The dataset describes variables collected from the DIVEGFOOD experiment (“DIVersified protected VEGetable systems in line with their FOOD supply chains”) that was set up at the agroecological vegetable systems experimental facility of the National Research Institute for Agriculture, Food and Environment (INRAE) in Alénya, in the South of France (42,63765°N; 2,971884°E). DIVEGFOOD consisted in assessing four cropping systems in a four-year system experiment from October 2018 to September 2022. A low-diversified system named REF, and three more diversified alternatives - named MODIV, DIVBANDE and DIVMIX - were implemented under walk-in plastic tunnels in organic farming in loamy-sandy soil under a Mediterranean climate. REF consisted of 11 and 2 cycles of cash crops and cover crops respectively. It was occupied by a sole cash crop at each growing cycle, was solarized every third summer and produced a small range of 4 species from 3 botanical families, with lettuce exclusively in winter. MODIV differed from REF by a wider diversity over time. As a highly strategic product locally, lettuce was maintained at a lower frequency, resulting in a more diversified pluriannual crop sequence. The DIVBANDE and DIVMIX systems differed significantly using no longer soil solarization, and introducing a higher spatial complexity of plants with continuous intercropping of three crops under the shelter. DIVBANDE used a strip intercropping system where there can be crop interactions but where strips can be cultivated independently under tunnel. The DIVMIX system was inspired by mixed intercropping but adapted to protected vegetable management requirements. It consisted in simultaneously growing three crop species in close interaction, managed as a whole. Each cropping system was implemented under one 400m² unheated plastic walk-in tunnel. One objective of DIVEGFOOD was to report on diversified systems’ exposure to negative hazards and assess whether diversified systems presented higher performances than the baseline system. The dataset provides information on winter crops’ exposure to major hazards such as disturbances or stresses that may negatively affect the vegetables cultivated. The hazard indicators were related to pressure from the main pests and diseases, injuries on the harvested part of the plants and non-optimal agronomic and technical conditions. The dataset also provides winter crop production performances with the marketable yields and the rate of unmarketable vegetables. Four indicators related to cropping system’s performances are also presented: ‘Treatment frequency index’ reflecting their dependency on plant protection products in organic farming, ‘labour time’, ‘operational costs’ referring to production factors, and ‘gross product’ expressing economic outputs.

The dataset is made publicly available to allow free and easy access for the scientific and professional community to enable innovative highly diversified market-gardening systems design, analysis and reuse.