A- A+
Alt. Display

Agricultural productivity, household poverty and migration in the Indian Sundarban Delta

Abstract

Deltas are highly sensitive to erosion, flooding, and salinization with consequential agricultural productivity losses and out-migration, which is a preferred adaptive measure for the inhabitants of deltaic islands. This study investigates the associations between agricultural productivity decrease, household poverty and the probability of out-migration in the Indian Sundarban Delta (ISD). Using newly collected survey data from randomly selected households within the ISD, we analysed these relationships by means of descriptive statistics and regression modeling. Results suggest the significant positive association between a decrease in agricultural productivity and out-migration. The results further show that ceteris paribus, out-migration is negatively associated with household poverty, which is likely to be explained by the effect of remittances. The results yield important policy implications at the local level and can contribute to the progress towards sustainable livelihoods in these deltaic islands.

Knowledge Domain: Earth & Environmental Science
How to Cite: Hajra, R. and Ghosh, T., 2018. Agricultural productivity, household poverty and migration in the Indian Sundarban Delta. Elem Sci Anth, 6(1), p.3. DOI: http://doi.org/10.1525/elementa.196
Published on 10 Jan 2018
Accepted on 24 Oct 2017            Submitted on 13 Jan 2017
Domain Editor-in-Chief: Oliver Chadwick; University of California, Santa Barbara, US
Guest Editor: Fabrice Renaud; United Nations University, Bonn, DE

