A comparison of spectral reflectance indices in response to water: A case study of Quercus aliena var. acuteserrata

Aims Using leaf spectral reflectance to detect plant status in real time and non-destructively is a new method of forest drought assessment, but each spectral index possesses considerable moisture sensitivity. Therefore, determining moisture index applicable to tree leaf and its sensitive spectral index are both very important. Methods This study selected Quercus aliena var. acuteserrata leaves in different growth stages and canopy positions as the research object, and measured leaf moisture index and its synchronous reflectance spectral response curve during the dehydration process, explored the relationship between changes of leaf spectral reflectance and water status, compared and evaluated the advantages and disadvantages of correlation between the moisture indices of leaves in different growth stages and space positions and different spectral reflectance indices. Important findings Results indicated: (1) The variability of relative water content (RWC) and equivalent water thickness (EWT) in different growth stages and canopy positions was smaller than specific leaf water content (SWC) and leaf moisture percentage on fresh quality (LMP) as measured by the four different moisture indices. RWC and EWT could steadily characterize the holistic water status of trees, and they had greater spectral sensitivity. Therefore, they were suitable for application in remote sensing detection. (2) Spectral reflectance difference analysis and spectral reflectance sensitivity analysis showed that the leaf spectral sensitivity is strongly influenced by growth stage. In short wave infrared region, spectral reflectance of mature leaves changed slightly in the initial stage of dehydration stress, but new expended leaves showed obvious spectral differences during the dehydration process. (3) Through the correlation analysis between 15 different spectral indices and moisture indices, we found that water index (WI)-RWC and double difference index (DD_{n (1530,525)})-EWT has higher correlation. The fitted relations of WI-RWC are greatly influenced by leaf growth stage and canopy position, while those of DD_n(1530,525)- EWT are more stable.