Recurrent food crises and climate change, along with habitat loss and pollution, have put food security and environmental sustainability at the top of the political agenda. Analyses of the dynamic linkages between food consumption patterns and environmental concerns have recently received considerable attention from the international and scientific community. Using the lens of a wide sustainability concept, this paper aims at developing a multidimensional framework for evaluating sustainability in food systems and diets applicable to the Mediterranean countries. The Mediterranean region - a geographically interlocked and heterogeneous area including South European, North African and South-East Mediterranean Countries – presents several conditions of vulnerability to food insecurity and unsustainability. Furthermore the demographic growth, in urban and coastal areas of the Basin, leads to an increasing pressure on natural resources and widening disequilibria with rural areas. Derived from natural disaster and sustainability sciences, a coupled domain/vulnerability approach has been applied to the analysis of the concepts of sustainable food security and diets. Within consensus-based preselected domains, the vulnerability methodology offers a coherent framework that disentangles exposure, sensitivity and adaptive capacities. A DELPHI method is applied to select the final set of indicators from the literature. The main result is the elaboration of an innovative conceptual evaluation framework for measuring sustainability in the agrofood system. The framework draws upon two existing approaches: a vulnerability-based assessment method combined with an analysis of detailed empirical domains relevant for Mediterranean countries. The DELPHI selection process, involving several international experts, has reduced the number of indicators to a reduced pool of indicators. Use of the participatory approach of the DELPHI method helps move beyond subjective evaluation and reach consensus. Recognizing the systemic dimension of sustainability, the vulnerability approach enables to investigate the causal factors dynamics, instead of targeting exclusively the final outcomes. The domain-based framework reflects the region- specific attributes that necessarily need to be identified to link scientific concepts with metrics. Urbanisation is a main domain of vulnerability as it is a key driver of change affecting both market dynamics and consumers’ behaviours, raising questions for food security.

Consumer preference for boiled potato in Uganda has been assessed to be prevalently based on soft (hardness) and mealy texture. However, harmonised standard procedures have not been hitherto developed to characterize the texture of boiled potato instrumentally in SSA. The RTBBreeding© project focussed on developing discriminant and sensory-correlated procedures based on comparative analyses of Extrusion, Penetration and Texture Profile Analysis (TPA) techniques for mid-throughput evaluation of texture of steamed potato from potato varieties. Tubers of nine popular landrace varieties cultivated in Kabale and Rakai districts of Uganda were used. The tubers were cut into 25 mm cubes, steamed for 15 min in banana leaves and analysed using a calibrated TA-XT texture analyser under standard conditions (Extrusion: test speed 1 mm s-1, strain 80 %, 5-blade grid Ottawa cell; Penetration: test speed 1 mm s-1, distance 10 mm, 60° cone probe; TPA: test speed 1 mm s-1, distance 5 mm, wait period 5 s, 75 mm cylindrical plate) at about 25 °C. For descriptive sensory analysis, twenty random tubers from each variety were steamed for 40 min and evaluated in duplicate by trained panellists for eleven selected sensory parameters on a 11-point scale ranging from 0 (minimum intensity) to 10 (maximum intensity). Results show that Area under curve/Extrusion work (73-236 N.mm), Maximum force/ hardness (6-17 N), and End force (6-15 N) were the more discriminant textural parameters for extrusion. Maximum force, End force and Extrusion work significantly correlated with sensory Moisture release and Hardness by hand. The discriminant textural parameters for penetration were Area under curve/ Penetration work (6-21 N mm) and Maximum force/ hardness (2-6 N). Significant correlations exist between penetration Hardness and Area under curve and the sensory Hardness by hand, fracturability, cohesiveness, and smoothness. Finally, the more discriminant textural parameters for TPA were Hardness (20-52 N), Gumminess (5-19 N) and Chewiness (5-18 N). There are significant correlations between TPA Adhesiveness and sensory moisture release and mealiness. Among the methods, the penetration method was more preferred as it correlated most with sensory evaluation, while TPA was least correlated with sensory.