Agriculture Land Conversion and Its Implications for Food Requirements and Farming in Vietnamese Northern Mountains45267

The paper aims at a quantitative monitoring and characterizing the conversion of agricultural to built-up land in a case study of Tran Yen Yen Bai by using satellite images focusing on t

EDESUS 2019, FOR PEER REVIEW https://edesus.ueb.edu.vn

Agriculture Land Conversion and Its Implications for Food Requirements and Farming in Vietnamese

Northern Mountains

Nguyen An Thinh (1)* , Nguyen Phung Quan (2)

(1) VNU University of Economics and Business, Vietnam National University, Hanoi, Vietnam

(2) The Committee on Ethnic Minority Affairs (CEMMA), Hanoi, Vietnam

* Correspondence: anthinhhus@gmail.com

Abstract: Vietnamese Northern Mountains have seen urbanization with slow pace in the relationship

with the effects the agricultural land conversion on local people's livelihoods, food and farming The paper aims at a quantitative monitoring and characterizing the conversion of agricultural to built-up land in a case study of Tran Yen (Yen Bai) by using satellite images focusing on the years 1994, 2004 and 2016 as well as examine the underlying relationships between specific land use land cover change (LULCC) and local livelihood diversification strategies Ethnographic fieldwork is combined with land-cover change mapping reveals hidden causes of LULCC Study results provided the status and drivers of the life and livelihood of people in Tran Yen district before and after land reclamation Our findings point to the importance of policy makers in approaching complementary approaches to integrating livelihood policies in the region to stabilize the lives of people Support policies should

be provided to households directly affected by these planning projects

Keywords: Agriculture Land Conversion; food requirements; farming; Vietnamese Northern

Mountains

1 Introduction

The conversion of agricultural land to non-agricultural land use is a common way to provide space for infrastructure development, urbanization and industrialization Therefore, it is almost an unavoidable trend during economic development and population growth (Tan et al., 2009) Not only the area of agricultural land which was used as residential land has declined, but also number of farmers have declined As a result of reduced livelihoods from agricultural, people have gradually moved to new livelihoods (Firdaus et al., 2011) According to Cruz (2010), majority (more than 80 per cent) of the smallholder farmers in the world are food insecure and depend on land as their primary source of livelihoods Three out of four poor people live in rural areas and depend on agriculture either directly or indirectly (WB, 2008) When livelihoods changed, demand for food and production patterns changed Food supply and distribution systems in most developing countries are undergoing major changes following rapid urban population growth (FAO, 2005) As cities and towns grow, in terms of physical size and population, the existing production systems and cropping patterns in the peri-urban areas intensify and the origin

of food supplies shifts, with supplies coming from areas further and further afield Land use land cover change (LULCC) could be detected by multiple means difference method In the 2000s, many studies applied effective tool for detecting changes in the space-time model of

the landscapes changes at various scales The results of studies indicate that the area of agricultural land is decreasing, but the area of construction land is increasing considerably However, all studies just carry out the image classification step and finding the variation of the area of the object class (Fazal, 2000; Mohammed et al., 2015; Mukhopadhyay et al., 2013; Statuto et al., 2016)

Since the start of economic renovation in Vietnam, socio-economic and LULC has changed more and more significantly With the emergence and development of industrial zones, economic zones, urban areas and other non-agricultural projects, a large amount of farmland has disappeared In the setting of accelerating conversion of farmland for urbanization and industrialization, a number of studies was conducted in major cities in Vietnam Applications of GIS for detecting and analyzing spatial changes as well as quantifying results show the urban growth processes, and its impact on the land-use distribution (Tran, 2008; Pham et al., 2010; Nguyen et al., 2012) An examination of landscape metrics in Hanoi indicates three trends of agricultural landscape transformation: loss of area, fragmentation and edge irregularity (Pham et al., 2014) Fragmentation and edge irregularity occurred mostly in new urban districts where there were expansion and linear branches along transport axes The impact of farmland loss to rural household livelihoods has been considered recently Turner et al (2015) mixed methods approach using remote sensing data and ethnographic fieldwork They examined how land-use and land-cover change (LULCC) has occurred across three borderland provinces – Lai Chau, Ha Giang and Lao Cai to analyze the underlying relationships between specific LULCC and local livelihood diversification strategies While the loss of agricultural land causes the loss of traditional agricultural livelihoods and threatens food security, it can also bring about a wide range of new opportunities for households to diversify their livelihoods and sources

of well-being A few studies argue that the impact is negligible, when people's income is not much difference after the acquisition of agricultural land (Nghi et al., 2012; Le and Tran, 2016)

