A	General model	700
2	B	Model providing illustrative mangrove images	350
3	C	Model without illustrative images	350

Source: Compiled from survey data (2022).

).

A summary of protest responses divided by commune is presented in

Table 6. Normal and protest responses, by commune.

Commune	Binh Khanh	An Thoi Dong	Long Hoa	Tam Thon Hiep	Ly Nhon	Total	Percentage (%)
Normal answer	12	14	8	9	15	58	8.29
Protest	128	126	132	131	125	642	91.7
Total	140	140	140	140	140	140	700

Source: Compiled from survey data (2022).

.

The results of WTP estimation using SPSS software 22.0 with binary logistic regression are shown in

Table 7. Parametric regression model results.

Variable	Model A	Model B	Model C
Constant	20.125 (6.53)	24.20 (6.72)	20.63 (7.81)
BID	−0.038 *** (0.007)	−0.041 *** (0.008)	−0.036 *** (0.005)
EDU	0.014 ** (0.059)	0.351 (0.082)	0.282 (0.097)
MEMBER	0.065 ** (0.045)	0.068 * (0.077)	0.045 (0.069)
INCOME	0.000 *** (0.000)	0.000 *** (0.000)	0.000 *** (0.000)
AGE	0.032 (0.008)	−0.017 (0.01)	−0.023 (0.015)
GENDER	0.186 (0.191)	0.202 (0.271)	0.469 (0.344)
−2 Log likelihood	633.730	323.064	300.116

Note: Standard deviation in parentheses. Significant at p < 0.1 (*), 0.05 (**), and 0.01 (***). Source: Compiled from this study analysis (2022).

.

The results show that in the three models, the coefficients of the BID variable were negative and significant at the 1% error level, consistent with findings in the literature that as the BID level increases, the likelihood of WTP at a certain BID will decrease.

The variable INCOME was also significant at the 1% error level and positive in all three models. Thus, in these models, household income has an effect on the WTP to conserve mangroves. Specifically, when family income increases by VND 1000/year, the probability of paying a given level of BID increases by no more than 1/1000 in models A and B.

The EDUCATION variable had a significant and positive impact on WTP with a 10% significance level in model A, and no significance in models B and C.

The MEMBER variable was only significant at the 5% error level in model A and 10% in model B. Hence, the number of people in the household affected the WTP to conserve mangrove forests. However, this variable had the opposite of the sign expected. The reason may be due to household psychology, assuming that more people benefiting from environmental services are willing to donate more money to conserve mangroves.

The estimation results indicated that the expected mean of the WTP in the general model was VND 529.6 thousand/household/year. In addition, there was no significant difference in means of WTP between model B (which provided images of the forest and its ecological value to the people) and model C (which did not provide images) (

Table 8. Estimation of WTP levels from the parametric regression model.

No.	Model	Mean of WTP (Thousand VND/Household/Year)
1	A	529.6
2	B	590.2
3	C	572.2

Source: Compiled from this study analysis (2022).

).

3.4.2. WTP Estimation with Non-Parametric Model

The study also estimated household WTP levels for mangrove conservation in Can Gio using a non-parametric model. The non-parametric model used information about the probability of agreeing to payment at different Bid levels to build a probability density function, thereby calculating the expectation of WTP. There were also three levels of WTP estimated for the three delineated models, as was the case with the parametric model.

Table 9. Probability distribution of agreement to pay at given Bid levels.

STT	Bid Levels (Thousand VND)	Probability of Answering “Yes” to Pay
		Model A	Model B	Model C
1	100	0.80	0.76	0.82
2	200	0.69	0.71	0.65
3	300	0.56	0.52	0.60
4	400	0.45	0.45	0.56
5	500	0.41	0.42	0.34
6	600	0.31	0.35	0.24

Source: Compiled from survey data (2022).

presents the results of the probability distribution of agreeing to payments at given Bid levels in the three non-parametric models. Accordingly, the simple hierarchy condition was not violated. Specifically, the smaller the Bid level, the higher the WTP.

The results of the non-parametric WTP estimation are presented in Table 9, according to which the expected level of WTP ranged from VND 563.5 thousand/family/year to VND 627.2 thousand/family/year. In the overall model (A), the expectation of WTP is 597.8 thousand VND/family/year. The WTP estimate with the non-parametric model was greater than that with the parametric model, which is theoretically suitable because the non-parametric model gave an upper bound estimate of WTP (

Table 10. Estimation of WTP levels from the non-parametric model.

No.	Model	Mean of WTP (Thousand VND/Household/Year)
1	A	597.8
2	B	627.2
3	C	563.5

Source: Compiled from this study analysis (2022).

).

4. Discussion

With regard to the protected areas in Vietnam, specifically CGBR, a nationally recognized biodiversity hotspot that is home to a number of unique mangrove species, CGBR was at the heart of the work of this study. The research was based on a CVM approach to determine local residents’ WTP for mangrove conservation and examine factors affecting WTP. The results show that nearly 90% of people were willing to pay for a conservation fund. This means that visitors are aware of the non-use value of protected areas, a finding consistent with those from other case studies around the world [19,22,30,31].

The value and ecological services of mangrove forests are well recognized by local people, and most believe that these benefits need to be preserved for future generations. Thus, people attach great importance to the sustainability of mangrove forests and are willing to pay part of their income to maintain them. In parallel with this study, other researchers also reached similar conclusions [25,26,29,32].

This study also revealed the factors affecting WTP for mangrove conservation. The research results show that when the level of a Bid increases, the probability of people agreement to pay will decrease. This is similar to the findings of most studies on WTP for the conservation of national parks and nature reserves around the world [16,26,30,31]. Among the factors affecting WTP, education level, income and number of family members were found to have a significant impact, among which education level has the strongest influence. As the level of education increases, people’s awareness of the value of mangroves will also increase, and this leads to an increase in the probability of people paying for conservation. This result is consistent with the studies of [17,25,26,30], which showed that education level is the most influential variable.

In addition, WTP based on the non-parametric model was higher than the WTP of the parametric model. This result is consistent with the results of [21,23,27] but contradicts the results of [32]. Ref. [31] also supported this result, but inconsistent results were reflected in the results of [18]. Although WTP in the non-parametric model was higher than that in the parametric model, the difference in this study was not large.

5. Conclusions and Recommendations

Through the CVM model, the amount that households in Can Gio were willing to pay for mangrove conservation ranged from VND 529 to 597 thousand/household/year. Research results showed that local people have a good awareness of the economic and ecological value of CGBR, especially that associated with their livelihoods and lives. People also want the mangroves to be preserved to benefit future generations, and most are willing to participate in conservation activities, including making financial contributions.

The results of the study have a number of policy implications. First, the financial mechanism for mangrove conservation should take into account the participation of communities and people who are attached to the environment and benefit from it. Second, the state and management agencies should conduct regular communication activities to increase public awareness of the role and value of mangrove forests. As people’s awareness increases, their conservation attitudes and behaviors will change in favor of the environment. In addition, the local communities need to organize activities and programs for mangrove conservation to give people direct access to information and strengthen the local government’s credibility. Third, through the investigation, it was found that it is possible to establish an environmental fund that is funded and managed by the community to conserve mangroves and contributes to building a sustainable financial mechanism for conservation in Vietnam.