Woody plant encroachment intensifies under climate change across tundra and savanna biomes.

Aim: Biomes worldwide are shifting with global change. Biomes whose extents are limited by temperature or precipitation, such as the tundra and savanna, may be particularly strongly affected by climate change. While woody plant encroachment is prevalent across both biomes, its relationship to temperature and precipitation change remains unknown. Here, we quantify the degree to which woody encroachment is related to climate change and identify its main associated drivers. Location: Tundra and savanna biomes. Time period: 1992 ± 20.27–2010 ± 5.62 (mean ± SD). 1876–2016 (range). Major taxa studied: Woody plants (shrubs and trees). Methods: We compiled a dataset comprising 1,089 records from 899 sites of woody plant cover over time and attributed drivers of woody cover change across these two biomes. We calculated cover change in each biome and assessed the degree to which cover change corresponds to concurrent temperature and precipitation changes using multiple climate metrics. Finally, we conducted a quantitative literature review of the relative importance of attributed drivers of woody cover change. Results: Woody encroachment was widespread geographically and across climate gradients. Rates of woody cover change (positive or negative) were 1.8 times lower in the tundra than in the savanna (1.8 vs. 3.2%), while rates of woody cover increase (i.e., encroachment) were c. 1.7 times lower in the tundra compared with the savanna (3.7 vs. 6.3% per decade). In the tundra, magnitudes of woody cover change did not correspond to climate, while in the savanna, greater cover change corresponded with increases in precipitation. We found higher rates of woody cover change in wetter versus drier sites with warming in the tundra biome, and higher rates of woody cover change in drier versus wetter sites with increasing precipitation in the savanna. However, faster rates of woody cover change were not associated with more rapid rates of climate change across sites, except for maximum precipitation in the savanna. Main conclusions: Woody encroachment was positively related to warming in the tundra and increased rainfall in the savanna. However, cover change rates were not predicted by rates of climate change, which can be partially explained by climate interactions in both biomes. Additional likely influences include site‐level factors, time‐lags, plant‐specific responses, and land use and other non‐climate drivers. Our findings highlight the complex nature of climate change impacts in biomes limited by seasonality, which should be accounted for to realistically estimate future responses across open biomes under global change scenarios. [ABSTRACT FROM AUTHOR]