Like all other parts of the world, land use patterns in Bangladesh especially of south-west part have been observed to change rapidly since late of 20th century. Lands of south-west region were generally used for rice farming since the middle of 20th century but polderization project of Bangladesh during 1970s caused major changes in land use pattern either through transformation or modification of land cover and cropping. Literature shows that single cropped rice areas of past decades have already been cultivated twice or thrice per year while some such lands have already been converted for shrimp farming. This paper examines the determinants of land use patterns and their corresponding changes (i.e. rice and shrimp farming) over time at pirozpur village of Kaligonj upazila under Satkhira district of Khulna division in Bangladesh. The study is being done on the basis of cross-sectional data collected from the decision maker or head of each sample household. Here data have been collected through questionnaire as well as focus group discussion from a sample size of 80 households; each forty from shrimp and rice farming. Here logistic regression considering rice farming land as the reference dummy as well as cost-benefit analysis is being done to know the extents of land use determinants. However, the study area being close to river Hariavanga, shrimp farming has become predominant in the study area and young people are more interested in shrimp farming than in any other land use alternatives. Analysis shows that cost free irrigation for shrimp farming as well as higher profit, lower cost and available inputs are the major factors of increased shrimp farming in the study area. The study also finds that if rice can be cultivated thrice per year then shrimp is less attractive while there lacks training facilities for the rice farmers which may cause dissatisfaction to land owners causing conversion of rice land into shrimp. Available land holders primarily decide their land use pattern based on short run cost benefit calculation rather than long run impact of land use in their livelihood as well as ecology. The study finds age, natural calamities, family type and availability of credit to be negatively related with shrimp farming while land engagement process, accessibility, economically active family number, proximity to service sector, neighborhood land use patterns, land ownership and land rent to be positively related.
Table of Contents
Acknowledgement
Abstract
List of Maps
List of Tables
List of Figures and Graphs
Acronyms
Abbreviations
Chapter One: Introduction
1.1 Background of the Study
1.2 Objective of the Study
1.3 Rationale of the Study
1.4 Scope of the Study
1.5 Operational Definitions
1.6 Limitation of the Study
1.7 Structure of the Study
Chapter Two: Theoretical Background
2.1 Land Use Models
2.2 History and Trends of Land Use Models
2.3 Land Use Modeling Approaches and Models
2.3.1 Agent-Based Perspective
2.3.2 Systems Perspective
2.3.3 Narrative Perspective
2.3.4 The Fitting Data Model
2.3.5 Simulation Processes
2.3.6 Structural Models
2.3.7 Statistical or Reduced Form Models
2.3.8 Geographic Models
2.3.9 Economic Models
2.3.10 Stochastic Markov Model
2.3.11 Ecological Models
2.3.12 Dyna-CLUE model
2.3.13 Spatial Economical Model
2.3.14 Cellular Automata Model
2.3.15 Species-distribution Model
2.4 Economics, Econometrics and Land Use Research
Chapter Three: Literature Review
3.1 Land
3.2 Land Use
3.3 Land Use Change
3.4 Land Cover
3.5 Land Cover Change
3.6 Land Use and Cover Change
3.7 Land Use Planning
3.8 Land Use Conflict
3.9 Methods Used to Identify Patterns and Changes of Land Use and Cover
3.10 Variable Used in Modeling Land Use and Cover Changes
3.11 Type and Scope of Land Use and Cover Change
3.12 Observed Land Use Pattern
3.13 Global Land Use and Cover Trends
3.14 Land Use Trends in Bangladesh
3.15 Causes of Land Use and Cover Change
3.16 Impact of Land Use and Cover Change
3.17 Initiatives for Land Use and Cover Changes
3.18 Findings and Results of Land Research
3.19 Problems and Limitation of Land Use and Cover Researches
3.20 Research Gap
Chapter Four: Methods and Materials
4.1 Conceptualization of the Research Problem
4.2 Study Area
4.3 Research Design
4.4 Target Group
4.5 Sample Design
4.5.1 Sampling Techniques
4.5.2 Sample Size
4.5.3 Data Collection Method
4.6 Type of Data Used
4.7 Variables and Indicators
4.8 Model Specification
4.8.1 Logistic Regression for Land Use Change
4.8.2 Empirical Analysis of Land Use Determinants
4.9 Data Collection
4.9.1 Primary Data Collection
4.9.2 Secondary Information
4.10 Data Processing and Analysis
4.11 Writing the Thesis Paper
Chapter Five: Land Use Patterns and Changing Trends
5.1 Global Land Use Patterns
5.2 Land Use Trends of Bangladesh
5.3 Trends of Land Availability in Khulna Division
5.4 Land Use Trend in South-west Part of Bangladesh
5.5 Land Use Policies in Bangladesh
Chapter Six: Overview of Study Area and Respondent
6.1 Overview of Study Area
6.2 Information of the Respondents
6.2.1 Age and Gender of the Sample Population
6.2.2 Educational Status
6.2.3 Family Size and Composition of the Respondents
6.2.4 Occupational Distribution
6.2.5 Engagement Process in Present Land Use Pattern
6.2.6 Land Ownership Pattern of Households
6.2.7 Scenario of Assets and Non-assets of the Sample Households
6.2.8 Household Yearly Income
6.2.9 Household Yearly Expenditure
6.2.10 Households’ Farming Experience
6.2.11 Training Facilities of Sample Population
6.2.12 Credit Facility
6.2.13 Plan to Change Land Use Pattern in Near Future
6.2.14 Pressure and Regulation on Current Land Use Pattern
Chapter Seven: Results and Discussion
7.1 Lands Cultivated over Time
7.2 Variation in Land Use Pattern
7.3 Change in Land Use Pattern
7.4 Location of Land
7.5 Land Elevation
7.6 Fertility of Land
7.7 Salinity and Sand in Land
7.8 Neighborhood Land Use Pattern
7.9 Water Management Facilities
7.10 Distance of Water Management Sources
7.11 Way Used for Water Management System
7.12 Cost of Water Management System
7.13 Proximity to Nearest Infrastructure
7.14 Land Rent
7.15 Accessibility to Land
7.16 Transport Mode and Available Facilities to Specific Land
7.17 Cost of Transportation per Trip
7.18 Availability of Input
7.19 Demand for Final Product
7.20 Market Location
7.21 Price Distribution of Final Output
7.22 Changes in Land Use Patterns of the Households
7.23 Conversion and Maintenance Cost
7.24 Cost-benefit of Land Use
7.25 Estimation of the Determinants of Land Use Change
Chapter Eight: Findings and Conclusion
8.1 Information through Focus Group Discussion
8.2 Findings of the Research
8.3 Comparison of Findings
8.4 Conclusion
8.5 Further Scope
List of References
List of Web References
Appendix I
Appendix II
List of Maps
illustration not visible in this excerpt
List of Tables
Table 4.1 Description of Independent Variable
Table 4.2 Explanation of Variables in Empirical Analysis
Table 5.1 Land Use Trends in Bangladesh during 1977-2008
Table 5.2 Scenario of per Capita Arable and Irrigated Land
Table 5.3 Total Land Area of Bangladesh during 1976-2010
Table 5.4 Rice and Shrimp Farming Area during 1976-2010
Table 5.5 Land Use Statistics of Khulna Division in 2008
Table 5.6 Land Use Pattern in Khulna Division during 1976-2010
Table 6.1 Khulna Division at a Glance
Table 6.2 General Information of Kaligonj Upazila
Table 6.3 Age and Gender Distribution
Table 6.4 Educational Status of the Decision maker
Table 6.5 Literacy Status of Sample Population
Table 6.6 Family Type of Sample Population
Table 6.7 Distribution of Economically Active Family Member
Table 6.8 Occupational Distribution of Sample Household
Table 6.9 Engagement Process in Current Land Use Pattern
Table 6.10 Information on Land and Non-land Assets
Table 6.11 Distribution of Income from Land and Non-land Assets
Table 6.12 Yearly Expenditure of Sample Household
Table 7.1 Amount of Land Cultivated over Time
Table 7.2 Variation in Land Use Pattern
Table 7.3 Distribution of Water Source
Table 7.4 Distances of Water Source and Disposal Location
Table 7.5 Way used for Water management
Table 7.6 Cost of Irrigation and Water Disposal
Table 7.7 Proximity to Nearest Infrastructures
Table 7.8 Land Rent Scenario per Year
Table 7.9 Cost of Input and Output Transportation
Table 7.10 Price Distribution of Final Output
Table 7.11 Summary Statistics
Table 7.12 Estimation of Determinants of Land Use Change
Table 7.13 Marginal Analysis of Determinants of Land Use Change
Table Annex_II.1 Description of Sample Data used in Logistic Regression
Table Annex_II.2 Summary of Sample Data used in Logistic Regression
Table Annex II.3 Summary Statistics of Categorical Variable
Table Annex II.4 Classification Table
Table Annex_II.5 Classification Table
Table Annex_II.6 Omnibus Tests of Model Coefficients
Table Annex_II.7 Hosmer and Lemeshow Test
Table Annex_II.8 Contingency Table for Hosmer and Lemeshow Test
Table Annex_II.9 Model Summary of Land Use Determinants
Table Annex_II.10 Wald Test of Sample Data
Table Annex_II.11 Test of Data Classification
Table Annex_II.12 Goodness-of-fit Test
Table Annex_II.13 Results of Binary Logit Model
Table Annex_II.14 Results of Logistic Regression
Table Annex_II.15 Marginal Analysis of Sample Data
Table Annex_II.16 Variables in the Equation
Table Annex_II.17 Observed and Probable Land Use Pattern of Each Sample
List of Figures and Graphs
Figure 2.01 Economic Dynamics of Land Use System
Figure 5.1 Land Use Statistics of Khulna Division in 2008
Figure 5.2 Percentage Land Uses during 1989-2010
Figure 6.1 Land Ownership Pattern of the Sample Population
Figure 6.2 Farming Experience
Figure 6.3 Training Facilities on Specific Land Use
Figure 6.4 Credit Facilities on Specific Land Use
Figure 6.5 Expectation of Change in Current Land Use
Figure 6.6 Expected Land Use Pattern in Future
Figure 6.7 Determinants of Expected Changes in Land Use
Figure 6.8 Pressure and Regulation Scenario on Land Use
Figure 7.1 Land Use Statistics of Sample Households during (2010-2014)
Figure 7.2 Changes in Total Land Size during 2010-2014
Figure 7.3 Location of Sample Land
Figure 7.4 Land Elevation Scenario of Sample Land
Figure 7.5 Fertility Scenario of Sample Land
Figure 7.6 Distributions of Salinity and Sand in Land
Figure 7.7 Neighborhood Land Use Patterns
Figure 7.8 Accessibility to Sample Land
Figure 7.9 Mode of Transport Used
Figure 7.10 Transport Facilities for Specific Land Use Pattern
Figure 7.11 Availability of Input for Specific Land Use
Figure 7.12 Demand Prototypes for Final Output
Figure 7.13 Distribution of Market for Final Product
Figure 7.14 Changes in Land use Patterns (early 2008- mid 2014)
Figure 7.15 Initial Conversion Cost for Specific Land Use Pattern
Figure 7.16 Yearly Land Maintenance Expenditure
Figure 7.17 Cost-benefit Analysis of Rice and Shrimp Farming
Figure 7.18 Change in Profit based on Cropping Frequency
Figure Annex_II.1 Area under ROC Curve
Figure Annex_II.2 Sensitivity and Specificity versus Probability Cutoff
Acronyms
illustration not visible in this excerpt
Abbreviations
illustration not visible in this excerpt
Acknowledgement
This thesis paper is prepared as a requirement of completing graduation in Economics from Khulna University since October, 2014. However, the author being grateful like to thanks Almighty because His great provision, protection and support throughout his whole life and especially during this research work.
