USGS Science for a Changing World USGS Science for a Changing World
Leetown Science Center
Leetown Science Center Leetown Science Center
Leetown Science Center Welcome Leetown Science Center About LSC Leetown Science Center Research Leetown Science Center Resources Leetown Science Center
Leetown Science Center Leetown Science Center Leetown Science Center Leetown Science Center



TWENTY-SECOND ANNUAL EASTERN FISH HEALTH WORKSHOP


ROYAL PAVILLION RESORT, ATLANTIC BEACH, NC
MARCH 18 - 20, 1997


An Evaluation of the Use of Biolog GN MicroplateTM Reactions in Constructing Taxonomic Trees for Classification of Bacterial Fish Pathogens.

Richard A. Robohm and Steven C. Pitchford.

Northeast Fisheries Science Center, NMFS, NOAA, Milford CT 06460.



Numerical taxonomy based on 40 or more phenotypic traits is reliable and widely used to classify bacteria. In earlier work, we found that the best taxonomic tree for bacterial fish pathogens was produced when we analyzed phenotypic traits using the Kulczynski no. 2 (K2) association coefficient in combination with the Complete Linkage (CL) clustering method. However, labor, cost, and time needed to generate phenotypic data using conventional biochemical tube-tests caused us to explore the 96-well microplate system developed by Biolog, Inc., (Hayward, CA) as a more rapid and convenient procedure. Microplates in this system contain 95 carbon sources and redox chemistry to indicate respiration when the carbon sources are utilized. We selected 22 fish pathogens from the Biolog data-base for which we had corresponding tube-test data. When we applied the K2 association coefficient and CL clustering technique to our tube-test data on the 22 isolates, we produced an accurate taxonomic tree consistent with known relationships among the pathogens. When we applied the same methods to Biolog GN Microplate data, the resulting tree had unacceptable splits among related species. The Biolog system can also construct a separate "user" data-base. Cluster analyses and dendrograms can then be generated using the algorithms built into the Biolog system. When microplate reactions for the 22 fish pathogens were analyzed in the "user" data-base, relationships among some of the bacteria were consistent; however, Aeromonas salmonicida subspecies were split and attached to inappropriate groups. Although the Biolog data-base does not contain all known bacterial fish pathogens, in our limited experience the system appears to accurately match unknown species with their counterparts which are in the data base. Nevertheless, identification of neighboring species or genera often appears to be inaccurate. Also, as detailed above, the Biolog microwell plate reactions could not be used to construct reliable taxonomic trees either using the system's algorithms or other methods.

Return to 22nd Annual Eastern Fish Health Workshop
Return to Leetown Science Center Home Page



U.S. Department of the Interior || U.S. Geological Survey
11700 Leetown Road, Kearneysville, WV 25430, USA
URL: http://www.lsc.usgs.gov
Maintainer: lsc_webmaster@usgs.gov
Last Modified: October 8, 2002 dwn
Privacy Policy and Disclaimers || FOIA || Accessibility