Evergreen species exhibit higher growth resistance under drought: insights from carbon–water relations

More frequent and extreme droughts under global climate change pose major threats to plant diversity and ecosystem productivity. Plant growth is constrained by the interplay between hydraulic failure and reduced carbon assimilation; however, how these carbon-water dynamics jointly regulate growth across functional types, particularly under varying drought intensity and duration, remains poorly understood. We conducted a meta-analysis of 249 studies covering 236 species across diverse biomes to examine differences in growth, carbohydrate allocation and hydraulic responses to drought among functional groups (e.g. evergreen vs deciduous, angiosperm vs gymnosperm, adult plants vs seedling, etc.). We also evaluated how carbon-water dynamics mediate plant growth under drought stress. We found that drought stress consistently reduced plant growth, photosynthetic rate, water potentials and the consequent hydraulic conductivity across species. Growth responses were strongly influenced by leaf phenology (evergreen vs deciduous) and drought intensity. Evergreen species showed greater growth resistance to drought than deciduous species, by maintaining photosynthesis and hydraulic function despite faster declines in water potential. Evergreen species exhibited linear reductions in growth, photosynthesis and water potentials with increasing drought intensity, reflecting gradual physiological adjustments indicative of drought resistance. In contrast, deciduous species showed significant limitation of photosynthesis and growth at drought onset. Our findings provide a quantitative framework linking plant traits related to carbohydrates and hydraulic to growth responses under drought. Understanding how drought affects carbon-water strategy based on leaf phenology advances predictive vegetation models of responses to climate extremes, with critical implications for ecosystem management and maintaining species diversity under global change scenarios.