Shrimp-prawn farming in Bangladesh : impacts on livelihoods, food and nutritional security

The overarching aim of this research was to improve understanding of the synergies and trade-offs between economic and domestic food security benefits associated export-oriented shrimp and prawn aquaculture in a fragile developing country environmental context. The research scope covers the entire ‘seafood system’ incorporating production, distribution, trading and consumption in the south-west coastal region under greater Khulna District, Bangladesh. The primary research objective was to understand causal factors in variation of seafood nutritional quality and health outcomes, exploring correlations with seafood consumption across a range of agro-ecological, aquatic farming systems and socio-economic conditions. The study also focused on differences in intra-household allocation with a special focus on adolescent girls, being amongst the most vulnerable members of society. These objectives necessitated a highly inter-disciplinary approach to understand complex interactions between biophysical aspects (e.g. where and which species are most effective in supplying essential nutrients) and the social norms of food allocation. Fieldwork was conducted in Satkhira, Khulna, and Bagerhat Districts of Khulna Province over 17 months during 2012 to 2015. Four villages across a peak-annual surface water salinity gradient; high saline (>10 ppt), medium saline (>5<10 ppt), low saline (<5 ppt) and freshwater (<0.5 ppt) were selected for case-studies. The thesis begins with a literature review of the evolution of shrimp and prawn farming in Bangladesh and the wider region and identification of knowledge gaps. Research resolved from district to community to household level. Key informant (KI) interviews were used to establish well-being criteria (based on a range of 5 social and economic assets) at community level. In each community a census of households (n=1082 households) were derived from the same interviews and KIs asked to ranked households on the established well-being criteria. Results were validated through a short interview of all the identified households (HH). This sample-frame provided the based for two concurrent survey efforts. Stratified-random selection of 160 HH on two well-being categories (better-off and worse-off) for ‘farm level’ analysis using a semi-structured questionnaire. Key topics included inputs/outputs characteristics, economic benefits and the fate of farmed products were evaluated. Another 240 households with single adolescent girls were selected from the same frame, again with randomized-stratified sampling based on well-being categories for ‘intra-household’ analysis. This resulted in selection of 60 HH per community consisting of 30 ‘better-off’ and 30 worse-off households (further analysis was conducted on a range of secondary sampling outcomes based on livelihood options, intra-household food distribution and aquatic farming assets). A 24-hour food recall method, food frequency questionnaire, food photography and measuring cup sets were used to estimate individual members’ food consumption at the household level. Anthropometric measures (stunting, wasting, BMI, MUAC) and biomarkers (omega-3 index in RBC and LC n-3 PUFA/LC PUFA in whole blood cell) were used to assess food security outcomes of adolescent girls (n=200 subject). In an entirely separate effort, samples of shrimp/prawn and fish polyculture species (57 species and 9 by-products, 1 live feed; n=672) were collected from the major agro-ecologies (four saline gradients; HS, MS, LS and FW) and culture systems (extensive, semi-intensive, intensive, organic and pocket gher). At least 3 sites from each saline gradient (3×4=12 sites) were selected for sampling and pooled samples to represent all the culture system in the region. The major macro and micronutrients of the collected species were analysed and these datasets were used to know the nutritional distribution among the family members in household level study. Two aggregate indices of wealth (or well-being) and aquaculture were developed based on a range of quantitative (ordinal and interval) measures. Principle Component Analysis (PCA) was used to understand how aquaculture influences on wealth gain. Wealth index of the same social well-being did not differ among the agro-ecologies. However, the aquaculture index was varied in agro-ecologies. Most of the households (60-80%) were involved directly in aquaculture however, a majority portion of the households was worse-off (48-64%). Both HS and LS area had more livelihood options due to their proximity to mangrove forest Sundarbans and city amenities, respectively compared to MS and FW area. Ownership of the land did not influence any involvement in aquaculture and yields but he willingness and risk-absorbing capacity of the farmers were the main factor to get involve in aqauculture. In aquatic farming system the importance of export-oriented shellfish yield gradually decreased from HS to FW area (55-20% by volume). The intentional stocking of finfish and PLs were common across the saline areas however the wild caught juveniles and hatchery originated fingerlings were usual in higher and lower saline areas, respectively. The low priced tilapia took the place of wild recruited mangrove fishes in MS area. Diseases, especially the devastating white spot virus (WSSV), frequency were higher in higher saline areas. The indicators like wild recruitment, salinity, water productivity and water management also a vital factor to gain yield. The integration of aquatic and terrestrial crops (rice and dyke crop) in the lower saline areas provide higher yield compared to higher saline areas. However, the net economic returns were largely determined by the aquatic products. The income of ghers in FW and MS area was sensitive to the lower prices of freshwater finfish and tilapia. The protein content in shellfish was higher than the other finfish, however, lower in other essential nutrients. Species living in the higher saline areas contained higher total n-3 PUFA (in weight) and LC n-3 PUFA/LC-PUFA compared to the same species living in lower saline areas. Small Indigenous Species (SIS) and Self-Recruiting Species (SRS) were proven to provide higher micronutrients and total n-3 PUFA than larger fish. Seafood that destined for the international markets contained less n-3 PUFA and micronutrients in comparison to the domestically consumed fish. Customary intra-household food distribution disparity (mainly fish) still exists at household levels where females, especially adolescent girls, were deprived. Fish consumption (>77 g /capita/day) and fish originated protein supply (>25% of total protein intake) was higher than the other part of Bangladesh. The protein consumption of adolescents was 2-3 times higher than the Recommended Nutritional Intake (RNI). However, the energy intake was lower than the required level. High protein, low energy consumption was not reflected in body mass. Micronutrients (zinc) consumption was above the RNI level. However, iron and calcium consumption was less than the RNI. The n-3 PUFA in RBC of adolescent girls accurately reflected their access to, and availability of, oily fish. In the omega-3 index (n-3 PUFA in red blood cell) both HS and MS areas, adolescent females were in the intermediate stage (4-8%), and rest of the two areas were in the undesirable stage (