塔柏刺形叶特征与叶面积估算模型

Leaf Characteristics and Leaf Area Estimation Model of the Spiny Leaf in Juniperus chinensis cv.Pyramidalis

【目的】刺形叶小且密生,传统的叶面积测定带来诸多不便,本研究旨在构建快速、准确测量塔柏刺形叶叶面积的模型。【方法】以塔柏刺形叶为材料,采用游标卡尺测量1270片塔柏刺形叶的叶长(LL)、叶基宽(LBW)、最大叶宽(LWmax)以及叶厚(LT),使用Photoshop CS5图像处理软件对叶片的图像进行处理并计算叶面积(LA)。应用SPSS统计软件对叶形态学指标和叶面积进行统计分析,探究刺形叶形态学指标与叶面积的关系,并构建塔柏刺形叶叶面积估算模型。【结果】在塔柏刺形叶5个形态学指标中,LA的变异系数最大(CV=0.301),数值分布在7.307~7.556mm^(2)(95%CI),且LA与LL、LW_(max)呈显著正相关(r=0.858,0.794)。塔柏LA的最优多变量线性回归模型为:Y=-3.879+0.718 X_(1)+5.679 X_(2)-1.177 X_(3)(R^(2)=0.914,RMSE=0.667,AIC=2060.969),其中X_(1)、X_(2)、X_(3)分别为LL、LW_(max)、LBW,预测精度为96.21%。LA的最优单变量模型为基于LL的模型:LA=-1.686+1.003X(R^(2)=0.725,RMSE=1.191,AIC=3238.133),预测精度为91.16%。【结论】本研究为准确估算塔柏刺形叶叶面积提供了简洁的方法,也为研究刺形叶性状指标之间的关系提供了数据基础。

【Objective】Leaf area is an important parameter in evaluating a plant's ability to photosynthesize and how well it can adapt to its surroundings.The tiny and dense growth of spiny leaves causes numerous difficulties in the traditional measurement of leaf area.Therefore,the aim of this study was to construct a rapid and accurate model for measuring the leaf area of the spiny leaves in Juniperus chinensis cv.Pyramidalis.【Method】The leaf length(LL),base width(LBW),maximum leaf width(LW_(max))and leaf thickness(LT)of 1270 leaves from J.chinensis were measured using a vernier caliper,and the leaf area(LA)was determined through leaf image using Photoshop CS5 software.Then the relationships between the morphological indexes and leaf area were analyzed to further construct an estimation model of leaf area for the spiny leaves of J.chinensis using SPSS statistical software.【Result】LA had the greatest coefficient of variation(CV=0.301)among the five morphological indexes and ranged from 7.307–7.556 mm^(2)(95%CI).LA was significantly and positively correlated with LL and LW_(max)(r=0.858,0.794).The optimal multivariate linear regression model for LA was Y=−3.879+0.718 X_(1)+5.679 X_(2)−1.177 X_(3)(R^(2)=0.914,RMSE=0.667,AIC=2060.969),where X_(1),X_(2) and X_(3) were LL,LWmax and LBW respectively,with a prediction accuracy of 96.21%.The optimal univariate model for LA was based on LL,Y=−1.686+1.003 X(R^(2)=0.725,RMSE=1.191,AIC=3238.133),with a prediction accuracy of 91.16%.【Conclusion】This study provides a concise method for accurately estimating the area of spiny leaf of J.chinensis and a basic data for studying the relationship between spiny leaf traits.