生長模式應用於臺灣林木生長研究之探討

Application of Growth Models in Tree Growth in Taiwan

邱祈榮(Chyi-Rong Chiou) ； 鄭景鵬(Ching-Peng Cheng) ； 楊勝驛(Sheng-I Yang)

《中華林學季刊》 46卷4期 (2013/12) Pp. 545-557

摘要

森林測計學中，生長模式廣泛地應用於估測和預測林木的生長表現，本研究整理1945至2013年國內生長模式的研究文獻，共計30篇，包含期刊報告23篇和碩博士論文7篇。依研究主題的階段性差異分為三個時期：1990年以前為生長模式引入期；1991年至2000年為生長模式試驗期；2001年以後為生長模式多元期。三個時期選用的生長模式分析，卡方檢定結果沒有顯著差異，Bertalanffy、Chapman-Richards和Schnute生長模式並沒有因為生長曲線的彈性較高，使研究者特別偏好選用而捨棄其他生長模式；反之，Mitscherlich生長模式和Gompertz生長模式的選用數目在多元時期時，明顯高於前面二個時期，反映出生長模式的選用係考量不同的因素後的結果。生長模式的應用對象分析，紅檜的生長模擬以單木生長為主，林齡範圍集中於30年生以前，而柳杉的生長模擬林分和單木生長皆有，且林齡範圍至90年生。Gompertz和Cbapman-Richards生長模式的模擬效果最好，可做為林木生長模擬的第一優先選用模式，也可依據生長模式與林木性態值對應表，選用最佳的生長模式。生長模式的應用地區分析，生長的資料來源七成以上來自中部地區，南部次之，北部與東部地區的生長模擬相當缺乏，目前該地區林木生長表現資訊仍難以有效地掌握。

關鍵字

生長模式 ； 林木生長 ； 生長模式與林木性態值對應表 ； 柳杉 ； 紅檜

並列摘要

Growth models are widely applied in forest mensuration to estimate and predict the growth of trees and stands. The study aimed to organize, classify and summarize past studies applying growth models in Taiwan from 1945 to 2013 as well as aimed to reinterpret the results in these studies. The results showed: I) We collected total 30 studies, including 23 from journal articles and 7 from dissertations. On the difference topics of these studies, there were 3 distinctive periods: Period 1 (before 1990) was introduction of growth model; Period 2 (1991-2000) was testing of growth model; Period 3 (after 2001) was diverse-topic of growth model. The Chi-square test, p = 0.212, showed that the selected growth models among 3 periods were no significant difference. Bertalanffy, Chapman-Richards and Schnute, which have high flexibility and versatility, were selected less frequently than other simpler models. Moreover, researchers in diverse-topic period preferred to use Mitscherlich and Gompertz to simulate growth. II) Simulation the growth of ＂Chamaecyparis formosensis＂ focused on tree level only and the fitting ages range from 0 to 30 years but simulation the growth of ＂Cryptomeria japonica＂ had both tree and stand level and the fitting ages range from 0 to 90 years. Gompertz and Chapman-Richards, applied in different forest parameters well, could be the first choice on simulation. Another approach to choose models was based on table of forest parameters and its fitness growth model. In addition, 70 percentage of growth data were from central of Taiwan and other were from south. There is no simulation of growth in north and east.

並列關鍵字

Growth model ； Tree growth ； Table of forest parameters and its fitness growth model ； "Cryptomeria japonica" ； "Chamaecyparis formosensis"