The study was carried out in fourteen transboundary river-bank villages of two local administrative units (Tahirpurn and Sunamganj Upazila) in Sunamganj district, Bangladesh (Figure 1 and Figure 2). The study focuses on assessing livelihood status in Jadukata and Nawagang transboundary riverine inhabitants of extreme northeastern Bangladesh. The regions are marked by extreme isolation, sandy soil, and hazard-prone conditions. The climate of this area is humid-subtropical with a predominantly hot and humid summer and experiences a relatively cool winter. The area also belongs to the monsoon climatic zone, with the annual average highest temperature of 23˚C (August-October) and the average lowest temperature of 7˚C (January). Nearly 80% of the annual average precipitation of 3800 mm occurs between May and September ([26]).

The Jadukata River-bank area, situated in Tahirpur Upazila, spans between 25˚01' and 25˚12' north latitude and 91˚02' and 91˚19' east longitude, bordering Bishwambarpur Upazila on the east, Dharmapasha Upazila on the west, Jamalganj and Dharmapasha Upazilas on the south, and Meghalaya on the north (Figure 3; [8]). The entire study area consisted of the following areas: Sahidabad, Lawrer Ghar, Daler Par, Razer Gaon, and Mader Gaon on the eastern side of the Jadukata River; Barek Tilla, Gagodia, and Gagra on the western side of this river (Figure 4).

The Nawagang River, originating in Meghalaya’s East Khasi Hills district, receives 11,500 mm of rainfall annually and joins the Surma in Bangladesh through Sunamganj Upazila (Figure 3; [65]). Sunamganj Sadar Upazila, spanning 290.71 square kilometers, is located between latitudes 24˚49' and 25˚10' north and longitudes 91˚14' and 91˚27' east. It is bordered by Meghalaya and Bishwambarpur Upazila, Dakshin Sunamganj Upazila, Dowarabazar Upazila, and Bishwambarpur Upazila ([8]). Figure 5 shows the sampling locations from the Nawagang River-bank areas, which include Dolora, Kalipor Dolora, Kalipor, Kair Gaon, Adang, and Porun Motorkandi on both sides of the river.

Figure 1. Transboundary rivers among Bangladesh, India and Mayanmar (Curtesy Join Rivers Commission-Bangladesh, Ministry of Water Resource, GoB, 2013).

Figure 2. Transboundary rivers at Sunamganj District, Bangladesh.

Figure 3. Map of study area.

The sample size calculation was conducted for an infinite population using the below [19] formula (1) and adjusted for a specified population ([52]).

Figure 4. Jadukata riverine area for collecting survey data.

Figure 5. Nawagang riverine area for collecting survey data.

The sample size (n₀) for an infinite population is determined using a z score of 1.96, with a population proportion (p) of 0.5 and a margin of error (d) of 4 to 8%, based on a 95% confidence level. The total households (N) in two Upazilas were 87,488 ([10]), and a sample size of 201 was calculated from this finite population by the following Cochran’s modified formula ([71]) to reduce the sample size for correction of population proportionately:

Finally, the sample size for 87488 populations was determined to be 200, purposively along the riverside (Table 1).

Fourteen villages nearer to the Jadukata and Nawagang Rivers under the Tahirpur and Sunamganj Upazila were purposively selected (Table1) on the basis of the number of households. A list of crop cultivation and fishing dependent people from the Upazila Agriculture Office was collected and updated list was prepared with the help of local leaders like chairman, members, SAAO (Sub-Assistant Agricultural Officer) and local NGO (Non-Government Organization) workers. About 87488 household heads ([10]) was considered as the population and using formula of sample size (n) for estimating sample proportion, 200 household heads was randomly selected as sample. A reserve list of 30 household heads was also prepared to use only if a respondent in the original list would not be available for interview. Data were gathered following face-to-face interviews using a semi-structured questionnaire from January to February 2022.

Table 1. Households from required villages at different riverine areas.

