Aim: Compared with gradual climate change, extreme climatic events have more direct and dramatic impacts on vegetation growth. However, the influence of climate extremes on important phenological periods, such as the end of the growing season (EOS), remains unclear. Here, we investigate the temporal trends of EOS across different biomes and quantify the response of EOS to multiple climate extreme indices (CEIs).
Location: Northern middle and high latitudes.
Time period: 2000–2020.
Major taxa studied: Plants.
Methods: Three phenology extraction methods were used to compute EOS from satellite, FLUXNET and Pan European Phenology Project PEP725 phenological datasets. Different stress states of cold, hot, dry and wet extremes were represented by 12 CEIs. Partial correlation and ridge regression analysis were used to quantify the response of EOS to climate extremes across latitudinal and biome scales.
Results: Our study showed a delayed EOS in boreal biomes, but a significantly advanced EOS in temperate biomes. The advanced EOS induced by cold stress was observed for c. 80% of the vegetated pixels. The warm-related CEIs delayed the EOS in high latitudes, and the delayed effect weakened or even reversed with decreasing latitude. In contrast, EOS exhibited opposite response patterns to dry days and wet-related CEIs. Overall, EOS exhibited higher sensitivity to extreme temperature in boreal biomes than in temperate biomes. Specifically, continuous drought and high heat stress induced an earlier EOS in some temperate forest biomes, whereas moderate heat stress delayed the EOS in most study biomes. In contrast, EOS was not sensitive to extreme drought in water-restricted biomes.
Main conclusions: EOS exhibited divergent responses to various climate extremes with different intensities and frequencies. Moreover, the response of EOS to extreme climate stress was dependent on the biome and latitude. These findings emphasize the importance of incorporating the divergent extreme climate effects into vegetation phenological models and Earth system models.