Nondestructive assessment of leaf chemistry and physiology through spectral reflectance measurements may be misleading when changes in trichome density co-occur

Reflectance indices are frequently used for the nondestructive assessment of leaf chemistry, especially pigment content, in environmental or developmental studies. Since reflectance spectra are influenced by trichome density, and trichome density displays a considerable phenotypic plasticity, we asked whether this structural parameter could be a source of variation in the values of the most commonly used indices. Trichome density was manipulated in detached leaves of three species having either peltate (Olea europaea and Elaeagnus angustifolius) or tubular (Populus alba) trichomes by successive removal of hairs. After each dehairing step, trichome density was determined by light or scanning electron microscopy and reflectance spectra were obtained with a diode-array spectrometer. Although species-specific differences were evident, most of the indices were considerably affected even at low trichome densities. In general, the less-affected indices were those using wavebands within the visible spectral region. The index that could be safely used even at very high hair densities in all species was the red edge index (lambda(RE)) for chlorophyll. The results indicate that changes in reflectance indices should be interpreted cautiously when concurrent changes in trichome density are suspected. In this case, the red edge for chlorophyll content may be the index of choice.