Remote detection of laser-induced autofluorescence on pure cultures of fungal and bacterial strains and their analysis with multivariate techniques

Remotely sensed laser-induced autofluorescence spectra of pure cultures of fungal strains (Aureobasidium pullulans, Verticillium sp.) and of bacterial strains (Bacillus sp., Pseudomonas sp.) are presented. The trains were isolated from samples collected in a Roman archaeological site (Tropaeum Traiani) near Constanta, Romania. The fluorescence spectra were detected in vivo from a distance of 25 m in the outdoor, using a high spectral resolution fluorescence LIDAR featuring a UV laser (XeCl@308 nm) as an excitation source. All the examined strains, except for the A. pullulans, showed fluorescence features such to allow their characterisation by processing data with multivariate techniques. Both Principal Component Analysis and Cluster Analysis were applied to the data set and compared to discriminate between the examined strains. Results demonstrate the feasibility of fluorescence-based detection and characterisation of fungi and bacteria in the outdoor with a high spectral resolution fluorescence LIDAR. In addition, they show that the proposed processing methods offer a means to discriminate between the fluorescence features due to the investigated samples and that of a fluorescence background of a known spectral shape, as that of the culture medium. This can be exploited for the remote fluorescence mapping of heterotrophic organisms on stone surfaces when the latter show a typical broad fluorescence band.