Uncertainties in measurements of leaf optical properties are small compared to the biological variation within and between individuals of European beech
Uncertainties in measurements of leaf optical properties are small compared to the biological variation within and between individuals of European beech
复制标题
DOI：
10.1016/j.rse.2021.112601
复制
发表时间：
2021-10
影响因子：
13.5
通讯作者：
F. Petibon;E. Czyż;Giulia Ghielmetti;A. Hueni;M. Kneubühler;M. Schaepman;M. Schuman
中科院分区：
文献类型：
--
作者：
F. Petibon;E. Czyż;Giulia Ghielmetti;A. Hueni;M. Kneubühler;M. Schaepman;M. Schuman
利用光的反射和散射特性测量叶片光学特性（LOP）可以连续、时间分辨和快速地表征许多物种的性状，包括水分状况、化学成分和叶片结构。个体表现的性状值的变化是生物和环境变化共同作用的结果。人们日益认识到，这种物种特征变异是生物多样性和生态系统特性的驱动因素和反应。然而，很少有人做了全面表征或监测这种变化，使用叶反射率，重点是更经常的物种平均值。此外，虽然存在各种各样的平台和协议来估计叶片反射率，但既没有标准方法，也没有处理测量不确定性的最佳实践。在这项研究中，我们调查什么水平的不确定性可以接受时，测量叶片反射率，同时确保在几个层面上检测物种性状变异：在个人，随着时间的推移，个人之间，种群之间。作为研究物种，我们使用了一种经济和生态上重要的欧洲优势树种，名为Fagussylvatica。我们首先使用织物作为标准材料，通过误差传播来量化与叶夹（0.0001至0.4反射率单位）和积分球测量（0.0001至0.01反射率单位）相关的测量不确定性。然后，我们量化光谱分辨的变化，从F的反射率。森林的叶子我们发现，测量不确定性与叶反射率，估计使用现场分光辐射计与附加的叶夹，平均代表一个单一的个人内的光谱变化的一小部分（2.7 ± 1.7%），在一个生长季节采样，或个人之间的采样超过一个星期（1.5 ± 1.3%或3.4 ± 1.7%，分别）在一组monitoredF。位于瑞士和法国森林中的森林。在所有森林中，个体之间的光谱变化超过了测量时单个个体的光谱变化。然而，随着时间的推移，对不同冠层位置处个体内变化的测量表明，抽样设计（例如，标准化抽样和样本大小）强烈影响我们测量个体间变异的能力。我们建议最佳实践方法对一个标准化的协议，允许严格量化的物种性状变异，利用叶片反射率。
The measurement of leaf optical properties (LOP) using reflectance and scattering properties of light allows a continuous, time-resolved, and rapid characterization of many species traits including water status, chemical composition, and leaf structure. Variation in trait values expressed by individuals result from a combination of biological and environmental variations. Such species trait variations are increasingly recognized as drivers and responses of biodiversity and ecosystem properties. However, little has been done to comprehensively characterize or monitor such variation using leaf reflectance, where emphasis is more often on species average values. Furthermore, although a variety of platforms and protocols exist for the estimation of leaf reflectance, there is neither a standard method, nor a best practise of treating measurement uncertainty which has yet been collectively adopted. In this study, we investigate what level of uncertainty can be accepted when measuring leaf reflectance while ensuring the detection of species trait variation at several levels: within individuals, over time, between individuals, and between populations. As a study species, we use an economically and ecologically important dominant European tree species, namelyFagus sylvatica. We first use fabrics as standard material to quantify measurement uncertainties associated with leaf clip (0.0001 to 0.4 reflectance units) and integrating sphere measurements (0.0001 to 0.01 reflectance units) via error propagation. We then quantify spectrally resolved variation in reflectance fromF. sylvaticaleaves. We show that the measurement uncertainty associated with leaf reflectance, estimated using a field spectroradiometer with attached leaf clip, represents on average a small portion of the spectral variation within a single individual sampled over one growing season (2.7 ± 1.7%), or between individuals sampled over one week (1.5 ± 1.3% or 3.4 ± 1.7%, respectively) in a set of monitoredF. sylvaticatrees located in Swiss and French forests. In all forests, the spectral variation between individuals exceeded the spectral variation of a single individual at the time of the measurement. However, measurements of variation within individuals at different canopy positions over time indicate that sampling design (e.g., standardized sampling, and sample size) strongly impacts our ability to measure between-individual variation. We suggest best practice approaches toward a standardized protocol to allow for rigorous quantification of species trait variation using leaf reflectance.