Impact of Collaborative Forest Management on Rural Livelihood: A Case Study of Maple Sap Collecting Households in South Korea

Abstract

Some forest-dependent rural communities participate in the Collaborative Forest Management (CFM) program in South Korea, which provides the local people with access to national forests for the collection of non-timber forest products (NTFPs) in return for their contribution to the management of national forests. This study investigated what factors influenced rural communities’ participation in CFM and how their participation in CFM affected livelihood strategies and income level. Households in 17 villages near the Seoul National University (SNU) forests owned by the Government were interviewed. The study found that CFM participating households tend to choose diversified livelihood strategies. CFM participating households with diversified livelihood strategies are likely to have a higher income than non-CFM participating households choosing sectoral focus strategies. Expansion of the CFM program is suggested as a policy option to improve forest-dependent rural livelihood. However, ageing and female-headed households are faced with difficulties in participating in CFM due to their physical ability of forestry work while new settlers restricted access to forest resources. There is a need for innovation in forest governance for equitable distribution of forest services for both original and new residents to achieve sustainable rural livelihoods.

1. Introduction

1.1. Research Background

Forests and the forest sector can contribute to addressing various dimensions of poverty such as income generation, subsistence and vulnerability, as well as contribute to improving energy efficiency, agricultural productivity, rural development, and inclusive governance [1]. They can help generate income through the production of timber and NTFPs for forest-dependent communities and rural people including poor households who rely on forests for their subsistence. Forests also play an important role in the reduction of community vulnerability to external shocks such as natural disasters and economic crises. Moreover, forests can provide the rural and urban poor with low-cost energy sources such as fuel wood and charcoal, and contribute to crop and livestock production by providing nutrients and fodder. The forest-dependent communities may be provided with more opportunities for participating in decision-making of resource management.

CFM refers to the co-management of forest resources between government forest agencies and other stakeholders [2]. This participatory and collaborative approach is an effective vehicle for sustainable forest management. Some examples of CFM are community forestry (CF) in Nepal, joint forest management (JFM) in India, and community-based forest management (CBFM) in the Philippines. In the CFM program, the rights to access forest resources and utilize NTFPs in national forests can be granted to local communities, which participate in protecting and managing the national forests. Therefore, CFM programs have often been in conjunction with objectives in poverty reduction in developing countries [3].

Studies have found that CFM can contribute to the reduction of poverty and income inequality as well as forest conservation in various parts of the world. In Ethiopia, forest resource use in forest areas owned or managed by communities, largely contributed to increased household income from the forest resources, reduced rural poverty, and decreased income inequality within rural households [4]. In India, social outcomes such as livelihood opportunities and wealth distribution were more equitable in community reserve forests than in private forests and ecological outcomes such as total basal area, average height, and genera richness were better in community reserve forests than in private forests [5]. After CFM approaches were implemented in China and Indonesia, livelihood assets such as natural, physical, human, financial, and social capital in villages were improved [6,7]. Also, household participation in CFM can become a viable strategy for their livelihood improvement. In Bhutan, CFM participating households had higher income levels, higher food security levels, and reduced food poverty as compared to non-CFM participating households [8]. Moreover, in the Philippines, CFM participating household groups had moderate resilience while non-CFM participating household groups had weak resilience in terms of resilience measured by the integrated indicator of socio-cultural capital, natural capital and economic capital [9]. CFM can provide financial rewards and yield intangible benefits for communities (i.e., social capital), as seen in two communities of Quezon Province of the Philippines [10].

Other studies have looked into how the characteristics of households can affect decision-making regarding their participation in CFM. Demographic and socio-economic factors such as age, gender, education, household size, and household income affected households’ participation in forest management and governance in Indonesia, Kenya and Nepal [11,12,13]. Social capital such as an advice and information provision about CFM, and perception of CFM also influenced households’ participation in CFM activities [11]. However, in Nepal, households at a low social stratum in caste and income level, as well as female-headed households are less likely to participate in the collection of forest products in CFM due to restricted access to forest resources [14,15]. Social heterogeneity such as sociocultural and economic differences can influence the access to forest resources and the perceptions of forest resources, and consequently affect the outcome of the CFM program [16]. Thus, case studies of a variety of countries are needed to investigate the outcome of CFM considering the social heterogeneity.

