EPA/600/R-92/020
                     Effects of Glean, a Sulfonylurea Herbicide,
                        on the Reproductive Biology and
                            Fruit Set in Cherry Trees
                          Progress Report to Region 10
                      U.S. Environmental Protection Agency
                                  January 1992
Thomas Pfleeger
John Fletcher
Hilman Ratsch

US EPA ERL-C
200 SW 35th St
Corvallis, OR  97333
The information in this document has been wholly funded  by the U.S. Environmental
Protection Agency.  It has been subject to the agency's peer and administrative review, and
it has been approved  for publication as an EPA document.

Mention  of trade names  or  commercial products does not constitute endorsement or
recommendation for use.

-------
INTRODUCTION

Approximately 85% of the 475 million pounds of pesticides used annually in the U.S. are
herbicides (Waddell and Bower,  1988; USDA, 1990).  These phytotoxic compounds are
applied to an estimated 270 million acres of cropland.  There has always been a concern
for damage to nontarget vegetation brought about by herbicide drift, believed  by some to
be as great as 25 to 50% of the applied chemical (Waddell and Bower, 1988). The concern
for drift-damage has intensified over the last 10 years with the introduction of new classes
of herbicides (sulfonylureas and  imidazolinones) which  are 100 tunes  more toxic than
formerly used compounds such as 2,4-D and atrazine (Beyer et ah, 1987). Approximately
10 sulfonylurea and imidazolinone herbicides are currently being sold and many more will
no doubt be registered in the future. The sulfonylurea compounds alone had over 230 U.S.
patents approved by 1987. (Beyer et ah, 1987).

Local problems in North Dakota,  Colorado, and Washington suggest that the sulfonylurea
herbicides have caused unexpected nontarget damage (Callihan and Lass, 1991; Westra et
ah, 1991; Mink and Howell, 1990). This is after the herbicides have been reviewed in the
registration process for nontarget plant damage (Lewis and Petrie, 1991). The total number
and nature of nontarget plant losses associated with these newly introduced herbicides are
uncertain because there is no tabulation of nontarget plant damage at the  national level.
The current pesticide registration process requires plant testing. However, it is limited to
germination, emergence and vegetative vigor tests (Table 1).  The process only tests a
limited portion of a plants life cycle (Figure 1) and only with a few agricultural species (i.e.,
no herbaceous perennials or woody plants).  The tests only involve germination and early
vegetative growth of a plants life cycle (Figure 1).  The registration process requires no
reproductive or life cycle test data and therefore, nontarget plant reproductive processes
may be  at risk from the new herbicides.

The complaints of orchard growers in south-central Washington are an excellent example
of farmer accusations regarding crop damage due to sulfonylurea drift from dry  land wheat
fields.  Some growers claim that drifting sulfonylurea herbicides  caused flower and fruit
abortion on their cherry, apricot, and plum trees in 1988 and 1990, whereby greater than
80% of their crops were lost. The total dollar loss of all crops in this area due to herbicide
drift has been estimated by local growers to be $40 million.  Although these  allegations
have been investigated by local and federal representatives, including the Washington State
Department of Agriculture and the  U.S. Environmental  Protection  Agency, the issue
remains unresolved for several reasons:  1) No data has ever been published pertaining to
the influence  of  the herbicides in question ('Glean' and 'Harmony') on reproduction
(flowering and/or fruit set) of orchard trees; 2) No plant reproduction data is currently
required by OPP (Office of Pesticide Programs) for herbicide registration (Table 1; Figure
1); and 3) It is conceivable that new  herbicides which  are very toxic to vegetative growth
may be  even more  toxic  towards reproductive  events.   If so, concentrations affecting
reproduction may be below the level of detection by conventional chemical analyses.  Thus,
the claims of orchard  growers in  Washington may be accurate, and if  so, the current
registration process may not be adequate with respect to these new chemicals.

