What represents a water source for the ecological restoration of a plant in an arid region is still up to debate. To address this issue, we conducted an in situ experiment in the Ulan Buh Desert of China, to study desert plants absorbing atmospheric water vapor. We selected Tamarisk, a common drought-salt-tolerant species in the desert, for ecological restoration as our research subject, used a newly designed lysimeter to monitor precipitation infiltration, and a sap flow system to track reverse sap flow that occurred in the shoot, branch, and stem during the precipitation event, and observed the precipitation redistribution process of the Tamarisk plot. The results showed that Tamarisk indeed directly absorbs precipitation water: when precipitation occurs, the main stem, lateral branch, and shoot all show the signs of reversed sap flow, and the reversed sap flow accounted for 21.5% of the annual sap flow in the shoot and branch, and 13.6% in the stem. The precipitation event in the desert was dominated by light precipitation events, which accounted for 81% of the annual precipitation events. It was found that light precipitation can be directly absorbed by the Tamarisk leaves, especially during nighttime or cloudy days. Even when the precipitation is absent, it was found that desert plants can still absorb water from the unsaturated atmospheric vapor; even the absorbed atmospheric water vapor was transported from the leaves to the stem, forming a reversed sap flow, as a reversed sap flow was observed when the atmospheric relative humidity reached 75%. This study indicated that the effect of light precipitation on desert plants was significant and should not be overlooked in terms of managing the ecological and hydrological systems in arid regions.

With global climate change, changes in vegetation phenology have become increasingly evident. Horqin Sandy Land is located near the eastern part of the West Liaohe River. It is the largest sandy land in China and its ecological environment is fragile. Investigating the changes in vegetation phenology in these sandy areas and determining the relationship between vegetation phenology and meteorological factors are of great importance for predicting the impacts of future climate change and understanding the response mechanisms of ecosystems. In this study, we used the time series of the Normalized Difference Vegetation Index (NDVI) from 2000 to 2021 and extracted the vegetation phenology in the Horqin Sandy Land using high-order curve fitting methods, including the start date of the growing season (SOS), the end date of the growing season (EOS), and the length of the growing season (LOS). We analyzed their temporal variation and used partial correlation analysis to determine their relationship with meteorological factors (temperature and precipitation). In addition, we compared the phenology and microclimate of forest and grassland within the study area. In the Horqin Sandy Land, the vegetation SOS was concentrated between the 115th and 150th day, the EOS was concentrated between the 260th and 305th day, and the LOS ranged from 125 to 190 days. Over the past 22 years, the SOS, EOS, and LOS of vegetation in the Horqin Sandy Land showed trends of delay, shift, and extension, with rates of change of 0.82 d/10a, 5.82 d/10a, and 5.00 d/10a, respectively. The start date of the growing season in the Horqin Sandy Land was mainly influenced by precipitation in April of the current year, while the end date was mainly influenced by precipitation in August of the current year. Overall, the SOS in the forested areas of the Horqin Sandy Land was slightly later than in the grasslands, but the EOS in the forested areas was significantly later than in the grasslands, resulting in a longer LOS in the forests. In addition, annual precipitation and the rate of precipitation increase were higher in the forested areas than in the grasslands, but soil temperature was higher in the grasslands than in the forests. Vegetation phenology in the Horqin Sandy Land has undergone significant changes, mainly manifested in the delayed end date of the growing season, the extended length of the growing season, and the differences between forest and grassland. This indicates that climate change has indeed affected phenological changes and provides a theoretical basis for subsequent ecological restoration and desertification prevention efforts in the region.