Overcoming barriers to adapt rice farming to recurring flash floods in haor wetlands of Bangladesh

Abstract

Climate change resultant hazards have become a major threat to farming, food production systems and agricultural sustainability globally. Like many other countries, Bangladesh is also the prey of climate change extremities. Haor wetlands of this country, a major rice growing area, are subjected to extreme climate tremors where millions of inhabitants lose their boro rice production due to recurring flash flood events. This study examined the barriers to adapt rice farming to recurring flash floods in the haor wetlands of Bangladesh. The ways of overcoming barriers to adapt rice farming to recurring flash floods in the haor wetlands of Bangladesh were also explored during the research work. The research was conducted in the Sunamganj district of Bangladesh and data was collected through a mixed-method approach. A survey was conducted with 115 haor farmers and FGD and key informant interviews were conducted with 32 and 4 respondents respectively. The results showed that the lack of availability of submergence tolerant variety (a rice variety that can survive and continue growing after being completely submerged in water for several days) is the major barrier to farmers' adaptation to flooding events followed by limited market access and lack of access to inputs. A total of 85% of respondents reported encountering moderate to severe barriers to adapt to flash flooding. Besides, some socio-economic traits, including annual family income, extension media exposure, and perception on climate change have been identified to be influencing farmers' adaptation behaviour to adapt their rice farming system to recurring flash flood events. This study elaborated pathways and suggested policy recommendations to adapt to flash flooding and to ensure sustainability in the agricultural system in the haor wetlands of Bangladesh.

Highlights

• •Climate change-driven sustainable production system is the prime concern nowadays.
• •Haor farmers of Bangladesh confronted several barriers due to climatic extremities.
• •Socio-economic issues significantly influence farmers' problem confrontation behaviour.
• •A distinctive, need-based and context-specific adaptation policy is an emergency today.

1. Introduction

Bangladesh, a member of the global south, is the earth's greatest deltaic plain area formed at the confluence of the Ganges, Brahmaputra, and Meghna rivers [1,2]. It is a predominantly agriculture-based country that contributes about 13.02% of the national GDP [3]. It is one of the dominant rice-growing countries producing about 34.7 million metric tons of rice annually [4]. It is the 4th largest rice-growing country in the world with almost 80% of the total cropped area covering 11.75 million ha used to grow rice [5,6]. So, rice production is of great importance to Bangladesh's national income, food and nutritional security, and agricultural economy because is the staple food of about 165 million people, contributing one-half of the agricultural GDP and accounting for about one-half of the country's rural employment [7,8]. Therefore, from Bangladesh's perspective, a continued and secure supply of rice is equivalent to attaining sustainable agriculture and a secure food system.

For Bangladesh, monsoon or seasonal floods and flash floods have turned into the most common climatic events [9]. On average, one-fifth of the total 8.0 million hectares (ha) of cultivable agricultural lands are annually damaged by flooding events [3]. Specifically, for rice growing areas, more than 2.5 million ha of land are flood prone of which about 1.0 million ha are highly prone to regular flooding events [7]. This changing climatic situation and uncertainty arise questions surrounding the future of the agricultural system and livelihood sources in the delta floodplain areas like Bangladesh [10,11]. Because, on average 4% of the total rice production is lost every year because of the flood events [7]. But flash flooding events are the most devastating causing almost 80% or even total rice production loss specifically in the haor wetlands of the country [12,13]. For instance, the flash flood in 2017 caused a loss of 0.88 million metric tons of boro rice in haor wetlands [14]. Hence, ensuring a safe production system under the crop loss threats due to flood events in rice-based agricultural regions is making attaining sustainability in agriculture a challenge [10]. Under this crisis adopting suitable climate change adaptation strategies is crucial to secure our rice production and the stability of our agricultural system, because according to IPCC ‘adaptation plays a key role in reducing exposure and vulnerability to climate change’ to meet sustainability in the vulnerable areas [15]. In addition, developing countries like Bangladesh are commonly more vulnerable than developed countries to climate change risks and its devastating impact on the crop production system will be felt more by smallholder farmers [16,17]. So, adopting adequate adaptive strategies by the farmers is the prerequisite to securing food availability for future generations.

The haor wetlands situated in the northeastern part of Bangladesh are the major rice growing area contributing about 18% of the total national rice production [3,6]. These haor wetlands are physically a bowl or saucer-shaped large tectonic depression prevailed by a subtropical monsoon climate. This unique physical setting and hydrology of haor areas had created a myriad of livelihood opportunities and abundant crop production [18,19]. However, they mostly dry up during the post-monsoon months. So, there is hardly any scope to grow crops other than boro rice in the dry season which covers about 85% of haor areas where only 15% area was allocated for aman rice and rabi crops [12]. Unfortunately, this area is also the victim of severe climatic threats, prominently early flash flood events [20,21]. In addition, the prime concern nowadays is, flood events in these north-eastern haor wetlands are expected to upsurge more in near future due to climate change, thereby making the sustainability of production and farming system even more uncertain in the future [20,22,23]. Since boro rice production greatly depends on nature, this crop remains under the constant risk of partial to complete damage from pre-monsoon flash floods just before harvesting [16,24]. Therefore, a safe harvest of boro rice would be the key to upkeep safe and secured rice production for the inhabitants. Unfortunately, these marginal and smallholder farmers in haor wetlands are more vulnerable to crop loss risk and generally have a low adaptive capacity [15]. Moreover, they have limited access to alternative means of production [25]. By this means, climatic changes are drastically threatening their production, food availability and their farming system thus arising the risk of acute food insecurity.