The study aims to identify what factors influenced households’ participation in CFM and how their participation in CFM affected livelihood strategies and income level. The case of 17 communities in Gwangyang-si and Gurye-gun of South Korea was studied for the objectives of testing a few theoretical hypotheses related to the aim of this study.

1.2. Collaborative Forest Management in South Korea

1.2.1. Forest Policy and Rural Community of South Korea

South Korea is a mountainous country where around 63% of the land is covered with forests as of 2019. Forestlands at large were deforested due to the overexploitation of shifting cultivation and the collection of fuelwoods. Deforestation continued until the mid-1950s and forest transition began in 1955 [17,18,19]. In order to restore the deforested and degraded forestland in the 1960s to 1980s, the government focused on providing economic incentives to forest owners and tree growers. The government’s forest rehabilitation programs were supported by the integrated policy for substituting fuel wood with fossil fuels and facilitating a demographic shift from rural to urban areas [18]. Policies reducing the demand for wood, providing substitutes for domestic wood with non-wood fuel, together with other changes such as increased timber importation and reduced fiber of forest resources, helped the forest landscape to be restored. Thanks to the success of reforestation, the forest growing stock of South Korea dramatically increased 8.7 m³/ha in 1955 to 125.6 m³/ha in 2010, which surpassed the world average. Nonetheless, the same policies and changes that helped to restore forests, may have had a negative impact on the rural communities in terms of their economic growth.

To sustain forest-dependent rural communities, they require livelihoods that are ecologically secure, economically efficient, and socially equitable [20,21]. However, opportunities to improve economic efficiency are limited even though there are abundant forest resources in the mountainous areas of South Korea. The overall forests of South Korea are managed according to six categories of forest functions, including the supply of wood, preservation of natural environment, development of water sources, prevention of forest disasters, preservation of human habitats, and forest recreation. Most of the forest area in the country is managed for the supply of wood (33.9%) and preservation of natural environments (21.7%). Only 9.4% of the total forest area is managed for forest recreation, although the demand for forest recreation for both rural and urban people is still on the rise. The number of visitors who come to the national parks increased from 41.0 million in 2012 to 43.2 million in 2019 [22]. The number of visitors who come to the recreation forests increased from 9.4 million in 2010 to 16.0 million in 2019 [23]. NTFP production is not included in the category of forest functions even though NTFPs are the most important products in the forestry sector of South Korea. Forest management focused on the wood supply and nature conservation limited the opportunities for employment or revenue generations of rural communities. Consequently, forest-dependent rural communities have lower income than urban communities in South Korea. The average annual income of households in the mountainous areas is only 47.5% of that of households in the urban areas in 2014 [24]. The poverty of forest-dependent communities, especially for the elderly, and income inequality between rural and urban households are a social issue.

1.2.2. CFM Program of South Korea

Following the establishment of the forestland ownership system in the early 1900s, local people were allowed to access and use forest resources only in designated areas within national forests. The State Forest Administration and Management Act was enacted in 2006 to promote the function of national forests and manage it efficiently. Under the Act and its Article 11, the Korea Forest Service (KFS) has implemented the State Forest Protection Agreement and Concession of State Forest Products since 2006, which provides local people with NTFPs such as mushrooms, wild vegetables, pine nuts and tree sap if they agree to participate in national forest protection activities such as forest patrols for preventing forest fires and illegal logging. Villagers living near national forests can collect NTFPs in national forests through annual contracts at the village level while engaging in the State Forest Protection Agreement for 5 years and extending the term of agreement by up to 5 years. However, KFS may terminate the agreement if the villagers violate conditions of the agreement. This agreement is the CFM program in South Korea, under which NTFPs can be an incentive for local people to participate in national forest management. In 2016, the national forest area covered by CFM was 38.8% of the total national forest area which is 25.5% of the total forest area (1.6 million ha) for 702 villages participating in the collection of NTFPs and generating NTFPs worth of 6.8 billion KRW (around 6 million USD, 1 USD = 1128.5 KRW).

