World Food Policy Volume/Issue 2-2/3-1 Fall 2015/Spring 2016

World Food Policy more pronounced in the 25% income percentile than in the whole sample. These differences largely stem from the fact that non-fishers in the 25% income percentile have considerably lower calorie and protein intakes than in the whole sample. For fishers the difference between the whole sample and the 25% percentile is not as pronounced. Almost all of the difference in protein intake between fishers and non-fishers in the 25% income percentile may be explained by the difference in proteins coming from fish. Meanwhile, only ~35% of the difference in the calorie consumption between fishers and non-fishers in the 25% percentile can be explained by differences in fish consumption. This confirms the opinion of Kawarazuka and Béné (2010; 2011) that fish is of special importance in poorer households. To sum up, the findings in this section indicate that fishing households are more engaged in subsistence activities than non-fishing households and have lower cash income. We further show that fishers in the overall sample had a more nutritious diet in terms of proteins and calories in the past week. This finding is confirmed by higher FCSs and the FCS components of fishers for the last week. For the 25% income percentile Figure 4: Comparison of nutrition variables of fishers and non-fishers in the whole sample and the 25% income percentile in StungTreng, Cambodia. All figures are per capita in adult equivalents. Pooled sample 2013 and 2014 17

World Food Policy Volume/Issue 2-2/3-1 Fall 2015/Spring 2016 | Page 17