To tackle climatic hazards and resultant crop loss, adaptation in the production system has evolved as a vital policy response for reducing climate vulnerability while protecting the livelihood and agri-food system of underprivileged farmers especially the ones from those developing countries [15,26]. This approach is also viable for the smallholder farmers in haor wetlands of Bangladesh. An extensive number of research works have already been carried out in the haor wetlands to evaluate different aspects of adaptation and climatic vulnerabilities. Worth mentioning of some are mainly focused on the situational explanation of climate variability [19,27], the impact of the flash flood events on crop production [18,28,29] and dominating adaptive factors and localities' vulnerability against changing climate [[30], [31], [32], [33]]. However, despite increasing recognition of the proficiency of adaptive cropping practices to secure food production and sustainability in changing climate [[34], [35], [36], [37]], significant research and their implications are yet to be done on the barriers confronted by haor farmers to adapt with the challenges in haor agricultural systems. Moreover, until recently no study is conducted on the ways of overcoming barriers to adapt rice farming to flash flooding in the haor wetlands of Bangladesh.

Thus, based on these findings and factual evidence, we hypothesized that rice farming in the haor wetlands of Bangladesh is greatly affected by climate change induced flash flood events. Concurrently we also hypothesized based on the field observations that farmers are facing different problems to adopt their farming system to secure their rice production. Hence, the objectives of this research are: 1) To identify the barriers to adapt rice farming to recurring flash floods in the haor wetlands of Bangladesh and 2) To explore the ways of overcoming barriers to adapt rice farming to recurring flash floods in the haor wetlands of Bangladesh. Fulfilment of these objectives gives new insights into farmers’ adaptive behaviour and the major obstacles they are facing in coping with climate change extremities. Furthermore, outcomes of this research work offer a pathway to attain a sustainable agricultural system in the haor wetlands that is shown in the discussion section.

2. Research methods

2.1. Study area

The data were collected through a multistage sampling technique from flash flood-prone haor areas of Bangladesh. Sunamganj is one of the five haor districts located in the north-eastern haor wetlands of Bangladesh [38]. This district covers a 3669.58 sq km area and is located between 24°34′ and 25°12′ north latitudes and between 90°56′ and 91°49′ east longitudes [39]. Sunamganj was selected purposively as it has more haor wetlands than any other district in Bangladesh. Due to this unique hydrology and geographical setting, this district is affected regularly by flash floods where about 67% of inhabitants are mostly dependent upon agriculture [12,40]. Research objectives were focused on the impact of flash flood events on the secured harvest, food availability and stable food production status, therefore, only flash flood-affected areas of the haor wetland region were considered in the research sampling frame. Tahirpur and Bishwamvarpur upazila of Sunamganj district, one of the major rice growing areas located in north-eastern haor wetlands of Bangladesh were selected as study areas where about 80% of the boro crops due to flash flood has been lost in recent years [12].

Two villages from Tahirpur (Dakhin Sreepur, Solemanpur) and Biswamvorpur upazila (Raipur, Bahadarpur) were selected on account of their degree of severity of flash flood damage every other year [12,41] (Fig. 1).

Fig. 1.

Dakhin Sreepur and Solemanpur villages of Tahirpur upazilla and Raipur and Bahadurpur villages of Bishwamvarpur upazilla of the Sunamganj district situated in the north-eastern part of Bangladesh where the study was conducted.

2.2. Sampling and data collection

Following the farmers’ household list of the Agricultural Extension office, 10% of the total population will be purposively selected as the sample. Because a good maximum sample size is usually 10% selected randomly which represents the total population [42]. The total household number in our study area is 1046 and we randomly selected 10% of the respondents (1046 × 10% = 104.6–105). We also kept 10 respondents as reserved. Thereby, 115 farmers were finally selected randomly representing ∼11% of the total sample population of 1046 households.

Data collection for this study was undertaken from October to January 2021 using a mixed-method research design. Several data collection methods such as Key Informant Interviews (KIIs), Focus Group Discussions (FGDs), Semi-structured interviews and a review of literature, Department of Agricultural Extension (DAE) official documents and NGO reports were utilized to collect both qualitative and quantitative data. Purposive sampling was done to select the respondents including haor farmers, farmer group leaders, union parished chairman, staff of the Department of Agricultural Extension Education (DAE) and local journalists, who participated in or facilitated at least one of the data collection methods. In total 4 KIIs and 4 FGDs were conducted with 4 and 32 respondents respectively. Hence, in each FGD 8 participants were involved following the work of Moser and Korstjens in 2018 [43]. Before data collection, a literature review was done to identify major problems faced by haor farmers in adapting to the changing climate. Then, Key Informant Interviews (KIIs) and Focus Group Discussions (FGDs) were conducted to explore necessary qualitative data. Initially, the respondents were asked generic questions regarding their socio-economic status, rice farming, climate change situation and its impact in the area, flash flood events and resultant damage and the major barriers that were pre-selected from the literature review to evaluate relevance and impact in the real field situation. Several follow-up questions were also asked in the later stage on the existing gap in policy implementation and future outlook to overcome the barriers and facilitate better adaptation of farming strategies.

