Efficient and accurate prediction of forest-stand productivity is critical for developing forest policy and management plans. Unfortunately, estimations of forest-stand productivity are complex because of interacting influences by diverse biological and environmental factors. Predictions of forest yield can provide a basis for improving the productivity of forest stands. Forest yield can be predicted by estimating the forest land productivity through the use of site index, which is the most common and quantitative method to reduce error.
Site index is the essential tool of forest management for classifying the forest land by its potential productive capacity, and site index equations provide a means to estimate the site quality of forest stands. The productivity of forest stands can estimate by adjusting the following factors: initial planting density, thinning regimes, cutting cycles, and rotation lengths. Currently, the site index equations are based on the relationship between stand age and dominant tree height; however, the use of the current site index equations has some limitations, especially in the unstocked forest. For this reason, an objective of this study was to develop improved site index equations for South Korea, based on various environmental factors, such as soil and topographical variables.
This study was conducted to develop site index equations based on ecoprovinces in South Korea, and estimate productive areas for common tree species in South Korea such as Pinus densiflora for. erecta, P. densiflora, P. rigida, P. thunbergii, Larix leptolepis, P. koraiensis, Cryptomeria japonica, Chamaecyparis obtusa, Quercus acutissima, and Q. mongolica using environmental and climatic factors obtained from both a digital forest site map and a climatic map. Topgraphical characteristics have been used to divide South Korea into five ecoprovinces: Mountain, Southeastern hills, Southwestern hills, Central hills, and Coastal. Using the classified large data set obtained from the maps, a total of 48 environmental factors including 19 climatic variables were regressed into site index by tree species in various ecoprovinces to develop improved site index equations.
Three to seven environmental factors were selected as independent variables in the final site index equations for each tree species and associated ecoprovince. The coefficients of determination for site index equations by species in different ecoprovinces were ranged from 0.29 to 0.89, which allowed estimation of forest stand productivity. The site index equations developed by this study were also verified by three evaluation statistics: model''s estimation bias, model''s precision, and mean square error of measurement. According to the evaluation statistics, the newly developed site index equations fit well to the test data sets and showed relatively low bias and variation. As a result, the site index equations developed by this study demonstrated estimation capability for site quality.
Based on the site index equations, the productive area for each species within each ecoprovince was estimated by applying GIS technique to the digital forest site map and climate map. In addition, the distributions of productive areas for each species in each ecoprovince was compared with its current distribution. The productive areas of each species within ecoprovinces were also estimated based on the site index equations. Furthermore, the various ecoprovinces distributions of productive areas by species were illustrated using GIS-based methods. The results from this study provide critical tools for guiding the forest management practitioners or policy makers in South Korea.