Crop choice and climate change : a case-study of increasing environmental pressure on agriculture in the Barishal Sadar Upazila, Bangladesh

(1)

Crop choice and climate change: a case-study of increasing environmental pressure on

agriculture in the Barishal Sadar Upazila, Bangladesh

Timeon Schot

11041269

Andres Verzijl (begeleider)

Josh Mayo (tweede lezer)

Sociale Geografie en Planologie

Internationale Ontwikkelingsstudies (specialisatie-richting)

timeon.schot@quicknet.nl

(2)

Bachelor thesis project

Crop choice and climate change:

a case-study of increasing environmental pressure on agriculture

in the Barishal Sadar Upazila, Bangladesh

(3)

(4)

3 Faculty of Social and Behavioural Sciences

Human Geography, Planning and International Development Studies Bachelor Thesis Project

Title: Crop choice and climate change:

a case-study of increasing environmental pressure on agriculture in the Barishal Sadar Upazila, Bangladesh

Name: Timeon Schot Email: timeon.schot@quicknet.nl

Date: June 16th, 2018 UVA ID: 11041269

Supervisor: Andres Verzijl Email: M.A.Verzijl@uva.nl

Faculty of Social and Behavioural Sciences University of Amsterdam, Netherlands

13821 words

(5)

(6)

5

1. Introduction

While doing fieldwork in a small local village near Barishal, I asked a farmer what he thought the main reason for all the changes in his environment were. He had already told me about his struggles in the cultivation of crops, as he experienced more drought, but at the same time an increase in water-related problems as well. The growth of his mango and coconut was hampered by the extended periods without sufficient precipitation and only a few years ago he lost all his crops due to water-logging. He begun by exclaiming the idea that the substantial cutting down of trees might be related to an increase of his problems. After a moment of silence he added: the new mobile-network towers in the city also cause the disasters, they especially ensure the intensification of the thunder and the storms.

Bangladesh, a small country on the Asian continent and largely surrounded by its large neighbour India, is considered to be an exceptionally vulnerable area in relation to climate change (Huq, 2001). A significant part of the country is located in the Green Delta (figure 1), a nickname given to the largest delta in the world. This Ganges-Brahmaputra delta is composed of soils made up of sediments deposited by the rivers, making the area particularly fertile and useful for agriculture. The northern part of Bangladesh is not located in the delta: this area borders the Himalaya and is due to its altitude too high to be part of the basin of the Ganges and Brahmaputra rivers.

Due to the fact that the country is located at the end of the arc-shaped delta, Bangladesh is just a few meters higher than sea-level. In other words, the area is highly susceptible to saltwater intrusion, but also to flooding due to greater wet season discharge (Collins, 2014). Secondary effects of climate change are longer periods of drought and more severe monsoons, which results in an abbreviated time span to produce and cultivate crops. This trend is a detrimental consequence to the infrastructural services and increase pressure on the supply of various ecosystem services (Lamarque et al., 2014).

Figure 1: The Ganges-Brahmaputra Delta. Source: Google Earth Pro, 2018

(9)

8 Since agricultural communities are dependent on natural resources, they are especially prone to future climate change (Kates, 2000). This means that alterations in the environment could put pressure on the food security of those agricultural communities; even more if these communities are autarkic or subsistent. However, also the global food security could possibly be endangered if supply from these regions recedes (Tai et al., 2014), as the area is an important worldwide agricultural producer (FAOSTAT, 2018). Bangladesh is a worldwide top five producer of rice, (goat)milk, (goat)meat and jute (FAOSTAT, 2018), being only the 92th country in size (UNSD, 2018).

The two threats, salt water intrusion and flooding, result in a peculiar situation for farming communities in Bangladesh. Additionally, higher temperatures, unreliable precipitation, more severe droughts and vigorous cyclones create a high risk to the food security of Bangladesh (Bates, 2008).

1.1 Problem Statement and Urgency

The area chosen to investigate the effects of climate change on the agriculture of Bangladesh is

Barishal Sadar Upazila. This district is extremely vulnerable to climate change due to geographical,

sociological and environmental factors. The Bangladesh Bureau of Statistics (BBS) provides multiple examples which illustrate the exposed position of this district:

The annual income per household of the district is 144.893 Tk (Bangladeshi taka) (BBS, 2018), which would be 1397,69 Euro each year (exchange-rate of June 1st_{, 2018). However, this income is the} average of all services provided within the district, including service-sector and manufacturing. The agricultural income; the average annual income per household which is derived from the agricultural sector; is only 35.485 Tk (342,30 Euro) (Ibid.). This illustrates the great disparities in income between households working in the agricultural sector and those working in others. Likewise, as only 0,94 Euro is made per day by the farming populace, the average income is only just above the extreme poverty line of 1 US Dollar (0,85 Euro), according to the UN standards (UNESCO, 2018). Being an average, this means that a significant portion of the households depending on the agricultural sector also lives in extreme poverty.

Nevertheless, the area of Barishal is a sizable producer of crops. Over 1,1 million hectares of land is used to produce almost 2400 billion tons of rice of the three main types (table 1): Aus, Aman and Boro (Ibid.). Aus and Aman are both sown in the pre-monsoon season and preform best under summer conditions (Alim, 1974), while the Boro-type holds a degree of cold tolerance, thus enabling its growth in the mild Bangladeshi winters (Zaman, 1980). However, the risk farmers are willing to take also plays a role: Aus grows faster, thus the chance that a disaster engulfs the harvest is lower. Yet Aman, which grows over a longer period, has higher yields and can be sold for more due to the higher quality.

(10)

9 Table 1: differences between the three main Rice varieties in Bangladesh.

Source: Author; adapted from Parsons, 1999

Furthermore, over 5,5 thousand hectares is used to produce jute, more than 9,5 thousand hectares for potato and almost 5,5 thousand hectares to produce wheat, for respectively 50 billion, 175 billion and 14 billion tons of crops (BBS, 2018). The bulk of the area is used to cultivate rice, while only a ‘small’ portion is used to grow potato and wheat. These two crops are mainly used for export, while the rice and jute are also domestically consumed, thus resulting in a higher need for cultivation (Ibid.).

However, the main problem is that the production of these crops is occasionally endangered by disasters. Especially unpredictable and devastating cyclones and storm/tidal surges are destructive to the agricultural sector (Ibid.). In a six-year period, from 2009 until 2014, a loss of 36974,39 million Tk (356,7 million Euro) in the Barishal district has been accredited to natural disasters. Nearly 6101,78 million Tk (58,86 million Euro) has been due to the loss of crops. This number only represents the value of the crops lost transferred to the selling price and not the additional damage of any related consequences, like loss of farmland, destruction of equipment or a limited food supply (Ibid.).