Farming activities, environmental, financial, and social questionnaires related to their living opportunities were collected by asking ten questions. The all-up opportunities were measured by respondents as 1 narrating to “strongly agree”, 2 narrating to “agree”, 3 narrating to “neutral”, 4 narrating to “disagree”, and 5 narrating to “strongly disagree” against ten statements. The range of possible mean scores for the main statements was 1 to 5, where less than 2.5, 2.5 to 3.5, and greater than 3.5 indicate statement contributions in all-up situations are positive, neutral, and negative, respectively. For assessing the livelihood index, the respondents were investigated about their livelihood status by asking twenty-seven modern questions that were designed in five capitals (Table2). Respondents were assessed on their living status by answering questions with a score range of 0 to 1, with a positive response indicating a positive status and a negative response indicating a negative status.

Table 2. Variable used for livelihood Capital assessment of transboundary riverine habitants.

The IBM SPSS program (Version 20) was used to perform proportion tests, box plots, spider diagrams, and cluster analysis through a laptop computer in the Central Lab, Bangladesh Agricultural University, Mymensingh. MS Excel (Version 2010) was used to calculate the livelihood assessment index by equal weighting. Arc GIS software (Version 10.8) was used to draw a research area map in the Central Lab, Bangladesh Agricultural University, Mymensingh. GPS Coordinates were used to measure latitude, longitude and altitude of the research areas in the selected field.

The Department for International Development’s (DFID) Sustainable Livelihood Framework served as the foundation for the development of the LAI ([24]). The transboundary riverine inhabitant’s livelihood can be specified as follows: The livelihood of a transboundary riverine inhabitant is determined by their human, financial, physical, natural, and social capital.

The livelihood status of transboundary riverine inhabitants was assessed using five livelihood capitals, based on a literature review and revisions after a pre-test. Table2 displays the livelihood indicators that were taken into consideration. The five capitals’ indicators were assessed using equal weighting to determine their livelihood status among transboundary riverine inhabitants, and a combined index was created. The balance weighting was used to ensure equal distribution of indicators across all capitals in the all-up index, despite different scales and units ([2]; [33]). Standardization was achieved using Equation (3):

where, the original value of indicators is S, the maximum and minimum values of indicators are S_max and S_min, and 100 for the indicators measured as a percentage ([33]). The average value of each capital was estimated using Equation (4) after standardizing.

where, M_cj means the value of one major capital j; index_si represents the index of i indicators, and n is the total number of indicators in each capital. Finally, the weighted average of LAI was estimated using Equation (5):

Equation (5) can be rewritten as

where, LAI means the livelihood assessment index, W_mj means the weightage of element j, W_H, W_F, W_P, W_N, and W_S are the weight values of the human capital, financial capital, physical capital, natural capital, and social capital, respectively.

[75] define human capital as the combination of knowledge, skills, attitudes, education, mental and physical health, work capacity, and leadership qualities that enable individuals to achieve their livelihood goals. Table3 presents the human capital findings from the two riverine communities, showing an average physical and mental fitness rate of 80.50%, which is notably higher than the 74.2% reported in Colombia ([31]). However, this level of fitness is still considered moderately low in the context of this study due to limited access to proper medical treatment and hospitality services (52.95%), inadequate sanitation facilities (78.33%) (Table4), and a rising prevalence of waterborne diseases (rules: positive and neutral; Table9). Mental well-being, a critical component of human capital, requires active engagement in recreational activities, which are currently lacking in the study areas ([32]). The literacy rate among residents in the extreme northeastern part of Bangladesh was found to be 69.16%, aligning closely with the national average ([12]) and significantly higher than the 6% educational attainment reported among floating fishermen ([62]). While primary schools are available locally, access to secondary and higher education requires travel to the Upazila level, as no such facilities exist within the immediate locality.

Table 3. Human capitals of respondent’s households.