According to World Bank figures (World Bank 2013), globally 10.7% of the world’s population lived below the international poverty line of US $1.90 a day (2011 Purchasing Power Parity). Globally an estimated 46.1% of the total rural population and 76% of extreme poor rural people are dependent on agriculture (Castañeda et al. 2016; World Bank and IMF 2016). Some evidence indicates direct and indirect association of rural poverty with agricultural production (Thirtle et al. 2001; World Bank 2008; De Janvry and Sadoulet 2009; Cervantes-Godoy and Dewbre 2010; Christiaensen et al. 2011). The incidence of the Green Revolution in Asia had put confidence in agriculture as a growth engine to reduce poverty in rural areas in particular (Christiaensen et al. 2011). A more dynamic and inclusive agricultural sector could dramatically reduce rural poverty, helping countries meet Goal 1 of the Sustainable Development Goals entitled “No Poverty: End poverty in all its forms everywhere” (UN 2015). Existing evidence suggests that agricultural productivity has been negatively influenced by a number of physical and non-physical factors, including climate change, extreme weather condition, and technological deficits (Reilly et al. 2003; Ewert et al. 2005; Battisti and Naylor 2009; Gornall et al. 2010; Hertel and Rosch 2010). Globally, the share of agriculture in total GDP has fallen from 9% in the early 1970s to 4% in 2006 (FAO 2006). The employment in the agricultural sector in the world has shown declining trend from 2005 (22.9% of total employment) to 2010 (19% of total employment) (World Bank 2010). A declining share of agriculture in national employment and GDP is an inevitable consequence of low production (FAO 2006). Farm income in India declined sharply from 22.4% in 2001 to 14.4% in 2011 (at constant 2004–05 prices), causing a simultaneous decrease of share in employment from 58.2% to 54.6% during this period (GOI Report, 2016). The weather sensitivity of agriculture and the increasing vulnerability of crop yields to both weather extremes and changing weather conditions are likely to further accelerate the rural to rural and rural to urban migration (Kumar and Viswanathan 2015). Migration can offer a pathway out of poverty for those who leave and also for those who stay behind through remittances, which reduced rural poverty (World Bank 2008). Remittances sent by migrated household members often increases the land, livestock, and human capital base of rural household members who stayed behind. Along with potential to improve well-being, migration and remittances have their effects on inequality, particularly in rural areas (Hass 2007). In India, 55% of male migrants send remittances in support of households (Bhagat 2014). There is a complex association between migration and agriculture as productivity can be negatively affected by migration of labour while remittances may have the opposite effect by helping households to overcome constraints on production (Miluka et al. 2010; de Brauw 2010; Damon 2010). Although earlier studies investigated the linkages between climate change, crop yield, migration, and the rural-urban wage differentials in other geographical locations with different geographical contexts (Feng et al. 2010 and 2012, Marchiori et al. 2012, and Nawrotzki et al. 2012), there is limited research investigating the connections between poverty, productivity and migration in Indian Sundarban Delta (ISD). ISD has limited access to resources, livelihoods, services, and infrastructure. In spite of being an agriculturally dependent economy, yields have been negatively affected by land loss and saline water inundation (Hazra et al. 2002). Around 34% of the 4.6 million people residing on different islands of the ISD are under extreme poverty and 75% of families there has at least one member working in other states of India (Hazra et al. 2014). Many of these issues have recently been investigated but there is limited research on linkages between productivity, poverty, and migration in the ISD. Our study aims to fill the gap of previous research by examining these associations. In this context, the main objective of this paper was to examine the significance and strength of the associations between key variables such as agricultural productivity decrease (PC), poverty (P), and the probability of out- migration (M) in the ISD. We have estimated the significance of association separately as the chosen key variables influence on each other. Findings of separate associations could be useful for specific policy recommendations. We conducted our analysis using logistic regression techniques with primary data from household surveys within the ISD community collected in 2012–2013. 2. Agricultural productivity and migration issues in the study area The western boundary of the ISD is the major focus of this study including Sagar, Ghoramara and Mousani Islands which are characterised by high population densities and reported vulnerability (Hazra et al. 2002; Hazra et al. 2014; Hajra et al. 2017) (Figure 1). The study islands are affected by coastal erosion and cyclonic surge causing saline water inundation, along with over-exploitation of resources. The consequential loss in agricultural crop productivity imposes severe challenges to the farming communities of these islands, which is why these islands were selected for this study. Administratively, Sagar Island is classified as a “CD-Block”, a collection of mouzas or villages which is the largest island in the ISD with a total population of 206,844 (Census Report of India, 2011). It is bounded by the Hooghly River to the north and west, the Muriganga River to the east, and the Bay of Bengal to the south. Overall, there are 42 villages on Sagar Island. Ghoramara Island, located to the north of Sagar Island covers an area of 4.8 km² (Jana et al. 2012) with a total population 5,193. The major villages on this island include Khasimara, Baishnabpara, Hathkola, Baghpara, Raipara, Mandirtala, Chunpuri, Lakshmi Narayanpur and Khasimara Char. Out of these, Khasimara Char, Lakshmi Narayanpur, Khasimara and Baishnabpara have already disappeared (Jana et al. 2012; Ghosh et al. 2003). Mousani Island covers 24 km2, and, according to 2011 census figures, is home to 3,340 families and 22,073 people. This island is surrounded by the Muriganga/Bartala River to the west and northwest, Pitt’s Creek/Chenayer River to the east, and the Bay of Bengal to the south (WWF 2010). Mousani Island is a single Gram Panchayat (GP) unit under Namkhana CD Block. Prevailing socio-economic conditions are not uniform across these islands. The population is principally engaged in farming, fishing, tourism and transport services, business and industry (Mondal, 2012). This study area is predominantly a rural region where 69% of the total workers (Census Report of India, 2011) is dependent on mono crop (Aman paddy) cultivation similar to other inhabited parts of the ISD (Hazra et al. 2002). Sagar Island enjoys better infrastructural facilities than the two other islands including improved road network, rural hospital, permanent jetty, etc. (HDR 2009). Figure 1 The study area-Sagar Island (Sample population 3038), Ghorama Island (Sample population 225), Mousani Island (Sample population 737) (Hajra et al. 2016). DOI: https://doi.org/10.1525/elementa.196.f1 The deltaic islands of the Indian Sundarban are highly vulnerable to frequent embankment failures, submergence and flooding, beach erosion, cyclone and storm surges. In addition, alarming growth of population in this environmentally sensitive and fragile ecological niche has posed a major threat to its very existence (Ghosh 2012). Paddy is the prime crop and consumable of the study islands and is considered as an indicator of the regional economy (Hazra et al. 2002; Hajra et al. 2016a). Several varieties of rice are cultivated including ‘C-R’, ‘Chinese’, ‘Talmugur’, ‘Dudheswar’, ‘Annada’, ‘Nona-Bokhra’ varieties which are salt resistant. Estimated agricultural land in Sagar, Ghoramara and Mousani Island is 85%, 1.2% and 17% of the total land area, respectively. The percentage of agricultural land is lowest in Ghoramara as this island has faced maximum land loss which includes a huge amount of agricultural land over time (Hajra et al. 2016b) and experienced large consequential out-migration (Ghosh et al. 2014). Cultivators from this sinking island are forced to migrate and search for other livelihood options, due to which the agricultural practice became less and those lands remain as fallow land. This non-practice and subsequent saline water inundation due to embankment failure further make these lands nonfertile. Reported agricultural land loss (Kumar et al. 2007; Hajra et al. 2016b) as a consequence of erosion generates severe challenges to the farmer communities of these islands. The rate of erosion estimated during 1990 to 2015 was 0.2 km2, 0.02 km2, and 0.08 km2 in Sagar, Ghoramara and Mousani Island, respectively (Hajra et al. 2016b). It is evident from time series shoreline change (Hajra et al. 2016b) (Figure 2) that considerable changes have occurred in shape and size in all three islands due to either erosion or accretion. Villagers are usually blaming salinization, erosion, and inundation to explain productivity losses (Hajra et al. 2016). There are strong associations between productivity loss with erosion and saline water inundation with consequential poverty, as poorest households experience that erosion and saline water