The author can’t but feel owe to supervisor, Mr. Md. Firoz Ahmed, for his constructive suggestion, criticism and encouragement throughout the research work to prepare such a representative research work by such a short span of time and despite all obstacles. Appreciation then goes to Economics Discipline as well as all the faculties and staff for their effort, suggestion and cooperation towards my progress of life since I have become a member of Economics Discipline of Khulna University and especially in this research work.
The author is also grateful to the respondents, the secretary as well as other staffs of Dhalbaria Union Parishad and the local representatives for the friendly behaviors and help towards my successful completion of the thesis paper. Moreover, the writer likes to empress his gratitude towards his friends, well-wishers and others who are not being mentioned here for their cooperation during the research work and whole life. It is author’s privilege to express gratefulness and deep sense of appreciation to all those individuals and institutions whose direct as well as indirect invaluable contributions and support have helped me in writing up this thesis.
Last but not the least, the author like to remember the devotion and contribution of his family members for their encouragement, support and help throughout the whole life. He is also grateful to all the teachers and others who have teach and support him in gaining knowledge and experience till now.
Finally the author like to ask reader and evaluator to take the mistakes as unnoticed by the author during the completion of this paper in such a short time. Moreover, the author being a new comer in research likes to acknowledge the errors in this paper because of his low experience and expertise in research.
Jahangir Alam
BSS 101502
Economics Discipline
Khulna University, Khulna
Abstract
Like all other parts of the world, land use patterns in Bangladesh especially of south-west part have been observed to change rapidly since late of 20th century. Lands of south-west region were generally used for rice farming since the middle of 20th century but polderization project of Bangladesh during 1970s caused major changes in land use pattern either through transformation or modification of land cover and cropping. Literature shows that single cropped rice areas of past decades have already been cultivated twice or thrice per year while some such lands have already been converted for shrimp farming. This paper examines the determinants of land use patterns and their corresponding changes (i.e. rice and shrimp farming) over time at pirozpur village of Kaligonj upazila under Satkhira district of Khulna division in Bangladesh. The study is being done on the basis of cross-sectional data collected from the decision maker or head of each sample household. Here data have been collected through questionnaire as well as focus group discussion from a sample size of 80 households; each forty from shrimp and rice farming. Here logistic regression considering rice farming land as the reference dummy as well as cost-benefit analysis is being done to know the extents of land use determinants. However, the study area being close to river Hariavanga, shrimp farming has become predominant in the study area and young people are more interested in shrimp farming than in any other land use alternatives. Analysis shows that cost free irrigation for shrimp farming as well as higher profit, lower cost and available inputs are the major factors of increased shrimp farming in the study area. The study also finds that if rice can be cultivated thrice per year then shrimp is less attractive while there lacks training facilities for the rice farmers which may cause dissatisfaction to land owners causing conversion of rice land into shrimp. Available land holders primarily decide their land use pattern based on short run cost benefit calculation rather than long run impact of land use in their livelihood as well as ecology. The study finds age, natural calamities, family type and availability of credit to be negatively related with shrimp farming while land engagement process, accessibility, economically active family number, proximity to service sector, neighborhood land use patterns, land ownership and land rent to be positively related. Whatever be the determinants of land use and their corresponding extents, mass awareness should be emphasized for optimal land use.
Key Words: Land Use Change, Determinants, South-west Region, Bangladesh
Introduction
Though land is an important component of nature to maintain ecological as well as bio-physical balance (Agarwal et al., 2001; Mohammad, 2009), there remains very little landscape on earth in their natural state (Zubair, 2006). Researchers have already reported that our universe has been changing rapidly through urbanization and industrialization with a corresponding decline of green lands and alteration of structure and functioning of ecosystem (Vitousek et al., 1997; Schneider and Pontius, 2001). Thus, changes of land use patterns i.e. forest into farmland, farmland into periphery; with shifting and melting of shorelines and glaciers have attracted global concern (NASA, 2006). Like many other parts of the world, land use patterns have been changing in Bangladesh (Qusem, 2011) with appalling impacts on livelihood pattern of her people chiefly who are dependent on land (Mondal, 2008). Moreover, south-west region of Bangladesh has already gone through dynamic but unsustainable changes in land uses (Rahman and Begum, 2011) because most of the areas are being observed to transfer its agro-land to non-agro uses i.e. rice to shrimp farming (Zubair, 2006; Quasem, 2011). Therefore, this paper is an attempt to address and discuss some of the existing land use patterns of south-west region and their determinants.
1.1 Background of the Study
Since 10,000 BCE, world population was less than 10 million with negligible land uses (NASA, 2006) but with the industrial revolution as well as rapid population growth especially in developing states (Lambin et al., 2003), researchers now claim that human footprint has affected 83% global terrestrial surface while degraded more than 60% ecosystem in last century (Nkonya et al., 2012). Moreover, settlements and sprawl development have become much influential both in underdeveloped, developed and developing countries (Oluseyi, 2006) with rapid and continuous changes in land use patterns (Minar et al., 2013). Therefore land has now been considered to have locally pervasive and globally significant influence on ecology and sustainability (Agarwal et al., 2001) mainly because of its limited size (Zubair, 2006). Humans have been altering land uses through clearance of patches of land (Shi, 2008) since the beginning of civilization and it is now claimed that during last three centuries, nearly 1.2 million sq km of forest and woodland as well as 5.6 million sq km of grassland and pasture have been transformed (Ramankutty and Foley, 1999) while cropland has increased by 12 million square km (Vitousek et al., 1997).
It is also demanded that most populated areas are located along coastal zones and major waterways in Indian Sub-continent, East Asia and western Europe (Lambin et al., 2003) and have witnessed major land use changes in last decades (Nkonya et al., 2012) through aggregated global influences in trade and commerce (Lambin et al., 2001). Researchers have also demanded that economy expands in size and growth with the expansion of population, invention and innovation (Houghton, 1994) which in turn causes a transfer of agro land to non-agro uses (Quasem, 2011). For instance, though by 1970 there were two megacities (e.g. populations more than 10 millions), today there are 23 megacities and is estimated to be 37 in 2025 of which most are in developing countries (Rui, 2013). Researches also show that land has both positive and negative influence on human life and environment based on the pattern of uses over time (Li, 1996; Vitousek et al., 1997; Zenga et al., 2008). In this point, Coleman (1987) and Lambin et al. (2001) has pointed out that large proportion of world’s problems observed recently have originated from the misuse, disuse, overuse, under use or abuse of land resources either directly or indirectly.
Coming from world scenario to Bangladesh, we observe that Bangladesh as an agricultural country with the involvement of more than 47.5% (MES, 2010; as cited in BBS, 2013) who contributes about 19.41% to total GDP in FY2011-12 (BBS, 2013) while that in FY2004-05 was 22% (BBS, 2005). Over the last 30-40 years, availability of agricultural land in Bangladesh has been declining at the rate of 1% per year (NFPCSP, 2011) while data available from BBS (2005) and BBS (2009) showed that between 1985 and 2007, net cropped area decreased by 11% (i.e. from 8770 to 7800 thousand ha). Moreover being a land of rivers, Bangladesh loses about 80 thousand ha of agro lands yearly (MoWR, 2005; as cited in Iftekhar, 2006) while nearly one percent land is being converted to non-agro uses particularly for settlements and infrastructure (Quasem, 2011) per year. In this regard, Islam et al. (2004; as cited in Mia and Islam, 2005) showed that about 220 ha arable land is being reduced daily due to uses such as road construction, industrialization and housing while at least, 86,000 ha of land has already been lost to river erosion during 1973-2000 (MES, 2001).
About 70% of total lands in Barisal and Khulna divisions are now affected by different degree of salinity (Mia and Islam, 2005) which are causing reduced agro production (PDO-ICZMP, 2004). PDO-ICZMP (2004) also showed that per capita agro land since 2003 was 0.056 ha (BBS, 2009) and will be decreased to only 0.025 ha by 2050 because of substitution by shrimp farming which started during eighties of last century (Mia and Islam, 2005). Therefore, marginal and small farmers are becoming more vulnerable (Quasem, 2011). Recent reports show that majority of her population being poor and exposed to various natural and artificial hazards (Iftekhar, 2006), frequent land use changes are occurring especially in south-west region (FAO, 1999; Mia and Islam, 2005; Minar et al., 2013). However, bio-physical, socio-economic and environmental objectives of sustainable development are not only affected by land use changes but also affect LUCC effectively (Müller, 2003).
1.2 Objective of the Study
Based on information through literature survey, the author has formulated a broad issue of land use problems under the objective of identifying the major land use patterns and their corresponding determinants in South-west region of Bangladesh over time. Moreover, author has reviewed the objective more critically as follows.
i. To trace out the major land use patterns and their corresponding changes
ii. To explore observed determinants of land use change from rice farming to shrimp farming and their respective extents
To achieve the above objectives, the author has collected information from various secondary sources to represent the land use patterns and their evolution both in regional and global context along with necessary local information collected through questionnaire survey with the aim of empirical and comparable analysis.
1.3 Rationale of the Study
Though the earth started her journey with few people (NASA, 2006), she now possesses millions of inhabitants and has been experiencing modern lifestyle and unplanned urbanization since industrial revolution (Chase et al., 1999; Schneider and Pontius, 2001). Moreover, land being one of the scarce natural resources as well as factors of production (Mohammad, 2009), has been observed to have diversified uses both in reversible and irreversible ways (Islam, 2000). Researches also show that economy experiences rapid expansion in size and growth to keep pace with the rapid increase in and expansion of population, invention and innovation (Houghton, 1994). Moreover, observations from last century particularly last decades have showed that changes in land use patterns are global in nature (Houghton, 1994; Dale et al., 2000) especially because of high dependency of human being on land (NFPCSP, 2011) for livelihoods, modern shelter in superb places, desired food for eating (NASA, 2006). Therefore, lands are becoming scarce natural resource (Mohammad, 2009) day by day causing acute conflicts (Ruben et al., 2008) especially due to lack of coordinated action amongst the various parties concerned with land management especially in developing nations (Mohammad, 2009). Change in land use patterns or the physical changes in land cover caused by human action is now a concern (Müller, 2003) due to its disastrous changes (Schneider and Pontius, 2001) at unparalleled rates, magnitudes and spatial scales (Turner, 1994a; Vitousek et al., 1997).
With high pressure on its natural resource base (NFPCSP, 2011), Bangladesh is in threat of declining agro lands (Zubair, 2006) with devastating consequences on country’s ability to sustainably achieve and maintain self-sufficiency in food and agro-based livelihoods (NFPCSP, 2011). Besides, shifting rate of agricultural land to non-agricultural uses is alarming with respect to crop production and food security in Bangladesh (PC, 2009; Rahman and Hasan, 2003). In this connection, SRDI (2010) estimated approximately 0.13% land was transferred from agro to non-agro sector per year during 1963 and 1983 (Rahman and Hasan, 2003) while PC (2009) demanded that at least one quarter of country’s agricultural land has already been lost since independence. Researches also show that shifting rate of productive lands towards unproductive purposes may be faster in 21st century because of rapid economic growth and infrastructural development (Hasan et al., 2013). Though shrimp farming was initially introduced in coastal as well as in the South-western part (Rahman et al., 2013), production of shrimp has now been growing at an annual rate of 20-30% since 1990 (Primavera, 1997). Moreover shrimps cultivated in coastal Bangladesh now accounts more than 2.5% of global production of shrimp with its position as the 7th exporter to the markets of Japan, EU and USA (Rahman et al., 2013). Despite all concerning reports on land use issues, very little attention has yet been paid to formulate a dynamic as well as suitable national land use policy to conserve and make best possible use of this highly scarce land (Mohammad, 2009).
