Study Region: This study was conducted in the typical agroforestry ecosystem named Linpan settlements (or simply Linpan) in the Chengdu Plain, located in southwestern China. Study Focus: The aims of this study were: (1) monitoring the seasonal variability in rainfall partitioning for tree plots in Linpan comprised of seven evergreen and four deciduous tree species from April 2018 to April 2019; (2) evaluating the effect of rainfall factors and Linpan tree structure on throughfall (TF), stemflow (SF), and canopy interception (I). New Hydrological Insights for the Region: Only a few studies have focused on the rainfall partitioning of trees in agroforestry ecosystems. Our results confirm that the mean TF, SF, and I values of Linpan trees, accounted for 57.4 %, 14.3 %, and 28.3 % of the total rainfall volume, respectively. Evergreen trees showed a lower TF% (56.51 %) than that of deciduous trees (59.16 %), but a slightly higher SF% (14.28 %) and I% (29.21 %) than those of deciduous trees (14.21 % and 26.64 %, respectively). The values of TF, SF, and I were very significantly associated with both gross precipitation (GP) and rainfall intensity. However, evergreen trees TF% was only significantly associated with GP, while evergreen trees SF% was only correlated with rainfall intensity. I % values of evergreen trees and deciduous trees exhibited negative non-linear correlations (p <= 0.05) with rainfall intensity and GP, respectively. Only SF and SF% were significantly negatively correlated with tree structure. DBH and tree height had opposite effects on SF. Both LAI and tree height showed a negative relationship with SF% of Linpan trees. These results have the potential to inform better regulation of runoff and maintaining a sustainable hydrological cycle of agroforestry ecosystem in the Chengdu Plain.
