Font Size: a A A

A PORTABLE MULTICHANNEL FLUORESCENCE SPECTROPHOTOMETER FOR CHARACTERIZATION OF MARINE PHYTOPLANKTON POPULATIONS (PATTERN RECOGNITION, ARC LAMP STABILIZATION, PHOTODIODE ARRAY, GEORGIA, CALIFORNIA, GULF OF MEXICO)

Posted on:1986-10-24Degree:Ph.DType:Dissertation
University:Texas A&M UniversityCandidate:OLDHAM, PHILIP BRYANFull Text:PDF
GTID:1478390017460655Subject:Chemistry
Abstract/Summary:
The rapid and sensitive monitoring of natural phytoplankton populations is commonly performed by detection of in vivo fluorescence. However, the non-scanning fluorescence spectrometers that are commonly used cannot easily acquire multiwavelength spectra. For this reason, a portable, multichannel fluorescence spectrophotometer (PMFS) is developed for the in situ characterization of marine phytoplankton populations. The PMFS combines the characteristics of rapid scanning and multichannel detection with sensitivity comparable to commercially available fluorescence spectrophotometers. Data is acquired by the PMFS in a two-dimensional format called an excitation-emission matrix (EEM). To ensure the acquisition of reproducible data, a novel method of arc lamp stabilization is described. This method inhibits arc wander by adding a small (< 2v) alternating current (AC) signal to the direct current (DC) power supply.;The effectiveness of the PMFS and the pattern recognition technique is illustrated by the analysis of natural phytoplankton populations at three different locations: San Diego Bay, California; Gulf of Mexico; coastal area near Skidaway Island, Georgia. Data from these locations are favorably compared to standard EEMs acquired in the laboratory.;The two-dimensional nature of the EEM in conjunction with the pigment systems specific to certain classes of algae make the EEM an effective "fingerprint" for the characterization of marine phytoplankton. A method of Fourier transform pattern recognition is employed to objectively correlate spectral similarity with sample taxonomy. More than 30 different species of marine algae from 7 different classes are examined to determine pattern recognition accuracy. The effects of cell physiology and binary mixtures of species on pattern recognition accuracy are also described.
Keywords/Search Tags:Pattern recognition, Phytoplankton populations, Fluorescence, Multichannel, Arc, Characterization, PMFS
Related items