Spatial patterns and climatic factors influence the branch xylem anatomical traits of Reaumuria soongarica in the desert region of northwestern China

The xylem is the water transport tissue of vascular plants, and its structure is closely related to the efficiency and safety of water transport. Studying how xylem anatomical traits and related functional traits vary with envi-ronmental gradients is important for understanding plant responses to environmental changes and ecological adaptation strategies. In this study, ten populations of the dominant shrub Reaumuria soongarica were selected along a west-east transect using an aridity index (AI, the ratio of annual precipitation to annual potential evapotranspiration) ranging from 0.02 to 0.09 in the desert region of northwest China. Based on AI, the sampling sites were divided into arid regions (0.05 <= AI < 0.2) and hyper-arid regions (AI < 0.05). Pearson correlation analysis, linear mixed models, and plant trait networks were used to analyze the spatial variation characteristics and climatic factors of 12 traits in three categories, including xylem anatomical traits, hydraulic functional traits, and mechanical strength. The results were as follows: (1) Vessel diameter contributing 95% hydraulic conduc-tivity (D95) and wood density (WD) significantly differed between arid and hyper-arid regions, whereas the other traits were not significantly different. (2) Mean temperature of the driest quarter was one of the most important climatic factors driving the most trait variation. With the increase in AI and the precipitation of the driest quarter, D95 and the vessel grouping index (Vg) significantly increased, but not WD. The mean temperature of the driest quarter was significantly negatively correlated with D95 and WD but not significantly correlated with Vg. (3) Correlations between most traits were significant, but Vg was not significantly correlated with all other traits. Theoretical hydraulic conductivity was significantly positively correlated with the vulnerability index, indicating a trade-off between the efficiency and safety of the xylem hydraulic system. (4) Coordination among multiple traits differed at the transect and regional climate levels, and the number of edges of the trait network of hyper-arid regions was more than that of arid regions. Moreover, network analyses showed that mean vessel diameter and theoretical hydraulic conductivity are central traits in arid and hyper-arid regions, respectively, because of the number and strength of the relationships with other traits, which could be the important predictor of related hydraulic functions and associations. In conclusion, the xylem anatomical structure of R. soongarica can adapt to changing environments by different combinations of functional traits. Plant trait networks provide a novel approach for exploring the adaptation strategies of desert shrubs along environmental gradients with future climate change.