-------
EPA, Corvallis Environmental Research Laboratory undertook this study for Region X to
help in their efforts to understand the issue of non-target plant damage in south-central
Washington from sulfonylurea herbicides. Specifically, the study was designed to determine
if low levels of 'Glean' interfere with the normal reproductive events in mature cherry trees.
This was accomplished by measuring the reproductive response (flower and fruit set) of
individual cherry trees to different exposure regimes in which both the amount and time of
exposure (bud development and expansion) have been measured. This is intended to be
a multi-year study; therefore, it is important to recognize that the results reported herein
are preliminary and must not be misconstrued to be conclusive.

METHODS

Fifteen different twenty year old Royal Anne Cherry trees (Prunus Avium L.) grown and
maintained by the Oregon State University Department of Horticulture, were used for each
spring and summer/fall exposure. The trees are  located at the Lewis-Brown Horticulture
Farm in  the Oregon Willamette Valley  approximately five miles southeast  of Corvallis,
Oregon.  The trees received the standard management practices used on all cherry  trees
under the  care of the Department of Horticulture, including insecticide and fungicide
treatments, pruning, fertilizing  and fruit harvesting (Fisher et al., 1990)

For each exposure, fifteen uniform trees (i.e., similar in height and diameter) were selected
from a row of twenty, in a block of 140 trees.  Six branches on the same side of each tree
were selected for treatment.  The branches were on the perimeter of the canopy and  were
approximately equal distance from the ground.  The  six branches were each randomly
assigned  one of six different treatments;  no treatment (control), water treatment (carrier
control),  0.1, 0.01, 0.002 and 0.001 of the field application  rate of 'Glean'.  The  field
application rate used was  1/3 of an ounce per acre (2.34 x 10~6 Kg/m2). The surfactant
'Unifilm  707' was used at a concentration of .05 percent of the spray solution. The carrier
control not only tested for  carrier effect, but  also the effect of enclosing the  treated
branches overnight.  Observations  of treatment  effects were  recorded both as notes and
photographs.

Bags made of Tyvek' were  placed over the branches, except the control  treatment, to
prevent cross contamination of treatments.  These were supported by a light weight frame
made of  sheet metal and fastened to the branch by tape. The bag and frame covered 90
cm of the branch of which the middle 50 cm was observed. The frames were placed on the
branch the morning of the treatment application.  The bags were placed on the frames late
in the afternoon just prior to spraying the treatments. The branches were sprayed with the
treatments for 15 seconds at 25 Ibs of pressure using a two gallon stainless steel hand pump
sprayer with a pressure gauge and a brass  extension fitted with dual nozzles.  The spray
nozzles consisted of a nylon core and a disc with  an orifice diameter of .61 mm (Spraying
Systems Co., Wheaton, IL). The spray apparatus was calibrated and verified to deliver 60
ml of solution per  spray event.  This procedure  ensured complete coverage of the 50 cm
treatment segment on each branch to  the point  that excess liquid dripped from the
branches. The bags were taped shut immediately following the spray treatment and  were
not removed until the next morning when the bag and the frame were also removed.   The

-------
bags remained on the tree over night to prevent cross contamination by volatilization of the
chemical and to also provide uniform exposure conditions at each stage of application (bud,
flower, fruit) in the event of rain. The treatment and bagging were conducted late in the
day to avoid any potential heat damage to the treated branch.

Fifty centimeters on each branch were treated and monitored.  The boundaries of the
treatment zone were marked on each branch with a permanent marker. The treated area
started 15 cm from the branch tip and extended toward the trunk for 50 cm.  This area was
chosen due  to its high concentration of flower buds. At different times during the spring
the numbers of buds, flowers and fruits  present on the treatment segment of each
designated  branch  were counted.  During  the  course of the study the foliage was
periodically  examined and descriptive assessments recorded.

This investigation has examined herbicide exposure at two different stages of perennial
reproduction (Figure 2).  One set of fifteen trees was exposed in the spring during flower
and fruit set, and a different row of trees was exposed in the late summer and early fall,
a time when the next year's reproductive buds were developing (Tuffs and Marrow, 1925).