The communities demonstrated a strong labor force capacity, with 80.50% of respondents able to perform work effectively—higher than the 60% fishing capacity among floating fishermen ([62]) and the 46.74% labor force capacity observed among farmers ([36]). Enhancing this capacity further would require improvements in public health, household wealth, and transportation infrastructure (Table9). However, leadership qualities (27.7%) and interpersonal attributes such as good behavior and humanity (44.35%) were found to be low across both study sites, hindering effective collaboration and community-driven initiatives. The absence of training and skill development programs has contributed to this gap. Therefore, capacity-building interventions focusing on leadership and behavioral development are essential to strengthen their ability to work cohesively. The average Human Capital Index (HCI) for the study area was calculated at 0.61, which is considerably higher than the 0.3364 reported in poverty-stricken regions of southwest China ([37]), reflecting comparatively better human capital conditions in the studied communities.

The physical capital indicators (Table4) reveal that 98.65% of respondents had access to electricity, 84.95% to safe drinking water, and 81.00% to transportation facilities. Approximately one-third of the respondents lived in semi-concrete houses. This aligns with earlier findings by [5], who reported that 26% of fish farmers in Hamirkustia and Katabari unions resided in half-built homes, and [7], who found that 34% of homes were semi-brick structures—figures comparable to those observed in the present study. The high electricity access rate (98.65%) is consistent with the findings of [2] and [7]. Respondents reported that drinking water was readily available, primarily from nearby tube wells and other local sources. In contrast, [62] noted that all floating fishers used fitkiri (potassium alum) to purify river water an unhygienic method. Transportation facilities were generally adequate, largely because roads and infrastructure in the study area are not submerged annually, unlike the haor regions. In monsoon-prone areas, agriculture and transportation infrastructure such as roads and culverts are often inundated for 6 - 7 months each year, posing significant challenges ([14]; [17]).

Table 4. Physical capitals of respondent’s households.

Access to healthcare services among respondents was limited, primarily due to lack of awareness and financial constraints. Although 52.95% sought treatment from district hospitals or Upazila health complexes, 47.05% could not access competent medical care due to financial limitations, relying instead on village doctors or traditional healers (Kobiras) (Table4). Similarly, [7] reported that 54% of fishing households depended on village doctors with limited qualifications, 28% received treatment at Upazila health complexes, and only 18% consulted licensed MBBS doctors. Sanitation facilities in the study area were also inadequate. As shown in Table4, 78.33% of respondents used pakka, semi-pakka, or kacha toilets, while 21.67% lacked any form of sanitary facility. Comparatively, [4] found better sanitary access among fishermen in Mymensingh district, with 12.5% using pakka toilets, 62.5% semi-pakka, and 25% kacha. Overall, the physical capital index for the study area was calculated at 0.72, which is notably higher than the index value of 0.4279 reported in poverty-stricken districts of southwestern China ([36]).

The respondents can use open water for their livelihood purposes, but the qualified water (18.50%) is deteriorating, making it unsuitable for drinking. In another research it was found that 11% of fishermen used open water for drinking purposes and others ([21]), which is lower than the result found in this study. However, they have minimal access to usable forest (11.605%), where they can use wood for building, repairing furniture, and other purposes (Table5). This is a higher percentage than the results reported by others ([2]; [79]). They need to buy extra wood from the market for their own family’s use. Per capita land use is 20.50 (decimal) for building houses and cultivation, which is more than the results reported by other researchers ([79]; [62]) because the respondents are not floating fishermen but riverine inhabitants. More importantly, their soil erosion protection (70.80%) is also good to protect their houses, roads, networks, and cultivations from flooding. There is no navigation in the rivers, but the surrounding air quality (55.70%) is good for living beings (Table5). Dhaka’s air quality index (AQI) is highest in winter months (January, February, and March) and lowest during the monsoon season (July, August, and September). This is due to increased pollution from anthropogenic sources and activities like brick kilns, motor vehicles, industries, and construction sites ([44]). The natural capital index of 0.36 is lower than the average of 0.4038 in southwest China, as reported by [37].

Table 5. Natural capitals of respondent’s households.

