DOI: 10.12688/f1000research.179180.1 ISSN: 2046-1402

How can food waste be reduced in the supply chain? Evidence from predictive modeling in wholesale food markets

Thayane Caroline Oliveira da Conceição, Dag Mendonça Lima, Andréa Leda Ramos de Oliveira, Matheus Araujo Maciel, Gleidisom Thyery Souza da Costa
Background The waste of fruits and vegetables along the supply chain poses logistical, economic, and environmental challenges, directly impacting food security and the achievement of Sustainable Development Goal (SDG) 12.3. The latest global monitoring indicates that progress toward achieving this goal by 2030 remains insufficient, requiring urgent integration of logistical efficiency and social responsibility strategies to mitigate systemic waste. Reducing fruit and vegetable waste in wholesale markets requires evidence of modifiable operational practices. Objective To assess whether logistical and commercial practices can predict waste levels in wholesale food markets. Methods We conducted a cross-sectional survey of traders operating in Brazilian wholesale markets in 2023. The dataset comprised 16 variables describing procurement, handling, storage, transport, and commercialization practices to predict self-reported waste, which was dichotomized into high and low. We compared logistic regression, a feedforward neural network, and a support vector machine (RBF kernel) and evaluated discrimination using the area under the ROC curve (AUC). Results Logistic regression showed the best discrimination among the tested models. A lower likelihood of high waste was observed among traders who reported using cold storage, adequate packaging, pre-sale processing, palletized handling, and adequate transportation. Conclusions A small set of operational practices can support predictive screening for waste risk in wholesale markets, inform infrastructure investment, and public policy to reduce waste. This study contributes by empirically validating analytical models that help identify operational bottlenecks and provide actionable evidence to inform strategies for reducing post-harvest losses and supporting environmental and social sustainability .

More from our Archive