Interactive effects of management and temperature anomalies on CO2 fluxes recorded over 18 years in a temperate upland grassland system

• Grassland CO 2 flux sensitivity to temperature anomalies is greatest in spring. • Warm rather than cold anomalies modify temperature-flux relationships across seasons. • Cold rather than warm anomalies promote asynchrony between C uptake and loss. • Cold anomalies promote divergence in fluxes between management treatments. • High management intensity promotes sensitivity of CO 2 fluxes to warm anomalies. Temperature anomalies such as heat waves and cold spells are increasingly common, posing threats to ecosystem functioning and delivered services. Although temperature anomalies have been shown to limit broad-scale patterns of carbon sequestration, fine-scale studies on CO 2 fluxes and temperature anomalies are limited and interactions between temperature anomalies and agricultural management remain poorly understood. Using long-term data on grassland CO 2 fluxes and air temperature (2003-2021), we investigated the effects of temperature anomalies (cold, warm, extreme cold, extreme warm) on gross primary productivity and ecosystem respiration in adjacent upland mesic grasslands subjected to either extensive or intensive cattle grazing management treatments. Both warm and cold temperature anomalies occurred throughout the study period, with 40 days per year on average classed as colder or warmer compared to normal. Irrespective of management treatment and flux type, we found that temperature anomalies had a significant effect on the magnitude of CO 2 fluxes, the relationship between air temperature and CO 2 fluxes, and the occurrence of anomalous flux events. The magnitude of CO 2 fluxes was most affected by temperature anomalies at the start of the growing season, and fluxes were generally more sensitive to warm rather than cold temperature anomalies. However, cold temperature anomalies promoted asynchrony between C uptake and loss. Temperature-CO 2 flux relationships were upregulated by warm temperature anomalies in spring but down-regulated in summer and autumn, with greater reductions in flux process rates in the intensive management treatment. Nevertheless, temperature anomalies were not found to be a strong driver of flux anomalies in this study system. Collectively, our results show that grassland management interacts with the timing and nature of temperature anomalies on CO 2 fluxes and suggest that low grassland management intensity may buffer CO 2 fluxes against warm air temperature anomalies in cool temperate systems, with implications for adaptation strategies for grasslands under climate change.