1.4 Scope of the Study
Land use pattern and its corresponding changes are in a close dependency with natural, socio-cultural and economic factors (Dale et al., 2000) and also influence the behaviors and decision making over time and space (Houghton, 1994; Dale et al., 2000; Ruben et al., 2008). Therefore, better understanding of the determinants of land use patterns as well as corresponding determinants is necessary (Agarwal et al., 2001; Lesschen et al., 2005) to assess present situation and possible future impact on sustainable development of environment, economy and society at large (Verburg et al., 2004; NASA, 2006). From this perspective, this study is primarily an attempt to consider what are the major determinants of lands used for rice and shrimp farming with an emphasis on the mode of interaction among the different driving forces of this two land uses. And for this purpose, collection of primary data, its analysis and presentation of analyzed data is being shown in a simple but effective way both using descriptive statistics and econometric models.
Broadly, data both on land use patterns and its trends of world as well as Bangladesh are being collected through secondary survey while about study area through questionnaire survey and face to face interviews. Moreover, households who have at least certain amount of personal lands for use (i.e. settlements, cultivation or any other purposes but must be personally owned) are the sample population and the decision maker of that specific household is treated as the target respondent. Data is also being collected from local representatives (i.e. chairman, union members, agro officer in charge) and from the oldest as well as large land holders for more accuracy of data. Though there is variation in the socio-economic status of the target population, only respondents living in the study area at least for five years or more are being selected as the target population. Moreover, the simplest as well as flexible procedures are being taken to complete the research work in time.
1.5 Operational Definitions
To avoid unnecessary confusion about the various used terms in this paper, here is the description of commonly used terminologies with their used meaning rather than traditional one as follows.
Household: Household is to be distinguished from family which comprises members having blood relationship while members of a family may live in different places but members of a household must live in the same place and share the same kitchen.
Illiterate: Respondent or decision maker who doesn’t have receive any education and can’t even write his name are treated here as illiterate.
Informal Learning: When respondents are able to read and write or at least can signature but didn’t participate in any formal institution (i.e. school, college) rather have learnt through participating in any informal learning centre (i.e. from friend, night courses offered by NGOs).
Land and Non-land Assets: Land assets include only the land resources possessed by each household while non-land assets are any resources (i.e. tress, furniture, business) except lands.
Land Owner and Farmer: Landowner and farmer are both used throughout this paper to refer to the person making land use decisions primarily. Broadly, to be land owner one must have his own land while farmers may or may not his own land.
Land Use: Land use refers to the manner in which people employ their land and its resources including cultivation or use of earth surface.
Land Use Pattern: land use pattern implies to all possible as well as existing manners in which humans are employing available land and its resources for the benefits both in the present as well as in future context.
Land Use and Cover Change: Land use and cover changes mainly refer to the replacement of natural lands (i.e. forests and grassland for agricultural use or agro lands for shrimp farming or settlements) over time either due to pressure or for expected benefits from any such conversion.
Mauza : Mauza is the lowest administrative unit having a separate jurisdiction list number in revenue records with its well-demarcated cadastral map.
Mixed Use: When lands are used in different ways over time and doesn’t follow any sequence, it is termed as mixed use lands. Mixed use here includes using the same lands either in more than one use at a single time (i.e. rice and shrimp farming) or using any lands in non-repetitive ways over some consecutive years.
Motorized, Non-motorized and Human Transport: Motorized transport takes account of motor cycle, private cars and auto-rickshaw while non-motorized one includes by-cycle, rickshaw (van). Human transport on the other hand includes human labor curt run by human force for transportation.
Neighborhood Characteristics: Neighborhood characteristics consist of different observed land use patterns in adjacent lands of the land under consideration.
Nuclear and Joint Family: Family which consists of only one spouse but may have members of any number while joint family refers to having more than one spouses under the control of single decision maker.
Other Occupation: In occupation, the terms others are being used to describe no certain sources of income that is transitory income by the households.
Primary, Intermediate and College Education: Here primary education ranges from preliminary stage (Class one or equivalent one) to till class eight (VIII), intermediate from class nine (IX) to twelve (XII) and college education refers to higher stages after intermediate education such as graduation, PhD.
Regular and Irregular Expenditure: Regular cost of household includes day to day transaction for maintaining each household while irregular expenditure refers to transitory expenditure (i.e. medical cost) by each household per year.
Remittance: Money (i.e. Bangladesh Taka) sent by other family member(s) who are working either abroad or far from his houses for at least six months.
Rice Farming: Using a certain piece of land only for cultivating rice whole year or any certain part of the year. All the rice farming lands under consideration are cultivated using traditional methods with little modern instruments like machinery, fertilizer while seeds are local.
Service: Service in this paper includes sale of labor hour at a single time and includes labor income, maid servant and teaching.
Shrimp Farming: When any land is used only for producing shrimp almost all the year round is treated here as the shrimp farming land. Shrimp farms are of different size but use factors of input from same sources and also sell their final output to same market at a more or less same price of both input and output.
Beside the above stated definitions as well as terminologies, some other terms are also used as described critically during the analysis or at the point where they need to define for easy understanding and to reduce ambiguity.
1.6 Limitation of the Study
In this study different types of data are being collected from similar types of work around the world and Bangladesh simultaneously together with the primary data from selected study area. Moreover, time series data are being given priority in order to understand the trends of changes in land. But in this regard, the author failed to manage enough time series data of land use pattern and corresponding changes due to lack of availability of secondary data especially of the study area. Besides, agriculture has a strategic function because it is the main food supplier for the people in Bangladesh (Hasan et al., 2013) and thus different estimation methods of agricultural statistics provide various data and information, so their reliability is questionable.
Moreover, the author couldn’t use sufficient econometric as well as statistical tools because of lack of expertise as it is the first time to do such a research for the author. The author has faced major problems in econometric analysis due to small sample size mostly in case of incorporating necessary variable and due to presence of several proxy or dummy variables in the study. Furthermore, similar answer by the respondents in several cases made the analysis contradictory despite the truth of such occurrence in the sample area. It is also to be noted that while calculating various continuous data there were some mismatch which are assumed to be the result of considering some factors but excluding some interrelated one.
The author for successful completion of the research work has used recall data where there may some lacking of consistency as well as accuracy of data on land use of the study area. And even in some cases there is variation in financial information despite other information being the same. Moreover, this paper hasn’t taken time value of money into consideration while dealing with time series cost and profit data.
1.7 Structure of the Study
The research work has been conducted in a systematic pattern which can be described in a well mannered way for quick overview of the paper. Primarily, this paper starts with writing of acknowledgement, abstract, table of contents for an easy understanding of the whole paper at a glance and then includes the main body of the research work, references and annex such as questionnaire, results of land use determinants.
The first chapter of the paper includes the background, objective, rationale with a clear definition of the scope of the study and faced limitations as well as problems. The paper then, Chapter Two, shows the theoretical background (i.e. theories and propositions on land use analysis) for explaining the research problem and associated issues in a systematic manner. The third chapter, named literature review has become informative with the arrangement of available literature and lastly existing research gap. The paper in next, Chapter Four, shows the materials and methods followed to complete the research work from research problem formulation till submission with especial emphasize on variables, model formulation, target group, research methods, tools of analysis and presentation process.
Description about the study areas and corresponding respondents are being enumerated in Chapter Six while Chapter Five includes some qualitative as well as quantitative overviews about land use and cover changes from global, national as well as local context. Chapter Seven constitutes the heart of the paper because here has been done the analysis of the collected data according to the objective. Presentation of major findings and comparison with literature along with concluding remarks and further scope of research are being enumerated in Chapter Eight.
Land use change is central to environmental management through its influence on biodiversity, water and radiation budgets, trace gas emissions, carbon cycling, and livelihoods (Lambin et al., 2000a; Turner, 1994). Wu and Li (2013) argued that world agriculture is going to face tremendous pressure for intensification over the next 50 years especially because of increase in demand for food dramatically. Therefore, land use modeling has attracted considerable attention (Gobim et al., 2002; Lambin, 1997; Serneels et al., 2001; Veldkamp and Fresco, 1996; Verburg et al., 2002; Wu and Yeh, 1997) to sanctify knowledge to recognize the determinants of land use (Yadav et al., 2012) over time and space. For example, the complexity of land use patterns and their changes over the last decades calls for multidisciplinary analyses (Veldkamp and Lambin, 2001) for a sustainable environment in future.
Chapter Two Theoretical Background
Land use and cover change (LUCC) issues have already attracted the interest of various researchers (Lambin et al., 2000; Verburg et al., 2004; Li, 2011; Wang, 2012; Silva and Wu, 2012) ranging from those modeling spatial and temporal patterns of land conversion (Verburg et al., 2008; Priess and Schaldach, 2008) to those trying to realize causes and penalties linked with these aspects (Irwin and Geoghegan, 2001; Burgi et al., 2004). Besides, land use analysis is complex for its dynamism as well as determinants (Lambin et al., 2003; Long et al., 2007) and asks for diverse approaches rather than single one for consistency and precision (Verburg and Veldkamp, 2001; Long, 2003; Cai, 2001; as cited in Long et al., 2007). Since, modeling land use issues represents part of the complexity of land use systems (Veldkamp and Lambin, 2001), reviews of different models on the basis of preferred variables (i.e. bio-physical and socio-economic) have been provided by numerous disciplines over time (Verburg et al., 2004; Priess and Schaldach, 2008; Trisurat and Duengkae, 2011).
Therefore, considering the importance of land use analysis in planning and decision making, this paper has given a nutshell but effective depiction of prime land researches undertaken so long to analyze land issues and to predict future problems.
2.1 Land Use Models
Models on land issues and problems range from simple system representations including a few driving forces to simulation systems based on a deep understanding of situation-specific interactions among a large number of factors at different spatial and temporal scales (Verburg et al., 2008; Verburg et al., 2004; Priess and Schaldach, 2008). Moreover, the term “model” in land use research refers to the sign of a system through mathematical, logical, physical and iconic methods (Rui, 2013) which can be categorized in multiple ways on the basis of the subject matter of the models, modeling techniques or methods used or actual uses of the models (Agarwal et al., 2001; Irwin and Geoghegan, 2001; Yang et al., 2008; Veldkamp and Lambin, 2001; Ducheyne, 2003; Torrens, 2006; Timmermans, 2003).
However, modeling methods have been developed to address when, where and why LUCC occurs (Baker, 1989; Riebsame et al., 1994a; Lambin, 1997; Theobald and Hobbs, 1998) to explore and predict the trends (Brown et al., 2000; Trisurat and Duengkae, 2011) especially involving empirical data on historical pattern of changes in land use patterns and then extending those for prediction (Brown et al., 2000). As a result, huge number of models on LUCC has been described over time because of different disciplinary perspectives and methodological approaches based on variations in data availabilities and modeling goals (Brown et al., 2000; Long et al., 2007).