The data obtained through FGDs and KIIs were recorded, transcribed and analyzed. Responses were coded manually to evaluate the major barriers to farmers' adaptation and the ways of overcoming the barriers. The transcriptions and the relevant secondary data were repeatedly read through line by line, and texts were identified as relevant to the barriers faced by haor farmers to adapt rice farming, the major reasons behind the barriers and the possible solutions to overcome the barriers. Then codes were assigned manually until saturation was reached. Thus, this process identified fourteen (14) major barriers proven to be relevant and significant and severely confronted by the haor farmers. These 14 barriers were divided into five aspects or themes and nine issues contributing to farmers’ problem confrontation behaviour were also identified in the process. These themes and issues were later used to describe research findings and to develop the semi-structured questionnaire for the survey.

The purposively selected 115 respondents were then surveyed with the aid of a semi-structured questionnaire to collect the quantitative data. To quantify the extent of problem confrontation regarding these fourteen (14) identified barriers, a Likert scale was used to represent the Problem Confrontation Index (PCI) levels (0 = problems not encountered, 1 = less confronted, 2 = moderately confronted, and 3 = severely confronted). Then the selected 14 barriers were ranked in order using descriptive statistics. Thereafter, the score of individual responses from the Likert scale was summed up and divided further to get the mean problem confrontation score of the farmers. In this way, the target farmers were categorized based on their PCI score to evaluate the state of problem confrontation behaviour among them.

2.3. Statistical analysis

The previously identified nine issues were explored as the explanatory variables for farmers' problem confrontation behaviour. Collected data on these socio-economic traits were coded and analyzed using descriptive statistics such as frequency, counts, mean, standard deviation and percentages. All statistical analyses like descriptive statistics, normality tests and correlation tests were performed using Statistical Package for the Social Sciences (SPSS, Version 26).

To check if the data is normally distributed or not, we checked the distribution of the residuals using a histogram. The shape of the distribution of residuals follows the shape of a normal curve. So, the histogram shows the normality of the standardized residuals of our data. The normal probability plot also proved the normal distribution of our data. Another way of checking normality is the Kolmogorov-Smirnov test or the Shapiro-Wilk test [44]. For datasets small than 2000 elements, the Shapiro-Wilk test is used where the null hypothesis is that the data is normally distributed [44]. In our case, since we have less than 2000 elements for our dependent variable (115 to be specific), the Shapiro-Wilk test is used. The Sig. value of the Shapiro-Wilk Test is greater than 0.05, so we can conclude that our data comes from a normal distribution. Hence, it is established that our dependent variable (problem confrontation) is normally distributed for each level of our independent variables.

To check the homoscedasticity, we run several tests. Firstly, to graphical check for homoscedasticity we used a scatterplot which shows that the homoscedasticity pattern is okay [45]. We also run Levene's test of homogeneity for testing the homogeneity of variance [46]. For our dependent variable (Problem confrontation behaviour), the p-value based on the mean is greater than the alpha value (0.05) against all the independent variables. So, it can be concluded that the variance in farmers' problem confrontation behaviour against the selected nine independent variables is homogenous.

The collinearity diagnostics from the SPSS output give a condition index value of less than 30. So, there is no strong presence of multicollinearity. The collinearity tolerance value from the coefficients output table is also greater than 0.5 with less than 10 statistics VIF (Variance Inflation Indicators) value. So, there is no strong multicollinearity between the independent variables.

As all the assumptions are satisfied, to conclude we did Pearson's correlation coefficient (r) analysis which was employed to predict the variance in farmers' problem confrontation behaviour because of these nine selected issues or traits.

2.4. Empirical approach

2.4.1. Problem confrontation behaviour of the haor farmers

For evaluating farmers' problem confrontation behaviour, respondents were asked to score the selected farming problems against a Problem Confrontation Index (PCI) formula modified from Masud et al., 2017 [32] and Uddin et al., 2014 [33]. The modified PCI used for this research was,

$$\mathrm{P}\mathrm{C}\mathrm{I}={\mathrm{P}}_{\mathrm{n}}\times 0+{\mathrm{P}}_{\mathrm{l}}\times 1+{\mathrm{P}}_{\mathrm{m}}\times 2+{\mathrm{P}}_{\mathrm{h}}\times 3$$

Where,

P_n = Frequency of farmers who rated the problem as not encountered.

P_l = Frequency of farmers who rated the problem as less confronted.

P_m = Frequency of farmers who rated the problem as moderately confronted.