Also other agricultural producers have been affected: the district is responsible for almost a third of all perished Bangladeshi livestock due to natural disasters, as well as more than 40% of all poultry (Ibid.). Damage to fisheries, during the same six-year period, has also staggered to a loss of 3416 million Tk (almost 33 million Euro). Although approximately 5,3% (2012) of the total Bangladeshi population lives in the Barishal division, almost 30% of all damage to housing in Bangladesh can be ascribed to the region (Ibid.). This illustrates the severity of the natural disasters in the Barishal district in comparison to the other divisions in Bangladesh. In total, already six million people have been displaced due to disasters which can be ascribed to climate change (UNFCCC, 2012), though the time-span of this displacement is obscure.

The natural disasters do not only affect the agriculture, it affects the households as well. Throughout a six-year period between 2009 and 2014, stunning amounts of households have had to deal with natural disasters. According to the BBS (2018) 12,5 thousand households have been affected by drought, 43 thousand by floodings, 32 thousand by water logging, 640 thousand by cyclones, 7,5

Aus Aman Boro

Sown March/April March October/November

Harvested July/August August/September February/March Duration 80 – 120 days 120 – 160 days 100 – 140 days

Water source Rainfed Rainfed Irrigation

Yield and quality Low yields; poor quality normal yields; high quality High yields; normal quality

(11)

10 thousand by tornadoes, 258 thousand by storms and tidal surges, 30 thousand by thunderstorms, 36 thousand by river- and coastal erosion and 7 thousand by salinity. Although some households have been affected by multiple disasters or manifold by the same, it should be noted that the data concerns only households. The number of individuals is thus much higher.

The data shows that the Barishal district in Bangladesh is exceptionally prone to natural disasters. As estimates suggest that climate change causes these disasters to occur more often (Collins, 2014; Kates, 2000), it is crucial to investigate how locals interact with these ominous predictions. The focus of this thesis research would be on the agricultural sector of Barishal, as this one is the most important sectors for the district.

1.2 Positioning of the research

In order to become more resilient, Bangladeshi farmers have only one option: adaptation. As they do not have sufficient (financial) capital or technical knowledge to enable mitigation, adaptation is the only possibility to readjust to a changing environment. Adaptation is to “reduce the potential for harm

or negative consequences and maximize or exploit any opportunities or positive consequences that may arise under climate change” (Leal Filho 2015, pp. 72), while mitigation is the act of using various

techniques to reduce the severity of global climate change (Schot & van Zelst, 2018). However, there are multiple approaches to facilitate adaptation.

One way to adapt in relation to the agricultural sector is changing crops. However, the current scientific debate about the role of climate change on crop choice is widely scattered and contradictory. While some researchers say that crop choice is primarily influenced by financial components and has little to do with climate change (Zivanomoyo & Mukarati, 2012), others state that global climate change actually plays a significant role (Moniruzzaman, 2015). However, by far the most research does not even take the hidden link between the crops planted and the role of climate change into account; they solely focus on the technicalities of the influence of environments on several crops, like drought-resistance or the ability to withstand certain extreme temperatures (Seo & Mendelsohn, 2008; Ali et al., 2007). Unfortunately, no author has considered how changes in the current environment may occur and how this results in accompanying implications for future crop-use. The research revolving crop-choice is reacting on climate change, but not anticipating on it.

This thesis will examine the underlying relation between climate change and crop choice. It will explore to what extent local Bangladeshi farmers are able to actively select the specific crop they deem successful, as numerous crop species are able to grow in the Bengal Delta (Shahid & Abdelfattah, 2013). As research on the direct connection between climate change and crop choice is limited, the thesis will add to the already existing literature.

(12)

11 A detailed investigation of this problem could help the creation of effective policies revolving around climate change in the area: whether officials are able to continue their policies on the willingness and knowledge of the agricultural sector, or if they first need to create awareness concerning the problem of climate change.

1.3 Research aims

The aspiration the thesis tries to achieve is, firstly, to add to the understanding of the possible relationship between crop choice and climate change. The main question the thesis will try to answer is ‘how does intensifying flooding and saltwater intrusion shape the crop choice of farmers in the

Barishal District in Bangladesh’. Therefore, the research could potentially be used in the debate on the

role of climate change and could validate or oppose the relation between the two, be it within a specific context.

Secondly, the outcomes of the research could provide policymakers with additional information. This eventually will help the implementation of newly formed policies, as policymakers get a transparent view on the motives and thought processes of local farmers.

Lastly, the findings of the thesis could help construct the assumptions of research on similar investigations. Although the research takes shape of a singular case study, the conclusions may be the same for comparable cases. Thus, the thorough analysis of the data gathered in this case study may benefit future explorations as well (Eisenhardt, 1989). Additionally, if more studies would look into the relationship between crop choice and climate change, a grounded theory approach could be possible.

1.4 Outline

In order to tackle these aims, the next section will address three distinct concepts. These concepts are vital for understanding the outcomes of the data. This chapter will be concluded by a conceptual model, illustrating the relations between the various components of the concepts. The following chapter will discuss the research- and sub-questions, being derived from the various concepts stated in the previous section.

Next, the questions will be succeeded by the Data and Methods section of the thesis. This passage will explore the collection, analysis and sampling of the gathered field-data.

Chapter 5, 6 and 7 will each discuss the outcomes of the literary review and field-data, although this will be incorporated in three distinct topics: Farming in Barishal, Flood Adaptation Measures and

Crop-choice, with chapter 5 also elaborating on the specific physical and social context of the chosen

(13)

12 Finally, the last chapters will be the conclusion in which will be reflected on the stated research question, the acknowledgements of specific individuals whom have been crucial for completing the thesis, the list of references and lastly the appendix.

(14)

13

2. Conceptual Framework

The conceptual framework will consist out of three key concepts: environmental (in)security, adaptation and crop-choice. These three concepts form the central approach needed to accurately analyse the data resulting from the fieldwork. Although they will not be specifically mentioned in the sub-questions of the research, they are important in order to fully comprehend the thought-process of constructing these sub-questions.

2.1 Environmental insecurity

The first concept is environmental (in)security. Throughout the literature, the concepts are defined in various ways. White (2014) states: “the notion of environmental insecurity is usually tied to actions and

conditions that undermine the ability to exploit or use nature sufficiently to meet human needs” (pp.

