A commercial electronic nose (e-nose) was used to discriminate healthy grapevine tissue from crown galled ones. Samples were collected from outdoor trials and from a nursery. Experimental samples were obtained by grapevine inoculation with Agrobacterium vitis suspensions: samples were collected after 9 months (young rootstocks) and 4 years (aged rootstocks). Field samples (1-year old shoots) were provided by Emilia Romagna Phytosanitary Service (Italy). E-nose analysis was carried out on rootstock and shoot segments. Samples were distributed in the principal component analysis score plot in two areas: on the left, healthy samples and those with small tumors were grouped, whereas high disease severity samples were localized on the right. Linear discriminant analysis correctly classified the samples with a total recognition percentage of 79%. The data obtained allowed building a database in which different sizes, organs, and disease severity of samples are taken into consideration.
Biondi, E., Blasioli, S., Fantini, M., Braschi, I., Bertaccini, A. (2015). Grapevine crown gall detection by electronic nose.
Grapevine crown gall detection by electronic nose
BIONDI, ENRICO;BLASIOLI, SONIA;BRASCHI, ILARIA;BERTACCINI, ASSUNTA
2015
Abstract
A commercial electronic nose (e-nose) was used to discriminate healthy grapevine tissue from crown galled ones. Samples were collected from outdoor trials and from a nursery. Experimental samples were obtained by grapevine inoculation with Agrobacterium vitis suspensions: samples were collected after 9 months (young rootstocks) and 4 years (aged rootstocks). Field samples (1-year old shoots) were provided by Emilia Romagna Phytosanitary Service (Italy). E-nose analysis was carried out on rootstock and shoot segments. Samples were distributed in the principal component analysis score plot in two areas: on the left, healthy samples and those with small tumors were grouped, whereas high disease severity samples were localized on the right. Linear discriminant analysis correctly classified the samples with a total recognition percentage of 79%. The data obtained allowed building a database in which different sizes, organs, and disease severity of samples are taken into consideration.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.