2 Methodology

2.1 Study area

The research is carried out in Tran Yen district, which is located in the northeast of Yen Bai province, with an area of 62,857.99 hectares, geographic coordinates from 21°31'48

N to 104°59'00"E The North borders of Van Yen district, Ha Hoa district and Phu Tho province in the south, Yen Binh district and Yen Bai city in the east, Van Chan district in the west Tran Yen district center is Co Phuc town, 13.5 km away from Yen Bai city Tran Yen also has the Hanoi - Lao Cai railway running through, with the Red River running in the direction of Northwest - South East, Highway 37, Highway 32C and 03 Provincial Road 163,166,172 are land roads of district

Under the industrialization and modernization policy, Tran Yen district have been changes in agricultural production in particular, and economic activities in general According to the report of 2015, the area of agricultural land decreased 1,170.22 hectares; Non-agricultural land increased 1,237.86 hectares Labor in agriculture, although tending to

decline, still accounts for a high rate of 78.39% (2015), while labor in industry, construction occupies, and commercial services in spite of the trend is quite slow, 6.25% The economic structure has shifted sharply in the direction of decreasing the proportion of agriculture and forestry from 52.7% in 2005 and 41.0% in 2010, the industry and construction increased from 27.4% in 2005 and 35% 0% in 2010, trade - services increased from 19.9% in 2005 and 24.0%

in 2010

With the rapid restructuring of the economy towards reducing the share of agriculture and forestry, the area of agricultural land has been transformed into residential land, roads, factories, etc On the other hand, with the proportion of labor in the agricultural sector still occupies a high proportion, the acquisition of agricultural land has a significant impact on the livelihood of the people here For these reasons, we have chosen Tran Yen as the area to investigate and analyze the impact of agricultural land transfer on livelihoods of people

2.2 Data collection

The study is based on both primary and secondary sources of data, and has an exploratory design The primary data were collected through field surveys, whereas the secondary data were obtained from government and institutions through published and unpublished reports, records and literature

In order to detect the change of agricultural land and urban growth in the study area,

we used two Landsat 5 TM satellite images acquired in October 1994, October 2004 and Landsat 8 October 2016 Because these days have low cloud cover <20%, the object is clearly visible In addition, we used Google Earth, a map of land use status as a basis for classification

2.2.1 LULCC detection

Among the methods of classifying remotely sensed images, the object-oriented approach offers the possibility to use information such as shape, contextual relationship of the objects and thematic knowledge to distinguish ecologically meaningful habitat types that are not necessarily distinguishable by spectral features (Bock et al., 2005) This approach

is based on the creation of image objects (also termed segmentation) and the classification

of objects Segmentation involves subdividing the image into separate regions of spatially grouped pixels that should represent meaningful objects in the real world (Benz et al., 2004; Bock et al., 2005) The classification like that used in eCognition software is based on fuzzy membership functions of object features The software offers a wide choice of object features, such as spectral statistics, texture, shape and topological features (neighbor objects, super-level objects, etc.) (Baatz et al., 2004) The object-oriented approach also allows us to extract meaningful objects, that is, similar to real-world objects and patches, and to convert them into vectors that are easily integrated into GIS for further analysis (Benz et al., 2004) Therefore, we favor this approach to classify land cover types of the study area from Landsat images, we identified four land cover types from the Landsat images: agriculture, built-up areas, water bodies and forest These are the main target areas of the study area, which can

be separated from remote sensing images The objects are decoded by eye and the subject has developed the key to image interpretation (Table 1)

Table 1 Image interpretation keys

1 Water bodies

2 Built-up areas

3 Agriculture

Segmentation parameters in eCognition (bands, scales, color/shape ratio, and compactness/smoothness ratio) were tested at different values Segmentations and rule-based classifications were undertaken at different scales using the same band composition (bands 1–5, 7 Landsat 5; band 1-7 Landsat 8), color/shape ratio (0.3/0.7), and compactness/smoothness ratio (0.5/0.5) Segmentation values of 10, 20, 50, 100, and 200 were tested, aiming at creating segments of different sizes Segments were visually examined to determine the visibility of the main land-cover classes We chose value scale is 50 that produced the most homogenous segments in terms of spectral values and texture

First set of criteria was used to classify those objects into water class Then the second set of criteria was used to classify segments into three vegetation classes of agriculture and forest Built-up pixels were assigned by evaluating Max different