P_h = Frequency of farmers who rated the problem as severely confronted.

After calculating the individual PCI score for the farmers, they were categorized into three groups, farmers who were less confronted = < (Mean – 1SD), moderately confronted = (Mean ± 1SD) and highly confronted = > (Mean + 1SD) during adapting different farming practices.

Moreover, to assess the significant factors that shape the problem confrontation behaviour of the farmers in climate change adaptation, a Pearson correlation analysis was used where farmers' socio-economic traits were fitted in the correlation matrix model. The correlation matrix model was estimated with a 95% confidence interval (CI) at a 5% level of significance.

3. Results

This section elaborates on the barriers to adapt rice farming to flash flooding, the extent of barriers faced by the Haor farmers to adapt to flash flooding and the traits of farmers' problem confrontation behaviour.

3.1. Barriers to adapt rice farming to flash flooding

A total of fourteen (14) barriers were identified to adapt rice farming to flash flooding in the Haor areas. These barriers were categorized into five aspects namely farming, economic, infrastructural, organizational and psychological aspects. The score for barriers ranged from 165 to 331 with a possible range of 0–345 as shown in Table 1.

Table 1.

The rank order of different barriers confronted by haor farmers (n = 115).

Aspects	Barriers/Factors	Extent of severity				Problem Confrontation Index (PCI)	Rank
		3	2	1	0
Farming problems	Lack of availability of submergence-tolerant rice variety	101	14	0	0	331	1st
	Labour shortage during pick agricultural operations	35	65	15	0	250	5th
	Lack of knowledge about different adaptation options	40	66	9	0	261	4th
Economic problems	Lack of access to inputs	68	45	2	0	296	3rd
	Limited access to the market system	95	20	0	0	325	2nd
	Lack of access to the credit facility	18	94	3	0	245	7th
Infrastructural problems	Transportation problem	6	72	33	4	196	13th
	Lack of proper storage facility	1	78	32	4	191	14th
Organizational problems	Lack of training facilities for farmers	8	103	4	0	234	10th
	Lack of provision of alternative income sources	27	60	28	0	229	11th
	Dominance of local elites thereby creating unavailability of land	38	55	22	0	246	6th
	Limited access to potential climate change information sources	11	100	4		237	9th
	Lack of extension service	11	101	3	0	238	8th
Psychological problem	Farmers' doubt/disinterest towards new technologies	30	52	33	0	227	12th

The results showed that the most prominent barrier to adaptation was the lack of availability of submergence-tolerant rice variety with a PCI score of 331. The next significant barriers were farmers' lack of market access (PCI = 325), lack of access to inputs (PCI = 296) and knowledge of adaptation options (PCI = 261). In addition, labour shortage during pick agricultural operations (PCI = 250) makes haor farmers unable to harvest their rice in time. Moreover, the dominancy of local elites in the haor leasing and tenancy system (PCI = 246) limits farmers' control and ownership over their productive land and thus making these marginal farmers and landless ones unable to adopt new practices and grow different varieties at different timeframes that further limits their adaptability. Supplementary to these barriers, farmers’ lack of access to credit facilities (PCI = 245), lack of access to extension services (PCI = 238), limited access to information sources on potential climate changes (PCI = 237), lack of training facilities (PCI = 234), lack of provision of alternative income sources (PCI = 229), farmers disinterest towards adopting new practices (PCI = 227), and existing transportation problem in haor wetlands (PCI = 196) and lack proper storage facility (PCI = 191) along with inadequate threshing and drying places also made it challenging for the farmers to secure their harvest during flooding seasons.

3.2. The extent of barriers faced by the haor farmers to adapt to flash flooding

The findings revealed that the PCI score for individual respondents ranged from 24 to 40 against the possible range of 0–45, while the mean score was 31.91 as presented in Table 2. It is evident that 85% of the respondents faced moderate to severe barriers to adapt rice farming to flash flooding.

Table 2.

The extent of barriers confronted by the haor farmers to adapt to flash flooding.

Categories (score)	Number	Percent	Mean	Standard deviation
Less confronted (up to 29)	30	26.1	31.91	3.243
Moderately confronted (30–36)	80	69.6
Severely confronted (>36)	5	4.3
Total	115	100

3.3. Determining traits of farmers' problem confrontation behaviour

The correlation matrix model identified several socio-economic traits of the haor farmers that influence their problem confrontation behaviour associated with the barriers faced to adaptation as shown in Table 3.

Table 3.

Issues contributing haor farmers’ problem confrontation behaviour to adapt rice farming to flash flooding.

Issues	Problem confrontation behaviour to adapt to flash flooding
	Correlation coefficient (r) values	p-value
Age	.018	.847
Level of education	−.044	.642
Family size	−.015	.876
Farm size	−.063	.505
Annual family income	−.206*	.027
Extension media exposure	−.228*	.014
Agricultural training experience	−.131	.164
Perception on climate change	−.184*	.049
Adaption behaviour	−.180	.054

Note: *Correlation is significant at the 0.05 level (2-tailed) with p-value ≤0.05.