Before the forestland ownership system was established in South Korea, forestland and forest resources were common-pool resources of local communities. A unique pine forest user group called Song-gye was organized by villagers in the late period of the Joseon dynasty (1392–1910) [25]. It was formed to secure the communal use of forest resources such as fuelwood and timber for heating and cooking [25]. Song-gye was based on collective decision-making membership, election of leader and committee members, accumulation and management of assets and penalties, and utilization and management of forest resources [25]. Members of Song-gye could benefit from logging and collecting fuelwood, timber, and fodder. In return of receiving these benefits, they were required to carry out forest protection activities including forest patrol to prevent forest fires, illegal logging, and shifting cultivation. Unfortunately, Song-gye has almost disappeared in South Korea due to changes in the forestland ownership system, and what remains of Song-gye are none of its original functions due to changes in energy sources.

The impact of the CFM program on the livelihood of the rural households collecting NTFPs from the national forests in South Korea was assessed recently. According to the study, households participating in CFM had higher revenues from NTFPs than households of non-participating villages in Gangwon Province and Gyeongsangbuk Province [26]. Households participating in CFM were more dependent on NTFPs than non-participating households [27]. Another study revealed that the CFM program increased the provision of ecosystem services such as NTFPs production, carbon storage, and the prevention of forest fires of national forests in South Korea [28].

The rural communities are disadvantaged for living with less developed infrastructure, and cultural and social services as compared to urban areas. The industrialization led the immigration of young and productive labor forces to industrialized cities. The population of rural societies are ageing and relatively poor to the urban. The average income of forest-dependent households in South Korea is only half of urban households [24]. However, there has been little research on the link between the CFM program and poverty alleviation despite the potential of CFM to alleviate poverty in forest-dependent communities in mountainous areas of South Korea.

1.3. Sustainable Rural Livelihood Framework

Sustainable rural livelihood (SRL) is defined as a level of wealth and of stocks and flows of food and cash that provide for physical and social well-being and security against becoming poorer [29]. The concept of SRL consists of a combination of capabilities, equity and sustainability [30]. In the SRL framework, given a particular context, the institutional process can affect access to livelihood assets as well as mediate the ability to carry out livelihood strategies, and achieve outcomes such as poverty reduction and sustainability of natural resources [31].

Based on the SRL framework, household livelihood strategies for income generation and poverty reduction are generally influenced by their given assets or available resources [32,33,34,35]. Moreover, the impacts of the institutional process on the livelihood asset, strategy, and outcomes of rural communities were identified in various parts of the world. After participating in effective agricultural practices from the Grain for Green program in China, farmers’ livelihood assets and strategies were diversified, and their incomes improved [36]. Similarly, participating in rural tourism as part of a rural revitalization strategy of the Chinese government, farmers’ livelihood assets increased, and their livelihood strategies tended to be more diversified, resulting in increased livelihood sustainability [37,38]. In Pakistan, households linked to forest enterprises who implemented a large-scale afforestation project, had more livelihood assets, generated more income, and came to have a higher wealth index as compared to households not linked to such enterprises [39].

The SRL framework is an appropriate framework to describe linkages between livelihood assets given to households, decision-making process in which households participate in CFM, livelihood strategies chosen by households, income level achieved as an outcome of livelihood strategies for the aims of this study.

2. Materials and Methods

2.1. Study Area and the CFM Arrangement

The study area is around the Southern Seoul National University (SNU) Forest in Gwangyang-si and Gurye-gun, Jeollanam Province. As of 2020, 293 villagers of 34 villages participate in the CFM program and 19.8 percent of the total Southern SNU Forest area (3203 ha of a total of 16,206 ha) is managed by villagers and used for maple sap collection. Out of 34 villages participating in CFM, 17 villages were randomly sampled.

Figure 1 shows the location of these study areas. The Southern SNU Forest comprises two areas: one of 10,961 ha located at an elevation of 20 m to 1222 m of Mt. Baekwoon including the protected areas for ecosystems and landscape conservation, and those of 5245 ha are located at an elevation of 20 m to 1732 m of Mt. Jiri within the national park. The Southern SNU Forest is a temperate deciduous forest (N 35°01′~35°20′, E 127°30′~127°43′) with an annual average temperature of 14.0 °C and an annual average precipitation of 1461 mm from 1967 to 2011.