Social capital is a crucial element that fosters trust, cooperation, and solidarity, ensuring shared needs and purposes in life. Social resources are defined by relationships and networks among neighborhoods, groups, and families ([20]). People use informal safety nets, based on these relationships, to maintain their livelihood plans during emergencies and challenging times ([51]). The study revealed that 87.3% of respondents shared their socially held knowledge with societal members due to a strong network relationship (67.70%) among them (Table6). [2] reported that sharing socially held knowledge is 80%, which is lower than the results found in this study. About 12.90% of the respondents had clubs, playgrounds, and community centers for amusement as a social feature based on faith, which is lower than the value found (32.7%) in Akwa Ibom State, Nigeria ([82]); 53.15% of the respondents were members of NGOs, union councils, and local institutes such as libraries, hospitals, schools, etc.; and 46.85% of them were not involved in any associations (Table6). In another research it was found that households have 81.8% NGO membership in Nigeria ([82]), which is higher than the finding of this research. The respondents’ social capital findings align with previous research findings ([42]; [2]). The social capital index of o.55 is lower than the floating fishermen’s index of 0.624 reported by [2].

Table 6. Social capitals of respondent’s households.

The analysis of financial capital revealed that the average household income among respondents was approximately US$134.23 per month, translating to US$33.56 per capita per month (Table7). This figure is significantly lower than the national per capita monthly income of US$94.69 ([11]). After covering household expenses, respondents were able to save an average of only US$7.59 per month—a figure typical of populations living just above the poverty line. Similar patterns of negligible savings have been reported among fishermen in Indonesia ([38]) and farmers in the Meherpur district of Bangladesh ([68]). Respondents had limited access to formal or semi-formal loans due to a lack of collateral. On average, they secured US$39.38 annually in collateral-free loans from various NGOs (Table7), a finding that closely aligns with [2], who reported a similar figure of US$39.59. Many respondents refrained from taking financial loans, citing poor credit histories, lack of loan security, and a complex application process ([53]; [78]; [90]). Consequently, only 14.3% were engaged in small business ventures, a rate much lower than the 83.6% reported by [82]. In addition to NGO loans, respondents borrowed informally from financially better-off relatives, neighbors, or friends, averaging US$11.24 per year (Table7). This is higher than the US$5.28 reported by [79] for small-scale fishermen.

Table 7. Financial capitals of respondent’s households.

Farming remains the primary source of income for the majority of respondents, with 84.61% involved in agriculture, poultry, dairy, or fish farming—higher than the 68.2% observed in Akwa Ibom State, Nigeria ([82]). The overall financial capital index was calculated at 0.198, which exceeds the values reported by [79] and [2]. These findings suggest that communities in impoverished regions generally lack adequate financial capital due to low income, restricted access to credit, and limited revenue-generating opportunities. Financial capital is a critical livelihood component, and its indicators, such as household income are influenced by factors like cultivable land, forest resources, education, access to loans, internet connectivity, and political engagement ([91]). Furthermore, age, training, and participation in NGO activities show a significant positive correlation with income levels ([79]). To compare income disparities between mining-affected and non-affected riverine inhabitants, box plots were utilized (Figure 6). The median lines for each group lie outside the interquartile boxes, indicating significant income variation. The box plot for the Nawagang River group was notably longer, suggesting greater income dispersion, while both the Jadukata and Nawagang groups exhibited outliers beyond the whiskers, indicating asymmetric distributions. On average, incomes among mining-affected communities were lower due to the substantial loss of fish and crop yields caused by AMD ([48]). Among all livelihood capitals assessed, financial capital scored the lowest with an index of 0.198 (Table7), whereas physical capital scored the highest at 0.72 (Table4). These results align with previous studies on small-scale and floating fishermen in the northeastern floodplains and river systems of Bangladesh ([79]; [62]), where riverine inhabitants receive modest improvements in transportation and income-generating opportunities.

Figure 6. Box plot for two riverine habitant’s monthly income.

