WEED MANAGEMENT
Research in 1999
Ken Hough, Director, Research and Market Development
At the same Woodstock location where testing was initiated in 1998 on
corn, the 10-acre field of soybeans was marked and GPS-referenced in 6 X 6 meter grids. Weed contour maps were
developed from weed counts at the intersection (corners) of the grid sections. The predominant weeds in 1999 were
lambs-quarters, field horsetail and spiny annual sowthistle.
The field was divided into 16 plots, randomly selected for one of four different spraying treatments, with four
replications each. Each of these plots was subdivided into 136 “decision units” of 3 X 5 meters (Figure A). The
four weed management treatments were:
Spraying was done with the Direct Injection Sprayer developed at the
University of Guelph School of Engineering. For site-specific treatments, decisions on whether to spray or not
within each 3 X 5 m “decision unit” was based on whether the targeted weed species were above the threshold density.
In the broadcast treatment, individual decision units were not assessed and the whole plot was sprayed with the
same herbicide treatment.
Relative to the broadcast treatment,
reduction in herbicide use varied among site-specific treatments, but was as much as 59 per cent in treatment 2
(results depended on distribution of weeds and the number of decision units sprayed within each treatment plot).
There was no difference in weed control efficacy achieved with any of the site-specific treatments (1 to 3) compared
to the conventional broadcast application in 1999. The site specific system, including the direct injection sprayer,
weed mapping, and computer directed spatially-variable application in the field, performed satisfactorily, and
provided significant reductions in herbicide use.
Comparing 1998 and 1999 weed maps shows that the location and shape of field horsetail patches were similar. Although
sowthistle was more widespread in 1998 than 1999, similarities existed in the maps of both years.
For further testing of site-specific herbicide applications, two other fields near Guelph were evaluated with scouting
and weed mapping. However, weed populations were found to be uniformly above the threshold levels, demonstrating
that site-specific application may not always be a useful alternative to broadcast application. The researchers
intend to continue their weed mapping in these fields in the future, to compare annual data on weed patchiness
and to explore the possibility of combining pre-emergence broadcast application with site-specific post-emergence
treatment. They also hope to obtain multi-year data on other weeds and other fields as the study progresses.
The species, density and heterogeneity (unevenness) of distribution of weeds within a spray-grid decision-unit
determines whether application is necessary. (Width of decision units generally matches sprayer width). Studies
to refine the decision process (after weed detection, before spraying) were undertaken in 1999 in soybeans at Woodstock
to determine how crop yield, harvest moisture, harvesting ease and crop dockage levels are affected by the heterogeneity
of weed patch distribution among decision units. Further studies are planned for 2000.
The researchers explored the use of aerial photographs for determining weed patches in the experimental sites,
but were unable to obtain satisfactory photographs, due to limited flying time availability and delayed emergence
from hot, dry weather. However, adequate aerial photographs were obtained from another field where the precision
sprayer was calibrated by spraying alternate patches of rye with Gramoxone. Further investigation of remote sensing
will be conducted in 2000.
Highlights of Ridgetown Research on Problem Weed
Control in Corn
(Excerpts from report submitted to OCPA by Peter Sikkema and Bill
Deen)
Glyphosate and amitrole both provided excellent pre-plant burndown of biennial wormwood. A number of post-emergence
treatments tested in corn provided good to excellent control of biennial wormwood, with dicamba-based treatments
tending to provide the highest control. Atrazine, glyphosate, and glufosinate provided only fair control.
All of the herbicides tested (Dual II Magnum, Primextra II Magnum, Prowl, BASF experimental applied pre-emergence;
Elim, Accent, DuPont experimental applied post) provided very good control of large crabgrass in corn.
Post-emergence application of Accent, Elim and Ultim gave excellent fall panicum control in corn (as did a pre-emergence
application of an experimental herbicide + atrazine + Prowl).
Dicamba or dicamba/atrazine provided good control of horsenettle at only one of three locations in corn, while
other treatments were ineffective.
Swamp smartweed was not effectively controlled by any of the pre-emergence or post-emergence treatments tested
in corn.
In sweet corn, excellent control of proso millet was obtained from post-emergence application of Ultim plus dicamba,
or two applications of glufosinate. Where proso millet population was less than 100 plants per square metre, post-emergence
application of Ultim, Elim or Accent all provided excellent control.
Highlights of Guelph Weed Management Trials in
Corn
(Excerpts from University of Guelph Plant Agriculture Department
“Weed Research 1999” report)
Roundup Ready corn exhibited complete tolerance to Roundup Transorb applied between the three and nine-leaf-stage.
Sequential applications of Roundup Transorb at the three and 7-8 leaf stages of corn, or tank-mix combinations
of Roundup with residual herbicides, were required to provide mid-season control of annual weeds.
Liberty Link corn also exhibited
excellent tolerance to Liberty herbicide applied sequentially at three and 7-8 leaf stages of corn or as a single
application at either of these timings. No-till Liberty Link (LL) corn grown in 30 inch rows yielded 23 per cent
higher than LL corn in 15 inch rows. Populations of 40,000 plants/acre yielded 24 per cent more than LL corn at
28,000 plants/acre. Averaged across all Liberty treatments, weed biomass was reduced by an average of 45 per cent
in corn grown at high populations, and was similar in wide and narrow rows. Single or sequential applications of
Liberty gave similar weed control, which was greater than the Dual + Banvel control. Yields with a single or sequential
Liberty treatment, or with Dual + Banvel were similar, and greater than the weedy check.
Three-seeded mercury was controlled by Converge applied pre-emergence, or post-emergence treatments of either Roundup
Transorb or Peak + Banvel II.
Quackgrass control in IR corn was good with Patriot, but was improved when tank-mixed with Accent or Summit. Mid-season
control of quackgrass was excellent with combinations of Accent or Ultim plus Distinct or Banvel II. Converge,
when tank-mixed with Roundup Transorb, provided excellent crop safety and control of quackgrass and annual weeds
when applied pre-plant in no-till corn.
Several weed trials were conducted to test weed control efficacy and crop tolerance of new herbicide compounds,
new tank mixtures, fertilizer carriers, adjuvants and application strategies.
1