SNU Forest is owned by the Ministry of Education of Korea and managed by SNU. The Southern SNU Forest management office was established in 1946 to manage the experimental forests. They have recognized local communities’ right to tap for sap from maple trees in the SNU Forest for a long time, officially since 1991. There are agreements every year between villagers’ groups and the SNU Forest management office concerning the amount of sap collection, conditions for maintaining forest health and protecting maple trees, and use fees [17]. Before these agreements, the Southern SNU Forest management office obtains permission to collect sap from the local government and the Korea National Park Service. Villagers tap a maple tree for sap in early spring for two months, and then manage and restore the forests for one month. They sell the fresh sap to tourists visiting their village in early spring and it is an important channel of distribution of sap [17,40]. This agreement is a case of the State Forest Protection Agreement and Concession of State Forest Products as the CFM program in Korea.

The sampled 17 villages consist of 54 households on average and 24% of the total households participate in the agreement on the maple sap collection in the SNU Forest. Households run tourism businesses for their income generation in 15 out of 17 villages.

2.2. Conceptual Framework

This study set up a logical structure using the SRL framework. The Southern SNU Forest management office has been providing the local communities located nearby the SNU forest with the access to maple sap in the SNU Forests through agreements signed with villagers since 1991. We postulated that the participation of rural households is determined as a function of their livelihood conditions (i.e., the level of assets including physical, cultural, financial, human and social capital). Each household can decide whether it participates in this agreement with SNU considering the assets available to their livelihood strategies. Whether or not households participate in the agreement may change their livelihood strategies. Once they agreed to collaborate in forest management with SNU, they can collect and sell maple sap from SNU forests, while cooperating in managing the SNU forests with an emphasis in the prevention of forest fires and illegal harvesting of forest products.

Figure 2 shows the SRL framework and the scope of this study. First, this study analyzes the factors affecting households’ participation in the agreement on maple sap collection in the SNU Forest in terms of livelihood assets, focusing on human and social capital. Second, this study analyzes the correlation between households’ participation in the agreement and their livelihood strategies. In this framework, livelihood strategies are categorized into two types: business diversification and sectoral focus business. Third, this study analyzes the impacts of livelihood strategies on livelihood outcomes, especially in terms of household income.

2.3. Data Analysis

2.3.1. Logistic Regression Model

To determine the factors affecting households’ participation in CFM and their impacts on the choice of livelihood strategy, the following equation is used in the estimation of the results:

$${\mathrm{Participation}}_{\mathrm{i}}={\mathsf{\alpha }}_0+\mathsf{\alpha }\mathrm{A}+{\mathsf{\epsilon }}_{\mathrm{i}}$$

(1)

where ${\mathrm{Participation}}_{\mathrm{i}}$ is a dummy variable that measures whether or not a respondent’s household participates in the agreement on maple sap collection in the SNU Forest for the ith individual. $\mathrm{A}$ represents the human and social capital as livelihood assets of a respondent’s household and ${\mathsf{\epsilon }}_{\mathrm{i}}$ is the error due to the unobservable variables affecting household’ participation in CFM.

Table 1. Definition and coding value of dependent and explanatory variables for the logistic regression model.

Variable			Definition	Coding Value
Dependent variable	Participation		Participation in the agreement for collecting maple sap in the SNU Forest	1 if respondent’s household participates in the agreement on maple sap collection in the SNU Forest, otherwise 0
Explanatory variables	Human capital	Age	Age of household head	Numerical age of head of respondent’s household
		Female-headed	Female-headed household living alone without spouse and children	1 if respondent belongs to a female-headed household, otherwise 0
		Family size	Number of family members	Number of family members in respondent’s household
	Social capital	Length of residence	Length of residence in the village	Number of years respondent has resided in the village

shows the definitions and coding values of participation in CFM as a dependent variable and livelihood assets as explanatory variables in Equation (1). The dependent variable in the equations is measured by a discrete value and the equation is estimated using the logistic regression model not linear regression model in STATA 14/IC.

2.3.2. Chi-Squared Test

A chi-squared test is used to analyze the correlation between households’ participation in the agreement on maple sap collection in the SNU Forest and their livelihood strategies. In this study, each variable is measured by a categorical variable. Participation is categorized into two types such as CFM participating and non-CFM participating household, and livelihood strategies are also categorized into two types such as diversification and sectoral focus strategy from the income sources (i.e., forestry, agriculture, tourism, labor and other business, and others).