Spring Exposure

The treatments were applied  once on each of the fifteen trees. Five randomly selected
trees had the treatments applied during the green tip stage of floral bud expansion (March
15,1991). An additional five trees were treated at the full bloom (April 2) and another five
at the post bloom stage  (April 26).
Cherry fruits were harvested on June 13.  They were harvested three and a half weeks prior
to full ripening to prevent possible loss to birds.  The fruit were bagged, placed on ice and
transported to the laboratory and refrigerated.  The fruit were  counted and weighed to
determine a mean cherry weight per tree for each treatment.  Cherries seeds were split to
determine seed viability. Seeds were considered  nonviable if the  embryo was aborted and
the endosperm was not developed. Nonviable cherries would have fallen off prior to the
normal harvesting date; therefore seed viability was used as the criteria for determining fruit
set.

The response variates (average cherry weight per tree and number of viable cherries) were
analyzed assuming  a randomized  complete block  design  with trees as  blocks  and
concentration levels randomly assigned to branches (experimental  units) with a block (tree).
Prior to the ANOVA and post-ANOVA analyses, the control treatments (carrier-control
and no  treatment) were compared for  each response variable  within each time period
(green tip, full bloom, post bloom) with no significant difference  detected (paired t-test, p
>  0.2).   The  carrier-control was  therefore removed from further consideration.   Post-
ANOVA analyses to determine specific pair-wise differences  was  accomplished using
Tukey's studentized range (HSD) test at an alpha level of 0.05.

Prior to the overall test for treatment effects, the number of viable cherries response was
transformed using the square root transformation  to correct for non-constant variance.

-------
Transformation of average cherry weight was not deemed necessary.

Summer / Fall Exposure

Because the loss of cherry fruit may be the result of injury during bud development in late
summer, and early fall rather than during bud expansion in the spring, another series of
experiments was initiated to investigate the effect 'Glean' has when it is applied during the
differentiation of flower tissues in buds (Figure 2). These experiments were performed in
exactly  the same manner as described for the spring experiments except that the spray
applications occurred on August 15, September 12 and October 10, and a different row of
trees in the same block of cherry trees was used.  The treatments were approximately one
month apart to cover a broad spectrum of bud development. Data from this experiment
(cherry fruit) will not be collected until late spring of 1992.

Pesticide incident reports submitted to the Washington State Department of Agriculture
show that herbicides are being applied throughout the growing season (State of Washington,
1987; 1988); therefore nontarget plants  may be subjected to multiple exposures  from
herbicide drift. Two experiments were set up to determine if repeated exposures to low
levels of 'Glean' had an effect on bud development.  Again, the same methodology was
used with the exception that only the 0.002 and both control treatments were applied.  In
one  experiment,  branches  from five trees  were treated  twice,  one month apart  on
September 26 and October 24. In the other experiment, branches from five trees  were
treated three times, one week apart on September 19, 26, and October 3 and another set
of branches from an additional five trees were similarly treated on October 10, 17 and 24.
Again, data from these  experiments will not be collected until late spring of 1992 when
the cherries are harvested.

RESULTS (Spring exposure 1991)

Cherry  Weight Data

When  'Glean' was  applied  at the green tip stage of bud development, no significant
differences (p  =  0.19) were found between any of the treatments and the controls in mean
cherry weight (Figure 3). In contrast to this, when 'Glean'  was applied at the full bloom
stage, the 0.1 application rate significantly reduced the mean cherry weight (p =  0.002) to
approximately one third of the control weight (Table 2). When the treatment was applied
at the post bloom stage the reduction in mean cherry weight was significant at both the 0.1
and 0.01 treatment levels (p = 0.0001) (Figure 3).  The 0.002 treatment level reduced mean
cherry weight by 17 percent, but was not significantly different from the controls (Table 2).

