Diverging trends in plant phenology and productivity across European mountains in a warming world

• Remote sensing data showed diverging phenological patterns in fragmented mountain ranges. • Mountain plant phenology trends varied by elevation and vegetation type. • Grassland spring advanced more at high elevations, forests at low elevations. • Climate variables predicted short-term phenology better than long-term trends. • Longer growing seasons will not necessarily translate into increasing productivity. Global warming is affecting phenology and productivity of terrestrial ecosystems, with large implications for carbon cycling. However, how plant phenological trends are shifting in climatically heterogenous mountains, and how these trends affect ecosystem productivity remains unclear. Using moderate resolution satellite data (500 m), we analyzed differences in phenological trends and productivity between vegetation types and along elevation in Europe’s major mountain ranges between 2001 and 2023. End-of-season shifts outpaced start-of-season changes in broadleaved forests (+0.15 vs. -0.05 d y⁻¹), while patterns in natural grasslands were opposite (+0.03 vs. -0.23 d y⁻¹). The magnitude of these shifts varied significantly with elevation: grassland spring phenology consistently advanced more at high than at low elevations, while broadleaved forest spring phenology exhibited mountain range-specific elevation responses—advancing more at low than at high elevations in the Alps and in the Carpathians, but not in the Pyrenees and in the Scandinavian Mountains. Autumn phenology of broadleaved forests showed greater delays at high than at low elevations, likely due to spring and summer droughts. Climate anomalies, calculated as Z-scores across the 23-year time-series, predicted phenological anomalies well (max R² = 0.51), although trends in climate over 23 years and phenological variables were related only weakly (max R² = 0.27), suggesting that plants adjusted to long-term differently than to short-term climate change. Growing season length (GSL) was strongly coupled with productivity (max R 2 = 0.60), especially in “cold-limited” vegetation. Nonetheless, temporal trends in GSL and productivity were not related (R² < 0.03). In the last 23 years, GSL significantly increased in only 22 % of forest and 16 % of grassland pixels, but productivity in 20 % of forest and 53 % of grassland pixels. Our results suggested that factors beyond GSL affect ecosystem productivity, indicating that longer growing seasons will not necessarily translate into increasing productivity across European mountains.