Objective This study aims to explore the root morphological characteristics， mechanical properties， and their relationships in herbaceous plants at different growth stages. Methods Root samples of Setaria viridis （fibrous root system） and Medicago sativa （tap root system） grown on purple soil slope land for 60， 90， 120， and 150 days were collected for root morphology scanning， single-root tensile tests， and whole-plant pull-out tests. Results 1） As growth time increased， the tensile and pull-out strength of the aforementioned herbaceous plants increased initially and then decreased. At the same growth time， the tensile and pull-out strengths of the fibrous root system were 1.21-2.89 times and 5.34-6.01 times greater than those of the tap root system， respectively. 2） The tensile strength showed a power function relationship with root diameter， and the pull-out strength showed a power function relationship with the diameter at the pull-out position. As growth time increased， the absolute values of the exponents in the function equation of the tensile strength showed an increasing trend. For the pull-out strength， the fibrous root system showed an increasing trend， while the tap root system showed a decreasing trend. 3） Random forest analysis showed that with the increase of growth time， the main influencing factors of tensile and pull-out strength of the fibrous root system changed from multiple root morphological indicators to a single indicator （root diameter）， while the tap root system showed the opposite trend. Conclusion The research findings provide a theoretical basis for the efficient utilization and scientific evaluation of biological measures in soil and water loss control in purple soil regions.

Objective This study systematically characterizes the spatial correlation network of new-type urbanization and examines its impact on territorial spatial utilization efficiency， aiming to provide theoretical support and references for improving new-type urbanization policies and promoting the efficient utilization of territorial space. Methods Based on panel data from 21 prefecture-level cities in Guangdong Province from 2005 to 2023， social network analysis was used to characterize the spatial correlation network of new-type urbanization. A super-efficiency model （SBM） was applied to measure territorial spatial utilization efficiency， and a panel data-based fixed effects model was constructed to empirically validate the impact of the spatial correlation network of new-type urbanization on territorial spatial utilization efficiency. Results 1） The spatial correlation network of new-type urbanization in Guangdong Province formed a core-periphery stratification structure， with the Pearl River Delta as the core area and the remaining regions as peripheral areas. 2） The network relevance score remained at 1 throughout the study period， and no island nodes were observed. Network density first increased and then decreased， but overall remained at a relatively low level. Network hierarchy showed a fluctuating declining trend， indicating a transition from a hierarchical structure to a flatter one. Network efficiency first increased and then gradually decreased， while network stability improved. 3） The degree centrality， closeness centrality， and betweenness centrality of Guangzhou， Shenzhen， Foshan， and Zhuhai remained at relatively high levels. 4） The spatial correlation network of new-type urbanization helped improve territorial spatial utilization efficiency. However， this enhancement effect was more pronounced in the Pearl River Delta region， plain areas， and in the dimension of urban spatial utilization efficiency. Conclusion It is necessary to internalize network-based thinking to guide the transformation of new-type urbanization and to encourage cities to formulate differentiated development strategies based on their own resource endowments and functional positions within the spatial correlation network of new-type urbanization， thereby strengthening the role of the network in promoting territorial space efficient utilization.