Seed Viability/Fruit Set

Data on the number of viable cherries  supports the results found  from the mean cherry
weight data. At the green tip stage none of the treatments caused a significant difference
in comparison to the controls (p = 0.68) (Figure 4). Treatments at full bloom resulted in
a significant difference  at the 0.1 concentration (p = 0.001).  However, this result is
confounded because the 0.1 treatment was  not significantly different from the 0.002

-------
treatment at an alpha level of 0.05 despite the fact that there was a 77 percent reduction
in the number of viable cherries between the 0.002 and the 0.1 treatments (Table 2).  At
the post bloom stage there were no viable cherries at either the 0.1 or 0.01 treatments
(Table 2).  Although the 0.002 treatment at the post bloom stage reduced the number of
viable  cherries by 19 percent in comparison to  the  control,  this  was not  significantly
different (alpha = 0.05). This is consistent with the analysis of the mean cherry weight data
(Table 2).

Foliage Response

Chemical application  at  the green  tip stage had  no observable influence on leaf
development or appearance. In contrast, treatments at  either the full bloom or post bloom
stage, when leaves were partially expanded, caused abnormalities in leaf morphology. The
abnormality symptoms consisted of yellowing, cupping and reduced leaf expansion. The
intensity and time at which these symptoms  appeared varied, depending on the stage of
treatment and the concentration of applied chemical. When the 0.1 'Glean' treatment was
applied at the full bloom stage, symptoms were first observed after four weeks; and by  ten
weeks  the  leaves had withered and appeared dead.   None of the other concentrations
applied at the full bloom stage altered leaf development.  At the post bloom stage, the 0.1,
0.01, and 0.002 treatments affected leaf morphology. The 0.1 and 0.01 treatments caused
leaf symptoms to appear  two and four weeks, respectively, after application.  The  0.1
treated leaves appeared dead six weeks following exposure; whereas the 0.01 treated leaves
remained alive, but permanently dwarfed. A few leaves on branches treated with the 0.002
application also showed the abnormal symptoms.

The only visible symptoms on branches treated in the fall were the yellowing of leaves.
This was only apparent on branches receiving the 0.1 treatment on August 15th. No clear
signs of injury were present on the other treatments made on September 12th and October
10th.  The apparent absence of leaf response to  'Glean' in late fall coincided with  the
general yellowing of all leaves as fall progressed.

DISCUSSION

The data suggests that the impact of 'Glean' on mean cherry weight and viability is related
to the time of application, with increasing sensitivity from the bud stage to the post bloom
stage.  Not only are the full bloom and post bloom stages more responsive to 'Glean', but
there is also an increased  sensitivity to lower concentrations of the chemical.  Significant
differences were found at the 0.01 treatment level at the post bloom stage, whereas  the
only significant difference  was at the 0.1 treatment during the full bloom stage.

The increased sensitivity of the cherry trees to 'Glean' was associated with leaf expansion.
At the  green tip stage the leaves were compressed within the swollen vegetative buds, by
full bloom they were beginning  expansion and by post bloom they were one third to one
half expanded. The preliminary results suggest that the  sensitivity of fruit development and
set on cherry trees exposed to 'Glean' is proportional to the extent of leaf expansion. The
more fully expanded the new leaves are, the more  susceptible the developing cherries are
to damage. It is conceivable that the most sensitive stage would be when new leaves are

-------
fully expanded and the fruits are still partially developed. This suggests that concentrations
that produced no significant  effects when applied early in leaf expansion may produce
significant effects when applied at later stages.  Our intention is to examine this possibility
during the spring of 1992.

The positive correlation between leaf expansion and cherry loss (fruit abortion or reduced
weight) to  'Glean' indicates  that the leaf, not the bud or flower, may be the primary
receptor surface for that portion of the applied  chemical which is detrimental  to fruit
development.  We propose that as the young leaves expand, their increased surface area
permits greater absorption of the applied chemical leading to elevated levels of 'Glean' in
the leaves.  The action of 'Glean' may be explained by  two  different hypotheses.  The
chemical may be transported through the phloem to the developing fruit, where it reaches
a concentration which proves to be inhibitory to cell division,  one of the reported modes
of action for 'Glean' (Beyer et al., 1987). A second possibility is that 'Glean' stays in the
leaf where it  prevents newly fixed carbon from entering the translocation  stream, as
reported by Bestman et al. (1990). If that occurred the developing fruits could be subjected
to carbohydrate starvation, leading to reduced  growth and abortion.

