Proximity to small-scale inland and coastal fisheries is associated with improved income and food security

Poverty and food insecurity persist in sub-Saharan Africa. We conducted a secondary analysis of nationally representative data from three sub-Saharan Africa countries (Malawi, Tanzania, and Uganda) to investigate how both proximity to and engagement with small-scale fisheries are associated with household poverty and food insecurity. Results from the analysis suggest that households engaged in small-scale fisheries were 9 percentage points less likely to be poor than households engaged only in agriculture. Households living in proximity to small-scale fisheries (average distance 2.7 km) were 12.6 percentage points more likely to achieve adequate food security and were 15 percentage points less likely to be income poor, compared to the most distant households. Households distant from fishing grounds (>5 km) were 1.5 times more likely to consume dried fish compared to households living close. Conserving the flow of benefits from small-scale fisheries is important for meeting the Sustainable Development Goals in the region.

Estimated marginal effects on the probability to be food insecure (households with below adequate food consumption scores), by quintile of distance to water bodies Supplementary Table 2 Summary (unweighted) statistics for the econometric analysis Supplementary Table 3a Fish and other animal-source food consumption by proximity to water bodies Supplementary Table 3b Fish and other animal-source food consumption by proximity to water bodies -by poor/non-poor HHs (above and below national poverty lines) Supplementary Table 3c Fish and other animal-source food consumption by proximity to water bodies -by dietary diversity score (food consumption score) Supplementary Table 4a Fish consumption -form (dried/fresh) and source (purchased/own production/other) -by proximity to water bodies Supplementary Table 4b Fish consumption -form (dried/fresh) and source (purchased/own production/other) -by proximity to water bodies and poor/non-poor HHs (above and below national poverty lines) Supplementary Table 5 Fish consumption -quantity -by quintile and proximity to water bodies Supplementary Table 6_Malawi Fishing HHs vs non-fishing HHs -fish consumption, dietary diversity and multiple poverty indicators Supplementary Table 6_Tanzania Fishing HHs vs non-fishing HHs -fish consumption, dietary diversity and multiple poverty indicators Supplementary Table 6_Uganda Fishing HHs vs non-fishing HHs -fish consumption, dietary diversity and multiple poverty indicators Supplementary Table 6_All countries Fishing HHs vs non-fishing HHs -fish consumption, dietary diversity and multiple poverty indicators Supplementary Table 7 Dietary diversity (food consumption score) by proximity to water bodies Supplementary Table 8 Multiple poverty indicators by proximity to water bodies Supplementary Table 9 Estimated fish consumption -in KG / per capita / year -compared to FAO Supply Data, and scenario for increasing 1% fish supply Supplementary Table 10 Prices of animal source food purchased by households Supplementary Table 11 Variable description Supplementary Table 12 Nutrient content of fish species caught from small-scale fisheries in the region and other animal source foods (per 100g edible food). Supplementary   Note: For each household, the estimated probability to be food insecure was calculated as: , where Φ is the cumulative normal distribution, X_i is the data vector consisting of observable characteristics for the i-th household, including the distance to water bodies, and β is the vector of coefficient estimates reported in Table 1B. Finally, e_i is the model error term. share of HHs who consumed over 7 days (*100) 0.11*** 0.08*** 0.1 0.12*** 0.09*** 0.11 0.09*** 0.06*** 0.08 Goat Number of days consumed (out of 7) 0.26*** 0.14*** 0.24 0.30*** 0.18*** 0.28 0.17*** 0.08*** 0.14 Pork

Supplementary
Households defined as fishing households were engaged in small-scale fisheries because the survey excluded commercial businesses, and livelihood activities were characterised by family labour inputs and a small volume of production.
Livelihoods: agriculture households Household with at least one member who engaged in agriculture (non-fishing) in the past 12 months. Livelihoods: non-agriculture/fishing households Household with no members who engaged in agriculture or fishing in the past 12 months. Average prices (constant 2010 in international US$) of household purchased food item per kg Price of food item [fish, fresh fish, dried fish, other fish, poultry, goat, pork, beef, eggs] purchased and consumed over past 7 days based on average prices (constant 2010 in international US$) per kg.

Capital assets -wealth index
Total number of assets owned by the household (durable goods -radio, bicycle, TV; utilities and infrastructure -access to protected water source and electricity) with an index developed using principal component analysis Households with cultivated and/or owned land Household with land Per-capita monthly expenditure (local currency unit) Monthly total expenditure by the household (divided by the total household members) Share of households living below the national poverty line (local currency unit) Percentage of households with per-capita monthly expenditure below the national poverty line based on cost of food and non-food basic needs (calculated by national authority) (World Bank, 2020).
Share (%) of households who consumed animalsource foods over the past 7 days Percentage of households who self-reported to consume animal-source foods [fish, poultry, goat, pork, beef, eggs] at least once over 7 days. Share (%) of households who consumed dried, fresh or other form of fish over the past 7 days Percentage of households who self-reported to consume dried, fresh or other form of fish over 7 days.
Share (%) of households who consumed fish over the past 7 days from purchased, own production or other source (including gifts).
The share of households who obtained fish for consumption over the past 7 days via purchases, own production or other sources, including gifts.
Number of days households consumed fish over the past 7 days Frequency of fish (general, dried, fresh and other) consumed over 7 days Food Security -diets and seasonality of subjective food insecurity Supplementary Table 11 Variable description All variables are obtained from the LSMS-ISA household survey, except for proximity to fishing grounds where geospatial data was obtained from the Global Lakes and Wetlands Database (GLWD) (Lehner & Döll, 2004), and the European Space Agency GlobCover databases for coastlines (ESA, 2009).

Demographic Food Environments -physical and economic access to food
Livelihoods: small-scale fishing households

Socio-economic Impacts -poverty
Quantity of fish (general, dried, fresh and other) consumed over the past 7 days (kg/per week/perhousehold) Quantity of fish consumed; general and by form (dried, fresh and other), by the households acquired from purchased, own consumption or gifts An index capturing food consumption profiles at the household level. Based on the diversity and frequency of food groups consumed over a 7-day recall period, with weights given to groups based on nutritional value. Validated as a proxy for energy sufficiency (quantity of food) (Leroy et al., 2015;World Food Programme, 2008). Food consumption profile: poor, borderline, acceptable Household with a food consumption score that is defined according to World Food Programme (2008) universal classification as poor (FCS<28), borderline (FCS>=28 and <41) or acceptable (FCS>=41), where acceptable represents food secure households.
Food insecure household over the past 12 months Subjective food insecurity defined as households who have experienced a situation in the last 12 months when they did not have enough food to feed the household. Total months household self-reported food insecurity over the past 12 months The number of months during over the past 12 months that a household reported not having enough food to feed the household  Fig 4. Difference in the quantities of fish consumed (kg/household/week) between richest (fifth quintile) and poorest (first quintile) households, by rural and urban areas and proximate (<5km) or distant (>5km) from water bodies in Malawi, Tanzania and Uganda. Graph depicts greater inequalities in consumption of fish in contexts on the left, and lower inequalities on the right. Green represents rural areas, and blue urban, whilst solid colours represent proximate to fishing grounds, and faded distant.