Food import diversification is an important strategy for improving the resilience of import-dependent food systems. This paper studies the allocation of food imports across multiple suppliers while considering three conflicting objectives: reducing supplier risk, reducing import cost, and limiting dependence on dominant suppliers. The problem is formulated as a constrained multi-objective optimization model and solved using two evolutionary approaches: NSGA II and MOEA/D. A projection step is used to ensure that all generated allocations satisfy demand and supplier capacity constraints. Computational experiments compare the two algorithms using hypervolume, inverted generational distance, runtime, and non-dominated solution quality. The results provide insight into the behavior of dominance-based and decomposition-based evolutionary methods for constrained food import allocation.