2.2 History and Trends of Land Use Models
Land use and cover change models allow testing the stability of linked social and ecological systems (Oluseyi, 2006) through scenario building and provide valuable information under a range of conditions despite failure of incorporating all aspects of reality (Veldkamp and Lambin, 2001). Thus over time, LUCC modeling has become more integrated, accurate and specialized (Nkonya et al., 2012) to ensure the modeling of ecological interrelationships of different land uses and sustainable development. Baker (1989) published the first reviews in the context of landscape ecology with explicit representation of human decision making but did not discuss models. However, with the passage of time researchers like Von Thünen (1826), Lösch (1940), Ducheyne (2003), Timmermans (2003) and Rui (2013) have used numerous forms theories, models and approaches to explore this issue.
Before mid nineties of last century, spatial economic theory was the base of most land use models (Wang, 2012) while the oldest was Von Thünen’s land rent theory of 1826 (Perraton and Baxter, 1974; Wang, 2012) showing that land close to the city centre is used intensively (Perraton and Baxter, 1974). However, over the last century, numbers of different clear-cut models on land issues have been made (Wang, 2012) especially following the first reviews in this context by Baker (1989). During the last century influential models such as Weber’s classical triangle of industrial location (1909) and Lösch’s theory of economic regions (1940) have also been formulated (Wang, 2012) while following the advances in computational facilities, computer-based urban models (i.e. Lowry model in 1964) arose with the domination of micro-economic theories focusing individual landowners making land use decision with the objective to maximize expected returns from the land (Wang, 2012).
Because of limitation of the then existing methods, spatial dimension was introduced into land use models (Wang, 2012) based on data about landowners’ economic decision and neighborhood conditions from the end of 1980s (Irwin, 2010; Wang, 2012). However, the most representative model of this group is CLUE model which simulates geographical pattern of land uses based on locations (Veldkamp and Fresco, 1996; Verburg et al., 1999; Verburg and Veldkamp, 2001; Verburg and Overmars, 2009; Verbug et al., 2012). Moreover, regression analysis based on various biophysical and socio-economic factors came into use in last century widely (Lambin et al., 2003; Alabi, 2011; Quasem, 2011; Wang, 2012).
2.3 Land Use Modeling Approaches and Models
Studies of land use and its changes over time can be arrayed in a number of dimensions such as theoretical versus empirical; structural versus reduced form; disaggregate versus aggregate; extensive-margin versus intensive-margin studies; drivers versus consequences-orientated studies, policy versus methods-orientated studies (Wu and Li, 2013). However, addressing and sorting all available data, the following shows a little but necessary details of how researchers have tried to deal with various land issues over time to keep pace with evolution and social objectives.
2.3.1 Agent-Based Perspective
Land use being typically based on suitability (Wang, 2012), agent-based models include various simulation models characterized by interacting autonomous agents who have ability to make decisions in changing situation (Parker et al., 2003; Wang, 2012; Oluseyi, 2006). Moreover, agent-based perception is based on general nature and rules of decision by individuals that range from rational decision making of neoclassical economics to socio-behavioral sciences (Lambin et al., 2003; Crooks, 2006). A familiar agent-based model is FEARLUS (Polhill et al., 2008; Wang, 2012).
2.3.2 Systems Perspective
Systems perspective explains changes through organization and institutions of society (i.e. governments, communities) that operate closely at diverse spatial and temporal scales; and is influenced by technical innovations, policy and institutional changes, rural-urban dynamics and macroeconomic changes (Lambin et al., 2003).
2.3.3 Narrative Perspective
Narrative perspective seeks depth of understanding LUCC patterns through historical details and on the same time, interpretation for a specific locality from the historical analyses of land in particular stochastic or non-random but unpredictable events that significantly affect it seriously (Lambin et al., 2003).
2.3.4 The Fitting Data Model
The fitting data model uses, theories of social sciences widely to represent decision making as well as biophysical processes to varying degrees and therefore, helps us understand where, how and why land are changing fast (Brown et al., 2000).
2.3.5 Simulation Processes
Simulation models are generative demonstrations of all essential practices of agent’s decision making based on socio-economic and biophysical settings with the intention of simulating the changes in expected outcome options (Brown et al., 2000).
2.3.6 Structural Models
Structural models are based on well established theoretical background and are being used for hypothesis formulation and to identify variables to be incorporated in a reduced form model based on the implicit assumption (Veldkamp and Lambin, 2001).
2.3.7 Statistical or Reduced Form Models
Statistical models are easier to put into practice because of its ability to deal with original changes in driving forces (i.e. neighborhood land uses, experience) over time in accordance with changes in system properties (Veldkamp and Lambin, 2001).
2.3.8 Geographic Models
Geographic models aims at optimal allocation of lands to ensure the best possible as well as optimal uses with minimal effect on ecosystems and ecology based on suitability of uses and spatial location of population (Nkonya et al., 2012).
2.3.9 Economic Models
Economic models stress on demand and supply of land based commodities and effectively reflect the effect of international trade and globalization on land issues through evaluation policies and socio-economic issues (Nkonya et al., 2012).
2.3.10 Stochastic Markov Model
Stochastic Markov Model combines both the stochastic processes as well Markov chain analysis techniques (Basharin et al., 2004) based on probabilities with discrete state space and continuous parameter space (Balzter, 2000). In this random process, the state of a system(s) at time (t+1) depends only on state of the system at time (t) not on previous states (Ahmed, 2011a).
2.3.11 Ecological Models
Ecological models link land allocation to species abundance and extinction, ecological footprints and other environmental concerns assuming that prices and other economic variables are exogenous factors (Nkonya et al., 2012).
2.3.12 Dyna-CLUE model
The Dyna-CLUE model is a spatial-explicit land use transition model that quantifies the location preferences of different land use patterns based on logistic regression models and determines relations between incidence of a land use pattern and physical as well as socio-economic settings (Trisurat and Duengkae, 2011). It is chosen because it explicitly addresses different future land demands driven by expansion of agriculture, plantation and biodiversity protection (Verburg et al., 2004).
2.3.13 Spatial Economical Model
Patterns and processes of LUCC are essentially spatial processes and gives valuable insights into associated processes and their underlying causes. Spatial economical model emphasizes on maximization of net income in determining the land use patterns of specific area over time (Li, 2002; Xie et al., 2014) and also account for socioeconomic, agro-ecological, geophysical and policy variables (Müller, 2003). Likewise, such models are useful to forecast changes (Serneels and Lambin, 2001).
2.3.14 Cellular Automata Model
Cellular Automata, originally invented by Von Neumann in the mid-1940s, provides a proper scaffold for investigating the self-reproducing features of biological systems (Alabi, 2011; Wang, 2012; Nkonya et al., 2012). They are more powerful for complex systems due to their ability to simulate dynamic spatial processes from a bottom-up perspective (Batty, 2007; Iltanen, 2012) and also for similarity to spatial allocation models in terms of using transition rules (Wang, 2012). Moreover, data from other models such as population growth model can easily be used (Wang, 2012; Li and Yeh, 2000; Santé et al., 2010; Li, 2011) also.
2.3.15 Species-distribution Model
Species-distribution models refer to relationship between given pattern(s) of interest and set of explanatory factors where the factors and associated results can be quantified properly in dynamic ways (Guisan and Zimmermann, 2000).
2.4 Economics, Econometrics and Land Use Research
Economics being the field of dealing with scarce resources; has already made enough involvement in land use and corresponding change analysis (Lambin, 1997; Serneels et al., 2001; Veldkamp and Fresco, 1997; Verburg, et al., 2002). Researches show that outputs are being used to reflect the value of the land use system as well as profit scenario (Dai et al., 2005; Veldkamp and Lambin, 2001) and keeping pace with this, equilibrium principle of microeconomics shows that under the condition of full competition as well as economic and technological stability, marginal benefit (MB) will decrease with the development of the land use system, whereas marginal cost (MC) will increase with demand for land (Houghton, 1994; Dai, 2002). Therefore, area under curve MB is the total benefit of that specific land use system and that under the curve MC is the total cost with expanded land use while E (i.e. as described in figure 2.01) is the point where maximum profits can be made from a land use (Dai et al., 2005). Moreover, rational behavior as well as random utility theory implies that transformations in use of lands are inevitable to maximize profits and to conserve limited resources (Veldkamp and Lambin, 2001; Serneels and Lambin, 2001) in particular when there is a divergence in suitability and target on land use (Dai, 2002; Mia and Islam, 2005). In a purely market oriented economy, a criterion for the transformation of land use type (LTC) can be expressed as (Dai, 2002) a point where land type i will be transformed to type j only and only if land use pattern j generates higher profit than that of i (Dai et al., 2005).
Figure 2.01 Economic Dynamics of Land Use System
illustration not visible in this excerpt
Source: Dai et al., 2005
Moreover with the passage of time, various econometric analyses are also being observed to be used along with economic theories (Lambin et al., 2003; Alabi, 2011). Most common as well as used economic tool used in land use analyses includes regression analysis which refers to method engaged in discovering empirical relationships between binary dependent and several independent categorical and continuous variables (McCullagh and Nelder, 1989). However with the passage of time, there are two basic approaches to assess spatial dependency within the regression framework- firstly, building a complex model known as autoregressive structure and secondly, designing a spatial sampling plot to enlarge distance interval between sampled points (Anselin, 1988). Here is to be noted that discrete choice model is one of the best-known ways of modeling land use patterns as well as changes based on the concept of utility (Koppelman and Wen, 1998) while logistic regression analysis is one of the most utilized approach during past decades (McCullagh and Nelder, 1989; Arsanjani et al., 2013) especially to predict land uses (Verhagen, 2007). When the dependent variable consists of more than two nominal outcomes, it is referred to as Multinomial logistic regression or Logit but in case of two possible outcomes logistic regression is called binary logit and when outcome may be ordered or ranked, ordered logit is being used (Heij et al., 2004; Ntantoula, 2013).
However based on random utility and profit maximization theory, distributions of the discrete states of land cover and use patterns in case of binary analysis can be linked with independent variables by the following equation (Long, 1997; Lambin et al., 2003; Alabi, 2011; Anselin, 2002).
The parameter in above equation represents a threshold and for observations of , the observed binary variable takes the value zero (0) and when , the dependent variable is equal to 1 i.e. land use pattern will be changed into type j. But as dependent variable is unobserved as well as discrete, ordinary least squares estimation (OLS) is not appropriate and therefore, researchers need to use maximum likelihood (ML) method (Long, 1997). ML estimation requires knowledge about the distribution of the error terms and if the error terms are assumed to be normally distributed, then probit model is used for a binary otherwise logit model is applicable (Lubowski et al., 2008; Rui, 2013; Hu and Lo, 2007).
As nations and areas are going towards urbanization rapidly, land use patterns and equivalent changes have gained increased importance by researches throughout the world (Mia and Islam, 2005) especially for sustainable development as well as to ensure optimal use of land and associated resources in more effective and efficient ways (Lambin et al., 2003). Thus developing realistic and dynamic models to explore vital drivers of changes in land use over time has no alternative (Veldkamp and Lambin, 2001). Keeping connection with this Lambin et al. (2003) has also emphasized on the integration of combined perspective for the best, valid and empirical study. Therefore for more accuracy and consistency, land use analyses should include best possible methods collectively (Zenga et al., 2008) with the inclusion of necessary socio-economic and other associated variables (Lambin et al., 2003; Hu and Lo, 2007; Lubowski et al., 2008; Rui, 2013).
