Estimation of throughfall with changing stand structures for Japanese cypress and cedar plantations

Abstract

Throughfall (TF) is a critical component of the hydrological and biogeochemical cycles and is greatly influenced by stand structures in forested watersheds. Previous studies have examined the relationships between TF and stand structures for different species and regions. However, there remains acknowledged difficulty in estimation of TF with changing stand structures due to forest management (e.g., thinning) for specific species and regions. This study conducted intensive thinning with 43–50% stem removal at seven experimental plots with various structures (e.g., stand density, canopy cover and basal area) across Japan that were covered by Japanese cypress (Chamaecyparis obtusa Endl.) and cedar (Cryptomeria japonica D. Don) plantations. TF was measured by a set of 20 tipping-bucket rain gauges placed in a lattice-like pattern within the plots before and after thinning. Results showed that during the study periods, thinning caused an increase in the mean TF rate from 63.2±7.0% to 75.4±6.0%. The TF rate was significantly negatively correlated with stand density, canopy cover, and basal area on a stand scale. Additionally, the slope (aTF) of event-based linear TF-Pg (gross rainfall) equation was significantly higher in post-thinning (0.782±0.077) than that in pre-thinning (0.663±0.078), whereas it’s intercept (bTF) showed no significant difference (−0.707±0.810mm in pre-thinning and −0.441±0.607mm in post-thinning). The aTF significantly decreased with increasing stand density and canopy cover, whereas the bTF was not significantly correlated with any forest-structure variables. Further, the TF rate and aTF were estimated, respectively, by their related stand-structural variables in multiple regression models with high determination coefficients and moderate relative errors. The bTF was assumed to be a constant and its mean value obtained from all the experimental plots in each (pre- and post-thinning) period. Thus, the models of stand-scale TF rate and event-based TF amount were developed with input of commonly forest inventories. These models were validated by using the dataset of this study and earlier publications for Japanese cedar and cypress plantations, and showed good fit between the estimated and measured values. These models are practical tools that can be readily used for assessing the changes in TF with changing stand structures at both stand and event-based scales, and have also potential implications in evaluating the spatial TF patterns at catchment scale and exploiting similar models in other species and regions.