Tecniche di remote sensing, GIS e geostatistica applicate alla mappatura ed all'epidemiologia del mal dell'inchiostro del castagno
Author(s)
Natili, Giuseppina
Date Issued
January 29, 2010
Type
Doctoral Thesis
Abstract
Remote sensing techniques were integrated with geographic information systems, and
applied to monitoring Ink diseases in central Italy. The multi-spectral high resolution images,
acquired with digital sensor mounted on an aircraft, were used to map the chestnut affected areas by Ink disease. The images definition allows to distinguish diseased plants and foci on the base of crown reflectance signature. A novel stratified method of separating dead tree crowns from bare soil, associated with a maximum likelihood classifier is proposed. The delivered product was a distribution map of Ink disease of chestnut that was validated with ground activities. Accuracy assessment is the procedure used to quantify product quality.
Distribution and gradient analysis of Ink disease caused by Phytophthora cambivora were
studied over a 2 years period in a chestnut forest in central Italy. Ink disease incidence, severity and tree mortality increased during the period over the studied chestnut forest. Disease descriptors and landform datasets were analysed by Geographic Information System software and displayed as a multilayer thematic map. Indices of dispersion and empirical dispersion models were used to study spatial distribution of disease in the investigated area. Semivariograms were used to interpolate spatial data and to map Ink disease. A logical spatial dependence of the studied variables was found.
Ink disease occurred preferentially along natural drainages routes where multicyclic inoculum build
up was likely to occur. A negative relationship was found between Ink disease incidence, severity
and tree mortality with respect to distance from the drainage.
Additional information
Dottorato di ricerca in Protezione delle piante
File(s)![Thumbnail Image]()
Name
gnatili_tesid.pdf
Size
17.42 MB
Format
Adobe PDF
Checksum (MD5)
01581d13d351bcdc8ee47fa983466fd1