Rainfall is one of the primary sources of chemical inputs in forest ecosystems, and the basis of forest nutrient cycling. Mixed evergreen and deciduous broadleaved forests are currently one of the most threatened ecosystems due to their sensitivity to anthropogenic climate change. As such, understanding the hydrochemical fluxes of these systems is critical for managing their dynamics in the future. We investigate the chemistry of bulk precipitation, stemflow and throughfall in a mixed evergreen and deciduous broadleaved forest in the Shennongjia region of Central China. Mean nutrient concentrations in throughfall and stemflow were higher than in bulk precipitation. Stemflow ion fluxes from deciduous tree species were greater than those for evergreen tree species because of the differences in bark morphology and branch architecture. Throughfall and stemflow chemistry fluctuated dramatically over the growing season. Nitrate nitrogen and ammonium nitrogen were retained, while other elements and compounds were washed off or leached via throughfall and stemflow pathways. Our findings will facilitate a greater understanding of nutrient balance in canopy water fluxes.

Interception loss accounts for a substantial portion of incident precipitation and evapotranspiration in forest ecosystems. Hence, identifying its magnitude is crucial for our understanding of biogeochemical cycling and related hydrological processes. In this study, gross rainfall partitioning into interception loss, throughfall and stemflow were measured and modelled using the revised Gash model for a mixed evergreen and deciduous broadleaved forest over the 2014 growing season. Field survey results revealed that interception loss accounted for 14.3% of gross rainfall, while understory rainfall was 84.8% throughfall and 0.9% stemflow. The revised Gash model produced a fairly good agreement between observed and estimated rainfall partitioning. The model underestimated interception loss by only 6.6%, while throughfall and stemflow were also slightly misestimated. Hence the interception loss predictions from the model were robust and reliable for this mixed evergreen and deciduous broadleaved forest. As quantified by the model, the vast majority of interception loss occurred as evaporation from the canopy under saturated conditions: 54.9% evaporated during rainfall events, and 38.3% after rainfall ceased. The sensitivity analysis indicated that predictions from the revised Gash model were most affected by changes in canopy storage capacity (S), followed by the mean evaporation rate (Ē) during rainfall events, the mean rainfall rate ( trueR¯) and last canopy cover (c). Model predictions were least sensitive to trunk parameters (St and pt). Copyright © 2016 John Wiley & Sons, Ltd.