inundation are responsible for maximum material losses including agricultural land and crop (Hajra et al. 2017). Figure 2 Morphological change analysis to identify the extent of erosion in Study Islands (1990 to 2015) (Hajra et al. 2016). DOI: https://doi.org/10.1525/elementa.196.f2 The estimated paddy production values of District Statistical Handbook, Government of West Bengal, from 2000–01 to 2011–12, showed a decrease in production rate (Figure 3). In line with the survey findings reducing productivity of the agricultural land has been influenced by soil quality degradation from saline water intrusion (Chand et al. 2012; Mandal and Mandal 2012; Hazra et al. 2002). In addition with soil quality degradation, external factors like excesive use of fertiliser, and irrigation deficit also hamper the productivity (Edmeades 2003; Ali et al. 2007; Das et al. 2013). High concentration of fertiliser has been used in the region mainly for betel leaf cultivation in order to boost production, compromising the sustainability of the system (Hajra et al. 2016a). Scarcity of irrigation infrastructure and dependence on monsoon rain are the main reasons behind less intensive mono-crop practices of the islands of the ISD (Hazra et al., 2014). Figure 3 shows a steady fall in production after 2009, mainly caused by Cyclone Aila. As a consequence of this cyclone, rice production was reduced to 32–40 quintal per 1.6 hectares from 64–80 quintal per 1.6 hectares produced before the event (Debnath 2013). Due to the absence of permanent crop market in these islands, farmers often sell their products to middlemen at lower prices that they would get from markets. Insufficient storage is another problem for Sagar, Ghorama and Mousani to be self-sufficient in food supply throughout the year (Hazra et al. 2014). Limited earning prompts the farmers to look for other livelihood options, particularly during non-monsoon months. This has compelled local cultivators to sell their lands and switch to other occupations such as daily labourers (Hajra et al. 2016a; Hazra et al. 2014). Low higher education achievement levels restrict the farmers in terms of involvement in any formal service sector rather than informal sector such as daily labour (Hajra et al. 2016b). Census figures show that the numbers of cultivators are decreasing in study area (Figure 4). Figure 3 Trend in paddy production in study area (Data Source: District Statistical Handbook, 1999–2012). DOI: https://doi.org/10.1525/elementa.196.f3 Figure 4 Trend in number of cultivators (Census of India, 1991, 2001, 2011). DOI: https://doi.org/10.1525/elementa.196.f4 A steadily denuding resource base is making it difficult for the people in this region to sustain themselves through the year whereas better income opportunities in other places like Gujarat and Chennai are being offered. Male migrants are dominant (39% in 2007–08 in India) in rural to urban labour migration flow (Bhagat 2014), which is mainly influenced by the difference between expected and actual wage rate (Harris and Todaro, 1970). The high degree of occupational dependence on agriculture and its rapidly declining income share is an indication of a higher incidence of poverty in these islands. Literature suggests that migration can have a positive effect on livelihoods through remittances (de Haas 2007; Erdal 2012; Viet 2008). Remittances may also contribute increasing inequality in income between households with and without migrant labourers (Barham and Boucher 1998; Acosta et al. 2008). 3. Data and method Data from primary surveys, conducted for the purpose of this study through direct interviews with households within the study area of the ISD, have been used. In order to assess actual socio-economic conditions of the area, a thorough household-level socio-economic survey was conducted in the sampled villages of the study area during 2012 to 2013. A two-stage cluster random sampling was used for this study. In the first stage, mouzas were chosen randomly (using random number table) from all three islands; in the second stage, a first house was selected using random number table (number assigned to households) from each mouza and then the remaining samples were selected using a systematic sampling scheme. Sagar Island has 42 villages under 9 Gram Panchayat. Ghoramara is a single mouza under Sagar Block. Mousani has 4 villages namely Bagdanga, Kusumtala, Baliara, and Mousani. The survey was carried out through direct interviews in 52% of the inhabited mouzas of Sagar Block, including Ghoramara (23 mouzas out of 42) and 100% of the inhabited mouzas (4 mouzas) of Mousani Gram Panchayat of Namkhana Block. The surveyed villages are distributed throughout the Sagar Island (other two islands were fully been surveyed) covering most of the coastal fringe and important central areas (Figure 5). One to one direct interviews were conducted between May 2012 and October 2013 with members of 783 households from 27 villages of the study area, consisting of a total number of surveyed populations of 4,500. In this way, almost 59% of mouzas were included and the margin of random error (Fox et al. 2007) was less than 3% at 95% confidence interval which is sufficient to reach the required precision of the survey, even though the number of households per mouza remained limited. Random number tables were used to select households which were either dispersed or clustered together in small groups (hamlets) or along the embankment (linear) or larger villages including Rudranagar, Gangasagar of Sagar Island. The questionnaire used was pre-tested through a small sample sized survey. The final questionnaire included questions related to household and individual-level characteristics, such as age and sex of the members of the family, family size, occupation of the earning members, education of the members, land use information, and some other livelihood details. Questions were also asked to get the perception of local farmers on agricultural productivity losses, poverty, and out-migration. Questions also covered the issues whether agricultural productivity is decreasing, what are the reasons behind the low production, if any member of the family migrated to other places and why. Respondents were also asked if they felt the impact of climate change. They mostly responded in the affirmative to this question but only mentioning temperature increases and shifts in rainfall as a consequence of climate change. Figure 5 Map of surveyed mouzas. DOI: https://doi.org/10.1525/elementa.196.f5 The main aim of this study has been to assess the linkages between key variables of interest such as household level poverty (P), agricultural productivity change (PC), and out-migration (M) through statistical models. Poverty status was defined in two different steps in this study. First, a household was classified as poor if its income fell in the lowest quantile based on the quantile distribution of normalised income. Secondly, a household was considered to be poor if it was categorised as such by the Indian Government measurement: for example, for a family of five, the all-India poverty line in terms of consumption expenditure would amount to about US$61.27 per month in rural areas and US \$75.09 per month in urban areas (GOI 2013). Agricultural productivity change, another key variable, takes the values 1 or 0 depending on whether productivity decreases or not according to respondents. In our sample, 41% of all households reported productivity losses. Out-migration is also a binary variable measuring whether or not a household has at least one internal migrant (native to the area but not necessarily the present village) or out migrant (other places of the country). Migration in this study has been considered as an internal or outside movement of one or more household members to get scope of alternative livelihoods for a minimum 3 months to a maximum of 1 year. Erosion (E), inundation (I), and Cyclones (C) were considered as key explanatory variables measuring whether or not a household suffered agricultural land and productivity loss from these hazards. Values of 1 or 0 were assigned if the response was yes or no, respectively. According to the respondents, inundation is considered as saline water intrusion due to extreme high tide overtopping or embankment failure. Other controlling variables considered in the study include standard socio-economic characteristics, i.e. household size (SZ), average age of the earning member of the family (A), literacy (L) of the earning members and occupation (O) of worker, amount of owned land area (LS) and geographical location of islands (IS). Literacy takes the value 1 or 0 depending on whether or not at least one of the earning member is literate. Also, a value of 1 was assigned if the occupation of any earning member is cultivation and 0 for other professions. We used 4 different models to analyse the result. Each model has a dependent variable. In model 1, agricultural productivity change was taken as dependent and inundation, erosion, and cyclone were considered independent variables, controlling for other variables included in the model. Likewise, in other models dependent and independent variables were used to identify the associations as stated in the aims of this study. Control variables remain the same for all the models. The dataset being binary in nature we performed logistic regression to test the effects. We used the Generalized Linear Model (GLM) code in R-Software for our analysis.

