Urbanization is one of the main threats to biodiversity. However, some urban green spaces could act as refuges for urban fauna if the composition of the flora were less horticultural and if a less intensive management strategy is adopted. Among the taxa, butterflies are experiencing a strong decline from European to regional scales. An ecological engineering project based on a plantation of host and nectariferous plants backed up by a well thought out management strategy was carried out in Marseille at the Parc Urbain des Papillons (the Butterflies Urban Park). We assessed its effectiveness by comparing the butterfly communities in this park before and after the engineering work, and we compared it to a neighboring wasteland with natural habitats. After 12 years of the project, the results show a significant change in the species composition. The species richness greatly increased from 25 to 42 species. Some specialist species we targeted appeared, and their numbers increased from one to five. However, three Mediterranean species are still absent compared to the wasteland with natural habitats. As the plant palette used and the management strategy implemented enabled us to significantly increase the number of species, we now plan to work on the structure of the vegetation.
Brownfield technosol restoration represents a real opportunity to minimize land consumption, but it often requires intensive intervention and reconstruction, leading to soil donor site destruction and potential pollution. Using a former oil refinery brownfield site, this research compares the short-term (one year) impact of localized restoration methods, including soil surface scarification, organic matter addition, earthworm inoculation (locally and commercially sourced), on the soil fauna and vegetation properties. Vegetation rapidly covered the bare soil, but few significant impacts were observed in terms of the soil or vegetation structure. The initial positive impact of soil scarification on surface compaction quickly faded and earthworm inoculations did not seem to impact soil characteristics. Natural soil colonization by vegetation or fauna was the major driver in soil modification. The analysis of an intermediate reference showed a delay of multiple decades between vegetation, fauna density, and soil structure improvement, as well as the achievement of a soil of “good biological quality”. We conclude that the use of engineer species on brownfield soil in its actual form is not an early restoration step but should be tested in later steps (e.g., when organic matter content is sufficient). However, additional research is needed in engineer species selection and multi-compartment approaches. A better understanding of the impacts of natural colonization in the early stages of technosol restoration is also essential for restoration techniques’ improvement.