Table 2. Definition and coding value of variables for the chi-squared test.

Variable	Definition	Coding Value
Participation	Participation in the agreement on maple sap collection in the SNU Forest	1 if respondent’s household participates in the agreement on maple sap collection in the SNU Forest, otherwise 0
Diversification	Diversification from more than two income sources	1 if income from more than two income sources constitutes the income of the respondent’s household, otherwise 0

shows the definition and coding value of participation and diversification variables.

2.3.3. Ordered Logit Model

To identify the impact of households’ participation in CFM on household income level, this study uses the following equation:

$${\mathrm{Income}}_{\mathrm{i}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{Strategy}}_{\mathrm{i}}+ \mathsf{\beta }\mathrm{D}+{ \mathsf{\epsilon }}_{\mathrm{i}}$$

(2)

where ${\mathrm{Income}}_{\mathrm{i}}$ is an ordinal variable that measures annual household income level for the ith individual. ${\mathrm{Strategy}}_{\mathrm{i}}$ is a dummy variable that measures by crossing between households’ participation in the agreement and their choice of livelihood strategy. $\mathrm{D}$ represents the demographic attributes of a respondent’s household as control variables and ${\mathsf{\epsilon }}_{\mathrm{i}}$ is error due to unobservable variables affecting household income.

Table 3. Definition and coding value of explanatory and control variables for the ordered logit model.

	Variable	Definition	Coding Value
Dependent variable	Income	Annual household income level	if under 10 million KRW a if between 10 to 20 million KRW if between 20 to 30 million KRW if between 30 to 40 million KRW if between 40 to 50 million KRW if more than 50 million KRW
Explanatory variable	Livelihood strategy	Type of livelihood strategy group decided by diversification and household participation in the agreement on maple sap collection in the SNU Forest	if respondent’s household chose the diversification strategy with participation in the agreement if respondent’s household chose the sectoral focus strategy with participation in the agreement if respondent’s household chose the diversification strategy with no participation in the agreement if respondent’s household chose the sectoral focus strategy with no participation in the agreement
Control variables	Age	Age of household head	Numerical age of head of respondent’s household
	Female-headed	Female-headed household living alone without spouse and children	if respondent belongs to a female-headed household, otherwise 0
	Family size	Number of family members	Number of family members in respondent’s household

Note: ^a 8868 USD = 10 million KRW on October 27, 2020.

shows the definition and coding value of income level as a dependent variable, strategy as an explanatory variable, and demographic attributes of a respondent’s household as control variables in Equation (2). Strategy is defined as the type of livelihood strategy decided by diversification and households’ participation in the agreement on maple sap collection in the SNU Forest. This equation has a dependent variable measured by a 6-scale ordinal value and were estimated using the ordered logit model in STATA 14/IC.

2.4. Data Collection

In this study, interviews were conducted with the household heads in villages participating in the agreement on maple sap collection in the SNU Forest. Interviewers trained by the authors, surveyed the villagers at the household level using a semi-structured questionnaire during the off-peak season for farming and maple sap collection in 2015, 2016, 2018 and 2019. The questionnaire consists of questions on the age and gender of the respondent, number of family members, length of residence in the village, income level, ratio of forestry income to total household income, ratio of agriculture income to total household income, ratio of tourism income to total household income, ratio of labor and other business income to total household income, ratio of income from other income sources to total household income, and whether the respondent’s household participates in the agreement on maple sap collection in the SNU Forest. The questionnaire can be seen in Supplementary Materials. Out of the 300 households in 17 villages interviewed, the data collected from 257 households were used in the analysis excluding 43 duplicated and missing data. Demographic, socio-economic, and ecological conditions are assumed to be constant in the villages during 2015 to 2019 considering that such conditions do not vary significantly in a mountainous area over a relatively short period of time.

3. Results

3.1. Characteristics of Respondent Households

The characteristics of 257 respondent households are shown in

Table 4. Characteristics of respondent households (N = 257).

