Steady-state natural fluorescence

This study employed a manifold analytical approach to the application of multiple excitation and emission wavelengths of steady-state natural fluorescence for the detection of stressed and damaged vegetation. This investigator has attempted to answer questions put forth in terms of the following null hypothesis: Natural fluorescence emission, measured in the steady-state from living, attached or freshly excised needles of eastern white pine (Pinus strobus L.) or leaves of common beans (Phaseolus vulgaris L.), cannot be used in the form of complete spectra or in ratios of discrete wavelengths as indicators of stress from abiotic sources such as ozone or cadmium. The null hypothesis was rejected in the case of white pine subjected to short-term ozone exposure. In experiments conducted in 1992 and 1993, mean red/far-red (R/FR) fluorescence ratios for pine measured 24 hours after fumigation were found to be significantly higher than the R/FR ratio means from pre-fumigation measurements. The fluorescence changes were most likely related to the effects ozone has on the water-splitting molecule and the thylakoid membrane rather than on a reduction in chlorophyll content. The null hypothesis was also rejected for the case of beans grown in cadmium doped soil. Strong correlations were found between R/FR ratio means and a series of cadmium treatments ranging from 0 to 50 mg Kg$\sp{-1}.$ The R/FR ratio means were also strongly correlated with bean chlorophyll content indicating that the cadmium was producing previsual chlorophyll losses. In a second experiment, strong correlation between fluorescence maxima and cadmium treated beans that did not have losses of chlorophyll suggest that fluorescence intensity may be used to detect stressed conditions prior to chlorophyll loss. A technique has been successfully demonstrated for detecting differences in pigment absorption in vivo in short- and long-term water stressed beans using the 3rd derivative of excitation spectra from excitation-emission-matrix data sets