| 摘要: |
| 利用阔叶图像分析系统WinFolia测定了我国南方阴香、三叉苦、九节和桃金娘4种植物的叶面积、叶周长、叶垂直长、叶水平宽和叶片长5项叶形态指标，通过方差分析和回归分析揭示各项指标的变异状况和指标间的相关性，探寻其叶面积无损测定方法及其可靠性。结果表明，叶面积的变异最大，其余4项指标变异程度接近，变异系数约为叶面积变异系数的1/2。叶面积与其余4项指标均呈极显著正相关（r>0.90，p<0.001）。综合考虑显示，叶垂直长、叶水平宽和叶片长均为准确可靠的叶面积无损测定指标。研究拟合出四种植物叶面积测定的回归方程，对今后开展叶面积测定及相关研究具有重要参考作用。 |
| 关键词: 叶面积 回归方程 叶周长 叶垂直长 叶水平宽 叶片长 |
| DOI： |
| 投稿时间：2009-10-19修订日期：2009-12-25 |
| 基金项目:广东省林业科技计划项目（2008-10）；华南农业大学科技创新项目(L06031) |
|
| Modeling for Nondestructive Leaf Area Measurement of Four Subtropical Plants in China |
| Ke Xian-Dong,Zhou Qing,Su Zhi-Yao |
| (College of Forestry, South China Agricultural University) |
| Abstract: |
| Five leaf morphometric indexes, namely Leaf area, leaf perimeter, leaf vertical length, leaf horizontal width and blade length of four plants including Cinnamomum burmannii, Evodia lepta, Psychotria rubra and Rhodomyrtus tomentosa were determined using WinFolia, an image analysis system of broadleaf. The result of variance analysis and regression showed that leaf area is the most variable index, and the others share a close coefficient of variation, about half of that in leaf area. leaf area has extremely significant correlation with other four characters (r>0.90, p<0.001). Considering in-situ measurement, leaf vertical length, leaf horizontal width and blade length were concluded to be appropriate characters for modeling. Four plants’ optimal regression equations of leaf area were given. Nondestructive leaf area measurement was proposed and would remarkably affect the future research. |
| Key words: leaf area, regression equation, leaf perimeter, leaf vertical length, leaf horizontal width, blade length |