In general, If y is a binary response (1 and 0), x1, x2, …, xk are the explanatory variables and π is the probability that y takes the value 1 = P(y = 1), then the log model is:

$\pi \text{\hspace{0.17em}\hspace{0.17em}}=\text{\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}}\frac{\mathrm{exp}\left({B}_{0}\text{\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{1}{x}_{1}+\text{\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}}\dots .\text{\hspace{0.17em}\hspace{0.17em}}+{B}_{k}{x}_{k}\right)}{1\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}\mathrm{exp}\left({B}_{0}\text{\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{1}{x}_{1}+\text{\hspace{0.17em}\hspace{0.17em}}\dots .\text{\hspace{0.17em}\hspace{0.17em}}+{B}_{k}{x}_{k}\right)}$

and hence,

$\mathrm{ln}\left(\frac{\pi }{1-\pi }\right)\text{\hspace{0.17em}\hspace{0.17em}}=\text{\hspace{0.17em}\hspace{0.17em}}{B}_{0}\text{\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{1}{x}_{1}+\text{\hspace{0.17em}\hspace{0.17em}}\dots .\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{k}{x}_{k}$

Here B0, B1, …, Bk are the coefficients of the regressors. The Maximum Likelihood Method is used to obtain these estimates.

For our analysis we have:

Model-1

Regression equation of productivity change on inundation and erosion:

$\begin{array}{l}\mathrm{ln}\left(\frac{\pi }{1-\pi }\right)\text{\hspace{0.17em}\hspace{0.17em}}=\text{\hspace{0.17em}\hspace{0.17em}}{B}_{0}\text{\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{1}I\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{2}E\text{\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{3}C+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{4}IS+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{5}A\text{\hspace{0.17em}\hspace{0.17em}}\\ \text{ \hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{6}SZ\text{\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{7}O\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{8}L\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{9}LS,\end{array}$

where π = P (agricultural productivity change (PC) = 1)

Model-2

Regression equation of poverty on productivity change:

$\begin{array}{l}\mathrm{ln}\left(\frac{\pi }{1-\pi }\right)\text{\hspace{0.17em}\hspace{0.17em}}=\text{\hspace{0.17em}\hspace{0.17em}}{B}_{0}\text{\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{1}PC\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{2}IS\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{3}A\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{4}SZ\text{\hspace{0.17em}\hspace{0.17em}}\\ \text{\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{5}O\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{6}L\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{7}LS\end{array}$

where π = P (poverty (P) = 1)

Model-3

Regression equation of out-migration on productivity change:

$\begin{array}{l}\mathrm{ln}\left(\frac{\pi }{1-\pi }\right)\text{\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}}=\text{\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}}{B}_{0}\text{\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{1}PC+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{2}IS+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{3}A\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{4}SZ\text{\hspace{0.17em}\hspace{0.17em}}\\ \text{\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{5}O\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{6}L\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{7}LS\end{array}$

where π = P (out-migration(M) = 1)

Model-4

Regression equation of poverty on migration:

$\begin{array}{l}\mathrm{ln}\left(\frac{\pi }{1-\pi }\right)={B}_{0}\text{\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{1}M\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{2}IS\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{3}A\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{4}SZ\\ \text{\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{5}O\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{6}L\text{\hspace{0.17em}\hspace{0.17em}}+\text{\hspace{0.17em}\hspace{0.17em}}{B}_{7}LS\end{array}$

where π = P (poverty(P) = 1)

P-values were calculated to test the influence of the factors. If the p-value for the test corresponding to the coefficients of the explanatory variables are less than certain standard fixed value, say α (0, 0.001, 0.01, 0.05, 0.1), then the explanatory variables have a significant effect at the significant level α. Generally, up to α = 0.05, we consider higher significant effect. However if α = 0.1 then we can also say the corresponding factor has a small significant effect. The Akaike Information Criterion (AIC) (Akaike 1973) is calculated for each model.

4. Results

Logistic regressions are considered to identify the associations among key variables such as poverty (P), agricultural productivity change (PC), and out-migration (M). Four separate tables (Tables 1, 2, 3, 4) show the results of four separate models that tested our hypotheses.

Table 1

Results of logistic regression of Model-1 to predict the dependence of productivity change on Inundation, Erosion, Cyclone and the control variablesa. DOI: https://doi.org/10.1525/elementa.196.t1

Coefficient Estimates Standard error z-value Number of observations Log likelihood Akike Information Criterion (AIC)

Intercept 0.77 0.52 1.48
Inundation 3.76 0.61 6.20*** 783 –436 891
Erosion 1.07 0.18 5.93***
Cyclone 0.42 0.17 2.46*
Island –0.49 0.13 –3.90***
Age 0.002 0.008 0.25
Size –0.01 0.052 –0.22
Occupation –0.48 0.21 –2.24*
Literacy –0.49 0.24 –2.02*
Land Size –0.041 0.05 –0.80

a Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1.

Table 2

Results of logistic regression of Model-2 to predict the dependence of poverty on productivity change and control variablesa. DOI: https://doi.org/10.1525/elementa.196.t2

Coefficient Estimates Standard error z-value Number of observations Log likelihood Akike Information Criterion (AIC)

Intercept 0.21 0.47 0.45 783 –519 1054
Productivity Change 0.39 0.16 2.50*
Island 0.13 0.10 1.26
Age –0.003 0.007 –0.42
Size –0.038 0.05 –0.79
Occupation –0.25 0.20 –1.25
Literacy –0.59 0.20 –2.88**
Land Size –0.01 0.05 –0.19

a Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1.

Table 3

Results of logistic regression of Model-3 to predict the dependence of migration on productivity change and the control variablesa. DOI: https://doi.org/10.1525/elementa.196.t3

Coefficient Estimates Standard error z-value Number of observations Log likelihood Akike Information Criterion (AIC)

Intercept –0.88 0.4856 –1.81. 783 –493 1001
Productivity Change 0.90 0.1649 5.44***
Island –0.20 0.1109 –1.82.
Age –0.001 0.0072 –0.20
Size of family 0.07 0.0050 1.45
Occupation –0.43 0.2119 –2.01*
Literacy –0.15 0.00001 –0.07
Land Size –0.10 0.2076 –1.86.

a Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1.

Table 4

Results of logistic regression of Model-4 to predict the dependence of poverty on migration and the control variablesa. DOI: https://doi.org/10.1525/elementa.196.t4

Coefficient Estimates Standard error z-value Number of observations Log likelihood Akike Information Criterion (AIC)

Intercept 0.34 0.47 0.75 783 –519 1054
Migration –0.41 0.15 –2.65**
Island 0.13 0.10 1.32
Age –0.003 0.007 –0.39
Size of a family –0.05 0.05 –0.95
Occupation –0.24 0.20 –1.20
Literacy –0.63 0.21 –3.09**
Land Size –0.02 0.05 –0.42

a Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1.

In order to examine the relationship between the productivity changes with inundation, erosion, cyclone, controlling for other variables included in the model, regression analysis was performed (Model 1 – Table 1). The values of the coefficients corresponding to the key variables including inundation, erosion, and cyclone are respectively 3.76, 1.07 and 0.42 with respective standard errors of 0.61, 0.18 and 0.17. These coefficients indicate that productivity change (decrease) is directly proportional to inundation and erosion (Table 1). p-values for these two factors are close to zero, highlighting the significant effect of these two key factors on productivity change. It is also noteworthy that the islands, occupation and literacy of worker have a significant effect. The coefficients of island, occupation, and literacy are –0.49 (standard error of 0.13), –0.48 (0.21) and –0.49 (0.24), respectively. The corresponding p-values are 0.0001, 0.039 and 0.031. We calculated the average of respondents with productivity change (decrease) reported and the proportions are 0.36, 0.41, and 0.45 in Sagar, Ghoramara, and Mousani, respectively. The condition is worse in Mousani and Ghoramara. Table 1 also illustrates the impact of literacy of workers. Educated farmers are more likely to face lesser loss from productivity change (Girgin 2011). The association between productivity change (decrease) and the people with the agricultural profession is significant (p < 0.05). The values of log likelihood and Akaike Information Criterion (AIC) of the Model-1 are respectively –436 and 891.