The findings revealed that farmers' annual family income, communication exposure to the extension media and their perception of climate change had a significant but negative correlation with their problem confrontation behaviour. On the other hand, farmers' level of education, family size, farm size, agricultural training experience and their adaptation behaviour had negative correlation value, but this does not represent a significant correlation with their problem confrontation behaviour. Now, the inverse significant correlation between some socio-economic traits and farmers' problem confrontation behaviour was caused by a multitude of factors. Here, the reason behind negatively significant correlation behind farmers' annual family income and problem confrontation behaviour lies in the fact that farmers with more economic resources or financial stability tend to be risk taker. Besides, because of their higher income and availability of resources, it is found that they comparatively suffer less in confronting the barriers to adapting to climate change risks to their crop production. On the other hand, in case of the negatively significant correlation between farmers' communication exposure and their problem confrontation behaviour, it is evident that farmers with more exposure to extension media often have the additional technical skill, access to the information source and logistic support available to cope with the barriers to adapting to climate change risk. So, in turn, they also face less confrontation in adapting their farming practices with changing climate condition. In case of the negative significant correlation between farmers’ perception on climate change and problem confrontation behaviour, it was found that, farmers with a clear and positive perception on climate change and its impact on the agricultural production system have confronted fewer problems to adapt to the climate change risks. Because they were more aware of the climate change uncertainty and risks in the rice production system and were more efficient in responding to mitigate the adverse effect of the changing climatic condition.

4. Discussion

4.1. Prominent barriers in boro rice production

To provide an appropriate adaptation policy or strategy framework for the haor wetland community, it is crucial to pinpoint the constraining factors. That's why this study identified the prominent barriers to adapting to the changing climatic scenario for secured boro rice production (Table 1). The unavailability of the submergence-tolerant varieties was ranked as the top most barrier to adopt to the flood resultant rice production loss by our respondents during the survey. This is also backed up by available literature and proven as a major adopting technique for rainfed and flood-prone rice production areas in Asia [7,47,48]. Because, there are previous studies indicating the need to adopt submergence tolerant variety in climate-prone delta regions of Bangladesh [7,49,50]. Findings from Dar et al., 2018 [51] also predicted adoption of flooding or submergence-tolerant rice varieties as the most viable solution for low-laying rice growing areas like the haor wetlands of Bangladesh. Till now, the Bangladesh Rice Research Institute (BRRI) and the Bangladesh Institute of Nuclear Agriculture (BINA) developed and promoted five submergence-tolerant (Sub1) rice varieties. But, their availability at the field level and adoption is still quite low [52,53]. So, this is an ever-present problem for the small-holder farmers located in marginal areas with limited extension service and access to input facilities like the haor wetlands of Bangladesh. But the situation is comparatively more extreme for the haor wetlands because, besides the seasonal flooding, they have extreme cases of sudden flash flood events resulting from the changed climatic condition [20]. In such a situation, unavailability or limited availability of flooding-tolerant rice varieties to the haor wetland farmers makes them more susceptible to crop loss by these climate resultant flash flooding events. Bairagi et al., 2021 [7] and Dar et al., 2021 [54] also stated unavailability or limited availability of flooding-tolerant rice varieties to the farmers makes them more susceptible to crop loss by flooding events which further proves the legitimacy of our study findings. On the other hand, though there are previous studies on monsoon flooding events, the devastating impact of flash flood in haor wetland areas are sometimes get ignored [55]. Moreover, with the changing climatic situation, the intensity and extent of the flash flood resultant crop loss is becoming severe. Farmers sometimes lose almost 100% of their ripened crops to sudden flash floods [20]. Therefore, it is crucial to investigate the hidden factors that are causing limited adoption of these stress-tolerant or submergence-tolerant varieties, even though these varieties were released almost a decade ago and there is scientific evidence to prove their profitability and relative advantage under changing climatic conditions.

The second-ranked barrier by the farmers is their limited access to the local market system which prevents them from getting the best price for their harvests. In addition to this factor, farmers' lack of access to input supply and farm machinery along with the high price of these inputs and lack of technical assistance is mostly responsible for hindering the farmers from strengthening their economic competence to adapt. Supporting statements from Akhtar et al., 2018 [56] and Fahad and Wang, 2018 [25] predicted farmers' lack of transportation facility, high price and unavailability of inputs hinders farmers' adaptation to climatic stress and crop damage further validating our findings. Contributing to these factors, farmers' lack of access to information on adaptation options and extension exposure often kept them in dark about what are the best options available at the current moment to minimize their crop loss. In that way, this cognitive knowledge gap makes it impossible for the farmers to adapt to the climate change extremities to secure their rice harvest [25,57].

Labour shortages during harvesting, and existing transport and storage problems in the haor areas are also making the farmers unable to harvest and secure the ripened crops timely before the onset of an early flash flood. In addition to this, leasing and tenancy systems in haor wetlands limit farmers' access over the cropping land and also hold them back from switching to or adopting new practices. This complex multi-stakeholder scenario also discourages farmers to take risks [58,59].