Chapter Three Literature Review
About half of the ice-free surface has been substantially modified over last 10,000 years (Lambin et al., 2003) while during last three centuries, nearly 1.2 million sq km of forest lands as well as 5.6 million sq km of grassland and pastures have been converted to other uses (Ramankutty and Foley, 1999). Land use changes, thus, have become locally pervasive and globally significant (Agarwal et al., 2001) as well as dynamic phenomenon (NASA, 2006; Mohammad, 2009) not only for its presence at almost everywhere but also for contribution to global ecology (Houghton, 1994).
People of Bangladesh are observed to shrink per capita land by 50 percent from 1970 and 1990 (Mohammad, 2009) and now have a per capita cultivable land of only 12.5 decimals or less (Quasem, 2011). As a result, with the passes of time land is becoming scarcer (Mohammad, 2009) especially with the growth and expansion of economy (Houghton, 1994; Quasem, 2011; Yadav et al., 2012) and increasing demand for non-farm commodities (Quasem, 2011). Moreover, land use changes have important implications for future changes in the earth climate and ecology (Agarwal et al., 2001) and therefore, understanding land use patterns has great role to facilitate ecological sustainability through improving land management, enhanced capability of assessing and predicting future trends (Veldkamp and Lambin, 2001; Wang, 2012).
3.1 Land
Land, the mother of resources (Mia and Islam, 2005; Iftekhar, 2006), is being considered as a prerequisite for all development purposes especially for sustainable development (Iftekhar, 2006). Land, therefore, refers to the basic natural resource that provides habitat and nourishment for living organisms (Mia and Islam, 2005) or the means for livelihood with potential revenue if properly utilized (Iftekhar, 2006).
Though, Stewart (1968) and Wolman (1987) defined land as the wide range of natural resources from the atmosphere above the land surface down to some meters below the surface, FAO (1992) defined not only as soil but also as landforms, climate and hydrology, plant and animal population, and the physical results of human activity like terraces and drainage works. Moreover, despite the similarity in physical characteristics across the universe (Zubair, 2006), its supports can vary over time and space according to the management conditions and uses (Mohammad, 2009).
3.2 Land Use
Land uses denote the purpose to which human puts land especially to fulfill all their needs (Turner and Meyer, 1991; Turner and Meyer, 1994; Skole, 1994). Moreover, land uses are considered as human activities linked with land, use of its resources (FAO/IIASA, 1993; Veldkamp and Fresco, 1997) which have potential ecological impact because of either permanent or cyclic interference (Vink, 1975). Precisely, land use describes alteration of each land cover (Prakasam, 2010) or how each parcel of land is being managed for alternative uses (FAO, 1992).
Land use, thus, is applied to the biophysical attributes of surface (Lambin et al., 2001) through various human induced activities (Prakasam, 2010) for different purposes i.e. habitation, forestry, agriculture (Ahmed, 2011; Yadav et al., 2012).
3.3 Land Use Change
Land use change is being considered as the single most important appearance of human interaction on atmosphere (Mohammad, 2009) and includes alteration of land covers (Lesschen et al., 2005) either in the form of agricultural intensification or changes in farming system over time (Farrow and Winograd, 2001) due to influence of population and economic expansion (Mohammad, 2009). Briassoulis (2000) has defined land use change as the quantitative increases or decreases in the area of a given type of land use while Wu and Li (2013) defined as any changes in arrangements, activities and inputs that people undertake in certain land cover type.
Precisely, land use change refers to changes in land use morphology over time with respect to particular socio-economic factors (Grainger, 1995; Zubair, 2006) which may include both temporal and spatial dimensions (Long et al., 2007).
3.4 Land Cover
Land cover is the most vital gears of ecology (Prakasam, 2010) attributable to functioning of ecosystem (Yadav et al., 2012). Meyer (1995) defined land cover as the kind and state of vegetation (e.g. forest or grass cover) but Zubair (2006) has widened the definition by including factors such as human structures, soil type, biodiversity and ground water. Land cover, thus, refers to assemblage of biotic and abiotic components on earth surface (Prakasam, 2010; Uddin and Gurung, 2010) or the set of spatial units each associated with attributes (Lambin et al., 2003).
Precisely, land cover can be described as the layer of soils and biomass that covers land surface (Fresco, 1994) with biota, soil, topography, surface, groundwater and human structures (Lambin et al., 2003) which together denotes the quantity and type of surface vegetation, water and earth materials (Turner and Meyer, 1994).
3.5 Land Cover Change
Land cover change refers to either changes in biophysical attributes (Lambin et al., 2001; Dale et al., 2000) or complete replacement of one cover type by another alternative (Lesschen et al., 2005). Precisely, it is the ultimate changes of the nature of soils, vegetation and water surfaces (Houghton, 1994; Wood et al., 2004) causing environmental modifications (Klooster and Masera, 2000; Mas et al., 2004).
3.6 Land Use and Cover Change
Land use and cover are separate terms often used interchangeably (Dimyati et al., 1994; as cited in Yadav et al., 2012) though are semantically equivalent (Brown et al., 2000) for their historic nature (Dale et al., 2000). However, together they refers to the likely changes in land cover with or without unaltered existing land uses (Turner and Meyer, 1994; Tiwari and Saxena, 2011) either directly or indirectly (Prakasam, 2010) from the interdependence between socio-economic, institutional, bio-physical, cultural and environmental forces (Lesschen et al., 2005).
3.7 Land Use Planning
Land is influenced by personal, economic, cultural, political and historical factors (Brown et al., 2000) and is used first and foremost for agriculture, industrial communication and settlement purposes (Mohammad, 2009). Therefore, coherent set of decisions about the use of land and ways needed to achieve the desired use and to ensure optimal productive capacity are the core of land use planning (FAO, 1992; Mia and Islam, 2005). Moreover, such planning shows fraction of total available lands for further uses either in productive or non-productive uses (Houghton, 1994).
3.8 Land Use Conflict
Nations advancing towards development, urbanization and industrialization face major land use conflicts in the form of converting valuable agro land to non-agro uses (Mohammad, 2009; Mia and Islam, 2005) despite the uniqueness in cover and attributes of each parcel of land (Zubair, 2006). About 1 to 2 million ha of croplands is being taken out of production every year in developing countries to meet demand for non-productive purposes (Houghton, 1994; Lambin et al., 2003). Moreover, most of the lands in Bangladesh are fit for more than one use (Mia and Islam, 2005) which leads to the diversified uses of limited land (Islam, 2000) causing acute conflict mostly between shrimp farming and other uses (Mia and Islam, 2005). Land use conflicts are acute under rapid population pressure and in mixed economies (Verheye, 1997) due to clumsy action among concerned parties (Mohammad, 2009).
3.9 Methods Used to Identify Patterns and Changes of Land Use and Cover
Land use research is devoted to analyze relationship among land use pattern, socio-economic as well as biophysical variables (Lesschen et al., 2005) that act jointly as driving forces and can be understood through monitoring and analyzing the trends regularly (NASA, 2006). As a result, researchers have used various methods based on existing data, techniques and facilities (Lambin et al., 2003) to explore the various land use patterns and corresponding changes over time and place.
Scientists and environmentalists have identified fast changing magnitude of land use patterns and corresponding changes across earth by observing and analyzing satellite images (Loveland et al., 1999) though have poor application especially in developing nations (IPCC, 2000). Despite all drawbacks, Mas et al. (2004) used map comparison based on GIS while NASA (2006) as well as Kamaruzaman and Manaf (1995) has used landsat satellites to explore changes through monitoring and analyzing data. Tefera and Sterk (2008) and Yadav et al. (2012) used satellite images and maps using GIS to analyze land use dynamics while Trisurat and Duengkae (2011) used Dyna-CLUE model with logistic regression and Xie et al. (2014) used spatially explicitly regression to describe economic drivers of agro land use change.
Brown et al. (2000) has used ‘Transition Probabilities’ while Veldkamp and Lambin (2001) have used a spatially explicit, integrated and multi-scale manner for the projection of alternatives into the future to test key processes and for describing the trends in quantitative terms. Lambin et al. (2001) used simple but elegance theme called ‘IPAT formulation’ showing interdependencies among population, affluence and technology. Ruben et al. (2008) used optimization models (Cost-benefit analysis based on opportunity cost of using or converting specific parcel of land at a specific time) of the agriculture and forestry sectors. Lubowski (2002) used econometric analysis through formulating Nested Logit model to include all major land use categories in both urban and non-urban land uses and examines a comprehensive set of transitions among the different land use categories. Lambin et al. (2003) have used regression to address land use as well as their changes while Lesschen et al. (2005), Alabi (2011) and Quasem (2011) have used empirical techniques to verify hypotheses through the application of statistical and econometric tools like goodness of fit, regression analysis, correlation analysis and descriptive statistics to predict actual landscape change. Zhang et al. (2001) used regression analysis with cross-sectional heteroscedasticy and simultaneous correlation analysis.
Mia and Islam (2005) in November 2004 used ‘Ground Truthing’ (an important aspect to check information incorporated in zoning exercise) to check land use patterns and their changes over time in southern part of Bangladesh while Uddin and Gurung (2010) used satellite remote sensing in Bangladesh with the use of change detection map (spatial location of changes) and change matrix (dimension of changes). Ahmed (2011a) have widely used Remote Sensing and GIS techniques to assess natural resources and environmental changes using time series of remotely sensed data and linking it with socio-economic and bio-physical data in Khulna city to detect, monitoring and mapping land cover change over time and hot spots. Rahman and Begum (2011) used remote Sensing and GIS Application to address the land use changes in Sundarbans areas in Khulna and Satkhira region.
3.10 Variable Used in Modeling Land Use and Cover Changes
Models of land use analysis are powerful tools to be aware of and analyze important linkage between socio-economic processes (Lesschen et al., 2005) linked with land and resource management and agricultural activities (Turner and Meyer, 1991; Brown et al., 2000). However, modeling land use change initially focuses on biophysical attributes (Veldkamp and Lambin, 2001) with various socio-economic drivers (Wilbanks and Kates, 1999). Therefore, researchers on the basis of accessible data, techniques and problems have used different variables as described below.
Ehrlich and Holdren (1974) and Lambin et al. (2001) used population, affluence, technology as variables despite an interdependencies and high risk of their separation while Quasem (2011) has shown total land (decimals), homestead land (decimals); proportion of non-crop land to total land owned (%), primary occupation and years of schooling (number); per capita annual income (Tk.); household assets other than housing (Tk.); disaster losses (Tk.). Agarwal et al. (2001) and Lambin et al. (2003) used population density, labor availability, quantity and sensitivity of resources, production costs, market prices, transportation costs and technology, subsidies, taxes, property rights, infrastructure, exposure to external perturbations while Alabi (2011) and Trisurat and Duengkae (2011) used elevation, soil type, income, proximity to near roads, water sources, infrastructure, drainage system, population density, road condition as major variable to quantify land use change.
3.11 Type and Scope of Land Use and Cover Change
Growing demand for urbanization as well as suburbanization is asking for frequent alteration in using the planet surfaces in diverse ways (NASA, 2006) and as a result, land use changes can be considered from two perspectives such as intended and unintended (Houghton, 1994) or progressive and gradual (Lambin et al., 2003) or reversible and irreversible (Islam, 2000). However, Lambin et al. (2001) have pointed out that about 26 researchers of various disciplines have worked on several issues of land use changes including tropical deforestation, rangeland modifications, agricultural intensification and urbanization supported by quantitative assessments with a deeper and more robust understanding of land use pattern and change especially to adopt appropriate policy intervention.