In Table 2 we consider Model 2 to establish a logistic regression of poverty on productivity change, controlling for other variables. i.e, island, the average age of the earning members of the family, size of the family, occupation, literacy and land size. Since the dependency between the productivity change and the control variables is very negligible, these interactions were ignored in model 2 to focus mainly on the poverty and productivity change dependence. Table 2 shows that the value of the coefficient of productivity change (decrease) is 0.39 with standard deviation 0.16 and the corresponding p-value is 0.0123, indicating that poverty increases as productivity decreases. Here also literacy has a significant effect on poverty. Households with literate workers are significantly more likely to be less poor than other households (p < 0.001). Household poverty is negatively associated with educational attainment of workers which is likely to be related to the fact that elementary and post-elementary education is necessary for poverty reduction particularly in rural regions (Tilak 2007; Wedgwood 2007). The values of log likelihood and AIC of Model-2 are respectively –519 and 1054.

Table 3 illustrates regression analysis considering out-migration, productivity change, and the control variables. The value of the coefficient and corresponding standard error of productivity change are 0.9 and 0.16, respectively. The corresponding p-value is 0. The value of the coefficient and corresponding p-value indicate a strong positive association between migration and productivity change. The positive association indicates that decrease in agricultural production accelerates the probability of migration (Kumar and Viswanathan 2015). It is also noteworthy that owned land area, occupation and geographical location of islands have a significant effect on migration. The households having some amount of agricultural land are likely to migrate least than those who have no land (p < 0.1). Model 3 also indicates that the migration trend is not equal among the islands. The coefficient of island is –0.20 with p-value 0.06. Workers from other occupation rather than farming is significantly more likely to migrate (p < 0.05). The values of log likelihood and AIC of the Model-3 are respectively –493 and 1001.

In Table 4 an attempt has been made to find out the associations between poverty and migration, controlling for other variables such as household size (SZ), average age of the earning members of the family (A), literacy (L) of the earning members and occupation (O) of workers, amount of owned land area (LS) and geographical location of islands (IS). Households having at least one member who migrated are significantly more likely to be less poor than the other households (p < 0.001). Migration has often been considered an alternative option where traditional practices including agriculture, forestry, fishing fail (Kumar and Viswanathan 2015; Iqbal and Roy 2014). The household poverty level is most likely reduced through remittances sending by migrants (World Bank 2014; Ghosh 2012). Controlling the other factors literacy has significant impacts on poverty. The results are consistent with the findings reported in Model 2. The values of log likelihood and AIC of the Model-4 are respectively –519 and 1054.

5. Discussion and policy implications

These estuarine islands of the Indian Sundarban Delta are under increasing stress due to a number of physical and non-physical changes. This fragile tidally-influenced ecosystem provides limited scope for livelihood expansion (Raha et al. 2012). Agriculture being the backbone of the ISD, any deterioration of productivity change further accentuates poverty (Hazra et al. 2002). In this context, the main objective of this paper was to examine the significance and strength of the associations between agricultural productivity decreases, household poverty and the probability of out-migration. The analysis has been done based on logistic regression analysis which determines the association between productivity change (decrease), poverty, migration and the control variables. The results of the study confirmed that erosion, inundation, and cyclones have the significant association with agricultural productivity decreases. A time series map shows a considerable change in shoreline due to erosion and accretion. Similar to the findings of previous studies (Mondal 2012; Kumar et al. 2007), this study also shows that agricultural land loss as a consequence of erosion generates threats to the farmer communities of these islands. Experiencing saline water inundation was shown to have the greatest impact on the productivity loss. The study of Hazra et al. (2002) also emphasized that inadequate drainage was a major problem explaining agricultural productivity loss. The problem is exacerbated during the monsoons when the yield of paddy crop is adversely affected by inadequate drainage and water logging. In line with other literature (Debnath 2013; Mondal 2012), this study also found significant impact of cyclonic storms on agricultural productivity. Controlling for socio-economic characteristics, the odds of suffering from productivity loss is higher on Mousani Island when compared with the other two islands. This can be explained by the differential developmental levels of these two islands, in particular, the considerably better infrastructure in Sagar Island (HDR 2009; Andersen 2010; Danda 2007).

The results of unadjusted model 2 show that the likelihood of the increase in poverty level following productivity decrease remains significant (p = 0.0123). In line with other literature (Hazra et al. 2014; Mandal and Mandal 2012; Ghosh 2012) low productivity accelerates the poverty among farmer communities. In addition, institutional shortcomings and limited governance such as the presence of middleman, absence of proper crop markets and cooperatives do not allow farmers to get better price of their produce (Mandal and Mandal 2012). Similar to other studies (Tilak 2007; Girgin 2011), we found negative associations between household poverty and educational attainment as literate farmers are likely to suffer least from the productivity change as they might have the know-how of the weather prediction and also modern techniques of farming. The trend in productivity change (Figure 3) and a decrease in cultivator number (Figure 4) indicates a change trend from cultivators to other livelihoods. Less profit from agriculture often compels farmers to sell their land and migrate to other places for livelihoods (Hazra et al. 2014). Low agricultural productivity does not allow to sustain farming families, triggering migration in the hope to generate alternative livelihoods (Hazra et al., 2014; Mistri, 2013). In line with this, model 3 also highlights the fact that decrease in agricultural production is significantly more likely to accelerate the probability of migration (p < 0.0). Model 3 also indicates that having some amount of land property at native areas may influence the decision of out-migration which is similar to findings of Feng and Heerink (2008) and VanWey et al. (2012). Also, the migration pattern is not the same in all the islands. Sagar Island being the most developed (HDR 2009) experiences less migration due to productivity decreases. Similar to the findings of Vatta and Sidhu (2010) this study also confirms that non-farm workers are likely to migrate more than farmers (p < 0.05), which may be attributed to the higher wage difference of non-farm sector among rural and urban areas.

