Page 229 - Plant Canada 2024 Proceeding

PLANT CANADA 2024

[P51] PATHOGENICITY OF VERTICILLIUM LONGISPORUM ISOLATES ON CANOLA AT THE
SEEDLING STAGE. Haitian Yu, Yixiao Wang, Sheau-Fang Hwang, Rudolph Fredua-Agyeman, and
Stephen E. Strelkov. Department of Agricultural, Food and Nutritional Sciences, University of Alberta,
Edmonton, AB T6G 2P5, Canada
Correspondence to: haitian7@ualberta.ca

The fungus Verticillium longisporum is responsible for Verticillium stripe, a soilborne disease affecting
canola (Brassica napus). Enhanced understanding of the pathogenicity of V. longisporum during the
seeding stage could aid in the development of improved disease management methods. In this study, 62
isolates of V. longisporum were collected from infected plant tissue samples across Alberta, Manitoba,
and Saskatchewan. These isolates were then evaluated for their effects on seedling emergence, plant
height and vigor on the canola cv. ‘Westar’ under controlled conditions. Inoculations were conducted at
seeding using low, medium and high concentrations of mycelial inoculum produced on potato dextrose
agar. Seedling emergence was assessed at 7 days post-inoculation (dpi), while plant height and vigor
were evaluated at 14 dpi. All three parameters declined following inoculation with V. longisporum, with
some fungal isolates inducing more severe reductions at higher inoculum concentrations. Correlation
analysis indicated that seedling emergence, plant height, and vigor were positively correlated with one
another, with correlation indices ranging from 0.41 to 0.52. Principal component analysis showed that,
collectively, these parameters explained 80% to 90% of the variation among isolates. There were no
geographic or year effects on the pathogenicity of V. longisporum isolates across different inoculum
concentrations. The work is ongoing, and disease severity will be assessed at 42 dpi. Nonetheless, the
results so far suggest that this fungus can significantly reduce canola emergence and growth at the
seedling stage.

[P52] ESTIMATING SOYBEAN YIELD LOSS TO WEED INTERFERENCE USING EARLY-SEASON
2
3
1 1
1
REMOTE-SENSING TOOLS. RH Gulden , CJ Henry , N Badreldin , and DI Benaragama . Dept. Plant
Science, University of Manitoba, Faculty of Agriculture and Food Sciences, 66 Dafoe Road, Winnipeg,
2
MB, Canada R3T 2N2; Dept. Computer Science, University of Manitoba, Faculty of Science, 75
Chancellors Circle, Winnipeg, MB, Canada R3T 2N2; and Dept. Soil Science, University of Manitoba,
3
Faculty of Agriculture and Food Sciences, 13 Freedman Crescent, Winnipeg, MB, Canada R3T 2N2
Correspondence to: rob.gulden@umanitoba.ca

Weed interference significantly reduces crop yield and blanket herbicide applications to manage these
weeds have resulted in the selection of herbicide resistant (HR) weed biotypes. While the technology for
site specific weed management exists, decision support systems that reduce the selection pressure for
herbicide resistant biotypes by site specific applications of herbicides only where a yield loss threshold is
met are lacking. Development of remote-sensed, site specific yield loss thresholds for data-driven
decision support systems can contribute to more sustainable weed management and herbicide use. In
2023, a soybean additive-series experiment was established with increasing densities of either a
surrogate HR broadleaf (canola) or a surrogate HR grassy weed (corn) sown in alternate rows to soybean
to facilitate image segmentation. All other weeds were managed as needed with glyphosate. Multispectral
digital images were captured using an Unmanned Aerial Vehicle (UAV) throughout the growing season to
generate orthomosaic images of the experiment that were segmented into the ground cover of the
respective crop and weed components for each experimental unit using a thresholding approach followed
by manual correction. Soybean yield loss based on weed density followed the well-established
rectangular hyperbola equation. Interestingly, the relationship between remote-sensed weed ground
cover and soybean yield loss was a much simpler linear relationship for both weeds. The greatest
regression coefficients for a single time point were obtained when yield loss was regressed against weed
ground cover at the 4-leaf stage of soybean (Broadleaf R2 = 0.69, Grass R2 = 0.92) with different slopes
between the two weed types. Inclusion of additional crop and weed ground cover data at other vegetative
soybean developmental stages further improved the fit of the models. Overall, the preliminary results from
this experiment show that early-season remote-sensed weed ground cover data is useful for predicting
yield loss in soybean and shows promise towards using this technology to develop data-based, decision-
support tools for weed management that can contribute to more sustainable crop production in a
changing world.

228