On the other hand, contrary to farmers' interest in adapting to the limited technical assistance and training facilities, credits and subsidies from the government, the lack of provision for alternative income or livelihood opportunities also limits their competency to grasp extensive adaptive strategies. Akhtar et al., 2018 [56] and Siddiquee et al., 2019 [59] also predicted farmers' lack of access to the training program and government subsidies diminishes their competency to adapt to climate change which is in line with the study findings. Moreover, farmers' disinterest towards adopting new practices generally results from the complex interaction between different socio-psychological and infrastructural obstacles and their age-old traditional mindset towards modern technologies [58]. This psychological factor adds to the explanation of why some recommendations or techniques are never adopted by farmers, despite their advantages [60,61]. Hence, this was also identified as one of the significant barriers to behavioural change amongst farmers [62]. Therefore, the issue also plays a major role in shaping their adeptness and in fighting the roadblocks to secured boro rice production.

These findings on significant barriers to boro rice adaptation to flooding are not only prominent in Bangladesh but also in other major rice-producing countries such as India [63], Pakistan [25], Nigeria [64], Africa [65], Philippines [66] and so on [67,68]. Based on this empirical evidence, it is evident that future policy interventions should focus more on overcoming these barriers for a well-suited boro rice adaptation package in haor wetlands.

4.2. The extent of barriers faced by the haor farmers to adapt to flash flooding

According to the prediction of the IPCC report 2022 [15], climate change risks in crop production are one of the major threats to the sustainability of the food system. To mitigate this risk, they suggested adaptation to climate change as a prerequisite for secured crop production and sustainability of the food system. Therefore, this research work focused on the barriers faced by farmers in haor wetlands in adapting their rice production system to climate change risks. These haor wetlands have a rice-based economy and rice is the dominant crop and staple food for millions of inhabitants. So, sustainable rice production means a stable food system, livelihood and overall sustainability in the haor wetlands. On the other hand, sudden losses of crops due to climate and weather events exposed millions of people in Asia to acute food insecurity. This is the harsh reality in haor wetlands of Bangladesh where flash flood events have devastating impacts on small-scale food producers and low-income households by destroying rice production every other year. This disastrous situation of crop loss is exposed in the reports published by the government organizations of Bangladesh, for example - Bairagi et al., 2021 [7]; BWDB, 2018 [12]; and Kamruzzaman et al., 2022 [14]. These articles showed the devastating impact of recurring flash food events on rice production and the influence of climate change in more frequent flash flood events and changing climatic patterns in haor wetlands. For the nature-based rice production system in the haor wetlands, these changes are a major threat to rice production and the sustainability of the agri-food system. So, adaptation to these climatic changes is a vital response for haor wetlands community to attain sustainability. Under this situation, before proposing any adaptation measure and strategic intervention, the state of the farmers should be taken into consideration. From the study findings (Table 2) it is obvious that a vast majority of the respondents (85%) faced moderate to severe barriers while adapting different farming practices to combat flash floods. Similar studies conducted by Dey et al., 2021 [20] and Ferdushi et al., 2019 [69] in haor areas also stated most of the farmers fought with indispensable constraints in adopting strategies to secure their harvest. So, necessary steps should be taken to boost their efficiency and adeptness to combat climate change-prompted crop damage. This can be done by safeguarding adequate access of these farmers to input supply and technical assistance while confirming better extension services and easy access to reliable information sources. Moreover, provision on credit supply and government subsidies, need-based training and consultation should also be ensured [64,70].

4.3. The influence of socio-economic traits on haor farmers’ problem confrontation behaviour to adapt to flash flooding

The findings of this research identified a number of socio-economic traits that significantly influenced haor farmers' problem confrontation behaviour to adapt to flash flooding. To support this significant correlation, a number of previous studies conceptually reported the link of socio-economic issues with farmers' climate change adaptation. For instance, the works of Comenetz and Caviedes, (2002) [71] and Thomas et al. (2007) [72] showed the effects of climate change on agriculture to understand how farm households or farming communities practice different adaptive strategies to overcome crop production problems due to environmental shocks. Some authors also linked sustainable farming practices with the community's resilience against the risk or with their capacity to adapt (e.g. Ericksen et al., 2009 [73]; Ericksen, 2008 [74]; Misselhorn et al., 2012 [75]). Scholars also have provided some evidence to support the influence of the socio-economic context on the conditions or barriers that enable farmers to adapt the agricultural systems to the effects of climate change [76]. Therefore, it is crucial to understand how socio-economic traits might affect the farmers' risk aversion decisions, and, therefore, might then help to transform the current situation into a less vulnerable cropping system with ensured farm sustainability.