Moreover, land use includes agricultural land, built up land, recreational area, wildlife management area (Zhang et al., 2001; and Prakasam, 2010) and its changes may involve shifting to a different use (i.e. from rice to built-up land) and/or expansion or intensification of an existing one (Morita et al., 1997).
3.12 Observed Land Use Pattern
Land use and cover changes have historical sets since civilization (Dale et al., 2000) due to growing trends of urbanization and innovation (NASA, 2006). The most observed and important human use of land includes agriculture, settlements, forests, water bodies, fisheries, salt production, industrial with infra-structural developments and tourism (Turner II et al., 1994; Mia and Islam, 2005; Mohammad, 2009; Islam, 2000; Iftekhar, 2006), mixed uses restricted and vacant land (Iftekhar, 2006). However, lands in south-west Bangladesh are being observed to be used for rice farming, shrimp cultivation and fish farming, forestry, salt production, ports, industries, human settlements and wetlands with some fellow lands (Alam et al., 2002; Islam et al., 2006; Mia and Islam, 2005; Flynn et al., 2009).
3.13 Global Land Use and Cover Trends
Major and historical changes in land use across the world occur since humans have controlled fire and domesticated plants and animals (Lambin et al., 2003) and especially with the growth of population and urbanization (Dale et al., 2000). Moreover, about half of the ice-free surface has been substantially modified by human activities over last 10,000 years (Lambin et al., 2003) while approximately one-third of the land surface were being converted to alternative uses (Houghton, 1994). Estimation shows that 10-15 percent of the transformed land surface is dominated by agricultural crop and urban-industrial areas while 6-8 percent is pasture (Vitousek et al., 1997). According to Ramankutty and Foley (1999), during the last three centuries, global cropland has increased by 12 million sq km.
3.14 Land Use Trends in Bangladesh
Bangladesh has a population of 153 million with an expected increasing rate of 1.37 percent (MoF, 2013) causing direct conversion of productive lands into non-productive uses (Mia and Islam, 2005). In last century, only 23 percent of total land area was cultivated by tenants or owner cum tenants and 45 percent by paid laborers (Hasan and Mulamottil, 1994). Mohammad (2009) showed that land has decreased by about 50% during 1970-1990 while arable land per economically active person is only 0.8 ha compared to more than 12 ha in developed countries (Graff, 1993; as cited in Mohammad, 2009). Moreover, land demand for non-agricultural purposes and urban uses has increased sharply in last decades though still agriculture is the major activity (Choudhury, 1987; as cited in Mohammad, 2009). Consequently, despite much fertile land Bangladesh is marginally deficient in food grains (BBS, 2006).
Trends of land use patterns in south-west part of Bangladesh are notable over last decades due to her major land uses (i.e. agriculture, shrimp and fish farming, forestry, urban development and settlement) and especially due to rising demand and huge populations in corresponding areas (Ahmed, 2011; Rahman and Begum, 2011).
3.15 Causes of Land Use and Cover Change
Land use changes can be described by the complex interaction of behavioral and structural factors (Verburg et al., 2004) which are driven by a combination of the so called land use drivers classified as socio-economic, political and biophysical factors (DeKonind et al., 1999; Stomph et al., 1994; Veldkamp and Fresco, 1997) along with some recent one like climatic and demographic factors, level of poverty and economic as well as institutional structure of the resource use (Mohammad, 2009). Therefore, driving forces are generally subdivided into two groups- proximate causes (Activities or actions that directly affect land use) and underlying causes (Fundamental forces that underpin the proximate causes including demographic, economic, technological, institutional and cultural factors) (Lesschen et al., 2005).
Researchers over time have pointed out numerous causes such as rapid growth and development of civilization (NASA, 2006), population and demands of food resources (Yadav et al., 2012), population and poverty driven deforestation, increased presence of shifting cultivators, triggering mechanisms for rapid development, globalization, low per capita land (Lambin et al., 2001), dam construction (Tefera and Sterk, 2008), economic growth and development, climate change, development of roads and electricity, improvements in irrigation, technologies, penetration of commercial forces (Uddin and Gurung, 2010), consumer tastes, international trade, weather, local rules (Lubowski et al., 2008), desire for profit, utility maximization, cost minimization, (Veldkamp and Lambin, 2001), soil suitability, population density, rainfall and accessibility, market conditions (Lesschen et al., 2005), increasing income, urbanization, infrastructural development, national and international policies, land tenure and property rights, bio-energy, land degradation (Nkonya et al., 2012), soils erosion, reduced rainfall, floods and siltation (Houghton, 1994), land ownership, non-agricultural occupation (Quasem, 2011), fertility (Mohammad, 2009). However, according to the words of Iftekhar (2006) land use change occurs because of the combined effect of social, political and economic conditions of a region or a country.
During past few decades Bangladesh has experienced rapid land use changes more or less for the above stated causes (Ahmed, 2011; Iftekhar, 2006; Mohammad, 2009) while south-west regions are being observed to have frequent changes due to the effects of increased salinity intrusion as well as natural disasters (Ahmed, 2011), intensive agriculture practices and changing land quality (Uddin and Gurung, 2010; Minar et al., 2013). However, Rahman and Begum (2011) showed two causes of land use changes in Khulna and Satkhira region such as natural (i.e. global warming, climate change, sea level rise (SLR), coastal flood, salinity intrusions, water logging) as well as anthropogenic forces (e.g. population growth, unplanned cultivations, salinity intrusions, water logging, misuse of Sundarbans, political unrest, illiteracy of local people about effect of land cover changes, poverty, higher expectation).
3.16 Impact of Land Use and Cover Change
Land use changes have come into view as one of the key drivers of ecological changes (Kueppers et al., 2004; Foley et al., 2005; Serneels and Lambin, 2002) because of its potential effect of causing various sudden but catastrophic environmental and socio-economic problems (Wang, 2012; Mia and Islam, 2005). Human use of land has altered structure and functioning of ecosystem (Vitousek et al., 1997) and keeping pace with this IPCC (2000) stated that expansion of agriculture have came into present form through conversion of forests and grassland during past 140 years. Kitamura and Kobayashi (1993) and Houghton et al. (1999) have pointed out that wrong land use has led to serious problems such as degradation and deforestation of tropical forests, climate change with the problems of greenhouse effect, loss of biodiversity and negative changes in regional hydrology and biogeochemical cycles (Chase et al., 1999; Mas et al., 2004).
However, researchers have pointed out some of the frequent impacts of land use and cover changes such as rapid conversion of potentially productive land to unproductive purposes (Houghton, 1994; Lambin et al., 2003), change in biotic diversity (Sala et al., 2000), important tradeoffs for sustainability, food security, vulnerability of people and ecosystems (Lesschen et al., 2005), deforestation, diminishing soil fertility, permanent degradation of land productivity (Islam and Weil, 2000), inundation of grazing lands, soil erosion, reduction of traditional farming, sedimentation (Tefera and Sterk, 2008), climate change, deforestation, natural hazards (NASA, 2006; Lubowski et al., 2008), climate variability, land degradation, vulnerability of places and people (Veldkamp and Lambin, 2001).
Here is to be remembered that all impacts are not negative because changes in land use patterns are also associated with increases in food and fiber production with more efficiency and well-being (Lambin et al., 2003; Vitousek et al., 1997) despite its externalities (Turner II et al., 1995; Lambin et al., 1999; Aylward, 2000).
3.17 Initiatives for Land Use and Cover Changes
Growing importance of land use and its policies has been approved by several international meetings (i.e. The World Forestry Congress, The Jakarta Declaration 1978 and Paris Declaration) through holding seminars and symposiums over time with the incorporation of socio-political and economic factors (Fresco et al., 1996; Veldkamp and Lambin, 2001). Recognizing the significance of land use issues, globally projects were prepared in 1994 for the first time (Verburg, 2006; Veldkamp, 2009; Wang, 2012) especially aiming at sustainable economic expansion and environmental protection (Wu and Li, 2013). Moreover, considering pervasive externalities of land use changes, a novel discipline named land use science has already emerged (Lubowski, 2002; Wang, 2012).
In recent years, significant progresses have been observed in land use planning in Bangladesh mainly in mapping shrimp and rice farming lands (Shahid et al., 1992), detection of changes in Sunderbans mangrove forest (Islam et al., 1997), shrimp-farming zone (Hossain et al., 2001), mapping suitable areas for saltpan development (Hossain et al., 2003a), mangrove afforestation (Hossain et al., 2003b), tilapia farming areas (Hossain et al., 2007), assessing suitable carp-farming areas (Hossain et al., 2009; Salam et al., 2005) and giant prawn farming area (Hossain and Das, 2010).
3.18 Findings and Results of Land Research
Land use and cover changes are extensive, accelerating and significant process driven by human actions (Xie et al., 2014) and also have influential effects on human activities (Agarwal et al., 2001). Moreover in most societies, use of land is more or less out of the owners’ hands and under the control of government or local authorities though their involvements vary much across time, region and culture (Kim, 2010; Ahmed, 2011a). Besides, when there is competition for residential land it is observed that financially deprived people are relegated to poor and bad terrains (Alabi, 2011) and agricultural intensification occurs at the intensive margin when more input is used for a given land or when a less input-intensive land is converted to a more input-intensive use i.e. conversions of grassland to crop production (Wu and Li, 2013).
Researchers over time have used various different methods on the basis of existing data, techniques and facilities (Lambin et al., 2003; Li and Zhao, 2011; Xie et al., 2014) and show that low income, low elevation and inefficient geography have negative effect on residential development while is induced through favorable ecological characteristic e.g. favorable road network, nearness to modern amenities and facilities (Skole and Davids, 2002; Gyawali et al., 2004; and Alabi, 2011). Lubowski (2002); Lubowski et al. (2008) and Alabi (2009) found that residential and industrial areas are now sited on areas which were once prime agricultural lands, wet lands and areas of physical constraints due to scarcity of land and found a significantly positive relationship with proximity to infrastructure while significantly negative relationship with elevation, road condition and population density and didn’t indicate any notable relationship between drainage, education, land price, soil type or flood potential. Rui (2013) showed higher value of commercial, industrial and public service areas than that of pasture and forest area. Built-up areas and urban greenbelts display positive relations with different centralities while agro and forest areas show negative relationships (Riebsame et al., 1994; Zubair, 2006; Lubowski, 2002).
3.19 Problems and Limitation of Land Use and Cover Researches
Unavailability of better data for improved models and projections of land use and cover changes especially to make a generalized conclusion (Lambin et al., 2001; Ochoa-Gaona and Gonza´lez-Espinosa, 2000; as cited in Mas et al., 2004) together with ignorance and misunderstanding about the cost and benefit of cropping or any other uses (CGCR, 1999; Oluseyi, 2006) is the major problems in dealing with land issues. Moreover, Lambin et al. (2001) and Long et al. (2007) have addressed the problem of application of micro scale data sets in global context because they are specific to time and place and have some common and popular myths regarding land use changes. Lesschen et al. (2005) and Lubowski et al. (2008) have pointed out that the misuse of different techniques described without a specific focus on land use change issues causes much probability of uncertainty in modeling land issues. Proxy variables, though easier to measure spatially complex variables (i.e. land management technologies, infrastructures and policies) generate acute problems in application of such results in policy makings (Wilbanks and Kates, 1999; Müller, 2003).