2.2.2 Food requirement and farming analysis

To analyze the changes in livelihoods under the impact of agricultural land use change, a top-down approach has been applied based on structured interviews with questionnaires structure According to the list of farmers in the acquired agricultural land and the list of industrial-oriented industrial development projects in the study area provided by the district People's Committee, two survey sites were selected: Bao Phat commune and Co Phuc town 75 households have been selected in 2 hotspots for land acquisition: 1 is the location of the garment factory and 2 is the location for the resettlement site The first interviews were conducted in December 2017, to assess whether questions in the questionnaire were appropriate The second interview was conducted in February 2018 After reviewing the data collected, two sets of responses did not prove reliable and were rejected Finally, 73 sets of feedback can be used as the basis for other data used and analyzed Structured interviews focused on basic household member information, land conversion, transgenic livelihoods, change in income from livelihoods, and expenditure on households before and after land acquisition

The whole sample was divided into two groups, corresponding to two different land acquisition areas, namely land for the construction of the garment factory in 2007-2009 and land for the 2016 road in Co Phuc town and land for construction Set up resettlement area

in 2015 in Bao Dap commune We have applied econometric methods to determine the degree of impact of agricultural land loss on the choice of household livelihood and household livelihood outcomes (Tuyen, 2013) Here we divide the livelihood activities into four main groups: formal workers, informal workers, agricultural workers and non-farm workers To analyze livelihood determinants, we apply a sustainable livelihoods framework (DFID, 2003) with five sources of capital: natural capital, social capital, human capital, material capital, and financial capital The change in funding is considered in terms of agricultural land loss In this research, we use two out of five source of capital

Two sources of capital which are natural capital, and human capital is used to analyze the effects of factors on the livelihood choices of households

Natural capital: covers the area of land owned / labor (100 square meter/labor) (can

be used as premise for household business), lost agricultural area (when the area Cultivation

is reduced, leading to increased idle time, the family is forced to move to new livelihoods.)

Human capital: expressed by household size and dependency ratio (this ratio is calculated by the number of family members under 15 and above 59, divided by the total number of members in the age group 15 to 59) (affects participation in wage work), age and level of household head (affect livelihood choices)

We also examine the impact of factors on livelihood outcomes by analyzing the remaining three sources, including:

Physical capital: This is reflected in changes in family facilities between before and after land acquisition (reflecting changes in livelihood outcomes, livelihoods are also improved)

Social capital: This is all social resources that can help people make a living (Ellis, 1999) Social capital is expressed through participation in social organizations Here, we

would like to see what kind of support groups are available to people when they lose agricultural land

Financial capital: in the form of access to formal and informal loans Households receiving formal or informal loans may use this resource to generate income or consumption (Tuyen, 2013)

In addition, we analyze the variation in household expenditures between before and after land acquisition, to assess the impact of the acquisition of livelihoods and ultimately

to influence to the life of the household

3 Results

3.1 Land use land cover change

Our land-cover mapping shows that the most important changes in terms of area

(Table 1) include an increase in Builtưup (6173.73 ha), followed by a decrease in Agriculture

land (roughly 667.81 ha), a decrease in bare soil (166.72 ha), and a decrease in forest (5975.02

ha) Examining the 1994, 2004 and 2016 maps (Figure 2), spatial patterns for these changes are visible (Table 2 and figure 1)

Table 2 Land use land cover change in the period 1994–2016

Land use land

cover

Area in

1994 (ha)

Area in

2004 (ha)

Area in

2016 (ha)

2016–1994 Area change (ha)

2016-1994 Percent Change (%)

Annual Rate of change (%/year)

Based on the results obtained after the classification of the image, we determined the level of urbanization and the variation in agricultural land use in the study area

Figure 1 Land use land cover in 1994, 2004 and 2016

3.2 Changes in food requirement and farming in agricultural land conversion areas

The research hold an interview in two locations in Bao Dap commune and Co Phuc town With 61 valid votes at Bao Dap and 12 votes at Co Phuc The interviewee were 78.1% male and 21.9% female, mostly in the 40-60s After synthesizing and processing forms, we have obtained the following results:

Factors influencing the choice of livelihoods

* Human capital

Water

Forest

Agricultural land Built - up

Table 2: Change the labor force in both agricultural and non-agricultural sectors