The study found that an average of 30% of people among surveyed population migrates (Figure 6) which is similar to other islands of the ISD (Ghosh 2012). Families which are having members working in other states get benefits from remittances. The other local residents are easily lured by the changed lifestyles of the migrants and follow them as they are being driven by poverty (Hazra et al. 2014). Model 4 indicates that households having at least one member who has migrated outside, are significantly more likely to be less poor than the other households (p < 0.001). A large part of the ISD population depends on remittances (Ghosh 2012). This confirms results from other studies (World Bank 2014; Hajra and Ghosh 2016; Ghosh 2012). A significant level of inequality in income distribution is prevailing in the study area which may be attributed to remittance (Barham and Boucher 1998; Acosta et al. 2008). Achieving a sustainable economy in low-elevation deltas is critical due to the relatively lower adaptive capacity of households which is coupled with frequent extreme natural events. Based on the conceptual framework of this study we found causal factors behind productivity change, poverty, and migration. In this context, strengthening links between these key variables should be the main focus for policy strategies. Few policy recommendations are provided for improving system sustainability, accounting for the complex socio-ecological system of the Sundarban.

First, agricultural training with special emphasis on climate resilient agricultural techniques needs to take place to allow households to cope with the saline water inundation impact. Appropriate, ecosystem-specific, cost-effective technologies including restoration of traditional seeds for farming can be introduced to the farmers. Crop diversification could be a measure to reduce crop failure and productivity losses. Horticulture practices could enhance the household’s income. Crop insurance with full subsidies also might be an effective measure to protect society from the financial vulnerability in case of natural hazards. Establishment of e-chaupal which is an ITC India initiatives to link directly with rural farmers via the internet for procurement of agricultural products may be an effective option for better economic returns, as the farmers can avoid middlemen. Establishing Block level market and flexible marketing strategies may also enhance profitability and ensure more involvement in the agricultural sector. The negative impact of erosion and sea water inundation could be limited through the construction of stable embankments along the sea and tidal rivers, and regular repair and maintenance of existing embankments.

Second, reduction of poverty may help lower the number of out-migration in search of livelihoods. Based on the UK Department for International Development (DFID)- Sustainable Livelihood Approach, a holistic, dynamic, people-oriented and sustainable framework focusing on poverty eradication actions through assets (human, social, natural, physical and financial capital) is necessary to provide special attention to the remote pockets with least developments (Speranza et al. 2014; Brocklesby and Fisher 2003; DFID 2000). New Self Help Groups can be set up or existing ones need to be strengthened through different training programmes preferably at mouza level to reduce poverty level through income generation. Economic poverty could be minimized by providing economic stability to all workers with the ensured occupation. Government employment schemes such as Mahatma Gandhi National Rural Employment Guarantee Act (MGNREGA), Swarnajayanti Gram Swarojgar Yojana (SGSY) might also prove to be effective measures. The inclusion of more BPL families to Antyodaya Anna Yojana (scheme for poorest of the poor) could be an effective measure at mouza level.

Third, migration has been considered as an adaptation factor to overcome poverty, and/or limited access to resources and livelihood opportunities. But unplanned migration are producing a huge number of unskilled labour. In line with other literature (Kabra 2003), this study also found that migrants are mostly becoming engaged in informal sectors as unskilled workers such as daily labour and construction labour. The absence of fixed wage structure and regulations at informal sector leads to exploitation and insecurity among workers employed here. This kind of insecurity may impact their expected economic condition. Extensive out-migration is creating scarcity of agricultural labour in these islands, and the situation is acute in Ghoramara. The traditional farm based economy is gradually being converted to labour based economy, due to all possible physical and non-physical changes. The necessity of skill development amongst the young generation for decent job or entrepreneurship is strongly felt. The potential of ecotourism and religious tourism can also be explored. Effective weather prediction and forecasting, along with the use of latest scientific and technological intervention may strengthen the traditional practices such as agriculture, forestry, fishing as the profit making livelihood to young generation through strategic management planning with appropriate stakeholder involvement.

Data Accessibility Statement

Data from primary surveys, conducted for the purpose of this study through direct interviews with households within the study area of the ISD, have been used.

Acknowledgements

We gratefully acknowledge Mr. Sanjib Kumar Gupta for his kind suggestion and help in data analysis. We would like to thank reviewers and editor for their useful comments and feedback, which has helped to improve the paper.

Competing interests

The authors have no competing interests to declare.

Author contributions

• Contributed to conception and design: RH, TG
• Contributed to acquisition of data: RH, TG
• Contributed to analysis and interpretation of data: RH, TG
• Drafted and/or revised the article: RH, TG
• Approved the submitted version for publication: RH, TG

References

1. Acosta, P, Calderon, C, Fajnzylber, P and Lopez, H 2008 What is the impact of international remittances on poverty and inequality in Latin America? World Development 31 , 36(1): 89–114.

2. Akaike, H 1973 Information theory as an extension of the maximum likelihood principle. In: Petrov, BN, and Csaki, F (Eds.), Second International Symposium on Information Theory . Akademiai Kiado, Budapest, 267–281.

3. Ali, MH, Hoque, MR, Hassan, AA and Khair, A 2007 Effects of deficit irrigation on yield, water productivity, and economic returns of wheat. Agricultural water management 16 , 92(3): 151–61.

4. Andersen, KS 2010 Enlightenment and empowerment through power and light? (Doctoral dissertation, University of Oslo). URL: http://hdl.handle.net/10852/32609.

5. Barham, B and Boucher, S 1998 Migration, remittances, and inequality: estimating the net effects of migration on income distribution. Journal of development economics 30 , 55(2): 307–31.

6. Battisti, DS and Naylor, RL 2009 Historical warnings of future food insecurity with unprecedented seasonal heat. Science 323: 240–244. DOI: 10.1126/science.1164363

7. Bhagat, RB 2014 ‘Urban migration trends, challenges and opportunities in India’, World Migration Report, 2015. International Organization for Migration (IOM).

8. Brocklesby, MA and Fisher, E 2003 Community development in sustainable livelihoods approaches–an introduction. Community development journal 38(3): 185–198. DOI: 10.1093/cdj/38.3.185

9. Castañeda, A, Doan, D, Newhouse, D, Nguyen, MC, Uematsu, H, Azevedo, JP and with Data for Goals Group 2016 Who Are the Poor in the Developing World? Policy Research Working Paper 7844, World Bank Group.