Land use pattern and corresponding changes have vital implications for future changes in earth climate as well as ecology (Agarwal et al., 2001; NASA, 2006) mainly in developing countries where per capita arable land is lower in contrast to that of developed countries (Graff, 1993; as cited in Mohammad, 2009). Moreover, changes in land use patterns occur not only for negligence and improper execution of land use policies but also for some misconceptions (Lambin et al., 2001). Researches also shows that despite accuracy and success of remote sensing data and GIS (Lesschen et al., 2005), these are rarely being used especially in developing nations (Ahmed, 2011a) and if used, the result of such studies on land use changes are placed in complex ways which shows variation from researchers to researchers because of geographic, demographic and climatic variations (Uddin and Gurung, 2010).
As a developing country Bangladesh lacks a well organized database both in national and regional levels as a result of improper coordination among different organizations (Oluseyi, 2006; Mohammad, 2009) and thus despite being a powerful tool, use of satellite image is limited here (Ahmed, 2011a).
3.20 Research Gap
Relationships between population increase, economic developments and land use changes have generated sufficient research interest recently (Agarwal et al., 2001; Oluseyi, 2006) but little has been done in predicting long term penalties in developing nations (Quasem, 2011). Though there are some researches in developed countries to check relationship of land use patterns as well as their changes with sustainability, smooth economic expansion; there has hardly any study in the area of conversion of farm land to non-farm uses in developing nations (Quasem, 2011; Ahmed, 2011a).
However, from the literature collected and discussed above shows that there occurs very little research on land use issues in south-west areas especially in Khulna and Satkhira areas where both natural as well as human induced forces are responsible for land use changes over time. Moreover, there is only some govt. information collected over time on land use and its changes at household level but there are enough gestation periods between data collection and publishing. Again despite being crucial, land use change is not taken into consideration significantly on national land policy and other policies where lands are used intensively. As a result, there are enough spaces for research on land use issues especially to know the extent of land use patterns and their corresponding changes in south-west region of Bangladesh.
Any activity (i.e. known as driver or determinant) associated with land use may be on side the causes and on the other side the result of changes in land use patterns and processes (Agarwal et al., 2001). Therefore, whatever is the planning or policies, success depends much more on the proper implementation of the policies which needs the establishment of integrated management through coordination, demarcation, better preparedness against adversity and introduction of modern land management systems (Ahmed, 2011). It is also to be noted that neither policies nor government regulation can ensure sustainable land use until the mass people become aware of the social cost and benefit of various alternative land use patterns and corresponding changes.
Chapter Four Methods and Materials
As this paper has already been described the rationale of the problem, objective as well as research question of the study (Chapter one), this chapter by this time describes all other necessary steps followed since research problem formulation to successful completion of the research work as follows.
4.1 Conceptualization of the Research Problem
After selecting the broad research area for investigation, search for and then reviewing of collected literature form offline (i.e. library, newspaper) and online sources (i.e. websites) are being made continuously for conceptualization of proposed problems as clearly as possible. Here the author has collected information with higher emphasizes on modeling and econometric issues (i.e. for clear and easy modeling of current study) as well as empirical analysis (i.e. for comparable findings) which have by now been discussed in chapter two and three. Moreover, the author has also concerned with resource persons for clear conceptualization on proposed problem.
Details but necessary information on different concepts, theories as well as their modeling approaches and findings over time, place and culture have been collected from previous studies such as books, journals, seminar papers, dissertations, organizational papers and various websites (i.e. outlined in reference part in details).
4.2 Study Area
Keeping pace with the title of the research work as well as after the process of conceptualization (i.e. developing theoretical as well as conceptual framework), the researcher has selected the study area to answer the research questions and compare with the existing findings in an empirical process. The author has used multi stage sampling process to select final study area within the south-west region and primarily, Khulna division, one of the seven divisions and the most influential coastal zones (Ahmed, 2011) of Bangladesh, has been chosen as the broad study area. After that, Satkhira districts out of 10 districts of Khulna division and then Kaligonj Upazila of Satkhira district have been selected conveniently as the study area. Finally, Pirozpur village (i.e. details in Chapter Six) of Dhalbaria union under Kaligonj upazila is being selected as the sample study area to collect data for empirical analysis.
4.3 Research Design
To keep pace with the objectives, author has proposed both exploratory and explanatory approaches in the study to address and then discuss the land use patterns as well as their corresponding determinants both in qualitative and quantitative approach. However, following Lambin et al. (2003); Parker et al. (2003); Oluseyi (2006); Torrens (2006); Polhill et al. (2008); Carrión‐Flores et al. (2009); Wang (2012) and Rui (2013), author has attempted to model land use conversion reasonably from a rich available literature emphasizing on the economic agent who is assumed to make an inter‐temporal, profit maximizing choice regarding the conversion of a parcel of land to some available but towards the most persuasive alternative use.
Moreover, author has used joint approach of various models to show link between changes in land use patterns (i.e. conversion of rice farming lands towards shrimp) and socio-economic, bio-physical, policy variables by following Verburg et al. (2004) and Trisurat and Duengkae (2011) on Dyna-CLUE model; Serneels and Lambin (2001); Müller (2003); Li (2002) and Xie et al. (2014) on Spatial Economical Model and Li and Yeh (2000); Batty (2007); Santé et al. (2010); Alabi (2011); Li (2011); Iltanen (2012); Wang (2012) and Nkonya et al. (2012) on Cellular Automata.
Therefore, agent based approach is being used based on single survey from the land owners or decision makers while some of the necessary but previous data (recall data) are being collected for the proper completion of the research.
4.4 Target Group
Agent based approach is based on rational agents who emphasize on profit maximization in choosing conversion of a parcel of land (Parker et al., 2003; Wang, 2012; Oluseyi, 2006). Hence for convenience of the study, households of the selected study area have been primarily treated as the target group while head or decision making individual of the each household is being taken as individual agent. It is to be noted here that households (i.e. respondents) who are living at least for five years in the study area are only being considered as the target sample population.
4.5 Sample Design
The author in this paper has used multistage sampling in selecting both study area and sample population. However, the author has used the following procedures for sampling technique, sample size and sampling methods (Next page).
4.5.1 Sampling Techniques
Systematic and stratified random sampling are the two agreed upon sampling methods in logistic regression (Arsanjani et al., 2013) because of its ability to reduce spatial dependency and complete pictogram of population (Huang et al., 2009). Hence, following Xie et al. (2005), the author has used systematic random sampling technique and during the survey the author had selected an initial point randomly (e.g. household) in the study area and then has visited each tenth (10th) household systematically for data collection. It is to be noted that when the respondent selected was found to be landless especially if no land even for household, then the author has taken next household as the sample for convenience. Here, head of each sample household (i.e. those living in the study area for at least five years) is treated as the sampling unit to conduct the research work.
4.5.2 Sample Size
As the total population (e.g. households) is not available in hand, the author has used systematic random sampling technique to collect data from a total of 80 households e.g. each 40 households engaged in rice and shrimp farming respectively in the study area. Here each group (i.e. both rice and shrimp farming households) is engaged in respective occupation at least for five years while sample shrimp farmers have changed from rice farming to shrimp farming at least five years ago.
4.5.3 Data Collection Method
After the selection of sample size and sampling technique, a semi-structured questionnaire (Appendix I) is being used during the interview session for data collection from target groups. Moreover, face to face interview (i.e. FGD) technique has been used for data collection from the local authorities and old persons of the study area. It is to be pointed here that author has used open ended as well as unstructured questions to have the FGD.
4.6 Type of Data Used
To achieve the objective, this paper has been prepared based on cross-sectional data primarily collected through a single survey from each respondent of selected area. However, here some of the necessary but previous data have also been collected from the households, local authorities and organizations for the completion of the research. Though primary data constitutes the heart of the study, some sorts of secondary data (e.g. time series data) are also being collected from necessary sources for more accuracy and validity of data and complete presentation of the research.
4.7 Variables and Indicators
Being an agent based approach to identify the existing land use pattern and their changing trends; author has used profit maximization theory and logistic regression in this study. Moreover, to reduce complexity and to ease interpretation, the author has chosen rice and shrimp farming land as two of the major land use pattern for subsequent econometric analysis as the dependent variables. Moreover, rice and shrimp farming land are being denoted by zero (0) and one (1) respectively where zero (0) means no change in land use (i.e. land is yet being used as rice farming) while one (1) means land use pattern has already shifted from rice farming to shrimp farming. On the other hand to trace out the extents of the determinants of land use patterns, influential socio-economic, cultural and bio-physical factors and decision variables (Table 4.1) are being treated as control variables.
Table 4.1 Description of Independent Variable
illustration not visible in this excerpt
Source: Author’s Compilation, 2014
4.8 Model Specification
This sub-section of methodology describes the best fitted econometric model of land use pattern as well as the corresponding process how parameters are to be estimated using the empirical data in following ways.
4.8.1 Logistic Regression for Land Use Change
Before land use modeling it is to be noted that discrete choice models are based on random utility theory which assumes that decision makers use their land in the form of optimal (i.e. land use pattern that gives highest return) alternative(s) and the decision-makers have perfect discriminating capability. Moreover, the author has used logistic regression because of binary or categorical nature of dependent variable and lack of normality in the distribution of error term while independent variables are mixture of continuous and categorical variables.
We have already discussed (Chapter Two) that logistic regression technique yields coefficient for each independent variable based on a sample of data and also identify the role and intensity of explanatory variables in the prediction of the probability of one state of the dependent variable (i.e. defined as a categorical variable ). Broadly, suppose is a vector of explanatory variables and p is the response probability to be modeled with, in the case of a dichotomous dependent variable, , with meaning rice farming land and meaning the presence of shrimp i.e. more critically land is converted from rice to shrimp farming. Therefore, the general linear logistic model may be as follows.
Here is the intercept and are slope parameters. The probability values can thus be quantitatively expressed in terms of explanatory variables by However, odds ratios are used to facilitate model interpretation as it is a measure of association which approximates how much more likely (or unlikely) it is for the outcome to be present for a set of values of independent variables (Serneels and Lambin, 2001). The probability, the odds and the logit are three different ways of expressing the same thing (Menard, 1995) which are computed as exponential of the parameter estimates (Serneels and Lambin, 2001) and be expressed as follows.
In this study, logistic regression technique is being performed using the logistic function in the STATA software while maximum likelihood estimates (MLE) are being used here for model estimation. Positive values of the parameter estimate indicate that larger values of the explanatory variable will increase the likelihood of the occurrence of the event while negative values indicate that larger values of the explanatory variable will decrease the likelihood of the occurrence of the event. The χ 2 statistic indicates the relative weight of each explanatory variable in the model and allows us to assess the role of each variable in the prediction of an event. In the case of logistic models, the goodness-of-fit measure is defined as the ratio of maximized log likelihood while pseudo- R 2 or ρ 2 is defined as follows.
Although ρ 2 ranges in the value from 0 to 1, its value tends to be considerably lower than the value of the coefficient of determination R 2 of conventional regression analysis. It should not be judged by the standards of what is normally considered a “good fit” in conventional regression analysis (Serneels and Lambin, 2001).
4.8.2 Empirical Analysis of Land Use Determinants
Keeping pace with above description, author has tried to formalize an econometric model with predetermined determinants to generate their impact on land use pattern (i.e. rice and shrimp) and their changes over time as follows.
Here, denotes the dependent variable; is a constant term while are the coefficients to be estimated and is the error term. The details are being enumerated in the next table.
