基于Dualex氮平衡指数测量仪的作物叶绿素含量估算模型

作物叶绿素含量的实时、无损及快速的监测，对及时掌握作物的胁迫状况、营养水平及环境适应性，进而对农田管理进行科学指导具有重要的意义。该研究论证是否可以通过Dualex氮平衡指数测量仪构建通用的叶绿素含量估算模型，以期实现叶绿素含量的快速及无损监测和估算。结果表明：1）Dualex估测叶绿素质量分数（Chl-M）和单位面积的叶绿素质量（Chl-S）具有较好的精度（决定系数R2分别为0.77和0.88），与SPAD叶绿素仪的估算模型（R2分别为0.66和0.79）相比，模型精度更高；2）Dualex估测Chl-S的精度明显高于Dualex对Chl-M的估测精度，Dualex与Chl-M的关系需要考虑叶片厚度的影响，而Dualex与Chl-S的线性关系更加一致；3）构建的Chl-S通用模型的R2，均方根误差和标准均方根误差分别为0.88，4.80 mg/dm2和8.33%，模型的精度较高，并且通用模型的数据范围为12.2～105.6 mg/dm2，较大的数据范围适用于冬小麦和玉米各关键生育期Chl-S的估算。该研究为Dualex实现冬小麦和玉米叶绿素含量监测和估算提供校准模型，为及时了解作物养分状况及作物营养诊断提供了参考。

Modelling of crop chlorophyll content based on Dualex

Abstract: Real-time non-destructive and quick estimation of leaf chlorophyll content can provide important information about plant stress, nutritional state and relationships between plants and their environment, and therefore will be of great significance in agriculture field management. In this study, 2 hand-held chlorophyll absorbance meters, Dualex 4 and SPAD-502, for relative chlorophyll content, and standard chemical method by the spectrophotometric for absolute chlorophyll content in wheat and maize were used for analyzing the relationships between relative chlorophyll content and absolute chlorophyll content. The absolute chlorophyll content can be expressed in terms of chlorophyll mass per leaf mass (Chl-M, mg/g) and chlorophyll mass per leaf square (Chl-S, mg/dm2). The objective of this study was to establish chlorophyll content models, Chl-M model and Chl-S model, and evaluate whether these models could be as general models for estimating leaf chlorophyll content. Four field experiments were carried out, including 2 experiments with 3 winter wheat cultivars during the growing season from 2014 to 2015 and 2 experiments with 3 maize cultivars during the growing seasons in 2013 and 2015. Time-course measurements were taken on relative chlorophyll content and absolute chlorophyll content. Linear regression analyses between relative chlorophyll content (Dualex value or SPAD value) and absolute chlorophyll content (Chl-M or Chl-S) were conducted. Three statistical indicators including determination coefficient (R2), root mean square error (RMSE) and normalized root mean square error (nRMSE) were employed to evaluate the performance of each model. The results showed that the performances of Chl-M and Chl-S model using Dualex (R2 values of 0.77 and 0.88, respectively) were better than those using SPAD (R2 values of 0.66 and 0.79, respectively). Chl-S models using Dualex at each growing stage were superior to Chl-M models. The relationship between Chl-M and Dualex values at each growing stage should consider the influence of leaf thickness. A strong relationship between Chl-S and Dualex values was demonstrated, with the R2 value of each Chl-S model ranging from 0.87 to 0.97 and the nRMSE value lower than 10%. The general model of Chl-S demonstrated a high performance, with the R2, RMSE, and nRMSE of 0.88, 4.80 mg/dm2, and 8.33%, respectively. The range of Chl-S was from 12.2 to 105.6 mg/dm2, and could be applied at any growing stage for winter wheat and maize. Therefore, this study proposes a consensus equation for the transformation of Dualex into leaf chlorophyll content of winter wheat and maize, achieves the real-time and non-destructive estimation of chlorophyll content, and provides theoretical basis and technical support for the acquisition of crop nutritional state and the field decision-making.