The experimental work described in this report covers only a small portion of the time
when cherry buds are undergoing development and expansion.  As shown in Figure 2 bud
development begins in the early summer and continues for many  months  (Tuffs and
Morrow, 1925).  According to pesticide  incident reports submitted  to the State  of
Washington, Department of Agriculture (1987;  1988) numerous spray events occur during
this period of bud development.  In view of the pronounced  response of leafed-trees to
sulfonylurea herbicides during spring exposures, it  became clear that it was very important
to conduct a summer-fall study. Several EPA experiments have  been initiated to investigate
the effects of 'Glean' when applied at different  stages of bud development.  In  all of these
experiments, damage to  cherry fruit  in the form of mean cherry weight decrease and
number of viable cherries will be measured.

CONCLUSIONS

The research described in this report is part of a three-year project.  The final data will be
collected on the fruit harvested in the spring of 1993.  By the conclusion of the project, all
treatments will have been examined twice.  Although we  believe that it is worthwhile to
make tentative conclusions at this time to help focus the remaining  research on  key
questions and issues, we feel strongly that as soon as our  experiments are completed and
the results thoroughly examined, the regulatory  process for new pesticides should be
reexamined.  This should occur no later than 1993.

Tentative conclusions

1. Application of 'Glean' directly  to swollen floral buds in the spring had no detrimental
influence on normal flower and fruit development of cherry trees.

2. Application of 'Glean' at  0.1 of the recommended  field concentration to fully open
flowers and partially expanded leaves (less than one third  expanded) caused 98 percent of

-------
the fruits to abort and the leaves to die.

3. Application of 'Glean' at 0.1 and 0.01 of the recommended field concentration (1/3 of
an ounce per acre) to post bloom fruit and partially expanded leaves (one third to one half
full size) caused 100 percent of the fruits to abort at both concentrations, leaf malformation
at 0.01 and dead  leaves at 0.1.

4. The adverse influence of 'Glean' on cherry fruit development appears to be proportional
to the extent to which new leaves have expanded.

Unanswered Questions

1. Are the responses to 'Glean' observed in  this study during the spring of 1991 typical, or
will they vary with yearly fluctuations in weather conditions?

2. What is the most sensitive stage in bud, flower and fruit development to 'Glean' damage?
Two important stages deserving attention are the developing floral bud during the summer-
fall season and the developing fruit in the spring after leaves are fully expanded.

3. How does the response to a multiple exposure of 'Glean' at low levels over several weeks
compare to the response noted in this study to 0.1 and 0.01 concentration administered at
the post bloom stage?

4. How do the pesticide concentrations used in these experiments compare to  actual drift
scenarios? The goal of the present work is to determine if the sulfonylurea herbicides have
an adverse reproductive effect at low concentrations. If this is established, then an  effort
should be made to correlate drift concentrations to the experimental dosages.

5. What is the mode of action by which 'Glean' reduces cherry fruit growth and promotes
abortion?

FINAL PRODUCT

When this research project is finished, we will have established the influence of low  levels
of 'Glean', a sulfonylurea compound, on the reproduction  (fruit set) of cherry trees. This
information will be valuable in evaluating the effectiveness of OPP's registration process;
specifically, the currently required  plant toxicity  tests which do not have reproductive
evaluations (see Lewis an Petrie, 1991).
ACKNOWLEDGEMENTS

We thank Karl Arne of US EPA Region 10 for making us aware of this problem and, with
Randy Bruins (Region 10), assisting in the garnering of funds to accomplish the work. We
thank Scott Robbins and his staff at the Oregon State University, Lewis Brown Horticulture
Farm for their invaluable help. We also thank Don Stufflebeem for field assistance and
Danny Kugler for statistical assistance.

                                        8

-------
REFERENCES

Bestman, H. D., M.D. Devine, and W.H. Vanden Born.  1990.  Herbicide clorsulfuron
decreases assimilate transport out of treated leaves of field pennycress (Thlaspi arvense L.)
seedlings.  Plant Physiol. 93:1441-1448.