The overall Livelihood Assessment Index (LAI) for transboundary riverine communities in northeastern Bangladesh was calculated at 0.487, based on the average of all five capital indices: human (0.61), physical (0.72), natural (0.36), social (0.55), and financial (0.198). This value aligns with findings from [79] and exceeds the LAI of 0.316 reported by [36]. Despite this moderate score, the overall livelihood status of respondents remains low, consistent with studies from disaster-prone riverine islands. This low livelihood status reflects serious barriers in accessing both formal and informal loans, stemming from limited opportunities for income generation, employment, access to usable forest resources, safe water, healthcare, education, and land use rights. The spider diagram (Figure 7) visually represents the LAI for each form of capital, ranging from 0 (lowest) to 1 (highest), and clearly indicates that the livelihood index of mine-affected communities is lower than that of non-mine-affected groups.

Figure 7. Index values of livelihood capitals of riverine inhabitants.

Table8 presents a comparative analysis of livelihood status based on different livelihood capitals namely human, physical, natural, social, and financial for both mine-affected and non-mine-affected riverine communities. The Livelihood Assessment Index (LAI) values for the Jadukata (mine-affected) riverine area were 0.57 (human), 0.67 (physical), 0.30 (natural), 0.52 (social), and 0.184 (financial), whereas for the Nawagang (non-mine-affected) riverine area, the values were 0.64, 0.76, 0.41, 0.58, and 0.212, respectively. Among these, significant differences were observed in physical, natural, and financial capital, while differences in human and social capital were statistically insignificant. Indicators such as soil erosion control, air quality, transportation infrastructure, healthcare facilities, and income levels were notably lower in the mine-affected Jadukata area. Although mining activities have not led to substantial changes in health service access or interpersonal relationships, they have contributed to an increase in waterborne and respiratory diseases, including tuberculosis, chronic cough, gastric issues, and eye infections ([39]). Environmental degradation from mining has also affected soil stability, waste management, water quality, and agricultural productivity, thereby intensifying food insecurity and disease prevalence.

Financial capital in the Jadukata region is particularly strained. Limited income sources and reduced access to productive assets such as livestock, small enterprises, and farming tools further hinder economic resilience. Similar findings were reported by [59], who noted that agricultural lands in the coalfield regions of Odisha had become infertile, severely impacting farming livelihoods. The overall LAI scores for mine-affected and non-mine-affected transboundary riverine communities in northeastern Bangladesh were 0.449 and 0.522, respectively, with the difference being statistically insignificant (Table8; Figure 8). Nevertheless, the data suggest that the mine-affected Jadukata riverine communities are experiencing greater socio-economic challenges compared to their Nawagang counterparts.

Table 8. Comparison of livelihood assessment index among mine and non-mine affected riverine areas.

Figure 8. Index values comparison of livelihood capitals for mine and non-mine affected riverine inhabitants.

The all-up livelihood opportunities were measured by asking ten questions as shown in Table9 and Table10, where the mean scores of each statement are less than 2.5, 2.5 to 3.5, and greater than 3.5, indicating that the statement’s contributions in all-up situations are positive, neutral, and negative roles, respectively. From the table, it was shown that improved environmental safety and increased modern technology use contribute positive, negative, and neutral rules in Jadukata riverine areas in all-up situations for livelihood opportunities at every cluster; and improved public health contributes the same result in both riverine areas also. Similarly, income generating and saving, increased waterborne diseases, increased fish and crop cultivation, and decreased public wealth are playing positive and neutral roles in the Jadukata and Nawagang riverine area’s inhabitants for different clusters. In addition, improved housing aspects play positive and neutral roles in Jadukata riverine areas, and positive roles in Nawagang riverine areas for all clusters, but the roles of it make family life painful, which is completely alternate. On the other hand, improved environmental safety and increased modern technology use play positive and neutral roles in Nawagang riverine areas across all clusters. There were significant (p < 0.001) different to contribute in living opportunities among the clusters in all riverine areas for every statement except income generating and saving in the Jadukata riverine area. The statement that it makes family life painful and improved housing aspects were also insignificant (p > 0.05) difference to contribute in living opportunities among the clusters in the Jadukata and Nawagang riverine areas (Table10).

Table 9. K-mean clustering or Non-hierarchical clustering for Jadukata and Nawagang riverine habitant’s observation.

Table 10. ANOVA among the different cases of riverine habitants for different clusters.