10. Census Report of India 2011 Primary Census Abstract. Goverenment of India.

11. Cervantes-Godoy, D and Dewbre, J 2010 “Economic Importance of Agriculture for Poverty Reduction”. OECD Food, Agriculture and Fisheries Working Papers, No. 23, OECD Publishing.

12. Chand, BK, Trivedi, RK, Biswas, A, Dubey, SK and Beg, MM 2012 Study on impact of saline water inundation on freshwater aquaculture in Sundarban using risk analysis tools. Explor. Anim. Med. Res . 2: 170–178.

13. Christiaensen, L, Demery, L and Kuhl, J 2011 The (evolving) role of agriculture in poverty reduction—An empirical perspective. Journal of development economics 96(2): 239–254. DOI: 10.1016/j.jdeveco.2010.10.006

14. Damon, A 2010 Agricultural Land Use and Asset Accumulation in Migrant Households: The Case of El Salvador. Journal of Development Studies 46(1): 162–189. DOI: 10.1080/00220380903197994

15. Danda, AA 2007 Surviving in the Sundarbans: Threats and responses. Unpublished PhD, University of Twente, 203. URL: http://doc.utwente.nl.

16. Das, S, Roy Choudhury, M, Das, S and Nagarajan, M 2013 Integrated Geospatial technologies for soil salinity assessment over South 24 PGS, West Bengal”. Int. J. Geo Sci. & Tech . 1(2): 41–85. ISSN: 2321-2144.

17. Debnath, A 2013 Condition of agricultural productivity of Gosaba C.D. Block, South 24 Parganas, West Bengal, India after severe cyclone Aila. International Journal of Scientific and Research Publications 3(7), ISSN 2250-3153.

18. de Brauw, A 2010 Seasonal Migration and Agricultural Production in Vietnam. Journal of Development Studies 46(1): 114–139. DOI: 10.1080/00220380903197986

19. De Janvry, A and Sadoulet, E 2009 Agricultural growth and poverty reduction: Additional evidence. The World Bank Research Observer, lkp015.

20. DFID 2000 Sustainable Livelihoods Guidance Sheets. Department for International Development.

21. Edmeades, DC 2003 The long-term effects of manures and fertilisers on soil productivity and quality: a review. Nutrient cycling in Agroecosystems 66(2): 165–180. DOI: 10.1023/A:1023999816690

22. Ewert, F, Rounsevell, MDA, Reginster, I, Metzger, MJ and Leemans, R 2005 Future scenarios of European agricultural land use I. Estimating changes in crop productivity. Agric. Ecosyst. Environ . 107: 101–116. DOI: 10.1016/j.agee.2004.12.003

23. FAO 2006 World agriculture: towards 2030/2050 . Rome.

24. Feng, S and Heerink, N 2008 Are farm households’ land renting and migration decisions inter-related in rural China? NJAS-Wageningen Journal of Life Sciences 31 , 55(4): 345–62.

25. Feng, S, Krueger, AB and Oppenheimer, M 2010 ‘Linkages among climate change, crop yields and Mexico–US cross-border migration’. Proceedings of the National Academy of Science 107(32): 14257–14262. DOI: 10.1073/pnas.1002632107

26. Feng, S, Oppenheimer, M and Schlenker, W 2012 ‘Climate change, crop yields, and internal migration in the United States’. NBER Working Paper No. 17734, NBER, Cambridge.

27. Fox, N, Hunn, A and Mathers, N 2007 ‘Sampling and sample size calculation’. The NIHR RDS for the East Midlands/Yorkshire & the Humber.

28. Ghosh, A 2012 Living with changing climate: Impact, vulnerability, and adaptation challenges in Indian Sundarbans. Centre for Science and Environment. Delhi. India.

29. Ghosh, T, Bhandari, G and Hazra, S 2003 Application of a ‘bio-engineering’ technique to protect Ghoramara Island (Bay of Bengal) from severe erosion. J Coast Conserv 9: 171–178. DOI: 10.1652/1400-0350(2003)009[0171:AOABTT]2.0.CO;2

30. Girgin, A 2011 The Role of Education in Agricultural Productivity: The Case of Village Institutes in Turkey, 1940–1966. Unpublished master thesis, Lunda University.

31. GOI 2013 Press Note on Poverty Estimates, 2011–12. Govt of India, Planning Commission, Month: July 2013.

32. GOI Report 2016 State of Indian Agriculture 2015–16, Ministry of Agriculture & Farmers Welfare Department of Agriculture, Cooperation & Farmers Welfare, Directorate of Economics and Statistics, New Delhi.

33. Gornall, J, Betts, R, Burke, E, Clark, R, Camp, J, Willett, K and Wiltshire, A 2010 Implications of climate change for agricultural productivity in the early twenty-first century. Philosophical Transactions of the Royal Society of London B: Biological Sciences 365(1554): 2973–2989. DOI: 10.1098/rstb.2010.0158

34. Haas, HD 2007 Remittances, migration and social development. Social Policy and Development Programme Paper no 34.

35. Hajra, R, Ghosh, A and Ghosh, T 2016b Comparative Assessment of Morphological and Landuse/Landcover Change Pattern of Sagar, Ghoramara, and Mousani Island of Indian Sundarban Delta Through Remote Sensing. In: Hazra, et. al. (Eds.), Environment and Earth Observation: case studies in India . Springer. ISBN: 978-3-319-46008-6. DOI: 10.1007/978-3-319-46010-9_11

36. Hajra, R and Ghosh, T 2016a Migration pattern of Ghoramara Island of Indian Sundarban-identification of push and pull factors. Asian Academic Research Journal of Social Sciences & Humanities 3(6): 186–195. ISSN: 2278-859X.

37. Hajra, R, Szabo, S, Tessler, Z, Ghosh, T, Matthews, Z, and Foufoula-Georgiou, E 2017 Unravelling the association between the impact of natural hazards and household poverty: evidence from the Indian Sundarban delta. Sustainability Science , 1–12. Springer. DOI: 10.1007/s11625-016-0420-2

38. Harris, JR and Todaro, MP 1970 Migration, unemployment and development: A two-sector analysis. The American economic review 60(1): 126–142.