836). However, he solely states the linkage between available resources and human needs. A second author, Barnett, describes it as the exposure of individuals towards shortages: “vulnerability of people

to environmental degradation” (2001, pp. 8). A third definition illustrates the broad range of different

interpretations researchers give to the environmental insecurity concept: “where rates of extraction

exceed the rate of recovery of ‘renewable’ resources (natural capital), or non-renewable resources are depleted” (Obi 2000, pp. 50). Other explanations also add the possibility of conflicts deriving from the

scarcity of resources to the definition.

The various concepts make it appear that the concept includes both a physical and a social component. The physical component, the crumbling and reduced presence of natural resources, seems to be linked to the social component of human susceptibility and demand. Additionally, social unrest and conflict may be derived from the degradation of natural resources. This becomes more evident in the two theories proven in the article Agbiboa (2013): (1) “environmental degradation is a potential

catalyst for conflict rather than a sole determinant” and (2) “the poor and marginalised communities suffer disproportionately from the effects of environmental degradation” (pp.449-450). Thus,

environmental insecurity comprises the notion of being a cause for conflict, with the addition that less resilient parts of society are hit harder.

However, although there are more than a few definitions on environmental insecurity, the thesis needs one, comprehensive description in order to demarcate the boundaries of the concepts. To do so, the concept of Hall (2013) will be used, as it provides the most complete incorporation of the various aspects, resulting in: “the concept tends to link environmental degradation and associated

scarcity of resources with human conﬂict at individual, group and state levels” (Hall 2013, pp. 36 in

(15)

14 A concluding remark needs to be made: it is possible for the environmental insecurity to be externalized towards other area’s or regions (Carolin, 2015). This provides the occurrence of the concept with far-reaching consequences, as other localities are also implicated. The case of Carolin (2015) shows, for example, how China externalises the effects of the overfishing of its own coast by redirecting its fishing fleet to other territories, thus decreasing the yields of the fishers of other countries.

2.2 Adaptation

Adaptation is used as the second concept. It is an essential concept for the entire thesis, as the thesis examines the adaptation of local Bangladeshi farmers by crop choice. Before exploring this concept, it should be noted that this thesis focusses on adaptation towards climate change, as there is a lot to adapt to. By solely focussing on climate change and its effects, the essence of adaptation is confined within certain limits, thus making it less abstract.

Throughout literature adaptation is almost always discussed together with mitigation. However, there are clear differences between the two. Where mitigation addresses measures to slow down the various processes which add to climate change, its counterpart acknowledges that climate change is already occurring and that adaptation is needed to decrease the actual impact of the event (Clark, 2012). However, Kinzig (2015) describes: “Adaptation is both easier and harder than mitigation: Easier

because it does not require collective global action and harder because it must address numerous sectors of society, not just energy and agriculture. Furthermore, most adaptation measures must be applied on the local level, where they are hampered by the lack of definitive predictions about the expected changes” (p.182). In other words, adaptation is the ability to create a symbiosis with a

potential harmful event and can be used at a local scale, although this brings forth its own complications.

In contrast to actual in-depth research of adaptation, which has only begun in the 1990’s after the formal concept had been coined by the International Panel on Climate Change, by publishing their assessment reports (Schmidt-Thomé, 2017), adaptation itself has been performed by communities and individual since the beginning of the human race. Kinzig (2015) states: “people have been adapting to

change, including climate change, for as long as humans have existed. Studies of some of those civilizations demonstrate that there may be different pathways to effective adaptation” (pp.181). This

suggest that multiple ways of adapting are possible in order to adept to a certain alteration in the climate. The adaptation occurs in a multiplex of scales, forms and at a variety of levels of authority (Looney, 2018).

(16)

15 The Bangladeshi farmers in the Barishal District are limited in their alternatives and are ought to adapt to their changing environments, as they do not have the financial-, nor cultural capital to establish mitigation techniques, nor have the ability to set up any structural protective measures. Before adaptations occurs, the local adaptive capacity needs to meet three aspects: ability, willingness and an

enabling context (Aggerwal, 2013). If these three criteria are adequately met, adaptation in a local

context in possible.

2.3 Crop-choice

The third concept discussed in the conceptual framework in crop-choice. Switching to another crop or resource to cultivate is one of many adaptation strategies. However, there are numerous factors which influence the decision of a farmer in whether or not to start cultivating another crop to ensure its own food supply. This section will state several factors which affect crop-choice and will support these by an additional review of literature.

A first aspect of crop-choice is the revenue on the local or global market. Evidence from Burkina Faso shows that many farmers switch to cash crops (in this case cotton), as these crops are worth far more than the commonly cultivated crops (Kazianga, 2016). Evidence shows that growing another crop boosts the financial situation of the households in relation to subsistent farmers (Kazianga, 2016). Klasen (2013) also reports that farmers are driven by financial incentives as he illustrates his case in Indonesia: “incomes from agriculture increased due to a shift in cropping patterns […] Higher output

volumes and more favourable commodity prices for cocoa than coffee help to explain most of the increase in agricultural incomes” (pp.362).

However, the highest revenue is not the only factor in play. Sociocultural factors also seem to have an effect on the crop choice. De Brauw (2014) has investigated the role of gender in African Mozambique in managing plots. She states that differences exist in how woman and men take care of their land: “when women manage plots, they tend to grow crops with less complicated production

techniques and are less likely to grow the main area cash crop” (pp. 435). This research makes it

apparent that not only the social aspect of gender plays a role, but also the cultural factor of what is grown in the area.

It should also be kept in mind that the cultivation of crops is limited by the ecological possibilities. The study of Wineman & Crawford (2017) shows that farmers are more eager to grow the most successful crop in terms of nutrition than other crops. The willingness to allocate land and inputs is higher for crops which are not - or to a small degree – affected by climate change, or to those crops

(17)

16 which are even able to benefit from it (Wineman & Crawford 2017). It is also said that farmers directly analyse the yields from their farmland in order to consider a possible adjustment to their growing patterns: “when farmers are faced with the estimated yields resulting from an altered climate, they

maximize calories by shifting” (Wineman & Crawford 2017, pp.28).

