Thirty years of nitrogen dynamics in European agri-food systems: a territorial metabolic perspective

Abstract European agri-food systems face the dual challenge of reducing their environmental footprint, particularly nitrogen losses, while meeting the demands for both food and non-food biomass. Because future responses will have to build on the legacy of past trajectories, addressing this challenge requires a sound, fine-grained understanding of those trajectories. In the present study, we address this challenge by compiling a dataset (1990–2019) quantifying nitrogen flows across the agri-food systems of 120 European territories. Using this dataset, we apply the generalized representation of agri-food systems framework to classify agri-food systems into six types representing different degrees of specialization. Our results show that agricultural systems became increasingly specialized, rising from 22% of specialized territories in 1990 to 36% in 2019, indicating a growing reliance on external feed and fertilizer inputs. Most of these transitions occurred in eastern Europe, predominantly toward stockless cropping systems. Although these evolutions unfolded within broader processes of market liberalization, transition patterns also displayed clear path-dependency effects rooted in the material configuration of agri-food systems. This is evidenced by the analysis of transition frequencies, which shows that some transitions occurred repeatedly while others never occurred, pointing to the influence of initial system configurations on subsequent trajectories. By contrast, changes in agro-environmental performance were only weakly linked to system transitions. The increases in nitrogen use efficiency (from 0.57 to 0.70) and declines in nitrogen surplus (from 68 to 49 kg N ha −1 ) rather coincided with higher yields and increased nitrogen input application rates, pointing to the dominant role of factors not explicitly captured here, such as shifts in agronomic practices and pedo-climatic conditions. Taken together, these findings provide both a critical dataset and a robust understanding of recent territorial agri-food trajectories, thereby offering a solid empirical basis for anticipating future transitions and developing scenario-based analyses.