North Atlantic Oscillation Modulates Long-Term ANPP Dynamics via 1 Precipitation or 2 Temperature, Depending on Soil Nutrient Levels

Precipitation and temperature are major controls of the inter-annual dynamics of Aboveground Net Primary Production (ANPP). However, the effect of humans, through fertilization, acts on ANPP in combination with environmental variations. This raises two questions that drive our research: (1) How do seasonal variations of global atmospheric patterns, such as the North Atlantic Oscillation (NAO), affect precipitation and temperature, and ultimately influence ANPP? (2) Does long-term fertilization modify the potential pathway from the NAO, through precipitation and temperature, to ANPP? We addressed these questions using data from plots with either ‘high’ or ‘low’ nutrient addition from the world longest ecological experiment (period used for this analysis: 1950 to 2018), ‘Park Grass’, at Rothamsted Research, England, UK. We evaluated the relationships between ANPP and interannual climate variations across both types of plots using a structural equation model, with the aim to understand how monthly variations in the NAO affected precipitation and temperature, and how these, in turn, affected ANPP. We observed a signal of NAO on ANPP. However, the signal differed between ‘low’ and ‘high’ nutrient addition plots. Under the ‘low’ nutrient addition level, the NAO signal from all months exerted some influence on ANPP (R2: 0.38, P-value: 0.001), and all pathways were associated with precipitation. Under ‘high’ nutrients addition, the NAO signal exerted a small influence on ANPP (R2: 0.17, P-value = 0.001) through April precipitation and May temperatures. These results shed light on the link between global atmospheric patterns and local ecosystem functioning. Our work also confirmed a shift from a precipitation-driven response to one mediated by temperature when shifting from ‘low’ to ‘high’ nutrient conditions. These contrasting patterns suggest there is no simple way to explain the mechanisms by which 34 global atmospheric patterns influence ecosystem functioning.