On March the 18th_{, 2018, The Kathmandu Post reported an interesting article. Bangladeshi} farmers in Joymoni (50 km south of Khulna) have started to switch to crab-farms, leaving their rice paddies and shrimp ponds behind (Reuters, 2018). This is a turn to make the farmers more resilient as the environment in the area is changing due to alterations in the climate. While their rice paddies are salinizing due to saltwater intrusion, crab-farms provide a viable way of living. This does not only happen in Joymoni: Seo & Mendolsohn already stated in 2008 that farmers will select the crops most appropriate to a new, changing climate. This means that the last factor determining if a farmer is willing to switch crops is its awareness of the environment. Evidence from peatlands, provided by Silvestri (2017), shows that farmers may knowingly choose to cultivate a certain species of crops if this crop proves to allow for sustainable agriculture. However, it appears that the basic needs of the farmer for this factor should first be met. In the Barishal Division various farmers are struggling with poverty, making it unlikely that they will think of the long-term consequences of their actions.

To conclude, multiple factors influence the choice for a certain crop. The revenue on the (local or global) market, sociocultural aspects, the most successful crop in the current environment, most appropriate to a new climate and awareness of the environment all seem to play a role. It is expected that multiple aspects at once are decisive to make a farmer switch crops. It should therefore be kept in mind that not a single, but a plural of factors in effect at the same time determine the alteration to a different cultivation.

2.4 Conceptual Model

As described in the previous paragraphs, the three concepts are all subject to multiple factors. Adaption is dependent on ability, an enabling context and willingness; environmental (in)security is related to social conflict and resource scarcity; and crop-choice is determined by socio-cultural factors, revenue, ecological possibilities, sustainability and the future environment. However, all these aspects of the three concepts also voice an influence on each other. Figure 2 shows the conceptual model in which the relationship between the various factors is depicted.

(18)

17 Figure 2: Conceptual Model

Source: Author; Mindmup

First of all, social conflict affects both the ability and willingness of adaptation. If social conflict endangers the lives of the farmers, they might be less eager to cultivate new, unknown crops, but rather stick to certainty in uncertain times. Furthermore, there might be less seeds available on the market, as the conflict could potentially harm the supply.

Secondly, resource scarcity influences the willingness of adaptation as well, but also the enabling

context. If resources are limited, a farmer might have a higher incentive to grow a new crop which

provides a high yield, instead of his traditional vegetables or fruits. On the other hand, the scarcity might also imply a constraint on the abilities of a farmer to grow alternative crops. If the previous harvests have been unsatisfactory, it could be possible that the farmer is not able to provide for new, more expensive goods.

Another relationship is the linkage between sustainability and resource scarcity. This might be implied to ensure that the - what is considered to be a problematic - event of resource scarcity will not occur. As sustainability entails the practice of growing crops most feasible for the upcoming years, this will also allow for the avoidance of resource scarcity.

Lastly, both ecological possibilities and socio-cultural affect the enabling context. The ecological possibilities, thus what is able to grow and what can not, is a mayor part of the enabling context. It decides to which extent a farmer can imply a certain adaptation method. Further, the socio-cultural context of a farmer might play a significant role in determining whether or not to adapt. If a certain practice is not allowed or if a farmer has to consult with the entire village before commencing implementation, he might choose not to implement at all.

(19)

18 Although these relationships are the ones that are visible within the model of figure 2, there is an abundance of imperceptible linkages. Revenue might for example also influences the willingness, as financial incentives are generally the main reason to initiate adaptation. However, the relations in the model are the most significant and are supplementary also used to illustrate the myriad of connections possible. A broad variety of relationships between the numerous factors exists.

(20)

19

3. Research questions

In this chapter the main thesis question and the three supporting sub-questions will be epitomized. The previous chapter has demarcated the various concepts and has elaborated on the degree to use them. The one before addressed the necessity of the thesis and explained for whom and why the research would be valuable. This section will extend on these chapters and shall construct the resulting research questions.

The previous chapter, conceptual framework, results in the following main thesis question:

How does the intensifying amount of flooding and saltwater intrusion shape the crop choice of farmers in Barishal Sadar Upazila, Bangladesh?

Three sub-questions which support the answering of the main thesis question from above are:

1) Which crops are currently cultivated and why are specifically these crops grown?

2) What are the biggest environmental insecurities concerning flooding and saltwater intrusion and how have farmers adapted over time?

3) How are the current crops affected and how does this shape the choice for alternate crops?

To create a clear starting point, it is necessary to know which crops are currently grown. Furthermore, it is required to gather an insight in the motives for harvesting these crops, as these could provide valuable information about the motivation and eagerness to eventually switch to another crop.

Secondly, to narrow the scope of the research, the data must be focused on flooding and saltwater intrusion. It is not possible to investigate all possible variables, thus the insecurities of the local Bangladeshi farmers concerning these two issues have to be examined. If it becomes apparent how their uncertainties have been shaped, it is possible to explore how far these farmers are willing to go before they feel another crop succeeds the one they are harvesting.

The two previous questions combined provide the data wanted as soon as it is known to what extend the crops are currently affected. If the scale and intensity is recognized and the motives for choosing a certain crop have been stated, it becomes clear how flooding and saltwater intrusion shape the crop choice of the farmers in the Barishal Division.

(21)

20

4. Data and Methods

4.1 Data collection

The method used to investigate the research question is that of a case study. As this is a qualitative approach of conducting a research, there are also several demands to guarantee the correct execution of this research. Bryman (2014) and Guba & Lincoln (1994) distinguish four criteria which such a qualitative research should meet: credibility, transferability, dependability and confirmability. When these four are executed into perfection, a completely untainted thesis research has been constructed. However, as there will always be minor flaws in achieving the four criteria, an actual unsullied study will be impossible.

To safeguard the reliability of the thesis, triangulation will be used. Triangulation is the usage of different sources to verify the same outcomes in data. If several sources state the same result, it is more like to be true than a single source presenting it (Bryman, 2014). The various sources will be interviews with the local farmers, observations and a literary review, on which shall be elaborated throughout this section.

Additional results of using a case study is that the outcome of the analysis holds a low external

validity, yet a high internal validity (Yin, 2014). This means that the conclusion of the analysis cannot

be applied to other cases, as these results are context-dependent (Ibid.). However, concerning the

internal validity, due to the in-depth investigation of the research, it is highly expectable that the

outcomes are also applying to other units within the studied case (Bryman, 2014). The external validity does nevertheless fulfil a more extensive purpose, as the whole of all the separate case studies may result in an inductive process of knowledge creation, thus possibly allowing a grounded theory

approach (Martina Perez et al., 2015): creating new generalities from a broad scale of similar outcomes

in different contexts.