Variable		Mean	Std. Dev.	Min.	Max.
Human capital	Age (Numerical age)	67.7	13.6	24	99
	Female-headed (Female-headed = 1)	0.3	0.5	0	1
	Family size (Number of members)	2.1	1.3	1	9
Social capital	Length of residence (Number of years)	46.5	24.8	1	98
Income source	Forestry (%)	22.3	31.8	0	100
	Agriculture (%)	16.7	32.2	0	100
	Tourism (%)	11.9	26.2	0	100
	Labor and other business (%) a	9.3	26.7	0	100
	Other income sources (%) b	39.7	43.1	0	100
Participation (participation in the agreement = 1)		0.38	0.49	0	1

Note: ^a Labor and other business are defined as all kinds of paid employment and self-employment using non-natural resources such as construction and transportation business. ^b Other income sources are defined as transfer income and unearned income.

. The average age of the respondent households’ head is 67.7 years old and the ratio of female-headed households to the total number of respondent households is 30%. Respondent households consisted of two family members on average and the head of respondent households had resided in the village for 46.5 years on average. In the study area, communities are ageing and show trends of young people moving out of the area to seek education and career opportunities, and elderly people moving into the village after their retirement. For this reason, there are two family members in each respondent household on average as the household only includes an older couple without children. The average length of residence in the village is 21 years shorter than the average age of the household head.

The income sources constituting the income of respondent households is shown in Table 4. In the study area, 39.7% of the total income of respondent households is constituted by the income from other income sources such as subsidies, remittances from their family or relatives, and the rent of land. The income from forestry activities constitutes 22.3% of the total income of respondent households, in which 16.7% of the total income of respondent households is constituted by the income from agriculture activities, 11.9% is constituted by the income from tourism activities, and 9.3% is constituted by the income from labor and other business activities.

Table 4 also shows that 38 percent of respondent households participate in the agreement for collecting maple sap in the SNU Forest. Collecting maple sap through the participation in the agreement and selling it to tourists is an important income source to villagers in the study area.

3.2. Factors Affecting Household Participation in CFM

Table 5. Results of estimating the logistic regression model for factors affecting household’s participation in the agreement.

Variable	Coefficient	Z Value
Age	−0.045 ***	−3.07
Female-headed	−0.684 *	−1.78
Family size	0.144	1.09
Length of residence	0.023 ***	3.11
Constant	1.339	1.40
Obs.	257
LR chi2 (prob > chi2)	33.37(0.000)
Pseudo R2	0.098
Log likelihood	−153.652

Note: *** p < 0.01, ** p <0.05, * p <0.1.

shows the estimation results of Equation (1) using the logistic regression model. The results indicate that the younger the household head, the higher the probability of participating in the agreement on maple sap collection in the SNU Forest. Male-headed households are revealed to be more likely to participate in the agreement as compared to female-headed households. On the other hand, household heads who have been residents of the village for a long period of time, are more likely to participate in the agreement as they have access to forest resources.

3.3. Choice of Household Livelihood Strategy through the CFM Program

Table 6. Results of the chi-squared test for the correlation between households’ participation in CFM and their livelihood strategies.

	Non-CFM Participating Households	CFM Participating Households	Total Households
Households who chose the sectoral focus strategy	108	15	123
Households who chose the diversification strategy	52	82	134
Total households	160	97	257

Note: Pearson’s chi² = 65.528, p-value = 0.000.

shows the tested results of the correlation between households’ participation in the CFM program and their livelihood strategies using the chi-squared test. In this test, households that participate in the agreement, tend to choose diversification strategies than households who do not participate.

3.4. Impact of Participation in CFM on Household Income

The study area was found to have a high poverty rate.

Table 7. Distribution of income of respondent households.

Household Income Level		Frequency	Percentage (%)
1	under 10 million KRW a	127	49.4
2	between 10 to 20 million KRW	55	21.4
3	between 20 to 30 million KRW	18	7.0
4	between 30 to 40 million KRW	15	5.8
5	between 40 to 50 million KRW	10	3.9
6	more than 50 million KRW	32	12.5
	Total	257	100.00

Note: ^a 8868 USD = 10 million KRW on 27 October 2020.

shows the distribution of income level across respondent households. Almost half of respondent households belong to the lowest income level. Moreover, income level of respondent households is largely distributed among the two lowest income levels and the highest income level. In other words, the distribution is slightly polarized.