Prior to land acquisition, 56 households accounted for 76.7% of the total labor force, while the remaining 17 were non-agricultural workers, accounting for 23.3% But only 47 households have their main source of income based on agricultural activity After land acquisition, laborers in households mainly work as hired laborers for individuals, factories, etc The households still have land recovered or not continue to cultivate on that land but transferred to business activities Non-agricultural, or farming to serve the needs of the household, but not agricultural products for sale, earn income As a result, the number of households with major labor force in agriculture decreased sharply to 16 households accounting for 21.9%, and the number of households having the main labor force in non-agricultural sectors increased to 57 households, accounting for 78.1% Main income generating activities of households are also changing, with only one household having a main source of income based on agricultural activity, which is the conversion of agricultural land after harvesting rice to grow mulberry to feed silkworms The rest of the main income sources are non-farm based non-farm activities, accounting for 98.6%

* Natural capital

In the survey areas, there were three times of land acquisition There were 61 households have been cleared by 2015 to build resettlement area in Bao Dap commune In

Co Phuc town, there were 8 households which have been recovered land to build garment factory in 2007, and 7 households have been recovered land for road construction in 2016,

of which 3 households in five 2007 has been withdrawn

Table 3: The average area of agricultural land of the household groups conforming the land loss

rate (Unit: square meters)

(2006)

After losing land (2017)

Compare 2006/2017

The results of the survey show that the area of agricultural land of households is significantly reduced due to land acquisition Of which, group 1 has 6,47% average area of agricultural losing land corresponding to 124.1 m2, group 2 decreased 46.5% and 1054.5m2, and group 3 of average land area decreased to 92% corresponding to 369.3 m2

Number of employees

% (Percent)

Main income earning

employees

income earning

%

Agricultural

labor

Non-agricultural

labor

The average area of agricultural land after acquisition of household groups decreased, leading to a significant decrease in the area of agricultural land on agricultural labor In which, special as group 3, the average area of agricultural land is 401.4

m2/agricultural labor, reduced to only 32.1 m2/agricultural labor, which indicates that significant means livelihood of farmer households have been reduced considerably after land acquisition This is a big shock to them

Using the method of monetary compensation Outside compensate damage, support for job change were also made by money Other side, many people wanted to compensate, support by this form because having a large cash flow is each farmer‘s dream

Such as, natural capital source automatically is moved to financial capital source

Previous, agriculture land is an important means of livelihood for farmers, now transferred

a money accounts Depending on the purpose of using money, financial capital could be convert into physical capital or human capital

In addition, the survey showed that 18 households still have land but no further cultivation, of which 11 households have an area of over 1,000 square meters but no more cultivation Some other households, although still a lot of land but only partially farming The reason is that after being recovered, the area of agricultural land is leveled to build factory or resettlement site, causing some sections of the canal system to be filled, affecting irrigation The dry season is short of water, flood season flooded, leading to crop failure Therefore, although there is still agricultural land, households can no longer farm and have

to work as hired laborers or as workers in factories

Factors affecting livelihood outcomes

* Social capital:

Figure 2 Proportion of households with at least one member participating

in associations/groups (Source: Own calculation from author’s survey)

None Communist Party Vietnamese Fatherland Front

Youth Union Women's Union Farmer's Union Veteran's Union Cooperative Red Cross Others

%

Social capital is considered by the number of groups and organizations that households participate in Each family can participate in different groups Of the interviewed households, 11 households did not participate in the group Farmers association has the largest number of households with 40 households The Red Cross has the smallest number of households with 0 households All respondents indicated that they did not receive cash assistance or shift orientation after land acquisition from all stakeholder groups they participated in This shows that social capital has no role in the selection of livelihoods after land acquisition

* Physical capital

Table 4: Purpose of using compensation money

households

8 Other (for children to school, repayment, medical expenses,

etc.)

21

Depending on the amount of compensation received and the condition of each household, the difference in using the money varies of each households It can be seen that compensation for land acquisition is largely used for the purpose of improving living conditions Accordingly, 24 households built or repaired their house after land acquisition, accounting for 16.44%, and 5 households bought equipment, accounting for 6.85% It can be seen that the recovery of material conditions of households has improved significantly There is rotation from financial capital to physical capital, however, mainly in the means of living, a small amount of investment in means of production

Using this money for the purpose of building houses, purchasing assets (physical capital) and other purposes, people's livelihood in the long run will be difficult When they are poor, they may lose the opportunity to develop themselves, their family, and their social life

With regard to the physical capital used for the community as the roads are being built, the future will create conditions for people to improve the conditions for exchanges with the outside environment

* Financial capital

Table 5 illustrates some statistical descriptions of the sources and total value of loans borrowed by households in the last two years About 27.4 percent of the total sample households reported having at least one loan from the rural credit markets (formal and informal) The participation rate in informal credit was higher than that in formal credit The most important source of borrowing is borrowing from relatives, and the least used loans