4.1.1 Literary review

The literary review has been done before going into the field. A part of the review has been to gather information on the context of the area, in order to prepare and create a more accurate set of expectations. The other section of the literary review had the goal to acquire data on the research topic. First literature was inquired by cross-referencing the area of interest, Barishal Sadar Upazila, with the concepts used in the conceptual framework. This created the starting point for the literary review. For both parts the information had further been amassed using a snowball-like method: while obtaining data from a certain set of documents, relevant information and notes were made bold. After finishing the set, these bold comments were looked up, thus creating a circle of knowledge-gathering.

(22)

21 4.1.2 Interviews

During the fieldwork, data has mostly been gathered from semi-structured interviews. The preference for a semi-structured interview had to do with the fact that it provides enough possibilities to go in-depth at certain interesting comments, while it also presents a standardised guideline to lead the interview in the right direction (Bryman, 2014). This allows for more information to be gathered, as the interviewer can address specific, interesting points without having to be obliged to follow a pre-made list. The list is available in the appendix (Appendix I)

The interviews have been conducted on four separate days. Accompanied by three field-assistants (Appendix II) we walked along the farming fields. On several occasions a farmer working in the field was approached and asked to participate in the research. There had been no rejections and all addressed farmer eagerly took part in the interview. Other farmers were found by strolling through small local markets. Shopkeepers were asked if they knew if any farmers were near the market, thus enabling us to specifically address these individuals pointed out by the owners of the local stores. A third part of the interviews have been conducted by farmers which were found along local roads. We positioned along a small road, one known for its frequent transport of locally produced goods, and approached farmers on their way to the local market, their farm or their homes. Again, there were no rejections. Ultimately, a total of 17 interviews (Appendix III) with the average length of 15 minutes were gathered.

Before starting the interview, the respondents were offered a seat in the shadow, as the outside temperature went well above the 30 degrees Celsius. During the interview, one of the field-assistants would ask the farmer the questions from the pre-made list, while another one would translate the local, Bengali language into English. I would then note the interesting comments and ask further if there were any interesting statements. At the end the interviewees were thanked for their participation: I exchanged a “dhonnobad” and “abar dekha hôbe” with the respondents, meaning “thank you” and “until we see each other again”.

Also, all the interviews have been recorded. The assumption had been that this would benefit the details of the transcripts produced later, and simultaneously allowed the interviewer only to make certain notes on specific remarks or quotes. Unfortunately, due to the poor quality of the recording device and the surrounding noise resulting from the roads, market and commotion produced by locals observing the interview, the actual audio recorded is of such a poor condition that it has not been able to translate the Bengali part of the conversation. The consequence has been that the recordings could only partly contribute to the transcript. Since the recorder had been positioned next to the author, instead of next to the interviewee, it is extremely challenging to listen to the Bengali. However, the English translation of the Bengali given by the field-assistants during the interview is mostly understandable. Thus, the English translation has been typed out in the transcript, instead of the initial

(23)

22 plan to ask the field-assistants to construct a transcript of the Bengali of the interviewee. Nevertheless, the resulting documents, containing a lot of irrelevant information, is considered to be used “as a set

of materials to be mined, accepting that a good deal will be dross.” (Weiss, 1994). Two transcripts are

shown in the appendix, relatively Appendix IV and Appendix V.

4.1.3 Observations

The observations are used to examine whether the outcomes of the data are apparent in the real world. If the observations show that the outcomes of the interviews are not perceptible in Barishal Sadar Upazila, the two different methods of gathering data are contradicting each other, thus impairing the reliability of this data (Flyvbjerg, 2006).

The observations made have been overt, meaning the individuals present in the area could know about the research. However, it is not necessary to conceal the observations, for instance by using

headnotes, as the observations are not of the social but of the physical world (Emerson, 1995). The

observations were carried out in the same district and the researcher made notes concerning these observations.

The data of the observations has been gathered alongside the conduction of interviews. While walking along the farms, notes have been made on the crops grown and the techniques used to let these crops grow. These notes were written down, however there had been no construction of a specific scheme to document them in a certain structure, as such a scheme could limit the observations by directing the researcher to specific aspects or components of the reality. No structure meant that the researcher was free to write every visible feature of importance down.

4.2 Data analysis

After the data had been collected, it was analysed. However, while doing so, several components of the data must be kept in mind: the interpretation of the researcher, attitudinal fallacy, interviewer

variability and reactive effects. Those four processes affect the data, be it only slightly or more

considerably, and should therefore be included while analysing the outcomes.

As Atkinson (1994) suggests: all social phenomena are subject to the interpretation of the researchers. This interpretation, based on the own socially constructed reality, results in a subjective value which could influence the data (Atkinson, 1994). Translated in other words: “it is not the case

that there is just “one truth” that the observer or interviewer either does or does not “see” or “hear.” Rather, each researcher implicitly draws on his or her common-sense cultural knowledge [...] and creates or constructs a truth or interpretation […] that will work for all practical (intellectual) purposes”

(24)

23 Another issue is attitudinal fallacy. Broadly defined, this concept implies that an interviewee declaring a certain statement does not specifically mean that the interviewee literally behaves in the way the interviewee says he or she does; behaviour is primarily dependent on the context of a situation, which is difficult to delineate during an interview (Jerolmack & Khan, 2014).

Interviewer variability is a third process which could influence the data. This obstacle implies that

the data gathered in the research is subconsciously affected by the ways in which a researcher works (Bryman, 2014). Both inter-interviewer variability as intra-interviewer variability may occur, respectively meaning a variance in data due to multiple interviewers, all with their own social traits which may influence the interview; or a single interviewer not being consistent throughout multiple interviews (Bryman, 2014). Gubrium (2012) adds a useful comment to the interviewer variability: “they

don't necessarily “hear” what their informants tell them but only what their own intellectual and ethical development has prepared them to hear” (pp.81). As this thesis only has one researcher and a limited

amount of only three field assistants, specific attention should be addressed to intra-interviewer

variability and the aim to be as constant as possible throughout multiple interviews. Standardization

of the interviews could positively stimulate the consistency between the interviews (Bryman, 2014). Subsequently, the last component which could have had an influence on the data are the reactive

effects. Bryman (2012) states that this concept indicates that respondents are aware that they are

being observed. As a result, they will alter their behaviour according to social standards, thus eroding the value of the observations of their manners of conducting themselves.

Analysing the transcripts and observations has been done in generally the same way. To examine and determine the transcripts, a first set of three groups had been made. Each group represented one of the sub-questions of the thesis research. Thus, the first took care of the crops cultivated, the second related to environmental insecurities concerning flooding and the third investigated the effects on crops and future crop-choice. The meaningful data from the transcripts had been divided over these three classes, while the useless components of the transcripts were left out. As a result, each group had an array of valuable data with a link to one of the sub-questions.