Table 4.2 Explanation of Variables in Empirical Analysis
illustration not visible in this excerpt
Source: Author’s Compilation, 2014
Here is to be noted that in case of major land use patter rice farming land is the reference category while tradition and belief, nuclear family, joint land ownership, dissimilar neighborhood land use, moderately accessible, no credit availability, no natural pressure are treated as reference category in case of engagement on current land use pattern, family type, land ownership pattern, neighborhood land use pattern, accessibility, availability of credit and natural pressure on current land respectively.
4.9 Data Collection
This study has adopted data from both secondary as well as primary sources. Here data form secondary sources (i.e. land use change in the world as well as Bangladesh, its scenario over the past years, policies on land use, pattern of urbanization, incentives for land use change and major macro impacts of land use) have been collected especially for conceptualization as well as to strengthen the discussion of the thesis. On the other hand, primary data through direct contract with the respondents have been collected to analyze and compare the findings of the research with the existing body of knowledge. However, three types of data were being used in this study which is national level data, local level data and household level data as described below on the basis of sources.
4.9.1 Primary Data Collection
A household survey was conducted to get data about land use patterns, needs and demand for land at micro level. In general, three methods have been used in collecting data from the sample population of study areas. Firstly, focus group discussions (FGD) were being conducted during the field study period for overall conceptualization on proposed field from the survey. Secondly, questionnaire survey was being conducted through a pre-tested but semi-structured questionnaire in the study area to assess the land use patterns and the role of different determinants. And thirdly, data has also been collected through monitoring of the farms and households about overall present land use information. Moreover, data have also been collected in from the authority i.e. chairman, member (local representative); govt. officials such as agricultural and fishery officers; organizations both govt. and NGOs.
4.9.2 Secondary Information
Secondary information and data were collected from Space Research and Remote Sensing Organization (SPARRSO), Forest Department (FD), Department of Agriculture Extension (DAE), Department of Fisheries (DoF), Department of Livestock Services (DLS), Bangladesh Water Development Board (BWDB), Bangladesh Agricultural Research Council (BARC), Soil Resources Development Institute (SRDI), International Union for Conservation of Nature (IUCN), Bangladesh Meteorological Department (BMD). Among the NGOs, information was collected from Bangladesh Resource Center on Indigenous Knowledge (BARCIK), Coastal Environment Conservation Center (CECC), Shushilon, Uttaran and various other wings of GoB. Moreover, various published and unpublished documents are also being reviewed for necessary data on the proposed field in recent years.
4.10 Data Processing and Analysis
After collection, data have been categorized and arranged according to their nature and type using Microsoft Excel, SPSS and STATA software for further analysis. Then, STATA as well as SPSS program and some manual procedures have been used to analyze the data already in hand to achieve the objective of research.
However, data have been analyzed using statistical tools like correlation, regression and dispersion analysis to present the results both in descriptive as well as in quantitative ways. Moreover, analyzed results are being interpreted using some of the common but well established economic theories associated with the proposed variables in terms of relationship.
4.11 Writing the Research Paper
After the sorting of raw data and completion of necessary analysis, results are being illustrated with the help of graph, tables, figures, charts and mostly through descriptive statistics. Research paper and associated analysis have been revised several times before the final submission to concerned authority.
A combined method of land use analysis is being used to complete the proposed research work while relevant data for describing land use patterns as well as corresponding changes are being collected directly through field survey using a combined method of questionnaire and interview including both structured and open-ended questions. The methodology adopted for the present study also makes extensive use of secondary material to build up and support the objectives as well as findings of the study.
Chapter Five Land Use Patterns and Changing Trends
Land use patterns and their changes over space and time being our main concern, this chapter describes global as well as national and local land use patterns and their changing trends based on secondary data. Here is to be noted that we have already summarized the major determinants of land use patterns and equivalent changes based on secondary survey (Chapter Three).
5.1 Global Land Use Patterns
Two important drifts are evident over last century- firstly, total lands devoted to human uses (e.g. settlement, agriculture) has increased radically; and secondly, increased production of goods and services has intensified both use and control of lands (Dale et al., 2000). Since early periods of civilization, about 30% lands were being used for cropping and rest 70% as permanent pastures which together comprise approximately 32% of earth (Houghton, 1994). But, historical changes in global land use patterns have increased total agro land whereas approximately one-third of the global land surface is devoted to croplands or pastures (FAO, 2001). Since humans have controlled fire and domesticated plants and animals, they have cleared forests to wring higher value (Lambin et al., 2003). Recent estimation also shows that undisturbed areas characterize 46% of earth’s total surface (Mittermeier et al., 2003) while recent forests covers only 30% which was 50% before 8000 years (Ball, 2001).
Agriculture has expanded into forests, savannas, and steppes in all parts of the world to meet the demand for food and fiber keeping pace with development of civilizations, economies and increasing populations (FAO, 2001). Global cropland has enlarged from 300–400 mha since 1700 to 1500–1800 mha in 1990 (Ramankutty and Foley, 1999) while area under pasture increased from around 500 mha since 1700 to about 3100 mha in 1990 (Goldewijk and Ramankutty, 2003). These increases led to decreases of forests from 6200 mha since 1700 to 4300 mha in 1990 (Ramankutty and Foley, 1999). Steppes, savannas and grasslands also experienced a rapid decline from around 3200 mha in 1700 to 1800 mha in 1990 (Lambin et al., 2003).
Moreover estimation also shows that 1-2 mha of cropland are being taken out of agro production per year in developing countries to meet land demand for housing, industry, infrastructure, and recreation (Lambin et al., 2003). Europe, Indo-Gangetic Plain and China experienced the most rapid cropland expansion during the eighteenth century while newly developed regions of North America and former Soviet Union in early nineteenth century (Goldewijk and Ramankutty, 2003). A very gradual cropland expansion occurred in Africa, south and South-east Asia, Latin America and Australia until 1850s, but since then these regions have observed dramatic increases mainly at second half of 20th century (FAO, 2001; Ramankutty et al., 2002).
On the basis of above description it may be concluded that land uses are changing since civilization especially to cope with basic needs as well as for more expected returns. Moreover, growing urbanization as well as globalization is causing more rapid changes in land use patterns than the era of industrial revolution (Lambin et al., 2003). Moreover, unplanned development in developing nations have intensified the situation more (Hails, 2002) while migration in search of better livelihood have caused much unplanned global development.
5.2 Land Use Trends of Bangladesh
Bangladesh, one of the poorest states with low resource base (ADB, 2000), falls under those regions having frequent changes in land uses in last decades (FAO, 2001; Lambin et al., 2003). Moreover, national income being very low (FAO, 2001), its residents are observed to alter land uses frequently (Quasem, 2011). Estimation shows that only 10% people hold more than 40% of total lands while 60% of total population is landless (ADB 2000; Kiron, 2011), as a result, most lands are cultivated by leaseholders (Quasem, 2011; BBS, 2013). However, though initially most of the lands in Bangladesh were being used for agricultural purposes (forestry, cropping), changes have occurred in land uses as well as production techniques (Mohammad, 2009). During the last decades of 20th century, majority areas of the south-western parts of Bangladesh have been observed to cultivate traditional shrimp culture which took the first but influential changes in land use patterns (Ahmed, 2011). However, salt intrusion and tidal surges were being then observed as the main obstacles in agro farming in south-west as well as coastal areas (Mia and Islam, 2005) which in turn causes heavy losses to cultivators and changes the behaviors in making the land use changes in those areas. Moreover, crop failures due to saltwater intrusion or lack of timely flooding in most areas (Ahmed 2011; Nishat, 1988) have caused major changes in land uses after population and migration (FAO, 2001; Ahmed, 2011).
Moreover, green revolution of 1960s influenced the then land owners to have a more intensive use of land for agriculture especially rice cultivation and as a result govt. emphasized the need to protect coastal areas through construction and repairs of embankments (Ahmed, 2011). Thus beside dominance of traditional agro sector, modern varieties and technologies were introduced along with salt production, mangrove forestry and traditional shrimp farming chiefly in south-west part (Rahman and Begum, 2011). In this aspect Ahmed (2011) pointed out that during the 1970s and 80s, continued polderization of coastal areas became part of the natural coastal setting and govt. established internal water management authority to enhance further agro production. Thus, there occurred major changes in land use largely due to introducing modern varieties and conversion of agro land to non-agro uses with the project of coastal afforestation to protect the coast from cyclones and erosion (FAO, 2001). Studies also show that attempts to boost rice production through large-scale polderization in 1970s resulted in artificial embankment which in later due to poor management were observed to hamper drainage system causing the low-lying marshy land water logged with salinity intrusion (Ahmed, 2011; Rahman and Begum, 2011). The acute salinity and drainage problem caused historical tradition of shrimp farming causing a gradual transfer of crop lands and mangrove forests into shrimp farming and fallow lands (Quasem, 2011). Moreover, agro lands declined by about 0.26% yearly during 1976-2011 while increased during 2000-11 by 0.14% yearly (Rahman, 2010; Ahmed, 2011). However, following table shows the land use trends since 1977-2008.
Table 5.1 Land Use Trends in Bangladesh during 1977-2008 Area in sq km
illustration not visible in this excerpt
Source: Uddin and Gurung, 2010; Rahman, 2010
Total amount of water bodies, bare land, shrub land, open forest have increased over time while agro lands, close forest, mangroves and grass lands are decreasing in Bangladesh (Table 5.1). Moreover, Mia and Islam (2005) have pointed out that there exist seasonal variations in land uses because though water bodies during wet or rainy season are being cultivated, during dry season they remain fallow. Thus, performance of agro sectors is continuously declining (Mohammad, 2009).
This paper by this time describes the per capita lands available over time in Bangladesh through following table.
Table 5.2 Scenario of per Capita Arable and Irrigated Land
Area in ha
illustration not visible in this excerpt
Source: Islam, 2000; IRC, 1996
Per capita cultivable lands are decreasing rapidly over time while irrigated lands increased from 1961-1990 but decreased from 1990 and towards (Figure 5.2). At this stage author has depicted changing trends of lands (Table 5.3).
Table 5.3 Total Land Area of Bangladesh during 1976-2010
Area in ‘000’ ha
illustration not visible in this excerpt
Source: Hasan et al., 2013
Bangladesh has gained a total area of 905 sq km (i.e. 90,512ha) during 1976-2010 due to accretion in southern coastal zone (Table 5.3) while lands used for non-agro lands have increased with the decrease of agro lands. However, here is the presentation of total sizes of rice and shrimp farming lands during 1976 and 2010.
Table 5.4 Rice and Shrimp Farming Area during 1976-2010
Area in ha
illustration not visible in this excerpt
Source: Hasan et al., 2013
Land use data during 1976-2010 presents that agricultural lands have decreased gradually over time while shrimp lands are observed to have positive change at much higher rate.
5.3 Trends of Land Availability in Khulna Division
Khulna division, known as the industrial area as well as the Kuwait city of Bangladesh (Kiron, 2011), plays an important role in agro production especially through aquaculture along with rice, vegetables and forest commodities (Rahman and Begum, 2011). However, in this stage, this paper is now concentrating on south-west part of Bangladesh to show total land use scenario as follows.
Table 5.5 Land Use Statistics of Khulna Division in 2008
illustration not visible in this excerpt
Source: BBS, 2010
[...]
-
-
-
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X. -
-
¡Carge sus propios textos! Gane dinero y un iPhone X.