© Benaki Phytopathological Institute
Weed control benefit to cost ratio with pendimethalin
63
ed some other herbicides that are registered
and commonly used by local farmers.
For a better evaluation of the actual her-
bicide efficacy and weed control costs, in
all experiments herbicide treatments were
used alone or combined to hand weeding at
the proper time to control escaping weeds.
The labour input was quantified by record-
ing the time and cost required for this sup-
plemental hand-weed control. All treat-
ments in an experiment were replicated
three times and arranged in a randomized
complete block design (RCBD).
With the exception of weed control prac-
tices, the crops in the trials received all oth-
er locally adopted practices of fertilization,
irrigation and plant protection. Crops were
grown to maturity and harvested. For each
treatment the marketable yield (kg/ha) was
recorded and the yield value (€/ha) was es-
timated based on actual farmer’s prices. For
each treatment, the cost of weed control (€/
ha, including value of the herbicide, cost of
application and/or cost of hand weeding,
depending on the treatment) was also es-
timated based on local prices. The benefit/
cost ratio in €/€, for each treatment, was cal-
culated by taking the difference in yield val-
ue between a treatment and the untreat-
ed control and dividing it by the respective
weed control cost. The benefit/cost ratio is
used in this study as a more useful expres-
sion of the weed control efficacy related di-
rectly to farmer’s economical interests.
Since hand weeding is sometimes used
by farmers in these countries as a rescue
method in situations of weed control failure
and more often as a supplemental measure
to situations of erratic herbicidal weed con-
trol, the labour return value for each treat-
ment in these trials was also determined. The
labour return value (in €/hr) for a treatment
was calculated by taking the difference in
yield value of the same treatment, with and
without hand weeding, and dividing it by
the respective time required for hand weed-
ing. This value could then be compared with
the actual hourly payment for hand weed-
ing to assess if a treatment has a positive or
negative effect on the weed control labour
inputs and if hand weeding can be econom-
ically justified.
The herbicide formulations used in the
trials were Stomp Aqua 455 CS (pendimeth-
alin 45,5%w/v), Goal 48 SC (oxyfluorfen 48%
w/v), Dual Gold 96 EC (s-metolachlor 96%
w/v), Sencor 35 WG (metribuzin 35% w/w)
and Cottonex 50 SC (fluometuron 50% w/v).
Herbicide application was made using AZO
sprayers, with a spraying boom of variable
length and nozzle number, delivering 300-
400 L/ha at a pressure of 3 bar.
The trials were conducted, under the
authors’ supervision, by contracted GEP-
certified bodies (Agro Unit and Agrolab) in
Greece and the GEP Test Facility of BASF Ita-
lia Srl in Italy.
The time of hand weeding and the yield
data were subjected to one-way ANOVA
and the statistically different means (denot-
ed with different letters in the same column
of the result tables) were determined using
the Duncan’s test at P=0.05. The SPSS statis-
tical pack PASW Statistics 18 was used for all
statistical analyses.
The processing tomato trial
The processing tomato trial was con-
ducted during the 2010 growing season in
Larissa (Central Greece). Seedlings of the
commercial hybrid Heinz 3402 were trans-
planted in the field on May 6. Experimen-
tal plots of 3.1x10 m (4 rows, 25 plants/row)
were used.
The two most widely applied pre-trans-
plant herbicides in transplanted processing
tomatoes in Greece, namely pendimethal-
in and s-metolachlor, were used as the ba-
si
c
treatments along with an untreated con-
trol. Pendimethalin was applied at the rate
of 1365 g/ha eight days before tranplanting
and was mechanically incorporated into the
soil as recommended. S-metolachlor was
applied at the rate of 1248 g/ha the same
day as pendimetalin but with no incorpo-
ration. The experimental area was irrigated
two days after tranplanting to activate the
soil applied herbicides.
The other experimental treatments in-
cluded the untreated control, and hand