The observations have been analysed as a control set in order to safeguard triangulation. They have first been sorted on the sort of observation: some were of the crops growing in the farm fields, whilst others revolved around the agricultural techniques used. A third category was one which showed contextual information of the area around the fields. After this reclassification, the observation of each category has been read to review if the observations were in line with the data from the transcripts and the literary review. But also to inquire if any observations stand out or were unexpected.

(25)

24

4.3 Sampling

Before commencing the collection of information, it is necessary to know how much data should be gathered. The requirements for the amount of data needed to justify a qualitative research vary from those of a quantitative investigation. Although a quantitative analysis needs to have a sizeable sample of the population (Bryman, 2014), qualitative research uses a different approach. “Qualitative

researchers justify their sample size on the grounds of quality data – something that should be clearly reflected in the presentation of the study’s findings” (Cleary et al., 2014, pp. 473). Cleary (2014)

furthermore adds: “stopping information gathering is dependent on ‘redundancy’ of information or

‘saturation’” (pp. 473). With ‘redundancy’ meaning the point in conducting interviews where no new

concepts are emerging and the ones already provided are numerous times repeated (Trotter, 2012), and ‘saturation’ the edge at which all questions have been thoroughly explored in detail and no new concepts or themes emerge in subsequent interviews (Trotter, 2012). It is thus clear that qualitative research is reliable when all information concerning the issue investigated is gathered.

The unit of analysis, the local farmers in Barishal Sadar Upazila, have been selected by either a “convenience sample” or a by “snowball-sampling”. A convenience sampling is a group available to the conductor of the research by the virtue of it accessibility (Bryman, 2014). “Snowball-sampling”, on the other hand, is when an interviewee points out or recommends the next respondent, based on the sample of the interview. My expectation was that not all farmers were willing to be a part of the analysis on the effects of climate change on crop-choice, thus making it more likely that those willing would be interviewed. However, while conducting the interviews, none of the approached farmers turned down the participation to the research. The sample has therefore been based on the farmers available in the area and able to cooperate in the study.

To conclude the sampling, beforehand it was expected that a total of 15 interviews was needed to provide substantial information about the chosen research question. Still, 17 interviews have eventually been conducted, in addition to three separate focus-groups between the three field-assistants and the researcher. To enhance the quality of the data gathered, the intention was to add three additional interviews experts or organizations, in order to be able to collect supplementary information about the context of the problem and the area. The understanding was that when those were done before the interviews with the local farmers, this extra contextual knowledge could have been constructive, as it would have created an opportunity to initiate more in-depth interviews with the unit of analysis. However, eventually only one of these interviews could have been checked-off. Due to the tight time-schedule it was not possible to make further appointments with other experts or organizations.

(26)

25

4.4 Practical and ethical challenges

The anonymity of interviewees is a first ethical challenge. Although the thesis does not revolve around a controversial topic, it would have been possible that some of the respondents made sensitive statements during the interview. These proclamations could be potentially harmful if they, for example, criticize the government when that would not be allowed. On the other hand, Yin (2014) states that choosing to disclose makes it easier to review the case, as “footnotes and citations can be

checked […] and appropriate external comments can be solicited about the published case” (p.197).

Nevertheless, if an interviewee states that it is appreciated that the identity will not be published, that request should be respected. In order to accomplish this, before commencing the interview, there must be informed consent: the respondent has to be informed as good as possible about the goals of the analysis in order to determine their participation (Bryman, 2014). Further, anonymizing respondents should also be done if potentially detrimental issues are discussed during the interview. Thus, the interviewees will only be anonymous if they request it or if the researcher deems the responses as potentially harmful. To make sure that none of the interviewees were mentioned against their will, all have been asked to grant permission to use their name. However, none rejected that their name would be used.

The second and perhaps most important challenge is the language barrier. Due to the fact that the research has been done in Bangladesh there is gap between the spoken and understood languages, as the interviewer does not speak Bengali, nor does the local population speak English. The solution has been a translator; however, using a translator also brings forth several disadvantages. While a translator has to interpret the interview with the local, he then has to transform this interview in another language. This means that a part of the data will be lost due to language differences (Santos. Jr. et al, 2015). The data will moreover be manipulated by the translator as well, since he is limited by his own perception and gives his own interpretation on the data (Santos. Jr. et al, 2015).

During the fieldwork this became apparent on several occasions. An example was when respondents were asked if they knew why changes in their environment or alterations in weather patterns had occurred. Whenever the interviewee would admit that they did not know, the field assistants would stay silent. Yet, when a respondent would make a suggestion and state that cutting down the trees might have some influence, the assistants would exclaim their satisfaction by heavy nodding, clapping or by saying: “Yes! This is the point!”. Several times the respondent would then start guessing for more processes which affect climate change, as they had tasted the gratification of the field assistants.

(27)

26 However, the analysis should also consider difficulties arising from both interpreter and interviewee. McCarthy (2016) states numerous of variables which could cause differences in responses. Although most originate at the respondents, the social-economic position of the translator should be considered as well: “during data collection, the researcher needs to be mindful of […] ethical and cultural issues,

as well as differences in age and language proficiency, peer and gender roles, emotional issues and levels of personal confidence” (pp. 37).

Other risks are that the translator only gives short translations, while the interviewee provides a whole lot more information. This reduces the richness of the final transcript and diminishes the amount of data each interviewee provides. To prevent this, there must be a good interpreter-interviewee match in terms of characteristics based on social traits (Temple, 2002). Temple (2002) states that “particular stress is laid on interpreter and interviewee being of the same sex, but culture,

religion and age are also seen as important within the hierarchy of suitability” (pp. 6). Beforehand it

was already known that it would be difficult to find an interpreter who would meet the requirements of Temple, however it had been kept in mind in order to minimise the chance of discussions or arguments between the translator and the researcher. The definition of the role of the translator would be in a nutshell: “the interpreter is a conduit linking the interviewer with the interviewee and

ideally is a neutral party who should not add or subtract from what the primary parties communicate to each other" (Freed 1988, pp. 316 in Temple et al., 2002)

While doing the interviews, the field assistants provided only short translations of what the interviewee told. However, this was mostly because the interview was already continuing: one would ask questions and another would translate. If the translations were to extensive, the next question would not have been able to be translated as well. This was countered by three separate focus groups, where the results of the interviews were discussed again and the assistants could add data that had not been translated earlier.