Table 8. Results of estimating the ordered logit model for the impact of household’s livelihood strategy on household income.

Variable		Coefficient	Z Value
Livelihood strategy a	CFM participating diversification	0.814 **	2.55
	CFM participating sectoral focus	−0.122	−0.22
	Non-CFM participating diversification	1.315 ***	3.68
Age		−0.074 ***	−6.35
Female-headed		−1.751 ***	−4.25
Family size		0.216 *	1.74
/cut1		−4.643
/cut2		−3.184
/cut3		−2.619
/cut4		−2.087
/cut5		−1.667
Obs.		257
LR chi2 (prob > chi2)		148.42(0.000)
Pseudo R2		0.204
Log likelihood		−289.710

Note: *** p < 0.01, ** p < 0.05, * p < 0.1; ^a Non-CFM participating sectoral focus strategy as a reference.

shows the estimation results of Equation (2) using the ordered logit model. In this model, households that participate in the agreement on maple sap collection in the SNU Forest and diversify their livelihood strategies, have a probability of belonging to a higher income level than households that are not a part of the agreement and do not diversify their livelihood strategies.

Despite not participating in the agreement, households that chose diversification strategies are likely to have a higher income level than those that did not choose diversification strategies. They can generate more income by utilizing their own livelihood assets such as agricultural and forest lands, tourism facilities, and jobs.

On the other hand, households that participate in the agreement on maple sap collection in the SNU Forest but do not diversify their livelihood strategies, are not likely to have a higher income level than households that are not a part of the agreement and do not diversify their livelihood strategies. The sectoral focus strategy with participation in the agreement does not generate a high income. However, income from maple sap collection in the SNU Forest can be only the income source to households who do not have enough livelihood assets.

4. Discussion

4.1. Forest Management to Increase Income of Forest-Dependent Households

Participation in the policy program and the project changed households’ livelihood strategies into more diversified strategies because it increases available assets and accessible resources in previous research [36,37,38,39]. As tapping maple sap in the SNU Forest in early spring only for 2 months, CFM participating households can engage in agriculture, tourism and other businesses for the rest of the year. Consequently, there is evidence that households’ participation in the agreement on maple sap collection in the SNU Forest diversifies their livelihood strategies.

As shown in Table 8, the CFM participating diversification strategy positively affected household income by generating additional income from NTFPs. This is similar to a previous study on the impact of CFM on household income and poverty where CFM participating households had higher income levels and reduced food poverty as compared to non-CFM participating households [8]. CFM participating diversification households can increase their income through additional income from NTFPs with running agriculture, tourism, labor and other businesses, and being supported by other income sources such as subsidies and remittances from their family or relatives similar to another previous study [4]. However, the distribution of forestry household income tends to be polarized in South Korea [41,42], and this trend results from the income disparity across different types of forestry activities such as NTFP collection, tree fruit production, ornamental tree production, and timber production [43]. Although the CFM participating sectoral focus strategy did not generate higher income than other livelihood strategies, the CFM program could contribute to the subsistence of the poor through NTFP collection in the study area. Thus, providing local people with NTFPs continuously are required to increase household income in mountainous areas.

The Government has implemented the State Forest Protection Agreement and Concession of State Forest Products, a CFM, to protect national forests and to increase local people’s income since 2006. In 2016, the national forest area covered by CFM was 38.8% of the total national forest area. This study found that expansion of the CFM program can contribute to rural policy by increasing income and reducing poverty of forest-dependent households.

Moreover, even though NTFPs are the main source of forestry revenue in South Korea [23], the functional categories of forests in the forestland using a planning system set by the KFS do not include NTFP production for rural communities. Rural people participating in CFM complain of decreasing NTFPs resources available in the national forests due to forest management practices with little consideration on the NTFPs production. The Government needs to manage national forests in a way that recognizes the importance of NTFPs as an economic output as well as a means of social policy of national forest management, in particular promoting their contribution to income generation for rural communities.