In this research, the identified contributing factors behind haor farmers' problem confrontation behaviour were their annual family income, exposure to the extension media and perception on climate change. These socio-economic traits found to have a significant and inverse correlations with the problem confrontation behaviour. Thus, any desirable change in these identified socio-economic traits are expected to result in a significant variation in their capability and magnitude to face different barriers to adapt to climatic shocks i.e., recurring flash floods in our research. To explain more specifically, the more one farmer has economic solvency, the better he can cope with the problem faced during the adoption of farming practices. This fact proves the negative significant correlation between farmers' annual family income and problem confrontation behaviour. On the other hand, regarding farmers’ exposure to different extension agents and communication media, farmers with better communication exposure have better capability to overcome the barriers to adapt to flash flooding. Because, with better and established media exposure, they have more knowledge on how to adapt to the changing climate and what are the best-suited farming practices. Moreover, regarding farmers' perception on climate change, a better and positive perception makes them aware of the climate change uncertainty and future threats to the cropping system. Therefore, they are compatible and more efficient in assessing their situation and acting on it to secure their harvest. These facts prove the negative significant correlation between farmers perception and their problem confrontation behaviour. In support of the significance of these specific socio-economic traits, evidences were found in some previous studies (e.g., Bryan et al., 2009 [67]; Hossain et al., 2011 [77]; Kim et al., 2017 [64]; Masud et al., 2017 [32]), where these specific traits of the farmers significantly shaped their problem confrontation behaviour in crop production. Hence, interventions from the policymakers focusing on these socio-economic traits of the farmers will be best fitted and precisely tailored to safeguard sustainable production in the haor wetlands. The conceivable interventions will also help to create new livelihood opportunities, providing micro-credit or financial assistance, providing agricultural subsidies, better media exposure and using local communication channels, creating awareness campaigns on climate change impacts, and timely and better access to the area-specific weather forecast and meteorological data for the farmers.

4.4. Pathways to adapt to flash flooding and ensure sustainability in rice farming in the haor wetlands

There is a significant gap in the literature world to investigate farmers' decision-making behaviour regarding the adoption of such climate-smart production practices [[78], [79], [80]]. Hence, in our research work, we focused on farmers' problem confrontation behaviour and linked it with their adaptation decision-making by exploring the ways of overcoming barriers faced by haor wetland communities to adapt against climate change to secure their rice production. Moreover, based on the literature review and collective findings from this study, we are proposing a pathway to attain sustainability in the haor wetlands while considering the socio-economic actors of the system (Fig. 2). Because, currently, there are no suggested pathways designed specifically for the haor wetlands to ensure rice farming sustainability. Besides, no significant research is done so far considering the unique rice-based agricultural system, climatic hazards and crop loss situation that is occurring every other year in haor wetlands of Bangladesh. Hence, this conceptual model proposed in the paper will contribute to the knowledge on flash flood adaptation and will also guide to formulation of future research works in the context of haor wetlands. This proposed pathway shows different barriers identified in this study (farming problem, economic problem, infrastructural problem, organizational problem and psychological problem) that shape farmer's problem confrontation behaviour. In addition, the socio-economic traits of the farmers more specifically their annual family income, exposure to the extension media and perception of climate change significantly influence their problem confrontation behaviour in boro rice production. These barriers and farmers' socio-economic characteristics influence their adaptability to the constraints in boro rice production. So, desirable change in their problem confrontation behaviour could be achieved through influencing farmers' socio-economic traits and targeted policy interventions. With the change in farmers' problem confrontation behaviour to mitigate these barriers, boro rice production could be secured from climate change's resultant loss. These changes will result in adaptive capacity building to mitigate crop loss due to the changing climatic situation. All the stated factors will, therefore, in the long run, will help in ultimately assuring the desired rice farming sustainability in the haor wetlands and in attaining the Sustainable Development Goals (SDGs) in the wetlands of Bangladesh [81,82] specifically SDG 1: No poverty (target 1:5 Build resilience to environmental, economic and social disasters), SDG 2: Zero hunger (target 2:4 Sustainable food production and resilient agricultural system), and SDG 13: Climate action (target 13:1 Strengthen resilience and adaptive capacity to climate-related hazards and natural disasters). To conclude, all the changes will result in a desirable shift towards a sustainable agricultural production system in haor wetlands of Bangladesh. In this way, this proposed pathway will help to explore ways to secure the sustainability of rice farming in the haor wetlands in face of climate change and weather climatic extremities [83,84]. Hence, this conceptual pathway links farmers' problem confrontation behaviour in adapting to the stress and uncertainties with their socio-economic traits and illustrated a possible way out through policy interventions and capacity building of the haor community to attain sustainability, availability and stability in food production in the long run and it will add significant directions to the future research. This conceptual pathway will also help to understand the interrelation and interdependence between farmers' socio-economic characteristics, their adaptation decision and the barriers they face to ensure food availability and stability in the haor wetlands. Moreover, it will provide a guideline to the policymakers to facilitate any future policy or strategic interventions while respecting the socio-ecological boundaries and situation prevailing in these haor wetlands. So, we can anticipate knowledge contribution from this conceptual pathway regarding flash flood adaptation strategy. In addition to addressing the current research gap, this will also facilitate discourses for future research works on the overall four pillars of the food security state that remain unexplored in this paper. Moreover, this conceptual pathway could be applied to any other wetland ecosystems facing similar climatic vulnerability to attain a secured and sustained agricultural production system.

Fig. 2.