Nevertheless, there have also been a couple of exasperation. One would be phone-calls during the interview. If an assistant was called during the interview, he or she would pick up the phone immediately and start a conversation. They would not hang up during the interview, as that would be rude from their cultural perspective. Furthermore, as all the assistants were significantly older and were used to a different habits and customs concerning conducting interviews, more than once there have been discussions on what would be the right way to collect the data. Instead of accepting my perspective for being the lead researcher, they would tend to argue. Nevertheless, this does not take away the immense help they have given me during the stay and data collection in Barishal.

(28)

27

5. Farming in Barishal

The research is conducted within a specific area. This chapter will provide contextual information on the study area and will furnish an insight in local farming in the chosen field.

The administrative structure in Bangladesh is, according to the BBS (2018), divided in five levels of legal power. The national government is the main body of power within Bangladesh. Below this level of governmental administrative power are several divisions. Bangladesh is divided in eight different divisions: Barishal, Chattogram, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur and Sylhet. The names of these divisions are derived from the city where the administrative power is housed, which is generally the largest city within the area. The eight divisions are split in a total of 492 upazilas. These are than again divided in 4554 Union Parishads, which are the smallest administrative units of local government and usually contain several villages or multiple neighbourhoods.

The study area of this thesis research will be in the Barishal division, in the Barishal Sadar Upazila. The upazila holds 128 villages, including the city of Barishal. Barishal Sadar Upazila is located in the centre of south-Bangladesh, only 80 kilometres to the Bay of Bengal and 110 kilometres to the capital of Bangladesh: Dhaka. The Kirtankhola river runs through the district and the city is located on the banks of this river as well. Because of the Kirtankhola, fishery is plenty and boats are one of the most important means of transport.

The interviews have not been gathered within the city of Barishal, but on the more remote fields laying in the east (figure 3). To get there, various kinds of transports have been used: boat, bus, normal- and motorised rikshaw and bike. Depending on the location, multiple of these were needed to translocate.

Figure 3: Study area; every yellow dot represents one interview. Source: Author; Google Earth Pro

(29)

28

5.1 Crops

In the fields of Barishal a multiplex of crops are grown (figure 4; 5; 6; 7; 8; 9; 10). During the research a broad variety had been encountered. Most farmers grew rice and pulse, however, local vegetables were popular as well. We came across Ladies Finger, Ceylon Spinach, Lotus Leaf, Betel leaf, Red Lentil, Green Chili, Sesame, Guava, Mango, Goya, Ridged Courds and Mumbai Chili. Some foreign crops were also cultivated. These include Corn, Wheat, Tomato, Mustard, Sunflower, Eggplant and Pumpkin.

Figure 4 – 7 (left row, from top to bottom): Corn, Betel leaf, Pumpkin and Goya. Figure 8 – 10 (right row, top to bottom): Pulse, Eggplant, Ridged Courds.

(30)

29 These crops were grown on farms varying in size and purpose. Some of the farmers had fields of only half an acre, while the largest had a total of eight acres of land. However, the average of land cultivated by the interviewees was around three acres. These acres were, depending on the position of the farmer, privately owned or a rent from another land-owner. Some also rent land in addition to the land they already owned, as it would be more productive to cultivate a larger field. On these fields the crops are cultivated, often using the same crops as long as the farmer has been practising agriculture.

In the Barishal division, agriculture represents with 29,01% a major portion of the main activity of household’s income, only second to day labour (32,18%) (BBS, 2018). However, as interviews with local farmers in the Barishal Sadar Upazila showed (Sha Alom Khan, 30; Motaleb Sikhder, 50), a significant amount of these day labourers are dependent on the agricultural sector and work as a ‘helping hand’ for the farmer. Especially during the sowing- and harvesting season these men exert themselves and support the farmer in his endeavour to ensure a proper yield of his crops, with a compensation of 500 taka per day (Sha Alom Khan, 30; Motaleb Sikhder, 50). Table 2 shows a list of current crops grown by the seventeen respondents. A significant part of the of the farmers grow rice and pulse. These two crops are also most apparent in the Barishal Sadar Upazila. However, it is more interesting to see that multiple species of rice are grown: Iri-rice and Aman-rice, but the new hybrid specie Dholot-rice as well. It is nevertheless uncertain to what extent these species are used, as not all farmers have specified the kind of rice they use.

Furthermore, cultivation in the study area also includes a broad variety of local crops. These have been placed between brackets in table 2, as their quantities on the lands are of considerably smaller proportions. The local crops are mainly used for own consumption, as the farmers are primarily subsistent. Additionally, these crops are worth less on the market, as they are widely produced. Farmers therefore assign just enough land to these crops to ensure that the production results in sufficient vegetables for their own families. The plot of land Interview

Number

Current crops

001 Pulse

002 Rice, pulse, corn (tomato, ladies

finger, Ceylon spinach)

003 Rice, pulse, jute (Betel leaf,

ladies finger)

004 Iri-rice, Aman-rice, mustard

005 Rice, pulse, Mumbai chili (Betel

leaf, tomato)

006 Rice, jute, pulse, mustard

(beans)

007 Rice, pulse, jute, mustard,

(Ladies finger, Betel leaf)

008 Pulse, green chili (Ceylon

spinach)

009 Pulse, green chili

010 Aman-Rice, pulse (ladies finger)

011 Rice (Betel leaf, Ceylon Spinach)

012 Rice

013 Rice, pulse

014 Rice, pulse

015 Rice, pulse, guava (Ladies

fingers, Ceylon spinach)

016 Iri-Rice, pulse (sesame)

017 Rice, pulse (Ladies finger, Betel

leaf)

Table 2: Crops grown per farmer. Source: Author

(31)

30 that remains is used to plant products which are worth more: rice, jute, green chili or pulse. Nevertheless, it may occur that a farmer grows additional local vegetables to sell these on the local market, however, it is more common to do not.

5.2 Environmental changes

Of the seventeen respondents, fifteen claim that the natural disaster rate has increased. Heavy rainfall increased in relation to the previous years, resulting in more water-logging and flooding of the land (Iddris Kholifa, 70; Huran Hawlader, 50; Jafar Ali Howleder, 45). Furthermore, there has also been significant more thunder and lightning (Motaleb Sikhder, 50). Although this does not immediately affect the yields of the harvest, it does pose a serious threat to the exposed cattle of the farmers. Besides, this thunder and lightning is often accompanied by storms or heavy rainfall, which could affect the crops.

