Plant-groundwater interactions in drylands: A review of current research and future perspectives

Global dryland areas are expanding due to climate change, and this expansion is accompanied by an increased frequency and intensity of drought occurrences. The degree of water stress in drylands is projected to further increase in the future, thereby threatening dryland ecosystem sustainability. Groundwater is among the most important water resources for dryland ecosystems to maintain growth and buffer against drought; thus, it is receiving increasing attention. Although plant-groundwater interactions (PGIs) have been directly observed in academia for 100 years, the mechanisms of their interactions remain unclear due to limited observations. Here, we review the research progress in the indication, intensity and related quantitative methods and mechanism of PGIs, with an emphasis on their divergence at various temporal and spatial scales, and we highlight the controversies or uncertainties that exist. Then, by overviewing the imbalance of plant-groundwater relationships that exist in several typical dryland ecosystems worldwide under climate change and anthropogenic disturbances, we identify the canonical ecological and environmental problems facing drylands both at present and in the future. Finally, we present the main limitations in the study of PGIs in drylands and offer insights into future research priorities. We argue that there are still great difficulties and uncertainties in accurately quantifying the strength of PGIs at present, both in ground monitoring and remote sensing inversion, especially at large scales and high temporal resolutions. While developing physical mechanism-based ecohydrological models is an effective way to address this problem, the current gaps in academia's understanding of the rhizosphere processes of groundwater-dependent plants are among the major constraints. We emphasize the importance of groundwater-dependent plants for dryland ecosystems despite their sparse distribution and that the causes and feedbacks of their changes are not often local but regional.