May
2004
, Volume
94
, Number
5
Pages
419
-
431
Authors
A. D.
Wilson
,
D. G.
Lester
,
and
C. S.
Oberle
Affiliations
U.S. Department of Agriculture, Forest Service, Forest Insect and Disease Research, Southern Research Station, Center for Bottomland Hardwoods Research, Southern Hardwoods Laboratory, 432 Stoneville Road, Stoneville, MS 38776
Go to article:
RelatedArticle
Accepted for publication 12 January 2004.
Abstract
ABSTRACT
Conductive polymer analysis, a type of electronic aroma detection technology, was evaluated for its efficacy in the detection, identification, and discrimination of plant-pathogenic microorganisms on standardized media and in diseased plant tissues. The method is based on the acquisition of a diagnostic electronic fingerprint derived from multisensor responses to distinct mixtures of volatile metabolites released into sampled headspace. Protocols were established to apply this technology specifically to plant disease diagnosis. This involved development of standardized cultural methods, new instrument architecture for sampling, sample preparation, prerun procedures, run parameters and schedules, recognition files and libraries, data manipulations, and validation protocols for interpretations of results. The collective output from a 32-sensor array produced unique electronic aroma signature patterns diagnostic of individual microbial species in culture and specific pathogen-host combinations associated with diseased plants. The level of discrimination applied in identifications of unknowns was regulated by confidence level and sensitivity settings during construction of application-specific reference libraries for each category of microbe or microbe-host combination identified. Applications of this technology were demonstrated for the diagnosis of specific disease systems, including bacterial and fungal diseases and decays of trees; for host identifications; and for determinations of levels of infection and relatedness between microbial species. Other potential applications to plant pathology are discussed with some advantages and limitations for each type of diagnostic application.
JnArticleKeywords
Additional keywords:
artificial neural network,
artificial olfaction,
bacterial wetwood,
early detection,
electronic nose,
homeland security,
oak wilt.
Page Content
ArticleCopyright
The American Phytopathological Society, 2004