39. Hazra, S, Das, I, Samanta, K and Bhadra, T 2014 Impact of climate change in Sundarban Area West Bengal, India. School of Oceanographic Studies. Earth Science and Climate Book . 9326/17.02.00. Report submitted to Caritus India, SCiAF.

40. Hazra, S, Ghosh, T, DasGupta, R and Sen, G 2002 Sea Level and associated changes in the Sundarbans. Science and Culture 68(9–12): 309–321. ISSN: 0036-8156.

41. HDR 2009 District Human Development Report. Development and Planning Department, Government of India.

42. Hertel, TW and Rosch, SD 2010 Climate change, agriculture, and poverty. Applied Economic Perspectives and Policy. DOI: 10.1093/aepp/ppq016

43. Iqbal, K and Roy, PK 2014 Examining the Impact of Climate Change on Migration through the Agricultural Channel: Evidence from District Level Panel Data from Bangladesh. SANDEE Working Papers, ISSN: 1893-1891. WP 84–14. ISBN: 978-9937-596-14-5.

44. Jana, A, Sheena, S and Biswas, A 2012 Morphological change study of Ghoramara Island, Eastern India using multi temporal satellite data. Research Journal of Recent Sciences 1(10): 72–81.

45. Kumar, DPK, Gopinath, G, Laluraj, CM, Seralathan, P and Mitra, D 2007 Change Detection Studies of Sagar Island, India, using Indian Remote Sensing Satellite 1C Linear Imaging Self-Scan Sensor III Data. Journal of Coastal Research 23(6): 1498–1502. DOI: 10.2112/05-0599.1

46. Kumar, KS and Viswanathan, B 2015 Weather Variability, Agriculture and Rural Migration: Evidence from State and District Level Migration in India. eSocialSciences.

47. Mandal, M and Mandal, S 2012 Impact of natural resource on socio-economic status of Sagar Island. West Bengal, India. The International Journal of Engineering and Science (IJES) . 2(1): 76–80. ISSN: 2319-1813.

48. Marchiori, L, Maystadt, JF and Schumacher, I 2012 ‘The impact of weather anomalies on migration in sub-saharan Africa’. Journal of Environmental Economics and Management 63: 355–74. DOI: 10.1016/j.jeem.2012.02.001

49. Miluka, J, Carletto, G, Davis, B and Zezza, A 2010 The vanishing farms? The impact of international migration on Albanian family farming. The Journal of Development Studies 46(1): 140–161. DOI: 10.1080/00220380903197978

50. Mistri, A 2013 Migration and sustainable livelihoods: A study from Sundarban Biosphere Reserve. Asia Pacific Journal of Social Sciences 5(2): 76–102.

51. Mondal, TK 2012 Environmental hazards and community responses in Sagar Island, India. Ocean and Coastal Management. DOI: 10.1016/j.ocecoaman.2012.10.001

52. Nawrotzki, RJ, Riosmena, F and Hunter, LM 2012 ‘Do Rainfall Defcits Predict U.S.-Bound Migration from Rural Mexico? Evidence from the Mexican Census’. Population Research Policy Review . DOI: 10.1007/s11113-012-9251-8

53. Raha, A, Das, S, Banerjee, K and Mitra, A 2012 Climate change impacts on Indian Sunderbans: a time series analysis (1924–2008). Biodiversity and Conservation 21(5): 1289–1307. DOI: 10.1007/s10531-012-0260-z

54. Reilly, J, Tubiello, F, McCarl, B, Abler, D, Darwin, R, Fuglie, K, Hollinger, S, Izaurralde, C, Jagtap, S, Jones, J and Mearns, L 2003 US agriculture and climate change: New results. Climatic Change. 1 , 57(1–2): 43–67.

55. Speranza, CI, Wiesmann, U and Rist, S 2014 An indicator framework for assessing livelihood resilience in the context of social–ecological dynamics. Global Environmental Change 28: 109–119. DOI: 10.1016/j.gloenvcha.2014.06.005

56. Thirtle, C, Irz, X, Lin, L, McKenzie-Hill, V and Wiggins, S 2001 Relationship between changes in agricultural productivity and the incidence of poverty in developing countries. Report commissioned by the Department for International Development, London.

57. Tilak, JB 2007 Post-elementary education, poverty and development in India. International Journal of Educational Development 31 , 27(4): 435–45.

58. UN 2015 Transforming our world: the 2030 Agenda for Sustainable Development. UN General Assembly Resolution A/RES/70/1.

59. VanWey, LK, Guedes, GR and D’Antona, ÁO 2012 Out-migration and land-use change in agricultural frontiers: insights from Altamira settlement project. Population and Environment 1 , 34(1): 44–68.

60. Vatta, K and Sidhu, RS 2010 Rural Non-Farm Employment, Income Distribution and Poverty: Micro Level Evidence from Punjab. Indian Journal of Agricultural Economics 65(4): 693–709.

61. Wedgwood, R 2007 Education and poverty reduction in Tanzania. International Journal of Educational Development 27(4): 383–396. DOI: 10.1016/j.ijedudev.2006.10.005

62. World Bank 2008 Agriculture for development. World development report 2008 . World Bank, Washington, DC. ISBN-13: 978-0-8213-6807-7.

63. World Bank 2010 World Development Indicators, Employment in agriculture (% of total employment). World Bank (data file). Retrieve from: http://data.worldbank.org/indicator/SL.AGR.EMPL.ZS on 25.12.2016.

64. World Bank 2013 World Development Indicators, Poverty headcount ratio. World Bank (data file). Retrieve from: http://data.worldbank.org/indicator/SI.POV.DDAY?locations=1W&start=1981&end=2013&view=chart on 25.12.2016.

65. World Bank 2014 Building resilience for sustainable development of the Sundarban: strategy report. Washington, DC: World Bank Group. URL: http://documents.worldbank.org/curated/en/2014/01/20162806/building-resiliencesustainable%20development-Sundarbans-strategy-report.

66. World Bank and IMF 2016 Global Monitoring Report 2015/2016: Development Goals in an Era of Demographic Change. Washington, DC: World Bank; International Monetary Fund. URL: https://openknowledge.worldbank.org/handle/10986/22547.

67. WWF 2010 Sundarbans: Future Imperfect Climate Adaptation Report. Danda, A (Ed.), WWF.