ABSTRACT
Increasing the overall use efficiency of nitrogen (N) and phosphorus (P) resources in food production while minimizing losses to the environment are required to meet the dual challenge of food security and sustainability. Yet studies quantifying the overall performance of different agro-system types and how these have changed over time remain rare, although they are essential to propose solution pathways. Here, we reconstructed fluxes of N and P within 78 watersheds of the St. Lawrence Basin (SLB) of eastern Canada between 1901 and 2011, using the Generalized Representation of Agro-Food System model (GRAFS). This analysis allowed us to classify different agro-food system types and to evaluate how agricultural specialization influenced nutrient efficiencies and potential losses to the environment over time. Using a cluster analysis, we identified four agro-food system types with different overall outcomes in efficiencies and losses. We show that agricultural practices in the SLB were similar until the 1950's and deemed unsustainable in several watersheds by depleting agricultural soils of their nutrients (particularly N). With the advent of manufactured fertilizers and the intensification of livestock farming, the SLB then rapidly shifted through the 1970s and 1980s to more intensified and highly unsustainable agro-food system types, where, in 2011, ~77 % of N and ~ 94 % of P inputs were lost to the environment. We also show that nutrient pollution continued to increase despite gains in the nutrient use efficiency of animal farming due to higher nutrient throughput from intensive production. The increased proportion of confined animals, disconnected from croplands, indeed resulted in inefficient nutrient recycling. While nutrient use efficiency may mitigate nutrient pollution, reducing the absolute nutrient flux through agro-food systems should be a priority, likely through a reconnection of crop and animal farming and an overall reduction of meat production, specifically from concentrated, intensive livestock systems.
