Page 39 - TriSociety Virtual Event

concentrations for all cultivars In addition dr y even under controlled conditions it is dificult is feasible using the five registered bands
weight and plant height decreased as to ascertain the nature of multiple pathogen reaching a precision and F1 Score of 97 and
inoculum concentration increased Inocula­ interactions This emphasiZes the challenge 98% respectively
tion of 1­week­old 2­week­old and 3­week­ of understanding pea root rot development
old seedlings of 'CS2000' and ' Westar' was under field conditions where more than three 40. Understanding the effects of land­
also compared by the root­dip method pathogens are ofen present scape structure on the abundance of
Symptoms were more severe when VL was cabbage seedpod weevil and lygus bugs,
inoculated at the later growth stages perhaps 39. Cucumber powdery mildew detection and associated crop damage
because infected seedlings died following using non­georeferenced multispectral P JEGATHEESWARAN H CARCAMO D
early inoculation It seems that the timing and images. JOHNSON S MEERS AND J BYRNE (P.J., D.J.,
method of inoculation with VL will need to be C I FERNANDEZ B LEBLON A HADDADI K J.B.) Department of Geography, University of
carefully considered with respect to the WANG AND J WANG Faculty of Forestry and Lethbridge, 4401 University Drive W., Lethbridge,
ob ectives of a particular study Environmental Management, University of New AB, Canada T1K 6T5 (H.C.) Agriculture and
;
Brunswick, 2 Bailey Dr, Fredericton, NB E3B 5A3, Agri-Food Canada, Lethbridge Research and
38. Disease dynamics of root rot complex Canada; (A.H & K.W.) A&L Canada Laboratories Development Centre, PO Box 3000, Lethbridge,
in field pea Inc. 2136 Jetstream Rd, London, ON N5V 3P5, AB, Canada T1J 4B1; (S.M) Alberta Agriculture
K BISCAGLIA­HORVATH AND S CHATTERTON Canada; (J.W.) Department of Geography, and Forestry, Crop Diversification Centre South,
Lethbridge Research and Development Centre, University of Western Ontario, 1151 Richmond Brooks, AB, T1R 1E6 Canada.
Agriculture and Agri-Food Canada, 5403 1 Ave St, London, ON N5G 2V4, Canada.
S, Lethbridge, AB, Canada; (K.B.) Department of Canola is a ma or oilseed crop grown in the
Biology, University of Lethbridge, 4401 Universi­ Powder y mildew which is caused by the fun­ Canadian Prairies Cabbage seedpod wee vils
ty Dr W, Lethbridge, AB, Canada. gus Podosphaera xanthii, is a ma or disease in and lygus bugs are two ma or insect pests
cucumber greenhouses and may lead to yield damaging canola and afecting yield These
Pea root rot is a disease complex comprised of losses between 30 to 50% of the total pro­ insects feed mainly on reproductive structures
multiple biotic and abiotic stressors result­ duction This study evaluated three geometric of canola such as buds flowers and pods
ing in decay of the root system Pathogens transformations in an image registration meth­ Chemical insecticides are the only available
contributing to root rot in field peas include od applied to non­georeferenced multispec­ option for canola growers to manage these
Aphanomyces euteiches and various Fusar­ tral images acquired at close range over green­ pests Given the expense and potential
ium spp However little is understood of house cucumber plants with a Micasense � harmful impacts of pesticides on the environ­
the nature of the interspecific interactions RedEdge camera The detection of matching ment and biodiversity biological control is an
between pathogens and disease severity points was performed using SURF features attractive option worth researching Biological
as well as environmental influence on these and outliers matching points were removed control has been investigated as a feasible
interactions The ob ective of our study was using the MSAC algorithm For each geometric option to keep these two canola pests below
to simulate multiple infection scenarios in a transformation (afine similarity and pro ec­ economic thresholds Recent research has
greenhouse experiment to (i) characteriZe the tive) we mapped the matching points of the focused on classical biological control using
nature and efect of pathogen interactions blue green red and NIR band images into the eficient parasitoids from Europe Previous
on disease severity and (ii) determine efect red­edge band space and computed the root studies suggest that surrounding landscape
of multiple pathogens on pathogen biomass mean square error (RMSE in pixel) to estimate structure and configuration could afect the
and host coloniZation patterns using quanti­ the accuracy of each transformation Then success of biocontrol of pests by natural
tative PCR (qPCR) 'CDC Meadow' was grown registered band images were used to create enemies Our study aimed to measure the
in soil inoculated with varying concentrations an aligned RGB image and diferent vegetation efects of landscape on the abundance of
of Fusarium avenaceum Fusarium redolens indices Using a pixel­by­pixel approach a fine cabbage seedpod weevil and lygus bugs and
and Aphanomyces euteiches Four weeks Gaussian support vector machine was trained associated crop damage in canola To assess
post seeding disease severity of roots was to classify healthy and infected pixels Our these efects independent canola field sites
rated using a visual scale and fresh biomass results achieved an RMSE of less than 1 pixel were selected and landscape structure and
measurements were taken DNA was isolated with the similarity and afine transformations configuration were documented Digital maps
from the tap and lateral roots for pending and of less than 2 pixels with the pro ective were constructed using ArcGIS Pro From
qPCR analysis Results showed a positive cor­ transformation whatever the band image We the resulting data and maps the proportion
relation between pathogen load and disease determined that the best transformation was of various habitats and non­crop habitats
severity in the case of all three pathogens The the afine transformation because it produc­ were calculated Cabbage seedpod weevil
presence of F. redolens a weak pathogen in es RMSEs of less than 1 pixel and having a and Lygus bugs abundance were estimated
the mixture resulted in a lower disease severity Gaussian distribution The classification using by sweep net sampling Weevil damage was
compared to single inoculations in some the RGB images presented a precision and F1 estimated by counting the exit holes in canola
trials but this was not consistent across all Score of 82 and 88% respectively Neverthe­ pods Preliminar y analysis was conducted
repeated trials These results highlight that less Cucumber powder y mildew detection using regression models The results will elu­

/ 39