植物個体間競争過程のモデリングと北方林への適用

• Main Authors: 中河, 嘉明, 横沢, 正幸
• Format: Article in Journal/Newspaper
• Language: Japanese
• Published: 北海道大学低温科学研究所
• Subjects: 400; Sakhalin
• Online Access: http://hdl.handle.net/2115/59080; https://doi.org/10.14943/lowtemsci.73.93

植物個体間競争過程のモデルに関する過去の研究例をレビューした結果,サイズ,個体間距離,競争のサイズ非対称性の関数である競争関数（CKF）は経験的に様々に表現されていた.我々は,北海道のトドマツ林の毎木データに基づいて,パラメータのベイズ推定およびモデル選択を行い新たなCKFを導出した.その関数型は個体間距離の－3.42乗に比例するものであり,従来,広く用いられてきた距離の－2乗に比例するものとは明らかに異なっていた.このようなデータに基づいた競争過程のモデルは,植物個体の空間パターンや多種共存機構の解明ならびに定量的予測に有効である. In this review, we summarize research on modeling competition processes among plants;that is, how an individual plant competes with its neighbors. Previous studies have empirically defined various types of competition kernel function (CKF), which is a function of the size difference, distance, and competitive asymmetry between two plants. In this study, we proposed a new CKF and determined the parameters based on census data of Sakhalin fir in Hokkaido using Bayesian inference with the Markov Chain Monte Carlo algorithm (MCMC). Model selection was conducted with Deviation Information Criteria (DIC)to obtain the most plausible model. The selected CKF attenuates with r-3.42, where r denotes the distance between plants, while most previous studies have assumed that CKF decreases with inverse squared distance. The data-oriented CKF derived in this study could be used to elucidate and quantitatively predict temporal changes in spatial pattern and size structure, conspecific clustering, and coexistence in plant communities.