A conceptual pathway explaining the nexus and collaboration between barriers in boro rice production and farmers' socio-economic profile to ensure sustainability in rice farming in haor wetland of Bangladesh.

4.5. Policy implications on overcoming the barriers to adapt rice farming to flash flooding

Our study elaborated on the current state of barriers to adapt to climatic vulnerability and different factors prevailing in the socio-ecological system that directly and indirectly influence the adaptive capacity of the haor community to flash flood events. Insights from the findings of this paper will motivate context-based policy formation and implementation to mitigate vulnerabilities of this delta floodplain and uncertainty in the agricultural production system. Besides, there is no ‘one-size-fit-for-all’ solution for all regions or communities. Different regions have different cropping patterns, geographical settings, different climatic risks and different needs and call for attention. So, as illustrated by the findings of this paper, identifying the major barriers that a community face in adapting and the contributing factors behind their adaptive decision-making will lead to need-based and target-oriented policy intervention based on the specific context i.e., haor wetlands of Bangladesh. Moreover, this paper also highlights the existing research gap and suggests some crucial target areas that are not given significant consideration till now and need urgent focus and interventions for development purposes. These target areas include an early warning system to make the farmers prepared before the onset of climatic hazards [20], governance of water management to deal with local environmental challenges like flood control and irrigation management for rice production [85], a damaged-based crop insurance scheme for the farmers after the onset of flooding to support them [86], promoting alternative income generating activities as a secondary income source to the rice growers [87] and coordination mechanism between different actors and stakeholders at local, regional and national level [88]. Hence, findings from this paper could motivate the government and policy makers on the necessity of allocating more resources, establishing effective coordination between government and local actors and providing logistic and technical support to the vulnerable communities to develop their adaptive capacity to climate change vulnerabilities.

4.6. Limitations and future outlooks

The threat of climate extremities to the sustainability of the rice production system is a major concern for the community resilience in the haor wetlands of Bangladesh. This study focused on the perspectives of boro rice farmers to explore their barriers to adapt to the effects of climate change and flash flooding. However, due to the seasonality of the agricultural production system in haor area, constraints in translating the local language and time and resource constraints, this research did not collect data from other farm stakeholders including input dealers, extension service staff and local public representatives. Besides, limited number of published reports on local farming system and farmers perspectives also restricted us to examine these aspects for this research work. There is also no up-to-date database to collect the environmental data or information on the physical attributes of the farming system like soil, water level, possible suitability of different varieties with the local ecosystem. Therefore, to follow up this work, we would suggest extensive research work on the overall picture of the barriers and their dynamics associated with sustainable rice farming in the haor wetlands. In addition, we suggest exploring the case study with qualitative data on the haor wetland community to better understand different dimensions of farmers' adaptation behaviour to overcome relevant barriers. Finally, the socio-psychological traits of the farmers that influence their barriers to overcoming behaviour could be inquired into in the future to provide more context on farmers’ viewpoints and offer contrasting views that could be valuable for policymaking. So, these windows could be taken into account for future studies.

5. Conclusions and recommendations

This study examined the barriers to adapt rice farming to recurring flash floods in haor wetlands of Bangladesh. The findings show that a number of barriers exist to adapt to flash flooding. These include a lack of availability of submergence-tolerant rice variety, lack of market access, lack of access to inputs, lack of knowledge of adaptation options and labour shortage during pick agricultural operations. The study also explored the ways of overcoming barriers to adapt rice farming to recurring flash floods in haor wetlands of Bangladesh. The results show that farmers’ barriers to adaptation and problem confrontation behaviour were also linked with their socio-economic traits. The socio-economic traits include annual family income, extension media exposure and perception of climate change.

Besides, this study elaborated a conceptual pathway to adapt to flash flooding and ensure sustainability in the haor wetlands of Bangladesh. This pathway indicates an opportunity to work on the existing shortcomings to minimize the vulnerability of haor wetland community to recurring flash flood events. Farmers in the haor region are all too familiar with flash flood events and are the worst victim of crop loss risks and livelihood vulnerabilities. But the policy options are limited to them due to their nature-dependent traditional boro rice production system and resource constraints to adapt to climate shocks. Therefore, effective collaboration and governance between the extension workers and policy makers and more logistic and financial support to the farmers are needed for fruitful adaptation of strategic interventions.

Finally, in addition to policy interventions, extensive research works are also needed to generate area-based strategies to overcome the barriers and boost rice farming in the study area. Future research should focus on farmers' cognitive abilities and competencies to act as risk-averse actors as this aspect remained unexplored in the study. These interventions would contribute to effectively managing climate change uncertainties with an assured sustainable farming system in the haor wetland region.

Author contribution statement

Smita Dash Baishakhy: Conceived and designed the experiments, performed the experiments, analyzed and interpreted the data and wrote the paper. Mohammad Ashraful Islam; Md. Kamruzzaman: Conceived and designed the experiments; Analyzed and interpreted the data and correction of the manuscript.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability statement

Data included in article/supplementary material/referenced in article.

Declaration of interest's statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.