Altered water uptake patterns of Populus deltoides in mixed riparian forest stands

Plant water uptake plays an important role in regulating ecosystem water balance and its productivity. However, previous studies regarding plant water uptake were primarily conducted in wet areas under seasonal drought conditions, with a limited understanding of the proportion and drivers of plant water uptake under humid conditions. Actually, climate change and variations in global precipitation patterns could simultaneously trigger seasonal drought and flooding. Therefore, it is critical to explore patterns and mechanisms for plant water uptake under humid conditions in wet regions. Here, we employed dual stable isotopes of hydrogen and oxygen coupled with a Bayesian mixing model (MixSIAR) to explore the water uptake patterns of Populus deltoides in two types of riparian forests (pure vs. mixed stand of P. deltoides), under different magnitudes of rainfall (7.9, 15.4 and 34.1 mm), in the Middle-Lower Reaches of the Yangtze River in China. We further used both partial correlation and variation partitioning analyses to determine the relative importance of soil variables and plant properties affecting the proportion of P. deltoides water uptake from different soil layers. Our results revealed that compared to pure stands, P. deltoides in mixed stands had a lower water uptake proportion from deep soil layers (60-80, 80-100 cm) and had higher water uptake from shallow soil layers (0-20, 20-40 cm) under 15.4 mm and 34.1 mm rainfall events. Our results also revealed that plant properties such as leaf biomass, fine root biomass, and diameter at breast height were the primary factors influencing water uptake by P. deltoides. This suggests that P. deltoides in mixed stands could increase the proportion of water uptake from shallow soil layers through altering plant attributes. These findings indicate that mixed stands could restrain frequent extreme rainfall events and subsequent flooding, suggesting more resilience towards future climatic variability.