Beyer, E.M., MJ. Duffy, J.V. Hay, and D.D. Schlueter.  1987. Sulfonylurea herbicides,
Chapter 3. In: Herbicides: Chemistry, Degradation, and Mode of Action, Marcel Dekker
Inc.

Callihan, R.H., and L.W. Lass. 1991. Investigating herbicide sensitivity thresholds,  pg. 91-
97. In: Plant tier testing: A workshop to evaluate nontarget plant testing in subdivision J
pesticide guidelines.  Fletcher, J. and H. Ratsch, eds.  EPA/600/9-91/041

Fisher, G., H. Homan, and A. Antonelli. 1990. Pacific northwest insect control handbook.
Cooperative Extension.  Oregon State University. Corvallis,  OR.  322 pg.

Gorsuch, J.   1991.  Difficulties in performing existing tier 1 and 2 test in subdivision J
guidelines,  pg. 48-57.  In: Plant tier testing: A workshop to evaluate nontarget plant testing
in subdivision J pesticide guidelines.  Fletcher, J. and H. Ratsch, eds. EPA/600/9-91/041

Lewis, C, and R. Petrie.  1991. Plant data analysis by ecological effects branch in the office
of pesticides program,  pg. 6-10. In: Plant tier testing: A workshop to evaluate nontarget
plant testing in subdivision  J pesticide guidelines.   Fletcher, J. and  H. Ratsch, eds.
EPA/600/9-91/041

Mink, G.I. and W.E. Howell.  1990.  An evaluation of problems alleged to be caused by
herbicide drift into Badger Canyon during April, 1990.  Washington State University.
Irrigated Agriculture Research and Extension Center. Prosser, WA.

State of Washington. 1987. Case Investigation Report 8D-87.  Department of Agriculture.
Pesticide Management Division. Yakima, WA.  98903.

State of Washington. 1988. Case Investigation Report 79-88.  Department of Agriculture.
Pesticide Management Division. Yakima, WA.  98903.

Tuffs,  W.P. and  E.B.  Morrow.   1925.   Fruit-bud  differentiation in deciduous  fruits.
Hilgardia 1:1-14

USDA. 1990.  Agricultural Resources Inputs Situation and Outlook Report.  U.S. Dept.
of Agriculture Economic Research Service.     AR-17 Feb. 1990.

Waddell, T.E. and T.B. Bower.  1988.  Agricultural production systems and discharge of
chemical residues, Chapter 3. In: Managing Agricultural Chemicals in the Environment.
The  Conservation Foundation.  Washington D.C.

-------
Westra, P., G. Franc, B. Cranmer and T. d'Amato.  1991.  Research report on 1988 potato-
herbicide  injury research,  pg. 98-104.  In:  Plant tier testing: A workshop to evaluate
nontarget plant testing in subdivision J pesticide guidelines.  Fletcher, J. and H. Ratsch,
eds.  EPA/600/9-91/041
                                         10

-------
Table 1. OPP test specifications for the Germination/Emergence and Vegetative Vigor
Tests required for pesticide registration (Gorsuch, 1991).
REQUIREMENTS
                                 Endpoint Measurements
                  Germination         Emergence       Vegetative Vigor
# Plant Species          10                10                10
Dicot Sp/Family         6/4               6/4               6/4
Monocot Sp/Family      4/2               4/2               4/2
Seeds/Replicate          10                10                 5
No. Replicates            3                 3                 3
Length (Days)            5                14               > 14
                                                (14-Day Post Germination)

Observations           Day 5           Day 10 & 14     Day 7, 14, Weekly
                                      [can be
                                      extended to
                                      28 days for
                                      soil vigor]
                                      11

