Analysis of Tree Stand Horizontal Structure Using Random Point Field Methods
O. P. Sekretenko1, P. Y. Grabarnik2
1V. N. Sukachev Institute of Forest, Russian Academy of Sciences, Siberian Branch, Akademgorodok, 50/28, Krasnoyarsk, 660036 Russian Federation 2Institute of Physicochemical and Biological Problems in Soil Science, Russian Academy of Sciences, Institutskaya str., 2, Pushchino, Moscow Oblast, 142290 Russian Federation
Keywords: горизонтальная структура древостоя, случайные точечные поля, пространственная статистика, экологическое моделирование, естественное изреживание насаждения, tree stand horizontal structure, random point fields, spatial statistics, ecological modeling, natural forest thinning
Abstract
This paper uses the model approach to analyze the horizontal structure of forest stands. The main types of models of random point fields and statistical procedures that can be used to analyze spatial patterns of trees of uneven and even-aged stands are described. We show how modern methods of spatial statistics can be used to address one of the objectives of forestry - to clarify the laws of natural thinning of forest stand and the corresponding changes in its spatial structure over time. Studying natural forest thinning, we describe the consecutive stages of modeling: selection of the appropriate parametric model, parameter estimation and generation of point patterns in accordance with the selected model, the selection of statistical functions to describe the horizontal structure of forest stands and testing of statistical hypotheses. We show the possibilities of a specialized software package, spatstat, which is designed to meet the challenges of spatial statistics and provides software support for modern methods of analysis of spatial data. We show that a model of stand thinning that does not consider inter-tree interaction can project the size distribution of the trees properly, but the spatial pattern of the modeled stand is not quite consistent with observed data. Using data of three even-aged pine forest stands of 25, 55, and 90-years old, we demonstrate that the spatial point process models are useful for combining measurements in the forest stands of different ages to study the forest stand natural thinning.
|