Additionally, there are longer periods of droughts as well (Abul Kasem Akondo, 60). The rain does not fall in the proper time anymore (Monirul Islam, 38; Selim Howleder, 42), with the ‘proper time’ being the period the farmers would expect precipitation, a calculated guess based on personal experience derived from the time where the rain used to fall. Consequently, this results in the fact that less water is available (Abul Kasem Akondo, 60; Altaf Hossain, 48), thus threatening the crops by water shortages. Another environmental change, directly linked to drought, is the fact that the temperature surpasses the upper-limit in which trees can survive. One respondent claims that “the heat of the air

is too heavy for the trees” (Muslim Howleder, 55). In other words, the temperature is too much for the

trees to bare, resulting in them dying or growing slower.

A lot of farmers also complain about the transitions between the seasons. There is more rain in the summer and there is no clear difference with the other seasons (Nurelem Badsha, 32). Only the ‘sunny season’ and the ‘rainy season’ remain; the others don’t exist anymore (Beleyat Hossen Gazi, 53). The results can be disastrous, as this development leads to an intensification of weather extremes: “during rainy season the crops were destroyed due to a flood; after the rainy season the crops died

because of a lack of rain" (Beleyat Hossen Gazi, 53). As a result of a lack of rain in the normal period, it

is not possible to grow as many crops as before (Sha Alom Khan, 30). This unpredictable redistribution of precipitation results in many farmers planting their crops at the wrong period, ultimately leaving them standing in a by water oversaturated land or watching their crops suffer and die because of the extended periods of droughts.

A consequence of the changing weather pattern is a less fertile soil. This result in a higher dependency on the usage of fertilizer. All the respondent have stated that their use of fertilizer has become more expensive. However, one also claimed that due to the over-use of fertilizer, the land has

(32)

31 lost its fertility. Production is not sufficient anymore and lower than the previous years, thus farmers use more and more to receive profitable yields (Muslim Howleder, 55). Nevertheless, a significant portion of the interviewees also has to use more fertilizer than before to get the same result (Selim Howleder, 42). The consequence is that farmer might end up in a downward spiral, where the over-use of fertilizer reduces the fertility, thus increasing the amount of fertilizer needed, which again reduces the fertility.

Another consequence of the different weather-pattern is that insects have been able to thrive over the past three years (Huran Hawlader, 50; Emayet Gati, 55). As the seasonal rain did not occur in the proper time, but was delayed, the amount of insects has dramatically increased (Berik Howleder, 46): “the insects live near the water, but mostly the seasonal rain floods these places and washes the insects

away. However, now that did not happen” (Altaf Hossain, 48). Due to the fact that the rains are more

severe and have a higher chance to create a flooding, the amount of insects may increase even more. This is because during the post-flood, when the flood slowly withdraws, there is a huge increase in breeding spots for insects in the newly created ponds and still pools of water (Beleyat Hossen Gazi, 53). However, most dangerous in relation to the insects might be the reports of the findings of several new species, which are resistant to the pesticides (Nurelem Badsha, 32).

The insects are regarded as one of the most harmful environmental influences on the crops. Some farmers note that they are potentially more disastrous than drought, flooding or any other natural disaster. Besides, they affect vegetables more than rice (Altaf Hossain, 48), and as the vegetables are mainly for private consumption, the farming families are immediately affected at their food supply. However, as Salam Talukder (44) states: “to prevent crops from getting eaten, he buys a lot of

pesticides. But this also increases his production cost and is not safe for the human body, so sometimes people get sick. But there is no other way to protect against the insects”. As most farmers are

subsistent, their families are the first to be harmed by these insecticides.

Within scientific literature, many agree on the increasing threats for the disaster-prone country of Bangladesh. Many researches use a guideline of 1,5ºC to 2,0ºC increase in temperature (Collins, 2014; Ali, 1999; Karim, 2008), stating that every degree will worsen the disaster-rate. However, most are more concerned about the future, as the whole of Bangladesh is expected to be affected by climate change: “considering the avenues for significant environmental change, and various scenarios, it is unlikely that any part of the country will remain untouched by climate change in this century” (Collins, 2014, pp. 290). Especially since predictions state that most threats will come from the south of Bangladesh, from the Bay of Bengal (Ali, 1999), as this area is the lowest and therefore most prone to sea-level rise; but is also situated at the mouth of the rivers; the area of Barishal Sadar Upazila faces a precarious future.

(33)

32 There are however two respondents who hold a different perspective on the alterations in the environment. One of them, Salam Talukder (44), states that the weather remained more or less the same. The other one, Dillow Hossain (38), even argues that there are less floods and cyclones than before. However, out of the seventeen interviewees, fifteen claim that an actual change has actually occurred and that they are victim to an increasingly degenerating environment.

(34)

33 Figure 11: Slums in Barishal

Source: Author

6. Flood Adaption Measures

The seventeen interviews have provided a lot of information concerning crop-choice and changing environments. However, some of the outcomes are contradictory to what other farmers claim, even though they all live within the same area. This chapter will elaborate on the perspectives of the interviewees on flooding and saltwater-intrusion. The data will be projected in relation to the conceptual framework, especially regarding the concept of adaptation and environmental insecurity. Within the city of Barishal the problems arising from environmental

changes are visibly apparent. Throughout the city, numerous slums (figure 11) have been build. These impoverished neighbourhoods predominantly house migrants from areas south of the city of Barishal. While having a dialogue with one of the field-assistants, we found that a lot of families have moved here as it is believed that a large city gives a better chance at finding a job. As Barishal is the largest city in the district it is the first place families will move to. However, when continuing the conversation and asking why these people have not been able to stay in their villages, it was said that they were mainly farmers who have lost their land to erosion of the rivers, floodings or salinization. As they were unable to maintain the cultivation of their land, they departed to the slums of Barishal. Although the slums might have existed without the loss of farmland as well, the vastness and extent of these ‘bidonvilles’ are tangible images of the severity of climate displacement within the division of Barishal.

6.1 Salt-water intrusion

An unexpected outcome of the interviews was the absence of any problems concerning salinization, resulting from salt-water intrusion. All the respondents exclaimed that they had never experienced any issue with decreasing fertility due to salt-water intrusion or sea-floodings. Other processes of salinization have not been encountered either. There are various reasons why these are not apparent in the area. A first is the distance to the sea. Although the land is only a few meters above the sea-level, the area is too far away for the salt-water to intrude. Especially as no groundwater is drained to provide in the needs of fresh water, the reserves are conserved, thus blocking the infiltration of salt-water in these underground basins. The lack of drainage can be accounted to the availability of the many rivers, but even more to the respondents practising rainfed farming. This means that the weather