-------
Table 2. Statistical data on cherry weight and number of viable cherry fruit compared to controls.
Stage   Concentration
        Cherry weight
   CV  Mean   SD
Decrease
From
Control (%)
Square root
# of viable fruits
Mean   SD
                                                                                     Decrease
                                                                                     From Control
                            CV
Bud    Control        8.77%  2.283    0.263
        0.001                   2.325    0.214
        0.002                   2.399    0.195
        0.01                    2.333    0.196
        0.1                     2.214    0.237
                              -1.84
                              -5.05
                              -2.18
                              3.06
               3.252   0.883
               3.978   1.31
               4.661   0.588
               4.049   1.120
               3.275   0.916
                                                                             26.2%
                     -23.64
                     -43.32
                     -24.51
                      -1.50
Flower  Control
        0.001
        0.002
        0.01
        0.1
25.6%  2.667   0.526                      4.436    1.285
        2.214   0.434          16.99        2.963    1.878         34.47
        2.428   0.422          8.96        2.685    0.701         40.62
        2.363   0.262          11.38        3.931    1.089         10.88
        0.992   0.828          62.82        0.629    0.869         86.08
                                                  39.3%
Young  Control
Fruit    0.001
        0.002
        0.01
        0.1
17.3%   2.408   0.495
        2.597   0.216
        1.997   0.1%
        0.941   0.193
        0.084   0.188
   -7.86
   17.05
   60.93
   96.51
3.363
4.081
2.754
0.0
0.0
1.056
1.197
1.738
0.0
0.0
                                                  43.3%
-20.58
 19.03
100.00
100.00
CV = Coefficient of Variation
SD = Standard Deviation
                                                 12

-------
                          Flowering
    Vegetative
      Growth
        is days
                                                             Pod
                                                         Development
55 days
                                  LIFE CYCLE

                                   SOYBEAN
                    t
             Germination
                  \
                     Seed
                 Development
                                                  120 days
Figure 1. The life cycle of a soybean plant. Current test protocols are designed to protect
only a small portion of the life cycle, seed germination and early seedling growth.
                                       14

-------
                                                                   Bud
                                                                      Flower
                                                                    Fruit
                                                          Fruit
                                                      Development
             Spray
         Applications	
             1991
   Spray
Applications
    1992
                Cherry Trees - Corvallis, Oregon

                                1991
Figure 2. The annual reproductive cycle in cherry trees with the times of
experimental treatments indicated. One set of fifteen trees was exposed in the spring
during flower and fruit set, and a different row of trees was exposed in the late
summer and early fall, a time when the next year's reproductive buds are developing
(Tuffs and Marrow, 1925).

-------
                  I
10-
Z5
ZO
1*
1.0-
OS-
X
X
X
X
X





.4H









A
-,









A
.^H









A
^••j




Gr*«nTip
(Bod)




4B^B







                    3.0
                    2,5
                    ZO
                  r
 0.0


 3.0


'Z5


 ZO
                    05
                    0.0
                                                     Fun Bloom
                                                      (Flower)
                                                      B
                                                   ru
                                                     Port Bloom
                                                    (Young Fruit)
                            o    0.001   0.002   aoi     ai
                                Concentration (% of field rate)

Figure 3.  Differences in mean cherry fruit weight when applications of 'Glean' where
sprayed on branches at different times of fruit bud expansion. Green tip stage (March 15,
1991).  Full bloom stage (April 2, 1991).  Post bloom stage (April 26,  1991).  Different
letters indicate significant differences between treatments at alpha = 0.05.

                                       14

-------
                     I
  4.0



  3.0



  ZO



  IX

  O5
                                                            GnxKiTip
                                                              (Bud)
  6.0

  4.5



I 3.5

| 3.0

1 Z5

i ZO
  1.0-

  0.5

  0.0-


  5.0

  4.5

  4.0

  3.5

  3.0



  ZO



  1.0

  0.5

  0.0
                                                            Ful Bloom
                                                             (Flower)
                                              A.B
                                                           Post Bloom
                                                           (Young Fruit)
                                                             .
                                0     0.001    0.002    0.01    0.1
                                     Concentration (% of field rate)
Figure 4. Differences in number of viable cherry fruit when applications of 'Glean' where
sprayed on branches at  different times of fruit bud expansion.  Green tip stage (March 15,
1991).  Full bloom stage (April 2, 1991). Post bloom stage (April 26, 1991).  Different
letters indicate significant differences between treatments at alpha = 0.05.
                                            14

-------