Regardless of household’s participation in the CFM program, diversified livelihood strategies positively affected household income as well. Thus, opening up more forests to the public for forest recreation activities and creating regular jobs in mountainous areas are required to increase household income. Especially, villages near the forests providing local people with NTFPs and providing the public with forest recreation services simultaneously have a synergy to increase income of forest-dependent households. This is because local NTFPs can help supply fresh foods and materials to households involved in restaurant businesses, and local consumption of NTFPs helps secure the market for households that produce NTFPs.

4.2. CFM Program for Sustainable Rural Livelihood

As compared to previous research on CFM in developing countries, the results of our study reveal that rural communities in South Korea are ageing and most of the residents are elderly people, especially in mountainous areas. Previous research [44,45] referred to the increase of the aging index in mountainous areas of South Korea from 1990 to 2018. This was attributed to the outflow of people in their 20s and 30s, who have moved to cities to pursue education and career opportunities [44,45]. These previous studies also anticipated that the population in mountainous areas would continuously decrease until 2050, with a high ratio of the 60s and 70s age groups and females. In Table 5, ageing and female-headed households could not participate in the CFM program. These results coincide with the findings of previous studies [8]. These two factors are related to the household’s capability of collecting maple sap and managing the forests. Ageing and female-headed households have physical limitations making it difficult to carry out labor-intensive work in the forest. Subsidies such as pensions for the aged play an important role in maintaining their subsistence. In this situation, ageing and female-headed households will not have livelihood strategies to generate income and to improve their livelihood. These households require safeguards from poverty to ensure that their livelihoods can be sustained and improved.

As populations in mountainous area continue to become aged and decline, the labor force becomes reduced. Ageing households no longer have the capability to collect NTFPs in national forests, and they are unable to participate in CFM activities. To sustain the CFM program, resettled households should be considered as new participants replacing the ageing labor force to manage national forests efficiently.

However, in Table 5, resettled households who have lived in the mountainous areas in Gwangyang-si and Gurye-gun for a short period, are less likely to participate in the agreement on maple sap collection in the SNU Forest. This is attributed to the restrictions on accessing maple sap. Some villages have full autonomy at the collective-choice level with their own rules to decide on membership, to allocate maple trees to members, and to impose fees and penalties. According to the rules, permits for collecting maple sap are distributed to villagers who have been residents for a long period of time, usually at least 5 or 10 years [17,40]. In previous research [14,15], rural societies in Nepal restrict the access to forest resources to those in the low social stratum and females. In South Korean rural societies, villagers tend not to recognize a person as a member of the village who has resettled in the village from other areas. This specific characteristic of a closed society influences the new comer’s access to maple sap, which is a common-pool resource. At the collective-choice level, autonomous rule affected action of users who utilize common-pool resources [46,47]. Creating an equitable benefit-sharing system in the CFM program can contribute to poverty alleviation and sustainable rural livelihoods [48]. The findings of this research can be discussed further with respect to social equity based on the social-ecological system (SES), which can be a theoretical framework about governance of common-pool resource management.

5. Conclusions

As more people return to mountainous areas from urban areas, developing diverse income sources is becoming important. The CFM program provides rural people with the opportunity to diversify their livelihood strategy. Also, the CFM program could help them to generate more revenue through the access to forest resources in a national forest regardless of their own livelihood assets. In light of the findings of this research, we suggest that the government should develop a forest and rural development policy expanding the area of national forests managed in collaboration with rural communities, which can contribute to rural livelihoods while designating more national forests for forest recreation and job creation in mountainous areas.

However, due to the customary village rules of restricting the access to NTFP resources of the national forests in the background, they face difficulties in participating in CFM for NTFP production. Given that the populations of mountainous areas are ageing, an innovative governance system providing fair opportunities for participation in the CFM program for both original residents and new comers in mountain villages will be an essential solution for sustainable rural livelihoods in mountainous areas. An equitable distribution system of forest benefits between older villagers and resettled villagers, in which resettled villagers can access NTFPs and donate some of the revenue from the NTFPs to older and poor villagers, can contribute to poverty alleviation and sustainable rural livelihoods as well as increased household income in South Korea. The autonomous rules can be modified to provide resettled villagers with a fair opportunity to access NTFPs. This can lead to the building of social capitals between original villagers and new comers and the success of resettlement policies in mountainous areas.